Modernization of Traditional Oriental Medicine: New Dosage Forms and Medical Instruments

Evidence-Based Complementary and Alternative Medicine

Lead Guest Editor: Gihyun Lee Guest Editors: Wonnam Kim, Woojin Kim, and Hanbing Li Modernization of Traditional Oriental Medicine: New Dosage Forms and Medical Instruments Evidence-Based Complementary and Alternative Medicine

Modernization of Traditional Oriental Medicine: New Dosage Forms and Medical Instruments

This is a special issue published in “Evidence-Based Complementary and Alternative Medicine.” All articles are open access articles distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Editorial Board

Mona Abdel-Tawab, Germany Ian Cock, Australia Maruti Ram Gudavalli, USA Rosaria Acquaviva, Italy Marisa Colone, Italy Narcis Gusi, Spain GabrielA.Agbor,Cameroon Lisa A. Conboy, USA Svein Haavik, Norway U. Paulino Albuquerque, Brazil Kieran Cooley, Canada Solomon Habtemariam, UK Samir Lutf Aleryani, USA Edwin L. Cooper, USA Abid Hamid, India M. S. Ali-Shtayeh, Palestine Maria T. Cruz, Portugal Michael G. Hammes, Germany Gianni Allais, Italy RobertoK.N.Cuman,Brazil Kuzhuvelil B. Harikumar, India Terje Alraek, Norway Ademar A. Da Silva Filho, Brazil Cory S. Harris, Canada Adolfo Andrade-Cetto, Mexico Giuseppe D’Antona, Italy Ken Haruma, Japan Isabel Andújar, Spain Vincenzo De Feo, Italy Thierry Hennebelle, France Letizia Angiolella, Italy Rocío De la Puerta, Spain Markus Horneber, Germany Makoto Arai, Japan Laura De Martino, Italy Ching-Liang Hsieh, Taiwan Hyunsu Bae, Republic of Korea AntonioC.P.deOliveira,Brazil BennyT.K.Huat,Singapore Giacinto Bagetta, Italy Arthur De Sá Ferreira, Brazil Helmut Hugel, Australia Onesmo B. Balemba, USA Nunziatina De Tommasi, Italy Ciara Hughes, Ireland Winfried Banzer, Germany Alexandra Deters, Germany Attila Hunyadi, Hungary Samra Bashir, Pakistan Farzad Deyhim, USA H. Stephen Injeyan, Canada Jairo Kennup Bastos, Brazil Claudia Di Giacomo, Italy Chie Ishikawa, Japan Arpita Basu, USA Antonella Di Sotto, Italy Angelo A. Izzo, Italy Sujit Basu, USA M.-G. Dijoux-Franca, France G. K. Jayaprakasha, USA Daniela Beghelli, Italy Luciana Dini, Italy Leopold Jirovetz, Austria Alvin J. Beitz, USA Caigan Du, Canada Takahide Kagawa, Japan Juana Benedí, Spain Jeng-Ren Duann, USA Atsushi Kameyama, Japan Bettina Berger, Germany Nativ Dudai, Israel Wen-yi Kang, China Maria Camilla Bergonzi, Italy Thomas Efferth, Germany Shao-Hsuan Kao, Taiwan Andresa A. Berretta, Brazil Abir El-Alfy, USA Juntra Karbwang, Japan Anna Rita Bilia, Italy Giuseppe Esposito, Italy Teh Ley Kek, Malaysia Yong C. Boo, Republic of Korea Keturah R. Faurot, USA Deborah A. Kennedy, Canada Monica Borgatti, Italy Nianping Feng, China Cheorl-Ho Kim, Republic of Korea Francesca Borrelli, Italy Yibin Feng, Hong Kong Youn C. Kim, Republic of Korea Gloria Brusotti, Italy Antonella Fioravanti, Italy Yoshiyuki Kimura, Japan Gioacchino Calapai, Italy Johannes Fleckenstein, Germany Toshiaki Kogure, Japan Giuseppe Caminiti, Italy Filippo Fratini, Italy Jian Kong, USA Raffaele Capasso, Italy Brett Froeliger, USA Tetsuya Konishi, Japan Francesco Cardini, Italy Siew H. Gan, Malaysia Karin Kraft, Germany Pierre Champy, France Jian-Li Gao, China Omer Kucuk, USA Shun-Wan Chan, Hong Kong Susana Garcia de Arriba, Germany Victor Kuete, Cameroon Kevin Chen, USA Dolores García Giménez, Spain Yiu-Wa Kwan, Hong Kong Evan P. Cherniack, USA Gabino Garrido, Chile Kuang C. Lai, Taiwan Salvatore Chirumbolo, Italy Ipek Goktepe, Qatar Ilaria Lampronti, Italy Jae Youl Cho, Republic of Korea Yuewen Gong, Canada Lixing Lao, Hong Kong K. B. Christensen, Denmark Susana Gorzalczany, Argentina Mario Ledda, Italy Shuang-En Chuang, Taiwan Sebastian Granica, Poland Christian Lehmann, Canada Yuri Clement, Trinidad And Tobago Settimio Grimaldi, Italy George B. Lenon, Australia Marco Leonti, Italy Yoshiharu Motoo, Japan Avni Sali, Australia Lawrence Leung, Canada Kamal D. Moudgil, USA Mohd Z. Salleh, Malaysia Chun-Guang Li, Australia Yoshiki Mukudai, Japan Andreas Sandner-Kiesling, Austria Min Li, China Sakthivel Muniyan, USA Manel Santafe, Spain Xiu-Min Li, USA MinKyun Na, Republic of Korea Tadaaki Satou, Japan Giovanni Li Volti, Italy Massimo Nabissi, Italy Michael A. Savka, USA Bi-Fong Lin, Taiwan Hajime Nakae, Japan Jana Sawynok, Canada Ho Lin, Taiwan Takao Namiki, Japan Roland Schoop, Switzerland Kuo-Tong Liou, Taiwan Srinivas Nammi, Australia Sven Schröder, Germany Christopher G. Lis, USA Krishnadas Nandakumar, India Veronique Seidel, UK Gerhard Litscher, Austria Vitaly Napadow, USA Senthamil R. Selvan, USA I-Min Liu, Taiwan Michele Navarra, Italy Hongcai Shang, China Monica Loizzo, Italy Isabella Neri, Italy Karen J. Sherman, USA Víctor López, Spain Pratibha V. Nerurkar, USA Ronald Sherman, USA Anderson Luiz-Ferreira, Brazil Ferdinando Nicoletti, Italy Yukihiro Shoyama, Japan Thomas Lundeberg, Sweden Marcello Nicoletti, Italy Morry Silberstein, Australia Dawn M. Bellanti, USA Cristina Nogueira, Brazil Kuttulebbai N. S. Sirajudeen, Malaysia Michel M. Machado, Brazil Menachem Oberbaum, Israel Francisco Solano, Spain Filippo Maggi, Italy Martin Offenbaecher, Germany Chang G. Son, Republic of Korea Valentina Maggini, Italy Ki-Wan Oh, Republic of Korea Con Stough, Australia Jamal A. Mahajna, Israel Yoshiji Ohta, Japan Annarita Stringaro, Italy Juraj Majtan, Slovakia Olumayokun A. Olajide, UK Shan-Yu Su, Taiwan Toshiaki Makino, Japan Ester Pagano, Italy Orazio Taglialatela-Scafati, Italy Nicola Malafronte, Italy Sokcheon Pak, Australia Takashi Takeda, Japan Francesca Mancianti, Italy Siyaram Pandey, Canada Ghee T. Tan, USA Carmen Mannucci, Italy Bhushan Patwardhan, India Norman Temple, Canada Arroyo-Morales Manuel, Spain Cláudia H. Pellizzon, Brazil Mencherini Teresa, Italy Fatima Martel, Portugal Florian Pfab, Germany Mayank Thakur, Germany Simona Martinotti, Italy Sonia Piacente, Italy Menaka C. Thounaojam, USA Carlos H. G. Martins, Brazil Andrea Pieroni, Italy Evelin Tiralongo, Australia Fulvio Marzatico, Italy Richard Pietras, USA Stephanie Tjen-A-Looi, USA Stefania Marzocco, Italy Andrew Pipingas, Australia Michał Tomczyk, Poland Andrea Maxia, Italy José M. Prieto, UK Loren Toussaint, USA James H. Mcauley, Australia Haifa Qiao, USA Luigia Trabace, Italy Kristine McGrath, Australia Xianqin Qu, Australia Yew-Min Tzeng, Taiwan JamesS.McLay,UK Roja Rahimi, Iran Dawn M. Upchurch, USA Lewis Mehl-Madrona, USA Khalid Rahman, UK Konrad Urech, Switzerland Gustavo B. Menezes, Brazil Danilo Ranieri, Italy Takuhiro Uto, Japan A. Guy Mensah-Nyagan, France Elia Ranzato, Italy Patricia Valentao, Portugal Oliver Micke, Germany Ke Ren, USA Sandy van Vuuren, South Africa Maria G. Miguel, Portugal Man Hee Rhee, Republic of Korea Luca Vanella, Italy Luigi Milella, Italy Luigi Ricciardiello, Italy Alfredo Vannacci, Italy Roberto Miniero, Italy Daniela Rigano, Italy Antonio Vassallo, Italy Letteria Minutoli, Italy José L. Rios, Spain Miguel Vilas-Boas, Portugal Albert Moraska, USA Barbara Romano, Italy Aristo Vojdani, USA Giuseppe Morgia, Italy Mariangela Rondanelli, Italy Almir Gonçalves Wanderley, Brazil Mark Moss, UK Omar Said, Israel Chong-Zhi Wang, USA Shu-Ming Wang, USA Jenny M. Wilkinson, Australia Ling Yang, China Jonathan L. Wardle, Australia D. R. Williams, Republic of Korea Albert S. Yeung, USA Kenji Watanabe, Japan Christopher Worsnop, Australia Armando Zarrelli, Italy J. Wattanathorn, Thailand Haruki Yamada, Japan Chris Zaslawski, Australia Silvia Wein, Germany Nobuo Yamaguchi, Japan Suzanna M. Zick, USA Janelle Wheat, Australia Junqing Yang, China Contents

Modernization of Traditional Oriental Medicine: New Dosage Forms and Medical Instruments Gihyun Lee , Wonnam Kim ,WoojinKim ,andHanbingLi Editorial (1 page), Article ID 6960125, Volume 2018 (2018)

Standardization of the Manufacturing Process of Bee Venom Pharmacopuncture Containing Melittin as the Active Ingredient Yoonmi L ee , Sung-Geun Kim, In-Su Kim, and Hwa-Dong Lee Research Article (7 pages), Article ID 2353280, Volume 2018 (2018)

Use of Western Medicine and Traditional Korean Medicine for Joint Disorders: A Retrospective Comparative Analysis Based on Korean Nationwide Insurance Data Boyoung Jung, Sukjin Bae, and Soyoon Kim Research Article (31 pages), Article ID 2038095, Volume 2017 (2018)

Kyungheechunggan-Tang-01, a New Herbal Medication, Suppresses LPS-Induced Inflammatory Responses through JAK/STAT Signaling Pathway in RAW 264.7 Macrophages Hee-Soo Han, Eungyeong Jang, Ji-Sun Shin, Kyung-Soo Inn, Jang-Hoon Lee, Geonha Park, Young Pyo Jang, and Kyung-Tae Lee Research Article (15 pages), Article ID 7383104, Volume 2017 (2018)

Development of a Cardiovascular Simulator for Studying Pulse Diagnosis Mechanisms Min Jang, Min-Woo Lee, Jaeuk U. Kim, See-Yoon Seo, and Sang-Hoon Shin Research Article (9 pages), Article ID 6790292, Volume 2017 (2018)

Experiences Providing Medical Assistance during the Sewol Ferry Disaster Using Traditional Korean Medicine Kyeong Han Kim, Soobin Jang, Ju Ah Lee, Bo-Hyoung Jang, Ho-Yeon Go, Sunju Park, Hee-Guen Jo, Myeong Soo Lee, and Seong-Gyu Ko Research Article (7 pages), Article ID 3203768, Volume 2017 (2018)

Modernization Trends of Infertility Treatment of Traditional Korean Medicine Jang-Kyung Park and Dong-Il Kim Review Article (9 pages), Article ID 4835912, Volume 2017 (2018)

Systems Pharmacological Approach to the Effect of Bulsu-san Promoting Parturition Su Yeon Suh and Won G. An Research Article (15 pages), Article ID 7236436, Volume 2017 (2018)

The Modulatory Effect of Acupuncture on the Activity of Locus Coeruleus Neuronal Cells: A Review Gihyun Lee and Woojin Kim Review Article (8 pages), Article ID 9785345, Volume 2017 (2018)

Understanding Mind-Body Interaction from the Perspective of East Asian Medicine Ye-Seul Lee, Yeonhee Ryu, Won-Mo Jung, Jungjoo Kim, Taehyung Lee, and Younbyoung Chae Research Article (6 pages), Article ID 7618419, Volume 2017 (2018) Effects of Combined Far-Infrared Radiation and Acupuncture at ST36 on Peripheral Blood Perfusion and Autonomic Activities Cheng-Chan Yang, Gen-Min Lin, Jen-Hung Wang, Hsiao-Chiang Chu, Hsien-Tsai Wu, Jian-Jung Chen, and Cheuk-Kwan Sun Research Article (7 pages), Article ID 1947315, Volume 2017 (2018)

Effective Heart Disease Detection Based on Quantitative Computerized Traditional Chinese Medicine Using Representation Based Classifiers Ting Shu, Bob Zhang, and Yuan Yan Tang Research Article (10 pages), Article ID 7483639, Volume 2017 (2018)

Clinical Effects of Laser Acupuncture plus Chinese Cupping on the Pain and Plasma Cortisol Levels in Patients with Chronic Nonspecific Lower Back Pain: A Randomized Controlled Trial Mu-Lien Lin, Jih-Huah Wu, Chi-Wan Lin, Chuan-Tsung Su, Hung-Chien Wu, Yong-Sheng Shih, I-Ting Chiu, Chao-Yi Chen, and Wen-Dien Chang Research Article (7 pages), Article ID 3140403, Volume 2017 (2018)

Ethanol Extract of Mylabris phalerata Inhibits M2 Polarization Induced by Recombinant IL-4 and IL-13 in Murine Macrophages Hwan-Suck Chung, Bong-Seon Lee, and Jin Yeul Ma Research Article (8 pages), Article ID 4218468, Volume 2017 (2018)

Long-Term Course to Lumbar Disc Resorption Patients and Predictive Factors Associated with Disc Resorption Jinho Lee, Joowon Kim, Joon-Shik Shin, Yoon Jae Lee, Me-riong Kim, Seon-Yeong Jeong, Young-jun Choi, Tae Kyung Yoon, Byung-heon Moon, Su-bin Yoo, Jungsoo Hong, and In-Hyuk Ha Research Article (10 pages), Article ID 2147408, Volume 2017 (2018) Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2018, Article ID 6960125, 1 page https://doi.org/10.1155/2018/6960125

Editorial Modernization of Traditional Oriental Medicine: New Dosage Forms and Medical Instruments

Gihyun Lee ,1,2 Wonnam Kim ,3,4 Woojin Kim ,2 and Hanbing Li5

1 National Development Institute of Korean Medicine, Gyeongsan, Republic of Korea 2Kyung Hee University, Seoul, Republic of Korea 3Yale University, New Haven, CT, USA 4Semyung University, Jecheon, Republic of Korea 5Zhejiang University of Technology, Hangzhou, China

Correspondence should be addressed to Gihyun Lee; [email protected]

Received 13 February 2018; Accepted 14 February 2018; Published 10 June 2018

Copyright © 2018 Gihyun Lee et al. Tis is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Traditional Oriental Medicine (TOM) has been developed for acupuncture plus Chinese cupping in chronic nonspecifc thousands of years; however recent developments in pharma- lower back pain treatment, which could further be a suitable ceutical technology promote modernization of TOM rapidly option for lower back pain treatment in clinical settings. to expand accessibility, ease of administration, and cost- Y.-S. Lee et al. and T. Shu et al. integrated technology to efectiveness. Various forms of herbal medicine are improved increase understanding of TOM. Y.-S. Lee et al. used a data for better efciency and consumer preferences including sof mining procedure, determined by the application of a term extracts, granules, intranasal administrations, eye drop, drop frequency-inverse document frequency weighting scheme pills, injection, and capsules. In-depth researches on safety to the cooccurrence table, to analyze relationships between and efcacy of these formulations are ongoing. Another trend emotions and the visceral system according to the principles on modernization of TOM is an integrated application with of East Asian medicine. T. Shu et al. proposed an efective technology. Progress on technology has developed oriental noninvasive computerized method based on facial images to medical instruments. For example, modern electrical engi- quantitatively detect heart disease. neering makes electroacupuncture, laser acupuncture, and In this special issue, we present 14 papers that address laser moxibustion possible. Experimental and translational the issue about modernization of TOM, focusing on new studies on safety and efcacy of these new methods are approaches by herbal medicine and medical instruments. underway. Here we highlight some of the key ongoing challenges Gihyun Lee published in this special issue. S. Y. Suh and W. G. An used a Wonnam Kim modernized method to understand herbal medicine, such as Woojin Kim herb-compound-target network and target-pathway network Hanbing Li analysis, to study Bulsu-san commonly used for pregnant women in East Asia. In the results, the authors report that most compounds in Bulsu-san work together with multiple target genes in a synergetic way. C.-C. Yang et al. and M.- L. Lin et al. used new treatment methods. C.-C. Yang et al. investigated vascular and autonomic impacts of combined acupuncture-far infrared radiation in improving peripheral circulation. M.-L. Lin et al. evaluated the efectiveness of laser Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2018, Article ID 2353280, 7 pages https://doi.org/10.1155/2018/2353280

Research Article Standardization of the Manufacturing Process of Bee Venom Pharmacopuncture Containing Melittin as the Active Ingredient

Yoonmi Lee , Sung-Geun Kim, In-Su Kim, and Hwa-Dong Lee

Traditional Korean Medicine Technology Division, R&D Development, Korean Medicine Preparation Team, National Development Institute of Korean Medicine (NIKOM), 94 Hwarang-ro, Gyeongsan-si, Gyeongsangbuk-do, Republic of Korea

Correspondence should be addressed to Hwa-Dong Lee; [email protected]

Received 25 August 2017; Revised 6 December 2017; Accepted 24 December 2017; Published 25 February 2018

Academic Editor: Woojin Kim

Copyright © 2018 Yoonmi Lee et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. Pharmacopuncture is a unique treatment in oriental medicine that combines chemical stimulation with conventional acupuncture. However, there are no standardized methods for preparing the herbal medicines used in pharmacopuncture, and it is not clear whether the active ingredients are safe and stable. Several studies have investigated nonstandardized preparation processes, but few investigations have addressed safety and preparation methods. Pharmacopuncture may provide an alternative treatment for incurable diseases. However, it must be as valid and safe as standardized medicine. In this way, the present project may contribute to the industrialization of medicine in Korea. It may also expand health insurance coverage by promoting evidence-based medical insurance benefits. Thus, the present study attempted to standardize and improve the raw materials, preparation, and efficacy of bee venom pharmacopuncture (BVP), which is a highly effective technique in oriental medicine. Method. To purify the crude bee venom, the extract was subjected to a stepped-gradient open column (ODS-A; 120 A,˚ 150 mesh). Using this method, the yield of melittin was significantly increased and the allergen proteins were effectively removed. The melittin content of the purified bee venom was determined using HPLC, and the product was then diluted to 0.1 mg/mL using injection water in preparation for BVP. Results. In the present study, we standardized the purification process to provide safe and stable BVP by increasing the main effective components and eliminating allergens. This study will be seminal in the industrialization and regulation ofBVP. Conclusion.We developed an effective strategy for melittin purification and allergen removal from bee venom to create safe BVP.

1. Introduction the healthcare field. Additionally, the term “pharmacopuncture” has been added to the new index of PubMed, which is Acupuncture has only been used for 60 years in Korean the world’s largest medical journal database. medicine. However, since the treatment was commercialized, Pharmacopuncture has strong anti-inflammatory and many studies have confirmed its efficacy. Although herbal pain-relieving effects because it directly treats the acupunc- acupuncture developed from acupuncture, its mechanism of ture point. In one survey of patients who had visited oriental action differs somewhat. Herbal acupuncture smooths the medicine hospitals, 48% of responders preferred pharma- flow of blood, which is referred to as “energy” in oriental copuncture to other oriental medicine treatments, because it medicine. Furthermore, the medicine contains concentrated caused a rapid decrease in pain [1, 2]. Moreover, the safety herbal ingredients that work simultaneously, thus surpassing investigation suggested that acupuncture/pharmacopuncture the efficacy of acupuncture itself. ledtoalowerrange,frequency,andseverityofsignificant Until recently, there was no proper English word for adverse events [3]. herbal acupuncture. However, the term “pharmacopuncture” The venom of the European honey bee (Apis mellifera) was registered in the 2017 medical academic information clas- comprises a mixture of proteins, peptides, and other small sification system (MeSH), which is used by the US National molecules. In bee venom pharmacopuncture (BVP), which Library of Medicine (NLM) to link academic information in has pain-relieving and anti-inflammatory effects, the venom 2 Evidence-Based Complementary and Alternative Medicine is injected at appropriate doses onto acupuncture points Table1:Comparingrawbeevenombyproductionarea. that are selected through syndrome differentiation [4]. BVP has significant therapeutic effects on degenerative knee and Division Apamin (%) PLA2 (%) Melittin (%) rheumatoid arthritis [5–9]. The main active component of Chung-Jin 37.240 12.631 32.245 bee venom pharmacopuncture (BVP) is melittin: a peptide Bi-sen 44.019 14.016 35.751 with antimicrobial, antitumor, and anti-inflammatory effects. Local 1 37.959 10.771 34.432 In oriental medicine, honey bee venom products containing Local 2 13.772 1.935 1.292 about 50% melittin are widely used for BVP. However, these Content Standards. Standard product apamin, PLA2, and melittin (0.1 mg products also contain the proteins phospholipase A2 (PLA2) content). and apamin, which are major allergens as they are capable of inducing the IgE response in susceptible individuals, according to the International Union of Immunological Societies 2.3.2. Isolation Scheme. Crude bee venom was isolated and (IUIS) [10–12]. Thus, to protect patients against side effects purified in a g/mL dilution. This 10% diluted sample was of BVP, both of these allergens must be effectively removed. subjected to a stepped-gradient open column (ODS-A, 120 A,˚ Previous studies have demonstrated that the purification and 150 meshes) that was eluted using 0%–80% ethanol, of bee venom is a challenging task, as it requires a series of affording 13 fractions. separation and purification steps [13–15]. Thus, researchers have not yet established the appropriate separation conditions 2.3.3. Isolation and Purity Verification. Each of the separated for completely removing the allergen proteins while still materials obtained through the open column, as well as their obtaining a high yield of melittin. In the present study, purity, were determined using HPLC. The separated compo- wedevelopedaneffectivestrategyformelittinpurification nents and their degree of purification were then compared from bee venom. Using this method, the yield of melittin with standard reagents. Apamin, PLA2, and melittin standard significantly increased, and the allergen proteins (apamin reagents were prepared at concentrations of 0.1 mg/mL, and and PLA2) were effectively removed. The current study may their contents were confirmed. HPLC was carried out using help researchers to develop high quality BVP medicines. a reversed-phase YMC C18 (5 𝜇m, 4.6 × 150 mm) that was It may also expand the coverage of medical insurance by eluted using a 10%–90% methanol-gradient menu system. providing a basis for quality control, standardization, and good manufacturing practice (GMP) of BVP drugs. 2.4. BVP Manufacturing. After removal of the allergen from raw bee venom and filtering of the purified melittin using 2. Materials and Methods membrane filters (pore size: 0.45–0.2 𝜇m), the melittin was subdivided into 2.25 mL vials. 2.1. Bee Venom. Crude bee venom was purchased from variousmanufacturersbasedonqualitytestresults.The 2.5. BVP Quality Management medicines were then compared with crude bee venom and with each other. Ultimately, four manufacturers were chosen: 2.5.1. Safety and Stability Evaluation. Changes in the melit- Chung-Jin Biotech, Bi-sen, and two local producers from tin’s composition were observed by applying the above man- Bong-hwa and Kyung Buk, South Korea. ufacturing process and the quality control method to the prototype product using the raw materials for distribution. 2.2. General. High-performance liquid chromatography To confirm the stability of the BVP using various additives, (HPLC) was performed with a C18-5E YMC packed column we used the pH compensator that was used for preparation. (5 𝜇m, 4.6 × 150 mm) using a Waters Alliance UV detector. Solvents for extraction, partition, thin-layer chromatography 3. Results (TLC), and HPLC were distilled from HPLC grade solvents. TheTLCplatesusedwereSilicagel60F254(Art.1.05554, 3.1. Bee Venom. To find high quality raw material, the Merck) and RP-18 F254s (Art. 1.05560, Merck). melittin content of different products was determined using HPLC analysis. Of the four crude bee venoms used, we found 2.3. Isolation and Purification that the Bi-sen product contained the highest amount of melittin (35.75%; Table 1). 2.3.1. Solvent Stability Test. Melittin, the main active ingredient of bee venom, is a protein that is reduced or destroyed 3.2. Isolation and Purification by heat, acids, bases, and so on. In the present study, ethanol was used as a solvent because it does not affect the melittin 3.2.1. Solvent Stability. Changes in melittin content were content during purification and analysis of raw bee venom. measured using ethanol, which does not affect melittin More specifically, the stability of melittin in 50% aqueous content during the analysis and purification of bee venom raw ethanol solution was investigated, and ethanol was used materials. HPLC confirmed that, in a 50% ethanol aqueous as a developing solvent in this experiment, because it did solution, the melittin content was stable, but the apamin not change the melittin content in aqueous solution. In content was significantly decreased (Figure 1). These results addition, the apamin content decreased in 50% aqueous suggest that ethanol is a good solvent for reducing the side ethanol solution. effects of BVP. Evidence-Based Complementary and Alternative Medicine 3

2500 2500 2000 2000 1500 1500 (mAU) (mAU) 1000 1000 500 500 0 0 0 10 20 30 40 50 0 10 20 30 40 50 (min) (min) (a) Crude bee venom in distilled water (b) Crude bee venom in 50% ethanol layer

Figure 1: Solvent stability: Ethanol was considered a good solvent for reducing the side effects of BVP therapeutics. Detection wavelength: UV 220 nm column (YMC C18; 5 𝜇m, 4.6 × 150 mm), flow rate: 0.4 mL/min, sample injection amount: 30 𝜇L, mobile phase conditions: 0.1% trifluoroacetic acid in H2O, and 0.1% trifluoroacetic acid in acetonitrile (gradient).

Table 2: Separate substances (detected compound) according to Table 3). The standard purity of the melittin was 99.4%, their solvent formulations. andmelittincontentofthepurifiedbeevenomwas99% Solvent gradient Amount higher than the commercial standard (Figure 4). The total Fr. Detected compound melittin yield was 63%, and its purity was about 92%–99% (H2O : ethanol) (mL) after separation and purification. 10 ND 20 ND 30 ND3.3. BVP Manufacturing. The purified bee venom was con- 40 NDcentrated and lyophilized (concentrated under reduced pres- 5 10% ND sure) to produce a powder. The melittin content in the 50 (each) 6 20% Apamin purified bee venom was determined using HPLC. The venom was then diluted to a concentration of 0.1 mg/mL, which is 7 30% Apamin, PLA2 used in BVP, using water that had been injected through a 8 40% Apamin, PLA2 0.2 𝜇m membrane. Vials were filled with 2.25 mL of this drug 950% PLA2 solution. All these procedures were performed at an aseptic 10 60% PLA2, melittin GMP facility (Figure 2). 11 70% 100 Melittin 12 70% 50 Melittin 3.4. BVP Quality Management 13 80% 100 Melittin ND: not detected. 3.4.1. Safety Evaluation. To ensure that the BVP was safe, we compared the efficacy and safety of original bee venom with those of purified bee venom that had been filtered for PLA2 3.2.2. Isolation of the Compounds. To isolate and purify the and histamine, as reported previously [10]. active component of crude bee venom (10 g/mL), the raw Bee venom for BVP is produced using a medicine venom was partitioned into 13 fractions (Table 2). According preparation process that ensures safety and lack of heavy to the corresponding HPLC profiles, Fractions 1–5 (∼10% metals. Thus, the evaluation items are the purity test and ethanol layer) contained null compounds. In Fraction 6, the heavy metal test. The purity test confirmed that the apamin appeared for the first time, and Fraction 7 contained herbicide had dissolved and that there were no heavy metals both apamin and PLA2. Melittin was eluted in Fraction 10; (lead, cadmium, arsenic, and mercury), insoluble particulate however, it was mixed with PLA2. Pure melittin was obtained matter, insoluble water, sterility, or endotoxins. Thus, based in the 70%–80% ethanol layer. on these standards, the purified bee venom appeared to be appropriate (Table 4). 3.2.3. Purity Verification. The composition of each fraction obtained through open column chromatography was deter- 3.4.2. Stability Evaluation. Changes in melittin composition mined by HPLC analysis, using apamin, PLA2, and melittin and purity were observed by applying the above manu- as standard compounds (Figure 2). Using the standard facturing process and the quality control method to the components, apamin was detected at 12 minutes, PLA2 in prototype product using the raw materials for distribution. two peaks at 18 and 19 minutes, and melittin at 27 minutes. The pH compensator was used to confirm the stability of Melittin was detected from Fraction 11 (70% ethanol layer) BVP produced using various additives (Figure 5). Changes and the peak area (%) was found to be about 98% (Figure 3, in the melittin content were examined for 6 months, and it 4 Evidence-Based Complementary and Alternative Medicine

Table 3: Component content eluted by fractions.

Change in component contents Fraction number 6 7 8 9 10 11 Apamin (%) 42.32 2.04 2.19 0.35 0.4 0.62 PLA2 (%) 0.59 20.17 24.64 6.16 1.59 0.16 PLA2 (%) 1.66 77.79 67.93 88.15 8.75 0.33 Melittin (%) - - 5.25 5.35 89.26 98.89

Raw bee venom powder (10 g)

ODS open column

H2O:EtOH

Frac. 1~4 567~8 91011~13

EtOH% 0% 10% 20% 30~40% 50% 60% 70% 90%

Detected Apamin PLA2 Apamin PLA2 Melittin compd PLA2 Melittin

Pure melittin

Purity test >90% melittin

Dilution 0.2 mg/mL

Bioactivity Cytotoxicity test

Ultrafiltration 2 times (0.4–0.2 uM)

Safety Stability test

Packing 2.2 mL

Figure 2: Purification process of bee venom pharmacopuncture (BVP) from raw material. The product was packaged at a good manufacturing practice (GMP) facility.

was found that melittin was highly stable in both pH-free and and quickly identify its cause. However, because pharma- salinity-free pharmacopuncture. copuncture has not been standardized, regulated, and indus- trialized, it is not clear whether the procedure is safe and 4. Discussion stable. The most important issue for pharmacopuncture is safety. Therefore, if pharmacopuncture is to be a pharma- As oriental medicine develops, social interest and research ceutical industry, safe medicine should be manufactured into its effects are growing. In addition, unlike injections, and standardized. Currently, China is actively producing pharmacopuncture uses acupuncture points to reduce pain and supplying medicinal herbs. To publicize this oriental Evidence-Based Complementary and Alternative Medicine 5

250 1750 200 1500 150 1250 100 1000

(mAU) 50 (mAU) 750 500 0 250 −50 0 0 10 20 30 40 50 0 10 20 30 40 50 (min) (min) (a) Fractions 1–5 (10% ethanol) (b) Fraction 6 (20% ethanol) 250 2000 1750 200 1500 150 1250 1000 100

(mAU) 750 (mAU) 50 500 0 250 0 −50 0210 0304050 0 10 20 30 40 50 (min) (min) (c) Fraction 7 (30% ethanol) (d) Fraction 8 (40% ethanol) 2500 2500 2000 2000 1500 1500 (mAU) (mAU) 1000 1000 500 500 0 0 0 10 20 30 40 50 0 10 20 30 40 50 (min) (min) (e) Fraction 9 (50% ethanol) (f) Fraction 10 (60% ethanol)

1400 2500 1200 2000 1000 800 1500 600 (mAU) (mAU) 1000 400 500 200 0 0 0 10 20 30 40 50 0 10 20 30 40 50 (min) (min) (g) Fraction 11 (70% ethanol) (h) Fraction 13 (80% ethanol)

Figure 3: Separation of sequential compounds according to solvent polarity (apamin, PLA2, and melittin). In total, 13 fractions were isolated from the crude bee venom. Pure melittin was obtained in the 70%–80% ethanol layer.

medicine, China is also actively investing in the medicine Therefore, to ensure the safety and the stability of this business. Chinese pharmacopuncture often uses two or more treatment, it is urgent that researchers standardize pharma- kinds of medicines from a single medicinal herb or material; copuncture. In this way, the therapy could be popularized these are administered to patients in various formulations. through the pharmaceutical industry. 6 Evidence-Based Complementary and Alternative Medicine

1000

(mAU) 500

0 0 20 40 (min)

Calculated

220 240 260 280 300 320 340 360 380 27 28 (nm) (min) (a) Melittin standard

600 400

(mAU) 200 0

0 20 40 (min)

Calculated

220 240 260 280 300 320 340 360 380 26 28 (nm) (min) (b) Isolated melittin from bee venom purity

Figure 4: Determination of the purity of the separated melittin (a) and comparison with standard commercial melittin (b). In the comparison with the melittin standard, the purity was 99.4%, and the melittin content of purified bee venom was 99% higher than the commercial standard.

To our knowledge, the present study was the first that by HPLC; the melittin was then diluted to 0.1 mg/mL in aimed to standardize and improve the raw materials, prepa- preparation for BVP. All these procedures were performed at ration, and efficacy of BVP,which is a highly effective oriental an aseptic GMP facility. medicinal treatment. Crude bee venom (Bi-sen) was isolated and purified in a 1 g/mL dilution. In total, 13 fractions 5. Conclusions were isolated from the crude bee venom. Pure melittin was This experiment aimed to separate melittin from crude bee obtained in the 70%–80% ethanol layer. In comparison with venomtoproducesafe,effective,andhigh-concentration the melittin standard, its purity was 99.4%, and melittin standardized medicines for pharmacopuncture. We stan- content of our purified bee venom was 99% higher than the dardized the manufacturing process to provide safe and commercial standard. Our total melittin yield was 63% and stable BVP by increasing the concentrations of the effective its purity was 92%–99% after separation and purification. The components and eliminating allergens. Thus, this study will content of melittin in our purified bee venom was determined be seminal in the industrialization and regulation of BVP. Evidence-Based Complementary and Alternative Medicine 7

Table4:Safetyevaluationofbeevenompharmacopuncture. [2] Korean Pharmacopuncture Institute, “Chapter 1, Definition and history,” in Pharmacopuncturology, Elsevier LLC, Seoul, Korea, Bee venom pharmacopuncture 2012. Purity test [3] M.-R. Kim, J.-S. Shin, J. Lee et al., “Safety of acupuncture Dissolution state ND and pharmacopuncture in 80,523 musculoskeletal disorder Lead 0 ppm patients: A retrospective review of internal safety inspection and Cadmium 0.0 ppm electronic medical records,” Medicine (United States),vol.95,no. 18, p. e3635, 2016. Arsenic 0 ppm [4] Y.-J. Ahn, J.-S. Shin, J. Lee et al., “Safety of essential bee venom Mercury 0.0 ppm pharmacopuncture as assessed in a randomized controlled Insoluble particulate matter ND double-blind trial,” Journal of Ethnopharmacology,vol.194,pp. Soluble particulate matter ND 774–780, 2016. 1 Sterility test ND [5] H. K. Ko, “Experimental Studies on the Effect of Bee Venom 2 Endotoxin test ND Theraphy on the Analgesic, Antipyretic and Anti-inflammatory 1 Sterility test: direct method using liquid thioglycolic acid medium and Action,” Koreanjournaloforientalmedicine,vol.no.1,pp.283– 2 soybean casein digestion medium. Endotoxin test: using Pierce5 LAL 292, 1992. Chromogenic Endotoxin Quantitation Kit (determination coefficient [𝑅2] ≥ [6] K. R. Kwon, H. K. Koh, Y. S. Kim, Y. B. Park, C. H. Kim, and 0.9932). S. K. Kang, “The Effect of Bee Venom Aqua-acupuncture on the Antitumor and Immune Response in the Epithelioma by 3- MCA,” JournalofKoreanAcupunctureMoxibustionSociety,vol. Melittin amount (g/mL) 2, pp. 151–172, 1997. 60 [7] J.-A. Lim, K. Sung-Nam, and L. Sung-Young, “The clinical study 50 on bee venom acupuncture treatment on osteoarthritis of knee 40 joint,” Journal of Pharmacopuncture,vol.8,no.2,pp.29–37, 30 2005. 20 [8] W. Wu-Hao, A. Kyu-Beom, L. Jin-Kang, and J. Hyoung-Seok, 10 “Clinical investigation compared with the effects of the bee- 0 venom acupuncture on knee joint with osteoarthritis,” Journal + + + of Pharmacopuncture,vol.3,pp.101–103,2001.

aCl [9] K. M. Lee, K. S. Lee, S. C. Yem et al., “A Clinical study of H aCl Ｊ

． Bee-venomacupuncturetreatmentonprotrusiondiscPatients,”

Ｊ（ + ． Journal of Korean Acupuncture Moxibustion Society,vol.5,pp. Crude powder

Crude powder Crude powder Crude powder 13–25, 2004. [10] M. Moreno and E. Giralt, “Three valuable peptides from bee and 0 days wasp venoms for therapeutic and biotechnological use: Melittin, 30 days apamin and mastoparan,” Toxins,vol.7,no.4,articleno.A020, 60 days pp. 1126–1150, 2015. Figure 5: Evaluation of the stability of the bee venom with different [11] E. Spoerri, J. Jentsch, and P. Glees, “Apamin from bee venom. additives. The changes of melittin components were examined for Effects of the neurotoxin on subcellular particles of neural 6 months; it was found that melittin was highly stable in pH- and cultures,” FEBS Letters,vol.53,no.2,pp.143–147,1975. salinity-free pharmacopuncture. [12] C. M. Freeman, C. R. A. Catlow, A. M. Hemmings, and R. C. Hider, “The conformation of apamin,” FEBS Letters,vol.197,no. 1-2, pp. 289–296, 1986. [13] Y. Maulet, U. Brodbeck, and B. W. Fulpius, “Purification from Conflicts of Interest bee venom of melittin devoid of phospholipase A2 contamination,” Analytical Biochemistry,vol.127,no.1,pp.61–67,1982. The authors have no conflicts of interest to declare regarding [14] W. Zhu, B. Wang, and X. Zhu, “Isolation and purification of the publication of this paper. BV I -2H from bee venom and analysis of its biological action,” Chinese Science Bulletin,vol.47,no.11,pp.910–914,2002. Acknowledgments [15] D.-O. Moon, S.-Y. Park, K.-J. Lee et al., “Bee venom and melittin reduce proinflammatory mediators in lipopolysaccharide- This work was supported by the Standardization Project stimulated BV2 microglia,” International Immunopharmacol- of Korean Medicine Acupuncture, which is funded by the ogy,vol.7,no.8,pp.1092–1101,2007. Korean Ministry of Health and Welfare (3243-302).

References

[1] Korean Pharmacopuncture Institute, Korean Pharmacopunc- ture Institute Compilation: A Pharmacopuncture Prepared from Herbs And Clinical Application, Korean Pharmacopuncture Institute, Seoul, Korea. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 2038095, 31 pages https://doi.org/10.1155/2017/2038095

Research Article Use of Western Medicine and Traditional Korean Medicine for Joint Disorders: A Retrospective Comparative Analysis Based on Korean Nationwide Insurance Data

Boyoung Jung,1,2 Sukjin Bae,2 and Soyoon Kim3,4

1 DepartmentofPublicHealth,GraduateSchool,YonseiUniversity,50Yonsei-ro,Seodaemun-gu,Seoul03722,RepublicofKorea 2Research Department, Research Institute of Korean Medicine Policy, 91 Heojun-ro, Gangseo-gu, Seoul 07525, Republic of Korea 3Asian Institute for Bioethics and Health Law (WHO Collaborating Centre for Health Law and Bioethics), Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea 4Department of International Health, Graduate School of Public Health, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul 03722, Republic of Korea

Correspondence should be addressed to Boyoung Jung; [email protected] and Soyoon Kim; [email protected]

Received 14 August 2017; Accepted 24 October 2017; Published 6 December 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Boyoung Jung et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This study aimed to compare the usage of Western medicine and traditional Korean medicine for treating joint disorders inKorea. Data of claims from all medical institutions with billing statements filed to HIRA from 2011 to 2014 for the four most frequent joint disorders were used for the analysis. Data from a total of 1,100,018 patients who received medical services from 2011 to 2014 were analyzed. Descriptive statistics are presented as type of care and hospital type. All statistical analyses were performed using IBM SPSS for Windows version 21. Of the 1,100,018 patients with joint disorders, 456,642 (41.5%) were males and 643,376 (58.5%) were females. Per diem costs of hospitalization in Western medicine clinics and traditional Korean medicine clinics were approximately 160,000 KRW and 50,000 KRW, respectively. Among costs associated with Western medicine, physiotherapy cost had the largest proportion (28.78%). Among costs associated with traditional Korean medicine, procedural costs and treatment accounted for more than 70%, followed by doctors’ fees (21.54%). There were distinct differences in patterns of medical care use and cost ofjoint disorders at the national level in Korea. This study is expected to contribute to management decisions for musculoskeletal disease involving joint disorders.

1. Introduction Korean medicine [5], an integral part of prevailing practice and belief systems throughout Korea’s history. Starting from Musculoskeletal diseases are one of the leading causes of disa- the late 19th century, Western medical practices were intro- bility worldwide. It is a major contributor to health burden duced by Christian missionaries to Korea. These practices and health care costs [1]. Korea has a rapidly aging popula- quickly supplanted traditional Korean medicine in institution due to the decrease in birth rate and increase in life expectancy. The percentage of the population aged 65 years tional health care. After the Korean War, the government or older will increase from 10.3% in 2008 to 15.6% in 2020 and revived interest in traditional Korean medicine and estab- 38.2% in 2050 [2]. Previous studies have reported findings lished colleges that specialize in that field, in addition to on the prevalence of musculoskeletal disease [3, 4]. As the colleges of Western medicine [6]. With this historical back- Korean population continues to age, the economic burden of ground, the Korean medical system is characterized by a dual musculoskeletal disease will continue to increase. The total [7, 8], mutually exclusive medical system consisting of West- economic burden of treating musculoskeletal is about 8.1 bil- ern medicine and traditional Korean medicine practices [9]. lion dollars [2]. Musculoskeletal diseases are the most com- In Korea, primary care physicians work mostly in solo private mon health problems that require the use of traditional practicesandarereimbursedonafee-for-servicebasis.This 2 Evidence-Based Complementary and Alternative Medicine system enables patients to choose and retain individual physi- patient). Therefore, the number of claims in this study was cians regardless of changes in employment status. There- higher than the total number of patients. A total of 7,996,903 fore,KoreanscanusebothWesternandtraditionalKorean claims for 1,100,018 patients with joint diseases with prefix medicine to treat musculoskeletal disorders. Previous studies codes of M17, M75, S63, and S93 in primary diagnoses were have assessed the prevalence and cost of Korean medicine included for analysis through discussion of a panel of three [5, 6, 10–13]. However, most of these studies did not focus on clinicians (one public health specialist, one Korean medical joint disorders [11, 14–16]. Particularly, the statuses of health doctor, and one statistician). A total of 1,100,018 patients were care utilization associated with joint disorders including finally included in our analysis (Figure 1). the scale of the whole population and health care costs of patients receiving treatment for joint disorders are currently 2.3. Episode Creating Process. Claims data provided by HIRA unknown. Therefore, the objective of this study was to analyze included raw data of treatment prescriptions for all patients claim data submitted to the Korean National Health Insur- who received medical services over the course of one year ance (NHI) and assessed by the Health Insurance Review after removing personal and corporate information. Because and Assessment Service (HIRA) to compare medical care use the claims were submitted monthly, charges in the statement between Western medicine and traditional Korean medicine. reflected up to one month of information. In other words, The results of this study will provide basic information for patients who had been hospitalized for more than one month future management decisions for musculoskeletal diseases wouldhavebeenchargedseparatelyforeachmonth.Insuch especially joint disorders in Korea. cases, errors such as overestimation of the number of inpatients and underestimation of medical expenses might occur 2. Materials and Methods when performing statistical analyses. Therefore, episodes, involving collecting and calibrating several claims charged 2.1. Data Source. This study used claims data from the monthly for one consecutive medical practice were used. In 2011–2014 National Patient Sample (NPS) dataset of HIRA. this study, separated claim forms of hospitalized patients were Datasets generated and/or analyzed for this study period bundled into one hospitalization episode. Variables used in are available from the HIRA-NPS repository [17]. The NPS the episode creating process included claims identification includes 3% sample data of 2011–2014 national insurance key, patient identification key, insurance type, main diagnosis billing data. It can represent the country as a whole (46 mil- code, treatment type, treatment start date, and treatment end lionpatients).Thetotalnumberoffiledclaimsandtotal date. health expenditure have increased steadily. As of 2011, the totalnumberoffiledclaimshasreached1.3billion,withatotal 2.4. Main Descriptive Variables. The main descriptive vari- health expenditure of ∼51.5 trillion KRW. Patients were strat- ables were frequency and cost of medical care without ified according to sex and age in 5-year intervals. These HIRA addressing a specific hypothesis. Frequency included the claim data are compiled by health care providers nationwide. number of hospitalizations and outpatient visits in Western They correspond to the number of claims submitted by medicine and traditional Korean medicine clinics, interven- patients. Claims from patients with the Medicaid program, tion (surgical and nonsurgical), and annual usage. Rehabilita- government expenditures, and veteran patients are also tion-related nonsurgical interventions were classified accord- included in these claim data [18]. All data were deidentified ing to National Evidence-based Healthcare Collaborating to ensure patient confidentiality. The HIRA Research Ethics Agency reports [20]. Committee of South Korea approved the study protocol. Cost included average cost per patient and cost per day (per diem) for joint disorders. Medical costs determined to 2.2. Study Population. After reviewing frequent diseases each beeligibleforreimbursementbyHIRAoutoftreatmentcosts year in traditional Korean medicine as described previously were indicated in the submitted insurance claim statement. [19], patients with the following four most frequent joint Medical costs, that is, the sum of benefits reimbursed by disorders were included in this study: M17 (gonarthrosis the insurer (Korean National Health Insurance Service) to [arthrosis of knee]), M75 (shoulder lesions), S63 (dislocation, the medical care institutions, were classified as INSUP and sprain, and strain of joints and ligaments at wrist and hand self-payment costs paid by the beneficiary (patient) as SLF. level), and S93 (dislocation, sprain, and strain of joints and It was expressed as total treatment cost in Korean Won ligaments at ankle and foot level). Although dorsalgia (M54) (1,000,000 KRW). Each patient’s medical costs were calcu- was at the top of the list, it was excluded from analysis since lated as the sum of costs listed on their claims. The average there was no change in ranking by year. To observe changes treatment amount was the amount of total medical expenses in the ranking of diseases by year, the remaining joints were for one year divided by the number of patients. Per diem was included in the study. We focused on musculoskeletal disor- the amount of total medical expenses for one year divided ders and injuries of the extremities. The diagnoses were coded bythenumberofdaysofhospitalizationorinanoutpatient using the 6th revision of the Korean Classification of Diseases clinic. (KCD-6) adapted from the International Classification of The number of reimbursed days included the number Diseases, 10th revision. Data from the billing statements for of hospitalized days or outpatient visits and in-care drug patients with missing cost data and those with zero total prescription days. These days were defined based on the cost were excluded. Patient might have visited more than number of visits (for outpatient departments) or the length of once during the study period (i.e., more than one claim per hospital stay (for inpatient departments) of patients indicated Evidence-Based Complementary and Alternative Medicine 3

Number of T20 claims with joint ICD codes M17, M75, S63, S93 in the NHP 2011–2014 (i) 2011: 1,885,740 (ii) 2012: 1,999,160 Excluded claims unmatched between (iii) 2013: 2,052,435 T20 and T30, T40, T53 (iv) 2014: 2,081,344 (i) 2011: 2,480 (ii) 2012: 3,314 Number of T20 claims in the NHP 2011–2014 (iii) 2013: 2,327 (i) 2011: 1,883,260 (iv) 2014: 2,342 (ii) 2012: 1,995,846

(iii) 2013: 2,050,108 Excluded claims with missing cost or 0 total cost (iv) 2014: 2,079,002 (i) 2011: 5,373 (ii) 2012: 1,818 Number of T20 claims in the NHP 2011–2014 (iii) 2013: 1,388 (i) 2011: 1,877,877 (iv) 2014: 2,724 (ii) 2012: 1,994,028 (iii) 2013: 2,048,720 Excluded claims of which main (iv) 2014: 2,076,278 disease is not M17 and M75, S63, S93

Aggregated claims connected within a Number of T20 claims in the NHP 2011–2014 patient, a main disease, a medical type, (i) 2011: 1,481,325 a medical institution, an insurance (ii) 2012: 1,559,336 type among hospitalization cases (iii) 2013: 1,594,147 (iv) 2014: 1,599,926

Patients included for analysis in T20 claims (i) 2011: 268,048 (ii) 2012: 273,752 (iii) 2013: 278,170 (iv) 2014: 280,048 Total: 1,100,018 patients

Figure1:Flowchartofthestudysample.

in the submitted insurance claim statement [14]. Days per (NHI,Medicaid,andothers)atthedateofvisitofahealthcare episode were calculated as total reimbursed days divided by institution and the most frequently visited ones. Individuals the total number of episodes. Patient and medical institution- were qualified for Medicaid if they had a household income of related characteristics are defined as follows. less than $600 per month. Medical services for veterans and beneficiaries were free of charge as government expenditure. Severity was measured using the Charlson Comorbidity 2.5. Patient Characteristics. Patient characteristics included Index (CCI) [21] defined as the sum of weights related to each gender, age, medical insurance type, severity of disease, condition for which a patient had available claim data. The existence of surgery, and type of medicine. The main attend- CCI score was determined based on the presence of specific ing hospital characteristics included hospital type, region, ICD-10 codes during one year [22]. In this study, CCI at ownership,thenumberofbeds,thenumberofWestern initiationwasusedastheCCIscoreofeachpatient. doctors, and the number of traditional medical practitioners. Patient demographic data obtained from the NHI claims 2.6. Medical Institution Characteristics. The types of medical database included gender, age, and medical insurance type practice were divided into three as follows: traditional, 4 Evidence-Based Complementary and Alternative Medicine

Western, and both traditional and Western. Hospital was Clinics showed a higher proportion than Western Clinic for the main attending medical institution which was visited outpatients. The admission rates were 96.90% in Western most frequently by the patient for care. If visit frequency medicine clinic and 3.10% in traditional Korean medicine per institution was the same, the main attending hospital clinics. Among all outpatient visits, 67.85% involved orthodox was the last health care institution that the patient visited. medicine while 32.15% involved traditional Korean medicine. Medical institutions included Tertiary and General Hospitals, In Western medicine clinics, patients were hospitalized most hospitals, long-term care hospitals, Western Clinics, Dental frequently in hospitals, followed by long-term care hospitals. Hospitals and Clinics, Public Health Hospitals (admission Hospitalization at hospital level gradually decreased from facility-equipped health center), Public Health Centers, Local 37.17% in 2011 to 36.37% in 2014. On the other hand, the Public Health Clinics, Traditional Hospitals, and Traditional percentage of patients who were hospitalized mainly in Tradi- Clinics. Region and ownership were the characteristics of the tional Hospitals increased from 2.12% in 2011 to 3.28% in 2014. medical institution that the patient visited. Outpatient visits accounted for most visits to hospitals (West- ern medicine: 56.26%; Traditional Korean medicine: 31.69%). While the percentage of outpatients at Western medicine 2.7. Statistical Analysis. Basic characteristics of the study clinics steadily increased from 66.19% in 2011 to 68.93% sample are presented as frequencies and percentages. They by 2014, the percentage of outpatients at traditional Korean are presented for each operational definition. Descriptive medicine clinics steadily decreased from 33.81% in 2011 to statistics are presented as type of care and hospital type. All 31.07% in 2014. Most (91.0%) institutions were privately statistical analyses were performed using SPSS version 21 for owned, and most (80.4%) of them were located outside Seoul. Windows (IBM Corp., Armonk, NY, USA). The majority (95.2%) of institutions had fewer than 5 beds. There was no significant difference among the four groups 3. Results according to year. The total average treatment cost (RPE) is thesumofINSUPandSLFpaidtomedicalcareinstitution. The general characteristics of the study population are sum- RPE per patient was 185,933 KRW in 2011, 192,290 KRW marized in Table 1. in 2012, 202,967 KRW in 2013, and 208,739 KRW in 2014. A total of 1,100,018 patients were included, including Women incurred more medical expenses in 2011 to 2014 456,642 (41.5%) males and 643,376 (58.5%) females. All four compared to men. Expenditure was increased with age. years (from 2011 to 2014) showed higher percentages of It peaked in patients in their 70s with a minimum of females than males. Patients under 29 years of age accounted 377,448 KRW to a maximum of 388,445 KRW. In terms of for the largest proportion (23.3%), followed by patients in expense by the type of joint lesion, knee lesions (M17) had their 50s (20.5%) and 60s (16.8%). A total of 1,050,691 (95.5%) the highest expense among the four types of joint disorders, patients were enrolled in the NHI scheme while the remain- followed by shoulder lesions (M75). As the severity of lesion ing 49,012 patients (4.5%) were enrolled in Medicaid. Patients was increased, the expense was also increased. However, with knee arthrodesis accounted for the most (24.2%), fol- the difference in expense was not statistically insignificant. lowed by those with foot joint disease. More than half (52.5%) Patients who were hospitalized spent 20 times more than of these patients had knee arthropathy. Approximately 70% those who were not hospitalized. The average treatment (70.5%) of patients had mild joint disorder with CCI score costs per patient in inpatient care and outpatient care were of 0. Among the 1,100,018 patients, 18,041 (1.6%) patients 192,414 and 65,319 KRW, respectively. Patients who used only underwent surgery while the majority (98.4%) of patients traditional Korean medicine spent twice less than those underwent nonsurgical procedures. For body regions where who only used Western medicine. The range of RPE for basic physical therapy was performed more than three times, Western medicine was from 181,225 KRW to 198,661 KRW. the knee and shoulder regions accounted for more than The range of RPE for traditional Korean medicine was from 25%. For regions that needed surgery, the knees accounted 82,019 KRW to 96,325 KRW. There were no differences in for the most. For regions that underwent acupuncture two costs over 400,000 KRW among hospitals that practiced times or more, the shoulder, hand, and foot areas accounted Western medicine (Tertiary and General Hospital, hospital, for 30% or more (Table 2). Regardless of disease type, only and Western Clinic). Costs for Western medicine hospitals 1.6% of patients underwent surgery, of which knee surgery were the highest, followed by that for Traditional Hospitals was the most frequently performed type (43%∼44%). The and long-term care hospitals (Table 5). results of nonsurgical intervention distribution are shown The frequency and total medical expenditures for West- in Table 3. The main attending medical institutions included ern medicine and traditional Korean medicine are shown 58,245 (5.3%) Tertiary and General Hospitals, 118,408 (10.8%) in Table 6. There were 21,894,252 claims with a cost of hospitals, 8,638 (0.8%) Western Clinics, 592,155 (53.8%) long- 168,024,474 (1000 KRW) for Western medicine. However, term care hospitals, 6,473 (29.6%) Traditional Hospitals, there were only 9,628,946 claims with a cost of 38,602,696 and (28.7%) Traditional Clinics. However, the results were (1000 KRW) for traditional Korean medicine. The medical different for hospitalization and outpatient visits (Table 4). expense per visit in an outpatient clinic was 22,000 KRW Hospitalization was mainly in the order of hospitals > long- for Western medicine and about 18,000 KRW for traditional term care hospital > Western Clinic. This is mainly due to the Korean medicine. The day per episode of traditional Korean characteristics of patients who require surgery. On the other medicine was longer than that of Western medicine. After hand, among the same primary clinic institutions, Traditional analyzing the medical cost of claims for Western medicine Evidence-Based Complementary and Alternative Medicine 5 (.0) (%) Total 𝑁 315 (%) 2014 𝑁 (%) 2013 51 (58.6) 163,318 (58.3) 643,376 (58.5) 49 (16.9) 47,162 (16.8) 184,831 (16.8) 52 (20.9) 58,201 (20.8) 226,023 (20.5) 𝑁 8,170 (100.0) 280,048 (100.0) 1,100,018 (100.0) 8,170 (100) 280,048 (100) 1,100,018 (100.0) (%) 95.5) 266,015 (95.6) 268,135 (95.7) 1,050,691 (95.5) 2012 𝑁 Table 1: General characteristics of patients and hospitals. (%) 2011 𝑁 60 (.0) 73 (.0) 86 (.0) 96 (.0) 1,631 (.6) 1,631 (.6) 1,628 (.6) 1,583 (.6) 6,473 (.6) 1,912 (.7) 2,138 (.8) 2,268 (.8) 2,320 (.8) 8,638 (.8) 5,041 (2.4) 4,792 (2.3) 4,874 (2.4) 11,556 (5.8) 26,263 (3.2) 3,837 (1.4) 4,427 (1.6) 4,758 (1.7) 5,019 (1.8) 18,041 (1.6) 11,630 (5.5) 11,245 (5.4) 11,097 (5.4) 21,684 (10.9) 55,656 (6.8) 111,101 (41.4) 113,592 (41.5) 115,219 (41.4) 116,730 (41.7) 456,642 (41.5) 27,303 (10.2) 27,272 (10.0) 27,047 (9.7) 27,176 (9.7) 108,798 (9.9) 16,677 (6.2) 16,401 (6.0) 17,223 (6.2) 16,858 (6.0) 67,159 (6.1) 37,600 (17.7) 37,236 (17.9) 36,331 (17.8) 49,180 (24.8) 160,347 (19.5) 53,565 (20.0) 56,005 (20.5) 58,2 40,901 (15.3) 40,824 (14.9) 41,482 (14.9) 40,793 (14.6) 164,000 (14.9) 25,672 (9.6) 29,657 (10.8) 30,968 (11.1) 32,111 (11.5) 118,408 (10.8) 35,964 (13.4) 38,443 (14.0) 41,092 (14.8) 42,915 (15.3) 158,414 (14.4) 64,453 (24.1) 63,426 (23.2) 63,234 (22.7) 61,159 (21.8) 252,272 (22.9) 63,564 (23.7) 65,754 (24.0) 66,825 (24.0) 66,647 (23.8) 262,790 (23.9) 80,423 (30.0) 79,507 (29.0) 79,357 (28.5) 76,779 (27.4) 316,066 (28.7) 42,286 (15.8) 41,452 (15.1) 41,118 (14.8) 40,898 (14.6) 165,754 (15.1) 46,659 (17.4) 47,900 (17.5) 48,776 (17.6) 49,269 (17.6) 192,604 (17.5) 64,226 (24.0) 64,564 (23.6) 63,712 (22.9) 63,696 (22.7) 256,198 (23.3) 44,704 (16.7) 46,016 (16.8) 46,9 141,148 (52.6) 143,697 (52.4) 145,346 (52.3) 147,274 (52.6) 577,465 (52.5) 157,575 (74.4) 154,798 (74.4) 151,540 (74.3) 115,851 (58.4) 579,764 (70.5) 264,211 (98.6) 269,325 (98.4) 273,412 (98.3) 275,029 (98.2) 1,081,977 (98.4) 145,278 (54.2) 146,694 (53.6) 149,037 (53.6) 151,146 (54.0) 592,155 (53.8) 162,550 (60.7) 169,502 (61.9) 173,454 (62.4) 178,096 (63.6) 683,602 (62.2) 156,947 (58.6) 160,160 (58.5) 162,9 268,048 (100) 273,752 (100) 27 268,048 (100.0) 273,752 (100.0) 27 ∗ ‡ † ) § ( CCI & 60–69 70–79 S63 S93 1 2 3+ Yes Both Western Western Clinic Long-term care hospital Traditional Hospital Traditional Clinic National Health Insurance 255,160M17 (95.2)0 261,381 ( No Traditional Tertiary and General HospitalHospital 13,100 (4.9) 14,125 (5.2) 14,911 (5.4) 16,109 (5.8) 58,245 (5.3) MedicaidOthers M75 12,828 (4.8) 12,298 (4.5) 12,069 (4.3) 11,817 (4.2) 49,012 (4.5) Female Male ≦ 29 30–39 40–49 50–59 Severity Medical insurance type Category Patient total Patient total Hospital type Surgery Type of medicine Lesion of joint Gender Age (yr) 6 Evidence-Based Complementary and Alternative Medicine (%) hen one patient Total W † 𝑁 (%) everity was measured using the Charlson S ‡ 2014 𝑁 of charge as a government expenditure. (%) 2013 𝑁 ospital type is the type of medical institution most frequently visited. H § (%) 2012 Table 1: Continued. 𝑁 nd others. Others include veterans and beneficiaries who receivecare free (%) case involving one or more of the other three joint disorders which were included in this study. A & 2011 𝑁 2,195 (.8) 2,444 (.9) 2,401 (.9) 2,542 (.9) 9,582 (.9) 1,894 (.7) 1,916 (.7) 2,018 (.7) 2,120 (.8) 7,948 (.7) 8,455 (3.2) 7,626 (2.8) 7,959 (2.9) 8,635 (3.1) 32,675 (3.0) 2,820 (1.1) 2,3435,436 (.9) (2.0) 2,552 1,796 (.9) (.7) 2,655 1,753 (.9) (.6) 10,370 1,686 (.9) (.6) 10,671 (1.0) 11,231 (4.2) 2,469 (.9) 2,976 (1.1) 2,474 (.9) 19,150 (1.7) 68,101 (25.4) 70,626 (25.8) 72,117 (25.9) 72,436 (25.9) 283,280 (25.8) 13,829 (5.2) 2,055 (.8) 1,897 (.7) 2,451 (.9) 20,232 (1.8) 20,395 (7.6) 22,750 (8.3) 23,386 (8.4) 24,272 (8.7) 90,803 (8.3) 52,548 (19.6) 53,682 (19.6) 54,538 (19.6) 54,907 (19.6) 215,675 (19.6) 70,428 (26.3) 11,177 (4.1) 12,686 (4.6) 14,624 (5.2) 108,915 (9.9) 147,399 (55.0) 149,444 (54.6) 151,515 (54.5) 152,705 (54.5) 601,063 (54.6) 172,528 (64.4) 257,710 (94.3) 260,076 (93.7) 260,036 (93.0) 950,350 (86.5) 245,726 (91.7) 249,084 (91.0) 252,765 (90.9) 253,654 (90.6) 1,001,229 (91.0) 262,580 (98.0) 270,211 (99.3) 274,766 (99.4) 276,741 (99.4) 1,084,298 (99.0) 254,565 (95.0) 261,339 (95.5) 265,257 (95.4) 266,216 (95.1) 1,047,377 (95.2) ≦ ≦ edical insurance type divided into National Health Insurance, Medicaid, a 0 1 ≦ ≦ 4 5–11 12–19 20 18 Metropolitan city Other (Rural) Seoul (Urban) Public Corporation Private ≦ 5 6–12 13–17 M Number of traditional doctors per 100 beds Number of Western doctors per 100 beds Category Region Ownership Number of beds had multiple joint diseases, the most frequent disease is indicated. ∗ Comorbidity Index (CCI), defined as the sum of weights related tocondition each fora which submitted patient claims. Evidence-Based Complementary and Alternative Medicine 7 2012 2011 761——33——— 96 32 39 23 2 24 12 9 3 — (.01) (.00) (.01) (.02) (.00) (.00) (.00) (.00) (.00) (.00) (.45) (.16) (.17) (3.50) (.06) (.53) (.22) (.31) (3.49) (.11) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) Total M17 M75 S63 S93 Total M17 M75 S63 S93 6,698 1,009 751 4,788 150 8,308 1,511 1,476 5,021 300 (9.23) (11.89) (7.52) (5.76) (7.27) (8.84) (10.63) (8.07) (5.88) (7.17) (2.20) (2.06) (1.47) (4.75) (2.47) (2.46) (2.17) (1.93) (5.15) (2.67) 32,582 13,360 6,568 6,504 6,150 38,311 14,819 9,081 7,424 6,987 (99.55) (99.84) (99.83) (96.50) (99.94) (99.47) (99.78) (99.69) (96.51) (99.89) (63.97) (58.11) (65.61) (73.91) (70.79) (64.16) (60.09) (63.40) (73.75) (70.85) (99.99) (100) (99.99) (99.98) (100) (100) (100) (100) (100) (100) (24.60) (27.93) (25.40) (15.58) (19.48) (24.55) (27.11) (26.61) (15.22) (19.31) 136,773 77,169 33,618 7,892 18,094 137,871 72,622 38,010 8,468 18,771 947,982 377,024 293,443 101,240 176,275 1,000,863 410,595 298,619 106,230 185,419 364,632 181,204 113,595 21,336 48,497 382,987 185,211 125,330 21,922 50,524 1,475,271 647,748 446,473 132,184 248,866 1,551,724 681,736 469,564 139,023 261,401 1,481,873 648,725 447,185 136,949 249,014 1,560,008 683,235 471,031 144,041 261,701 1,481,962 648,751 447,223 136,972 249,016 1,560,029 683,244 471,040 144,044 261,701 1,481,969 648,757 447,224 136,972 249,016 1,560,032 683,247 471,040 144,044 261,701 ) ∗ 𝑁 % % % % 𝑁 𝑁 𝑁 % 𝑁 𝑁 % 𝑁 % 𝑁 % 𝑁 % 𝑁 % Total ( 𝑁 1 0 2 0 1 ≦ 0 1 ≦ 3 ≦ 1 ≦ 0 § Table 2: Distribution of nonsurgical intervention in both WM and traditional medicine according to the corporal name. † ‡ Basic physical therapy Simple rehabilitation Professional rehabilitation Nonsurgical intervention WM WM WM WM Rehabilitation of CNS 8 Evidence-Based Complementary and Alternative Medicine 1 1 47 37 8 5 Table 2: Continued. 47 29 10 3 113 45 30 8 30 47 22 13 2 10 257 88 109 8 52 188 115 58 4 11 (.01) (.01) (.01) (.01) (.01) (.00) (.00) (.00) (.00) (.00) (.02) (.01) (.02) (.01) (.02) (.01) (.02) (.01) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.01) (.00) (.00) (.00) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (2.74) (1.97) (3.15) (3.21) (3.78) (2.88) (2.26) (3.13) (3.28) (3.85) 40,678 12,784 14,083 4,402 9,409 44,984 15,421 14,765 4,728 10,070 (99.98) (99.99) (99.98) (99.99) (99.98) (99.99) (99.98) (99.99) (100) (100) (99.99) (99.99) (99.99) (99.99) (99.99) (100) (100) (100) (100) (100) (30.96) (18.97) (38.46) (35.44) (46.29) (29.03) (18.70) (33.21) (35.19) (45.11) (66.29) (79.06) (58.40) (61.35) (49.94) (68.08) (79.04) (63.66) (61.53) (51.04) 982,471 512,929 261,161 84,032 124,349 1,062,109 540,051 299,859 88,627 133,572 458,820 123,044 171,980 48,538 115,258 452,939 127,775 156,416 50,689 118,059 1,481,712 648,669 447,115 136,964 248,964 1,559,844 683,132 470,982 144,040 261,690 1,481,922 648,728 447,214 136,969 249,011 1,559,985 683,210 471,032 144,043 261,700 1,481,856 648,712 447,194 136,964 248,986 1,559,985 683,225 471,027 144,042 261,691 % % 𝑁 % 𝑁 % 𝑁 𝑁 𝑁 𝑁 % 𝑁 % % 𝑁 % 𝑁 % 1 ≦ 1 ≦ 0 0 0 0 1 1 ≦ 2 ≦ TM Acupuncture TM Moxibustion TM Cupping TM Heat & cold therapy Evidence-Based Complementary and Alternative Medicine 9 2014 2013 Table 2: Continued. 725——11——— 35 19 16 — — 38 32 4 1 1 9,115 1,858 1,498 5,445 314 10,066 2,138 2,001 5,568 359 (.57) (.27) (.31) (3.62) (.12) (.63) (.30) (.43) (3.65) (.13) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) (100) Total M17 M75 S63 S93 Total M17 M75 S63 S93 (7.91) (8.97) (7.32) (5.99) (7.27) (7.65) (8.45) (7.07) (6.00) (7.50) (2.94) (3.08) (2.16) (4.75) (2.94) (2.82) (2.63) (2.16) (5.00) (3.20) 46,856 21,507 10,255 7,147 7,947 45,063 18,589 10,106 7,624 8,744 (24.32) (26.57) (26.95) (14.89) (19.14) (24.97) (27.70) (27.61) (15.01) (18.92) (99.43) (99.73) (99.69) (96.38) (99.88) (99.37) (99.70) (99.57) (96.35) (99.87) 126,126 62,627 34,839 9,016 19,644 122,500 59,726 33,139 9,146 20,489 (64.83) (61.39) (63.57) (74.38) (70.64) (64.57) (61.22) (63.15) (73.99) (70.39) 387,943 185,606 128,210 22,432 51,695 399,645 195,711 129,354 22,863 51,717 1,594,914 698,521 475,749 150,621 270,023 1,600,736 706,585 468,439 152,358 273,354 1,585,834 696,682 474,267 145,176 269,709 1,590,708 704,479 466,442 146,791 272,996 1,034,024 428,800 302,461 112,026 190,737 1,033,566 432,591 295,844 112,726 192,405 1,594,942 698,538 475,760 150,621 270,023 1,600,773 706,616 468,443 152,359 273,355 1,594,949 698,540 475,765 150,621 270,023 1,600,774 706,617 468,443 152,359 273,355 ) ∗ 𝑁 % 𝑁 % % 𝑁 𝑁 % % 𝑁 𝑁 % 𝑁 % % 𝑁 𝑁 % 𝑁 % 2 3 ≦ 1 1 ≦ Total ( 𝑁 0 0 0 1 ≦ 0 1 ≦ § † ‡ Basic physical therapy Simple rehabilitation Professional rehabilitation Nonsurgical intervention WM WM WM WM Rehabilitation of CNS 10 Evidence-Based Complementary and Alternative Medicine 2 Simple ‡ 1 rical stimulation therapy, study was higher than the number of patients. , massage therapy, and simple therapeutic exercise. 40 33 4 exercise, motor point block, pneumatic compression, complex decongestive abilitative dysphagia therapy. M17, knee lesions; M75, shoulder lesions; S63, Table 2: Continued. t therapy, UV irradiation, transcutaneous electrical nerve stimulation 33 11 — 8 33 11 — 8 256 190 46 2 18 itative breathing therapy, rehabilitative functional training, and reh 52 52 73 56 9 2 6 107 95 7 — 5 (.00) (.01) (.00) (.00) (.00) (.01) (.01) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.00) (.02) (.03) (.01) (.00) (.01) (100) (100)(100) (100) (100) (100) (100) (100) (99.98) (100) (99.97) (100) (99.99) (100) (100) (100) (100) (100) (100) (100) (3.46) (3.29) (3.50) (3.44) (3.81) (3.15) (2.84) (3.39) (3.06) (3.60) 55,124 23,007 16,632 5,186 10,299 50,447 20,075 15,875 4,656 9,841 (28.50) (18.03) (31.87) (36.94) (44.95) (27.95) (17.91) (30.58) (36.55) (44.57) (68.04) (78.68) (64.63) (59.62) (51.24) (68.90) (79.25) (66.03) (60.39) (51.83) 454,601 125,946 151,649 55,635 121,371 447,355 126,560 143,257 55,694 121,844 (100.00) (99.99) (100) (100) (100) (99.99) (99.99) (100) (100) (100) 1,594,876 698,484 475,756 150,619 270,017 1,600,667 706,522 468,436 152,359 273,350 1,085,224 549,587 307,484 89,800 138,353 1,102,972 559,982 309,311 92,009 141,670 1,594,897 698,507 475,7541,594,897 150,621 698,507 270,015 475,754 1,600,518 150,621 706,427 270,015 468,397 1,600,734 152,357 706,584 273,337 468,439 152,358 273,353 rehabilitation included pool therapy, occupational therapy, activities of daily living training, neurogenic bladder training, functional elect % 𝑁 % 𝑁 % 𝑁 % 𝑁 % 𝑁 % 𝑁 % 𝑁 𝑁 % 𝑁 % 1 ≦ 0 1 2 ≦ 0 0 1 ≦ 0 1 ≦ Professional § physical therapy included superficial heat therapy,cold therapy, deep hea patient could be hospitalized more than once during the study period, which resulted in more than one claim per patient. Thus, the number of claims in the A Basic TM Acupuncture TM Moxibustion TM Cupping TM Heat & cold therapy rehabilitation included paraffin bath, hydrotherapy, intermittent traction therapy, electricalphysical stimulation therapy,therapy, and iontophoresis. therapy, laser therapeutic ∗ † myofascial trigger point injection, rehabilitative socialwrist work, and rehabil hand level lesions; S93, ankle and foot level lesions; WM, Western medicine; TM, Korean traditional medicine. Evidence-Based Complementary and Alternative Medicine 11

Table 3: Comparison of rate of surgery by diagnostic code.

Shoulder lesions Wrist and hand Ankle and foot Total Knee lesions [M17] Year [M75] level [S63] level [S93] ∗ 𝑁 % 𝑁 % 𝑁 % 𝑁 % 𝑁 % Total Unit 1,481,969 100.00 648,757 43.78 447,224 30.18 136,972 9.24 249,016 16.80 𝑁 1,477,896 646,826 445,926 136,436 248,708 No 2011 % (99.73) (99.70) (99.71) (99.61) (99.88) 𝑁 4,073 1,931 1,298 536 308 Yes % (.27) (.30) (.29) (.39) (.12) Total Unit 1,560,032 100.00 683,247 43.80 471,040 30.19 144,044 9.23 261,701 16.78 𝑁 1,555,319 681,193 469,334 143,450 261,342 No 2012 % (99.70) (99.70) (99.64) (99.59) (99.86) 𝑁 4,713 2,054 1,706 594 359 Yes % (.30) (.30) (.36) (.41) (.14) Total Unit 1,594,949 100.00 698,540 43.80 475,765 29.83 150,621 9.44 270,023 16.93 𝑁 1,589,924 696,372 473,895 150,072 269,585 No 2013 % (99.68) (99.69) (99.61) (99.64) (99.84) 𝑁 5,025 2,168 1,870 549 438 Yes % (.32) (.31) (.39) (.36) (.16) Total Unit 1,600,774 100.00 706,617 44.14 468,443 29.26 152,359 9.52 273,355 17.08 𝑁 1,595,458 704,410 466,431 151,772 272,845 No 2014 % (99.67) (99.69) (99.57) (99.61) (99.81) 𝑁 5,316 2,207 2,012 587 510 Yes % (.33) (.31) (.43) (.39) (.19) ∗ A patient could be hospitalized more than once during the study period, which resulted in more than one claim per patient. Thus, the number of claims in the study was higher than the number of patients. and traditional Korean medicine, the proportion of each outpatient cases occurred in the following order based on item was different. For Western medicine, the proportion the location of the disease: hand (11.24%) > knee (8.49%) > ofpsychiatriccostswasthehighest(28.78%),followedby ankle (9.63%) > shoulder (4.35%). According to time (year), doctors’ fees (27.7%), injections (16.59%), radiotherapy costs difference in current usage patterns was especially different (8.74%), and laboratory costs (7.09%). For traditional Korean between Western medicine and traditional Korean medi- medicine, the proportion of doctors’ fees was the highest cine. Particularly, hospitalization increases for the knee and (26.48%), followed by procedural costs (25.16%), injections shoulder areas (shoulder: 105.00%, knee: 250.00%) in tradi- (13.52%), admission costs (9.23%), and psychiatric costs tional Korean medicine were higher than those in Western (7.26%). Regarding traditional Korean medicine, most (70%) medicine (shoulder: 77.99%, knee: 58.04%). The proportion medical treatment costs were procedural costs and treatment of outpatient visits for the hand region in traditional Korean costs. Doctors’ fees accounted for only 21.54% of the total cost, medicine increased steadily (2012: 4.31%, 2013: 14.49%, 2014: similar to doctors’ fees for Western medicine. Procedural 13.46%). However, the shoulder area showed steady decrease costs accounted for the most (56.45%) among total cost for (2011: −8.31%, 2012: −9.73%, 2014: −16.45%) (Table 9). Western medicine. The second largest proportion was doc- The costs and length of hospitalization by year are shown tors’ fees (40.49%). Admission costs, medication costs, and in Tables 10–17. Basic physical therapy was the most com- laboratory costs comprised less than 1% (Table 7). mon nonsurgical intervention in Western medicine while In Table 8, it was not possible to use the inspection and acupuncture was the most common nonsurgical intervention image capturing system of 0 only in the Traditional Clinic in traditional Korean medicine. Both procedures are steady because of legal restrictions. According to the region of treatments that require two or more treatments. The propor- disease,thekneeaccountedforthemost,followedbythe tions of acupuncture and basic physical therapy are almost shoulder, foot, and hand in terms of hospitalization and out- equal (Table 18). patient visits. As the years progressed, the number of inpatient and outpatient visits was also increased for all body regions. Among the hospitalized patients, the number of 4. Discussion claims for all years after 2011 increased the most for shoulder joints (78.43%) compared to 2011, followed by knees (61.93%), This study assessed the prevalence and costs of most fre- foot (50.72%), and hands (16.29%). On the other hand, quently used treatments for joint disorders in Korea to 12 Evidence-Based Complementary and Alternative Medicine 2011–2014 2014 2013 2012 2011 2 1,253 1 1,151 3 946 1 1,060 7 4,410 108108 81 0140380110180 0 1,845 0 2,159 0 2,441 0 2,865 0 9,310 0 442 0 487 0 245 0 236 0 1,410 HOHOHOHOHO 12 493,842 35 485,603 37 497,972 52 483,833 136 1,961,250 184 4,035 286 7,932 366 7,680 469 8,818 1,305 28,465 196 497,877 321 493,535 403 505,652 521 492,651 1,441 1,989,715 355 7,860 348 10,697 383 11,051 455 10,817 1,541 40,425 1,123 29,569 1,263 46,307 1,543 47,052 1,677 52,028 5,606 174,956 2,491 3,778 3,504 8,461 4,237 8,142 4,925 8,247 15,157 28,628 1,296 867,535 1,383 858,553 1,493 874,403 1,508 881,056 5,680 3,481,547 3,232 62,241 4,061 127,047 4,493 131,644 5,193 136,668 16,979 457,600 8,695 1,472,414 10,881 1,548,435 12,555 1,581,587 14,280 1,585,646 46,411 6,188,082 8,499 974,537 10,560 1,054,900 12,152 1,075,935 13,759 1,092,995 44,970 4,198,367 ∗ % 14.91 58.92 12.71 55.45 11.89 55.29 10.56 55.56 12.24 56.26 %% 4.08% 12.92 0.53% 37.17 2.01 3.20 28.65 4.23% 11.61 0.26 0.69% 37.32 0.00 2.99 32.20 3.05 0.00 8.20 0.00 12.29 0.55 0.70 0.13 35.79 0.00 2.97 33.75 3.19 8.32 0.00 0.00 11.74 0.51 36.37 0.68 0.14 0.00 3.28 34.49 8.62 3.32 0.00 0.00 12.08 0.52 36.58 0.65 0.15 0.00 2.83 32.66 7.39 0.00 0.00 0.46 0.18 0.00 0.00 0.00 0.15 % 97.75 66.19 97.05 68.13 96.79 68.03 96.35 68.93 96.90 67.85 % 2.25 33.81 2.95 31.87 3.21 31.97 3.65 31.07 3.10 32.15 %% 0.00 0.02 0.03 0.09 0.00 0.01 0.03 0.07 0.00 0.02 0.02 0.06 0.00 0.01 0.01 0.07 0.00 0.02 0.02 0.07 %% 2.12 0.14 0.27 33.54 2.63 0.32 0.51 31.36 2.92 0.29 0.49 31.49 3.28 0.36 0.56 30.51 2.81 0.29 0.46 31.69 % 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 𝑁 Table 4: Number of hospitalizations and outpatient visits by type of medicine in 2011–2014. Total Total Hospital Western Clinic Dental Hospital General Hospital Tertiary Hospital Traditional Clinic Traditional Hospital Public Health Center Public Health Hospital Long-term care hospital Local Public Health Clinic with overlapping records were tallied as one patient (overlap was not allowed). H, hospitalization; O, outpatient; WM, Western medicine; TM, Korean traditional medicine. Patients Type of medicine Hospital type Unit WM TM Total ∗ Evidence-Based Complementary and Alternative Medicine 13 ‡ SLF † 2014 INSUP ∗ RPE ‡ SLF † 2013 INSUP ∗ RPE ‡ SLF † 2012 INSUP ∗ RPE ‡ Table 5: Average treatment amount and benefit per patient by year. SLF † 2011 INSUP ∗ RPE 61,372 44,201 16,944 58,572 42,469 15,878 61,113 44,077 16,623 65,319 47,296 17,866 82,019 61,845 20,154 86,266 65,161 21,076 90,827 68,827 21,979 96,325 73,323 22,966 72,576 50,798 21,766 75,827 53,013 22,798 80,484 56,216 24,252 86,426 60,587 25,825 59,589 41,837 17,681 60,432 42,435 17,970 62,989 44,297 18,650 67,402 47,452 19,899 80,779 56,892 23,850 83,765 58,891 24,844 88,453 62,166 26,253 94,310 66,510 27,765 84,267 58,835 25,372 89,666 62,667 26,939 92,840 64,659 27,997 98,049 68,492 29,482 121,156 86,493 34,503 127,488 91,277 36,107 134,378 96,345 37,875 141,772 101,542 40,027 511,061 393,692 115,248 536,884 410,668 124,195 562,200 430,102 129,674 572,730 437,335 132,812 193,314 139,945 52,833 196,220 141,450 54,582 207,876 150,072 57,232 216,450 157,025 59,170 214,173 161,693 51,469 228,679 172,504 55,286 246,375 185,404 59,464 250,494 188,780 60,578 181,225 135,985 44,240 186,151 139,155 46,017 194,305 145,102 48,164 198,661 147,900 49,621 167,073 123,186 43,449 156,721 114,800 41,323 165,529 121,119 43,911 172,643 125,955 46,092 172,322 129,370 42,262 178,367 133,526 44,130 187,228 140,111 46,235 192,414 143,780 47,833 185,933 140,190 45,052 192,290 144,552 47,009 202,967 152,456 49,617 208,739 156,519 51,398 318,775 247,050 70,286 321,522 247,790 72,013 334,876 258,239 74,930 335,457 257,336 76,387 225,139 171,769 53,297 232,227 176,546 55,573 243,557 185,039 58,411 248,061 187,859 60,108 199,087 150,625 47,563 211,708 159,868 50,817 223,060 168,461 53,604 273,424 207,024 65,595 203,081 153,881 48,398 209,020 157,743 50,440 221,676 167,186 53,420 157,944 115,659 42,148 120,367 88,290 31,661 121,813 89,057 32,334 126,287 92,267 33,586 131,932 96,429 35,035 212,462 161,156 50,725 204,987 155,273 49,011 230,624 174,488 55,662 371,890 287,348 82,390 130,549 95,580 33,405 135,980 99,443 34,934 145,561 106,375 37,180 153,724 112,670 39,211 377,448180,246 305,038 132,778 71,346 46,862 382,637 306,720 187,175 74,275 137,685 391,056 48,839 198,249 313,167 146,009 76,574 388,445 51,511 309,676 204,759 77,258 150,730 53,296 145,409 103,545 41,771 154,690 110,097 44,542 162,736 116,024 46,700 173,306 123,653 49,639 204,016 154,551 48,973 247,400 188,624 57,910 244,925 186,833 57,180 470,263 368,719 93,844 273,798 204,867 68,361 287,629 215,195 71,603 299,415 223,468 75,254 322,791 240,597 81,185 309,243 236,773 70,029 311,086 237,807 70,698 321,981 245,078 73,736 322,749 244,490 75,304 302,890 0 0 281,712 720 0 193,558 0 0 285,501 0 0 368,924 280,760 87,644 382,499 290,307 91,415 410,128 310,688 97,825 421,279 318,877 101,784 298,500 288,278 9,262 300,467 291,371 8,384 307,033 295,660 8,220 298,439 289,135 8,745 3,192,333 2,519,960 659,078 2,972,861 2,334,061 625,995 2,932,509 2,294,974 620,271 2,689,568 2,097,376 579,924 & & 70–79 NHI Medicaid M17 M17 60–69 Others 50–59 30–39 40–49 ≦ 29 1 ≦ 1 Both 2-3 4 ≦ Traditional Western Female Male M75 S63 S63 S93 0 1 2 3+ No Yes Medical insurance type Lesion of joint Age (yr) Inpatient visit Outpatient visit Type of medicine Gender Category Total Severity (CCI) Hospital admission 14 Evidence-Based Complementary and Alternative Medicine ‡ SLF † 2014 INSUP : the self-payment amount SLF ‡ ∗ RPE ‡ nce Review and Assessment Service) out of the SLF † 2013 INSUP ∗ RPE ‡ SLF † 2012 INSUP Table 5: Continued. : the cost reimbursed by the Korean National Health Insurance Service as the insurer. ∗ RPE INSUP † ‡ SLF † 2011 INSUP ∗ RPE 170,319 116,816 53,366 190,266 132,261 57,678 206,204 143,719 62,073 254,941 178,466 76,299 193,889 144,031 48,933 204,091 151,668 51,127 207,939 153,673 52,466 217,649 161,097 55,268 173,384 131,459 41,609 181,856 137,378 44,157 191,688 144,969 46,317 198,235 149,414 48,438 123,954 93,265 30,612 126,000 94,782 31,157 133,370 100,470 32,823 138,372 104,463 33,798 166,882 126,175 40,558 171,357 129,080 42,132 181,359 136,620 44,574 185,744 139,770 45,822 439,283 346,145 92,816 490,846 388,042 102,269 549,396 431,928 117,009 601,268 473,599 125,127 356,865 272,144 78,988 502,172 389,016 108,319 458,967 355,555 96,233 446,919 343,020 100,664 399,808 296,961 96,104 395,387 293,375 95,243 414,433 306,100 100,105 428,235 315,244 105,371 400,853 302,498 97,522 400,137 301,007 97,878 403,412 303,041 99,363 401,838 299,704 101,105 206,956 156,123 49,511 205,397 154,325 49,914 222,903 167,268 54,395 224,130 168,026 54,704 case involving one or more of the other three joint disorders which were included in this study. A & sum of INSUP and SLF paid to the medical care institution. The total costs of items determined to be eligible for reimbursement by the HIRA (Health Insura Seoul (Urban) Metropolitan city Other (Rural) Private Corporation Public Long-term care hospitalTraditional Hospital Traditional Clinic 151,730 115,386 36,252 152,247 115,222 36,955 161,165 121,970 39,115 164,468 124,275 40,104 Tertiary and General HospitalHospital 489,902Western Clinic 358,403 120,547 499,923 365,372 124,248 521,851 379,551 129,039 513,518 373,917 129,108 The Region Ownership Category Hospital type paid by the beneficiary. ∗ total treatment amount were indicated in the submitted insurance claim statement. Evidence-Based Complementary and Alternative Medicine 15

Table 6: Comparison of medical costs by type of medicine in 2011–2014.

Frequency Insurance charge LOS Unit Year Type of medicine ∗ † ‡ 𝑁 %Cost % Per diem Days per episode§ WM 8,499 97.75 16,542 99.37 162,276 10.98 2011 TM 196 2.25 105 0.63 53,097 12.23 Total 8,695 100 16,647 100 159,815 11.01 WM 10,560 97.05 18,063 99.03 157,605 9.67 2012 TM 321 2.95 178 0.97 57,256 12.18 Total 10,881 100 18,240 100 154,644 9.75 H WM 12,152 96.79 19,969 98.65 155,222 9.31 2013 TM 403 3.21 216 1.35 62,217 10.78 Total 12,555 100 20,185 100 152,237 9.35 WM 13,759 96.35 20,077 98.65 154,829 8.51 2014 TM 521 3.65 276 1.35 60,435 10.72 Total 14,280 100 20,353 100 151,385 8.59 WM 974,537 66.19 23,827 71.81 22,030 1.2 2011 TM 497,877 33.81 9,354 28.19 18,573 1.04 Total 1,472,414 100 33,181 100 20,861 1.15 WM 1,054,900 68.13 24,848 72.24 23,555 1.12 2012 TM 493,535 31.87 9,551 27.76 19,352 1.03 Total 1,548,435 100 34,399 100 22,215 1.09 O WM 1,075,935 68.03 26,203 72.24 24,354 1.11 2013 TM 505,652 31.97 10,071 27.76 19,917 1.03 Total 1,581,587 100 36,274 100 22,935 1.08 WM 1,092,995 68.93 27,887 73.19 25,514 1.1 2014 TM 492,651 31.07 10,218 26.81 20,740 1.03 Total 1,585,646 100 38,104 100 24,031 1.08 ∗ A patient could be hospitalized more than once during the study period, which resulted in more than one claim per patient. Thus, the number of claims in † the study was higher than the number of patients. Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. Costs are expressed in Korean Won (1,000,000 KRW). ‡ Per diem is the average daily cost of services covered by National Health Insurance. It is expressed in Korean Won. §Days per episode are the total number of reimbursed days divided by the total number of episodes. The number of reimbursed days includes the number of hospitalized days or outpatient visits and in-care drug prescription days. H, hospitalization; O, outpatient; WM, Western medicine; TM, Korean traditional medicine.

provide basic information for future usual care guidelines joint disease has been ranked the 6th among reasons for that may reduce health expenditures and help solve National inpatient care visit and the 5th among reasons for outpatient Health Insurance deficits. This study used the 2011–2014 care visit among the population aged 65 years or older [26]. HIRA-NPS data consisting of 3% age-stratified and gender- The incidence of gonarthrosis has been steadily increasing in stratified random samples. It appropriately reflected the Korea.Itsrateinwomenwasmuchhigherthanthatinmen South Korean population of 2011–2014 to capture real-world [27]. medical use and cost in joint disorders. While Western Clinics were the most frequently visited The results of the study showed that the proportion medical institution type between 2011 and 2014, the finding of female patients was higher compared to that of male that Traditional Clinics were the next most frequently visited patients. This is consistent with previous findings showing inthisstudywasnoteworthy.TheKoreanmedicalsystemis that women are more likely to utilize health care than characterized by both Western and traditional Korean med- men [23, 24]. This might be due to gender role differ- ical practices. In 2014, the number of claims from Western ences such as occupation, hours of work, and occupa- medicine was 49,031 for Tertiary and General Hospitals and tional activities including housework and biological fac- 16,935 for hospitals and clinics. On the contrary, the number tors. Women are typically responsible for childcare and of claims from traditional Korean medicine was 14,729 for housework while men are typically expected to have a job hospitals and 7,690 for clinics [26]. These circumstances [25]. reflect the high proportion of traditional Korean medicine Theshoulderandkneejointsaccountedforthemosthos- use for joint disorders [9]. Our results are consistent with pital visits and increased steeply. In Korea, musculoskeletal previous results showing that Traditional Clinics are the disease accounted for 28.2% of National Health Insurance second most visited institution by patients with nonspecific Corporation (NHIC) inpatient and outpatient claims. Knee low back pain [14]. 16 Evidence-Based Complementary and Alternative Medicine ∗ None Total costs Radiotherapy ated in the submitted insurance claim statement. costs Laboratory 3.61 7.09 8.74 0.68 100 24.5 4.95 6.07 1.34 100 31.01 2.63 4.88 1.91 100 costs 25.16 3.21 6 2.14 100 71.99 0.09 0 2.86 100 56.45 0.08 0 0.9 100 Procedural 3 6,932,215 8,192 0 275,017 9,628,946 0 0 0 0 0 0 31 21,791,571 31,038 0 347,946 38,602,696 costs Psychotherapy ts by type of medicine (2011–2014). 897 3,929 64,069,631 5,424,922 10,074,191 3,937,011 206,627,170 0 0 0 0 costs Physiotherapy view and Assessment Service) out of the total treatment amount were indic costs Anesthesia costs Injection Table7:Comparisonofsubgroupcos costs Medication costs Admission fees Doctors’ 8,138,295 185,481 946,907 3,632,780 697,677 6,300,815 197 7,723,344 1,559,874 1,914,274 423,554 31,523,198 2,074,061 6,648 332,810 0 0 6,064,234 178,833 614,097 3,632,780 697,677 6,300,815 194 791,129 1,551,682 1,914,274 148,537 21,894,252 % 29.1 7.8 1.3 10.99 4.49 5.91 % 25.82 0.59 3 11.52 2.21 19.99 % 27.7 0.82 2.8 16.59 3.19 28.78 % 26.48 9.23 1.47 13.52 5.52 7.26 %% 21.54 40.49 0.07 1.53 3.46 0.54 0 0 0 0 𝑁 𝑁 𝑁 KRW 60,131,134 16,107,497 2,686,093 22,718,121 9,271,746 12,202, KRW 44,499,914 15,515,383 2,477,317 22,718,121 9,271,746 12,202,897 3,897 42,278,059 5,393,884 10,074,191 3,589,065 168,024,474 KRW 15,631,220 592,114 208,776 0 0 determined to be eligible for reimbursement by the HIRA (Health Insurance Re Costs Type Unit WM TM Total They are expressed Korean as Won (1,000 KRW). ∗ Evidence-Based Complementary and Alternative Medicine 17 Total costs Radiotherapy costs Laboratory costs Injection costs/procedural costs Psychiatric costs Psychotherapy related to type of medicine. 𝑡 (.16) (18.89) (.00) (29.61) (2.62) (6.79) (100) costs Anesthesia costs Procedural costs Medication Table 8: Comparison of subgroup costs Admission fees (17.52) (2.44) (5.93) (6.24) (2.56) (14.34) (.00) (17.23) (19.82) (13.94) (100) (25.94) (.59) (3.89) (2.52) (2.25) (20.08) (.00) (33.67) (4.95) (6.10) (100) Doctors’ Total Hospital Western Clinic (31.99) (.18) (1.29) (2.58) (3.54) (35.12) (.00) (16.82) (1.98) (6.50) (100) Dental Hospital (17.57) (.00) (.42) (.00) (.00) (.00) (.00) (82.01) (.00) (.00) (100) General Hospital (17.14) (2.52) (9.43) (6.95) (1.87) (8.68) (.01) (15.35) (21.44) (16.61) (100) Tertiary Hospital (11.96) (2.36) (6.43) (7.60) (2.34) (3.56) (.01) (15.63) (30.35) (19.76) (100) Traditional Clinic (21.61) (.01) (6.44) (.00) (.09) (.00) (.00) (71.86) (.00) (.00) (100) Traditional Hospital (17.99) (3.57) (2.61) (.08) (.46) (1.44) (.00) (68.04) (4.17) (1.64) (100) Public Health Center (.00) (.00) (.08) (1.48) (.00) (96.83) (.00) (.35) (1.21) (.03) (100) Public Health Hospital (35.19) (.20) (4.47) (2.06) Long-term care hospital (23.67) (3.96) (3.39) (.76) (.87) (17.46) (.00) (44.73) (1.85) (3.31) (100) Local Public Health Clinic (.00) (.00) (49.13) (.34) (.00) (47.29) (.00) (2.68) (.55) (.00) (100) Type Hospital type TM WM WM, Western medicine; TM, Korean traditional medicine. 18 Evidence-Based Complementary and Alternative Medicine % ∗ 47 5.18 𝑁 907 100.00 860 94.82 123,147 49.65 124,891 50.35 248,038 100.00 † Ankle and foot level [S93] − 21.28 37 3.67 %Growth ∗ 5 2.26 221 100.00 𝑁 216 97.74 53,106 38.84 83,624 61.16 136,730 100.00 † Wrist and hand level [S63] − 5.43 209 100.00 11.03 1,007 100.00 − 5.56 204 97.61 12.79 970 96.33 − 20.00− 20.00 4 4 1.50 1.56 36.17 51.06 64 71 5.41 5.19 %Growth ∗ 40 1.64 𝑁 2,439 100.00 2,399 98.36 rate compared to 2011. H, hospitalization; O, outpatient; WM, Western medicine; TM, traditional Korean medicine. 185,699 41.76 258,936 58.24 444,635 100.00 Growth † † Shoulder lesions [M75] − 8.31 170,259 36.40 4.31 55,394 38.51 2.56 128,092 49.15 − 9.73 167,639 35.53 14.49 60,801 40.44 5.53 131,801 49.04 − 16.45 155,153 0.72 13.46 60,252 39.61 5.55 131,824 48.48 %Growth Table 9: Comparison of lesion frequency by type of medical practice in 2011–2014. ∗ 𝑁 104 2.03 5,128 100.00 5,024 97.97 134,181 20.87 643,011 100.00 508,830 79.13 † Knee lesions [M17] Growth TM TM 250.00 364 4.38 105.00 82 1.88 TM 150.96 261 3.55 85.00 74 1.97 TM 104.81 213 3.26 65.00 66 2.10 0.00 5 2.39 TM TM 5.90 142,092 20.37 TM 4.18TM 139,790 8.37 20.67 145,411 21.06 WM 25.72 6,316 96.74 27.97 3,070 97.90 WM WM WM 58.04 7,940 95.62 77.99 4,270 98.12 17.13 253 98.44 50.70 1,296 94.81 WM 41.04 7,086 96.45 53.61 3,685 98.03 21.76 263 98.50 30.00 1,118 94.59 WM 9.18 555,535 79.63 17.99 305,510 65.84 9.83 91,843 60.39 13.77 140,107 51.52 WM 5.41WM 536,381 7.14 79.33 545,161 14.91 78.94 297,536 17.50 63.60 304,245 5.76 64.47 88,439 7.08 61.49 89,546 59.56 7.63 11.24 132,544 50.85 136,983 50.96 Total Total 27.32Total 6,529 43.27Total 100.00 7,347 61.93 28.58 100.00 8,304 3,136 54.12 100.00 100.00 78.43 3,759 100.00 4,352 20.81 100.00 16.29 267 100.00 257 30.32 100.00 50.72 1,182 100.00 1,367 100.00 Total Total 8.49 697,627 100.00 4.35 463,993 100.00 11.24 152,095 100.00 9.63 271,931 100.00 Total 7.40 690,572 100.00 6.13 471,884 100.00 9.96 150,347 100.00 8.36 268,784 100.00 Total 5.16 676,171 100.00 5.21 467,795 100.00 5.19 143,833 100.00 5.08 260,636 100.00 2011 2011 2013 2013 2012 2012 2014 2014 with overlapping records were tallied as one patient (overlap was not allowed). Patients Class Year Type H O ∗ Evidence-Based Complementary and Alternative Medicine 19

Table 10: Numbers of hospitalizations for gonarthrosis [M17] patients by hospital type.

Frequency Hospitalization costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 %Cost % Per diem episode§ Tertiary Hospital 204 3.98 1,225 9.50 460,420 14.20 General Hospital 589 11.49 3,215 24.93 254,225 22.53 Hospital 1,689 32.94 7,044 54.62 225,222 17.68 Western Long-term care hospital 1,857 36.21 553 4.29 94,819 4.42 Western Clinic 683 13.32 794 6.16 81,749 13.64 2011 Public Health Hospital 2 0.04 1 0.01 39,420 11.00 Total 5,024 97.97 12,832 99.50 170,393 12.66 Traditional Hospital 99 1.93 62 0.48 57,349 13.10 Traditional Traditional Clinic 5 0.10 2 0.02 33,235 13.80 Total 104 2.03 64 0.50 56,190 13.13 Total 5,128 100.00 12,896 100.00 168,077 12.67 Tertiary Hospital 187 2.86 1,190 8.81 490,276 14.06 General Hospital 615 9.42 3,188 23.60 236,169 22.07 Hospital 2,011 30.80 7,435 55.05 212,520 15.80 Western Long-term care hospital 2,792 42.76 700 5.18 99,643 3.50 Western Clinic 710 10.87 868 6.42 78,343 13.69 2012 Public Health Hospital 1 0.02 1 0.01 42,477 19.00 Total 6,316 96.74 13,383 99.08 158,039 10.69 Traditional Hospital 190 2.91 114 0.84 62,827 12.07 Traditional Traditional Clinic 23 0.35 11 0.08 33,646 15.57 Total 213 3.26 125 0.92 59,676 12.45 Total 6,529 100.00 13,507 100.00 154,830 10.74 Tertiary Hospital 195 2.65 1,287 8.79 454,263 16.49 General Hospital 725 9.87 3,681 25.16 234,228 21.27 Hospital 1,997 27.18 7,678 52.49 215,432 16.46 Western Long-term care hospital 3,393 46.18 888 6.07 98,884 3.62 Western Clinic 774 10.53 939 6.42 84,784 12.97 2013 Public Health Hospital 2 0.03 1 0.01 44,311 16.00 Total 7,086 96.45 14,474 98.95 153,802 10.42 Traditional Hospital 234 3.18 140 0.96 64,218 11.45 Traditional Traditional Clinic 27 0.37 14 0.10 42,487 13.56 Total 261 3.55 154 1.05 61,970 11.67 Total 7,347 100.00 14,628 100.00 150,539 10.47 Tertiary Hospital 272 3.28 1,641 11.46 459,732 14.81 General Hospital 727 8.75 3,723 26.01 233,047 22.00 Hospital 2,152 25.92 6,911 48.29 205,374 14.52 Western Long-term care hospital 4,022 48.43 1,045 7.30 102,035 3.44 Western Clinic 766 9.22 783 5.47 75,945 13.00 2014 Public Health Hospital 1 0.01 0 0.00 63,640 4.00 Total 7,940 95.62 14,103 98.54 151,771 9.46 Traditional Hospital 329 3.96 193 1.35 64,804 11.00 Traditional Traditional Clinic 35 0.42 16 0.11 38,910 13.77 Total 364 4.38 209 1.46 62,314 11.26 Total 8,304 100.00 14,312 100.00 147,849 9.54 ∗ † Patients with overlapping records were tallied as one patient (overlap was not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are ‡ expressed as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of hospitalized days divided by the total number of hospitalizations. 20 Evidence-Based Complementary and Alternative Medicine

Table 11: Number of outpatients with gonarthrosis [M17] by hospital type. Frequency Outpatient costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 % Cost % Per diem episode§ Tertiary Hospital 3,771 0.59 199 1.35 39,490 1.76 General Hospital 13,999 2.18 689 4.69 32,279 6.26 Hospital 28,259 4.39 1,667 11.33 28,169 2.04 Long-term care hospital 1,872 0.29 95 0.65 15,504 3.51 Western Clinic 458,027 71.23 9,641 65.55 21,049 1.02 Public Health Center 1,495 0.23 5 0.04 3,474 1.00 2011 Local Public Health Clinic 389 0.06 5 0.03 12,991 7.70 Public Health Hospital 1,018 0.16 27 0.18 26,723 1.67 Total 508,830 79.13 12,329 83.83 21,823 1.24 Traditional Hospital 1,022 0.16 49 0.33 16,099 2.81 Traditional Traditional Clinic 133,159 20.71 2,330 15.84 17,497 1.04 Total 134,181 20.87 2,379 16.17 17,486 1.05 Total 643,011 100.00 14,707 100.00 20,918 1.20 Tertiary Hospital 4,817 0.71 215 1.42 44,547 1.55 General Hospital 20,949 3.10 720 4.76 34,367 4.58 Hospital 54,982 8.13 1,858 12.27 33,789 1.12 Long-term care hospital 4,519 0.67 99 0.66 22,003 1.42 Clinic 448,226 66.29 9,691 64.01 21,621 1.02 Western Public Health Center 1,592 0.00 6 0.00 3,811 1.01 Local Public Health Clinic 439 0.24 3 0.04 7,361 4.12 2012 Public Health Hospital 855 0.06 26 0.02 30,431 1.46 Dental Hospital 2 0.13 0 0.17 24,970 1.00 Total 536,381 79.33 12,618 83.34 23,525 1.18 Traditional Hospital 2,332 0.34 50 0.33 21,369 1.16 Traditional Traditional Clinic 137,458 20.33 2,472 16.33 17,986 1.03 Total 139,790 20.67 2,522 16.66 18,042 1.03 Total 676,171 100.00 15,140 100.00 22,391 1.15 Tertiary Hospital 4,949 0.72 218 1.37 44,048 1.57 General Hospital 21,512 3.12 783 4.92 36,416 4.20 Hospital 55,199 7.99 1,879 11.81 34,034 1.12 Long-term care hospital 4,780 0.69 104 0.66 21,854 1.49 Western Clinic 456,160 66.06 10,203 64.13 22,367 1.01 Public Health Center 1,678 0.24 6 0.04 3,442 1.05 2013 Local Public Health Clinic 191 0.03 2 0.02 12,633 6.90 Public Health Hospital 692 0.10 23 0.15 33,579 1.97 Total 545,161 78.94 13,219 83.09 24,248 1.16 Traditional Hospital 2,306 0.33 52 0.32 22,375 1.13 Traditional Traditional Clinic 143,105 20.72 2,639 16.59 18,443 1.03 Total 145,411 21.06 2,691 16.91 18,505 1.03 Total 690,572 100.00 15,910 100.00 23,039 1.14 Tertiary Hospital 5,377 0.77 243 1.43 45,133 1.24 General Hospital 22,999 3.30 856 5.06 37,234 3.85 Hospital 57,441 8.23 1,986 11.74 34,571 1.11 Long-term care hospital 4,748 0.68 106 0.63 22,320 1.48 Clinic 461,684 66.18 10,962 64.81 23,742 1.01 Western Public Health Center 2,372 0.00 8 0.00 3,298 1.00 Local Public Health Clinic 171 0.34 2 0.05 12,230 6.67 2014 Public Health Hospital 732 0.02 21 0.01 28,470 1.31 Dental Hospital 11 0.10 0 0.12 18,469 1.00 Total 555,535 79.63 14,183 83.86 25,531 1.15 Traditional Hospital 2,516 0.36 59 0.35 23,541 1.13 Traditional Traditional Clinic 139,576 20.01 2,670 15.79 19,127 1.03 Total 142,092 20.37 2,729 16.14 19,205 1.03 Total 697,627 100.00 16,912 100.00 24,242 1.12 ∗ † Patients with overlapping records were tallied as one patient (overlap was not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They is ‡ expressed as a mean and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of outpatient visit days including drug prescription days divided by the total number of outpatient visits. Evidence-Based Complementary and Alternative Medicine 21

Table 12: Number of hospitalizations for shoulder lesion [M75] patients by hospital type.

Frequency Hospitalization costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 %Cost % Per diem episode§ Tertiary Hospital 135 5.54 278 9.55 377,378 6.08 General Hospital 381 15.62 765 26.31 210,523 12.83 Western Hospital 1,101 45.14 1,608 55.27 203,071 8.86 Long-term care hospital 589 24.15 69 2.37 94,749 1.45 2011 Western Clinic 193 7.91 167 5.74 99,580 11.10 Total 2,399 98.36 2,886 99.24 179,142 7.69 Traditional Hospital 39 1.6 22 0.75 53203 14.26 Traditional Traditional Clinic 1 0.04 0 0.01 21440 14 Total 40 1.64 22 0.76 52409 14.25 Total 2,439 100.00 2,908 100.00 177,064 7.80 Tertiary Hospital 142 4.53 318 8.32 374,833 6.37 General Hospital 468 14.92 989 25.85 208,846 12.43 Western Hospital 1,547 49.33 2,168 56.66 200,108 8.36 Long-term care hospital 661 21.08 75 1.96 98,519 1.25 2012 Western Clinic 252 8.04 242 6.33 138,525 9.81 Total 3,070 97.90 3,792 99.13 182,594 7.48 Traditional Hospital 62 1.98 32 0.84 58,945 11.53 Traditional Traditional Clinic 4 0.13 1 0.03 31,056 14.25 Total 66 2.10 33 0.87 57,255 11.70 Total 3,136 100.00 3,826 100.00 179,956 7.57 Tertiary Hospital 169 4.50 388 8.65 379,396 6.41 General Hospital 582 15.48 1,212 27.01 188,152 13.07 Hospital 1,920 51.08 2,524 56.27 193,851 7.58 Western Long-term care hospital 739 19.66 100 2.22 95,331 1.68 Western Clinic 274 7.29 223 4.97 156,193 8.29 2013 Public Health Hospital 1 0.03 0 0.00 89,370 1.00 Total 3,685 98.03 4,446 99.12 178,875 7.26 Traditional Hospital 68 1.81 37 0.82 64,497 10.57 Traditional Traditional Clinic 6 0.16 3 0.06 40,844 14.17 Total 74 1.97 40 0.88 62,579 10.86 Total 3,759 100.00 4,485 100.00 176,585 7.33 Tertiary Hospital 161 3.70 333 6.97 377,875 5.68 General Hospital 671 15.42 1,214 25.45 202,746 10.41 Western Hospital 2,298 52.80 2,773 58.12 188,633 6.86 Long-term care hospital 817 18.77 134 2.80 99,291 2.07 2014 Western Clinic 323 7.42 276 5.79 164,902 8.50 Total 4,270 98.12 4,730 99.12 179,097 6.58 Traditional Hospital 76 1.75 39 0.83 60,405 10.09 Traditional Traditional Clinic 6 0.14 2 0.05 23,633 17.17 Total 82 1.88 42 0.88 57,714 10.61 Total 4,352 100.00 4,772 100.00 176,810 6.66 ∗ † Patients with overlapping records were tallied as one patient (overlap not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are expressed ‡ as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of hospitalized days divided by the total number of hospitalizations. 22 Evidence-Based Complementary and Alternative Medicine

Table 13: Number of outpatients with shoulder lesions [M75] by hospital type. Frequency Outpatient costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 % Cost % Per diem episode§ Tertiary Hospital 3,279 0.74 147 1.55 29,079 2.14 General Hospital 8,754 1.97 420 4.42 26,730 4.09 Hospital 16,434 3.70 803 8.45 21,712 2.21 Long-term care hospital 1,120 0.25 47 0.50 16,906 2.84 Clinic 228,892 51.48 4,559 47.99 19,916 1.01 Western Dental Hospital 8 0.00 0 0.00 18,765 1.63 Public Health Center 301 0.07 1 0.01 3,204 1.00 2011 Local Public Health Clinic 22 0.00 0 0.00 5,281 1.91 Public Health Hospital 126 0.03 3 0.03 23,603 1.25 Total 258,936 58.24 5,980 62.96 20,345 1.21 Traditional Hospital 1,327 0.30 62 0.66 17,637 2.58 Traditional Traditional Clinic 184,372 41.47 3,456 36.39 18,746 1.04 Total 185,699 41.76 3,519 37.04 18,738 1.05 Total 444,635 100.00 9,499 100.00 19,674 1.14 Tertiary Hospital 4,805 1.03 150 1.54 31,203 1.33 General Hospital 16,148 3.45 418 4.29 25,911 2.07 Hospital 42,956 9.18 1,004 10.29 23,362 1.05 Long-term care hospital 2,372 0.51 52 0.53 21,931 1.40 Clinic 230,518 49.28 4,790 49.14 20,780 1.01 Western Dental Hospital 5 0.00 0 0.00 26,800 1.00 Public Health Center 552 0.12 2 0.02 3,290 1.00 2012 Local Public Health Clinic 12 0.00 0 0.00 9,058 3.92 Public Health Hospital 168 0.04 4 0.04 22,856 3.33 Total 297,536 63.60 6,420 65.85 21,577 1.08 Traditional Hospital 2,348 0.50 52 0.53 22,053 1.31 Traditional Traditional Clinic 167,911 35.89 3,277 33.62 19,517 1.03 Total 170,259 36.40 3,329 34.15 19,552 1.04 Total 467,795 100.00 9,749 100.00 20,840 1.07 Tertiary Hospital 5,053 1.07 167 1.63 33,032 1.29 General Hospital 15,611 3.31 434 4.26 27,831 2.06 Hospital 46,514 9.86 1,108 10.85 23,823 1.05 Long-term care hospital 2,105 0.45 49 0.48 23,288 1.45 Clinic 234,099 49.61 5,077 49.73 21,688 1.01 Western Dental Hospital 6 0.00 0 0.00 17,590 1.00 Public Health Center 701 0.15 2 0.02 2,487 1.00 2013 Local Public Health Clinic 24 0.01 0 0.00 6,914 2.96 Public Health Hospital 132 0.03 3 0.03 25,864 1.74 Total 304,245 64.47 6,841 67.01 22,486 1.08 Traditional Hospital 2,163 0.46 51 0.50 23,727 1.09 Traditional Traditional Clinic 165,476 35.07 3,317 32.49 20,047 1.03 Total 167,639 35.53 3,369 32.99 20,095 1.03 Total 471,884 100.00 10,210 100.00 21,636 1.06 Tertiary Hospital 4,474 0.96 149 1.42 33,281 1.34 General Hospital 17,733 3.82 502 4.79 28,320 2.11 Hospital 48,171 10.38 1,136 10.85 23,590 1.04 Long-term care hospital 2,106 0.45 50 0.47 23,507 1.36 Western Clinic 232,309 50.07 5,315 50.74 22,881 1.01 Public Health Center 471 0.10 2 0.02 3,848 1.37 2014 Local Public Health Clinic 35 0.01 0 0.00 7,430 3.80 Public Health Hospital 211 0.05 6 0.06 28,187 1.34 Total 305,510 65.84 7,160 68.35 23,437 1.09 Traditional Hospital 3,330 33.44 73 30.96 21,781 1.03 Traditional Traditional Clinic 155,153 0.72 3,243 0.69 20,903 1.03 Total 158,483 34.16 3,316 31.65 20,921 1.03 Total 463,993 100.00 10,476 100.00 22,578 1.07 ∗ † Patients with overlapping records were tallied as one patient (overlap was not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are ‡ expressed as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of outpatient visit days including drug prescription days divided by the total number of outpatient visits. Evidence-Based Complementary and Alternative Medicine 23

Table 14: Number of hospitalizations for wrist and hand level lesions [S63] by hospital type.

Frequency Hospitalization costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 %Cost % Per diem episode§ Tertiary Hospital 11 4.98 13 7.68 278,227 5.45 General Hospital 45 20.36 43 25.65 179,511 6.20 Western Hospital 97 43.89 74 44.40 139,886 6.36 Long-term care hospital 11 4.98 1 0.84 61,642 2.18 2011 Western Clinic 52 23.53 34 20.56 82,895 9.83 Total 216 97.74 165 99.13 137,482 6.90 Traditional Hospital 4 1.81 1 0.59 50,891 4.75 Traditional Traditional Clinic 1 0.45 0 0.28 32,616 14.00 Total 5 2.26 1 0.87 47,236 6.60 Total 221 100.00 166 100.00 135,440 6.90 Tertiary Hospital 9 4.31 10 6.32 248,187 4.89 General Hospital 32 15.31 31 20.03 181,560 8.19 Western Hospital 112 53.59 87 55.96 127,213 7.51 Long-term care hospital 13 6.22 1 0.83 99,160 1.00 2012 Western Clinic 38 18.18 25 15.80 91,169 9.32 Total 204 97.61 154 98.93 132,573 7.42 Traditional Hospital 2 0.96 1 0.61 53,661 8.00 Traditional Traditional Clinic 3 1.44 1 0.46 41,900 7.33 Total 5 2.39 2 1.07 46,604 7.60 Total 209 100.00 156 100.00 130,517 7.43 Tertiary Hospital 11 4.12 13 7.25 342,630 3.73 General Hospital 50 18.73 47 26.27 156,802 6.76 Western Hospital 111 41.57 85 46.85 130,977 6.95 Long-term care hospital 44 16.48 4 1.98 80,871 1.09 2013 Western Clinic 47 17.60 30 16.67 88,549 10.30 Total 263 98.50 179 99.03 128,774 6.40 Traditional Hospital 4 1.50 2 0.97 53,481 8.00 Traditional Total 4 1.50 2 0.97 53,481 8.00 Total 267 100.00 180 100.00 127,646 6.42 Tertiary Hospital 13 5.06 12 5.82 210,379 4.00 General Hospital 65 25.29 71 33.81 167,548 8.48 Western Hospital 118 45.91 97 46.30 137,949 6.69 Long-term care hospital 14 5.45 1 0.54 81,609 1.00 2014 Western Clinic 43 16.73 27 12.62 88,134 8.91 Total 253 98.44 208 99.08 137,691 7.08 Traditional Hospital 3 1.17 1 0.59 64,672 8.00 Traditional Traditional Clinic 1 0.39 1 0.33 36,168 19.00 Total 4 1.56 2 0.92 57,546 10.75 Total 257 100.00 210 100.00 136,444 7.13 ∗ † Patients with overlapping records were tallied as one patient (overlap was not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are ‡ expressed as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of hospitalized days divided by the total number of hospitalizations. 24 Evidence-Based Complementary and Alternative Medicine

Table 15: Number of outpatients for wrist and hand level lesions [S63] by hospital type. Frequency Outpatient costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 % Cost % Per diem episode§ Tertiary Hospital 251 0.18 19 0.60 58,745 1.58 General Hospital 2,612 1.91 155 4.95 44,742 1.62 Hospital 7,082 5.18 319 10.18 29,974 1.58 Long-term care hospital 321 0.23 12 0.37 18,678 1.99 Western Clinic 73,276 53.59 1,618 51.67 22,082 1.00 Public Health Center 39 0.03 0 0.01 5,699 1.00 2011 Local Public Health Clinic 13 0.01 0 0.01 18,520 11.46 Public Health Hospital 30 0.02 1 0.02 18,767 1.23 Total 83,624 61.16 2,124 67.81 23,546 1.08 Traditional Hospital 545 0.40 21 0.68 18,048 2.18 Traditional Traditional Clinic 52,561 38.44 987 31.51 18,777 1.03 Total 53,106 38.84 1,008 32.19 18,770 1.04 Total 136,730 100.00 3,132 100.00 21,691 1.06 Tertiary Hospital 306 0.21 16 0.48 51,987 1.36 General Hospital 3,404 2.37 146 4.41 42,810 1.19 Hospital 11,477 7.98 357 10.80 31,071 1.04 Long-term care hospital 667 0.46 13 0.40 19,954 1.10 Western Clinic 72,525 50.42 1,680 50.86 23,163 1.01 Public Health Center 1 0.00 0 0.00 5,790 1.00 2012 Local Public Health Clinic 13 0.01 0 0.00 8,726 4.00 Public Health Hospital 46 0.03 1 0.04 29,252 1.43 Total 88,439 61.49 2,213 66.99 25,022 1.02 Traditional Hospital 838 0.58 18 0.53 20,947 1.05 Traditional Traditional Clinic 54,556 37.93 1,073 32.48 19,664 1.02 Total 55,394 38.51 1,090 33.01 19,683 1.02 Total 143,833 100.00 3,303 100.00 22,966 1.02 Tertiary Hospital 297 0.20 19 0.54 65,123 1.40 General Hospital 3,512 2.34 163 4.58 46,354 1.19 Hospital 10,971 7.30 360 10.13 32,801 1.07 Long-term care hospital 484 0.32 10 0.28 20,811 1.07 Clinic 74,188 49.34 1,758 49.50 23,690 1.00 Western Public Health Center 34 0.00 0 0.00 3,086 1.00 Local Public Health Clinic 7 0.02 0 0.00 7,674 3.00 2013 Public Health Hospital 52 0.00 1 0.00 24,083 1.23 Dental Hospital 1 0.03 0 0.04 24,540 1.00 Total 89,546 59.56 2,311 65.09 25,808 1.02 Traditional Hospital 956 0.64 22 0.62 22,855 1.05 Traditional Traditional Clinic 59,845 39.80 1,218 34.30 20,350 1.02 Total 60,801 40.44 1,240 34.91 20,390 1.02 Total 150,347 100.00 3,551 100.00 23,617 1.02 Tertiary Hospital 332 0.22 21 0.57 63,661 1.34 General Hospital 3,886 2.55 180 4.80 46,221 1.44 Hospital 11,366 7.47 372 9.96 32,755 1.05 Long-term care hospital 646 0.42 14 0.38 21,995 1.20 Clinic 75,535 49.66 1,868 49.94 24,725 1.00 Western Public Health Center 10 0.00 0 0.00 4,050 1.00 Local Public Health Clinic 14 0.01 0 0.00 8,487 3.43 2014 Public Health Hospital 50 0.01 1 0.00 24,903 1.18 Dental Hospital 4 0.03 0 0.03 19,768 1.00 Total 91,843 60.39 2,456 65.69 26,745 1.03 Traditional Hospital 892 0.59 20 0.55 22,893 1.08 Traditional Traditional Clinic 59,360 39.03 1,263 33.77 21,273 1.02 Total 60,252 39.61 1,283 34.31 21,297 1.02 Total 152,095 100.00 3,740 100.00 24,587 1.03 ∗ † Patients with overlapping records were tallied as one patient (overlap was not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are ‡ expressed as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of outpatient visit days including drug prescription days divided by the total number of outpatient visits. Evidence-Based Complementary and Alternative Medicine 25

Table 16: Number of hospitalizations for ankle and foot level lesions [S93] by hospital type.

Frequency Hospitalization costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 %Cost % Per diem episode§ Tertiary Hospital 5 0.55 8 1.24 173,270 10.40 General Hospital 108 11.91 123 18.23 98,380 12.77 Western Hospital 345 38.04 280 41.41 84,468 10.59 Long-term care hospital 34 3.75 6 0.96 65,373 3.26 2011 Western Clinic 368 40.57 241 35.54 56,559 12.45 Total 860 94.82 659 97.38 74,034 11.37 Traditional Hospital 42 4.63 17 2.47 49,012 9.38 Traditional Traditional Clinic 5 0.55 1 0.15 34,441 6.80 Total 47 5.18 18 2.62 47,462 9.11 Total 907 100.00 677 100.00 72,657 11.25 Tertiary Hospital 10 0.99 12 1.61 193,439 6.00 General Hospital 148 14.70 156 20.70 108,791 11.49 Western Hospital 391 38.83 305 40.51 88,346 9.74 Long-term care hospital 38 3.77 12 1.61 82,398 5.34 2012 Western Clinic 383 38.03 250 33.20 59,570 11.50 Total 970 96.33 734 97.63 80,954 10.49 Traditional Hospital 32 3.18 16 2.17 46,740 12.50 Traditional Traditional Clinic 5 0.50 2 0.20 32,100 9.80 Total 37 3.67 18 2.37 44,761 12.14 Total 1,007 100.00 752 100.00 79,624 10.55 Tertiary Hospital 8 0.68 13 1.41 206,721 7.75 General Hospital 186 15.74 192 21.49 117,806 9.83 Western Hospital 465 39.34 390 43.78 104,115 8.94 Long-term care hospital 61 5.16 14 1.57 93,890 3.52 2013 Western Clinic 398 33.67 262 29.42 64,569 11.39 Total 1,118 94.59 871 97.67 92,491 9.65 Traditional Hospital 60 5.08 20 2.23 66,044 6.93 Traditional Traditional Clinic 4 0.34 1 0.11 22,968 11.50 Total 64 5.41 21 2.33 63,352 7.22 Total 1,182 100.00 891 100.00 90,913 9.52 Tertiary Hospital 9 0.66 10 0.92 200,250 6.00 General Hospital 214 15.65 234 22.08 127,396 10.31 Western Hospital 625 45.72 533 50.31 105,591 8.72 Long-term care hospital 72 5.27 24 2.27 92,879 5.67 2014 Western Clinic 376 27.51 235 22.22 63,569 10.69 Total 1,296 94.81 1,036 97.81 96,951 9.36 Traditional Hospital 61 4.46 20 1.90 58,725 7.28 Traditional Traditional Clinic 10 0.73 3 0.29 25,945 12.80 Total 71 5.19 23 2.19 54,108 8.06 Total 1,367 100.00 1,059 100.00 94,726 9.30 ∗ † Patients with overlapping records were tallied as one patient (overlap not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are expressed ‡ as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of hospitalized days divided by the total number of hospitalizations. 26 Evidence-Based Complementary and Alternative Medicine

Table 17: Number of outpatient visits for ankle and foot level lesions [S93] by hospital type. Frequency Outpatient costs LOS Type of Year Hospital type ∗ † ‡ Days per medicine 𝑁 % Cost % Per diem § episode Tertiary Hospital 559 0.23 46 0.79 71,456 1.84 General Hospital 4,204 1.69 285 4.88 54,910 1.60 Hospital 10,466 4.22 571 9.77 35,586 1.69 Long-term care hospital 465 0.19 18 0.31 23,088 1.90 Western Clinic 107,340 43.28 2,471 42.30 23,025 1.00 Dental Hospital 6 0.00 0 0.01 16,350 2.67 2011 Public Health Center 10 0.00 0 0.00 3,012 1.00 Local Public Health Clinic 18 0.01 0 0.00 7,036 2.56 Public Health Hospital 79 0.03 2 0.04 31,605 1.52 Total 123,147 49.65 3,395 58.10 25,402 1.09 Traditional Hospital 1,141 0.46 45 0.78 18,915 2.03 Traditional Traditional Clinic 123,750 49.89 2,403 41.12 19,416 1.03 Total 124,891 50.35 2,448 41.90 19,412 1.04 Total 248,038 100.00 5,843 100.00 22,386 1.06 Tertiary Hospital 769 0.30 50 0.80 64,899 1.51 General Hospital 5,806 2.23 310 5.00 53,426 1.27 Hospital 17,632 6.76 625 10.07 35,457 1.06 Long-term care hospital 903 0.35 19 0.31 21,587 1.12 Western Clinic 107,284 41.16 2,589 41.71 24,132 1.00 Dental Hospital 31 0.01 1 0.01 26,883 1.00 2012 Public Health Center 14 0.01 0 0.00 4,871 1.00 Local Public Health Clinic 23 0.01 0 0.00 9,318 4.30 Public Health Hospital 82 0.03 3 0.04 31,415 1.39 Total 132,544 50.85 3,597 57.96 27,142 1.03 Traditional Hospital 2,414 0.93 54 0.86 22,190 1.06 Traditional Traditional Clinic 125,678 48.22 2,556 41.18 20,336 1.03 Total 128,092 49.15 2,609 42.04 20,371 1.03 Total 260,636 100.00 6,207 100.00 23,814 1.03 Tertiary Hospital 752 0.28 50 0.76 66,703 1.68 General Hospital 6,417 2.39 357 5.40 55,586 1.22 Hospital 18,960 7.05 675 10.22 35,603 1.05 Long-term care hospital 773 0.29 17 0.25 21,454 1.10 Western Clinic 109,956 40.91 2,731 41.35 24,835 1.00 Dental Hospital 4 0.00 0 0.00 25,545 1.00 2013 Public Health Center 28 0.01 0 0.00 2,908 1.00 Local Public Health Clinic 23 0.01 0 0.00 6,697 2.83 Public Health Hospital 70 0.03 2 0.03 30,855 1.23 Total 136,983 50.96 3,832 58.02 27,972 1.02 Traditional Hospital 2,255 0.84 52 0.79 23,228 1.05 Traditional Traditional Clinic 129,546 48.20 2,720 41.18 20,994 1.02 Total 131,801 49.04 2,772 41.98 21,032 1.02 Total 268,784 100.00 6,604 100.00 24,569 1.02 Tertiary Hospital 634 0.23 53 0.76 83,468 1.66 General Hospital 7,410 2.72 415 5.95 56,052 1.27 Hospital 19,690 7.24 715 10.25 36,321 1.05 Long-term care hospital 747 0.27 17 0.25 23,249 1.08 Clinic 111,528 41.01 2,884 41.34 25,857 1.00 Western Dental Hospital 3 0.00 0 0.00 19,330 1.00 2014 Public Health Center 12 0.00 0 0.00 4,125 1.00 Local Public Health Clinic 16 0.01 0 0.00 10,259 4.63 Public Health Hospital 67 0.02 2 0.03 28,492 1.03 Total 140,107 51.52 4,087 58.58 29,169 1.03 Traditional Hospital 2,080 0.76 51 0.73 24,517 1.04 Traditional Traditional Clinic 129,744 47.71 2,839 40.69 21,880 1.02 Total 131,824 48.48 2,890 41.42 21,922 1.02 Total 271,931 100.00 6,977 100.00 25,656 1.03 ∗ † Patients with overlapping records were tallied as one patient (overlap not allowed). Costs determined to be eligible for reimbursement by the HIRA (Health Insurance Review and Assessment Service) out of the total treatment amount were indicated in the submitted insurance claim statement. They are expressed ‡ as means and are in Korean Won (1,000,000 KRW). Per diem is the average daily cost of services covered by National Health Insurance. §Days per episode are the total number of outpatient visit days including drug prescription days divided by the total number of outpatient visits. Evidence-Based Complementary and Alternative Medicine 27

Table 18: Distribution of nonsurgical interventions in Western medicine and traditional Korean medicine.

Year 2011 2012 2013 2014 Nonsurgical intervention Total ∗ 𝑁 1,481,969 1,560,032 1,594,949 1,600,774 𝑁 947,982 1,000,863 1,034,024 1,033,566 0 % (63.97) (64.16) (64.83) (64.57) 𝑁 32,582 38,311 46,856 45,063 1 † % (2.20) (2.46) (2.94) (2.82) WM Basic physical therapy 𝑁 136,773 137,871 126,126 122,500 2 % (9.23) (8.84) (7.91) (7.65) N 364,632 382,987 387,943 399,645 3≦ % (24.60) (24.55) (24.32) (24.97) 𝑁 1,475,271 1,551,724 1,585,834 1,590,708 0 ‡ % (99.55) (99.47) (99.43) (99.37) WM Simple rehabilitation 𝑁 6,698 8,308 9,115 10,066 1≦ % (.45) (.53) (.57) (.63) 𝑁 1,481,873 1,560,008 1,594,914 1,600,736 0 % (99.99) (100.00) (100.00) (100.00) WM Professional rehabilitation§ 𝑁 96 24 35 38 1≦ % (.01) (.00) (.00) (.00) 𝑁 1,481,962 1,560,029 1,594,942 1,600,773 0 % (100.00) (100.00) (100.00) (100.00) WM Rehabilitation of CNS 𝑁 7371 1≦ % (.00) (.00) (.00) (.00) 𝑁 982,471 1,062,109 1,085,224 1,102,972 0 % (66.29) (68.08) (68.04) (68.90) 𝑁 40,678 44,984 55,124 50,447 TM Acupuncture 1 % (2.74) (2.88) (3.46) (3.15) 𝑁 458,820 452,939 454,601 447,355 2≦ % (30.96) (29.03) (28.50) (27.95) 𝑁 1,481,922 1,559,985 1,594,876 1,600,667 0 % (100.00) (100.00) (100.00) (99.99) TM Moxibustion 𝑁 47 47 73 107 1≦ % (.00) (.00) (.00) (.01) 𝑁 1,481,712 1,559,844 1,594,897 1,600,518 0 % (99.98) (99.99) (100.00) (99.98) TM Cupping 𝑁 257 188 52 256 1≦ % (.02) (.01) (.00) (.02) 𝑁 1,481,856 1,559,985 1,594,897 1,600,734 0 % (99.99) (100.00) (100.00) (100.00) TM Heat & cold therapy 𝑁 113 47 52 40 1≦ % (.01) (.00) (.00) (.00) ∗ A patient could be hospitalized more than once during the study period, resulting in more than one claim per patient. Thus, the number of claims in † the study was higher than the number of patients. Basic physical therapy included superficial heat therapy, cold therapy, deep heat therapy, ultraviolet ‡ irradiation, transcutaneous electrical nerve stimulation, massage therapy, and simple therapeutic exercise. Simple rehabilitation included paraffin bath, hydrotherapy, intermittent traction therapy, electrical stimulation therapy, laser therapy, therapeutic exercise, motor point block, pneumatic compression, complex decongestive physical therapy, and iontophoresis. §Professional rehabilitation included pool therapy, occupational therapy, activities of daily living training, neurogenic bladder training, functional electrical stimulation therapy, myofascial trigger point injection, rehabilitative social work, rehabilitative breathing therapy, rehabilitative functional training, and rehabilitative dysphagia therapy. WM, Western medicine; TM, traditional Korean medicine.

Despite the high demand for traditional Korean medicine subgroup costs related to the type of medicine in total for musculoskeletal diseases, traditional Korean medical treatment cost. Apart from procedures such as acupuncture, practitioners are precluded from diagnosing joint disorders moxibustion, and cupping, many treatments were not cov- independently due to regulatory restrictions in imaging ered by the NHI (Table 7). Large variations in diagnostic device use. We confirmed this fact again by comparing and therapeutic management between Western medicine 28 Evidence-Based Complementary and Alternative Medicine and traditional Korean medicine indicate that more items Recently, research results have been utilized as basis for in Korean medicine need to be covered and developed. policies by utilizing health-related big data. However, there is Among the hospitalized patients, the number of claims for all alackofdataanalyzingvariouspatternsoftraditionalmedical years after 2011 increased mostly for shoulder joints (78.43%) services [30]. Although there are studies that use NHI claims compared to 2011, followed by knees (61.93%), foot (50.72%), data, they are limited to a single year or disease range [8, 13]. and hands (16.29%). While the rate of use of traditional This study is novel in that it compared the utilization of West- Korean medicine for the shoulder region slightly decreased ern medicine and traditional Korean medicine for the treat- in outpatient care, the number of hospitalizations increased ment of joint disorders in Korea. We believe the current sharply (Table 9). It is interesting that the proportion of tradi- study would serve as a good reference for countries with tional hospitalization increase for the knee and shoulder similar medical systems as that of Korea and would be able to regions (shoulder: 105.00%, knee: 250.00%) was higher than contribute to international literature. Further research is that of Western medicine clinic hospitalization (shoulder: required, such as analysis of factors influencing the use or 77.99%, knee: 58.04%). In Korea, the medical delivery system frequency of Korean traditional medicine using multivariate of traditional medicine is not strict. Individuals can choose statistics. to visit primary medical institutions (Traditional Clinic) and Second, while fee-for-service for nationally covered higher medical institutions (Traditional Hospital). Shoulder health care service was comprehensively recorded in the and knee joint diseases are common musculoskeletal dis- claim database, nonreimbursable items such as traditional eases. They are usually treated in primary care settings. How- drugs did not generate billing data. In addition, we only ever, if there is no response or a lack of effectiveness in calculated direct medical costs based on information in primary care, a Traditional Hospital is attended for a more the claim database. In general, there are nonmedical costs accurate diagnosis and evaluation. This might be the reason why there is increase in hospitalization in Traditional Hospi- such as transportation costs and lost productivity due to tals that hire orthodox medical practitioners who can use X- morbidity because joint diseases tend to be chronic. In addi- rays and magnetic resonance imaging. Besides, because tra- tion, the costs did not uncover items based on claims data ditional medical practitioners cannot use these examination that only contained information about medical services pro- devices in a Traditional Clinic, outpatient care in Traditional vided under the NHI. If uncovered items were included, the Clinics was much lower compared to that in Western Clinics. costs for traditional Korean medicine might be higher than Although simple radiology is helpful in the diagnosis of that of Western medicine. In a previous study [31] comparing joint disease [28], patient might suffer the inconvenience of Western and traditional Korean medicine users, it was found going to both Western and Traditional Clinics for accurate that the traditional Korean medicine user group paid signif- diagnosis due to legal restrictions [29]. icantly more medical expenses than the Western medicine Traditional Korean medicine had lower medical expen- user group. In a study on the determinants of traditional ditures than Western medicine (inpatient care cost: Western Korean medicine use based on panel data [32], the number of medicine clinic, 160,000 KRW; traditional Korean medicine patients using traditional Korean medicine was significantly clinic, 50,000 KRW; outpatient care cost: Western medicine, higher than those using Western medicine. This is because 22,000 KRW; traditional Korean medicine, 18,000 KRW). The the insurance benefit for traditional Korean medicine is lower average treatment cost for traditional Korean medicine was than that for Western medicine. lowercomparedtothatforWesternmedicine.RPEranged Third, we did not include essential factors influencing from 181,225KRW to 198,661KRW for Western medicine and choice of medical practice, such as education level, income, 82,019 KRW to 96,325 KRW for traditional Korean medicine. residence, severity, and health-related risk factors (e.g., alco- In addition, out-of-pocket expenses for Western medicine hol consumption, smoking, and exercise) [33]. Previous (44,240 KRW to 49,621 KRW) were higher than those of studies have shown that factors affecting the use of traditional traditional Korean medicine (20,154 KRW to 22966 KRW). Koreanmedicinearenotrelatedtoeducationlevelorincome Although expenditures for traditional Korean medicine were significantly lower compared to those for Western medicine, level [34, 35], high education level [14], low education level daily cost amount showed no significant difference between [36, 37], and low or high-income level [32]. In previous the two depending on the year. These results are similar to studies, factors such as the use of a therapist [33], confidence those of a previous study [6]. in oriental medical institution [32], recognition of therapeutic Most patients had mild joint diseases (more than 70%) effect [38], and coping attitude of the oriental medical treat- with CCI scores of 0 and underwent nonsurgical treatment. ment [39] have been found to be significant factors. Although Regardless of disease type, the proportion of surgery was less these essential variables are important parameters in choos- than 1%. Therefore, traditional medical care can serve as inghospitals,theyweremissedduetothenatureoftheclaims an alternative to Western medical care. We found that the data. To overcome the omission of disturbance variables, proportion of acupuncture was slightly higher than basic medical insurance type (NHI or Medicaid) and region of hos- physical therapy (Table 18). However, further research is pital were used as surrogates of income and residence in this needed to confirm that traditional Korean medicine is cost- study. In addition, we considered disease severity by using effective for managing joint diseases. CCI because severity of disease might greatly affect hospital This study has several limitations. First, the study was choice. The remaining factors have been judged due to descriptive in nature. It reported sociodemographic charac- their impacts on health care utilization. We believe that the teristics, procedures, medication, and average cost for treat- direction of the analysis of this study will not change. Due ing joint disorders without addressing a specific hypothesis. to the limitation in data characteristics, these elements were Evidence-Based Complementary and Alternative Medicine 29 not included in this study. Future studies considering these were produced by physicians based on diagnosis at the first factors are needed to confirm our findings. visit to the hospital. Related claim data were then produced, Finally, the accuracy of diagnoses has been an issue due including the diagnosis code, the date of initiation of treat- to the nature of claims data collected with the purpose of ment (hospitalization or outpatient), and personal informa- reimbursing health care services and not for clinical purposes tion (age, sex). To use epidemiological data, it is necessary [18]. The accuracy of diagnosis in the KNHI claims data has to link the billing statement classified for administrative pur- been reported to be about 70% [40]. Moreover, the accuracy poses to the same hospitalization case [45]. In this study, the of disease classification has been reported to be higher same patient filed a billing statement with the same hospital for inpatients than for outpatients. It is higher for severe in the same medical institution for inpatient care. The date disorders than for common mild disorders. It is also higher was connected to one hospitalization case. in General Hospitals than in clinics [41]. Nonetheless, in the Our study has several policy implications. As disease process of designing this study, physicians in current practice structure can change from acute to chronic degenerative dis- concurred that these codes were not clearly differentiated ease, interest in traditional medicine is increasing with aging. for diagnosis in actual clinical practice settings in Korea. The main purpose of traditional medicine use in South Korea Therefore, analysis was performed in primary and secondary diagnoses in accordance with the opinion that various issues is to prevent disease and promote general health [10]. Though shouldbetakenintoaccount(e.g.,privateinsurance,medical traditional medicine plays a substantial role in the Korean care institution characteristics, and individual differences in health care system, the annual number of health insurance physicians) in category division. In addition, primary and claims from traditional Korean medicine institutions has secondary diagnoses are generally used in conjunction [14]. stagnated and decreased since 2012 [46]. Medical use is Therefore, in defining medical care usage due to joint disease, affected by demographic, socioeconomic, and psychocultural we reviewed not only the major diagnosis code, but also factors. It has been reported that these factors can affect secondary diagnosis code. Despite our efforts, the diagnosis health care utilization by interactions between factors rather accuracy for joint diseases in this study might be challenged. than independent factors [47, 48]. Therefore, it is important Despite these limitations, our study has several strengths. to grasp the current position of traditional Korean medicine First, we analyzed age-stratified and gender-stratified ran- in order to prepare policy and directives. Currently, difference dom samples of the KNHI claims database representing in standards of practice underlies mistrust for traditional 98% of the South Korean population. Claims statements Korean medicine among Western medical practitioners [6]. covered extensive information on health care interventions To overcome conflicts among orthodox and traditional prac- (e.g., treatment, procedures, diagnostic tests, and prescrip- titioners, we need an effective health care delivery system that tion drugs), diagnosis, NHI payment cost, beneficiaries’ encourages consultation for both Western and traditional out-of-pocket expenses, sociodemographic characteristics, Korean medicine with accessibility. Further discussion must and medical institutions, thus providing useful nationwide be considered by providing consultation programs for other epidemiological data. Its representativeness, reliability, and chronic diseases and joint diseases. validity have been confirmed previously [41]. However, there is a lack of data necessary to understand various consumption patterns and supply patterns of traditional medical services 5. Conclusions [30]. Second, there are studies in other countries that analyze This study provided objective information about epidemio- the status of traditional Korean medicine utilization by using logic characteristics of patients with joint disorders treated representative data source [30, 42–44]. Unlike most previous with Western medicine and traditional Korean medicine. clinical studies whose duration was less than one year, we It provided an understanding about the recent status and attempted to analyze the change over four years for joint trends. It will provide a basis for further expansion of tradi- diseases based on the type of medicine used (Western or tional Korean medicine for patients with muscular disorders. traditional Korean medicine). Until now, no studies have Based on HIRA data, medical use for joint disorders showed reported national data on the management of joint disorders significant difference between the groups. It provides basic for 2011 to 2014. This study holds significance in that itis information for future usual care guidelines linked to health the first study that reports distinct differences in patterns of policy and budget appropriation. Timely and accurate infor- medical care use and costs between Western medicine and mation is essential for policy-makers to make decisions. The traditional Korean medicine. results of our study will contribute to management decisions An added strength of this study was that it provided patterns of complementary and alternative medicine treat- for musculoskeletal diseases involving joint disorders. ment for joint disorders in Korea by covering traditional Korean medicine treatments as acupuncture, moxibustion, Abbreviations andcuppingintheNHI. Third, we constructed pilot medical episode data consid- HIRA: Health Insurance Review & Assessment Services ering characteristics of health claim data for joint diseases. NHIS: National Health Insurance Sample Thiscanbeusedasadataprocessingtechniquetocalculate WM: Western medicine basic dynamics information. Health insurance claim data TM: Traditional Korean medicine. 30 Evidence-Based Complementary and Alternative Medicine

Additional Points the Korean Rheumatism Association,vol.16,no.4,pp.281–290, 2009. Availability of Data and Material. Datasets generated and/or [5]S.Brage,J.F.Nygard,andG.Tellnes,“Thegendergapin analyzed in the current study are available from the HIRA- musculoskeletal-related long term sickness absence in Norway,” NPS repository. HIRA data are third-party data not owned Scandinavian Journal of Social Medicine,vol.26,no.1,pp.34–43, by the authors. HIRA data are available upon visit or by mail 1997. upon direct, email, or fax submission of dataset request form [6]A.D.WoolfandK.Akesson,˚ “Understanding the burden of and declaration of data use downloadable from the “HIRA” musculoskeletal conditions,” British Medical Journal,vol.322, website: [http://opendata.hira.or.kr/op/opc/selectPatDataAplIn- no. 7294, pp. 1079-1080, 2001. foView.do]anduponpaymentofthetransferofdatarequest [7] G. Bodeker, “Lessons on integration from the developing fee (300,000 KRW per dataset). world’s experience,” British Medical Journal,vol.322,no.7279, pp.164–167,2001. Ethical Approval [8] Y.-J. Ahn, J.-S. Shin, J. Lee et al., “Evaluation of use and cost of medical care of common lumbar disorders in Korea: The HIRA Research Ethics Committee of South Korea Cross-sectional study of Korean Health Insurance Review and approved the protocol of this study. Assessment Service National Patient Sample data,” BMJ Open, vol. 6, no. 9, Article ID e012432, 2016. [9]J.KimandM.Son,“NationalHealthInsuranceStatistical Disclosure Yearbook,” Seoul: Health Insurance Review and Assessment Service and National Health Insurance Service, 2017. Written informed consent was not obtained from participants [10] E. M. Badley, I. Rasooly, and G. K. Webster, “Relative impor- because only clinical records were used in this study based on tance of musculoskeletal disorders as a cause of chronic health national billing data submitted to HIRA. Patient information problems, disability, and health care utilization: Findings from was anonymized and deidentified by HIRA prior to analysis the 1990 Ontario Health Survey,” The Journal of Rheumatology, in this study. vol.21,no.3,pp.505–514,1994. [11] P. C. Coyte, C. V. Asche, R. Croxford, and B. Chan, “The eco- Conflicts of Interest nomic cost of musculoskeletal disorders in Canada,” Arthritis &Rheumatology,vol.11,no.5,pp.315–325,1998. The authors declare that there are no conflicts of interest [12] D. P. Lubeck, “The costs of musculoskeletal disease: Health regarding the publication of this article. needs assessment and health economics,” Best Practice & Research Clinical Rheumatology,vol.17,no.3,pp.529–539,2003. [13] H. Joo, Y.J. Lee, J. Shin et al., “Medical service use and usual care Authors’ Contributions of common shoulder disorders in Korea: a cross-sectional study Soyoon Kim constructed study design, Sukjin Bae analyzed using the Health Insurance Review and Assessment Service National Patient Sample,” BMJ Open,vol.7,no.7,ArticleID thedata,andBoyoungJunginterpretedtheresults.Soyoon e015848, 2017. Kim was scientific advisor and directed this study. Boyoung [14] J. M. Gill and A. G. Mainous III, “The Role of Provider Con- Jungwrotethemanuscriptandperformedthestatistical tinuity in Preventing Hospitalizations,” Archives of Family analysis. All authors read and approved the final manuscript. Medicine,vol.7,no.4,pp.352–357,1998. [15] E. Yelin and L. F. Callahan, “Special article the economic cost Acknowledgments and social and psychological impact of musculoskeletal conditions,” Arthritis & Rheumatism,vol.38,no.10,pp.1351–1362, ThisstudywassupportedbyResearchInstituteofKorean 1995. Medicine Policy. [16] L. J. Weiss and J. Blustein, “Faithful patients: The effect of long- term physician-patient relationships on the costs and use of References health care by older Americans,” American Journal of Public Health,vol.86,no.12,pp.1742–1747,1996. [1] P.M. Brooks, “The burden of musculoskeletal disease—a global [17] J. McLeod, J. McMurray, J. D. Walker, G. A. Heckman, and P. perspective,” Clinical Rheumatology,vol.25,no.6,pp.778–781, Stolee, “Care transitions for older patients with musculoskeletal 2006. disorders: continuity from the providers’ perspective,” Interna- [2] I.-H. Oh, S.-J. Yoon, H.-Y. Seo, E.-J. Kim, and Y. A. Kim, “The tional Journal of Integrated Care, vol. 11, no. 2, 2011. economic burden of musculoskeletal disease in Korea: A cross [18] L. Kim, J.-A. Kim, and S. Kim, “A guide for the utilization sectional study,” BMC Musculoskeletal Disorders,vol.12,article of Health Insurance Review and Assessment Service National no.157,2011. Patient Samples,” Epidemiology and Health,vol.36,ArticleID [3]N.Hur,C.Choi,W.Uhm,andS.Bae,“ThePrevalenceand e2014008, 2014. Trend of Arthritis in Korea: Results from Korea National Health [19]D.T.Felson,A.Naimark,J.Anderson,L.Kazis,W.Castelli, and Nutrition Examination Surveys,” The Journal of the Korean and R. F. Meenan, “The prevalence of knee osteoarthritis in Rheumatism Association,vol.15,no.1,pp.11–26,2008. the elderly. The Framingham Osteoarthritis Study,” Arthritis & [4] H. Choi, W.Han, J. Im, and H. Baek, “The Prevalence and Clin- Rheumatism, vol. 30, no. 8, pp. 914–918, 1987. ical Features of Musculoskeletal Diseases in Incheon: Results [20] G. Tellnes and T. Bjerkedal, “Epidemiology of sickness certi- from Chronic Disease Management Surveys,” The Journal of fication - A methodological approach based on a study from Evidence-Based Complementary and Alternative Medicine 31

Buskerud county in Norway,” Scandinavian Journal of Social women,” Journal of Orthopaedic Science,vol.13,no.5,pp.413– Medicine,vol.17,no.3,pp.245–251,1989. 418, 2008. [21] V. Sundararajan, T. Henderson, C. Perry, A. Muggivan, H. [38]Y.Zhang,L.Xu,M.C.Nevittetal.,““Comparisonofthe Quan, and W. A. Ghali, “New ICD-10 version of the Charlson prevalence of knee osteoarthritis between the elderly Chinese comorbidity index predicted in-hospital mortality,” Journal of population in Beijing and whites in the United States: The Clinical Epidemiology, vol. 57, no. 12, pp. 1288–1294, 2004. Beijing Osteoarthritis Study,” Arthritis & Rheumatology,vol.44, [22] H. Quan, V. Sundararajan, P. Halfon et al., “Coding algorithms no.9,pp.2065–2071,2001. for defining comorbidities in ICD-9-CM and ICD-10 adminis- [39] H. Du, S.-L. Chen, C.-D. Bao et al., “Prevalence and risk factors trative data,” Medical Care, vol. 43, no. 11, pp. 1130–1139, 2005. of knee osteoarthritis in Huang-Pu District, Shanghai, China,” [23] P.Sloane and C. Egelhoff, “The relationship of continuity of care Rheumatology International,vol.25,no.8,pp.585–590,2005. to age, sex, and race,” JournalofFamilyPractice,vol.16,no.2,pp. [40] J. Kim, “Basis of using health insurance data, strategic, and 402–405, 1983. assignment from the computation of health statistic,” 2005. [24] M. F. Fleming, E. J. Bentz, E. J. Shahady, A. Abrantes, and C. [41] A. D. Woolf, “The bone and joint decade 2000-2010,” Annals of Bolick, “Effect of case mix on provider continuity,” Journal of the Rheumatic Diseases,vol.59,no.2,pp.81-82,2000. Family Practice,vol.23,no.2,pp.137–140,1986. [42] M. Cross, E. Smith, D. Hoy et al., “The global burden of hip and [25] R. C. Lawrence, C. G. Helmick, F. C. Arnett et al., “Estimates knee osteoarthritis: estimates from the global burden of disease of the prevalence of arthritis and selected musculoskeletal 2010 study,” Annals of the Rheumatic Diseases,vol.73,no.7,pp. disorders in the United States,” Arthritis & Rheumatology,vol. 1323–1330, 2014. 41, no. 5, pp. 778–799, 1998. [43] J. S. Hong, H. C. Kang, and J. Kim, “Continuity of care for [26] J. Kim and M. Son, “National Health Insurance Statistical elderly patients with diabetes mellitus, hypertension, asthma, Yearbook. Seoul: Health Insurance Review and Assessment and chronic obstructive pulmonary disease in Korea,” Journal Service and National Health Insurance Service,” 2014. of Korean Medical Science,vol.25,no.9,pp.1259–1271,2010. [27] H.-A. Kim, S. Kim, Y. I. Seo et al., “The epidemiology of [44] M.C.Sokol,K.A.McGuigan,R.R.Verbrugge,andR.S.Epstein, total knee replacement in South Korea: national registry data,” “Impact of medication adherence on hospitalization risk and Rheumatology,vol.47,no.1,pp.88–91,2008. healthcare cost,” Medical Care,vol.43,no.6,pp.521–530,2005. [28] J. H. Kellgren and J. S. Lawrence, “Radiological assessment of [45] G. Peat, R. McCarney, and P. Croft, “Knee pain and osteoarthri- osteo-arthrosis,” Annals of the Rheumatic Diseases,vol.16,no. tis in older adults: a review of community burden and current 4, pp. 494–501, 1957. use of primary health care,” Annals of the Rheumatic Diseases, [29] A. D. Woolf and B. Pfleger, “Burden of major musculoskeletal vol. 60, no. 2, pp. 91–97, 2001. conditions,” Bulletin of the World Health Organization,vol.81, [46] Y. J. Song, “The south korean health care system,” Japan Medical no. 9, pp. 646–656, 2003. Association Journal,vol.52,no.3,pp.206–209,2009. [30] K. Han, I. Ha, and J. Lee, “Application of Health Care Big data [47] C.-H. Yoon, S.-J. Lee, S. Choo, O.-R. Moon, and J.-H. Park, and Necessity of Traditional Korean Medicine Data Registry,” “Continuity of care of patient with diabetes and its affecting Journal of Korean Medicine for Obesity Research,vol.17,no.1, factors in Korea,” JournalofPreventiveMedicine&Public pp.46–53,2017. Health,vol.40,no.1,pp.51–58,2007. [31]X.Kang,M.Fransen,Y.Zhangetal.,“Thehighprevalenceof [48] J. Hong, J. Kim, and H. Kang, “Continuity of Ambulatory knee osteoarthritis in a rural chinese population: the wuchuan Care among Adult Patients with Type 2 Diabetes and Its osteoarthritis study,” Arthritis Care Research,vol.61,no.5,pp. Associated Factors in Korea,” Korean Journal of Health Policy 641–647, 2009. and Administration,vol.19,no.2,pp.51–70,2009. [32] I. Kim, H. A. Kim, Y.-I. Seo, Y. W. Song, J.-Y. Jeong, and D. H. Kim, “The prevalence of knee osteoarthritis in elderly community residents in Korea,” Journal of Korean Medical Science,vol. 25,no.2,pp.293–298,2010. [33] M.Blagojevic,C.Jinks,A.Jeffery,andK.P.Jordan,“Riskfactors for onset of osteoarthritis of the knee in older adults: a sys- tematicreviewandmeta-analysis,”Osteoarthritis and Cartilage, vol.18,no.1,pp.24–33,2010. [34] Y.-J. Choi, S.-H. Kang, and Y.-I. Kim, “Association of higher continuity of primary care with lower risk of hospitalization among children and adolescent patients,” Korean Journal of Health Policy and Administration,vol.18,pp.85–107,2008. [35]J.M.Gill,A.G.Mainous3rd,J.J.Diamond,andM.J.Lenhard, “Impact of provider continuity on quality of care for persons with diabetes mellitus,” Annals of Family Medicine,vol.1,no.3, pp. 162–170, 2003. [36] D. A. Christakis, L. Mell, T. D. Koepsell, F. J. Zimmerman, and F. A. Connell, “Association of lower continuity of care with greater risk of emergency department use and hospitalization in children,” Pediatrics,vol.107,no.3,pp.524–529,2001. [37]A.Sudo,N.Miyamoto,K.Horikawaetal.,“Prevalenceand risk factors for knee osteoarthritis in elderly Japanese men and Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 7383104, 15 pages https://doi.org/10.1155/2017/7383104

Research Article Kyungheechunggan-Tang-01, a New Herbal Medication, Suppresses LPS-Induced Inflammatory Responses through JAK/STAT Signaling Pathway in RAW 264.7 Macrophages

Hee-Soo Han,1,2 Eungyeong Jang,3 Ji-Sun Shin,1 Kyung-Soo Inn,4 Jang-Hoon Lee,3 Geonha Park,2 Young Pyo Jang,2,5 and Kyung-Tae Lee1,2

1 Department of Pharmaceutical Biochemistry, College of Pharmacy, Kyung Hee University, Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea 2Department of Life and Nanopharmaceutical Sciences, College of Pharmacy, Kyung Hee University, Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea 3Department of Internal Medicine, College of Korean Medicine, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea 4Department of Pharmaceutical Science, College of Pharmacy, Kyung Hee University, Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea 5Department of Oriental Pharmaceutical Sciences, College of Pharmacy, Kyung Hee University, 26 Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea

Correspondence should be addressed to Kyung-Tae Lee; [email protected]

Received 25 August 2017; Revised 3 November 2017; Accepted 9 November 2017; Published 29 November 2017

Academic Editor: Ho Lin

Copyright © 2017 Hee-Soo Han et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Medicinal plants have been used as alternative therapeutic tools to alleviate inflammatory diseases. The objective of this study was to evaluate anti-inflammatory properties of Kyungheechunggan-tang- (KCT-) 01, KCT-02, and Injinchunggan-tang (IJCGT) as newly developed decoctions containing 3–11 herbs in LPS-induced macrophages. KCT-01 showed the most potent inhibitory effects on LPS-induced NO, PGE2,TNF-𝛼, and IL-6 production among those three herbal formulas. In addition, KCT-01 significantly inhibited LPS-induced iNOS and COX-2 at protein levels and expression of iNOS, COX-2, TNF-𝛼, and IL-6 at mRNA levels. Molecular data revealed that KCT-01 attenuated the activation of JAK/STAT signaling cascade without affecting NF-𝜅BorAP- 1 activation. In ear inflammation induced by croton oil, KCT-01 significantly reduced edema, MPO activity, expression levels of iNOS and COX-2, and STAT3 phosphorylation in ear tissues. Taken together, our findings suggest that KCT-01 can downregulate the expression of proinflammatory genes by inhibiting JAK/STAT signaling pathway under inflammatory conditions. This study provides useful data for further exploration and application of KCT-01 as a potential anti-inflammatory medicine.

1. Introduction system activator can induce large amounts of proinflammatory mediators, nitric oxide (NO), prostaglandin E2 (PGE2), Inflammation is a complicated biological response caused and cytokines such as interleukin- (IL-) 1𝛽,IL-6,andtumor by many harmful stimuli, including pathogens, bacteria, and 𝛼 𝛼 irritants [1]. It is also a tissue protective reaction of immune necrosis factor- (TNF- ). These proinflammatory media- cells including macrophages. Initiation of inflammation is tors and cytokines can lead to inflammation and various triggered by interactions between surface receptors such as clinical manifestations [3]. Overexpressed proinflammatory Toll-like receptor- (TLR-) 4 and TLR-2 and their ligands, mediators further exacerbate immune responses in many including lipopolysaccharides (LPS) derived from bacteria acute and chronic inflammatory diseases, including arte- [2]. When activated by LPS, macrophage as a potent immune riosclerosis, inflammatory bowel disease, arthritis, infectious 2 Evidence-Based Complementary and Alternative Medicine

Table 1: Prescription of KCT-01, KCT-02, and IJCGT.

Formula Scientific name KCT-01 Artemisia Capillaris Herba, Sanguisorbae Radix, and Curcuma longa Radix KCT-02 Artemisia Capillaris Herba, Sanguisorbae Radix,Curcuma longa Radix, Rubi Fructus, and Salviae Miltiorrhizae Radix Artemisia Capillaris Herba, Sanguisorbae Radix, Rubi Fructus, Atractylodis Rhizoma Alba, Poria Sclerotium, Polyporus IJCGT Sclerotium, Alismatis Rhizoma, Glycyrrhizae Radix, Raphani Semen, Citrus Unshiu Immaturi Pericarpium,andZingiberis Rhizoma Crudus

diseases, and cancer [4]. Accordingly, materials or com- effect in vitro using LPS-induced RAW 264.7 macrophages pounds that can inhibit these proinflammatory mediators and in vivo using croton oil-induced ear edema rat model. have been considered as potential anti-inflammatory agents. Janus kinase- (JAK-) signal transducer and activator 2. Materials and Methods oftranscription(STAT)cascadeisacriticalinflammatory signaling pathway that mediates immune responses [5]. In 2.1. Materials and Reagents. The herbal plants were pur- particular, STATs have been reported to play a pivotal role chased from Kyung Hee Herb Pharm (Wonju, Republic in inflammatory signaling cascades triggered by LPS and of Korea), a licensed company with Good Manufacturing several cytokines such as interferon gamma (IFN-𝛾)[6– Product (GMP) facilities. Briefly, 149.66 g of KCT-01, 198.34 g 8]. TLR-4 activation results in phosphorylation of receptor- of KCT-02, and 307.13g of IJCGT were extracted twice associated enzymes JAKs known to be activators of STATs [9]. in 1500 ml, 2000 ml, and 3000 ml of 30% EtOH for 3 h ∘ Binding of ligands to their receptors leads to phosphorylation at 80–85 C. After each extracted solution was processed of JAKs, which will induce phosphorylation of STATs that through a filtration process, we could get 29.45 g of KCT- in turn leads to the release of STATs from the receptor 01 (yield = 19.63%), 44.26 g of KCT-02 (yield = 22.13%), complex. Released STATs can form homo- or heterodimers and 56.36 g of IJCGT (yield = 18.78%) in forms of freeze- and translocate into the nucleus to regulate transcription dried powders of each formula. HPLC-grade formic acid of target genes encoding proinflammatory cytokines and was purchased from Wako (Osaka, Japan) and HPLC- inducible enzymes such as inducible nitric oxide synthase grade acetonitrile was obtained from Fisher Scientific Korea [1] and cyclooxygenase-2 (COX-2) [10–12]. Since JAK-STAT (Seoul, South Korea). Reference standards of chlorogenic signaling pathway is involved in the mediation of proinflam- acid, isochlorogenic acid A, isochlorogenic acid B, and matory gene expression, it is important to tightly regulate its ziyuglycosideIwerepurchasedfromChemfaces(Wuhan, activity to prevent inflammatory responses. Hubei, China). Ellagic acid was purchased from Chromadex Viral infection, high consumption of alcohol, fat accu- (Irvine, CA, USA) and hyperoside, jaceosidin, eupatilin, and mulation, and toxic agent are the main causes of chronic glycyrrhizic acid were purchased from Chengdu Biopurify inflammation in the liver [13]. Chronic inflammation causes Phytochemicals Ltd. (Chengdu, Sichuan, China). Scoparone pathological changes in liver function and, therefore, might and luteolin were purchased from Sigma-Aldrich (St. Louis, progress severe problems such as liver fibrosis and cancer MO, USA) and curcumin was purchased from HWI Ana- [14]. Thus, controlling inflammatory responses in the liver lytik GmbH (Rheinzabern, Germany). Dulbecco’s modified is desirable in the management of severe hepatic diseases. eagle’s medium (DMEM), fetal bovine serum (FBS), and Many traditional herbal medicines have been revealed to have penicillin-streptomycin (PS) were obtained from Life Tech- pharmacological properties against inflammatory injury [15]. nologies Inc. (Grand Island, NY, USA). COX-2 antibody Injinchunggan-tang (IJCGT) is a herbal medicine containing was purchased from Cayman (MI, USA), and phospho- 11 medicinal herbs. It has been widely used to treat hepatic STAT1 and phosphor-STAT3 antibodies were purchased diseases in Kyung Hee University Korean Medicine Hospital from Cell Signaling Technology Inc. (Beverly, MA). iNOS, (Seoul, Republic of Korea) [16–21]. Based on IJCGT, we reor- phospho-JAK1/2, JAK1/2, STAT1/3, and 𝛽-actin antibodies ganized new herbal medications called Kyungheechunggan- and peroxidase-conjugated secondary antibody were pur- tang-(KCT-)01whichcontainedArtemisia Capillaris Herba, chased from Santa Cruz Biotechnology, Inc. (CA, USA). Sanguisorbae Radix, and Curcuma longa Radix.KCT-02 The enzyme immunoassay [4] kits for PGE2,TNF-𝛼,and is an extended decoction after adding Rubi Fructus and IL-6 were obtained from R&D Systems (MN, USA). Ran- Salviae Miltiorrhizae Radix to KCT-01 (Table 1). Although dom oligonucleotide primers and M-MLV reverse tran- hepatoprotective effect of IJCGT and clinical case report have scriptase were purchased from Promega (WI, USA). SYBR been demonstrated previously, scientific evidence to support green ex Taq was purchased from TaKaRa (Shiga, Japan). the effect of KCT-01 or KCT-02 and their fundamental mech- COX-2, iNOS, TNF-𝛼, IL-6, IFN-𝛽,and𝛽-actin oligonu- anism of action are currently unclear. Therefore, the objective cleotide primers were obtained from Bioneer (Seoul, Korea). ofthepresentstudywastocompareanti-inflammatoryeffects 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bro- ofKCT-01,KCT-02,andIJCGT.Afterthat,themostpotent mide (MTT), NS-398, LPS (Escherichia coli, serotype 0111:B4), medicine among the three was selected to unravel the detailed and all other chemicals were purchased from Sigma Chemical molecular mechanism involved in its anti-inflammatory Co. (MO, USA). Evidence-Based Complementary and Alternative Medicine 3

2.2. UPLC-PDA-ESI-MS Analysis. AWatersAcquity6 H- for Animal Care and Use of the Kyung Hee University class ultraperformance liquid chromatography (UPLC) sys- according to an animal protocol (KHUAPS(SE)-16-013). Male tem (Waters Corp., Milford, MA, USA) with photodiode C57BL/6 mice (6–8 weeks old) or male Sprague-Dawley array (PDA) detector and JMS-T100TD (AccuTOF-TLC) (SD) rats (6 weeks old) were purchased from the Orient Bio (JEOL Ltd., Tokyo, Japan) spectrometer equipped with elec- Inc. (Seongnam-si, Korea) and maintained under constant ∘ trospray ionization (ESI) source were used for chromato- conditions (temperature: 20 ± 2 C, humidity: 40–60%, and graphic and spectrometric (MS) analysis. The chromato- light/dark cycle: 12 h). graphic separation was carried out on an ACQUITY UPLC 𝜇 × BEH C18 Column (130 A,˚ 1.7 m, 2.1 mm 50 mm, Waters 2.5. Cell Viability Assay. Cell viability was evaluated by MTT Corp., Milford, MA, USA) equipped with an ACQUITY assay. RAW 264.7 macrophage cells or BMDMs were plated 𝜇 5 UPLC BEH C18 VanGuard Precolumn (130 A,˚ 1.7 m, 2.1 mm at a density of 2 × 10 cells per well in 24-well plates and × 5mm).Themobilephaseconsistedofacetonitrile(solvent then treated with samples (KCT-01, KCT-02, or IJCGT) at A) and 0.1% formic acid water (solvent B). The gradient various concentrations 1 h prior to LPS (1 𝜇g/mL) treatment. condition of the mobile phase was 0–3 min, 5%; 3–5 min, After 24 h of LPS stimulation, 20 𝜇l MTT solution (5 mg/mL) 5% to 10%; 5–15 min, 10% to 15%; 15–20 min, 15% to 50%; was added to each well, and the cells were further incubated 20–30 min, 50% to 100%; and 30–35 min, 100% as percent for an additional 4 h. The supernatant was removed and the ofsolventA.Theflowratewas0.6mL/minandthecolumn ∘ formazanwasresolvedwith1mL/wellofDMSO.Theoptical oven temperature was maintained at 40 C and the detection density was measured at 540 nm by microplate reader. wavelength was 330 nm. The injection volume was 2.0 𝜇L. The conditions of MS analysis in the positive ion mode were 2.6. Nitrite Determination. RAW 264.7 macrophage cells or as follows: scan range, 𝑚/𝑧 50–1000; desolvating chamber 5 ∘ ∘ BMDMswereplatedat2× 10 cells per well in 24-well plates temperature, 250 C; orifice 1 temperature, 80 C; orifice 1 andthenincubatedwithorwithoutLPS(1𝜇g/mL) in the voltage,80V;orifice2voltage,10V;ringlensvoltage,5V; absence or presence of various concentrations of samples peak voltage, 1500 V; detector voltage, 2200 V; and nitrogen (KCT-01, KCT-02, or IJCGT) at various concentrations for gas flow rate, 1.0 L/min (nebulizing gas) and 3.0 L/min (des- 24 h. Nitrite levels in culture media were determined using olvating gas). theGriessreactionassayandpresumedtoreflectNOlevels. For the UPLC-MS analysis, ten milligrams of KCT-01 or The optical density was measured at 540 nm by microplate IJCGT extract was dissolved in one milliliter of 30% ethanol reader. or 70% ethanol, respectively. Reference standard compounds (1.0 mg/mL) were dissolved in methanol and then mixed into 𝛼 a cocktail solution which was used as reference standards 2.7. PGE2, TNF- ,andIL-6Assay. RAW 264.7 macrophage × 5 solution. The sample solutions were filtered through a 0.2 𝜇m cells or BMDMs were plated at 2 10 cells per well in 24-well 𝜇 polyvinylidene fluoride syringe filter (Whatman, Maidstone, plates and then incubated with or without LPS (1 g/mL) in UK) and reference standards solutions were filtered through the absence or presence of various concentrations of samples. a0.2𝜇m polytetrafluoroethylene syringe filter (Whatman, Dilutions of the cell culture medium were assayed for PGE2, 𝛼 𝛽 Maidstone, UK) before injection into UPLC system. TNF- , IL-6, and IL-1 .PGE2 levels in cell culture medium were determined using a colorimetric competitive enzyme- 2.3. Cell Culture and Sample Treatment. The RAW 264.7 linked immunosorbent assay (ELISA) kit (Enzo Life Science, 𝛼 murine macrophage cell line was purchased from the Korea NY, USA) according to manufacturer’s instructions. TNF- Cell Line Bank (Seoul, Korea) and cultured in DMEM and IL-6 levels in cell culture medium were quantified using containing 10% fetal bovine serum, penicillin-streptomycin mouse DuoSet kit (R&D Systems, MN, USA) according to ∘ (100 units/mL) at 37 Cwith5%CO2. manufacturer’s instructions. Murine bone marrow derived macrophages (BMDMs) were isolated from femur of C57BL/6 mice. Cells in bone 2.8. Protein Extraction and Western Blot Analysis. RAW 2 marrow were washed several times with cold phosphate 264.7 macrophage cells were seeded in 60 mm dish and bufferedsaline(PBS).Theisolatedcellswerefilteredthrough incubated for 24 h and then added to KCT-01 1 h prior sieve mesh, centrifuged, and resuspended in DMEM to LPS (1 𝜇g/mL) treatment. The cells were collected by (supplemented with 10% FBS, 100 units/mL penicillin- centrifugation and washed three times with PBS. Washed 6 streptomycin). The cells were incubated at2 × 10 in Petri dish cell pellets were resuspended in protein extraction solution with 15% L929-conditioned medium in DMEM for 7 days PROPREP (Intron Biotechnology, Seoul, Korea) and then ∘ to differentiate into macrophages. The culture medium was incubated for 30 min at 4 C. Cell debris was removed by addedatthe3rddayandreplacedatthe6thdayofincubation. microcentrifugation and supernatants were quickly frozen. After being differentiated, the cells were seeded in 24-well The protein concentration was determined using the Bio- culture plates. In all experiments, cells were incubated with Rad protein assay reagent (Bio-Rad Laboratories Inc., CA, samplesatvariousconcentrationsthatwasalwaysadded1h USA) according to manufacturer’s instruction. Proteins were prior to LPS (1 𝜇g/mL) treatment for the indicated time. electroblotted onto a PVDF membrane following separation on an 8% or 10% SDS-polyacrylamide gel electrophoresis. 2.4. Animals. All experiments in the present study were The immunoblot was incubated for 1 h with blocking solution conducted under university guideline of ethical committee (5% skim milk) at room temperature and then incubated 4 Evidence-Based Complementary and Alternative Medicine

∘ ∘ overnight with a 1 : 1000 dilution of primary antibody at 4 C. The other ear biopsies were immediately frozen (−70 C) Blots were washed three times with Tween 20/Tris-buffered for the determination of myeloperoxidase (MPO) activity, a saline (T/TBS) and then incubated with a 1 : 2000 dilution marker of neutrophil influx into the tissue, and intracellular of horseradish peroxidase-conjugated secondary antibody protein expression including iNOS, COX-2, and STATs. The (Santa Cruz Biotechnology Inc. CA, USA) for 2 h at room tissuewasthawedandhomogenized.Thehomogenatewas temperature. Blots were again washed three times with then centrifuged at 1500 ×gfor15min,andtheresulting T/TBS and then developed by enhanced chemiluminescence supernatant was assayed for MPO assay and Western blotting. (Amersham Life Science, IL, USA). 2.12. Statistical Analysis. Data are presented as mean ± 2.9. Total RNA Extraction and Quantitative Real-Time RT- SD. Comparison between groups was made with SigmaPlot PCR (qRT-PCR). Total cellular RNA was isolated by Easy followed by Student’s 𝑡-test. 𝑝 values of 0.05 or less were Blue kits (Intron Biotechnology, Seoul, Korea). 1 𝜇gofRNA considered statistically significant. was reverse-transcribed (RT) using MuLV reverse transcriptase, 1 mM deoxyribonucleotide triphosphate (dNTP), and 𝜇 𝜇 3. Results 0.5 g/ l oligo (dT12–18). Real-time PCR was performed using Thermal Cycler Dice Real-Time PCR System (Takara, 3.1. Phytochemical Identification of KCT-01 and IJCGT by Shiga, Japan). The primers used for SYBR green real-time UPLC-PDA-ESI-MS. To identify phytochemicals of KCT- reverse transcription-PCR were as follows: for iNOS,sense 󸀠 󸀠 01 and IJCGT, chromatographic fingerprint analysis was primer, 5 -CAT GCT ACT GGA GGT GGG TG-3 ,antisense 󸀠 󸀠 carried out of KCT-01 or IJCGT by UPLC-PDA-MS and primer, 5 -CAT TGA TCT CCG TGA CAG CCC-3 ;for 󸀠 MS/MS. Chromatogram of KCT-01 detected at 330 nm is COX-2, sense primer, 5 -TGC TGT ACA AGC AGT GGC 󸀠 󸀠 shown in Figure 1(a). Ten peaks (chlorogenic acid, hypero- AA-3 , antisense primer, 5 -GCA GCC ATT TCC TTC TCT 󸀠 󸀠 side, scoparone, isochlorogenic acid A, isochlorogenic acid CC-3 ;forTNF-𝛼, sense primer, 5 -AGC ACA GAA AGC 󸀠 󸀠 B, luteolin, jaceosidin, eupatilin, ziyuglycoside I, and cur- ATG ATC CG-3 , antisense primer, 5 -CTG ATG AGA GGG 󸀠 󸀠 cumin) were confirmed by direct comparison with their AGG CCA TT-3 ;forIL-6, sense primer, 5 -GAG GAT ACC 󸀠 󸀠 corresponding reference standards (Figure 1, Supplementary ACT CCC AAC AGA CC-3 , antisense primer, 5 -AAG TGC 󸀠 Figure 1). Eleven peaks were identified in chromatogram of ATC ATC GTT GTT CAT ACA-3 ,for𝛽-actin, sense primer, 󸀠 󸀠 IJCGT shown in Figure 2(a) and Supplementary Figure 2(A). 5 -ATC ACT ATT GGC AAC GAG CG-3 , antisense primer, 󸀠 󸀠 Qualitative identification using standard solutions confirmed 5 -TCA GCA ATG CCT GGG TAC AT-3 .Theresultswere the presence of chlorogenic acid, hyperoside, scoparone, 𝛽 expressed as the ratio of optimal density to -actin. isochlorogenic acid A, isochlorogenic acid B, luteolin, jaceosidin, eupatilin, ellagic acid, ziyuglycoside I, and glycyrrhizic 2.10. Plasmid, Transient Transfection, and Luciferase Assay. acid (Figures 1(b) and 2(b), Supplementary Figures 1(B) and RAW 264.7 macrophages were cotransfected with pNF-𝜅B- 2(B)). Chlorogenic acid, hyperoside, scoparone, isochloro- luc or pAP-1-luc (Clontech, Shiga, Japan) plasmid plus the genic acid A, isochlorogenic acid B, luteolin, jaceosidin, phRL-TK plasmid (Promega, WI, USA) using Lipofectamine and eupatilin have been previously identified in Artemisia LTX6 (Invitrogen, CA, USA) as instructed by the manufac- Capillaris [22]. Ziyuglycoside I in Sanguisorba officinalis turers. After 24 h of transfections, cells were pretreated with [23], curcumin in Curcuma longa [24], ellagic acid in Rubus KCT-01for1hpriortoLPS(1𝜇g/mL) stimulation. After 18 h, coreanus [25], hesperidin in Citrus unshiu [26], ziyuglycoside each well was washed with cold-PBS and cells were lysed IinSanguisorba officinalis [23], and glycyrrhizic acid in and the luciferase activity was determined using the Promega Glycyrrhiza uralensis [27] have been previously identified. luciferase assay system (WI, USA). Detailed UPLC-MS data of these peaks are listed in Tables 2 and 3. Based on UPLC-PDA-MS analysis, representative phy- 2.11. Ear Edema Animal Model. Sprague-Dawley (SD) male tochemicals from each herbal formulation were successfully rats weighing 180–200 g were divided into three groups (𝑛 = identified. 6); croton-oil-alone group, KCT-01-50 mg/kg-treated group, andKCT-01-100mg/kg-treatedgroup.SDmaleratswere 3.2. Effect of KCT-01, KCT-02, or IJCGT on NO and PGE2 pretreated with KCT-01 (50 or 100 mg/kg, p.o.) and after 1 h, Production, and Cell Viability in LPS-Induced RAW 264.7 ear edema was induced on inner surface of the right ear by Macrophages. To evaluate inhibitory properties of KCT-01, topical application of croton oil (5% solution in 100 𝜇Lof KCT-02, or IJCGT on production of NO and PGE2,RAW acetone). The left ear was used as a control and received the 264.7 macrophages were pretreated with various concen- same amount of the vehicle (acetone). Two hours after croton trations (25, 50, or 100 𝜇g/mL) of KCT-01, KCT-02, or oil application, rats were sacrificed using CO2,andboth IJCGTfor24hinpresenceofLPS(1𝜇g/mL). As shown in ear tissues were collected using 6 mm punching. Ear punch Figure 3(a) and Table 4, KCT-01, KCT-02, and IJCGT each biopsies were immediately measured for thickness to assess suppressed NO production in a concentration-dependent ear edema. Ear biopsies were fixed in 4% paraformaldehyde manner. At concentration of 100 𝜇g/mL, KCT-01, KCT-02, overnight and embedded in paraffin. Ear biopsy sections and IJCGT suppressed NO production by 70.71%, 32.31%, were stained with hematoxylin and eosin (H&E) at the and 42.02%, respectively (IC50 of KCT-01, KCT-02, and Seoul Medical Science Institute (SCL Co. Ltd., Seoul, Korea). IJCGT: 64.93 𝜇g/mL, >100 𝜇g/mL, and >100 𝜇g/mL, resp.). Evidence-Based Complementary and Alternative Medicine 5

330 nm 0.16

(1) (4, 5)

0.08 (3)

(AU) (2)

(7) (6) (8) (10) 0.00 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Time (min) (a)

(3) (4, 5) (7) (10) 0.16 (1) (6) (2) (8) 0.08 (AU)

0.00

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Time (min) (b)

Figure 1: UPLC chromatograms of KCT-01 extract (a) and reference standards (b). (1) Chlorogenic acid; (2) hyperoside; (3) scoparone; (4) isochlorogenic acid A; (5) isochlorogenic acid B; (6) luteolin; (7) jaceosidin; (8) eupatilin; (9) ziyuglycoside I; and (10) curcumin.

330 nm 0.16 (5, 6)

0.08 (1) (AU) (3) (2) (4) (7) (9) (10) (12) (8) 0.00

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Time (min) (a)

(5, 6) (9) (13) 0.16 (4) (1) (8)

0.08 (3) (10) (AU) (2) (7) (12) 0.00

0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 Time (min) (b)

Figure 2: UPLC chromatograms of IJCGT extract (a) and reference standards (b). (1) Chlorogenic acid; (2) ellagic acid; (3) hyperoside; (4) scoparone; (5) isochlorogenic acid A; (6) isochlorogenic acid B; (7) hesperidin; (8) luteolin; (9) jaceosidin; (10) eupatilin; (11) ziyuglycoside I; (12) glycyrrhizic acid; (13) curcumin. 6 Evidence-Based Complementary and Alternative Medicine

Table 2: Retention times (Rt), precursor ions, molecule weights, and UV Maxima (𝜆 max) of identified peaks of KCT-01.

Compound Rt (min) Precursor ion (𝑚/𝑧) Molecule weight 𝜆 max (nm) + 355.10 [M+H] (1) Chlorogenic acid 1.78 + 354.10 218, 242, 325 377.08 [M+Na] + (2) Hyperoside 7.19 465.10 [M+H] 464.10 203,255,353 + (3) Scoparone 7. 9 2 2 0 7. 0 6 [M+H] 206.06 203,229,343 + (4) Isochlorogenic acid A 9.90 517.13 [M+H] 516.13 218, 244, 327 + (5) Isochlorogenic acid B 9.90 517.14 [M+H] 516.13 218, 244, 327 + (6) Luteolin 14.86 287.06 [M+H] 286.05 253, 349 + (7) Jaceosidin 18.68 331.08 [M+H] 330.07 214, 272, 345 + (8) Eupatilin 20.33 345.10 [M+H] 344.09 214, 273, 343 + 767.46 [M+H] (9) Ziyuglycoside I 20.61 + 766.45 784.48 [M+NH4] + (10) Curcumin 22.68 369.13 [M+H] 368.12 197, 264, 428

Table 3: Retention time (Rt), precursor ion, molecule weight, and UV Maxima (𝜆 max) of identified peaks of IJCGT.

Compound Rt (min) Precursor ion (𝑚/𝑧) Molecule weight 𝜆 max (nm) + 355.10003 [M+H] (1) Chlorogenic acid 1.78 + 354.10 218, 242, 325 377.08276 [M+Na] + (2) Ellagic acid 6.53 303.02 [M+H] 302.01 196, 253, 366 + (3) Hyperoside 7.19 465.11 [M+H] 464.10 203, 255, 353 + (4) Scoparone 7. 9 2 2 0 7. 0 6 [M+H] 206.06 203,229,343 + (5) Isochlorogenic acid A 9.90 517.14 [M+H] 516.13 218, 244, 327 + (6) Isochlorogenic acid B 9.90 517.13 [M+H] 516.13 218, 244, 327 + (7) Hesperidin 12.28 611.20 [M+H] 610.19 199,283,329 + (8) Luteolin 14.86 287.06 [M+H] 286.05 253, 349 + (9) Jaceosidin 18.68 331.08 [M+H] 330.07 214, 272, 345 + (10) Eupatilin 20.33 345.10 [M+H] 344.09 214, 273, 343 + 767.46 [M+H] (11) Ziyuglycoside I 20.61 + 766.45 784.49 [M+NH4] + (12) Glycyrrhizic acid 21.49 823.41 [M+H] 822.40 252

Table 4: Effect of KCT-01, KCT-02, or IJCGT on NO, PGE 2,TNF-𝛼, and IL-6 production in LPS-induced RAW264.7 macrophages.

a IC50 (𝜇g/ml) Cell viability NO PGE2 TNF-𝛼 IL-6 KCT-01 >100 64.93 ± 0.77 18.18 ± 2.73 22.88 ± 0.96 35.47 ± 3.77 KCT-02 >100 >100 51.75 ± 3.02 31.78 ± 4.59 57.93 ± 3.31 IJCGT >100 >100 27.70 ± 1.94 48.14 ± 6.71 >100 a Data are presented as means ± SD of three independent experiments.

At concentration of 100 𝜇g/mL, KCT-01, KCT-02, and IJCGT effects on NO and PGE2 production were not attributable to also suppressed PGE2 production by 97.67%, 84.47%, and their nonspecific cytotoxicity. 93.22%, respectively (IC50:18.18𝜇g/mL, 51.75 𝜇g/mL, and 27.70 𝜇g/mL, resp.) (Figure 3(b)). L-NIL (40 𝜇M) and NS398 (10 nM) were used as NO and PGE2 inhibitors, respectively. 3.3. Effect of KCT-01, KCT-02, or IJCGT on Production of TNF- 𝛼 𝛼 AsshowninTable4,KCT-01moreeffectivelyblocked andIL-6inLPS-InducedRAW264.7Macrophages. TNF- production of NO and PGE2 than KCT-02 or IJCGT. The and IL-6 are important inflammatory cytokines secreted by potentially cytotoxic effect of KCT-01, KCT-02, or IJCGT macrophages [28]. Therefore, we examined effects of KCT- on RAW 264.7 macrophages was determined by MTT assay. 01, KCT-02, or IJCGT on production of proinflammatory Resultsshowedthatviabilityofcellswasnotsignificantly cytokines (TNF-𝛼 and IL-6) in LPS-induced RAW 264.7 affected by KCT-01, KCT-02, or IJCGT at concentrations up macrophages. Cells were pretreated with KCT-01, KCT-02, to 100 𝜇g/mL (Figure 3(c)), indicating that their suppressive or IJCGT at various concentrations (25, 50, or 100 𝜇g/mL) Evidence-Based Complementary and Alternative Medicine 7

120 120 # # # ## # 100 ∗∗ ∗∗∗ 100 ∗∗∗ 80 80 ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ 60 ∗∗∗ 60 ∗∗∗

production (%) production ∗∗∗ 40 2 40 ∗∗∗ ∗∗∗ ∗∗∗∗∗∗∗∗∗ NO production (%) production NO ０＇％ 20 20 ∗∗∗ ∗∗∗ ∗∗∗ ∗∗∗ 0 0  LPS (1 g/ml) − + + + + − + LPS (1 g/ml) − + + + + − +  − − − − − − ，-NIL (40 M) + NS-398 (10 nM) − − − − − − +  − − − Extract ( g/ml) 25 50 100 100 Extract (g/ml) − − 25 50 100 100 − KCT-01 KCT-02 KCT-01 IJCGT KCT-02 IJCGT (a) (b) 120

Cell viability (%) Cell viability 30

0 LPS (1g/ml) + + + + Extract (g/ml) − 25 50 100

KCT-01 KCT-02 IJCGT (c)

Figure 3: KCT-01, KCT-02, or IJCGT inhibited production of NO and PGE2 in LPS-induced RAW264.7 macrophages at noncytotoxic concentrations. Cells were treated with KCT-01, KCT-02, or IJCGT at various concentrations (6.25–100 𝜇g/mL) plus LPS (1 𝜇g/mL) or LPS alone for 24 h. (a) Cell viability was measured by MTT assay and NO production was measured using Griess reaction assay. (b) PGE2 # production was measured using an EIA kit. Values are expressed as means ± SD of three independent experiments. 𝑝 < 0.05 versus control, ∗∗ ∗∗∗ 𝑝 < 0.01 and 𝑝 < 0.001 versus LPS-treated cells.

for 1 h followed by stimulation with LPS for 24 h. KCT- primary macrophage cells, we examined the inhibitory effect 01, KCT-02, or IJCGT each concentration dependently and of KCT-01 on production of proinflammatory mediators and significantly suppressed the production of TNF-𝛼 and IL-6 cytokines in LPS-induced BMDMs. We found that KCT- (Figure 4). Of these three, KCT-01 showed the most potent 01 significantly and concentration dependently inhibited inhibition for TNF-𝛼 (IC50:22.9𝜇g/mL) and IL-6 (IC50: production of NO, PGE2,TNF-𝛼, and IL-6 in LPS-stimulated 35.5 𝜇g/mL) production. Collectively, our results indicate that BMDMs (Figure 5), indicating that the anti-inflammatory KCT-01 is a more potent herbal medicine that can suppress effect of KCT-01 might not be cell specific responses. proinflammatory mediators compared to KCT-02 or IJCGT (Table 4). Therefore, we selected KCT-01 as the most effective 3.5. KCT-01 Inhibits Expression Levels of iNOS, COX-2, TNF- anti-inflammatory agent and conducted further experiments 𝛼, and IL-6 in LPS-Induced RAW 264.7 Macrophages. Next, to determine the underlying molecular mechanism involved we evaluated whether the inhibitory effect of KCT-01 on the in its anti-inflammatory effect. production of NO and PGE2 was related to reduced expression of iNOS and COX-2 by Western blotting and qRT-PCR. 3.4. KCT-01 Inhibits Production of NO, PGE2, TNF-𝛼,and As shown in Figure 6(a), iNOS and COX-2 protein levels were IL-6 in LPS-Induced BMDMs. To confirm our findings in markedly increased by LPS. However, KCT-01 significantly 8 Evidence-Based Complementary and Alternative Medicine

6 50

### 5 ### 40 ∗ 4 ∗∗ ∗ ∗ 30 ∗∗∗

3 ∗∗∗ ∗∗ 20

2 ∗∗∗ ∗∗ ∗∗ ∗∗∗ ∗∗∗ 10 1 (ng/ml) production IL-6 ∗∗∗ production (ng/ml)  production TNF-

∗∗∗ 0 0 LPS (1 g/ml) − + + + + − LPS (1 g/ml) − + + + + − Extract (g/ml) − − 25 50 100 100 Extract (g/ml) − − 25 50 100 100 KCT-01 KCT-01 KCT-02 KCT-02 IJCGT IJCGT (a) (b)

Figure 4: Effects of KCT-01, KCT-02, or IJCGT on proinflammatory cytokines production in LPS-induced RAW264.7 macrophages. Cells were treated with KCT-01, KCT-02, or IJCGT at various concentrations (12.5–100 𝜇g/mL) plus LPS (1 𝜇g/mL) or LPS alone for 24 h. (a) Productions of TNF-𝛼 andIL-6weremeasuredusingEIAkits.Valuesareexpressedasmeans± SD of three independent experiments. # ∗ ∗∗ ∗∗∗ 𝑝 < 0.05 versus control; 𝑝 < 0.05, 𝑝 < 0.01,and 𝑝 < 0.001 versus LPS-treated cells.

16 150 60 # 120 12 45 # 90 ∗∗∗ ∗∗∗ ∗∗∗ 8 30 60 production (ng/ml) production 2 4 (%) Cell viability 15 NO production (  M) production NO ∗∗∗ ∗∗∗ 30 ∗∗∗ ０＇％ ∗∗∗ 0 0 0 LPS (1 g/ml) − + + + + − + LPS (1 g/ml) − + + + + − + ，-NIL (10 M) − − − − − − + NS-398 (10 nM) − − − − − − + KCT-01 (g/ml) − − 25 50 100 100 − KCT-01 (g/ml) − − 25 50 100 100 − (a) (b) 1 16 # # 0.8 ∗∗ 12

0.6 8 ∗∗∗ ∗∗∗ 0.4 ∗∗∗ ∗∗∗ 4 0.2

∗∗∗ (ng/ml) production IL-6 production (ng/ml)  production TNF- 0 0 LPS (1 g/ml) − + + + + − LPS (1 g/ml) − + + + + − KCT-01 (g/ml) − − 25 50 100 100 KCT-01 (g/ml) − − 25 50 100 100 (c) (d)

Figure 5: KCT-01 inhibited NO, PGE2,TNF-𝛼, and IL-6 productions in LPS-induced mouse bone marrow derived macrophages. Cells were treated with KCT-01, KCT-02, or IJCGT at various concentrations (25–100 𝜇g/mL) plus LPS (1 𝜇g/mL) or LPS alone for 24 h. (a) Cell viability was measured by MTT assay and NO production was measured using Griess reaction. ((b), (c), and (d)) PGE2,TNF-𝛼, and IL-6 productions # ∗∗ were measured by EIA kit. Values are expressed as means ± SD of three independent experiments. 𝑝 < 0.05 versus control; 𝑝 < 0.01 and ∗∗∗ 𝑝 < 0.001 versus LPS-treated cells. Evidence-Based Complementary and Alternative Medicine 9

LPS (1 g/ml, 24 h) − + + + + − KCT-01 (g/ml) − − 25 50 100 100

iNOS

COX-2 1.2 # # 1 -Actin ∗∗ 0.8 1.2 # # 0.6 0.9 ∗ ∗∗∗ ∗ 0.6 0.4 ∗ ∗∗∗ ∗ 0.3 0.2 ∗∗∗ Relative mRNA expression mRNA Relative ∗∗∗ Ratio of density of Ratio ∗ ∗∗∗ 0 0 LPS (1 g/ml, 4 h) − + + + + − iNOS/actin KCT-01 (g/ml) − − 25 50 100 100 COX-2/actin iNOS COX-2 (a) (b) 1.2 1.2 # # 1 1 ∗ 0.8 0.8 ∗∗∗ 0.6 0.6 ∗∗∗ mRNA expression  mRNA 0.4 0.4 ∗∗∗ 0.2 ∗∗∗ 0.2

Relative IL-6 mRNA expression mRNA IL-6 Relative ∗∗∗ Relative TNF- Relative 0 0 LPS (1 g/ml, 4 h) − + + + + − LPS (1 g/ml, 4 h) − + + + + − KCT-01 (g/ml) − − 25 50 100 100 KCT-01 (g/ml) − − 25 50 100 100 (c) (d) Figure 6: KCT-01 inhibits the expression of inflammatory genes in LPS-induced RAW264.7 macrophages. Cells were treated with KCT-01 plus LPS (1 𝜇g/mL) or LPS alone for indicated time. (a) Total cellular proteins were prepared and subjected to Western blotting to determine protein expression levels of iNOS and COX-2. Levels of iNOS and COX-2 were normalized against 𝛽-actin expression. ((b), (c), and (d)) Total cellular RNA were prepared and subjected to qRT-PCR to determine mRNA expression levels of iNOS, COX-2, TNF-𝛼, and IL-6. Levels of iNOS, COX-2, TNF-𝛼, and IL-6 were normalized against 𝛽-actin expression. Values are expressed as means ± SD of three independent # ∗ ∗∗ ∗∗∗ experiments. 𝑝 < 0.05 versus control; 𝑝 < 0.05, 𝑝 < 0.01,and 𝑝 < 0.001 versus LPS-treated cells. suppressed these upregulations in a concentration-dependent NF-𝜅BorAP-1activityinRAW264.7macrophagesby manner. KCT-01 also markedly inhibited LPS-induced iNOS luciferase reporter gene assay using pNF-𝜅B-luc or pAP-1-luc and COX-2 mRNA expression levels (Figure 6(b)). These vector.OurresultsrevealedthatKCT-01didnothaveany findings demonstrate that KCT-01 can downregulate the effect on NF-𝜅B or AP-1-dependent transcriptional activity expression of LPS-induced iNOS and COX-2 and lead to (Figure 7). When transcription factors are phosphorylated, decreased production of NO and PGE2.KCT-01alsodown- these may translocate to the nucleus where they can bind to regulated TNF-𝛼 and IL-6 mRNA expression levels in a their consensus DNA binding sites to regulate transcription concentration-dependent manner (Figures 6(c) and 6(d)). of target genes [31]. Accordingly, we performed Western blot- Taken together, these results suggest that KCT-01 possesses ting to determine the effect of KCT-01 on phosphorylation anti-inflammatory activity by inhibiting expression levels of of p65 (a subunit of NF-𝜅B) or c-fos and c-jun (subunits various LPS-induced proinflammatory mediators. of AP-1). Our results showed that KCT-01 did not have any inhibitory effect on LPS-induced phosphorylation of p65, c- 3.6. Effects of KCT-01 on NF-𝜅B and AP-1 Activation in LPS- fos, and c-jun (data not shown). Induced RAW 264.7 Macrophages. NF-𝜅BandAP-1arekey transcriptional factors regulating inflammatory responses 3.7.EffectofKCT-01onActivationofJAK/STATSignal- mediated by LPS or proinflammatory cytokines [29, 30]. ing Pathway in LPS-Induced RAW 264.7 Macrophages. Therefore, we explored the effect of KCT-01 on LPS-induced JAK/STAT signaling pathway is involved in immunity and has 10 Evidence-Based Complementary and Alternative Medicine

40 50 # # 32 40

24 30

16 20

8 10 AP-1 luciferase activity (RLU) (RLU) activity luciferase AP-1 B luciferase activity (RLU) activity  B luciferase NF- 0 0 LPS (1 g/ml) − + + + + − LPS (1 g/ml) − + + + + − KCT-01 (g/ml) − − 25 50 100 100 KCT-01 (g/ml) − − 25 50 100 100 (a) (b)

Figure 7: KCT-01 has no inhibitory effect on activation of NF-𝜅B or AP-1 in LPS-induced RAW264.7 macrophages. Cells were transfected with p-NF-𝜅B-luc reporter or pAP-1-luc reporter. phRL-TK vector was used as an internal control. Cells were then treated with KCT-01 plus LPS or LPS alone. After 18 h of treatment, luciferase activity levels were determined using luciferase assay. Values are expressed as the means # ± SD of three independent experiments. 𝑝 < 0.05 versus control.

LPS (1g/ml, 2 h) − + + + + − KCT-01 (g/ml) − − 25 50 100 100

p-STAT1 (Ser 727) LPS (1g/ml, 60 min) − + + + + − KCT-01 (g/ml) − − 25 50 100 100 p-STAT1 (Tyr 701) p-JAK1

STAT1 JAK1

p-STAT3 (Tyr 705) p-JAK2

STAT3 JAK2

 -Actin -Actin 1 2.5 # # # 0.8 2 ∗ ∗ ∗ # ∗ 0.6 # 1.5 ∗ ∗ ∗ ∗ ∗ ∗ 0.4 ∗ ∗ 1 ∗ ∗

Ratio of density ∗ 0.2 density Ratio of 0.5

0 0

p-STAT1 (Ser727)/STAT1 p-JAK1 (Tyr1022/1023)/JAK1 p-STAT1 (Tyr701)/STAT1 p-JAK2 (Tyr1007/1008)/JAK2 p-STAT3 (Tyr705)/STAT3 (a) (b)

Figure 8: KCT-01 suppresses activation of JAK/STAT signaling cascades in LPS-induced RAW264.7 macrophages. Cells were treated with KCT-01 plus LPS (1 𝜇g/mL) or LPS alone for indicated time. (a) Total cellular proteins were prepared and subjected to Western blotting to determine phosphorylation levels of STAT1 (Ser727 or Tyr 701), STAT3 (Tyr705), and JAK1/2. Levels of STATs and JAKs were normalized # ∗ against 𝛽-actin expression. Values are expressed as means ± SD of three independent experiments. 𝑝 < 0.05 versus control; 𝑝 < 0.05 versus LPS-treated cells. also affected inflammatory signaling cascades triggered by pathway. Binding of LPS ligands to their receptors induces LPS, IFN-𝛾, and other cytokines [7, 32]. Since KCT-01 had no phosphorylation of receptor-associated JAKs followed by effect on NF-𝜅B and AP-1 activation, we investigated whether phosphorylation of STATs [9]. As shown in Figure 8(a), KCT- the inhibitory effect of KCT-01 on the expression of proin- 01 concentration dependently downregulated the phospho- flammatory mediators was mediated by JAK/STAT signaling rylation of STAT1 (Ser727 and Tyr701) and STAT3 (Tyr705) Evidence-Based Complementary and Alternative Medicine 11

140 900 # 800 120 700 100 ∗∗∗ 600 80 500 ∗∗∗ ∗∗∗ 60 400 300 Ear edema (%) 40 ∗∗∗ 200

20 thickness (  M) Epidermal 100 0 0 Croton oil + + + Croton oil − + + + KCT-01 (mg/kg) − 50 100 KCT-01 (mg/kg) − − 50 100 (a) (b)

KCT-01 KCT-01 CON Croton oil 50 mg/kg 100 mg/kg (c) Croton oil − − + + + + + + KCT-01 (g/ml) − − − − 50 50 100 100 160 140 # iNOS 120 COX-2 100 ∗ 80 -Actin ∗∗ 60

MPO activity (%) MPO activity 40 p-STAT3 (Tyr 705) 20 STAT3 0 Croton oil − + + + -Actin KCT-01 (mg/kg) − − 50 100 (d) (e)

Figure 9: KCT-01 inhibits ear inflammation induced by croton oil in rats. (a) KCT-01 (50 or 100 mg/kg, p.o.) was administered 1hpriorto croton oil treatment and 6 mm ear punch was measured for thickness at 3 h after croton oil application to assess ear edema. ((b) and (c)) Epidermal thickness of ear tissues and histological analysis of croton oil-induced ear edema. Representative H&E sections of ear tissues were obtained from vehicle-treated control rat’s left ear (control), croton oil treated rat’s right ear, and KCT-01 (50 or 100 mg/kg, p.o.) plus croton oil treated rat’s right ear. ((d) and (e)) Ear tissues were homogenized and lysates were used to determine MPO activity, protein expression, # ∗ ∗∗ and phosphorylation of iNOS, COX-2, and STAT3. Values are expressed as means ± SD. 𝑝 < 0.05 versus control; 𝑝 < 0.05, 𝑝 < 0.01,and ∗∗∗ 𝑝 < 0.001 versus croton oil treated group. at 2 h after LPS stimulation. KCT-01 also significantly and we used a rat model of acute inflammatory ear edema induced concentration dependently blocked the phosphorylation of by croton oil. As expected, ear thickness increased by croton JAK1 (Tyr1022/1023) and JAK2 (Tyr1007/1008) at 1 h after LPS oil application was reduced by pretreatment with KCT-01. challenge (Figure 8(b)). Therefore, KCT-01 potently inhibited At 50 and 100 mg/kg, KCT-01 decreased ear thickness by JAK1 and JAK2, subsequently downregulating phosphoryla- 45.00±6.08%and70.45±8.86%, respectively (Figure 9(a)). To tion of STAT1 and STAT3. examine histopathological changes during ear edema, cross sections of ear discs were stained with hematoxylin and eosin. 3.8.EffectofKCT-01onCrotonOil-InducedEarEdemainRats. Consistent with its inhibitory effect on ear thickness, KCT- To determine the anti-inflammatory effect of KCT-01 in vivo, 01 also inhibited epidermal ear edema after pretreatment 12 Evidence-Based Complementary and Alternative Medicine

(Figures 9(b) and 9(c)). Next, neutrophil migration into without causing cytotoxicity. These results indicate that KCT- croton oil-induced ear was indirectly determined using MPO 01, KCT-02, and IJCGT exhibit anti-inflammatory effects on activity assay. As expected, application of croton oil to LPS-induced RAW 264.7 macrophages. We selected KCT-01 ear increased MPO production in tissues. However, oral asthemostpotentprescription/formulationamongthethree administration of KCT-01 suppressed croton oil-induced and elicited molecular mechanisms involved. The inhibitory MPO production (Figure 9(d)). Consistent with our findings effectofKCT-01onthereleaseofNO,PGE2,TNF-𝛼, in macrophages, KCT-01 also inhibited iNOS and COX-2 and IL-6 was also proved in murine bone marrow derived expression levels and STAT3 phosphorylation (Figure 9(e)). macrophages. NO and PGE2 are major proinflammatory mediators produced by enzymes iNOS and COX-2, respec- 4. Discussion tively. These two enzymes are responsible for cell damage and tissue destruction in inflammation [45, 46]. Consistently, Many studies have reported that traditional herbal resources we observed that KCT-01 also inhibited mRNA expression can benefit the management of various diseases, including of iNOS, COX-2, TNF-𝛼,andIL-6,indicatingthatKCT- arthritis [33], atopic dermatitis [34], and hepatic fibrosis [35]. 01 downregulated the expression of inflammatory genes at According to classic literature such as well-known herbal transcription level. medical books “Sang han lon” and “Geum gwe yo lyag,” To further explore the intracellular mechanism under- a variety of herbal medications have been used to treat lying the anti-inflammatory effect of KCT-01, we focused many diseases in Asian countries. In particular, “Geum gwe on various transcription factors such as NF-𝜅B, AP-1, and yo lyag” recommended IJORS (injinoryung-san), a herbal STATs involved in the regulation of inflammatory genes. decoction consisting of Artemisia Capillaris Herba, Alismatis Mechanistically, LPS induces TLR4 to activate NF-𝜅B, an Rhizoma, Poria Sclerotium, Atractylodis Macrocephalae Rhi- important transcription factor for iNOS and COX-2 expres- zoma, Polyporus Sclerotium,andCinnamomi Cortex as a typ- sion in various cells including macrophages [47]. LPS can ical prescription to treat jaundice. Based on IJORS expelling also activate AP-1 which is stimulated by mitogen-activated jaundice, IJCGT is made by omitting Cinnamomi Cortex but protein kinases (MAPKs), including ERK1/2, p38 MAPKs, adding Sanguisorbae Radix, Rubi Fructus, Glycyrrhizae Radix, and JNK, thus enhancing proinflammatory gene expression Raphani Semen,andCitrus Unshiu Immaturi Pericarpium, in macrophages [48–50]. However, KCT-01 did not influ- and Zingiberis Rhizoma Crudus (Table 1). Currently, IJCGT ence NF-𝜅B or AP-1 activation. Besides NF-𝜅BandAP-1 is mainly applied to treat patients with various liver diseases, signaling, some studies have reported that the JAK/STATs such as viral hepatitis, cirrhosis, and hepatocellular carci- signal pathway is crucial for the expression of genes encoding noma in Kyung Hee University Korean Medicine Hospital inflammatoryenzymessuchasiNOSandCOX-2[5,8,9]. (Seoul, Korea). Among 11 individual medicines, Artemisia The JAK-STATs pathway is also critical for cytokine activated Capillaris Herba and Sanguisorbae Radix are regarded as signaling in immune response [51–53]. Thus, it is reasonable keycomponentsofIJCGTforitsvariouspharmacological to speculate that the inhibitory effect of KCT-01 on LPS- effects for hepatic diseases, including liver cirrhosis and stimulated inflammatory response might be attributed to its hepatoma usually caused by chronic inflammation followed suppression on JAK-STATs signaling pathway. Accordingly, by pathological changes in liver function [17, 36–40]. In we found that KCT-01 blocked phosphorylation of both addition, increasing evidence has revealed that curcumin, an JAK and STATs in a concentration-dependent manner which active ingredient of Curcuma longa Radix, possesses anti- in turn inhibited translocation of STATs to the nucleus to inflammatory properties [41, 42]. Thus, we developed acom- bind to promoters of target genes for their transcription pressed herbal prescription consisting of Artemisia Capillaris activation [6]. These results demonstrated that KCT-01 could Herba, Sanguisorbae Radix,andCurcuma longa Radix to treat restrain LPS-elevated release of proinflammatory mediators hepatic inflammation. It was named KCT-01. Additionally, via blocking STAT1 and STAT3 activation. KCT-02 was expanded from KCT-01 by adding Rubi Fruc- To verify whether KCT-01 could ameliorate acute inflam- tus and Salviae Miltiorrhizae Radix to strengthen its anti- matory symptoms in vivo,weusedcrotonoil-inducedear inflammatory activities [20, 43, 44]. In the present study, we edema rat model. As expected, we found that oral admin- evaluated anti-inflammatory effects of KCT-01, KCT-02, and istration of KCT-01 significantly suppressed the swelling of IJCGT and elucidated underlying molecular mechanisms of ears and MPO activity in ear tissues, indicating that KCT- KCT-01 in LPS-induced RAW 264.7 macrophages. Based on 01couldinhibitacuteinflammationviainhibitinginfiltra- their inhibitory potencies on LPS-induced NO, PGE2,TNF- tion of inflammatory cells. Consistent with our findings 𝛼,andIL-6production,weselectedKCT-01tofurtherinves- in macrophages, KCT-01 inhibited protein expression of tigate its anti-inflammatory effects and underlying molecular inflammatory mediators (including iNOS and COX-2) and mechanism in vitro (LPS-induced macrophages) and in vivo phosphorylation of STAT3 in ear tissues. (croton oil-induced acute inflammation in ear edema of rat). Herbal medicine has limitations in practical uses as a LPS, an inflammatory stimulator, can induce various drug because it is difficult to standardize herbal medicine or proinflammatory mediators and cytokines. Our data revealed find obvious active components of each formulation. Thus, that proinflammatory mediators (NO, PGE2) and cytokines most herbal medicines are underestimated although they are (TNF-𝛼, IL-6) were induced by LPS stimulation in RAW effective in treating various diseases. As part of our effort to 264.7 macrophages. However, KCT-01, KCT-02, and IJCGT standardize herbal medicine, we analyzed phytochemicals attenuated the expression of NO, PGE2,TNF-𝛼,andIL-6 of KCT-01 and IJCGT using chromatographic fingerprint Evidence-Based Complementary and Alternative Medicine 13 analysis. However, we need to investigate active components mediated NF-𝜅BandMAPKssignalingpathways,”Journal of of KCT-01. If we can scientifically verify and prove Ethnopharmacology,vol.145,no.1,pp.193–199,2013. therapeutic effects of this herbal medicine, the value of [4] Y.YamamotoandR.B.Gaynor,“IkappaBkinases:keyregulators KCT-01 will be increased. of the NF-kappaB pathway,” Trends in Biochemical Sciences,vol. 29, no. 2, pp. 72–79, 2004. 5. Conclusions [5] E. J. Park, S. Y. Park, E.-H. Joe, and I. Jou, “15d-PGJ2 and rosiglitazone suppress Janus kinase-STAT inflammatory signal- In conclusion, we found that KCT-01, KCT-02, and IJCGT ing through induction of suppressor of cytokine signaling 1 suppressed inflammatory mediators, with KCT-01 being the (SOCS1) and SOCS3 in glia,” TheJournalofBiologicalChemistry, most effective one among the three. We also proved that the vol. 278, no. 17, pp. 14747–14752, 2003. inhibitoryeffectofKCT-01oninflammatoryproteinsand [6]D.E.LevyandJ.E.DarnellJr.,“STATs:transcriptionalcontrol genes was accompanied by suppression of phosphorylation of and biological impact,” Nature Reviews Molecular Cell Biology, JAK1/2 and STAT1/3. Therefore, this study provides evidence vol. 3, no. 9, pp. 651–662, 2002. that KCT-01 might exhibit anti-inflammatory effect via sup- [7]P.J.Murray,“TheJAK-STATsignalingpathway:Inputand pressing JAK/STATs activation in LPS-induced RAW 264.7 output integration,” The Journal of Immunology,vol.178,no.5, macrophages. Such mechanism of action also contributed to pp. 2623–2629, 2007. the pharmacological potential of KCT-01 in vivo using croton [8] S.Okugawa,Y.Ota,T.Kitazawaetal.,“Januskinase2isinvolved oil-induced ear edema model. Our findings suggest that KCT- in lipopolysaccharide-induced activation of macrophages,” 01 may have potential as a herbal medicine for treating a American Journal of Physiology-Cell Physiology,vol.285,no.2, variety of inflammatory diseases. pp. C399–C408, 2003. [9] Z. Qi, F. Yin, L. Lu et al., “Baicalein reduces lipopolysaccharide- Conflicts of Interest induced inflammation via suppressing JAK/STATs activation and ROS production,” Inflammation Research,vol.62,no.9,pp. The authors state that they have no conflicts of interest to 845–855, 2013. declare. [10] P. Kovarik, M. Mangold, K. Ramsauer et al., “Specificity of signaling by STAT1 depends on SH2 and C-terminal domains that regulate Ser727 phosphorylation, differentially affecting Acknowledgments specific target gene expression,” EMBO Journal,vol.20,no.1- 2, pp. 91–100, 2001. ThisresearchwassupportedbyagrantoftheKoreaHealth Technology R&D Project through the Korea Health Industry [11] R. W. Ganster, B. S. Taylor, L. Shao, and D. A. Geller, “Complex regulation of human inducible nitric oxide synthase gene Development Institute (KHIDI), which is funded by the transcription by Stat 1 and NF-𝜅B,” Proceedings of the National Ministry of Health and Welfare, Republic of Korea (Grant no. AcadamyofSciencesoftheUnitedStatesofAmerica, vol. 98, no. HI14C0955). 15, pp. 8638–8643, 2001. [12] Z. Wen, Z. Zhong, and J. E. Darnell Jr., “Maximal activation of Supplementary Materials transcription by statl and stat3 requires both tyrosine and serine phosphorylation,” Cell,vol.82,no.2,pp.241–250,1995. Figure S1. UPLC chromatograms of KCT-01 (A) and reference [13]K.Q.deAndrade,F.A.Moura,J.M.dosSantos,O.R.P.de standards (B) at 203 nm. 1, chlorogenic acid; 2, hyperoside; Araujo,´ J. C. D. F. Santos, and M. O. F. Goulart, “Oxidative stress 3, scoparone; 4, isochlorogenic acid A; 5, isochlorogenic acid and inflammation in hepatic diseases: therapeutic possibilities B; 6, luteolin; 7, jaceosidin; 8, eupatilin; 9, ziyuglycoside I; 10, of N-acetylcysteine,” International Journal of Molecular Sciences, curcumin. Figure S2. UPLC chromatograms of IJCGT extract vol. 16, no. 12, pp. 30269–30308, 2015. (A) and reference standards (B) at 203 nm. 1, chlorogenic acid; [14] L. A. Zenewicz, G. D. Yancopoulos, D. M. Valenzuela, A. J. 2, ellagic acid; 3, hyperoside; 4, scoparone; 5, isochlorogenic Murphy, M. Karow, and R. A. Flavell, “Interleukin-22 but not acid A; 6, isochlorogenic acid B; 7, hesperidin; 8, luteolin; 9, interleukin-17 provides protection to hepatocytes during acute jaceosidin; 10, eupatilin; 11, ziyuglycoside I; 12, glycyrrhizic liver inflammation,” Immunity,vol.27,no.4,pp.647–659,2007. acid; 13, curcumin. (Supplementary Materials) [15]M.Ghasemian,S.Owlia,andM.B.Owlia,“ReviewofAnti- Inflammatory Herbal Medicines,” Advances in Pharmacological References Sciences, vol. 2016, Article ID 9130979, 2016. [16] C. Yun, Y. Jung, W. Chun et al., “Anti-inflammatory effects of [1] L. Ferrero-Miliani, O. H. Nielsen, P. S. Andersen, and S. E. artemisia leaf extract in mice with contact dermatitis in vitro Girardin, “Chronic inflammation: importance of NOD2 and and in vivo,” Mediators of Inflammation,vol.2016,ArticleID NALP3 in interleukin-1𝛽 generation,” Clinical & Experimental 8027537, 2016. Immunology,vol.147,no.2,pp.227–235,2007. [17] J.-H. Yang, J.-M. Yoo, W.-K. Cho, and J. Y. Ma, “Anti- [2] O. Takeuchi and S. Akira, “Toll-like receptors; their phys- inflammatory effects of Sanguisorbae Radix water extract on iological role and signal transduction system,” International the suppression of mast cell degranulation and STAT-1/Jak-2 Immunopharmacology,vol.1,no.4,pp.625–635,2001. activation in BMMCs and HaCaT keratinocytes,” BMC Comple- [3] Y.Fu, B. Liu, N. Zhang et al., “Magnolol inhibits lipopolysaccha- mentary and Alternative Medicine,vol.16,no.1,articleno.347, ride-induced inflammatory response by interfering with TLR4 2016. 14 Evidence-Based Complementary and Alternative Medicine

[18] M. Rana, P. Maurya, S. S. Reddy et al., “A standardized [33] A. Mobasheri, “Intersection of inflammation and herbal chemically modified Curcuma longa extract modulates IRAK- medicine in the treatment of osteoarthritis,” Current Rheuma- MAPK signaling in inflammation and potentiates cytotoxicity,” tology Reports, vol. 14, no. 6, pp. 604–616, 2012. Frontiers in Pharmacology,vol.7,articleno.223,2016. [34] M. P. Sheehan, M. H. A. Rustin, C. Buckley et al., “Efficacy of [19] H. Gao, W. Sun, J. Zhao et al., “Tanshinones and diethyl traditional Chinese herbal therapy in adult atopic dermatitis,” blechnics with anti-inflammatory and anti-cancer activities The Lancet,vol.340,no.8810,pp.13–17,1992. from salvia miltiorrhiza bunge (danshen),” Scientific Reports, [35] I. Shimizu, Y.-R. Ma, Y. Mizobuchi et al., “Effects of Sho-saiko- vol. 6, Article ID 33720, 2016. to, a Japanese herbal medicine, on hepatic fibrosis in rats,” [20] J.-S. Shin, E.-J. Cho, H.-E. Choi et al., “Anti-inflammatory effect Hepatology,vol.29,no.1,pp.149–160,1999. of a standardized triterpenoid-rich fraction isolated from Rubus [36] L. Liu, J. Zhao, Y. Li et al., “Artemisia capillaris formula inhibits coreanus on dextran sodium sulfate-induced acute colitis in hepatic steatosis via an MIR-122-induced decrease in fatty acid mice and LPS-induced macrophages,” Journal of Ethnopharma- synthase expression in vivo and in vitro,” Molecular Medicine cology,vol.158,pp.291–300,2014. Reports,vol.13,no.6,pp.4751–4758,2016. [21]E.G.Jang,Y.C.Kim,H.J.Woo,andJ.H.Lee,“Theeffects [37] K.-S. Kim, H. Yang, J. Lee et al., “Effects of 𝛽-sitosterol derived of Injincheonggan-tang on two patients with HBeAg-Negative from Artemisia capillaris on the activated human hepatic stellate chronic hepatitis B: case report,” JournalofKoreanMedicine,vol. cells and dimethylnitrosamine-induced mouse liver fibrosis,” 35,no.2,pp.69–78,2014. BMC Complementary and Alternative Medicine,vol.14,no.1, [22]H.Yang,D.Y.Lee,M.Jeon,Y.Suh,andS.H.Sung,“Deter- article363,10pages,2014. mination of five active compounds in Artemisia princeps [38] J. Chen, J. Chen, X. Wang, and C. Liu, “Anti-tumour effects of and A. capillaris based on UPLC-DAD and discrimination of polysaccharides isolated from Artemisia Annua L by inducing two species with multivariate analysis,” Archives of Pharmacal cell apoptosis and immunomodulatory anti-hepatoma effects Research,vol.37,no.5,pp.617–625,2014. of polysaccharides,” African Journal of Traditional, Complemen- [23] Y. H. Kim, C. B. Chung, J. G. Kim et al., “Anti-wrinkle tary and Alternative Medicines, vol. 11, no. 1, 2013. activity of ziyuglycoside I isolated from a Sanguisorba officinalis [39] E. Jang, B.-J. Kim, K.-T. Lee, K.-S. Inn, and J.-H. Lee, “A survey root extract and its application as a cosmeceutical ingredient,” of therapeutic effects of Artemisia capillaris in liver diseases,” Bioscience, Biotechnology, and Biochemistry,vol.72,no.2,pp. Evidence-Based Complementary and Alternative Medicine,vol. 303–311, 2008. 2015, Article ID 728137, 10 pages, 2015. [24] X.-G. He, L.-Z. Lin, L.-Z. Lian, and M. Lindenmaier, “Liquid [40] S. Jo, J. Ryu, H. Kim et al., “Anti-inflammatory effects of San- chromatography-electrospray mass spectrometric analysis of guisorbae Radix on contact dermatitis induced by dinitroflu- curcuminoids and sesquiterpenoids in turmeric (Curcuma orobenzene in mice,” ChineseJournalofIntegrativeMedicine, longa),” Journal of Chromatography A,vol.818,no.1,pp.127– 2015. 132, 1998. [25]M.Gasperotti,D.Masuero,U.Vrhovsek,G.Guella,andF. [41] J. S. Jurenka, “Anti-inflammatory properties of curcumin, a Mattivi, “Profiling and accurate quantification of Rubus ellag- major constituent of Curcuma longa: a review of preclinical and itannins and ellagic acid conjugates using direct uplc-q-tof clinical research,” AlternativeMedicineReview,vol.14,no.2,pp. hdmsandhplc-dadanalysis,”Journal of Agricultural and Food 141–153, 2009. Chemistry,vol.58,no.8,pp.4602–4616,2010. [42] Y. Rivera-Espinoza and P. Muriel, “Pharmacological actions of [26] D.He,Y.Shan,Y.Wu,G.Liu,B.Chen,andS.Yao,“Simultaneous curcumin in liver diseases or damage,” Liver International,vol. determination of flavanones, hydroxycinnamic acids and alka- 29, no. 10, pp. 1457–1466, 2009. loids in citrus fruits by HPLC-DAD-ESI/MS,” Food Chemistry, [43] J. E. Lee, S.-M. Cho, E. Park et al., “Anti-inflammatory effects of vol. 127, no. 2, pp. 880–885, 2011. Rubus coreanus Miquel through inhibition of NF-𝜅BandMAP [27]T.X.Duan,M.M.Yu,C.S.Liu,C.H.Ma,W.Q.Wang,and kinase,” Nutrition Research and Practice,vol.8,no.5,pp.501– S. L. Wei, “Simultaneous determination of glycyrrhizic acid, 508, 2014. liquiritin and fingerprint of licorice by RP-HPLC,” ChineseTra- [44] S. Lee and S. Cho, “Anti-inflammatory effects of Salviae Mil- ditionalPatentMedicine, vol. no. 2, p. 001, 2006. tiorrhizae Radix extract on RAW264.7 cell. via anti-oxidative [28] C. Scheinecker, K. Redlich, and J. S. Smolen, “Cytokines as activities,” The Korea Journal of Herbology,vol.30,no.4,pp.89– Therapeutic Targets: Advances and Limitations,” Immunity,vol. 94, 2015. 28,no.4,pp.440–444,2008. [45] Z. Liu, Y. Fan, Y. Wang et al., “Dipyrithione inhibits [29] G. Bonizzi and M. Karin, “The two NF-𝜅B activation pathways lipopolysaccharide-induced iNOS and COX-2 up-regulation and their role in innate and adaptive immunity,” Trends in in macrophages and protects against endotoxic shock in mice,” Immunology, vol. 25, no. 6, pp. 280–288, 2004. FEBS Letters,vol.582,no.12,pp.1643–1650,2008. [30] M.-W. Lin, L.-T. Tsao, L.-C. Chang et al., “Inhibition of [46] C. Tsatsanis, A. Androulidaki, M. Venihaki, and A. N. lipopolysaccharide-stimulated NO production by a novel syn- Margioris, “Signalling networks regulating cyclooxygenase-2,” thetic compound CYL-4d in RAW 264.7 macrophages involv- International Journal of Biochemistry & Cell Biology,vol.38,no. ing the blockade of MEK4/JNK/AP-1 pathway,” Biochemical 10, pp. 1654–1661, 2006. Pharmacology,vol.73,no.11,pp.1796–1806,2007. [47] J.-B. Kim, A.-R. Han, E.-Y. Park et al., “Inhibition of LPS- [31] M. Karin and Y. Ben-Neriah, “Phosphorylation meets ubiq- induced iNOS, COX-2 and cytokines expression by poncirin uitination: the control of NF-𝜅Bactivity,”Annual Review of through the NF-𝜅B inactivation in RAW 264.7 macrophage Immunology,vol.18,pp.621–663,2000. cells,” Biological & Pharmaceutical Bulletin,vol.30,no.12,pp. [32] N. Fujiwara and K. Kobayashi, “Macrophages in inflammation,” 2345–2351, 2007. Current Drug Targets—Inflammation & Allergy,vol.4,no.3,pp. [48] K.-J. Yun, J.-Y. Kim, J.-B. Kim et al., “Inhibition of LPS-induced 281–286, 2005. NOandPGE2productionbyasiaticacidviaNF-𝜅B inactivation Evidence-Based Complementary and Alternative Medicine 15

in RAW 264.7 macrophages: Possible involvement of the IKK and MAPK pathways,” International Immunopharmacology,vol. 8, no. 3, pp. 431–441, 2008. [49] J. H. Kim, D. H. Kim, S. H. Baek et al., “Rengyolone inhibits inducible nitric oxide synthase expression and nitric oxide production by down-regulation of NF-𝜅Bandp38MAPkinase activity in LPS-stimulated RAW 264.7 cells,” Biochemical Phar- macology, vol. 71, no. 8, pp. 1198–1205, 2006. [50] E. D. Chan and D. W. H. Riches, “IFN-𝛾 + LPS induction of iNOS is modulated by ERK, JNK/SAPK, and p38mapk in a mouse macrophage cell line,” American Journal of Physiology- Cell Physiology,vol.280,no.3,pp.C441–C450,2001. [51] E. Tedeschi, M. Menegazzi, D. Margotto, H. Suzuki, U. Forstermann,¨ and H. Kleinert, “Anti-inflammatory actions of St. John’s Wort: Inhibition of human inducible nitric-oxide synthase expression by down-regulating signal transducer and activator of transcription-1𝛼 (STAT-1𝛼) activation,” The Journal of Pharmacology and Experimental Therapeutics,vol.307,no.1, pp.254–261,2003. [52] M.Galdiero,M.Vitiello,M.D’Isanto,K.Raieta,andE.Galdiero, “STAT1 and STAT3 phosphorylation by porins are independent of JAKs but are dependent on MAPK pathway and plays a role in U937 cells production of interleukin-6,” Cytokine,vol.36,no. 5-6, pp. 218–228, 2006. [53] K. Shuai and B. Liu, “Regulation of JAK-STAT signalling in the immune system,” Nature Reviews Immunology,vol.3,no.11,pp. 900–911, 2003. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 6790292, 9 pages https://doi.org/10.1155/2017/6790292

Research Article Development of a Cardiovascular Simulator for Studying Pulse Diagnosis Mechanisms

Min Jang,1 Min-Woo Lee,2 Jaeuk U. Kim,3 See-Yoon Seo,4 and Sang-Hoon Shin2

1 Department of East-West Medical Engineering, Sangji University, Wonju 26339, Republic of Korea 2Department of Oriental Biomedical Engineering, Sangji University, Wonju 26339, Republic of Korea 3Korea Institute of Oriental Medicine, Daejeon 305-811, Republic of Korea 4Department of Korean Medicine, Sangji University, Wonju, Republic of Korea

Correspondence should be addressed to Sang-Hoon Shin; [email protected]

Received 25 August 2017; Accepted 5 November 2017; Published 15 November 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Min Jang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This research was undertaken to develop a cardiovascular simulator for use in the study of pulse diagnosis. The physical (i.e., pulse wave transmission and reflection) and physiological (i.e., systolic and diastolic pressure, pulse pressure, and mean pressure) characteristics of the radial pulse wave were reproduced by our simulator. The simulator consisted of an arterial component and a pulse-generating component. Computer simulation was used to simplify the arterial component while maintaining the elastic modulus and artery size. To improve the reflected wave characteristics, a palmar arch was incorporated within the simulator. The simulated radial pulse showed good agreement with clinical data.

1. Introduction Conventional simulators have been developed to perform functional evaluations of medical devices, such as left ven- Throughout history, pulse diagnosis has been a diagnostic tricular assist devices and artificial organs; thus, they are technique used in Oriental medicine to yield many clinically mainly focused on cardiac characteristics [6, 7]. Since the significant results. The technique provides a way to monitor physical properties of arteries are important in pulse diag- the overall health of the body by sensing the radial artery nosis research, heart-oriented cardiovascular simulators are pulse. inappropriate for investigations of pulse diagnosis. To modernize pulse diagnosis, the mechanisms of pulse Some studies have used simulators that focus on arte- diagnosis need to be explained in terms of modern science. rial characteristics. For example, Pahlevan and Gharib [8] Previous studies in this area [1–3] can be classified into three researched the effect of aorta tapering on a physical heart categories: clinical studies, mathematical simulation studies, model, and Knierbein et al. [9] developed an arterial tube and physical simulator studies. Conducting clinical research model using one mother branch divided into two daughter is expensive and time-consuming, and manipulating biolog- branches. However, the tube did not include the radial artery, ical variables is very challenging [4]. Mathematical modeling which is critical for pulse diagnosis. Full-branch elastic tube may also be impractical in some situations (e.g., arterial models have also been used to model the cardiovascular shunt) due to the assumptions made in the models. On the system [10, 11]. In humans, the elastic modulus of arteries other hand, cardiovascular simulations are inexpensive and increases as it reaches the periphery, but these simulators did variables can be easily controlled using computer programs. not reflect the central-to-peripheral elastic modulus gradient. Therefore, of the various types of research methods, research Because pulse diagnosis measures the radial artery pulse involving a cardiovascular simulator can be effective, mini- wave, this study aimed to develop a cardiovascular simulator mize the time and money required for the study, and solve that focused on the radial pulse. A key requirement for other practical issues [5]. investigating pulse diagnosis is the ability to replicate pulse 2 Evidence-Based Complementary and Alternative Medicine

Start Start

Q1 =F[q1(t)] Tree structure conguration P1 =(ZI)1 ·Q1

−1 Read vascular tube characteristics p1(t) =F [P1]

−1 pi (t) =F [Pi] Read ascending aorta pattern Qi =Pi/(ZI)i i=i+1

−1 Backward calculation of input impedance qi(t) =F [Qi]

i>NB Forward calculation of ow and pressure Stop

End Figure 2: Computational algorithm for the calculation of the pressure and the flow in the artery [12]. Figure 1: Computational algorithm for hemodynamics [12].

this study, human vascular dimensions and elastic constants wave transmission and the superposition of the forward were obtained from the literature [13]. The input impedance and reflected waves. Additionally, the characteristics of the (𝑍𝐼) was defined as simulated pulse wave should be within the physiological range. 1+Γ𝑒−2𝛾𝐿 𝑍 =𝑍 , (1) 𝐼 𝐶 1−Γ𝑒−2𝛾𝐿 2. Method where Γ is the reflection constant, 𝛾 is the wave propagation Thepulsewaveattheradialarteryisthesuperpositionofthe constant, and 𝑍𝐶 is an impedance characteristic that was forwardwave,producedbytheheart,andthereflectedwave defined as from the aortic bifurcation at the iliac artery [15]. The factors affecting pulse wave velocity include the geometric properties 𝜌𝑐0 −1/2 𝜙 𝜙 𝑍𝐶 = (1 − 𝐹10) (cos +𝑗sin ), (2) and elastic modulus of the relevant arteries. Additionally, the 𝐴√1−𝜎2 2 2 aortic bifurcation also impacts pulse wave reflection. In this study, the simulated arteries were designed to have human where 𝜌 is the fluid density in a tube, 𝑐0 is the pulse wave arterial characteristics. However, human arterial trees are so velocity in an inviscid fluid, 𝐴 is the cross-sectional area of complicated that simplification was needed. Thus, computer the tube, 𝜎 is Poisson’s ratio of the arterial wall, 𝐹10 is the simulation was used to simplify the arterial tree. To improve Womersley function [16], and 𝜙 represents the phase lead of thecharacteristicsofthereflectedwave,ashunt(palmararch) pressure in relation to wall displacement [13]. Blood pressures was incorporated into the simulator. The radial pulse of the and blood flows were calculated in every branch from the simulator was compared with that obtained from clinical ascending aorta to the peripheral artery. The calculation data. process is shown in Figure 2. 𝑞1(𝑡) is the given time-domain flow pattern in the ascend- 2.1. Arterial System Simplification ing aorta, 𝑝1(𝑡) is the 𝑝 time-domain pattern in the ascending 𝐹[ ] 𝐹−1[] 2.1.1. Hemodynamic Analysis Program. Figure 1 shows the aorta, is a Fourier transformation, is an inverse hemodynamic computational algorithm. After the arterial Fourier transformation, and NB is the number of branches. branch structure was defined, the arterial characteristics (e.g., length, elastic modulus, thickness, and radius) were input 2.1.2. Arterial System Simplification Using Simulation. The [12].Theinputimpedancewascalculatedasthatbetweenthe arterial properties described by Avolio [13] were used in this peripheral artery and the ascending artery. After calculating study. Some arterial branches are too complex and have too the input impedance, blood pressure and blood flow were small diameters to be included in the simulator. The arterial calculated from the ascending aorta to the peripheral artery tree was simplified using a hemodynamic analysis simulation using the measured flow or blood pressure patterns, as (Figure 3). Figure 4 shows a comparison between the two appropriate, at the ascending aorta [12]. Each branch was arterial tree models; the anatomical names of the arteries numbered for identification in the analysis, starting from the are shown in Table 1. With the same stroke volume input ascending aorta and continuing to the peripheral arteries. For into both models, the simplified arterial model showed a Evidence-Based Complementary and Alternative Medicine 3

(2) (10) (11) (12) (3) (13) (1) (4) (14) (5) (15) (6) (16) (7) (17) (8) (18) (9) (20) (19) (21) (22)

(a) (b)

Figure 3: Simplification of the arterial tree. (a) Full-branch model [13], (b) simplified model.

140 140

120 120

100 100 Pressure (mmHg) Pressure Pressure (mmHg) Pressure 80 80

60 60 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Time (s) Time (s)

Full-branch model Full-branch model Simplied model Simplied model (a) (b)

Figure 4: Comparisons of arterial tree models. (a) Ascending aorta, (b) radial artery. mean pressure increase compared with that in the full-branch component. The pulse-generating component consisted of arterial model. However, the pulse waves did not change. a rotating motor (1), slider-crank (2), cylinder-piston (3), Although the size of the reflected wave decreased slightly, the heart compliance chamber (4), and check valve ((5), (12)). overall structure of the reflected wave was retained. The arterial component consisted of arteries (6), peripheral resistance ((7), (8)), and a reservoir ((9), (10), and (11)). The 2.2. Cardiovascular Simulator rotation of the motor was transformed into linear motion by the slider-crank mechanism, producing a pulsatile flow 2.2.1. System Composition. The composition of the cardio- atthepiston-cylinder.Thefluidinthearterialsystemwas vascular simulator is shown in Figure 5. The simulator is drained to the reservoir, flowing back to the pulse-generating composed of a pulse-generating component and an arterial component. 4 Evidence-Based Complementary and Alternative Medicine

(11)

(12)

(4) (2) (1) (10) M (8) (6) (5) (3) (9) (7) (a) (b)

Figure 5: Simulator. (a) Photograph, (b) schematic diagram.

Table 1: Anatomical names of the simplified arterial tree. a sampling rate of 100 Hz. A laser distance-sensor (DT20- N244B, SICK) was used to measure the linear motion of the Branch number Anatomical name piston. All data were gathered and synchronized using a NI- (1) Ascending aorta DAQ (NI USB-6008, National Instruments). (2) Aortic arch 1 (3) Aortic arch 2 2.2.2. Arterial Model Construction and Simulation. Asim- (4) Thoracic aorta 1 plified arterial model was also manufactured. The length, (5) Thoracic aorta 2 thickness, diameter, and elastic modulus of each artery were (6) Thoracic aorta 3 individually determined following the guidance of Avolio (7) Abdominal aorta 1 [13]. The elastic modulus of one model artery was4 × 6 2 (8) Abdominal aorta 2 10 dyne/cm ,anditwasmadeofsilicon(SortaClear40, (9) Abdominal aorta 3 Smooth-On) mixed with a hardener. Another model artery × 6 2 (10) Subclavian artery 1 hadanelasticmodulusof8 10 dyne/cm ,anditwasalso (11) Subclavian artery 2 made of silicon (Smooth Sil 950, Smooth-On) mixed with a hardener. Figure 6 shows the developed arterial model; the (12) Axillary artery 1 vessel properties are shown in Table 2. Figure 7 shows the (13) Axillary artery 2 waveform results generated by the simulator. (14) Brachial artery 1 (15) Brachial artery 2 2.2.3. Arterial Tree Improvement. Figure 7(b) shows the radial (16) Brachial artery 1 artery pressure wave. However, the waveform is different (17) Brachial artery 2 from a clinically measured radial waveform because of the (18) Radial artery 1 superposition of the reflected wave from the end of the radial (19) Radial artery 2 artery. There is a shunt connecting the radial artery to the (20) Ulnar artery 1 ulnar artery [19, 20]. Figure 8 shows arterial tree with the (21) Ulnar artery 2 shunt, and the properties of the arteries are shown in Table 3. (22) Ulnar artery 3 3. Results and Discussion Figures 9, 10, and 11 show the simulator results using the A stepping motor (A200K-M599-G10, Autonics) was simplified arterial model. Specifically, Figure 9 shows the used for the rotating motor. A healthy heart has a normal pressure wave measured simultaneously at the ascending cardiac output of approximately 5.3 L/min [17]. Given that aorta and the radial artery. Figure 10 shows how changes in the size of the simplified model used in this study was 55% peripheral resistance affect the pressure at the radial artery of the full-branch arterial model, the cardiac output was set as well as the effects of the reflection ratio from the aortic at 2.9 L/min. Therefore, the pulse-generating component was bifurcation to the radial artery. Figure 11 shows a comparison setat82.8bpm(heartrate)witha35mLstrokevolume.To of the pressure wave from the simulator with clinical data [14]. mimic blood viscosity, the working fluid was a mixture of The cardiovascular simulator was developed to satisfy the 37% glycerin and 63% water [18]. The pressure sensor was following three conditions: (1) a focus on the radial pulse, an invasive type of pressure sensor (1620 pressure sensor, (2)apulsewaveproducedwithsuperpositionoftheforward MSISensors),anddatawereacquiredatasamplingrateof wave from the heart and the reflected wave from the aortic 1000 Hz. An ultrasound sensor (Bidop ES-100V3, Hadeco) bifurcation, and (3) a pressure wave within the physiological wasusedtomeasurebloodflow,anddatawereacquiredat range. Evidence-Based Complementary and Alternative Medicine 5

(1)

(2) (3)

(4)

(5)

(6) (7)

(a) (b)

Figure 6: Arterial tree. (a) Schematic model, (b) photograph.

140 140

120 120

100 100 Pressure (mmHg) Pressure (mmHg) Pressure 80 80

60 60 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Time (s) Time (s) (a) (b)

Figure 7: Simulated pressure wave. (a) Ascending aorta, (b) radial artery.

Table 2: Parameters used in the arterial tree. Elastic Number Branch Thickness in (cm) Thickness out (cm) Radius in (cm) Radius out (cm) Length (cm) modulus 6 2 (×10 dyne/cm ) (1) Aorta 0.16 0.05 1.47 0.58 41.40 4 (2) L. subclavian 1 0.08 0.07 0.44 0.40 10.70 4 (3) L. subclavian 2 0.07 0.06 0.40 0.33 10.60 4 (4) L. subclavian 3 0.06 0.05 0.33 0.33 8.80 4 (5) L. subclavian 4 0.05 0.04 0.33 0.33 9.00 4 (6) L. radial 0.04 0.03 0.20 0.20 27.20 8 (7) L. ulnar 0.04 0.03 0.22 0.22 26.90 8 L, left. 6 Evidence-Based Complementary and Alternative Medicine

(1)

(2) (3)

(4)

(5)

(6) (7)

(10) (8) (9)

(a) (b)

Figure 8: Improvement of arterial tree (with shunt). (a) Schematic model, (b) photograph.

140 140

120 120

100 100 Pressure (mmHg) Pressure Pressure (mmHg) Pressure 80 80

60 60 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Time (s) Time (s) (a) (b)

Figure 9: Simulator results using a simplified arterial model. (a) Ascending aorta, (b) radial artery.

Table 3: Parameters used in the arterial tree (with palmar arch).

Elastic Number Branch Thickness in (cm) Thickness out (cm) Radius in (cm) Radius out (cm) Length (cm) modulus 6 2 (×10 dyne/cm ) (1) Aorta 0.16 0.05 1.47 0.58 41.40 4 (2) L. subclavian 1 0.08 0.07 0.44 0.40 10.70 4 (3) L. subclavian 2 0.07 0.06 0.34 0.33 10.60 4 (4) L. subclavian 3 0.06 0.05 0.33 0.33 8.80 4 (5) L. subclavian 4 0.05 0.04 0.33 0.33 9.00 4 (6) L. radial 0.04 0.03 0.20 0.20 27.20 8 (7) L. ulnar 0.04 0.03 0.22 0.22 26.90 8 (8) Deep palmar arch 1 0.03 0.02 0.69 0.69 14.00 8 (9) Deep palmar arch 2 0.03 0.02 0.22 0.22 14.00 8 (10) Superficial palmar arch 0.03 0.02 0.21 0.21 14.00 8 L, left. Evidence-Based Complementary and Alternative Medicine 7

1.2

1.0

0.8

0.6

0.4

0.2 Normalized pressure Normalized

0.0

0.0 0.2 0.4 0.6 0.8 1.0 Time (s)

Decreasing Normal Increasing

Figure 10: The effects of the reflection ratio of the aortic bifurcation to the radial artery.

160 160

140 140

120 120

100 100 Pressure (mmHg) Pressure Pressure (mmHg) Pressure

80 80

60 60 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 Time (s) Time (s) (a) (b)

Figure 11: Comparison of the radial pulse. (a) Simulator, (b) clinical data [14].

The properties of young arteries were used in the simula- and 63 cm from that at the aortic bifurcation. Differences tor [13]. Figure 7 shows the simulator results in the absence in the forward wave arrival time and the difference in ofashunt.Inthefigure,aradialarterypressurewaveshowsa locations are seen in Figure 9. Pulse wave velocity (PWV) sharp peak at the beginning of systole. Because the end of the was calculated using a foot-to-foot method and found to be radial artery is closest to the measuring point, the reflected 5.9 m/s [21], which is within the clinical range of 5–8 m/s [22]. wave from the end of the radial artery comes first. However, The elastic modulus used for the simulator was based on data this does not reflect the physiological situation. To reduce from a young individual; therefore, the PWV was less than this discrepancy between the model and actual physiology, a that associated with adults [23]. The superposition time was shunt was incorporated into the simulation. Figure 7 shows then calculated to be 0.13 s; the incisura point is shown in results before adding the shunt and Figure 9 shows results Figure 9. According to Figure 9, the superimposition of the afteraddingtheshunt.ThepressurewaveinFigure9shows reflected wave occurred after the incisura point. that the sharp peak in early systole had disappeared, resulting In the pulse wave, the first peak indicates the maximum in a waveform similar to what might be seen for a young value of the forward wave, and the second peak indicates person [14]. The shunt adaptation allowed dissemination of the maximum value of the reflected wave. By comparing the the peripheral reflection, solving the artefact observed in difference in time between the two peaks, the gap between Figure 7. the first and second peak times in the ascending aorta was The distance between the pulse-generating component closer than that for the radial artery. This was due to the and the measurement site at the ascending aorta was 4 cm, faster superposition time at the ascending aorta than at 8 Evidence-Based Complementary and Alternative Medicine the radial artery [14]. Although the reflected wave traveled Conflicts of Interest 129.8cmtotheradialartery,ittraveledonly74.8cmtothe ascending aorta, resulting in the time difference. Therefore, The authors declare that there are no conflicts of interest thereflectedwaveshownintheradialarterywasconcludedto regarding the publication of this article. have come from the aortic bifurcation, as was demonstrated experimentally. Acknowledgments Figure 10 shows the pulse wave of the radial wave with different levels of aortic periphery resistance. With Thisstudywassupportedbyagrant(K17021)fromthe increasing resistance, an increase in mean pressure and a Korea Institute of Oriental Medicine, funded by the Korean decrease in pulse pressure were observed. To compare the government. sizes of the forward and reflected waves, the measured pressure waveforms were normalized. As a result, as aortic References periphery resistance increased, there was an increase in the gap between the incisura point and the second peak. This [1] Z. F. Fei, Contemporary Sphygmology in Traditional Chinese shows that the increase in the reflected wave was larger than Medicine, People’s Medical Publishing House, Beijing, China, 2003. that in the forward wave. The waveform and the location of the reflection remained unchanged, meaning that aortic [2] J.-H. Bae, J. Y. Kim, J. U. Kim, and Y. J. Jeon, “New assessment model of pulse depth based on sensor displacement in pulse peripheral resistance was only increased in the reflected diagnostic devices,” Evidence-Based Complementary and Alter- wave. An animal experiment conducted by Alastruey et native Medicine,vol.2013,ArticleID938641,9pages,2013. al. [24] also demonstrated that an increase in peripheral [3] J.-Y. Lee, M. Jang, and S.-H. Shin, “Study on the depth, rate, resistance accounted for the increase in the reflected wave. shape, and strength of pulse with cardiovascular simulator,” Thus, the developed simulator replicated the superposition of Evidence-Based Complementary and Alternative Medicine,vol. the forward wave of the pulse-generating component and the 2017, Article ID 2867191, 11 pages, 2017. reflected wave of the aortic bifurcation. [4] K.-W. Gwak, B. E. Paden, J. F. Antaki, and I.-S. Ahn, “Exper- The pressure wave of the ascending aorta at the sim- imental verification of the feasibility of the cardiovascular ulator was 71.0–103.0 mmHg, similar to clinically acquired impedance simulator,” IEEE Transactions on Biomedical Engi- data (73.0–116.0 mmHg). Similarly, the pressure wave at neering,vol.57,no.5,pp.1176–1183,2010. the radial artery was 86–140.9 mmHg, compared with [5] R. Zannoli, I. Corazza, and A. Branzi, “Mechanical simulator of 70.0–148.0 mmHg for clinical data, indicating that the sim- the cardiovascular system,” Physica Medica,vol.25,no.2,pp. ulated data were within the data range for a healthy person 94–100, 2009. [14].Asthepulsewavetraveledfromtheascendingaorta [6] D. Timms, M. Hayne, K. McNeil, and A. Galbraith, “Acomplete to the radial artery, the pulse pressure at the simulator mock circulation loop for the evaluation of left, right, and biventricular assist devices,” Artificial Organs,vol.29,no.7,pp. increased by 22.3 mmHg (range, 2.6–54.9 mmHg); in a clini- 564–572, 2005. calexample,thepulsepressureincreasedby25mmHg(range, 43.0–78.0 mmHg). An increase in the pulse pressure may be [7] T. I. Marx, B. R. Baldwin, and C. F. Kittle, “A cardiovascular simulator for the evaluation of prosthetic aortic valves,” The explained by a decreased arterial diameter and an increased Journal of Thoracic and Cardiovascular Surgery,vol.38,pp.412– elastic modulus, leading to increased systolic pressure [25]. 418, 1959. This phenomenon was also shown in the developed simulator. [8] N. M. Pahlevan and M. Gharib, “Low pulse pressure with high This study had a few limitations. There are many studies pulsatile external left ventricular power: influence of aortic regarding the geometric structure of the palmar arch, which waves,” Journal of Biomechanics,vol.44,no.11,pp.2083–2089, was developed at the distal end of the radial artery, in 2011. this study. However, the anatomical parameters such as the [9] B. Knierbein, H. Reul, R. Eilers, M. Lange, R. Kaufmann, and length, internal diameter, thickness, and Young’s modulus G. Rau, “Compact mock loops of the systemic and pulmonary were difficult to determine. We expect that the simulator can circulation for blood pump testing,” International Journal of be enhanced by determining precise boundary conditions Artificial Organs,vol.15,no.1,pp.40–48,1992. using more accurate parameters for the palmar arch artery. [10] K. S. Matthys, J. Alastruey, J. Peiroetal.,“Pulsewavepropa-´ gation in a model human arterial network: assessment of 1-D numerical simulations against in vitro measurements,” Journal 4. Conclusions of Biomechanics,vol.40,no.15,pp.3476–3486,2007. In this study, a cardiovascular simulator was developed [11] J.-Y. Lee, M. Jang, S. Lee, H. Kang, and S.-H. Shin, “Acardiovas- for use in studying the mechanisms associated with pulse cular simulator with elastic arterial tree for pulse wave studies,” Journal of Mechanics in Medicine and Biology,vol.15,no.6,8 diagnoses. The radial pulse from the simulator showed pages, 2015. superposition of the forward wave from the heart and the [12] S.-H. Shin and Y.-B. Park, “Effects of changes in the physical backward wave from the aortic bifurcation. Additionally, the properties of the central elastic artery on haemodynamic generated pulse wave was within the physiological ranges. characteristics during ageing,” Proceedings of the Institution Thus, this cardiovascular simulator may be useful for future of Mechanical Engineers, Part H: Journal of Engineering in pulse diagnosis studies. Medicine,vol.223,no.5,pp.525–535,2009. Evidence-Based Complementary and Alternative Medicine 9

[13] A. P. Avolio, “Multi-branched model of the human arterial system,” Medical and Biological Engineering and Computing,vol.18, no. 6, pp. 709–718, 1980. [ 14 ] W. W. Ni c h o l s a n d M . F. O R o u r k e , McDonald’s Blood Flow in Arteries, Hodder Arnold, London, UK, 5th edition, 2005. [15] R. D. Latham, N. Westerhof, P. Sipkema, B. J. Rubal, P. Reud- erink, and J. P. Murgo, “Regional wave travel and reflections along the human aorta: a study with six simultaneous micro- manometric pressures,” Circulation,vol.72,no.6,pp.1257–1269, 1985. [16] J. R. Womersley, “Oscillatory flow in arteries: the constrained elastictubeasamodelofarterialflowandpulsetransmission,” Physics in Medicine and Biology,vol.2,no.2,pp.178–187,1957. [17] P.Segers, E. R. Rietzschel, M. L. De Buyzere et al., “Noninvasive (input) impedance, pulse wave velocity, and wave reflection in healthy middle-aged men and women,” Hypertension,vol.49, no. 6, pp. 1248–1255, 2007. [18]D.Legendre,J.Fonseca,A.Andradeetal.,“Mockcirculatory system for the evaluation of left ventricular assist devices, endo- luminal prostheses, and vascular diseases,” Artificial Organs, vol. 32, no. 6, pp. 461–467, 2008. [19]V.P.S.Fazan,C.T.Borges,J.H.daSilva,A.G.Caetano,and O. A. R. Filho, “Superficial palmar arch: An arterial diameter study,” Journal of Anatomy,vol.204,no.4,pp.307–311,2004. [20] M. P. Suma, S. V. Kumar, and P. Ranganath, “An anatomical study of superficial palmar arch,” International Journal of Anatomy and Research,vol.2,no.4,pp.735–739,2014. [21]S.C.Millasseau,A.D.Stewart,S.J.Patel,S.R.Redwood,and P. J. Chowienczyk, “Evaluation of carotid-femoral pulse wave velocity: influence of timing algorithm and heart rate,” Hyper- tension,vol.45,no.2,pp.222–226,2005. [22] W. R. Milnor, Hemodynamics, Williams and Wikins, Pennsyl- vania, Pa, USA, 2nd edition, 1982. [23] M. P. Baldo, R. S. Cunha, A. L. Ribeiro et al., “Racial differences in arterial stiffness are mainly determined by blood pressure levels: results from the ELSA-Brasil study,” Journal of the American Heart Association,vol.6,2017. [24] J. Alastruey, S. R. Nagel, B. A. Nier, A. A. E. Hunt, P. D. Weinberg, and J. Peiro,´ “Modelling pulse wave propagation in the rabbit systemic circulation to assess the effects of altered nitric oxide synthesis,” Journal of Biomechanics,vol.42,no.13, pp. 2116–2123, 2009. [25] S. S. Franklin, W. Gustin IV, N. D. Wong et al., “Hemodynamic patterns of age-related changes in blood pressure: the Framing- ham heart study,” Circulation, vol. 96, no. 1, pp. 308–315, 1997. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 3203768, 7 pages https://doi.org/10.1155/2017/3203768

Research Article Experiences Providing Medical Assistance during the Sewol Ferry Disaster Using Traditional Korean Medicine

Kyeong Han Kim,1 Soobin Jang,2 Ju Ah Lee,3 Bo-Hyoung Jang,4 Ho-Yeon Go,5 Sunju Park,6 Hee-Guen Jo,7 Myeong Soo Lee,8 and Seong-Gyu Ko4

1 Department of Preventive Medicine, College of Korean Medicine, Woosuk University, Jeollabuk-do, Republic of Korea 2KM Fundamental Research Division, Korea Institute of Oriental Medicine, Daejeon, Republic of Korea 3Internal Medicine, College of Korean Medicine, Gachon University, Seongnam, Republic of Korea 4Department of Preventive Medicine, College of Korean Medicine, Kyung Hee University, Seoul, Republic of Korea 5Internal Medicine, College of Korean Medicine, Semyung University, Chungbuk, Republic of Korea 6Department of Preventive Medicine, College of Korean Medicine, Daejeon University, Daejeon, Republic of Korea 7Department of Pharmaceutical Affairs, The Association of Korean Medicine, Seoul, Republic of Korea 8Clinical Research Division, Korea Institute of Oriental Medicine, Daejeon, Republic of Korea

Correspondence should be addressed to Seong-Gyu Ko; [email protected]

Received 28 June 2017; Revised 7 September 2017; Accepted 1 October 2017; Published 7 November 2017

Academic Editor: Hanbing Li

Copyright © 2017 Kyeong Han Kim et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Background. This study aimed to investigate medical records using traditional Korean medicine (TKM) in Sewol Ferry disaster in 2014 and further explore the possible role of traditional medicine in disaster situation. Methods. After Sewol Ferry accident, 3 on- site tents for TKM assistance by the Association of Korean Medicine (AKOM) in Jindo area were installed. The AKOM mobilized volunteer TKM doctors and assistants and dispatched each on-site tent in three shifts within 24 hours. Anyone could use on-site tent without restriction and TKM treatments including herb medicine were administered individually. Results.Thetotalof1,860patients were treated during the periods except for medical assistance on the barge. Most patients were diagnosed in musculoskeletal diseases (66.4%) and respiratory diseases (7.4%) and circulatory diseases (8.4%) followed. The most frequently used herbal medicines were Shuanghe decoction (80 days), Su He Xiang Wan (288 pills), and Wuji powder (73 days). Conclusions. TKM in medical assistance can be helpful to rescue worker or group life people in open shelter when national disasters occur. Therefore, it is important to construct a rapid respond system using TKM resources based on experience.

1. Background industry, transportation, and higher mobility. Therefore, the response to man-made disasters has become more important A disaster is a social emergency that occurs when damage to than ever before [3]. the lives and possessions of people in a nation outweighs the capacity of the society to absorb the damage [1]. Worldwide, Also disasters have become complex and take various humans have experienced numerous disasters throughout forms, and it has become more difficult for today’s disasters to history,andwehavecontinuouslymadeeffortstoreducethe be efficiently dealt with using the existing response paradigm damage from various disasters, although there are differences [4]. In particular, it can be difficult to use medical devices that among individuals, cultures, and societies [2]. In the past, are essential to preserve life in an emergency due to a lack disasters mostly took the form of natural disasters, although of electricity, gas, tap water, medical products, and effective we have seen the possibility of man-made disasters increasing communication [5]. Groups of people who are evacuated as societies have become more developed, with advances in from an emergency area into an open shelter with poor 2 Evidence-Based Complementary and Alternative Medicine hygiene can experience minor symptoms, such as colds, Pang-Mok harbor, where the victims’ families and rescue diarrhea, and myalgia [6]. workers were on stand-by. On May 22, the third on-site tent Korea has experienced various man-made and com- was installed to provide medical assistance to rescue workers plex disasters during the last decades. These disasters have on the barge. On July 8, the assistance work being performed included subway arson: “a fire broke out in subway and 192 at the three on-site tents was transferred to the public medical people died in Daegu on February 18, 2003,” the collapse of clinic in the Jindo district, and the tents were removed [8] a department store: “the Sampoong Department Store col- (Figure 1). The process for providing medical assistance is lapsed and 501 people died in Seoul on June 29, 1995,” a bridge showninFigure2. collapse:“theSungsubridgewasbrokedownand32people died in Seoul on October 21, 1994,” and the sinking of a ship: 2.3. Patients. Anyone, victims’ families, rescue workers, vol- “SuhaeFerrysankand292peoplediedinBusanonOctober unteers,andsoon,whofeltabnormalinthefield,could 10, 1993” and have had a significant impact on Korean use the on-site tents without restriction. Also, both conven- society. Recently the Sewol Ferry sinking accident occurred tional medicine and traditional Korean medicine tent were in Korea and was a devastating disaster for the nation in installed,soitwasabletochoosewherepatientswantedto 2014. go. The data of the patient’s age and gender using traditional Particularly in Sewol Ferry disaster, there had been two Korean medicine tent were not provided by the government kinds of medical assistance operated separately: conventional due to the protection of personal information. medicine and TKM service. The goal of this study was to survey examples of medical assistance involving TKM pro- 2.4. Operation of On-Site Tents. The first tent was installed vided to the Sewol Ferry sinking victims and to further at a Jindo public gym, where most of the victims’ families explore the possible role of TKM in medical assistance. waited to receive treatment. Medical professionals treated 1,230 people or 15.4 people per day on average. Medical 2. Methods professionals at another tent installed at Pang-Mok harbor treated 630 people or 8.3 people per day on average. Addi- 2.1. Accident Outline. The Sewol Ferry (a 6,000 ton passenger tionally, three other on-site tents were installed at Pang- ship) was carrying 476 passengers. It departed for Je-Ju Island Mok harbor to supply conventional medical treatment to fromIncheonat9:00PMonApril15,2014.Theferrybeganto a total of 3,288 people. The number of patients treated lose its balance on the sea 1.7 miles away from the southwest wasfarlowerthanthenumbertreatedatthetwoother coast of Gwangmae-do, Jindo-gun, Jeollanam-do, at 8:50 AM on-site tents, which were installed for no longer than 23 on April 16, 2014, and completely sank on April 18, which days,where34.3peopleand62.9peopleweretreated led to 295 passenger deaths and 9 missing passengers. The per day on average [4], but the number of patients was Ministry of Health and Welfare and the Central Emergency similar to the number treated at an on-site tent that was Medical Center were first contacted about the accident installed for no longer than 90 days and provided Korean through the media at 9:30 AM on the 16th before a disaster medicine, treating 11.6 people per day on average [4]. control room was opened. They reinforced their response efforts by installing the first on-site emergency tent at 9:40 2.5. Supply of Medical Services. The AKOM mobilized vol- AM and supplying medical products as well as other facilities. unteer Korean medicine doctors and assistants, dispatching In the first rescue, 47 passengers were sent to the on-site seven Korean medicine doctors along with four assistants emergency tent until 3:00 PM for triage based on the severity on average in three shifts within 24 hours. Total 67 Korean of their injuries and were then transported to local medical medicine doctors participated (male : female = 46 : 21). Six institutions. With confirmation that there would be no more Korean medicine university hospitals (Kyung Hee, Won rescues, efforts to provide medical assistance began [4]. Gwang, Domg Guk, Daegu Hanny, Dong Shin, and Dong Eui) and four private Korean medicine hospitals (Jaseng, 2.2. Installation of On-Site Tents for the Administration of Chungyeon, Dongseo, and Jooenraphas) dispatched medical Korean Medicine. There was sporadic mobilization of med- staffs regularly. Also doctors of private Korean medical clinics ical assistance from national Korean medicine doctors and participated. All of Korean medicine doctors had national institutions beginning at the time of the accident. On April license and more than 5-year experience. Medical supplies, 17, the day after the accident, the AKOM formed a special products, and other consumables were purchased voluntarily disaster committee to integrate Korean medicine, medical by members of the AKOM or through fund raising. At the volunteers, and institutions. The AKOM consulted with the medical office, Korean medicine doctors diagnose and treat Ministry of Health and Welfare to install the first on-site patients independently. There was no specific guideline for tent at Jindo public gym, where the victims’ families were, treatment and doctor decided the treatment method, herbal and dispatched Korean medicine doctors and assistants [7]. medicine individually. Acupuncture, moxibustion, cupping, Beginning on April 20, the on-site tent was operated by the pharmacoacupuncture, and chuna were performed, and Central Emergency Medical Center, which reported on the patching and taping treatments were administered. Herbal medical activities twice a day to the medical assistance board medicines, including decoctions, powders, pills, and extracts, of the Ministry of Health and Welfare. At the request of the were prescribed depending on symptoms. The list of herbal Center, the second on-site tent was installed on April 23 at medicines supplied at that time can be found in Table 3. Evidence-Based Complementary and Alternative Medicine 3

Figure 1: Distribution of on-site tents in Jindo.

April 15 April 16 April 17 April 20 April 23 May 22 July 8

Departure of Sinking Establishment First on- Second Third on- The three Sewol Ferry in the of a special site on-site site tents were from Incheon Jindo sea disaster medical medical medical disassembled to Jeju committee by tent tent tent the Traditional installed at installed at installed Korean a gym in Pang-Mok on a Medical Jindo port barge Association Figure 2: Process for providing medical assistance following the Sewol Ferry disaster using traditional Korean medicine.

The questionnaire was not implemented because of the 3. Results hectic situation directly after the disaster. Also the patients’ residential areas were scattered throughout the country, and 3.1. Medical Activities. Excludingthemedicalassistancepro- treatment effect of each patient could not be judged suffi- vided on the barge, for which there were no medical records, a total of 1,860 patients were treated during the disaster period. ciently. The first on-site tent installed at the Jindo public gym treated 1,230 patients, and the second on-site tent at Pang-Mok 2.6. Ethics and Consent. We collected the medical records harbor treated 630 patients. Regarding the medical activities with the cooperation of both the AKOM and Ministry of at the first on-site tent, 453 patients (36.8%) were treated Health and Welfare. In this study, we could not receive an starting from the day of the accident on the 5th through the original individual medical chart. We received, however, the 14th; 341 patients (27.7%) were treated from the 15th through medical activities that were reported twice a day to the med- the 30th; and 436 patients (35.5%) were treated after the 30th ical assistance board of the Ministry of Health and Welfare. (Table 1). An institutional review board at the Kyung Hee University At the medical office, Korean medicine doctor diagnosis approved this study (number KHSIRB-16-049) for use of the and treat patients independently. There was no specific guide- medicinal records. line for treatment and doctor decided to treatment method, 4 Evidence-Based Complementary and Alternative Medicine

Table 1: Operational results of medical assistance provided using Table 2: Chief complaints of patients treated with traditional traditional Korean medicine following the Sewol Ferry disaster. Korean medicine at the first on-site tent.

Classification Number of patients (%) Chief complaints Number of patients (%) Total treated patients (2014. 4. 20-7.8) 1,860 (100.0) Musculoskeletal disorders Site distribution Neck stiffness 112 (11.2) 1st on-site tent (2014. 4. 20-7.8) 1,230 (66.1%) Sprain 32 (3.2) 2nd on-site tent (2014. 4. 23.-7.8) 630 (33.9%) Low back pain 166 (16.6) 3rd on-site tent (2014. 5. 22.-7.8) Not reported Upper limb pain 35 (3.5) ∗ Period distribution Knee pain 47 (4.7) Acute (5th–14th) 453 (36.8%) Shoulder pain 118 (11.8) Subacute (15th–30th) 341 (27.7%) Lower limb pain 103 (10.3) Chronic (after 30th) 436 (35.5%) Hip joint pain 19 (1.9) ∗ 1st on-site tent/incident onset. Others 11 (1.1) Subtotal 643 (64.1) Gastrointestinal disorders herbal medicine individually. Acupuncture, moxibustion, Dyspepsia 58 (5.8) cupping, pharmacoacupuncture, and chuna were performed, Constipation 10 (1.0) and patching and taping treatments were administered. Abdominal pain 13 (1.3) Herbal medicines, including decoctions, powders, pills, and Diarrhea 6 (0.6) extracts, were prescribed depending on symptoms. The list Subtotal 87 (8.7) of herbal medicines supplied at that time can be found in Table 3. Respiratory disorders Acute pharyngitis 33 (3.3) 3.2. Chief Complaints of Patients Involved in the Sewol Ferry Common cold 26 (2.6) Disaster. There were 1,003 diagnostic medical records from Acute rhinitis 20 (2.0) the first on-site tent in the Jindo public gym, and the Subtotal 79 (7.9) diagnoses are shown in Table 2. The major reported symp- Psychological problems toms included musculoskeletal disorders (64.1%), followed by Anxiety 47 (4.7) gastrointestinal disorders (8.7%) and psychological problems Depression 53 (5.3) (8.7%), respiratory disorders (7.9%), neurological disorders Insomnia 19 (1.9) (6.1%), exhaustion (3.4%), and circulatory disorders (0.2%), Subtotal 87 (8.7) as well as other symptoms (1.2%). The detailed symptoms Neurological disorders were as follows: (1) lower back pain (16.6%), shoulder pain Headache 43 (4.3) (11.8%), neck stiffness (11.2%), lower limb pain (10.3%), knee pain (4.7%), upper limb pain (3.5%), sprain (3.2%), hip joint Nausea 5 (0.5) pain (1.9%), and other (1.1%) in the category of musculoskele- Dizziness 13 (13.0) tal disorders; (2) dyspepsia (5.8%), abdominal pain (1.3%), Subtotal 61 (6.1) constipation (1.0%), and diarrhea (0.6%) in the category Exhaustion 34 (3.4) of gastrointestinal disorders; (3) depression (5.3%), anxiety Circulatory disorders 2 (0.2) (4.7%), and insomnia (1.9%) in the category of psychological Others problems; (4) acute pharyngitis (3.3%), common cold (2.6%), Tinnitus 1 (0.1) and acute rhinitis (2.0%) in the category of respiratory disor- Acute hemiplegia 1 (0.1) ders; and (5) headache (4.3%), dizziness (13.0%), and nausea Poststroke syndrome 1 (0.1) (0.5%) in the category of neurological disorders. In addition, in the category of miscellaneous symptoms, pruritus (0.3%), Edema 2 (0.2) edema (0.2%), superficial burns (0.2%), tinnitus (0.1%), acute Pruritus 3 (0.3) hemiplegia (0.1%), poststroke syndrome (0.1%), lip tremor Lip tremor 1 (0.1) (0.1%), and hand tremor (0.1%) were recorded. The records Hand tremor 1 (0.1) on major symptoms of other on-site tents were not written in Superficial burns 2 (0.2) report. Subtotal 12 (1.2) Total 1,003 (100) 3.3. Herbal Medicines Used for Sewol Ferry Disaster Medical Assistance. It was reported that herbal medicines were pre- scribedatthefirston-sitetentattheJindopublicgymfrom April 20 to May 16. The formulae of the herbal medicines He Xiang Wan (288 pills), and Wuji powder (73 days). The included decoctions, pills, and powders. The most frequently details of the amounts used and the effectiveness of every usedherbalmedicineswereShuanghedecoction(80days),Su herbal medicine are presented in Table 3. Evidence-Based Complementary and Alternative Medicine 5

Table 3: Herbal medicines used in the medical assistance provided following the Sewol Ferry disaster.

∗ Classification Herbal medicine Efficacy Amount Fragile habitus, fatigue recovery, overwork, during Shuanghe decoction 80 days illness or convalescence Guy ZhiMa Huang Ge Ban Tang Cold, cough, itch 10 days Xiang Sha Ping Wei San Anorexia with dyspepsia, gastric atony 41 days Guizhi decoction Initial cold with physical deterioration 12 days Ganmaidazao decoction Sudden loss of consciousness and convulsions 22 days Decoction Ganjie decoction Swelling and pain in throat 16 days Insampaedok-san Fatigue cold, fever, headache 16 days Bronchitis, bronchial asthma, runny nose, cough Xiao Qing Long Tang 31 days with dilute phlegm, rhinitis Summer cold, anorexia caused by heat, diarrhea, Huoziang-Zhengqi powder 1day tiredness Shoulder pain, fatigue, and anxiety of weak Modified Xiaoyao powder 17 days woman Jiawei Wendan decoction Insomnia with stomach weakness, nervousness 9 days Feeling heavy with worries, anger, suddenly Su He Xiang Wan 288 pills falling down Feeling of helplessness, physical degradation, Herbal Gongjin-Dan dizziness due to deterioration of liver function, 137 pills medicinal Pill headache, chronic fatigue products Insomnia, anxiety, thirsty, palpitation, shortness Tianwangbuxin Dan 107 pills of breath, neurasthenia, forgetfulness, hot chest Uwhangchungsimwon Palpitation, nervousness, autonomic imbalance 53 pills Buzhongyiqi decoction Fragile habitus, tiredness, convalescence, anorexia 2 days Gastroenteritis, low back pain, neuralgia, Five Retention powder 73 days arthralgia, feeling of cold, common cold Stomach weakness, feeling of cold in lower limbs, BanxiaBaizhuTianma Decoction 6days Herbal dizziness, headache medicinal Powder products Shensuyin Cold, cough 2 days Taorenchengqi decoction Neurasthenia, hypertension 16 days Xiang Sha Ping Wei San Anorexia with dyspepsia, gastric atony 14 days Headache caused by cold, neck stiffness, fever, JiuweiQianghuo decoction 14 days arthralgia ∗ Data from the Korea Pharmaceutical Information Center, http://www.health.kr/.

4. Discussion psychological discomfort. At 2nd on-site tent and 3rd on- site tent, mainly ground rescue agents and divers those who Themajorsymptomsthatweretreatedattheconventional suffered from cold and musculoskeletal disorders used TKM medicine on-site tents were the musculoskeletal system service.Thereasonforthesimilarpatternsofbothdaily (54.8%), followed by the digestive system (6.9%) and the patients and their symptoms between the Korean medical respiratory system (6.1%) [4], whereas the major symp- on-site tents and medical on-site tents was that most of the toms treated at the Korean medical on-site tents were the patients were family members or relatives of dead or missing musculoskeletal system (61.1%), followed by the digestive victims, not the victims themselves. These trends indicate system (8.7%). The only difference was that the Korean that Korean medicine may play a role in addressing medical medicine on-site tents treated mental disorders (5.1%) and needs other than surgical needs during disaster situations. exhaustion (2.5%), whereas the medical on-site tents treated This indication was supported by a previous study regarding dermatological diseases (11.1%). The dermatological diseases the effects of Korean medicine treatments following disasters that required rapid treatment were given preference at the to treat cold symptoms, diarrhea, and myalgia [6]. Given that medical on-site tents, and Korean medicine was chosen there must be other medical needs in addition to the need for the treatment of psychological and physical symptoms, for first-aid treatment during disasters [7, 9, 10], the role of such as mental disorders and exhaustion. At 1st on-site tent, Korean medicine treatments is important. patients were mostly family of victims and their main There was confusion among the government, the AKOM, symptoms were cold, diarrhea, and decreased physical and and volunteers regarding the installation and operation of an 6 Evidence-Based Complementary and Alternative Medicine on-site tent in the early stages of the Sewol Ferry accident be listed by disaster in advance and primarily supplied to response because Korean medicine is not included in the the disaster area; support could then be provided from the national disaster medical assistance system. The actual work surrounding areas when supplies are low. at the on-site tent began four days after the accident, and all In a disaster situation, medical doctors can be responsible the staff who were mobilized were volunteers. Any medical for first-aid treatment of patients, and Korean medicine supplies, products, and consumables used were provided by doctors can treat rescue workers. There is a report showing Korean medicine doctors who donated money for assistance. that prescriptions of Maimendong decoction and Xiao Qing Some beds and desks were provided by a public health clinic Long Tang were effective for individuals at a rescue site in Jindo district. Korean medicine doctors sent decoctions who had a runny nose, stuffy nose, or dry cough, and these and pills for prescriptions for expected disorders to the on- prescriptions can be used instead of antihistamines, which site tents, with Shuanghe decoction being the most popular have side-effects that reduce attention and promote sleepiness prescription. The AKOM subsequently minimized the pos- [6]. In a group facility, such as a shelter with poor hygiene, sibility of any overlapping prescriptions provided in large minor symptoms, including chills, fever, cough, runny nose, quantities through daily checks and online reports of defi- diarrhea, constipation, and insomnia, are frequently preva- cient prescription supplies. In addition, no standardized form lent, and Korean medicine treatments could help to alleviate for medical records was prepared, and every doctor used his these symptoms. or her own form; thus, all of the records were different from There are some limitations in this study, and overcoming oneanother.TheAKOMlaterprovidedamedicalrecordform these is prerequisites in order that TKM could be included in only to record simple medical information, but it was not disaster medical system. First of all, the effectiveness of TKM designed to record information reflecting the disaster situa- treatment was not measured because of its special and urgent tion itself. Moreover, there was no collection of basic informa- situation. Although detailed investigation might have been tionsuchasthesexandageofthevictimsandtheirfamilies difficult, appropriate and simple evaluating measurements who received treatment. Therefore, it was apparent that the for disaster should be applied. Second, many medical data use of such information was limited. were missing in report and statistical analyses could not be No education was implemented before the medical staff done. From now on, medical records in disaster situations were dispatched, and they therefore failed to adequately need to be fully recorded and archived. Third, a scheme respond to the typical PTSD symptoms, including extreme for disaster response education with certification must be physical and mental anxiety and drastic deterioration of phys- developed for Korean medicine doctors, and these doctors ical status. Furthermore, medical products and consumables must be regularly required to participate in refresher training were not properly supplied due to a lack of consideration of courses. the nature of medical needs following the disaster. Although this limitation, the effect of the treatment was Under these circumstances, the second on-site tent was indirectly measured in two respects. installed at Pang-Mok to meet the growing need for Korean First, patients treated in traditional Korean medicine on- medicine at the site after installation of the first on-site tent site tent in Jindo public gym (15.4 people per day on average) atthegymintheJindodistrict.Thethirdon-sitetentwas and it is similar to number treated conventional medicine installed on a barge because of health concerns raised about on-site tent (11.6 people per day on average) which operated the staff, since a diver who was conducting a search on May forasimilarperiod.Theuseofmorepatientscomparedto 6, 2014, died due to loss of consciousness. conventional medicine on-site tent can indirectly show the To prepare to provide Korean medicine assistance more traditional Korean medicine treatment’s therapeutic effect. rapidly and efficiently in a similar situation, we suggest Also considering that the victim’s family mainly lives in Jindo the following policies. Since the Sewol Ferry accident, the publicgym,itcanbeassumedthattraditionalKoreanmedical national safety authority was launched in November 2014 assistance could contribute to group life in a disaster situ- to meet nationwide need for reform of the national disaster ation. Acupuncture and moxibustion treatments are timely, response system. The national safety authority has integrated handy,andeasytouse.Inaddition,theyhavethemeritofpro- the existing response systems, which were divided into sys- viding medical services without the need of any special med- tems for land, sea, natural, and social disasters, in an attempt ical facilities as far as the operator doctor is skilled in emer- to establish itself as a control center for national disasters [11]. gency.Thus,theycanbeappliedimmediatelyinadisastrous However, Korean medicine is currently not included in the situations. national disaster response system, which could again lead to Secondly, the first on-site tent in Jindo public gym was confusion in terms of the installation and operation of on-site installedbytheAKOM,butthesecondon-sitetentinPang- tents and other responses to incidents such as the Sewol Ferry Mokharborandthethirdon-sitetentinthebargewereall accident. Therefore, the inclusion of Korean medicine in the installed at the request of the government. This indicates national disaster response system is highly recommended. that the patients were satisfied with the first on-site tent at To this end, the government should develop an integrated the time and requested to install additional on-site tents [7]. medical assistance model including Korean medicine in Therefore, it can be assumed that the patient’s satisfaction of the disaster response, and based on this model, local gov- traditional Korean medical assistance was high. The use of ernments should work with related organizations in their Korean traditional medicine assistance during Sewol disaster regionstoestablishadisasterresponseplanandconductjoint was similar to that of conventional medicine assistance, and training. Regarding medical products and supplies, they must the satisfaction of the patients was also high. Therefore, it Evidence-Based Complementary and Alternative Medicine 7

is possible to confirm the possibility of traditional Korean [11] G. J. Jung, Improvement of disaster safety R&D promoting medicine assistance in a disaster situation. Particularly, as the system according to establishment of Ministry of Public Safety and case of Sewol disaster, emergency patients rarely occurred, Security, Korea Institute of Science & Technology Evaluation have to live a group life, and need medical assistance for and Planning, 2014. rescue workers, traditional Korean medicine assistance also needed with conventional medicine assistance. In Korea, especially, Korean medicine is included in the public medical system among the medical system, so it can be considered that it is easy to provide medical service in the state of emergency. In the future, it will be necessary to systematically manage the traditional Korean medicine assistance within the national disaster medical system. In addition, it will be necessary to nurture a Korean medicine doctor who has expert knowledge to cope with the disaster situation.

Conflicts of Interest The authors declare no conflicts of interest.

Acknowledgments This work was supported by the Traditional Korean Medicine R&D Programme funded by the Ministry of Health and Welfare through the Korea Health Industry Development Institute (HI12C1889).

References

[1] K. L. Koenig and C. H. Schultz, Koenig and Schultz’s disaster medicine: comprehensive principles and practices, Cambridge University Press, New York, NY, USA, 2009. [2]K.J.BaekandY.S.Hong,“CurrentstatusofKoreanDisaster Medicine - Analysis of Railroad Collapsed Accident of Gupo,” Journal of the Korean Society of Emergency Medicine,vol.4,no. 2, pp. 40–46, 1993. [3] J. Admas, Emergency Medicine: Clinical Essentials,Saunders, Philadelphia, Pa, USA, 2nd edition, 2013. [4] W. Hong, I. Kim, and S.-J. Wang, “Experiences and lessons of the disaster medical assistance in Korea,” Journal of the Korean Medical Association,vol.57,no.12,pp.999–1007,2014. [5]J.W.Kim,S.H.Park,Y.S.Jung,andJ.P.Cho,“Analysisof Medical Service In the 1999 Turkey Earthquake,” Journal of the Korean Society of Emergency Medicine,vol.12,no.3,pp.330–337, 2001. [6] S. Takayama, R. Okitsu, K. Iwasaki et al., “The Role of Oriental Medicine in the Great East Japan Earthquake Disaster,” Kampo Medicine,vol.62,no.5,pp.621–626,2011. [7]Y.J.Kang,W.J.Kim,andJ.O.Park,“Characteristicsofinjured patients related with typhoon Nari,” Journal of the Korean Society of Emergency Medicine,vol.19,pp.462–473,2008. [8] T.A.o.K. Medicine, “White paper of medical support in Sewol Ferry disaster,”Tech. Rep., The Association of Korean Medicine, Seoul, Korea, 2014. [9] K. C. You, M. E. Ahn, Y. J. Cho, J. M. Chaeng, and K. S. Lim, “Injury type in Sampung collapse,” Journal of the Korean Society of Emergency Medicine, vol. 8, pp. 185–192, 1997. [10] K. W. Lee, “Clinical analysis of the stadium stampede in Sang- ju, Korea,” JournaloftheKoreanSocietyofEmergencyMedicine, vol. 18, pp. 367–374, 2007. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 4835912, 9 pages https://doi.org/10.1155/2017/4835912

Review Article Modernization Trends of Infertility Treatment of Traditional Korean Medicine

Jang-Kyung Park1 and Dong-Il Kim2

1 College of Korean Medicine, Dongguk University, Seoul, Republic of Korea 2Department of Obstetrics and Gynecology, College of Korean Medicine, Dongguk University, Seoul, Republic of Korea

Correspondence should be addressed to Dong-Il Kim; [email protected]

Received 24 July 2017; Accepted 24 September 2017; Published 5 November 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Jang-Kyung Park and Dong-Il Kim. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Despite the development of assisted reproductive technology (ART), it is difficult to increase the implantation rate. In Korea, Traditional Korean Medicine, including herbal medicine, is an important component of infertility treatment. Korean medical doctors who are treating infertility often use herbal medicine to promote implantation. In this article, as one of the research works on modernization of Traditional Korean Medicine, we investigated the experimental studies to clarify the effects of herbal medicines that are traditionally used to promote pregnancy. We searched for experimental studies over the past 10 years of improvement of endometrial receptivity in herbal medicine using six domestic and international sites. We analyzed 11 studies that meet the selection criteria. We found that herbal medicines demonstrably improved endometrial receptivity and increased pregnancy rates.

1. Introduction receptivity in herbal medicine. Studies on embryo implantation are mainly focused on experimental studies because The modernization of Traditional Korean Medicine means they can cause ethical problems; therefore, only experimental combining or converting it into modern technology and studieswereselectedinthisstudy[8]. modern scientific culture, which includes studies of the Through review of studies, we tried to clarify the effect efficacy, effect, and mechanism of herbal medicine [1–3]. of the herbal medicine on the endometrial receptivity and In 2012, South Korea’s infertility rate is 32.3%, an annual ascertain the significance as an effective treatment modality average growth rate of 7.7%. The overall fertility rate is for infertility. the lowest in the world, causing a serious social problem [4]. Despite advances in assisted reproductive technology 2. Materials and Methods (ART) and national support program for infertile couples, the success rate remains 25% to 30% because of implantation 2.1. Data Sources and Searches. We searched experimental dysfunction [5]. Infertile couples in South Korea often turn studies in six domestic and foreign search sites (NDSL, to Traditional Korean Medicine as an important component KSTUDY, RISS, OASIS, JKOM, and MEDLINE). With of primary or secondary treatment of infertility [6]. June1,2017,astheenddate,thestudywaslimited The implantation process is through the interaction of to the last 10 years, and the language was limited to the embryo and the endometrium and occurs in a very Korean and English. In domestic search sites, we searched limited period of time, called the window of implantation. for the keywords “endometrial receptivity”, “Traditional Ovarian stimulation, though essential for IVF,may itself have Korean Medicine”, and “Herbal medicine”. In Medline, detrimental effects on endometrial receptivity, embryonic we used the mesh term of “medicine, east Asian tradi- implantation, and perhaps pregnancy outcomes [7]. tional”, “medicine, Chinese traditional”, “medicine, Korean As part of the study on modernization of Korean traditional”, “medicine, Kampo”, “plant extracts”, “herbal Medicine, we reviewed experimental studies of endometrial medicine”, “uterine receptivity”, “uterine endometrium”, 2 Evidence-Based Complementary and Alternative Medicine

Articles identiﬁed through database searching (n=51) Excluded after screening the abstract (i) Duplicated publication (n=2) 49 articles screened Publication excluded after screening the text (i) not related to Traditional Korean medicine (n=2) (ii) not related to endometrial receptivity (n=14) (iii) not related to herbal medicine (n=5) (iv) related to Anti-infertility or Anti-implantation (n=12) (v) related to Toxicity or Drug safety (n=5) 11 Studies included in analysis

Figure 1: The process of data selection and extraction.

“endometrial receptivity”, and “embryo implantation”. We control (progynova [10], aspirin [16]) was also established. To searchedforeachofthetermsandfinallysearchedfor induce the COH model, pregnant mare serum gonadotropin intersection combination. (PMSC) and human chorionic gonadotropin (hCG) were used. Mifepristone was used in four cases and indomethacin 2.2. Study Selection. We searched experimental studies of the in three studies to derive EID. Li et al. described a COH positive effect of endometrial receptivity in herbal medicine. model,butitwasjudgedtobeanEIDmodelinducedby Exclusions included studies of the mechanism by which indomethacin [12]. Traditional Korean Medicine interferes with pregnancy and safety and toxicity studies. Studies where the intervention 3.4.2. Outcome Measures. Immunohistochemical assay, real was not herbal medicine and where the subject was not time polymerase chain reaction (RT-PCR) assay, TUNEL endometrial receptivity were also excluded (Figure 1). method, biomarkers associated with endometrial receptivity such as endometrial thickness, angiogenesis, and pinopodes 3. Results by microscopic examination were measured by microsocpic exam. Polypeptides associated with endometrial receptiv- 3.1. Study Description. No articles were found in domestic ity and microscopic observation with pregnancy were the search sites. In Medline, 51 studies were found. Out of these, most frequently used indicators of evaluation. Number of 11studieswereselected.Twooftheauthors(PJK,KDI)read implanted blastocysts, pregnancy rate, and expression of inte- the title and abstract and reviewed the 51 initial papers. grin 𝛽3 mRNA were used in four studies. Expression of LIF After selecting 49 papers, excluding two duplicate papers, mRNA was used in three studies, expression of LIF protein, two authors (PJK, KDI) reviewed the original articles and and implantation sites were used in two studies, and COX- extracted the key information. When opinions differed, the 2, IFN-𝛾,IL-10,integrin𝛼] mRNA, integrin 𝛽3proteins,LIF decision was agreed through discussion. mRNA/𝛽-actin, MMP-9, NF-𝜅jOPN mRNA, PGI2, PPARd, IL-11 mRNA, TIMP-3, apoptotic index, proliferative index, 3.2. Analysis of the Publishing Year. The review found that one LCM-DE-MS, microvessel density, endometrial thickness, paper on the topics of interest was published yearly from 2008 andpinopodesintheepitheliumwereusedinonestudy to 2014, and two were published yearly in 2015 and 2016. (Table 1).

3.3. Analysis of the Publishing Country. Ten out of eleven 3.5. Analysis. In all 11 studies, herbal medicines improved studies were from China, and one was from South Korea. endometrial receptivity compared to the model group. There was an improvement in the expression of polypeptides 3.4. Analysis of Study Design related to endometrial receptivity compared to aspirin and progynova (Table 1). 3.4.1. Control Group. In all 11 studies, in vivo studies were performed rather than in vitro. All 11 studies included in vivo 3.6. Herbal Medicine. In one study, the efficacy of a single studies; Choi et al. [9] conducted both in vivo and in vitro herbal medicinal material was assessed, and, in ten studies, studies. the efficacy of a prescription consisting of a herbal combina- Each study included at least two control groups, including tion was evaluated. a normal group and model group. The models were either COH induction or EID induction. Embryo implantation 3.6.1. Herbal Formula. Bushenantai prescription was used dysfunction (EID) model was used in six studies [6, 9– in two studies [6, 14], and Xianziyizhen [10], Bushenyiqi- 13] and controlled ovarian hyperstimulation (COH) model hexue [11], DS-1-47 [12], Bangdeyun [13], Er’zhi Tiangui [15], in four studies [14–17]. In one study, both EID model and Shoutaiwai [16], Yiqixue Buganshen [17], and Zhuyun [18] COH model were examined [18]. In two studies, a positive prescriptions were each used in one study (Table 2). Evidence-Based Complementary and Alternative Medicine 3

Table 1: Summary of experimental studies on the improvement of endometrial receptivity in herbal medicine.

Control Experimental Number Author Study design Outcome Results group group Pregnancy EID+TCMversusEID:↑ (𝑃 < 0.05) rate Implantation EID+TCMversusEID:↑ (𝑃 < 0.01) sites EID + TCM versus EID on Pd 4, 6: ↑ Normal (𝑃 < 0.01) In vivo: KM (𝑛=20) EID + TCM (1) Choi et al. [6] ∗ EID + TCM versus EID on Pd 5 ↓ mice EID (𝑛=19) LIF mRNA 𝑃 < 0.05 𝑛=23 ( ) ( ) EID + TCM versus normal: not significant𝑃 ( > 0.05) EID + TCM versus Normal on Pd 6: not significant (𝑃 > 0.05) LIF protein EID versus EID + TCM: ↓ (𝑃 < 0.05) EID versus normal: ↓ (𝑃 < 0.05) PPARd, IL-11 TCM at doses of 1.82 g/kg, 3.64 g/kg mRNA versus progynova: ↑ (𝑃 < 0.05) EID + TCM, EID + Progynova versus EID: ↑ In gland (𝑃 < 0.05) EID + TCM versus COX-2 Progynova: ↑ (𝑃 > 0.05) not significant In stroma (𝑃 > 0.05) EID + TCM, EID + Progynova versus EID: ↑ (𝑃 < 0.05) EID + TCM In gland (0.91 g/Kg) EID + TCM at high Normal (𝑛=12) doses versus EID + (𝑛=12) PGI2 † EID + TCM Progynova: ↑ (𝑃 < 0.05) In vivo: KM EID (𝑛=12) (2) Xu et al. [10] (1.82 g/Kg) mice EID + EID + TCM versus EID: (𝑛=12) In stroma ↑ 𝑃 < 0.05 Progynova ( ) EID + TCM (𝑛=12) EID + TCM, EID + (3.64 g/Kg) In gland Progynova versus EID: ↑ 𝑛=12 ( ) (𝑃 < 0.05) MMP-9 EID + TCM, EID + In stroma Progynova versus EID: ↑ (𝑃 < 0.05) EID + TCM, EID + Progynova versus EID: ↑ (𝑃 < 0.05) In gland TIMP-3 EID + TCM at high dose versus EID + Progynova: ↑ (𝑃 < 0.05) not significant In stroma (𝑃 > 0.05) ‖ COH + BS§,COH+HX,COH+BH¶ versus COH: ↑ (𝑃 = 0.0004, <0.0001, Endometrial COH + TCM <0.0001) thickness (BS) (𝑛=10) COH + BH¶ vs COH + BS§: ↑ Normal COH + TCM (𝑃 = 0.02) 𝑛=10 § ‖ ¶ In vivo: KM ( ) (HX) COH + BS ,COH+HX,COH+BH (3) Cui et al. [14] ‡ 𝑛=10 Microvessel mice COH ( ) versus COH: ↑ (𝑃 = 0.0046, 0.0003, 𝑛=10 density ( ) COH + TCM 0.0004) 𝑛=10 ‖ (BH) ( ) COH + HX , COH + BH versus COH: Pinopodes in ↑ (𝑃 = 0.0011, 0.0009) the ‖ COH + HX ,COH+BHversusCOH epithelium +BS§: ↑ (𝑃 = 0.0374, 0.033) 4 Evidence-Based Complementary and Alternative Medicine

Table 1: Continued. Control Experimental Number Author Study design Outcome Results group group Pregnancy EID + TCM versus EID: ↑ rate (𝑃 < 0.05) Number of EID + TCM versus EID: ↑ implanted (𝑃 < 0.01) blastocysts EID + TCM versus EID: ↓ (𝑃 < 0.01) Proliferative In gland EID + TCM versus index Normal normal: not significant In vivo: KM (𝑛=30) EID + TCM 𝑃 > 0.05 (4) Huang et al. ∗∗ ( ) mice EID (𝑛=30) [11] Proliferative not significant (𝑛=30) In stroma index (𝑃 > 0.05) EID + TCM versus EID: ↑ (𝑃 < 0.01) Apoptotic In gland EID + TCM versus index normal: not significant (𝑃 > 0.05) Apoptotic not significant In stroma index (𝑃 > 0.05) COH + TCM versus COH on Pd 2, 4: ↓ Integrin 𝛽3 (𝑃 < 0.01) Normal mRNA Among 3 groups on Pd 0, 6, 8: not In vivo: KM (𝑛=5) COH + TCM significant𝑃 ( > 0.05) (5) Sun et al. [15] ‡ 𝑛=5 mice COH ( ) COH + TCM versus COH on Pd 4, 6, 𝑛=5 ( ) 8: ↓ (𝑃 < 0.01) OPN mRNA Among 3 groups on Pd 0, 2: not significant𝑃 ( > 0.05) Integrin 𝛽3 COH + TCM versus aspirin: ↑ Normal mRNA (𝑃 < 0.01) (𝑛=10) Chen et al. In vivo: KM ‡ COH + TCM LIF (6) COH + 𝑛=30 COH + TCM versus aspirin: ↑ [16] mice ( ) mRNA/𝛽- aspirin (𝑃 = 0.02) (𝑛=30) actin COH + TCM versus COH: ↑ (𝑃 < 0.01) Pregnancy EID+TCMversusEID:↑ (𝑃 < 0.01) rate TCM versus normal: not significant difference (𝑃 > 0.05) Normal TCM Implantation TCMversusEID:↑ (𝑃 < 0.05) (𝑛=18) (𝑛=21) ‡ sites EID+TCMversusEID:↑ (𝑃 > 0.05) In vivo: KM COH COH + TCM (7) ↑ 𝑃 < 0.05 Yu et al. [18] mice (𝑛=15) (𝑛=22) TCM versus COH, EID: ( , ∗∗ LIF proteins 𝑃 < 0.01 EID EID + TCM ) (𝑛=16) (𝑛=23) Integrin 𝛽3 COH + TCM, EID + TCM versus proteins COH, EID: ↑ (𝑃 < 0.01, 𝑃 < 0.05) TCM versus COH + TCM: ↑ LIF mRNA (𝑃 < 0.05) TCM versus EID + TCM: ↑ (𝑃 < 0.05) TCM versus COH + TCM: ↑ Integrin 𝛽3 (𝑃 < 0.05) mRNA TCM versus EID + TCM: ↑ (𝑃 < 0.05) Normal COH + TCM group showed significant In vivo: KM (𝑛=6) COH + TCM LCM-DE- changes in 23 proteins: 7 proteins (8) Li et al. [12] ∗ †† ‡‡ mice COH (𝑛=6) MS downregulated and 16 proteins (𝑛=6) upregulated§§ Evidence-Based Complementary and Alternative Medicine 5

Table 1: Continued. Control Experimental Number Author Study design Outcome Results group group EID + TCM versus EID: ↑ (𝑃 < 0.05) Pregnancy EID + TCM versus normal: ↓ rate (𝑃 < 0.05) EID + TCM versus EID: not significant Number of (𝑃 > 0.05) Implanted EID + TCM versus normal: ↓ embryos 𝑃 < 0.05 Normal ( ) In vivo: KM (𝑛=10) EID + TCM EID versus Normal on Pd 5, 6, 7: ↑ (9) Wu et al. [13] ∗ mice EID (𝑛=10) (𝑃 < 0.01) (𝑛=10) EID versus EID + TCM on Pd 5, 6, 7: ↑ NF-𝜅j (𝑃 < 0.01) EID + TCM versus normal during whole time: not significant (𝑃 > 0.05) EID + TCM versus normal: not significant𝑃 ( > 0.05) IFN-𝛾,IL-10 EID + TCM versus EID on Pd 5, 6, 7: ↑ (𝑃 < 0.01) PL-PP group showed significant In vitro: LIF mRNA increase dose-dependently (𝑃 < 0.05) Ishikawa cells Adhesion of PL-PP versus Control: ↑ (𝑃 < 0.05) (10) Choi et al. [9] JAr spheroids Normal No of In vivo: KM (𝑛=11) EID + PL-PP PL-PP versus RU486: ↑ (𝑃 < 0.05) ∗∗ implanted mice EID (𝑛=11) PL-PP versus normal: ↑ (𝑃 < 0.05) embryos (𝑛=11) Number of TCM + model versus model: ↑ implanted (𝑃 < 0.05) Normal blastocyst In vivo: KM (𝑛=60) TCM 𝛼] ↓ (11) Li et al. [17] ‡ 𝑛=60 Integrin Model versus TCM, control: mice COH ( ) mRNA (𝑃 < 0.05) (𝑛=60) Integrin 𝛽3 Model versus TCM, control: ↓ mRNA (𝑃 < 0.05) ∗ † ‡ Indomethacin; Hydroxyurea + Mifepristone; Pregnant mare serum gonadotropin (PMSG) + human chorionic gonadotropin (hCG); §BS: Bushen ‖ (Semen Cuscutae, Herba Taxilli Chinensis, Radix Dipsaci); HX: Huoxue (Radix Astragali, Radix Angelicae Sinensis, Radix Salviae); ¶BH: Bu-Shen-An-Tai; ∗∗ †† ‡‡ Mifepristone; Laser capture microdissection-double dimensional electrophoresis-mass spectrum; collagen 훼-1 (VI) chain, keratin 7, keratin 14, myosin regulatory light chain 12B, myosin light polypeptide 9, heat shock protein 훽-7, and C-U-editing enzyme APOBEC-2; §§apolipoprotein A-I, calcium regulated protein-3, proliferating cell nuclear antigen, L-xylulose reductase, and calcium binding protein.

Eight out of ten studies using combination prescriptions 4. Discussion described the Oriental Medicinal efficacy of the prescription. Among the efficacy of prescription, invigorating or tonifying Traditional Korean Medicine has historical evidence accumu- the kidney [6, 10–14, 16, 17] was the most common, followed lated over thousands of years. In the recent years, efforts are by replenishing or supplement Qi [11–13, 17], nourishing under way to modernize and reconstruct individual experi- blood [6, 13, 17], and activating or promoting blood [6, 11, 14]. ential knowledge and to improve the quality of Traditional Korean Medicine. The modernization of Traditional Korean 3.6.2. Herbal Medicinal Materials. Twenty-two herbal medic- Medicine means converting it into a reasonable and empirical inal materials were used in 11 studies. The most frequently system that fits the evidence-based medicine model, combin- used herbal medicinal material was Astragali Radix, followed ing it with modern technology, modern academic thinking, by Dipsaci Radix, Cuscutae Semen, and Angelicae Gigantis and modern scientific culture [1, 2]. This may include exper- Radix (Figure 2). imental studies, observational studies, and clinical studies By oriental herbal efficacy [19], the highest rate was to determine the efficacy, effect, and mechanism of herbal for the tonifying and replenishing medicinal effects (68%). medicine [3]. Among the tonifying and replenishing medicinals, five yang- Infertility in a couple is defined as the failure to conceive tonifying medicinals, four yin-tonifying medicinals and after an arbitrary period of 12 months without the use of blood-tonifying medicinals, and three qi-tonifying medici- contraception. The infertility rate in Korea is estimated to be nals were identified (Figure 3). more than 13%, increasing every year [5], and there are an 6 Evidence-Based Complementary and Alternative Medicine

Table 2: Composition of prescription.

No Prescription Medicinal materials Cuscutae Semen, Loranthi Ramulus et Folium, Dipsaci Radix, Salviae Miltiorrhizae Radix, (1) Bushenantai [6] Astragali Radix, Angelicae Gigantis Radix (2) PL-PP [9] Polysaccharides depleted-water extract of Paeonia lactiflora Pallas Rehmanniae Radix, Ligustici Fructus, Ecliptae Herba, Dipsaci Radix, Cuscutae Semen, (3) Xianziyizhen [10] Epimedii Herba Loranthi Ramulus et Folium, Salviae Miltiorrhizae Radix, Astragali Radix, Angelicae Gigantis (4) Bushenyiqihexue [11] Radix, Cnidium officinale Makino (5) DS-1-47 [12] Astragali Radix, Atractylodes macrocephala Koidzumi, Scutellariae Radix, Dipsacales Dipsaci Radix, Astragali Radix, Salviae Miltiorrhizae Radix, Scutellariae Radix, Atractylodes (6) Bangdeyun [13] macrocephala Koidzumi Cuscutae Semen, Loranthi Ramulus et Folium, Dipsaci Radix, Astragali Radix, Angelicae (7) BuShenAnTai [14] Gigantis Radix, Salviae Miltiorrhizae Radix Lycium chinense Miller, Ligustici Fructus, Ecliptae Herba, Cuscutae Semen, Angelicae Gigantis (8) Er’zhi Tiangui [15] Radix, Paeonia lactiflora Pallas, Cnidium officinale Makino, Rehmanniae Radix, Cyperi Rhizoma, Glycyrrhiza uralensis Fischer and so forth Astragali Radix, Loranthi Ramulus et Folium, Dipsaci Radix, Cnidium officinale Makino, (9) Shoutaiwai [16] Atractylodes macrocephala Koidzumi, Adenophorae Radix Astragali Radix, Adenophorae Radix, Atractylodes macrocephala Koidzumi, Eucommiae Yiqixue Buganshen (10) Cortex, Glycyrrhiza uralensis Fischer, Rehmanniae Radix, Cinnamomi Cortex, Angelicae [17] Gigantis Radix, Paeonia lactiflora Pallas, Cnidium officinale Makino (11) Zhuyun [18] Epimedii Herba, Morindae Radix, Cuscutae Semen, Eucommiae Cortex

8 7 6 5 4 3 2 1 0 Miller Dipsaci RadixDipsaci Ecliptae Herba Ecliptae Astragali Radix Astragali Epimedii Herba Epimedii Morindae Radix Morindae Cyperi Rhizoma Cuscutae SemenCuscutae Ligustici Fructus Scutellariae RadixScutellariae Rehmanniae RadixRehmanniae Eucommiae Cortex Eucommiae Cinnamomi Cortex Cinnamomi Adenophorae Radix Adenophorae Lycium chineseLycium Paeonia lactiﬂora Pallas lactiﬂora Paeonia Angelicae Gigantis Radix Gigantis Angelicae Cnidium officinale Makino officinale Cnidium Salviae Miltiorrhizae Radix Miltiorrhizae Salviae Loranthi Ramulus et Folium Ramulus Loranthi Glycyrrhiza Fischer uralensis Atractylodes macrocephala Koidzumi macrocephala Atractylodes

Figure 2: Kinds and Frequencies of Medicinal Materials. Among the 11 studies, the most used herb was Astragali Radix which was included in 7 studies, followed by Dipsaci Radix included in 6 studies, Cuscutae Semen and Angelicae Gigantis Radix included in 5 studies, Loranthi Ramulus et Folium, Salviae Miltiorrhizae Radix, Atractylodes macrocephala Koidzumi, and Cnicium officinale Makino included in 4 studies, and Rehmanniae Radix and Paeonia lactiflora Pallas included in 3 studies. increasing number of patients seeking clinical evaluation and hormones at doses higher than the physiological dose reduces treatment for infertility [10]. hormone levels leading to decreases in endometrial receptiv- Multiple treatment approaches are available for infertility, ity and perhaps also pregnancy outcomes [6, 15, 20–23]. and assisted reproductive technology has become an impor- In Korean Medicine clinics, herbal medicines have been tant choice for infertile couple [10]. COH is a key determinant widely used to prevent miscarriage and increase the implan- of the success of IVF-ET. However, the use of exogenous tation rate. Korean Medicine as an adjunctive therapy for Evidence-Based Complementary and Alternative Medicine 7

Wind-dampness dispelling medicinal 4.5% (1)

Qi-regulating medicinal 4.5% (1) Interior warming medicinal 4.5% (1)

Dampness-draining diuretic medicinal 4.5% (1)

Heat-clearing medicinal 4.5% (1)

Blood-activating and stasis-resolving medicinal 9% (2)

Tonifying and replenishing medicinal 68% (15)

Figure 3: Classified by Single Medicinal Materials Type. According to oriental herbal efficacy classification criteria, the herb medicinal belonging to tonifying and replenishing medicinal was the most common with 15 species. It is followed by blood activating and stasis- resolving medicinal, heat-clearing medicinal, dampness-draining diuretic medicinal, interior-warning medicinal, qi-regulating medicinal, and wind-dampness dispelling medicinal.

assisted reproductive technology is effective in terms of medicine is widely used in clinical practice, and it should be induction of ovulation and development and maintenance backed up by experimental studies. of endometrium. In traditional Korean Medicine, pattern The herbal medicine and herbal medicinals which have identification of infertility is divided into kidney deficiency, tonifying effect were most commonly used in experimental liver depression, dampness-phlegm, blood deficiency, and studies, which is similar to herbal medicine used in the blood stasis. Clinically, the most common cases of infertility Korean Medicine clinic. These herbal medicines may improve were in the category of kidney deficiency [24]. Recently, endometrial receptivity by reinforcing the expression of according to the menstrual cycle, treatment to help the endometrial LIF and integrin 𝛽.Amongoutcomemeasures, recovery of the ovulation cycle is performed in the follicular integrin 𝛽3 is a specific molecule for evaluating uterine phase, and supplement with function of the luteum to receptivity during implantation, primarily expressed in the increase the rate of implantation is performed in the luteal cytoplasmofendometrialglandularepithelialcellsduringthe phase [25]. The treatment for enhancing the function of implantation window [15, 16]. Integrin is an important cell theluteumisprescribedbasedonthetonifyingkidney, adhesion molecule that can recognize extracellular matrix such as Sutaehwan [24, 26]. However, there is no review of proteins based on the integrin arginine-glycine-aspartic acid the experimental studies on how certain herbal medicine sequence, which mediates cell adhesion to the extracellu- improves implantation. lar matrix, promotes production of angiogenesis factors, Therefore as part of the study on modernization of Tra- mediates intracellular and extracellular transduction, and ditional Korean Medicine, we analyzed studies that clarified increases the blood supply to the endometrium, thereby the effects of herbal medicines used to improve endometrial improving endometrial receptivity [27, 28]. LIF has been receptivity and promote pregnancy. Studies on implantation identified as a key mediator in promoting endometrial trans- are conducted on animals for ethical reasons, so we investi- formation. It can induce COX, enhance vascular permeability gated only experimental studies. and angiogenesis, and reconstruct endometrial blood vessel Through analyzing the publication year of 11 selected [10]. Proper angiogenesis is fundamental to implantation, and papers, the number of experimental studies on improvement pregnancy will not occur when the endometrial thickness of endometrial receptivity in herbal medicine tended to is below a particular thickness [29]. Pinopodes are classi- increase slightly, but most studies were published in China. In cal biomarkers of endometrial receptivity [30]. Nikas and Korea,forthepurposeofpromotingtheimplantation,herbal Makrigiannakis reported that the coverage of pinopodes on 8 Evidence-Based Complementary and Alternative Medicine the endometrial epithelium during the implantation window Given that there is only domestic Korean study, evaluating of normal fertile women is more than 50%, while women the effectiveness of single herbal medicinal materials and with a lower coverage of pinopodes, particularly those with studies on the prescription of clinical practice should begin. acoverage<10%, tend to experience multiple implantation The safety and efficacy of combination of traditional Korean failures [31]. Proteomics is very important for understanding Medicine and Western Medicine for infertility should be the action mechanism of traditional Chinese medicine. Li et studied. Observational studies on combination therapy for al. suggested that these regulated proteins were important infertility treatment are also needed. in regulating the uterine environment for the blastocyst implantation [12]. Conflicts of Interest After reviewing these studies, we confirmed that herbal medicines have positive effects on factors associated with The authors declare that there are no conflicts of interest endometrial receptivity, leading to increase in endometrial regarding the publication of this paper. receptivity and perhaps pregnancy outcomes. Therefore it is recommended that women who plan to become pregnant Authors’ Contributions should be given treatment to improve the endometrial receptivity and to improve the implantation environment with Jang-Kyung Park and Dong-Il Kim were responsible for the the recovery of the ovulation cycle. In particular, when the study concept and design. Jang-Kyung Park and Dong-Il Kim endometrium is thin or repeatedly failed to be implantd, it is searched and selected studies for inclusion and performed the thought that Korean Medicine will be more necessary. data analysis and interpretation. Jang-Kyung Park wrote the By confirming the enhancement of endometrial receptiv- paper. Dong-Il Kim revised the manuscript. All authors were ity of herbal medicine in the COH model, we confirmed the involved in the entire study process. use of herbal medicine to increase implantation rates by the adjunctive use of assisted reproductive technology. In partic- Acknowledgments ular, in two studies, herbal medicines have been shown to have a comparable effect compared to progynova and aspirin, ThisstudywassupportedbyagrantoftheTraditional which are used to promote implantation. As is known, the Korean Medicine R&D Project, Ministry of Health & Welfare, window of implantation moves forward during the con- Republic of Korea (HB16C0062). trolled ovarian hyperstimulation, which negatively affects the implantation. If the Korean Medicine treatment which References improves endometrial receptivity is performed together with COH, it may contribute to the increase of implantation rate. [1] C. Y. Lee, “Discussion on the issues of the modernization of the fundamental theories and terms in korean medicine,” Korean However, these studies are limited to experimental J. Oriental Physiology & Pathology,vol.27,no.5,pp.540–552, research, and most were published in China. It is necessary 2013. to investigate the current use of prescriptions to enhance [2] W.-F. Li, J.-G. Jiang, and J. Chen, “Chinese Medicine and Its implantation in Korean Medicine clinics and to conduct Modernization Demands,” Archives of Medical Research,vol.39, experimental studies on the effects of those herbal medicines no. 2, pp. 246–251, 2008. on endometrial receptivity. Reviews and experimental studies [3] Q. Xu, R. Bauer, B. M. Hendry et al., “The quest for moderni- are needed as well to inform safety and efficacy in combina- sation of traditional Chinese medicine,” BMC Complementary tionwithWesternMedicineprescribedinassistedreproduc- and Alternative Medicine, vol. 13, article 132, 2013. tive technology. Furthermore, observational studies on the [4]S.H.Kim,Research on risk factors and diseases associated with combined treatment effects of traditional Korean Medicine infertility in Korea, Seoul National University RDB Foundation andassistedreproductivetechnologyinthetreatmentof Corporation, 2014. infertility are needed. [5] D. K. Gardner, A. Weissman, C. M. Howles et al., Textbook of Assisted Reproductive Techniques,Taylor&FrancisGroup, 5. Conclusions London, UK, 2nd edition, 2004. [6] M. Choi, D. Lee, and D. Kim, “A survey study on use condition From 2008 to 2016, experimental studies on herbal medicine of Korean Medical Institution and demand of Korean Medical on endometrial receptivity yielded the following conclusions. Treatment Project of infertility in infertile couple,” The Journal Experimental studies of the effects of herbal medicine on of Oriental Obstetrics & Gynecology,vol.26,no.2,pp.151–165, 2013. endometrial receptivity are gradually increasing. The experi- [7]M.Zhang,Y.Huang,G.Zhu,G.Huang,L.Dong,andJ.Zhang, mental studies on herbal medicine on endometrial receptivity “Effect of Bushenantai recipe on the expression of endometrial evaluate outcomes by two types of experimental models. LIF in mice with embryonic implantation dysfunction,” Journal The effects of herbal medicine used in experiments were of Huazhong University of Science and Technology (Medical invigorating kidney, replenishing Qi, nourishing blood, or Sciences),vol.28,no.1,pp.65–68,2008. activating blood. The most frequently used herbal medicinal [8] H. S. Song, “Lessons from mouse models to understand uterine materials were Astragali Radix and tonifying and replen- receptivity for embryo implantation,” KoreanJournalofFertility ishing medicinals. Herbal medicines appeared to improve and Sterility,vol.33,no.3,pp.139–147,2006. endometrial receptivity in the EID or COH model and [9] H.-J. Choi, T.-W. Chung, M.-J. Park et al., “Paeonia lactiflora showed efficacy similar to progynova and aspirin. enhances the adhesion of trophoblast to the endometrium via Evidence-Based Complementary and Alternative Medicine 9

induction of leukemia inhibitory factor expression,” PLoS ONE, ovarian hyperstimulation therapy and their importance,” Chi- vol.11,articlee0148232,no.2,2016. nese Journal of Practical Gynecology and Obstetrics,vol.24,no. [10] M. Xu, Y.-Z. Tian, X.-J. Zhu et al., “Effect of Xianziyizhen 5,pp.383-384,2008. recipe capsule on PGI2-PPAR𝛿 signaling pathway in embryo [24]C.Choi,S.Kim,M.Song,D.Hwang,H.Cho,andD.Kim, implantation dysfunction mice,” AmericanJournalofReproduc- “Analysis of Korean traditional medicinal treatment trend of tive Immunology,vol.73,no.6,pp.545–556,2015. infertility and development of questionnaire for infertility [11] D.-M. Huang, G.-Y.Huang, and F.-E.Lu, “Effects of Bushenyiqi- treatment,” TheJournalofOrientalObstetrics&Gynecology,vol. hexue formula on the endometrial gland apoptosis in mice with 29,no.2,pp.29–46,2016. blastocyst implantation dysfunction,” Journal of Traditional [25]I.Ahn,D.Kim,M.Choi,S.Jang,andJ.Jeong,“AStudyon Chinese Medicine, vol. 30, no. 3, pp. 196–200, 2010. Factors Influencing Pregnancy in the Pilot Project for Korean [12]Y.L.Li,X.Y.Zhang,Y.Lengetal.,“Globalproteinexpression Medical Treatment for Subfertility,” The Journal of Oriental analysis of molecular markers of DS-1-47, a component of Obstetrics & Gynecology,vol.26,no.3,pp.73–84,2013. implantation-promoting traditional chinese medicine,” Journal [26] D. I. Kim, “Oriental medical treatment as supplementary treat- of Huazhong University of Science and Technology,vol.36,no.6, ment method in ART: Clinical review,” The Journal of Oriental pp. 910–915, 2016. Obstetrics & Gynecology,vol.23,pp.67–79,2010. [13] Y. Wu, C. Zheng, L. Hu, and J. Li, “Local immune regulatory [27] P.-C.Chan, S.-Y.Chen, C.-H. Chen, and H.-C. Chen, “Crosstalk effects of Bangdeyun on the endometrium of mice with embryo between hepatocyte growth factor and integrin signaling path- implantation dysfunction during the implantation time,” Jour- ways,” Journal of Biomedical Science,vol.13,no.2,pp.215–223, nalofHuazhongUniversityofScienceandTechnology(Medical 2006. Sciences),vol.29,no.3,pp.372–376,2009. [28] Y. Zhang, J. H. Wang, and X. Liu, “Arg-Gly-Asp-motif toxins [14] D.-D. Cui, C.-H. Zheng, P. Gong et al., “Effects of Bu-Shen-An- using integrin as a target: research advances,” International Tai recipe and its two components on endometrial morphology Journal of Pharmaceutical Research,vol.36,no.6,pp.406–411, during peri-implantation in superovulated mice,” Journal of 2009. Huazhong University of Science and Technology (Medical Sci- [29]N.Noyes,H.-C.Liu,K.Sultan,G.Schattman,andZ.Rosen- ences), vol. 34, no. 5, pp. 768–774, 2014. waks, “Endometrial thickness appears to be a significant factor [15] Z.-G. Sun, F. Lian, Q. Jia et al., “Effects of Er’zhi tiangui in embryo implantation in in-vitro fertilization,” Human Repro- granule on sequential expressions of integrin 𝛽 3 and its ligand duction,vol.10,no.4,pp.919–922,1995. osteopontininmouseendometriumduringcontrolledovarian [30] U. Bentin-Ley, A. Sjogren,¨ L. Nilsson, L. Hamberger, J. F.Larsen, hyperstimulation,” ChineseJournalofIntegrativeMedicine,vol. and T. Horn, “Presence of uterine pinopodes at the embryo- 18,no.11,pp.846–849,2012. endometrial interface during human implantation in vitro,” [16]X.Y.Chen,J.Chen,Z.Y.Wang,X.H.Yu,B.X.Wei,andX.H. Human Reproduction,vol.14,no.2,pp.515–520,1999. Wu, “Effects of modified Shoutaiwai recipe on integrin 𝛽3and [31] G. Nikas and A. Makrigiannakis, “Endometrial pinopodes and leukemia-inhibitory factor in endometrium of controlled ovar- uterine receptivity,” Annals of the New York Academy of Sciences, ian hyperstimulation mice during the implantation window,” vol. 997, pp. 120–123, 2003. Genetics and Molecular Research,vol.14,no.2,pp.2970–2977, 2015. [17] H.-X. Li, X.-Y. Guo, Y. Xie et al., “Yiqixue buganshen recipe regulates the expression of integrin 𝛼]𝛽3 in the endometrium of controlled ovarian hyperstimulation mice,” Chinese Journal of Integrative Medicine,vol.19,no.10,pp.755–760,2013. [18] N. Yu, J. Yang, and T. Yin, “Extracts from a traditional Chinese herbal remedy (Zhuyun recipe) improve endometrial receptivity in mice with embryonic implantation dysfunction and ovulation stimulation,” Journal of Ethnopharmacology,vol.137, no. 1, pp. 389–395, 2011. [19] B. I. Seo, D. Y. Kwon, H. Y. Choi et al., “Medical Herbalogy,” Seoul: Younglim-Sa,pp.10–19,2012. [20] G. S. Basir, O. Wai-sum, E. H. Y. Ng, and P. C. Ho, “Morpho- metric analysis of peri-implantation endometrium in patients having excessively high oestradiol concentrations after ovarian stimulation,” Human Reproduction,vol.16,no.3,pp.435–440, 2001. [21] C. Bourgain and P. Devroey, “The endometrium in stimulated cycles for IVF,” Human Reproduction Update,vol.9,no.6,pp. 515–522, 2003. [22] G. Ertzeid and R. Storeng, “The impact of ovarian stimulation on implantation and fetal development in mice,” Human Repro- duction, vol. 16, no. 2, pp. 221–225, 2001. [23] Z. Z. Sun, Y. K. Ma, Z. R. Wang et al., “Expressions of integrinß3 and epidermal growth factor receptor in the endometrium during implantation window phase after different controlled Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 7236436, 15 pages https://doi.org/10.1155/2017/7236436

Research Article Systems Pharmacological Approach to the Effect of Bulsu-san Promoting Parturition

Su Yeon Suh and Won G. An

Department of Pharmacology, School of Korean Medicine, Pusan National University, Yangsan, Gyeongnam 50612, Republic of Korea

Correspondence should be addressed to Won G. An; [email protected]

Received 28 July 2017; Accepted 25 September 2017; Published 29 October 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Su Yeon Suh and Won G. An. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Bulsu-san (BSS) has been commonly used in oriental medicine for pregnant women in East Asia. The purpose of this research was to elucidate the effect of BSS on ease of parturition using a systems-level in silico analytic approach. Research results show that BSSis highly connected to the parturition related pathways, biological processes, and organs. There were numerous interactions between most compounds of BSS and multiple target genes, and this was confirmed using herb-compound-target network, target-pathway network, and gene ontology analysis. Furthermore, the mRNA expression of relevant target genes of BSS was elevated significantly in related organ tissues, such as those of the uterus, placenta, fetus, hypothalamus, and pituitary gland. This study used a network analytical approach to demonstrate that Bulsu-san (BSS) is closely related to the parturition related pathways, biological processes, and organs. It is meaningful that this systems-level network analysis result strengthens the basis of clinical applications of BSS on ease of parturition.

1. Introduction management of uterine smooth muscle contractions [8]. Although BSS has therapeutic effects on various pathological The name of Bulsu-san (BSS) originated from its therapeutic symptoms in pregnant or childbearing aged women, this effects that help to promote easy labor as if being touched research focused on the molecular mechanisms and impact by merciful Buddha’s hand [1]. BSS is composed of Angelicae of BSS on easing parturition and the acceleration of labor. Sinensis Radix (Danggui, DG) and Cnidium officinale Makino In terms of parturition onset, numerous studies have (Cheongung, CG), which is one of the most commonly described the complex hormone interactions between estro- used herb pairs in Traditional Medicine of East Asia and gen, progesterone, oxytocin, corticosteroid, and prostaglan- the usual component ratio is 2 : 3 (CG : DG) or 1 : 1 [2]. din. Among these, corticotrophin releasing hormone (CRH) BSS is widely used in women’s medicine in East Asia; its isregardedasatriggerthatinitiatesthelabor[9].Theplacenta recognized therapeutic effects are as follows: removal of releases substantial amounts of CRH, which stimulates the impure blood, blood making, easy parturition, acceleration pituitary glands of both mother and fetus to secrete adreno- of labor, elimination of dead fetus or placenta, amelioration of corticotropin hormone [10]. This in turn induces the release pain, nourishing blood, and promoting blood circulation [3]. of estrogen precursor, which is converted into estrogen by What is more, recent experimental research on the CG- the placenta that induces smooth muscle contraction [10]. DG herb pair indicated that they affect the nourishment of Additionally, dilatation of cervical connective tissue and blood [4], activate blood circulation, and prevent blood stasis smooth muscle is induced by the following changes: a shift [5]. In addition, the CG-DG herb pair showed significant from progesterone to estrogen dominance, increased respon- inhibitory effects on the proliferation and protein synthesis of siveness to oxytocin via the upregulation of myometrial oxy- vascular smooth muscle cells [6]. It was suggested BSS could tocin receptor, increased prostaglandins synthesis in uterus, affect the activities of Akt kinase and eNOS by increasing increased myometrial gap junction formation, decreased 2+ intracellular Ca and reducing ROS levels [7] and regulate nitric oxide activity, and increased influx of calcium into menstruation and provide relief from pain by enabling the myocyte [11]. 2 Evidence-Based Complementary and Alternative Medicine

Angelicae Sinensis Radix Cnidium ocinale Makino

ADME Screening

transcription, DNA-templated positive regulation of transcription from RNA polymerase II promoter negative regulation of transcription from RNA polymerase II promoter chloride transport gamma-aminobutyric acid signaling pathway peptidyl-serine phosphorylation protein autophosphorylation regulation of cell proliferation synaptic transmission, cholinergic adenylate cyclase-inhibiting G-protein coupled acetylcholine receptor signaling pathway Active Compound phospholipase C-activating G-protein coupled acetylcholine receptor signaling pathway positive regulation of MAPK cascade positive regulation of sequence-specic DNA binding transcription factor activity regulation of blood pressure regulation of smooth muscle contraction adenylate cyclase-activating adrenergic receptor signaling pathway cyclooxygenase pathway

H-C-T Network GO Analysis

Target Fishing

-aminobutyric acid receptor subunit -1 GABRA1 -aminobutyric-acid receptor -2 subunit GABRA2 -aminobutyric-acid receptor -3 subunit GABRA3 -aminobutyric-acid receptor subunit -6 GABRA6 Glucocorticoid receptor NR3C1 Potential Target Glutamate receptor 2 GRIA2 Glycogen synthase kinase-3  GSK3B Heat shock protein HSP 90 HSP90AB1 Ig -1 chain C region IGHG1 T-P Network Leukotriene A-4 hydrolase LTA4H Target organ location map

Figure 1: The workflow: the network pharmacological approach of Bulsu-san (BSS), namely, active compounds screening, target fishing, network analysis, and relevant organ location mapping was performed in this study.

The hypothesis of this study was that BSS may promote database were based on Tanimoto coefficient formula [17] as the positive-feedback of hormone loops as well as a series follows: of myometrial and cervical changes to ease parturition and 𝐴×𝐵 𝐹 (𝐴,) 𝐵 = , safely accelerate labor. A network based in silico approach 𝐴2 +𝐵2 −𝐴×𝐵 (1) wasusedtoidentifytheeffectofBSSonparturitionrelated systems and the aim of this study was to elucidate the effect of where 𝐴 represents the molecular parameters of herbal com- BSS on the parturition by system-level analysis. The workflow pounds and 𝐵 is the average molecular parameters of all com- of the network pharmacological study is summarized in pounds in the Drugbank database (http://www.drugbank.ca/) Figure 1. [18]. In the present study, we excluded compounds with a DL of <0.08. Other previous researches of herbal formulas set a 2. Material and Methods higherthresholdintherangeof0.1to0.18.However,wefound out that most compounds of DG have low DL. In detail, only 2.1. Identification of Active Compounds. Compounds in CG 36 compounds of 125 in DG show higher or equal DL value and DG were identified using a phytochemical database than 0.08. For this reason, this study sets a lower threshold of that is the Traditional Chinese Medicine Systems Phar- DL than other previous researches to see the most potential macology (TCMSP, http://ibts.hkbu.edu.hk/LSP/tcmsp.php). targets of BSS. We applied parameters related to absorption, distribution, metabolism, and excretion (ADME), namely, human drug- 2.1.2. Oral Bioavailability (OB) Prediction. OB is defined as likeness (DL) [12], oral bioavailability (OB) [13], and Caco- the ratio of active compounds’ absorption into the systemic 2 permeability (Caco-2) to screen the Potential active com- circulation, which represents the convergence of the ADME pounds in BSS [14]. process [13]. OB values are dependent on drug dissolution in the gastrointestinal (GI) tract and hepatic and intestinal first- 2.1.1. Drug-Likeness Evaluation. DL helps filter “drug-like” pass metabolism, as well as on intestinal membrane perme- compounds in oriental herbs, as DL represents a qualitative ation, which makes it a major pharmacokinetic parameter for concept for valuations based on how “drug-like” a prospec- drug evaluations [16]. In this study, the OB threshold was set tive compound is [15]. Accordingly, a high DL may lead to as ≥15%. a greater possibility of therapeutic success, and compounds withahigherDLvaluearemorelikelytopossesscertain 2.1.3. Caco-2 Permeability Screening. Caco-2 permeability is biological properties [16]. The calculations of DL in TCMSP used to predict the absorption of an orally administered Evidence-Based Complementary and Alternative Medicine 3 drug [14]. Surface absorptivity of the small intestine is network) to extract the pathways from KEGG database maximized with the presence of villi and microvilli, for this (http://www.genome.jp/kegg/), and the terms highly associ- reason most orally administered drug absorption occurs in ated with parturition with 𝑃 values < 0.05 were selected as the small intestine [19]. Moreover, the movement of orally the related pathways of targets in this work. Related targets administered drugs across the intestinal epithelial barrier were mapped onto relevant pathways, which resulted in the determines the rate and extent of human absorption and T-P network. Both networks were generated in Cytoscape ultimately affects drug bioavailability [20]. In the present 3.5.1, an open-source biological network visualization and study, compounds with OB, DL and Caco-2 values of greater data integration software package [27]. than 15%, 0.08, and >−0.4, respectively, were regarded as active compounds and subjected to further analysis. 2.5. Target Organ Location Map. Tissue-specific patterns of mRNA expression can indicate important associations with 2.1.4. Lipinski’s Rule (LR) Screening. In addition, the screen- biological events or gene functions [28]. To explore the ing standard used was defined based on Lipinski’s rule (LR), beneficial effects of BSS during parturition, it is important which identifies druggable compounds as having molecular that the tissue mRNA expression profiles of target genes weight (MW) of ≤500 Da (MW ≤ 500), chemical composition at the organ level be known [29]. The target organ loca- with ≤5 hydrogen-bond donors, ≤10 hydrogen-bond accep- tion map was built according to the Dataset: GeneAtlas tors, and an octanol-water partition coefficient, AlogP of ≤ U133A, gcrma (http://biogps.org). BioGPS database provides 5 [21]. AlogP can be used to estimate local hydrophobicity, expression data acquired by direct measurements of gene to produce molecular hydrophobicity maps, and to evaluate expression obtained by microarrays analysis [30]. First, the hydrophobic interactions in protein-ligand complexes [22]. mRNA expression patterns of each target gene in 176 parts HdonandHaccarethenumberofpossiblehydrogen-bond of organ tissues were obtained. Second, average values were donors and acceptors, and the hydrogen-bonding capacity of calculated for each gene. Third, frequency of above average a drug solute is recognized as a crucial determinant of perme- mRNA expression tissue organs was inspected. Forth, based ability; moreover high hydrogen-bonding potential is often on the result from the third step and parturition mechanism relatedtolowpermeabilityandabsorption[23].Eventually,in theory, mRNA expression data of relevant organ tissues the present study, we selected active compounds satisfying the were extracted and categorized into 6 groups, namely, uterus following criteria: OB ≥ 15%; DL ≥ 0.08; Caco-2 ≥−0.4; MW and/or uterus corpus, fetus and/or placenta, hypothalamus ≤ 500; H-bond donors ≤ 5; H-bond acceptors ≤ 10; AlogP ≤ 5. and/or pituitary, smooth muscle, and whole blood.

2.2. Target Fishing. Aside from filtering active compounds, 3. Results we also sought to identify the molecular targets of these active compounds. Compound-target interaction profiles 3.1. Identification of Active Compounds. 314 compounds of were established based on a systematic prediction of mul- BSSwereidentified,including189moleculesinCGand125in tiple drug-target interactions tool which employs random DG (as shown in Supplementary Material Table S1 in Supple- forest(RF)andsupportvectormachine(SVM)methods mentary Material available online at https://doi.org/10.1155/ and integrates chemical, genomic, and pharmacological 2017/7236436) and active compounds met the criteria OB information for drug targeting and discovery on a large ≥ 15%, Caco-2 ≥−0.4, and DL ≥ 0.08,aswellasthe scale [24]. Compound-target interactions satisfying SVM standards of Lipinski’s rule (LR) (as shown in Table 1). In score ≥ 0.8 and RF score ≥ 0.7wereselectedforfurther detail, 60 active compounds were initially chosen, but 8 com- study. Additionally, filtered compound-target interaction pounds were present in both herbs, namely, 3-butylidene- profile mapping was performed using the UniProt database 7-hydroxyphthalide, adenine, BdPh, beta-selinene, palmitic (http://www.uniprot.org/) [25]. acid, senkyunolide-C, senkyunolide-D, and senkyunolide- E, and 14 had no target protein information and were 2.3. Gene Ontology (GO) Analysis. Biological process (BP) thus excluded from the list of active compounds, whereas of gene ontology (GO) analysis was employed to determine 27 compounds with lower ADME properties than above the biological properties of target genes [26]. GO annotation thresholds were included, which were reported to be related indicates the possibility of direct statistical analysis on gene to oxytocin. In total, 65 active compounds were filtered. function information. In this research, GO BP terms with Although ligustilide and ferulic acid have a DL of <0.08, 𝑃 values < 0.01 were employed and the data was collected both were included in this study. Since ligustilide (C12, DL = using the DAVID 6.8 Gene Functional Classification Tool 0.07, OB = 53.72, Caco-2 = 1.3) was reported to be the main (http://david.abcc.ncifcrf.gov/). compound of DG in uterine contraction [31], and ferulic acid (C42,DL=0.06,OB=54.97,Caco-2=0.53)hasbeenreported 2.4. Network Construction and Analysis. In order to under- to be useful for the treatment of vascular diseases [6, 32] and stand the multiscale interactions between the active com- blood deficiency syndrome [33] in China and to suppress pounds of BSS and targets, two types of networks were inflammatory responses and tumor progression [34]. Some built: (1) the herb-compound-target network (H-C-T net- other compounds also have been shown experimentally to work), in which nodes represent either compounds, tar- have various biological activities; for example, crysophanol get genes, or herbs and edges indicate herb-compound- (C42, DL = 0.21, OB = 18.64, Caco-2 = 0.62) can be used target connections; and (2) the target-pathway network (T-P to treat menorrhagia and thrombocytopenia [35]. Perlolyrine 4 Evidence-Based Complementary and Alternative Medicine

Table 1: 65 Potential active compounds of BSS (compound with ∗ was present in both herbs).

ID Active compounds OB (%) Caco-2 DL Herb C1 ()-alpha-Terpineol 46.3 1.28 0.03 DG C2 ()-Aromadendrene 55.74 1.81 0.1 CG C3 ()-Terpinen-4-ol 81.41 1.36 0.03 CG C4 (+)-alpha-Funebrene 52.87 1.79 0.1 CG C5 (+)-Ledol 16.96 1.43 0.12 DG C6 (1R,5R,7S)-4,7-Dimethyl-7-(4-methylpent-3-enyl)bicyclo[3.1.1]hept-3-ene 16.23 1.86 0.09 CG (1S,4aR,8aR)-1-Isopropyl-7-methyl-4-methylene-2,3,4a,5,6,8a-hexahydro-1H- C7 19.8 1.86 0.08 DG naphthalene C8 (1S,4E,8E,10R)-4,8,11,11-tetramethylbicyclo[8.1.0]undeca-4,8-diene 21.69 1.86 0.08 CG C9 (3E)-3-butylidene-7-hydroxy-2-benzofuran-1-one 42.17 1.03 0.08 DG C10 (L)-alpha-Terpineol 48.8 1.39 0.03 CG C11 (R)-Linalool 39.8 1.33 0.02 CG C12 (Z)-Ligustilide 53.72 1.3 0.07 CG C13 1-Acetyl-beta-carboline 67.12 1.18 0.13 CG C14 1-beta-Ethylacrylate-7-aldehyde-beta-carboline 28.53 0.45 0.31 CG 1H-Cycloprop(e)azulen-7-ol, decahydro-1,1,7-trimethyl-4-methylene-, C15 82.33 1.37 0.12 CG (1aR-(1aalpha,4aalpha,7beta,7abeta,7balpha))- C16 1-Terpineol 49.83 1.24 0.03 CG C17 2,6-Di(phenyl)thiopyran-4-thione 69.13 1.74 0.15 DG C18 2-[(2S,5S,6S)-6,10-Dimethylspiro[4.5]dec-9-en-2-yl]propan-2-ol 37.62 1.44 0.09 CG ∗ C19 3-Butylidene-7-hydroxyphthalide 62.68 1 0.08 CG&DG C20 4,7-Dihydroxy-3-butylphthalide 106.09 0.69 0.1 CG C21 49070 FLUKA 85.51 1.29 0.12 CG C22 4-Hydroxy-3-butylphthalide 70.31 0.9 0.08 CG C23 58870 FLUKA 49.01 1.82 0.1 CG ∗ C24 Adenine 62.81 −0.3 0.03 CG&DG C25 ADO 15.98 −1.56 0.18 CG C26 alpha-Cubebene 16.73 1.83 0.11 CG C27 alpha-Selinene 31.81 1.82 0.1 CG C28 Aromadendrene oxide 2 65.1 1.56 0.14 CG ∗ C29 BdPh 42.44 1.32 0.07 CG&DG C30 beta-Chamigrene 31.99 1.82 0.08 DG ∗ C31 beta-Selinene 24.39 1.83 0.08 CG&DG C32 beta-Cubebene 32.16 1.82 0.11 CG C33 Cadinene 17.12 1.88 0.08 DG C34 Caffeic acid 25.76 0.21 0.05 CG C35 Carotol 149.03 1.46 0.09 CG C36 Cedrene 51.14 1.82 0.11 CG C37 Chuanxiongol 22.19 0.94 0.1 CG C38 cis-Thujopsene 56.43 1.84 0.12 DG C39 Coniferyl ferulate 4.54 0.71 0.39 DG C40 Crysophanol 18.64 0.62 0.21 CG C41 FA 68.96 −1.5 0.71 CG C42 Ferulic acid (CIS) 54.97 0.53 0.06 DG InChI=1/C15H24/c1-10-7-8-15-9-12(10)14(3,4)13(15)6-5-11(15)2/h7,11-13H,5-6,8- C43 55.56 1.79 0.1 DG 9H2,1-4H C44 L-Bornyl acetate 65.52 1.29 0.08 CG C45 Methyl palmitate 18.09 1.37 0.12 CG C46 Myricanone 40.6 0.67 0.51 CG C47 Nicotinic acid 47.65 0.34 0.02 DG Evidence-Based Complementary and Alternative Medicine 5

Table 1: Continued. ID Active compounds OB (%) Caco-2 DL Herb C48 Oleic acid 33.13 1.17 0.14 CG ∗ C49 Palmitic acid 19.3 1.09 0.1 CG&DG C50 Perlolyrine 65.95 0.88 0.27 CG C51 PLO 14.07 0.69 0.43 CG C52 Scopoletol 27.77 0.71 0.08 DG C53 Senkyunolide A 26.56 1.3 0.07 CG C54 Senkyunolide G 39.52 0.63 0.08 CG ∗ C55 Senkyunolide-C 46.8 0.87 0.08 CG&DG ∗ C56 Senkyunolide-D 79.13 0.12 0.1 CG&DG ∗ C57 Senkyunolide-E 34.4 0.55 0.08 CG&DG C58 Senkyunolide-K 61.75 0.52 0.08 CG C59 Sinapic acid 64.15 0.48 0.08 CG C60 Sphingomyelin 0.31 −0.46 0.51 DG C61 Stearic acid 17.83 1.15 0.14 CG C62 Stigmasterol 43.83 1.44 0.76 DG C63 Succinic acid 29.62 −0.44 0.01 DG C64 Sucrose 7.17 −2.89 0.23 CG C65 Wallichilide 42.31 0.82 0.71 CG

(C52, DL = 0.27, OB = 65.95, Caco-2 = 0.88) was confirmed related to parturition process as estrogen and progesterone to have a protective effect on injured human umbilical vein play important roles in pregnancy and parturition, and endothelial cells [36], and myricanone (C48, DL = 0.51, OB estrogen induceS the principal stimulatory myometrial con- = 57.61, Caco-2 = 0.67) was found to best inhibit mouse skin tractility [40]. Also, estradiol takes key place in parturition tumor progression [37]. process [41]. It was identified that increased ERK activation is observed at the onset of labor, and it promotes myometrial 3.2. Target Fishing. The 65 active compounds interact with contractility and development of parturition [42, 43]. To sum 185targetproteins,asshowninTable2;inotherwords, up, the target genes of BSS are highly associated with the on average, each compound on average interacts with 2.85 biological process (BP) of parturition. target proteins. This result confirms the polypharmacological character of oriental medicine and demonstrates the syn- 3.4. Network Construction and Analysis. Network analysis is ergistic effects of multiple compounds on multiple targets an efficient tool for visualizing and understanding multiple [38]. Different compounds in CG and DG can directly affect targeted drug actions and demonstrates drug actions within common targets, for example, the target protein “calmodulin the context of the whole genome [44, 45]. For a better insight (CALM1)” interacts with crysophanol from CG and coniferyl of therapeutic impacts, H-C-T and T-P networks were con- ferulate from DG at the same time, which implies the structedanddisplayedinFigures3and4,respectively.Inthe synergetic or cumulative effects of herbal medicine. H-C-T network, nodes represent herb names, compounds, and targets. Also in the T-P network, circular nodes represent 3.3. GO Analysis. 397 biological process terms with 𝑃 values targets and triangle nodes represent pathways. Besides node of <0.01 were sorted using the functional annotation chart of size is relative to the degree and edges show interactions the DAVID 6.8 Gene Functional Classification Tool, based between nodes. on 185 filtered target genes, and 𝑃 values were adjusted H-C-T network confirmed that there were 739 interac- using the Benjamini-Hochberg method. 30 enriched GO BP tions between 185 targets and 65 active compounds of CG terms extracted by 𝑃 value and gene counts are displayed in and DG: oleic acid (C48, degree = 42) with the highest Figure 2. It is meaningful that most of the target genes are number of interactions with targets, followed by succinic significantlyrelatedtothevariousBPinvolvedinparturition. acid (C63, degree = 40) and stigmasterol (C62, degree = For instance, 30 extracted GO BP terms include “MAPK 37). It shows that single molecules target multiple recep- signaling pathways,” “steroid hormone mediated signaling tors [46]. Also, some compounds from CG and DG were pathway,” “response to glucocorticoid,” “response to estra- found to share common targets. Likewise, prostaglandin diol,” and “positive regulation of ERK1 and ERK2 cascade.” G/H synthase 2 (PTGS2, degree = 56) displayed the most “MAPK signaling pathways” were reported to be activated affinitive connections with compounds, followed by gamma- in human uterine cervical ripening during parturition [39]. aminobutyric acid receptor subunit alpha-1 (GABRA1, degree “Steroid hormone mediated signaling pathway” is highly = 48), prostaglandin G/H synthase 1 (PTGS1, degree = 37), 6 Evidence-Based Complementary and Alternative Medicine

Table 2: Related targets of potential compounds in BSS.

UniProt ID Target name Gene Name P80404 4-aminobutyrate aminotransferase, mitochondrial ABAT P33121 Long-chain-fatty-acid--CoA ligase 1 ACSL1 O60488 Long-chain-fatty-acid--CoA ligase 4 ACSL4 P00813 Adenosine deaminase ADA P07327 Alcohol dehydrogenase 1A ADH1A P00325 Alcohol dehydrogenase 1B ADH1B P00326 Alcohol dehydrogenase 1C ADH1C P29274 Adenosine A2a receptor ADORA2A P35348 Alpha-1A adrenergic receptor ADRA1A P35368 Alpha-1B adrenergic receptor ADRA1B P25100 Alpha-1D adrenergic receptor ADRA1D P08913 Alpha-2A adrenergic receptor ADRA2A P18089 Alpha-2B adrenergic receptor ADRA2B P18825 Alpha-2C adrenergic receptor ADRA2C P08588 Beta-1 adrenergic receptor ADRB1 P07550 Beta-2 adrenergic receptor ADRB2 Q5SY84 Adenylosuccinate synthetase ADSS P21549 Serine--pyruvate aminotransferase AGXT O43865 Putative adenosylhomocysteinase 2 AHCYL1 P15121 Aldose reductase AKR1B1 P13716 Delta-aminolevulinic acid dehydratase ALAD P51649 Succinate semialdehyde dehydrogenase, mitochondrial ALDH5A1 P04745 Alpha-amylase 1 AMY1A P04746 Pancreatic alpha-amylase AMY2A P04114 Apolipoprotein B-100 APOB P10275 Androgen receptor AR P06576 ATP synthase subunit beta, mitochondrial ATP5B P06276 Cholinesterase BCHE P10415 Apoptosis regulator Bcl-2 BCL2 Q06187 Tyrosine-protein kinase BTK BTK P00915 Carbonic anhydrase I CA1 P62158 Calmodulin CALM1 P42574 Caspase-3 CASP3 P04040 Catalase CAT P06307 Cholecystokinin CCK P20248 Cyclin-A2 CCNA2 P30305 M-phase inducer phosphatase 2 CDC25B P24941 Cell division protein kinase 2 CDK2 P11597 Cholesteryl ester transfer protein CETP P28329 Choline O-acetyltransferase CHAT O14757 Serine/threonine-protein kinase Chk1 CHEK1 P36222 Chitinase-3-like protein 1 CHI3L1 P11229 Muscarinic acetylcholine receptor M1 CHRM1 P08172 Muscarinic acetylcholine receptor M2 CHRM2 P20309 Muscarinic acetylcholine receptor M3 CHRM3 Q15822 Neuronal acetylcholine receptor subunit alpha-2 CHRNA2 P36544 Neuronal acetylcholine receptor protein, alpha-7 chain CHRNA7 Q99966 Cbp/p300-interacting transactivator 1 CITED1 P02452 Collagen alpha-1(I) chain COL1A1 Q02388 Collagen alpha-1(VII) chain COL7A1 P17538 Chymotrypsinogen B CTRB1 P07339 Cathepsin D CTSD P04798 Cytochrome P450 1A1 CYP1A1 Evidence-Based Complementary and Alternative Medicine 7

Table 2: Continued. UniProt ID Target name Gene Name P05177 Cytochrome P450 1A2 CYP1A2 Q9ULA0 Aspartyl aminopeptidase DNPEP P27487 Dipeptidyl peptidase IV DPP4 P21728 Dopamine D1 receptor DRD1 P14416 D(2) dopamine receptor DRD2 P25101 Endothelin-1 EDNRA Q07075 Glutamyl aminopeptidase ENPEP Q6UWV6 Ectonucleotide pyrophosphatase/phosphodiesterase family member 7 ENPP7 P04626 Receptor tyrosine-protein kinase erbB-2 ERBB2 P03372 Estrogen receptor ESR1 Q92731 Estrogen receptor beta ESR2 P00742 Coagulation factor Xa F10 P00734 Thrombin F2 P08709 Coagulation factor VII F7 P07148 Fatty acid-binding protein, liver FABP1 P01100 Proto-oncogene c-Fos FOS P15408 Fos-related antigen 2 FOSL2 P35575 Glucose-6-phosphatase G6PC P14867 Gamma-aminobutyric acid receptor subunit alpha-1 GABRA1 P47869 Gamma-aminobutyric-acid receptor alpha-2 subunit GABRA2 P34903 Gamma-aminobutyric-acid receptor alpha-3 subunit GABRA3 P48169 Gamma-aminobutyric-acid receptor subunit alpha-4 GABRA4 P31644 Gamma-aminobutyric-acid receptor alpha-5 subunit GABRA5 Q16445 Gamma-aminobutyric-acid receptor subunit alpha-6 GABRA6 P17677 Neuromodulin GAP43 P47871 Glucagon GCGR P14136 Glial fibrillary acidic protein GFAP Q2TU84 Growth-inhibiting protein 18 GIG18 P23415 Glycine receptor alpha-1 chain GLRA1 P00367 Glutamate dehydrogenase 1, mitochondrial GLUD1 P17174 Aspartate aminotransferase, cytoplasmic GOT1 P00505 Aspartate aminotransferase, mitochondrial GOT2 P42262 Glutamate receptor 2 GRIA2 P49841 Glycogen synthase kinase-3 beta GSK3B Q15486 Putative beta-glucuronidase-like protein SMA3 GUSBP1 P19367 Hexokinase-1 HK1 P04035 3-hydroxy-3-methylglutaryl-coenzyme A reductase HMGCR P00738 Haptoglobin HP O14756 Oxidoreductase HSD17B6 P08238 Heat shock protein HSP 90 HSP90AB1 P28223 5-hydroxytryptamine 2A receptor HTR2A P01344 Insulin-like growth factor II IGF2 P01857 Ig gamma-1 chain C region IGHG1 P22301 Interleukin-10 IL10 P05231 Interleukin-6 IL6 P01308 Insulin INS Q12809 Potassium voltage-gated channel subfamily H member 2 KCNH2 Q12791 Calcium-activated potassium channel subunit alpha 1 KCNMA1 P35968 Vascular endothelial growth factor receptor 2 KDR P09848 Lactase-phlorizin hydrolase LCT Q32P28 Prolyl 3-hydroxylase 1 LEPRE1 Q8IVL6 Prolyl 3-hydroxylase 3 LEPREL2 P06858 Lipoprotein lipase LPL 8 Evidence-Based Complementary and Alternative Medicine

Table 2: Continued. UniProt ID Target name Gene Name P09960 Leukotriene A-4 hydrolase LTA4H P21397 Amine oxidase [flavin-containing] A MAOA P27338 Amine oxidase [flavin-containing] B MAOB Q16539 Mitogen-activated protein kinase 14 MAPK14 Q00266 S-adenosylmethionine synthetase isoform type-1 MAT1A P31153 S-adenosylmethionine synthetase isoform type-2 MAT2A P23368 NAD-dependent malic enzyme, mitochondrial ME2 Q16798 NADP-dependent malic enzyme, mitochondrial ME3 Q3SYC2 2-acylglycerol O-acyltransferase 2 MOGAT2 P05164 Myeloperoxidase MPO Q15788 Nuclear receptor coactivator 1 NCOA1 Q15596 Nuclear receptor coactivator 2 NCOA2 P29475 Nitric-oxide synthase, brain NOS1 P35228 Nitric oxide synthase, inducible NOS2 P29474 Nitric oxide synthase, endothelial NOS3 P16083 NRH dehydrogenase [quinone] 2 NQO2 Q14994 Nuclear receptor subfamily 1 group I member 3 NR1I3 P04150 Glucocorticoid receptor NR3C1 P08235 Mineralocorticoid receptor NR3C2 Q16620 BDNF/NT-3 growth factors receptor NTRK2 P04181 Ornithine aminotransferase, mitochondrial OAT P00480 Ornithine carbamoyltransferase, mitochondrial OTC Q9BYC2 Succinyl-CoA:3-ketoacid-coenzyme A transferase 2, mitochondrial OXCT2 O15460 Prolyl 4-hydroxylase subunit alpha-2 P4HA2 P49585 Choline-phosphate cytidylyltransferase A PCYT1A 󸀠 󸀠 Q14432 CGMP-inhibited 3 ,5 -cyclic phosphodiesterase A PDE3A O00330 Pyruvate dehydrogenase protein X component, mitochondrial PDHX P52945 Pancreas/duodenum homeobox protein 1 PDX1 P06401 Progesterone receptor PGR P48736 Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit, gamma isoform PIK3CG P11309 Proto-oncogene serine/threonine-protein kinase Pim-1 PIM1 P61925 cAMP-dependent protein kinase inhibitor alpha PKIA P14618 Pyruvate kinase isozymes M1/M2 PKM2 P04054 Phospholipase A2 PLA2G1B P00749 Urokinase-type plasminogen activator PLAU P00747 Plasminogen PLG P00491 Purine nucleoside phosphorylase PNP P27169 Serum paraoxonase/arylesterase 1 PON1 Q07869 Peroxisome proliferator-activated receptor alpha PPARA Q03181 Peroxisome proliferator-activated receptor delta PPARD P37231 Peroxisome proliferator activated receptor gamma PPARG Q9UBK2 Peroxisome proliferator-activated receptor gamma coactivator 1-alpha PPARGC1A P17612 mRNA of PKA Catalytic Subunit C-alpha PRKACA P05771 Protein kinase C beta type PRKCB P35030 Trypsin-3 PRSS3 P60484 Phosphatidylinositol-3,4,5-trisphosphate 3-phosphatase and dual-specificity protein phosphatase PTEN PTEN P43115 Prostaglandin E2 receptor EP3 subtype PTGER3 P23219 Prostaglandin G/H synthase 1 PTGS1 P35354 Prostaglandin G/H synthase 2 PTGS2 P18031 mRNA of Protein-tyrosine phosphatase, non-receptor type 1 PTPN1 P10082 Peptide YY PYY P63000 Ras-related C3 botulinum toxin substrate 1 RAC1 P50120 Retinol-binding protein 2 RBP2 P08100 Rhodopsin RHO Evidence-Based Complementary and Alternative Medicine 9

Table 2: Continued. UniProt ID Target name Gene Name P19793 Retinoic acid receptor RXR-alpha RXRA O00767 Acyl-CoA desaturase SCD Q14524 Sodium channel protein type 5 subunit alpha SCN5A P31040 Succinate dehydrogenase [ubiquinone] flavoprotein subunit, mitochondrial SDHA P16109 P-selectin SELP P05121 Plasminogen activator inhibitor 1 SERPINE1 P14410 Sucrase-isomaltase, intestinal SI O76082 Solute carrier family 22 member 5 SLC22A5 Q9UBX3 Mitochondrial dicarboxylate carrier SLC25A10 P11168 Solute carrier family 2, facilitated glucose transporter member 2 SLC2A2 P23975 Sodium-dependent noradrenaline transporter SLC6A2 Q01959 Sodium-dependent dopamine transporter SLC6A3 P31645 Sodium-dependent serotonin transporter SLC6A4 P35610 Sterol O-acyltransferase 1 SOAT1 P00441 Superoxide dismutase [Cu-Zn] SOD1 P08047 Transcription factor Sp1 SP1 P12931 Proto-oncogene tyrosine-protein kinase SRC SRC P36956 Sterol regulatory element-binding protein 1 SREBF1 Q12772 Sterol regulatory element-binding protein 2 SREBF2 Q9P2R7 Succinyl-CoA ligase [ADP-forming] beta-chain, mitochondrial SUCLA2 Q99973 Telomerase protein component 1 TEP1 P01375 Tumornecrosisfactor TNF Q16881 Thioredoxin reductase, cytoplasmic TXNRD1 P55851 Mitochondrial uncoupling protein 2 UCP2 P55916 Mitochondrial uncoupling protein 3 UCP3 and muscarinic acetylcholine receptor M1 (CHRM1, degree catalytic subunit alpha (PRKACA, degree = 20), protein = 37). Except for C60 (PLA2G1B, degree = 1), the rest of kinase C beta type (PRKCB, degree = 18), and calmodulin the64activecompoundsareconnectedwithmorethan (CALM1,degree=11).Wecanconfirmthesameresultinthe one target; likewise, 73 (39.5%) target genes out of 185 previous researches. For instance, PI3-kinase subunit gamma interacted with more than one compound. This result demon- plays the key role in regulating cAMP, calcium cycling, strates the multicompounds and multitarget properties of and beta-adrenergic signaling [49]. Moreover, during the herbal compounds and there was a report that compounds labor, calmodulin-calcium complex activates myosin light- with multiple targets could have greater therapeutic efficacy chain kinase, which causes the generation of ATPase activity; [47]. eventually, uterine contraction is promoted [50]. In addition, the top 40 pathways were extracted based H-C-T network explains the multitarget, multicom- on gene counts and 𝑃 value (<0.05), and 𝑃 value was pounds properties and accumulates effect of herbal adjusted by Benjamini-Hochberg method. T-P network using medicines and T-P network shows that target genes of BSS relevant targets of herbal compounds is demonstrated in are highly related to the pathway associated with parturi- Figure 4. There were 485 interactions between the top 40 tion process. pathways and 135 of 185 target genes. “Metabolic pathways” (degree = 49) and “neuroactive ligand-receptor interaction 3.5.TargetOrganLocationMap. It is important to confirm pathway” (degree = 32) had the highest and the second the tissue mRNA expression profiles of the target genes at highest numbers of connections with the targets, followed the organ level to identify the effects of BSS on parturi- by “calcium signaling” (degree = 21), “cAMP signaling path- tion. Since there was no mRNA expression information in way” (degree = 17), and “cGMP PKG signaling pathway” BioGPS of muscarinic acetylcholine receptor M1 (CHRM1), (degree = 15). These are compelling results that parturition putative beta-glucuronidase-like protein SMA3 (GUSBP1), processes are the complex hormone interactions and it is and retinol-binding protein 2 (RBP2), excluding these 3 well known that calcium signals within the myometrium are targets from 185 filtered targets, totally 182 genes mRNA pivotal for uterine contractions [48]. In the same manner, expression profiles were analyzed in this study. There were 519 some target genes demonstrated higher degree centrality interactions between target genes and organ locations. The with top 40 pathways, namely, PI3-kinase subunit gamma networks of target genes tissue mRNA expression profiles and (PIK3CG, degree = 23), cAMP-dependent protein kinase compounds of BSS are shown in Figure 5. 10 Evidence-Based Complementary and Alternative Medicine

GO Analysis

response to drug 34 23.3 response to hypoxia 20 13.6 positive regulation of transcription from RNA polymerase II promoter 28 5.3 aging 18 11.7 oxidation-reduction process 22 5.8 positive regulation of cell proliferation 20 6.2 cell proliferation 18 6.4 positive regulation of transcription, DNA-templated 19 5.0 response to glucocorticoid 12 10.1 cell-cell signaling 15 6.2 adenylate cyclase-activating adrenergic receptor signaling pathway 9 11.0 transcription initiation from RNA polymerase II promoter 12 6.2 response to estradiol 10 6.3 lipid metabolic process 11 5.1 response to lipopolysaccharide 11 5.0 positive regulation of ERK1 and ERK2 cascade 11 4.7 platelet activation 10 5.4 cholesterol metabolic process 9 6.2 response to cold 8 7.0 response to nicotine 8 7.0 positive regulation of MAPK cascade 9 5.6 positive regulation of protein kinase B signaling 9 5.5 response to hydrogen peroxide 8 6.0 response to ethanol 9 4.8 steroid hormone mediated signaling pathway 8 5.6 cellular response to lipopolysaccharide 9 4.6 glucose transport 7 5.9 response to nutrient 8 4.9 circadian rhythm 8 4.8 response to fatty acid 6 6.1 0 1020304050

Count − ＦＩＡ 10 (P-value)

Figure 2: GO analysis: 30 enriched biological process (BP) of gene ontology (GO) terms sorted by 𝑃 value < 0.01 and gene counts are displayed. The 𝑦-axis represents enriched biological process (BP) terms for the target genes, and the 𝑥-axis shows gene counts and − log 10 (𝑃 value).

As a result, 159 of 182 target genes displayed beyond aver- in all of 6 groups. For instance, muscarinic acetylcholine age mRNA expression in relevant organ tissues, such as uterus receptor M2 (CHRM2), neuronal acetylcholine receptor and/or uterus corpus, fetus and/or placenta, hypothalamus subunit 𝛼-2 (CHRNA2), gamma-aminobutyric acid receptor and/or pituitary, smooth muscle, and whole blood. The rest of subunit alpha-3 (GABRA3), NO synthase, inducible (NOS2), 󸀠 󸀠 23 genes of 182 targets did not display above average mRNA cGMP-inhibited 3 ,5 -cyclic phosphodiesterase A (PDE3A), expression in above organ tissues, for example, gamma- and sodium-dependent dopamine transporter (SLC6A3) aminobutyric acid receptor subunit alpha-6 (GABRA6) and recorded beyond average mRNA expression in all six groups. coagulation factor X (F10). Furthermore, 79% of targets were expressed in two or more Nevertheless, most genes of 159 demonstrated high organ tissues, which suggests that those organs and target expression patterns in several organs of parturition related genes of BSS are closely correlated. tissues at the same time. In detail, 60 genes showed most significant mRNA expression in the uterus and/or uterus cor- 4. Discussion pus group, 130 for placenta and/or fetus, 86 for hypothalamus and/or pituitary, 82 for smooth muscle, 80 for pituitary, and 81 In this study, network pharmacology method with DL, OB, for whole blood. Besides, 30 of 159 genes showed expression Caco-2, and LR evaluation, multiple drug-target prediction, Evidence-Based Complementary and Alternative Medicine 11

COL7A1 HK1 CHI3L1 NR1I3 AMY1A SI ACSL1 SLC25A10 FOS SREBF2 LCT FOSL2 SREBF1 ME2 ACSL4 ADSS BCHE G6PC NOS1 AMY2A ALAD ADA SUCLA2 GOT1 MAT2A SLC22A5 CDC25B GLUD1 TXNRD1 COL1A1 CYP1A2 AGXT PCYT1A GUSBP1 P4HA2 SRC C64 GIG18 MAT1A RHO PTEN NR3C1 ADORA2A ABAT ME3 OAT PTPN1 LEPREL2 LEPRE1 BCL2 SP1 HSD17B6 PTGER3 PKM2 OTC IL6 ENPP7 PNP SDHA OXCT2 CTSD GOT2 IL10 C25 C63 AKR1B1 C45 C60 GLRA1 ALDH5A1 PLA2G1B C24 AHCYL1 CASP3 C49 C65 ADH1B ADH1A PRSS3 C38 ATP5B PPARG DG C61 C23 C43 C15 APOB C32 PDHX C28 C21 UCP3 C26 C5 ADH1C CCK C56 GABRA2 C47 NR3C2 C2 DNPEP C8 C35 C58 C18 CHRNA2 SOAT1 LPL C36 GABRA1 SOD1 PPARA NCOA2 C27 C4 GABRA6 GAP43 HP MOGAT2 C3 PPARGC1A C51 PGR C54 TEP1 NOS3 IGHG1 C33 C31 C48 C53 CHRM3 UCP2 TNF PTGS2 CHRM2 PRKACA GABRA5 CAT RXRA C7 SCD C6 CHRM1 GABRA3 CETP CG C30 C16 INS C44 SLC2A2 FABP1 C50 PTGS1 C11 DRD1 C62 GABRA4 MPO RBP2 ERBB2 C19 GRIA2 ADRA1B C41 C14 C55 NTRK2 SERPINE1 PON1 C57 C10 HMGCR PDX1 PYY KCNMA1 C17 C1 EDNRA NCOA1 C13 C9 GSK3B PLAU C37 C29 SLC6A2 GCGR ENPEP C40 CITED1 C20 C12 PLG C59 C42 PPARD ESR2 ESR1 KDR CHRNA7 C22 ADRA2B C46 NOS2 KCNH2 C52 ADRA1A CDK2 HSP90AB1 C39 SCN5A SLC6A3 MAOB HTR2A F2 SLC6A4 ADRA2A PDE3A ADRB2 PKIA F10 DRD2 CHEK1 ADRB1 PIM1 LTA4H MAOA MAPK14 CALM1 F7 CCNA2 PRKCB C34 CTRB1 ADRA2C DPP4 PIK3CG BTK AR ADRA1D SELP RAC1 CHAT CA1 CYP1A1 NQO2 IGF2 GFAP

Figure 3: H-C-T network: herb-compound-target (H-C-T) network demonstrated multicompound, multitarget property of BSS. In this network, red and blue nodes represent herbs, green nodes show compounds, and pink nodes indicate targets and node size is relative to the degree and edges demonstrate interactions between nodes.

network analysis, and relevant organ location mapping was the delivery time, the amount of bleeding, and the residual used to explain the targets of BSS in relation to the parturition rate of placenta [52, 53]. In addition, BSS targets six genes of process. There is no denying that network based analysis is GABA receptor and NOS, which was reported to be related powerful approach for identifying the actions of multitarget- oxytocin neurons at the time of parturition in rats [54]. Also, ing herbal medicines at the systems level and our study shows BSS targets NOS and NO (nitric oxide) which are involved in target genes of BSS are strongly connected to parturition the regulation of uterine contractility during pregnancy and related pathways, biological processes, and organs. It was is a key factor for the onset of labor [55], and iNOS (inducible confirmed that 98% of the active compounds of BSS were nitricoxidesynthase)canbeupregulatedaccordantlyby interacted with more than two targets and 39.5% of the similar inflammatory mediators during ripening [11]. targets related to more than one compound. The synergetic multitarget properties of BSS were visualized, but further In fact, rather than DG, Angelicae Gigantis Radix (Dang- discussion about differentiated drug action based on degree gwi,AGR)growsnaturallyinKorea;forthatreason,the centrality and simultaneous targeting effect of more than one combination of AGR and CG is commonly used as BSS compound is required [51]. Also, detailed potential pathways in Korea. Instead, DG is named as Chinese Danggwi for ofBSSshouldbeexploreddeeplyinthefuture. accurate classification in Korea. Several studies have shown Similar findings were identified in a few RCT researches AGR is differs from DG in terms of its main active con- in China that using BSS in induction of labor can reduce stituents and genetic form. AGR is mainly composed of 12 Evidence-Based Complementary and Alternative Medicine

CDK2 CCNA2 HSP90AB1 ESR2 AR

ESR1 IL10 PTEN PPARG CDC25B Prostate.cancer SCD Progesterone.mediated.oocyte.maturation

SP1 PPARD Hepatitis.B LPL IL6 Chagas.disease..American.trypanosomiasis. PPARGC1A PGR AMPK.signaling.pathway SERPINE1 RXRA Estrogen.signaling.pathway SREBF1 Prolactin.signaling.pathway COL1A1 Pathways.in.cancer PPAR.signaling.pathway yroid.hormone.signaling.pathway HMGCR Insulin.resistance CTSD Pertussis SCN5A NCOA2 Amoebiasis PTPN1 IGF2 PLAU NCOA1 TNF Sphingolipid.signaling.pathway

G6PC SLC2A2 Proteoglycans.in.cancer INS Adrenergic.signaling.in.cardiomyocytes FABP1 ERBB2 GSK3B ACSL4 BCL2 Non.alcoholic.fatty.liver.disease..NAFLD. PIK3CG ADRA1A KCNMA1 EDNRA ACSL1 SI Carbohydrate.digestion.and.absorption HIF.1.signaling.pathway CASP3 FOS cGMP.PKG.signaling.pathway Insulin.signaling.pathway AMY2A MAPK14 PTGER3 ADRA1D CALM1 LCT PPARA RAC1 SRC ADH1C KDR Vascular.smooth.muscle.contraction ADRA1B ADH1A ENPP7 NOS2 PRKACA ADRB2 ADA AMY1A NOS3 ADRA2B VEGF.signaling.pathway Regulation.of.lipolysis.in.adipocytes Salivary.secretion ADRA2C SDHA HK1 ADSS ADRB1 HSD17B6 ALAD PRKCB CHAT Calcium.signaling.pathway ADRA2A PTGS2 Amyotrophic.lateral.sclerosis..ALS. Cholinergic.synapse PDHX ATP5B PLA2G1B CHRM2 SLC6A3 ADORA2A cAMP.signaling.pathway NR3C1 SOD1 Dopaminergic.synapse CHRM3 MAT2A SUCLA2 NOS1 ADH1B PDE3A Amphetamine.addiction PRSS3 OTC Metabolic.pathways PTGS1 CHRNA7 AGXT CAT PNP CHRM1 ME2 Serotonergic.synapse

ME3 Arginine.and.proline.metabolism DRD1 Neuroactive.ligand.receptor.interaction Retrograde.endocannabinoid.signaling Carbon.metabolism AHCYL1 HTR2A PLG ALDH5A1 MAT1A SLC6A4 GABAergic.synapse AKR1B1 MAOB DRD2 F2 GCGR CYP1A2 CHRNA2 MAOA GRIA2 Morphine.addiction GLUD1 PCYT1A GLRA1 GOT1 GABRA2 LTA4H CYP1A1 P4HA2 PON1 GOT2

OAT GABRA1 ABAT

GABRA6 GABRA3

GABRA5 GABRA4

Nicotine.addiction

Figure 4: T-P network: in target-pathway (T-P) network, circular nodes represent compounds and triangles indicate pathways. Node size is relative to the degree and edges demonstrate interactions between nodes.

water soluble polysaccharide but coumarin, which is liposol- remodeling and parturition by estrogen [58]. Further study uble including nodakenin (1), peucedanone (2), marmesin is needed in terms of the effect of BSS on inflammatory (3), decursinol (4), 7-hydroxy-6-(2R-hydroxy-3-methylbut- reactions and parturition. 3-enyl) coumarin (5), demethylsuberosin (6), decursin (7), Furthermore, the CG-DG herb pair has other names, such decursinol angelate (8), and isoimperatorin (9) [56]. Of as, Gunggui-tang (weight ratios of 2 : 3 or 1 : 1), Ogeum-san these, decursin and its isomer decursinol angelate have been (1 : 1), Iphyo-san (1 : 1), and Sinmyo Bulsu-san (1 : 2), those reportedtobetheactivecompoundsinAGR[57].Itwas are prepared at different weight ratios [3]. Accordingly, identified in the experimental studies that AGR and DG weight ratio should be determined based on considerations act via different mechanisms in the cardiovascular, central of targeted symptoms for relevant clinical applications. nervous system, and anticancer activity but both have similar pharmacological effects [57]. Since the compositions of DG 5. Conclusion and AGR differ, further study on BSS with AGR is required. Currently, BSS is commonly prescribed to treat cerebra ThisstudyresultsshowthatBulsu-san (BSS) is highly vascular and cardiovascular diseases in China [33], but, in connected to the parturition related pathways, biological Korea, BSS is widely applied in obstetrics. processes, and organs. Most compounds in BSS work together The similarity between cervical ripening during partu- with multiple target genes in a synergetic way, and this was rition and inflammatory reaction has been pointed out in confirmed using herb-compound-target network and target- earlier studies; this has been attributed to the induction pathway network analysis. The mRNA expression of relevant of leukocyte migration into tissue, thus promoting cervical target genes of BSS was elevated significantly in parturition Evidence-Based Complementary and Alternative Medicine 13

FOS

IGHG1 ENPEP

uterus and/or uterus corpus COL7A1

ADRB1 ADH1B CDK2 KCNH2 NR3C2

MAOB KDR EDNRA CHI3L1 DPP4 CHEK1 P4HA2 ATP5B ADRA1D SUCLA2 COL1A1 ME3 PGR ADRA1A TEP1 ADRA1B SOAT1 PPARGC1A BCHE ADRA2C OAT ADRA2B PTGER3 CHAT IGF2 RAC1 AGXT DRD2 AR PIM1 ALDH5A1 ADRA2A CA1 PDX1 PDE3A SREBF2 UCP3 SELP GABRA3 CHRNA7 NR3C1 CHRM2 NOS1 ACSL4 GABRA1 ESR2 AHCYL1 ADSS ERBB2 SP1 LEPREL2 CHRNA2 ENPP7 SRC ALAD PDHX SLC6A3 NOS2 GAP43 ESR1 GLRA1 HTR2A SOD1 hypothalamus and/or pituitary CHRM3 NR1I3 Pituitary SCD SDHA SmoothMuscle GLUD1 PNP SCN5A PTGS1 SREBF1 fetus and/or placenta CETP PKIA PIK3CG MAT1A HSP90AB1 NCOA2 TXNRD1 WholeBlood IL10 PPARG GFAP LEPRE1 GOT1 GABRA4 PPARD OTC PRKACA PCYT1A GIG18 PKM2 ADH1A SLC22A5 GOT2 ME2 GCGR PPARA AKR1B1 CITED1 NTRK2 APOB KCNMA1 MOGAT2 FOSL2 CTSD LPL HK1 LCT F2 ABAT ACSL1 PON1 SLC6A2 CASP3 IL6 CCNA2 PTPN1 NOS3 BCL2 ADORA2A DRD1 SERPINE1 PTEN MAOA GABRA5 PRKCB LTA4H G6PC GRIA2 MAPK14 RXRA GSK3B SLC2A2 NQO2 MAT2A UCP2 CALM1 HMGCR PTGS2 CAT CDC25B PLAU GABRA2 NCOA1 ADRB2

FABP1 F7 SLC6A4 HP PLG

Figure 5: Target organ location map: it shows that tissue-specific patterns of mRNA expression are highly active in relative organs of parturition process such as uterus, fetus, placenta, hypothalamus, pituitary, and smooth muscle. Yellow nodes show compounds and pink nodes indicate targets and node size is relative to the degree and edges demonstrate interactions between nodes.

relatedorgantissues,suchasthoseoftheuterus,placenta, Conflicts of Interest fetus, hypothalamus, and pituitary gland. This study employed the network analytical methods The authors declare that there are no conflicts of interest. to show the multicompound, multitarget properties of BSS. The results not only support clinical applications of BSS on Acknowledgments easing childbirth but also suggest the related target genes and pathways of BSS on promoting parturition according to a ThisresearchwassupportedbyBasicScienceResearch systems-level in silico analytic approach. However, detailed Program through the National Research Foundation of mechanisms and other functions of BSS should be discussed Korea (NRF) funded by the Ministry of Education (NRF- further. 2015R1D1A1A01059994). 14 Evidence-Based Complementary and Alternative Medicine

References herb Folium eriobotryae,” International Journal of Molecular Sciences, vol. 16, no. 2, pp. 2913–2941, 2015. [1] S. Kim and Y. Lee, “The prescriptions of enriching blood and [17] P.Willett,J.M.Barnard,andG.M.Downs,“Chemicalsimilarity nourishing vital essence (Fill yin blood prescription),” Journal searching,” Journal of Chemical Information and Computer of Korean Medical Classics,vol.20,pp.67–77,2007. Sciences, vol. 38, no. 6, pp. 983–996, 1998. [2] Y. Jin, C. Qu, Y. Tang et al., “Herb pairs containing Angelicae [18] H. Liu, J. Wang, W. Zhou, Y. Wang, and L. Yang, “Systems Sinensis Radix (Danggui): a review of bio-active constituents approaches and polypharmacology for drug discovery from and compatibility effects,” Journal of Ethnopharmacology,vol. herbal medicines: an example using licorice,” JournalofEthno- 181, pp. 158–171, 2016. pharmacology,vol.146,no.3,pp.773–793,2013. [3] J. Lyu and C. Jeong, “Constitution of prescription and medicinal [19] K. S. Pang, “Modeling of intestinal drug absorption: roles of effect adaptation diseases of ’bullsoosan (berghean)’in korean transporters and metabolic enzymes (for the gillette review medical books,” Journal of Korean Medical classics,vol.29,pp. series),” Drug Metabolism & Disposition,vol.31,no.12,pp.1507– 17–41, 2016. 1519, 2003. [4] W.Li, Y. Tang, J. Guo et al., “Enriching blood effect comparison in three kinds of blood deficiency model after oral administra- [20] T. Pei, C. Zheng, C. Huang et al., “Systematic understanding tion of drug pair of angelicae sinensis radix and chuanxiong the mechanisms of vitiligo pathogenesis and its treatment by rhizoma and each single herb,” China Journal of Chinese Materia Qubaibabuqi formula,” Journal of Ethnopharmacology,vol.190, Medica,vol.36,no.13,pp.1808–1814,2011. pp. 272–287, 2016. [5]W.Li,M.Huang,Y.Tang,J.Guo,E.Shang,andX.Liu,“Estab- [21]C.A.Lipinski,F.Lambardo,B.W.Dominy,andP.J.Feeney, lishment and optimization of acute blood stasis rat model,” “Experimental and computational approaches to estimate sol- Chinese Pharmacological Bulletin,vol.12,p.032,2011. ubility and permeability in drug discovery and development settings,” Advanced Drug Delivery Reviews, vol. 64, pp. 4–17, [6] Y.-Z. Hou, G.-R. Zhao, Y.-J. Yuan, G.-G. Zhu, and R. Hiltunen, 2012. “Inhibition of rat vascular smooth muscle cell proliferation by extract of Ligusticum chuanxiong and Angelica sinensis,” Journal [22] A. K. Ghose, V. N. Viswanadhan, and J. J. Wendoloski, “Predic- of Ethnopharmacology,vol.100,no.1-2,pp.140–144,2005. tion of hydrophobic (lipophilic) properties of small organic molecules using fragmental methods: an analysis of ALOGP [7]C.W.C.Bi,L.Xu,X.Y.Tianetal.,“FoShouSan,anancient and CLOGP methods,” TheJournalofPhysicalChemistryA,vol. chinese herbal decoction, protects endothelial function through 102,no.21,pp.3762–3772,1998. increasing endothelial nitric oxide synthase activity,” PLoS ONE, vol. 7, no. 12, Article ID e51670, 2012. [23] P. Rajasethupathy, S. J. Vayttaden, and U. S. Bhalla, “Systems modeling: a pathway to drug discovery,” Current Opinion in [8]W.Li,Y.Hua,Y.Tang,H.Wang,L.Qian,andM.Gao,“Effects Chemical Biology,vol.9,no.4,pp.400–406,2005. of radix angelicae sinensis and rhizoma chuanxiong on mouse uterine contractions in vitro,” Journal of Nanjing University of [24] H. Yu, J. Chen, X. Xu et al., “A systematic prediction of multiple Traditional Chinese Medicine,vol.2,p.014,2010. drug-target interactions from chemical, genomic, and phar- [9] M. McLean and R. Smith, “Corticotrophin-releasing hormone macological data,” PLoS ONE,vol.7,no.5,ArticleIDe37608, and human parturition,” Reproduction,vol.121,no.4,pp.493– 2012. 501, 2001. [25] C. H. Wu, R. Apweiler, A. Bairoch, D. A. Natale, W. C. Barker, [10]S.K.Kota,K.Gayatri,S.Jammulaetal.,“Endocrinologyofpar- B. Boeckmann et al., “The universal protein resource (UniProt): turition,” Indian Journal of Endocrinology and Metabolism,vol. an expanding universe of protein information,” Nucleic Acids 17,no.1,pp.50–59,2013. Research,vol.34,pp.D187–D191,2006. [11] M. Ledingham, A. J. Thomson, I. A. Greer, and J. E. Norman, [26] G. Bindea, B. Mlecnik, H. Hackl et al., “ClueGO: a Cytoscape “Nitric oxide in parturition,” BJOG: An International Journal of plug-in to decipher functionally grouped gene ontology and Obstetrics & Gynaecology,vol.107,pp.581–593,2000. pathway annotation networks,” Bioinformatics,vol.25,no.8,pp. 1091–1093, 2009. [12] M. Shen, S. Tian, Y. Li et al., “Drug-likeness analysis of traditional Chinese medicines: 1. property distributions of drug- [27] M. E. Smoot, K. Ono, J. Ruscheinski, P. L. Wang, and T. Ideker, like compounds, non-drug-like compounds and natural com- “Cytoscape 2.8: new features for data integration and network pounds from traditional Chinese medicines,” Journal of Chem- visualization,” Bioinformatics,vol.27,no.3,pp.431-432,2011. informatics,vol.4,no.1,article31,2012. [28] J.-B. Pan, S.-C. Hu, D. Shi et al., “PaGenBase: a pattern gene [13] X. Xu, W.Zhang, C. Huang et al., “Anovel chemometric method database for the global and dynamic understanding of gene for the prediction of human oral bioavailability,” International function,” PLoS ONE,vol.8,no.12,ArticleIDe80747,2013. Journal of Molecular Sciences,vol.13,no.6,pp.6964–6982,2012. [29] W. Zhang, Q. Tao, Z. Guo et al., “Systems pharmacology [14]I.Hubatsch,E.G.E.Ragnarsson,andP.Artursson,“Determi- dissection of the integrated treatment for cardiovascular and nation of drug permeability and prediction of drug absorption gastrointestinal disorders by traditional chinese medicine,” in Caco-2 monolayers,” Nature Protocols,vol.2,no.9,pp.2111– Scientific Reports,vol.6,p.32400,2016. 2119, 2007. [30] Z. Prevorˇsek, G. Gorjanc, B. Paigen, and S. Horvat, “Congenic [15] W. Tao, X. Xu, X. Wang et al., “Network pharmacology-based and bioinformatics analyses resolved a major-effect Fob3b QTL prediction of the active ingredients and potential targets of on mouse Chr 15 into two closely linked loci,” Mammalian Chinese herbal Radix Curcumae formula for application to Genome,vol.21,no.3-4,pp.172–185,2010. cardiovascular disease,” Journal of Ethnopharmacology,vol.145, [31] J. Du, B. Bai, X. Kuang et al., “Ligustilide inhibits spontaneous + no.1,pp.1–10,2013. and agonists- or K depolarization-induced contraction of rat [16] J. Zhang, Y. Li, S.-S. Chen et al., “Systems pharmacology dis- uterus,” Journal of Ethnopharmacology,vol.108,no.1,pp.54– section of the anti-inflammatory mechanism for the medicinal 58, 2006. Evidence-Based Complementary and Alternative Medicine 15

[32] K. J. Chen and K. Chen, “Ischemic stroke treated with ligus- [48] S. Wray, K. Jones, and S. Kupittayanant, “Calcium signaling ticum chuanxiong,” Chinese Medical Journal,vol.105,pp.870– and uterine contractility,” Journal of the Society for Gynecologic 873, 1992. Investigation,vol.10,no.5,pp.252–264,2003. [33] W.Li,J.Guo,Y.Tangetal.,“Pharmacokineticcomparisonoffer- [49] G. Y. Oudit and Z. Kassiri, “Role of PI3 kinase gamma in ulic acid in normal and blood deficiency rats after oral admin- excitation-contraction coupling and heart disease,” Cardiovas- istration of angelica sinensis, ligusticum chuanxiong and their cular and Hematological Disorders,vol.7,no.4,pp.295–304, combination,” International Journal of Molecular Sciences,vol. 2007. 13, no. 3, pp. 3583–3597, 2012. [50] R. Smith, “Parturition,” The New England Journal of Medicine, [34]W.Karl,W.Cathy,X.Li,andG.Amy,“Pharmacokinetic vol. 356, no. 3, pp. 271–283, 2007. analyses of ferulic acid in rat plasma by liquid chromatog- [51] S. Y. Suh and W. G. An, “Systems pharmacological approach raphy?tandem mass spectrometry: a synergistic action of an of pulsatillae radix on treating crohn’s disease,” Evidence-Based ancient herbal decoction fo shou san,” Pharmaceutica Analytica Complementary and Alternative Medicine,vol.2017,ArticleID Acta,vol.6,no.5,p.2,2015. 4198035, 21 pages, 2017. [35] D. Huang, “8 purgative herbs,” Chinese Materia Medica: Chem- [52] C. l. Chen and J. Y. Chen, “Analysis on the effect of foshousan istry, Pharmacology and Applications,p.231,1998. combined with ethacridine lactate in mid-term pregnancy,” [36] X. Liu, Q. Zhao, C. Li, R. Zhang, X. Wang, and W. Xu, Journal of Practical Traditional Chinese Medicine,vol.30,pp.511- “Synthesis of chuanxiong perlolyrine and its protective effect 512, 2014. on injured vascular endothelial cells,” Journal of Shandong [53] Y. Sun, “Modified Xiong Gui Tang with auricular acupoint press University (Health Sciences),vol.41,pp.485–487,2003. for 70 cases of late pregnancy,” Guide of China Medicine,vol.7, [37] J. Ishida, M. Kozuka, H.-K. Wang et al., “Antitumor-promoting no.20,pp.92-93,2009. effects of cyclic diarylheptanoids on Epstein-Barr virus activa- [54] J.-J. Koksma, J.-M. Fritschy, V. MacK, R. E. Van Kesteren, tion and two-stage mouse skin carcinogenesis,” Cancer Letters, and A. B. Brussaard, “Differential GABA A receptor clustering vol. 159, no. 2, pp. 135–140, 2000. determines GABA synapse plasticity in rat oxytocin neurons [38] J. Liu, T. Pei, and J. Mu, “Systems pharmacology uncovers around parturition and the onset of lactation,” Molecular and the multiple mechanisms of Xijiao Dihuang decoction for the Cellular Neuroscience, vol. 28, no. 1, pp. 128–140, 2005. treatment of viral hemorrhagic fever,” Evidence-Based Comple- [55] H.Maul,M.Longo,G.R.Saade,andR.E.Garfield,“Nitricoxide mentary and Alternative Medicine,vol.2016,ArticleID9025036, and its role during pregnancy: From ovulation to delivery,” 17 pages, 2016. Current Pharmaceutical Design,vol.9,no.5,pp.359–380,2003. [39] H. Wang and Y. V. Stjernholm, “Plasma membrane receptor [56] M.-J. Ahn, M. K. Lee, Y. C. Kim, and S. H. Sung, “The simul- mediated MAPK signaling pathways are activated in human taneous determination of coumarins in Angelica gigas root by uterine cervix at parturition,” Reproductive Biology and Endo- high performance liquid chromatography-diode array detector crinology,vol.5,article3,p.1,2007. coupled with electrospray ionization/mass spectrometry,” Jour- [40] S. Mesiano and T. N. Welsh, “Steroid hormone control of myom- nal of Pharmaceutical and Biomedical Analysis,vol.46,no.2,pp. etrial contractility and parturition,” Seminars in Cell & Develop- 258–266, 2008. mental Biology,vol.18,no.3,pp.321–331,2007. [57] S. Kim, H. Oh, J. Kim, J. Hong, and S. Cho, “A review of phar- [41] P.W.Nathanielsz, S. L. Jenkins, J. D. Tame, J. A. Winter, S. Guller, macological effects of angelica gigas, angelica sinensis, angelica and D. A. Giussani, “Local paracrine effects of estradiol are acutiloba and their bioactive compounds,” The Journal of central to parturition in the rhesus monkey,” Nature Medicine, Korean Oriental Medicine,vol.32,pp.1–24,2011. vol. 4, no. 4, pp. 456–459, 1998. [58] J. E. Norman, S. Bollapragada, M. Yuan, and S. M. Nelson, [42] Y. Li, H. Je, S. Malek, and K. G. Morgan, “ERK1/2-mediated “Inflammatory pathways in the mechanism of parturition,” phosphorylation of myometrial caldesmon during pregnancy BMC Pregnancy and Childbirth,vol.7,supplement1,articleS7, and labor,” AmericanJournalofPhysiology—Regulatory,Integra- 2007. tive and Comparative Physiology,vol.284,no.1,pp.R192–R199, 2003. [43] Y. Li, H.-D. Je, S. Malek, and K. G. Morgan, “Role of ERK1/2 in uterine contractility and preterm labor in rats,” American Journal of Physiology-Regulatory, Integrative and Comparative Physiology,vol.287,no.2,pp.R328–R335,2004. [44] S. I. Berger and R. Iyengar, “Network analyses in systems pharmacology,” Bioinformatics,vol.25,no.19,pp.2466–2472,2009. [45] P. Li, L.-W. Qi, E.-H. Liu, J.-L. Zhou, and X.-D. Wen, “Analysis of Chinese herbal medicines with holistic approaches and integrated evaluation models,” TrACTrendsinAnalyticalChemistry, vol. 27, no. 1, pp. 66–77, 2008. [46] L. M. Espinoza-Fonseca, “The benefits of the multi-target approach in drug design and discovery,” Bioorganic & Medicinal Chemistry,vol.14,no.4,pp.896-897,2006. [47] W. Zhang, Y. Bai, Y. Wang, and W. Xiao, “Polypharmacology in drug discovery: a review from systems pharmacology perspective,” Current Pharmaceutical Design,vol.22,no.21,pp.3171– 3181, 2016. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 9785345, 8 pages https://doi.org/10.1155/2017/9785345

Review Article The Modulatory Effect of Acupuncture on the Activity of Locus Coeruleus Neuronal Cells: A Review

Gihyun Lee and Woojin Kim

Department of Physiology, College of Korean Medicine, Kyung Hee University, Seoul 02447, Republic of Korea

Correspondence should be addressed to Woojin Kim; [email protected]

Received 27 July 2017; Accepted 17 September 2017; Published 18 October 2017

Academic Editor: Gabino Garrido

Copyright © 2017 Gihyun Lee and Woojin Kim. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The Locus Coeruleus (LC) is a small collection of noradrenergic neurons located in the pons. In the brain, noradrenaline (NE) is primarily produced by noradrenergic cell groups in the LC, which is the largest group of noradrenergic neurons in the central nervous system. Acupuncture, including the electroacupuncture which is a modified acupuncture method, is known to be effective in various kinds of diseases, and the involvement of noradrenergic system in the central nervous system has been reported by previous studies. However, on whether acupuncture can modulate the LC neuronal cells activities, results vary from studies to studies. In this paper, we included twelve articles, which observed the effect of acupuncture on the activities of LC in humans and animals. Our study shows that, among twelve included studies, six reported decrease of LC activities, whereas six showed increase of LC activities after acupuncture treatment. Although it is difficult to draw a firm conclusion, the authors suggest that the difference of frequencies may play an important role in the modulatory effect of acupuncture on LC. Further studies are needed to clarify the precise mechanism of acupuncture on LC, as it can lead to a new therapeutic method for various LC-NE related diseases.

1. Introduction of 𝛼2-adrenoceptors was demonstrated to inhibit the follower cells [5]. 𝛼2-Adrenoceptors are widely distributed in the The Locus Coeruleus (LC), meaning the blue spot in Latin, spinal cord as well as in the brain [7]. Via these receptors, is a small collection of noradrenergic neurons (about 16,000 NE produced in LC exerts many different functions in our per hemisphere in the human), located just behind the body, and their function is known as the LC-NA system. The periaqueductal gray (PAG) in the dorsorostral pons [1]. LC-NA system was reported to be important in learning and In the brain, noradrenaline (NE) is primarily produced by noradrenergic cell groups classified as A1–A7, which projects memory [8] and sleep-wake cycle [6, 9]. Also, by stimulating NE to widespread area of the brain. Among them, A5–A7 the sympathetic nerve, they could regulate the blood pressure groups project not only to the brain but also to the spinal [10] and play an important role in the stress [11]. Furthermore, cord [2], and the A6, which is the LC, is the largest group of theyarealsocloselyrelatedtopain[4],asNEisknownto noradrenergic neurons in the central nervous system (CNS) inhibit the transmission of the pain, by acting through the [3]. NE, as other catecholamines, dopamine, and epinephrine, 𝛼2-adrenoceptors present at the spinal cord, at the pre- and possesses two hydroxyl groups and one amine group bound postsynaptic neurons. As one of the major NE producing sites to a benzene ring, and it is biosynthesized from tyrosine. in the brain, LC along with the PAG and rostral ventromedial Tyrosine is first converted into dopamine, and dopamine is medulla (RVM) play an important role in pain modulation, further converted into NE by dopamine-beta-hydroxylase as they are known to produce NE, endogenous opioid, and (DBH) present in noradrenergic cells [4]. serotonin, respectively. These neurotransmitters are involved NE exerts its effect in various parts of the CNS by its in pain inhibition by acting through their receptors at the receptors present on cells. The most widely known noradren- spinal cord [12]. ergic receptors are 𝛼1-, 𝛼2-, or 𝛽-adrenoceptors. Activation of Acupuncture is a treatment method that has a long his- 𝛼1-adrenoceptors and 𝛽-adrenoceptors by NE is reported to tory, and nowadays acupuncture along with electroacupunc- generally excite the follower cells [5, 6]. In contrast, activation ture, which is a modified method by providing electrical 2 Evidence-Based Complementary and Alternative Medicine current through acupuncture, is used throughout the world. whereas six reported the increase of LC activities after EA According to the report released by the World Health Orga- treatment (Table 1). Thus, in this part, we will divide articles nization (WHO), more than 40 disorders, including stress into two parts: increased and decreased activity of LC after and insomnia, can benefit from acupuncture treatment [13]. acupuncture treatment. Furthermore, it is recommended for various types of pain, such as low back pain [14], knee pain [15], and headache [16]. 2.1. Increased LC Activity by Acupuncture. To assess whether The analgesic effect of acupuncture is widely known, and, by the low (4 Hz) or high (100 Hz) frequencies of EA adminis- numerous studies conducted both in humans and in rodents tered at ST36 could affect the number of Fos-Like Immunore- [17,18],itseffecthasbeenproven.Studiesperformedon active (FLI) neurons in the LC and in the spinal cord, Lee animals have shown that acupuncture can significantly relieve and Beitz [25] used lightly anesthetized rats. For control, behavioral signs such as hyperalgesia and allodynia in periph- acupuncture was administered at ST36 without any electrical eral nerve injury-induced neuropathic pain models [19–21]. stimulation. Their results show that three hours of both low Also, it was reported to reduce hypertension [22] and stress and high frequency of EA treatment exhibited a significantly [23]. However, although the curative effect of acupuncture is greaternumberofFos-labeledneuronsinthedorsalhornof continuously reported by clinical and experimental studies, theL2spinalcordsegmentandtheLC. the mechanism that lies behind it is not fully understood, Kwon et al. [26] also demonstrated the effect of EA by especially in the brain. using similar protocol to Lee and Beitz, as low (4 Hz) or high Over the last several decades, researchers have clarified (100 Hz) frequencies of EA were used to see the cellular activ- the involvement of NE in the action of acupuncture [24], ity of central catecholaminergic (CA) synthesizing neurons in by using diverse animal models. In pain, by reporting that the LC. Immunohistochemistry with double labeling method the action of acupuncture was blocked by 𝛼2-adrenoceptors between FLI neurons and DBH- or tyrosine hydroxylase- antagonist yohimbine, they demonstrated the involvement of (TH-) positive neurons was used, as TH-positive neurons are NE in acupuncture analgesic mechanism [19]. In accordance an indicator of CA. They used na¨ıve rats and anesthetized with these results, many articles mentioned the LC as the the animals with isoflurane throughout the acupuncture treatment. For EA administration, bilateral ST36 acupoints source of the NE action in the spinal cord after acupunc- were stimulated for 120 min. Immunohistochemistry was ture treatment [19, 21, 24]. However, results of experiments conducted two hours after the electrical stimulation. Their conducted to clarify the effect of acupuncture in the LC results demonstrate that both frequencies of EA increased the were not consistent. Some reported that LC activities were number of FLI neurons in the LC as well as other parts of the increased by acupuncture treatment, whereas some reported brain such as the dorsal raphe (DR), hypothalamic arcuate that they were decreased by acupuncture treatment. Thus, in nucleus(Arc),A5,andA7.Furthermore,boththelowand this review, we will first state the differences of experimental high frequencies of EA increased the number of FLI neurons methods and the results of all twelve included papers, and, and the cellular activities of DBH/TH-positive neurons in in the discussion, we will try to analyze the results and will the LC. Although both the low and high frequencies induced further try to draw a conclusion. Given the importance of LC- significant increase in the LC, high frequency had a stronger NA system in our body and the implication of acupuncture effect compared to low frequency𝑝 ( < 0.01). in NE system, we believe that a timely review is important, Before observing the change in the LC induced by EA, to guide future efforts in the advancement of acupuncture Medeiros et al. used repeated immobilization protocols to treatment, as well as in the acupuncture related researches. exclude the effect of anesthesia or acute immobilization stress Basedonthepreviouslypublishedstudies,wewillproceed on the c-Fos expression. Firstly, they observed the change in to expand on clarification of the effect of acupuncture on the brain c-Fos expression following repeated immobilization for activity of the LC neuronal cells. two hours/day for 13 days. They found that this method was effective as the c-Fos expression was significantly different 2. Modulation of LC Neuronal from rats which did not undergo repeated immobilization Cells by Acupuncture (𝑝 < 0.001). The 60 min bilateral stimulation of EA at ST36 significantly increased the level of c-Fos expression in theLC In our review, we have included twelve articles which ana- compared to the control, where the acupuncture was admin- lyzed the effect of acupuncture on LC neuronal cells. From istered 5 mm lateral to the midline of the posterior surface the twelve included studies, one was assessed in humans, of the hind limb (𝑝 < 0.05). However, the effect was not whereas others were assessed in animals such as rats, rab- significant in the group where rats did not undergo repeated bits, and goats. Most articles used ST36 acupoint; however, immobilization. The EA neither increased nor decreased the other acupoints such as GB30 and LI4 were also used. The activity of the LC. stimulation frequency and duration were also different, as Li et al. [27] have observed the effect of EA on LC well as the methods they used to assess the effect on the by using inflammatory pain rat model. Inflammatory pain LC. Studies conducted in human used functional magnetic was induced by injecting complete Freund's adjuvant (CFA) resonance imaging (fMRI), and most studies conducted subcutaneously into one hind paw of rats. For EA treatment, with animals used c-Fos expression method. The results GB30 acupoint was chosen, and GB30 was stimulated for were also different. Among twelve studies, six reported the 20 min. For control, acupuncture needles were inserted bilat- decrease of LC activities following acupuncture stimulation, erally into GB30 without electrical or manual stimulation. Evidence-Based Complementary and Alternative Medicine 3 versus 15.75 ± 4.17 versusgroup 𝑝 < 0.02 ). Control (𝑝 < 0.001) . 𝑝 < 0.05 (control); pg/ml; 𝑝 < 0.05 ). mA 1.898 ± 0.38 versus group without EA (100 Hz). 𝑝 < 0.05 6.00 ± 2.55 6.9 ± 1.02 ; 59.41 ± 7.8 1.36 ± 0.22 𝑝 < 0.05 ). 𝑝 < 0.05 ). versus mA to versus normal saline injected group). pg/ml to 13.50 ± 3.07 15.8 ± 1.28 156 ± 29 𝑝 < 0.05 EA had a significant inhibitoryfeedback effect increase in on L-dopa-caused the synthesis( of DBH in the LC groupwithoutEAdidnothavesignificantchange.EA increased the NE level inthe the spinal perfusate cord. of The dorsal pain horn0.36threshold of ± increased 0.09 from EA decreased the NE level in( perfusate of LC Repeated immobilization (6 days: 1 h/day;2 13 h/day) days: significantly reduced the immobilization induced c-Fos expression in LC comparedundisturbed to rats ( EA increased this decreased level ofthe c-Fos LC expression compared in to control3.148 ( ± 0.43 ; without EA treatment) EA decreased the frequent micturition andbasal increased bladder pressure in L-dopa injected( group Both low (4 Hz) and highexhibited (100 a Hz) significantly frequency greater of number EA ofneurons Fos-labeled in the LC. (4 Hz); Both low (4 Hz) and highincreased (100 the Hz) Fos frequency expression of in EA catecholaminergic neurons in the LC ( treatment). Colocalization of FLI neurons and DBHneurons or all TH-positive increased after EA ( IHC Methods used Findings 2-3 Hz 20 min. Brain perfusate 100 Hz 60 min. c-Fos expression 4 or 100 Hz 180 min4 or c-Fos 100 expression Hz 120 min. c-Fos expression EA stimulations ST36 ST36 ST36 Acup. Freq. Durat. Bilateral Bilateral Bilateral LI4, SJ05 Ipsilateral Bilateral B29 1 Hz 30 min. Table 1: Summary of studies on the effect of acupuncture on the activity of neuronal LC cells. rat SDrat SD rat Models Albino rabbit hyperactivated bladder/SD rat L-Dopa-induced Immobilization stress induced/Albino Wistar Authors Cao et al. 1983 LeeandBeitz1992 Kwon et al. 2000 S. Wang and X. Wang 2002 Medeiros et al. 2003 4 Evidence-Based Complementary and Alternative Medicine 𝑝 < 0.01 ) 𝑝 < 0.001 ). 𝑝<0.001 ). versus Hz ( versus normal 𝑝<0.05 20.5 ± 0.8 1.1 ± 0.8 to 𝑝 < 0.01 15.27 ± 2.40 ; 𝜇 msection).Untreatednaiverats 2.9 ± 1.5 versus 𝑝 < 0.01 ). FLI, Fos-Like Immunoreactive; fMRI, functional magnetic EA induced abundant Fos expression in( 110.4 the ± bilateral 9 /per LC 50 expressed little Fos in this nucleus. Double-immunofluorescence staining demonstrated that Fos was completely colocalized with TH in the LC. 180 min of immobilization stressincrease produced a in significant FLI in thegroup). LC EA ( attenuated immobilization stress inducedin FLI the LC compared to control ( Immobilization stress induced for 21TH-immunoreactive days. expression The in the LC was significantly decreased in the ST36group ( versus stress induced group). c-Fos and c-Jun activity of thefrequencies LC (2, increased 60, in and all 100 Hz) three control. compared Pain to threshold the was blank measured usingpotassium the iontophoresis method. Pain threshold induced by 60 Hzthan was that higher by ( 0, 2, or 100 Hz. Compared with blank control, EAacupoints at increased two sets c-Fos of expression( 9.31 in ± the 2.39 LC 15.8 ± 1.0 ; Firing rate of noradrenergic LC neurons decreased significantly from Stimulation with EA deactivated LC activitysham compared EA to (8cm above the proximal edge of the patella). IHC fMRI recording expression Extracellular Methods used Findings c-Fos and c-Jun 1min. 30 min 30 min. Table 1: Continued. 3 HZ 30 min. c-Fos expression 15 Hz 10 Hz 20 min. c-Fos expression 2and 1.0–1.5 turns/s 0, 2, 60, Manual rotation and 100 Hz EA stimulations TE8 ST36 GB30 sacral Acup. Freq. Durat. Bilateral and PC6 AH1, and Ipsilateral MA at the segment (S3) Bilateral HT3 GV20, Santai, Bilateral ST36 2 Hz 10 min. rat SD rat Models Human CFA-induced induced/SD rat induced/SD rat Immobilization stress Immobilization stress inflammatorypain/SD Authors Lee et al. 2004 Li et al. 2007 Wang et al. 2007 Napadow et al. 2009 Park et al. 2013 Qiu et al. 2015 Two-year-old male goats Hu et al. 2016 One-year-old male goats GV20, Santai 60 Hz 30 min. c-Fos expression resonance imaging; GV20, Baihui; GB30, Huan Tiao; TH, tyrosine hydroxylase. Italics font style: decreased activityelectroacupuncture; of Freq, LC; frequency; roman IHC, font immunohistochemistry; style: LC, increased Locus activity Coeruleus; of LC; L14, Hegu; Acup., Acupoints; SJ05, AH1, Wai Guan; Ergen; TE8, CFA, Sanyangluo; Complete ST36, Freund’s Adjuvant; Zusanli; DBH, Dopamine-P-hydroxylase; Durat., duration; EA, Evidence-Based Complementary and Alternative Medicine 5

Paw withdrawal latency to a noxious thermal stimulus was NE level also decreased in PAG and NRM areas. However, at measured before and after 20 min of EA treatment. Com- the dorsal horn of the spinal cord, the NE level was increased pared to sham EA, EA significantly𝑝 ( < 0.05)increased compared to the level of before administrating EA (𝑝 < 0.02). withdrawal latency of the inflamed hind paws in the sham- Cao et al. supposed that EA cause a central inhibition, by operated rats. EA, compared to the sham EA, also signifi- decreasingtheNEreleasedinboththebrainandtheplasma, cantly inhibited c-Fos expression in laminae I-II of the spinal but increasing the NE level in the spinal cord, and suppose cord (58.4 ± 6.5 versus 35.2 ± 5.4 per section). However, EA that different pathway of NE is involved in the effect of EA. activated serotonin- and catecholamine-containing neurons S. Wang and X. Wang [32] observed the effect of 15 min in the nucleus raphe magnus (NRM) and LC that project to EA stimulation in rats with hyperactivated bladder induced the spinal cord. by intraperitoneally (i.p.) injected L-dopa. Injection of L- Qiu et al. [28] used two-year-old healthy hybrid male dopa induced a frequent micturition and increased basal goats to evaluate the levels of c-Fos and c-Jun expression bladder pressure compared to normal saline injected group induced by EA in the brain. EA were administered at a set of in rats. However, this increased frequency and pressure were GV20, Santai, Ergen (AH1), and Sanyangluo (TE8), and four decreased by bilateral B29 treatment for 30 min. Further- different frequencies, such as 0, 2, 60, and 100 Hz, were used. more, EA decreased the upregulated feedback in the synthesis The results show that three frequencies (2, 60, and 100 Hz) of of DBH in the LC, which resulted in the decreased release EA induced significant increase of c-Fos and c-Jun expression of NE from LC, after L-dopa injection. In the L-dopa group, in the LC, whereas 0 Hz did not induce any significant change DBH significantly increased to 10.7 ± 1.64 and 15.8 ± 1.28 intheLCcomparedtotheblankcontrol.Also,theexpressions for 8 and 24 h after the injection, respectively, compared with of c-Fos or c-Jun in the LC were not significantly different thenormalsalineinjectedcontrolgroup(𝑝 < 0.05 each). EA among the frequencies. However, pain threshold measured by treatment at B29 had an inhibitory effect on L-dopa-caused using the potassium iontophoresis test showed that the pain feedback increase in the synthesis of DBH. In the EA group, threshold induced by 60 Hz was higher (𝑝 < 0.01) than that DBH significantly decreased to 6.9 ± 1.02,24hafteraL-dopa induced by 0, 2, or 100 Hz. injection followed by 15 min after EA treatment as compared 𝑝 < 0.05 Hu et al. [29] also used healthy crossbred male goats to with that in the drug group ( ). assess the effect of EA on pain threshold and LC. EA was Yang et al. reported that EA at Shaohai and Neiguan administered at both GV20 and Santai acupoints for 30 min. (HT3-PC6) points significantly attenuated stress induced Firstly, by measuring the pain threshold using iontophoresis peripheral responses, such as increased blood pressure, heart method, they showed that the pain threshold induced by rate, and plasma CA [33]. In a subsequent study [23], they EA at set of GV20-Santai acupoints was 44.74% ± 4.56% reported that 180 min of immobilization stress significantly higher than that by EA at set of ST36 acupoints (32.64% ± increased FLI in various areas of the brain related to stress, 5.04%). Furthermore, they reported that, compared with such as the paraventricular hypothalamic nucleus (PVN), blank control, EA at two sets of acupoints increased c-Fos arcuate nucleus (ARN), supraoptic nucleus (SON), and the expressionintheLC(9.31±2.39 versus 15.27 ± 2.40; 𝑝 < 0.05). LC. In their study, bilateral combination stimulation of HT3 and PC6 for 30 min attenuated the increased number of FLI neurons induced by 180 min of immobilization stress in the 2.2. Decreased LC Activity by EA. Napadow et al. [30] LC compared to control group (20.5 ± 0.8 versus 15.8 ± 1.0; observed the effect of EA in humans. Ten healthy volun- 𝑝 < 0.01). teers’ brains were analyzed by using noninvasive fMRI. For Wang et al. examined the effects of acupuncture stimula- treatment, EA was administered unilaterally at ST36 and was tion to the sacral segment on electroencephalograms (EEGs) stimulated for 30 min, whereas, for sham acupuncture, the and activity of LC neurons in urethane-anesthetized rats same electrical stimulation was given to a nonacupoint site, [34]. A fine acupuncture needle (diameter, 0.35 mm) was 8 cm above the proximal edge of the patella, on the midline positioned at the periosteum of the sacral segment (S1–S4) by of the thigh. The difference of electrical current in theEA palpation and was rotated manually 1.0–1.5 turns/s for 1min. and the control group was not significant. They stated that LC Among the four sacral segments, the stimulation of acupunc- activity was decreased after acupuncture treatment. However, ture was more effective in S3 segments. This acupuncture the activity of PAG increased after EA treatment. treatment decreased the frequency of the wave from small- Cao et al. [31] conducted studies both in humans and in amplitude faster waves to large-amplitude slow waves, which rabbits. In human, they checked the change related to sym- demonstrate that acupuncture stimulation may change the pathetic nervous system functions such as palm temperature, state from light anesthesia to deep anesthesia, and also pulse rate, and pain tolerance threshold. Using rabbits, they decreased the activity of noradrenergic neurons in the LC. conducted biochemical studies. They observed whether EA Park et al. [35] experimented with rats undergoing could elevate or decrease NE level in the LC, PAG, and NRM immobilization stress for six hours a day for 21 consecutive andinthedorsalhornofthespinalcordbyusingperfusate days. The stress increased the response of the anxiety-related method.ThebaselineofNElevelinperfusateofLCwas behavior, and the serum level of corticosterone and the 156 ± 29 pg/ml; however, after 20 min of EA at LI4 and SJ05 number of TH-immunoreactive cells in the LC were also points, it dropped markedly to 59.41 ± 7.8 pg/ml (𝑝 < 0.02). increased. 10 minutes of EA treatment to the ST36 decreased At the same time, the pain threshold to stimulation increased the anxiety-related behavioral response, compared with from 0.36 ± 0,09 mA to 1.36 ± 0.22 mA (𝑝 < 0.001). The the stress group. Moreover, the serum corticosterone level 6 Evidence-Based Complementary and Alternative Medicine and TH-immunoreactive expression were also decreased to conclude whether the acupuncture decrease or increase in the ST36 group, compared to the group treated to the theactivityoftheLC.However,itmaybespeculatedthat nonacupuncture point in the tail. The anxiety-related behav- the frequency used in acupuncture stimulation may play an ior was tested using the elevated plus maze and the Vogel important part in the effect of acupuncture. Most studies in test on day 22. The serum concentration of corticosterone which the effect of acupuncture on LC induced the decrease was determined using an enzyme-linked immunosorbent of LC activities used low frequency (<4 Hz) with lesser than assay kit. The expression of TH in the LC was measured by 30 minutes of stimulation, whereas most studies, where the immunohistochemistry. acupuncture induced an augmentation of LC activity, used relatively high frequency (≥10 Hz) and longer stimulation 3. Discussion duration. By previously published studies conducted on pain, it was shown that the frequency of acupuncture stimulation For more than a thousand years, acupuncture has been used involved different mechanism in our body. For example, low as a treatment method. Through numerous clinical [36, 37] frequency of acupuncture induced analgesia was abolished and experimental studies [38, 39], its therapeutic effect has following lesions of the arcuate nuclei but not high frequency, been proven in various spectrums of diseases. However, com- whereasselectivelesionsoftheparabrachialnucleiattenuated pared to the large number of studies which investigated on high frequency induced analgesia but not low frequency [43]. its therapeutic effect, studies to clarify its mechanism in the These results stated that the mechanism of low and high brain are still relatively small in numbers [40]. Particularly, frequency acupuncture may be different. Furthermore, the in the LC, not only are the published experimental studies intensity of the stimulation may also play an important role. small, but also no review paper has ever been written, and to In our included studies, most of low-frequency stimulated our knowledge this is the first review paper. EA decreased the activity of the LC whereas high frequency Over the last several decades, researchers have supposed increased the activity of the LC, except in the studies of Kwon that acupuncture not only acts at the spinal level but also acts et al. [26], Lee and Beitz [25], and Qiu et al. [28]. In the studies at the supraspinal level, involving the activation of different of Kwon et al. and Lee and Beitz, the low frequency (4 Hz) of brain areas. By using pain animal models, researchers have EA increased the c-Fos expression in the LC. However, both suggested that acupuncture could decrease different types of of these studies used relatively longer stimulation (120 and acute and chronic pain through the descending inhibitory 180 minutes, resp.) compared to other studies where only 20 pathway, involving brain areas such as PAG, RVM, and tha- to 30 minutes of stimulation were given. Also, in the study lamus[21].LC,asthelargestNEproducingpartintheCNS, of Qiu et al., 2 Hz of acupuncture was administered in a set was also suggested to be related to acupuncture mediated pain of four acupoints simultaneously. So, we suppose that strong inhibition, as the acupuncture analgesic effect was blocked stimulation such as longer duration of the stimulation or by 𝛼2-adrenergic receptors antagonist at the spinal level [19]. increased number of stimulated acupoints may have led to an Also, by using animal stress model, it was reported that the incensement of the LC c-Fos expression, as strong stimulation acupuncture could decrease stress by modulating the NE level wasreportedtoincreasetheactivityoftheLC[44]. of our body [31]. However, although, in the spinal cord and in Duetothelimitednumbersofthestudiesandthe the peripheral site, various studies showed that NE is playing differences of the results, in this review, we cannot draw afirm an important role in the effect of acupuncture, the effect of conclusion. More high quality clinical as well as experimental acupuncture on the activities of the LC neuronal cells is still studies are needed, to draw any firm conclusions. Further- not clear [24, 41]. more, research not only in the LC but in other areas of the Inourreviewalso,allincludedtwelvearticlesshow brain should help the understanding of LC functions, as in different results. LC activities were found to decrease after the brain different areas are closely related to one another. acupuncture administration in some studies, whereas, in Also, for the advancement of the research in acupuncture, other studies, the activities of LC increased as demonstrated it might be helpful for researchers of acupuncture to adjust in Table 1. Among our twelve included studies (Table 1), one the stimulation criteria (duration, amplitude), so that all study observed the effect of acupuncture on humans [30], researchers could compare the difference of acupoint specific whereas eleven observed its effect on various animals. Eight and frequency specific effects of the acupuncture. studies used rats, one used rabbits [31], and two used goats [29]. Furthermore, the animal model they used were all different; seven studies assessed the effect on healthy¨ naıve human Conflicts of Interest and animals [26, 29–31, 34], three used immobilization stress The authors declare that there are no conflicts of interest to decrease [42] or increase [23, 35] the activity of LC before regarding the publication of this paper. acupuncture treatment, one used L-dopa induced hyperactivated bladder model [32],and finally one study used CFA- induced inflammatory pain model [27]. Also, the assessment Acknowledgments methods were all different. Most studies observed the change of c-Fos expression; however, Wang et al. and Napadow et al. The authors thank Dr. Sun Kwang Kim and Dr. Hidemasa used electrophysiology and fMRI, respectively. Furue for their kind support and advice on this manuscript. With all the differences of the results and the implicated This work was supported by a Basic Science Research treatment methods involved in the studies, it is difficult Program through the National Research Foundation of Evidence-Based Complementary and Alternative Medicine 7

Korea (NRF) funded by the Ministry of Education (NRF- [17] T. Murotani, T. Ishizuka, H. Nakazawa et al., “Possible 2016R1D1A1A02937335). involvement of histamine, dopamine, and noradrenalin in the periaqueductal gray in electroacupuncture pain relief,” Brain Research,vol.1306,pp.62–68,2010. References [18]C.Deluze,L.Bosia,A.Zirbs,A.Chantraine,andT.L.Vischer, [1] G. Aston-Jones and J. D. Cohen, “An integrative theory of locus “Electroacupuncture in fibromyalgia: Results of a controlled coeruleus-norepinephrine function: adaptive gain and optimal trial,” British Medical Journal,vol.305,no.6864,pp.1249–1252, performance,” Annual Review of Neuroscience,vol.28,pp.403– 1992. 450, 2005. [19] S. K. Kim, J. H. Park, and S. J. Bae, “Effects of electroacupuncture [2]W.D.WillisandK.N.Westlund,“Neuroanatomyofthepain on cold allodynia in a rat model of neuropathic pain: mediation Experimental system and of the pathways that modulate pain,” Journal of by spinal adrenergic and serotonergic receptors,” Ne Clinical Neurophysiology,vol.14,no.1,pp.2–31,1997. urology,vol.195,no.2,pp.430–436,2005. [20] S. K. Kim, H. J. Moon, J. H. Park et al., “The maintenance of [3] A. A. Dahlstrom, “Evidence for existence of monoamine con- individual differences in the sensitivity of acute and neuropathic taining neurons in the central nervous system. 1. Demonstration pain behaviors to electroacupuncture in rats,” Brain Research of monoamines in the cell bodies of brain stem neurons,” Acta Bulletin, vol. 74, no. 5, pp. 357–360, 2007. Physiologica Scandinavica,vol.62,pp.1–55,1964. [21] W. Kim, S. K. Kim, and B.-I. Min, “Mechanisms of [4] A. Pertovaara, “Noradrenergic pain modulation,” Progress in electroacupuncture-induced analgesia on neuropathic pain in Neurobiology,vol.80,no.2,pp.53–83,2006. animal model,” Evidence-Based Complementary and Alternative [5] B. E. Jones, “Activity, modulation and role of basal forebrain Medicine, vol. 2013, Article ID 436913, 11 pages, 2013. cholinergic neurons innervating the cerebral cortex,” Progress [22] K. C. Tam and H. H. Yiu, “The effect of acupuncture on essential in Brain Research,vol.145,pp.157–169,2004. hypertension,” American Journal of Chinese Medicine,vol.3,no. [6] C. W. Berridge and B. D. Waterhouse, “The locus coeruleus- 4, pp. 369–375, 1975. noradrenergic system: modulation of behavioral state and state- [23] H.-J. Lee, B. Lee, S.-H. Choi et al., “Electroacupuncture reduces dependent cognitive processes,” Brain Research Reviews,vol.42, stress-induced expression of c-Fos in the brain of the rat,” no. 1, pp. 33–84, 2003. American Journal of Chinese Medicine,vol.32,no.5,pp.795– [7] S. P. Hume, A. A. Lammertsma, J. Opacka-Juffry et al., “Quan- 806, 2004. tification of in vivo binding of [3H]RX 821002 in rat brain: [24] J. S. Han and L. Terenius, “Neurochemical basis of acupuncture Evaluation as a radioligand for central 𝛼2-adrenoceptors,” Inter- analgesia.,” Annual Review of Pharmacology and Toxicology,vol. national Journal of Radiation Applications and Instrumentation. 22, pp. 193–220, 1982. Part B. Nuclear Medicine and Biology,vol.19,no.8,pp.841–849, 1992. [25] J. H. Lee and A. J. Beitz, “The distribution of brain-stem and spinal cord nuclei associated with different frequencies of [8] S. J. Sara, “The locus coeruleus and noradrenergic modulation electroacupuncture analgesia,” PAIN,vol.52,no.1,pp.11–28, of cognition,” Nature Reviews Neuroscience,vol.10,no.3,pp. 1993. 211–223, 2009. [26] Y.-B. Kwon, M.-S. Kang, C.-J. Ahn, H.-J. Han, B.-C. Ahn, and J.- [9] G. Aston-Jones and F. E. Bloom, “Activity of norepinephrine- H. Lee, “Effect of high or low frequency electroacupuncture on containing locus coeruleus neurons in behaving rats anticipates the cellular actitivy of catecholaminergic neurons in the brain fluctuations in the sleep-waking cycle,” The Journal of Neuro- stem,” Acupuncture & electro-therapeutics research,vol.25,no. science,vol.1,no.8,pp.876–886,1981. 1, pp. 27–36, 2000. [10] A. F. Sved and G. Felsten, “Stimulation of the locus coeruleus [27] A. Li, Y. Wang, J. Xin et al., “Electroacupuncture suppresses decreases arterial pressure,” Brain Research,vol.414,no.1,pp. hyperalgesia and spinal Fos expression by activating the 119–132, 1987. descending inhibitory system,” Brain Research,vol.1186,no.1, [11] R. J. Valentino and E. van Bockstaele, “Convergent regulation pp.171–179,2007. of locus coeruleus activity as an adaptive response to stress,” [28] Z.-Y. Qiu, Y. Ding, L.-Y. Cui, M.-L. Hu, and M.-X. Ding, “The EuropeanJournalofPharmacology,vol.583,no.2-3,pp.194– expression patterns of c-Fos and c-Jun induced by different 203, 2008. frequencies of electroacupuncture in the brain,” Evidence-Based [12] M. J. Millan, “Descending control of pain,” Progress in Neurobi- Complementary and Alternative Medicine,vol.2015,ArticleID ology,vol.66,no.6,pp.355–474,2002. 343682, 10 pages, 2015. [13]G.Stux,B.Berman,andB.Pomeranz,Basics of Acupuncture, [29] M.-L. Hu, Z.-Y. Qiu, K. Hu, and M.-X. Ding, “Analgesic Springer, Berlin, Germany, 2003. neural circuits are activated by electroacupuncture at two sets [14]C.P.O.CarlssonandB.H.Sjolund,¨ “Acupuncture for chronic of acupoints,” Evidence-Based Complementary and Alternative low back pain: a randomized placebo-controlled study with Medicine, vol. 2016, Article ID 3840202, 10 pages, 2016. long-term follow-up,” The Clinical Journal of Pain,vol.17,no. [30] V. Napadow, R. Dhond, K. Park et al., “Time-variant fMRI 4, pp. 296–305, 2001. activity in the brainstem and higher structures in response to [15] A. White, N. E. Foster, M. Cummings, and P. Barlas, “Acupunc- acupuncture,” NeuroImage,vol.47,no.1,pp.289–301,2009. ture treatment for chronic knee pain: a systematic review,” [31] X.-D. Cao, S.-F. Xu, and W.-X. Lu, “Inhibition of sympa- Rheumatology, vol. 46, no. 3, pp. 384–390, 2007. thetic nervous system by acupuncture,” Acupuncture & Electro- [16] A. J. Vickers, R. W. Rees, C. E. Zollman et al., “Acupuncture for Therapeutics Research,vol.8,no.1,pp.25–35,1983. chronic headache in primary care: large, pragmatic, randomised [32] S. Wang and X. Wang, The Inhibitory Effect of Acupuncture trial,” British Medical Journal, vol. 328, no. 7442, pp. 744–747, on L-Dopa-Induced Hyperactivity of Rat’s Bladder,International 2004. Congress Series, Elsevier, 2002. 8 Evidence-Based Complementary and Alternative Medicine

[33] C. H. Yang, B. B. Lee, H. S. Jung, I. Shim, P. U. Roh, and G. T. Golden, “Effect of electroacupuncture on response to immobilization stress,” Pharmacology Biochemistry & Behavior, vol. 72, no. 4, pp. 847–855, 2002. [34] H. Wang, Y. Tanaka, H. Seki et al., “Acupuncture stimulation to the sacral segment affects state of vigilance in rats,” Neuroscience Research,vol.57,no.4,pp.531–537,2007. [35]H.-J.Park,H.Y.Kim,D.-H.Hahm,H.Lee,K.-S.Kim,andI. Shim, “Electroacupuncture to ST36 ameliorates behavioral and biochemical responses to restraint stress in rats,” Neurological Research, vol. 32, no. 1, pp. S111–S115, 2010. [36] A. M. Wong, T.-Y. Su, F.-T. Tang, P.-T. Cheng, and M.-Y. Liaw, “Clinical Trial of Electrical Acupuncture on Hemiplegic Stroke Patients,” American Journal of Physical Medicine & Rehabilitation,vol.78,no.2,pp.117–122,1999. [37] J. Shen, N. Wenger, J. Glaspy et al., “Electroacupuncture for control of myeloablative chemotherapy-induced emesis: a randomized controlled trial,” Journal of the American Medical Association,vol.284,no.21,pp.2755–2761,2000. [38]G.-T.Gim,J.-H.Lee,E.Parketal.,“Electroacupunctureatten- uates mechanical and warm allodynia through suppression of spinal glial activation in a rat model of neuropathic pain,” Brain Research Bulletin, vol. 86, no. 5-6, pp. 403–411, 2011. [39] S. K. Kim, G. Lee, M. Shin et al., “The association of serum leptin with the reduction of food intake and body weight during electroacupuncture in rats,” Pharmacology Biochemistry &Behavior,vol.83,no.1,pp.145–149,2006. [40] R. Torres-Rosas, G. Yehia, G. Pena, P. Mishra, M. Del Rocio Thompson-Bonilla, and M. A. Moreno-Eutimio, “Dopamine mediates vagal modulation of the immune system by electroacupuncture,” Nature Medicine, vol. 20, no. 3, pp. 291–295, 2014. [41] Z.-Q. Zhao, “Neural mechanism underlying acupuncture analgesia,” Progress in Neurobiology, vol. 85, no. 4, pp. 355–375, 2008. [42] M. A. Medeiros, N. S. Canteras, D. Suchecki, and L. E. A. M. Mello, “c-Fos expression induced by electroacupuncture at the Zusanli point in rats submitted to repeated immobilization,” Brazilian Journal of Medical and Biological Research,vol.36,no. 12, pp. 1673–1684, 2003. [43] Q. Wang, L. Mao, and J. Han, “The arcuate nucleus of hypothalamus mediates low but not high frequency electroacupuncture analgesia in rats,” Brain Research,vol.513,no.1,pp.60–66,1990. [44]D.Sugiyama,S.W.Hur,A.E.Pickeringetal.,“Invivopatch- clamp recording from locus coeruleus neurones in the rat brainstem,” The Journal of Physiology,vol.590,no.10,pp.2225– 2231, 2012. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 7618419, 6 pages https://doi.org/10.1155/2017/7618419

Research Article Understanding Mind-Body Interaction from the Perspective of East Asian Medicine

Ye-Seul Lee,1 Yeonhee Ryu,2 Won-Mo Jung,1 Jungjoo Kim,1 Taehyung Lee,1 andYounbyoungChae1

1 Acupuncture and Meridian Science Research Center, College of Korean Medicine, Kyung Hee University, Seoul, Republic of Korea 2KM Fundamental Research Division, Korea Institute of Oriental Medicine, Daejeon, Republic of Korea

Correspondence should be addressed to Younbyoung Chae; [email protected]

Received 8 June 2017; Accepted 20 July 2017; Published 22 August 2017

AcademicEditor:MarkMoss

Copyright © 2017 Ye-Seul Lee et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Objective. Attempts to understand the emotion have evolved from the perspective of an independent cognitive system of the mind to that of an interactive response involving the body. This study aimed to quantify and visualize relationships between different emotions and bodily organ systems from the perspective of East Asian medicine. Methods. Term frequency-inverse document frequency (tf-idf) method was used to quantify the significance of Five Viscera and the gallbladder relative to seven different emotions through the classical medical text of DongUiBoGam. Bodily organs that corresponded to different emotions were visualized using a body template. Results. The emotions had superior tf-idf values with the following bodily organs: anger with the liver, happiness with the heart, thoughtfulness with the heart and spleen, sadness with the heart and lungs, fear with the kidneys and the heart, surprise with the heart and the gallbladder, and anxiety with the heart and the lungs. Specific patterns between the emotions and corresponding bodily organ systems were identified. Conclusion. The present findings will further the current understanding of the relationship between the mind and body from the perspective of East Asian medicine. Western medicine characterizes emotional disorders using “neural” language while East Asian medicine uses “somatic” language.

1. Introduction instantaneousinteractionswithinthebodythatallowfor the experience of an emotion to be defined as a mental Throughout the history of science, efforts and discussion recognitionaswellasafeelingwithinthebody. aimed at comprehending emotions in relation to the body The intangible nature of emotion has been a subject of havefocusedonthemanifestationofemotionsthroughphys- interest since the beginning of recorded history. Although iological responses [1–4]. Subsequent to James and Lange scholars from diverse fields such as medicine, neuroscience, proposing that physiological feedback plays a role in emotion and anthropology have taken many different approaches [5, 6], more recent findings have indicated that there are to address this issue, the relationship between emotion close relationships between emotions and bodily responses (psychologicalprocesses)andthebody(somaticsystem)has during the formation of mental experiences. This process, been a primary focus. The thoughts and language of modern which may also be referred to in terms of feelings, provides Western scholars are strongly influenced by the dichotomy evidence that physiological reactions such as the activation of the psyche and the soma, which implies the superiority of of the cardiovascular, skeletomuscular, neuroendocrine, and the intentional mind over the body [10]. On the other hand, autonomic nervous systems and their somatosensory feed- East Asian perspectives understand emotions and emotional back mechanisms trigger emotional experiences [4, 7, 8]. disorders through the body and associate emotions with Furthermore, it has been demonstrated that specific patterns different parts of the bodily system, such as the liver or of bodily sensations are related to categorical emotions [9]. heart [11–13]. Moreover, East Asian medicine interprets the A recent perspective on the relationships between emotion human body and its disorders, including emotional and and bodily responses indicates that there are direct and psychosomatic disorders, using a holistic approach [14]. In 2 Evidence-Based Complementary and Alternative Medicine other words, East Asian medicine understands emotions in widely considered to be the representative doctrine of Korean terms of their relationships with different visceral systems medicine [21]. This text serves as the fundamental doctrinal in the body [15]. Although imaginary in nature, metaphors source of medical classics such as the Huangdi Neijing involving internal organs that evoke imaginary bodily images and ShangHan Lun as well as more recent medical texts, are not really arbitrary but appear to have bodily and/or including the Wanbinghuichun (1587) from China and the psychological bases [16]. This approach enables one to under- HyangYakJipSeongBang (1443) and EuiRimChwalYo (1567) standthemindandbodywithinaunifiedframeworkin from Korea. In fact, the DongUiBoGam was written as the which different functions and components of the bodily summation of East Asian medicine knowledge up to the year system are related to corresponding categorical emotions. of its compilation and served as the core text for medical theo- In East Asian medicine, visceral systems do not exactly ries and perspectives as understood in Korean medicine [22]. correspond to anatomical organs as described in modern The essence of health as depicted in the DongUiBoGam human anatomy but, rather, reflect functional systems that emphasizestheharmonyoftheFiveVisceraandthecircu- encompass many different functional realms [17]. The close lation of 𝑞𝑖 throughout the body. The picture of ShinHyung- relationships among these systems allow for the direct asso- JangBuDo on the first page of the DongUiBoGam illustrates ciation of emotions with specific physical symptoms within this concept by emphasizing the body organs, or viscera, and certain visceral systems as well as an understanding of the major pathway through which the 𝑞𝑖 passes by breathing emotional regulation via the visceral system, as perceived in [22]. Thus, in the DongUiBoGam,emotionsarecategorized East Asian medicine by the movement of 𝑞𝑖 [18]. Two different into seven different groups: anger, happiness, thoughtfulness, exemplar sentences from Huangdi Neijing,whichhasbeenthe sadness, anxiety, fear, and surprise. Furthermore, each of fundamental doctrinal source of East Asian medicine for two these categorical emotions is described in terms of its close millennia, illustrate a singular relationship between emotion 𝑞𝑖 and the visceral system: “the liver is in charge of anger” and relationship to the visceral system and the flow of . “anger damages liver.” These sentences imply a bidirectional and instantaneous association between an emotion and the 2.2. Database Construction. We included three chapters of body under normal and pathogenic conditions of health. The text in the DongUiBoGam:thefirstchapter,NaeGyeong Huangdi Neijing further describes the manifestation of this (internal bodily elements), addresses internal body elements relationshipinthebodyasitismediatedby𝑞𝑖 movement such such as essence, 𝑞𝑖,spirit,andtheFiveVisceraandSixBowels; that “anger (is) the 𝑞𝑖 movement surging upwards.” From the the second chapter, Waehyung (external bodily elements), perspective of East Asian medicine, the liver is involved in discusses disorders and symptoms that manifest in external sending 𝑞𝑖 upwards during its circulation, and excerpts from body parts; and the last chapter, Japbyung (miscellaneous this classical medical text elaborately describe the manner in disorders), includes descriptions of miscellaneous diseases which the liver, anger, and the ascending movement of 𝑞𝑖 and details the causes of these symptoms or disorders. are different phenomena that are part of a single systemic Different parts of the chapters and subchapters elaborately function within the body [15]. Although it is invisible, 𝑞𝑖 is discuss categorical emotions and some passages illustrate the felt throughout the body via sensations such as heat and cold manner in which emotions can affect the bodily state and or numbness and pain. Furthermore, according to East Asian result in different symptoms that might also cause various medicine, because the movement of 𝑞𝑖 is governed by the health-related issues. Other passages describe how the vis- visceral systems of the liver, heart, spleen, lungs, and kidneys, ceral system regulates emotions and shows how sicknesses in itcanbedefinedasapartofthebodysystemthatoperates these visceral systems can cause irregular emotional states or through different processes and that is recognized through mental experiences within a patient. bodily sensations [19]. While recent scientific research has Using these passages as the data source, information indicated that specific bodily sensations are related to specific on emotions from the perspective of Korean medicine emotions, East Asian medicine has demonstrated that certain was collected as the dataset. Subsequently, mentions of the movements of 𝑞𝑖 can be recognized as specific categorical visceral system in these passages were collected and the emotions [9, 15]. Additionally, East Asian medicine relates to data were categorized according to different emotions and emotions through the Five Viscera rather than the brain [20]. different viscera. The collected data consisted of various Thus,thepresentstudyexaminedtherelationship groups of passages mentioning different categorical emotions between emotion and the body from the perspective of (anger, happiness, thoughtfulness, sadness, anxiety, fear, and East Asian medicine using a data mining process in a surprise) according to the visceral systems of the liver, representative classical medical text. This study aimed to heart, spleen, lung, kidneys, and gallbladder; notably, the describe how the relationship between emotion and the body gallbladder is the only organ system among the Six Bowels is understood as a whole by quantifying and visualizing these that is described in relation to emotions. The text of the relationships in the classical medical text of DongUiBoGam. DongUiBoGam explains the symptoms of disorders that are caused by either excessive or deficient activity within the 2. Methods gallbladder system through certain emotions, such as anger and surprise; as a result, the present study included the 2.1. Data Source. The classical medical text known as five visceral organs and the gallbladder in its analyses. After DongUiBoGam was first published in 1613 by a royal physi- removing duplicate passages, a total of 334 sentences were cian, Heo Jun (1539–1615), under the Joseon Dynasty and is analyzed in the present study. Evidence-Based Complementary and Alternative Medicine 3

Anger Happiness oughtfulness Sadness Fear Surprise Anxiety 1 Liver 0.8 Heart

Spleen 0.6

Lungs 0.4

Kidneys 0.2

Gallbladder 0 Figure 1: Characteristics of the relationships between emotions and the visceral systems. Emotions are presented on the 𝑥-axis of the array and the organs of the visceral system are presented on the 𝑦-axis of the array. The following emotions had superior tf-idf values with the following bodily organs: anger with the liver, happiness with the heart, thoughtfulness with the heart and spleen, sadness with the heart and lungs, fear with the kidneys and the heart, surprise with the heart and the gallbladder, and anxiety with the heart and the lungs.

2.3. Data Mining and Visualization. Using the above-de- visceral system was only described if it exhibited a tf-idf scribed data construction, data regarding the cooccurrence of value > 0.2. The calculated values were overlaid on a human frequencies and visceral systems within emotional categories body template using matplotlib, which is a plotting library were extracted to understand relationships between these for python (http://matplotlib.org/); a variety of relationships variables. In other words, the specific viscera that were between specific emotions and specific visceral systems were more meaningfully associated with one category of emotion labeled according to the tf-idf(V,𝑒) values. were determined by the application of a term frequency- inverse document frequency (tf-idf) weighting scheme to the cooccurrence table. 3. Results The tf-idf method is one of the most widely used weight- 3.1. Characteristics of the Relationships between Emotions ing schemes in the data mining research field, especially and the Visceral Systems. The bodily organ systems, namely, for information retrieval systems, because it quantifies the the Five Viscera and the gallbladder, were highly associated significance of particular terms in a document [23]. Using with the seven emotions (Figure 1). More specifically, anger this system, the present study quantified the significance of showed superior tf-idf values with the liver (0.94), happiness the associations between mentions of the body system and showed superior tf-idf values with the heart (0.99), thought- emotions. In the tf-idf scheme, term frequency (tf [𝑡 ⋅ )𝑑] fulness showed superior tf-idf values with the heart (0.84) refers to the number of times that the term “𝑡”occursin and spleen (0.54), sadness showed superior tf-idf values with document “𝑑” and, therefore, tf (𝑡,𝑑) represents how relevant the heart (0.81) and lungs (0.56), fear showed superior tf- term “𝑡”istodocument“𝑑.” Document frequency (𝑑𝑓𝑡)is idf values with the kidneys (0.76), heart (0.45), liver (0.33), thenumberofdocumentsthatcontaintheterm“𝑡”and, and gallbladder (0.33), surprise showed superior tf-idf values therefore, 𝑑𝑓𝑡 represents the rarity of a term within the system with the heart (0.97) and the gallbladder (0.23), and anxiety of documents. Across the document system, rare terms are showed superior tf-idf values with the heart (0.90) and lungs more informative than frequent terms and, thus, the inverse (0.41). document frequency of “𝑡”(idf𝑡) is positively related to the informativeness of “𝑡.” Arithmetically, idf is defined as log(𝑁/𝑑𝑓𝑡) instead of 𝑁/𝑑𝑓𝑡 where 𝑁 is the number of whole 3.2. Visualization of Specific Patterns between an Emotion documents in order to diminish the effect of idf. Thus, in and the Human Body. The visualization of emotions ona the present study, tf-idf(V,𝑒) was defined by assigning visceral human body template created for the present study (Figure 2) organs as the “term” and emotions as the “document” so that revealed that specific patterns existed between the visceral the equation quantified the significance of the relationship system and corresponding emotions such that anger corre- between a specific visceral system and a specific emotion. sponded with the liver, happiness with the heart, thought- Based on the tf-idf(V,𝑒) values, each emotion was repre- fulness with the heart and spleen, sadness with the heart sented by a vector of the tf-idf weights in a 6-dimensional and lungs, fear with the kidneys, heart, liver, and gallbladder, vector space. Next, the calculated tf-idf weights of each emo- anxiety with the heart and lungs, and surprise with the heart tion were normalized using the cosine normalization. All data and gallbladder. Furthermore, the present findings showed mining was done using the R package (https://r-project.org) that the heart had significant associations with most of the andtherelationshipbetweenanemotionandaparticular emotions listed, other than anger and fear. 4 Evidence-Based Complementary and Alternative Medicine

1 Anger Happiness oughtfulness Sadness Fear Surprise Anxiety 0.8 Lung 0.6 Heart Liver Spleen 0.4

Kidney Gallbladder 0.2 0 Figure 2: Visualization of specific patterns between an emotion and the human body. Seven categorical emotions were visualized onahuman body template that included the lungs, heart, liver, gallbladder, spleen, and kidneys. Specific patterns existed between the visceral system and corresponding emotions.

4. Discussion mind as illustrated by previous studies. By understanding the emotions in the context of the body itself rather than in a The present study used a data mining procedure to analyze dichotomized model, the relationship of the mind and the relationships between emotions and the visceral system body is horizontal [10]. These findings also showed that the according to the principles of East Asian medicine. Based on heartwassignificantlyprevalentinmostemotions,which the normalized tf-idf values for the frequency of cooccur- implies that this organ is considered to be a common visceral rences between the seven categorical emotions and six bodily system involved in the experience of emotions. Interestingly, organs, including the Five Viscera and the gallbladder, each many other East Asian medical texts also consider the heart of the categorical emotions was related to a specific bodily organ. Anger was related to the liver, happiness to the heart, to be the center of the mind and emotions. For example, thoughtfulness to the heart and spleen, sadness to the heart the sentences “[The] heart stores the mind” and “Sadness, and lungs, fear to the kidneys, heart, liver, and gallbladder, thoughtfulness, and worries all damage the heart” signify the surprise to the heart and the gallbladder, and anxiety to the role of the heart in the processing of thoughts and emotions. heart and the lungs. It is also important to note that Porkert pointed out that The present findings also demonstrated that specific pat- the visceral systems in East Asian medicine, such as the terns existed between the visceral system and corresponding “liver,” “heart,” and “spleen,” do not refer to these specific emotions, which suggests that each emotion is primarily anatomical substrates but, rather, to a certain pattern of associated with a corresponding body system and can also functions within the body [26, 27]. The organ-based nosology be explained by the principles of East Asian medicine. used in East Asian medicine is a metaphor for which the Additionally, the patterns observed in the present study primary referent is not a particular anatomic organ but were similar to theories from East Asian medicine [15]. For an emotion, diagnosed based on the patterns of somatic example, the DongUiBoGam states the following: “Liver is symptoms. Thus, East Asian medicine is built on a symptom- in charge of anger, heart is in charge of happiness, spleen is basedlanguageratherthananorgan-basedlanguage[10]. in charge of thoughtfulness, lungs are in charge of sadness, Dualistic thought limits the bodily perceptions as well as and kidneys are in charge of fear.” The quantification of the bodily awareness, which dichotomizes our selfhood from our terms used to explain the relationships between emotion body, that is, the mind thinking of the body rather than the and bodily organs in this classical medical text produced body perceiving itself [10]. resultsthatweresimilartotheprinciplesofEastAsian Although the direct relationship between emotions and medicine. Furthermore, imbalances in emotions that can the body is characterized differently by East Asian and lead to illnesses in their corresponding organs can also be Western medicine traditions, efforts to overcome the limits of explained by the DongUiBoGam, which describes in detail the mind-body dichotomy have been continuously suggested the manner in which damage to these organ systems and in the West. For example, Merleau-Ponty termed the lived the gallbladder are manifested through excessive or abnormal body as the “body in human experience” or the “perceived emotional states. For example, this text states the following: body” and goes on to explain that this refers to the projection “Anger damages liver, happiness damages heart, thoughts ofthebodyintotherealworld[28].Perceptionisthestarting damage spleen, anxiety damages lungs, fear damages kidneys, point of consciousness and begins in the body before it is and surprise damages gallbladder.” projected to other objects. The somatization of emotional Thus, the present study provides evidence, from the symptoms is not unique in East Asian countries and can perspective of East Asian medicine, that emotions are related be identified across a broad range of cultural backgrounds to visceral organs, a finding that differs from the perspective [19]. Efforts to understand the close and complex relationship of Western medicine in which emotions are understood in between the mind and body are not limited to philosophy and terms of their relationship with specific brain areas, such anthropology. Recent studies in the field of neuroscience have as the amygdala [24, 25]. The comprehensive role of the suggested that there are mutual interactions between bodily heart in East Asian medicine which includes the role of the responses and emotions in which physical functions trigger brain in Western medicine implicates the somatization of the emotional experiences and emotional experiences lead to Evidence-Based Complementary and Alternative Medicine 5 particular spatial patterns of sensation throughout the body Western medicine limits the manifestation of these symptoms [4, 8, 9]. to being described as “cultural” disorders. When translating As Kleinman explained, medical approaches involve “the “the notions about an episode of sickness and its treatment notions about an episode of sickness and its treatment that that are employed by all those engaged in the clinical areemployedbyallthoseengagedintheclinicalprocess” process,” a comprehensive definition of the sickness and its [15, 29]. During this process, notions about illness and symptoms is required, within its medical understanding [29]. treatment are derived from the medical understanding of a By focusing on the manifestation of the mind in the body particular culture, and East Asian medicine has traditionally using the principles of East Asian medicine, the present study employed a unique understanding of the body through its was an attempt to translate the understanding of emotion cultural and social conditions. Furthermore, it is interesting as one aspect of bodily functions. A careful translation and to note that the concept of emotions manifesting through interpretation of the knowledge and theories are necessary the body, as elaborated in the classical medical texts of regarding the systems and functions of the body require East Asian medicine and observed in the present study, has attentionbeforethetwoperspectivesaretranslatedintoeach evolved over time and now exhibits an understanding that other’s languages. Previous study has shown through a time incorporates the influence of Western medicine. Additionally, series analysis that the emotions and physical states are medical perspectives that understand emotions as an aspect related in a long-term observation even in healthy volun- of bodily function or dysfunction within visceral systems teers, implying for clinical relevance of emotional states and have triggered further questions from the perspective of physical states [31]. Furthermore, manipulations of muscular Western medicine in East Asia. The modern articulation feedback from facial expression can modulate the emotional of emotional symptoms in biomedical disease categories, states, including autonomic responses to the emotional cues such as depression, is an example of the response to the [32]. In this regard, a discussion about these two medical embodiment of an illness. Previous study pointed out that viewpoints needs to be established with a comprehensive the historical and modern views of East Asian medicine understanding of the concepts of East Asian medicine in regarding the relationship between emotional disorders and terms of sicknesses and psychosomatic symptoms. the body indicate that this relationship is man-made and The present study has several limitations. First, the not an objectively defined fact [30]. Emotional diseases that characteristics of emotions as they are related to bodily are related to bodily organ systems, such as “Hwa-byung” in organ systems were extracted from a single classical Korean Korea and “Utsu” in Japan, which are thought to be based medical text with limited bibliographical data. Thus, it will be on the influences of modernized East Asian medicine, are necessary to further investigate the characteristics of emotion examples of “cultural” emotional disorders that have been from the perspective of East Asian medicine using other newly defined in the modern era. classical medical texts including Huangdi Neijing.Second, Attempts to understand emotional symptoms through the available data only allowed for the present data mining somatization, or in relation to bodily organs, in East Asian method to be applied to the relationship between emotion medicine have resulted in the classification of emotional and the visceral system. Studies on the relationships between diseases within frameworks that are similar to those of emotion and bodily changes, including physical symptoms Western medicine, even though the direct association of and 𝑞𝑖 movement, need to be conducted to understand the emotional symptoms with bodily organs has been deemed manifestation of emotion in East Asian medicine and to “cultural.” However, attempts to understand such disorders understand the different approaches to emotion used by in the modern East using Western medical terms have also Western medicine and East Asian medicine. led to diagnoses of neurasthenia and depression and the In summary, the present study identified specific patterns conceptualization of these disorders as “cultural” disorders between emotions and corresponding bodily organ systems. has only recently been suggested; in fact, the uniquely Korean East Asian medicine directly associates an emotion with attributes of Hwa-byung have never been established by the visceral system using the language of symptoms rather a clear consensus [12, 13]. Similarly, the interpretation of than that of organs, which can be divided into visible parts. emotion-related disorders as predominantly organ system Thus, the different understanding of emotions and their disorders began in modern Japan during the Republican Era, relationship to the body in Western medicine and East which led to the definition of Utsu as “cultural” disorder [13]. Previous studies have argued that the issue of translation Asian medicine have led to distinct interpretations of illness in cultural medical history has not only brought about a such that Western medicine understands emotional disor- linguistic expansion but also led to the conceptual fusion of ders using neural language while East Asian medicine uses new ideas [13]. For example, the defining of both Hwa-byung somatic language. Understanding these phenomena as well as and Utsu raised questions about what was specifically cultural the actors of medicine can offer a more comprehensive per- to disorders and in what ways they related to biomedical dis- spective when examining the human body and its symptoms. ease categories and traditional disease concepts of emotional disorders [30]. Disclosure Western medicine strives to understand humans based on a dichotomy of the psyche and the soma and, as a result, The funders had no role in study design, data collection trying to understand emotional symptoms using East Asian and analysis, decision to publish, or preparation of the practices based on the interpretations and classifications of manuscript. 6 Evidence-Based Complementary and Alternative Medicine

Conflicts of Interest [16] N. Yu, “Body and emotion: body parts in Chinese expression of emotion,” Pragmatics and Cognition,vol.10,no.1-2,pp.341–367, The authors declare that no competing financial interests or 2002. conflicts of interest exist. [17] Y.-T. Gao, H.-W. Pan, and S.-B. Wu, “Gross conception of anatomical structure of zang-fu viscera in Huangdi Neijing,” Acknowledgments ZhongXiYiJieHeXueBao,vol.4,no.4,pp.339–342,2006. [18]D.-Z.Sun,S.-D.Li,Y.Liu,Y.Zhang,R.Mei,andM.-H. ThisresearchwassupportedbytheKoreaInstituteofOriental Yang, “Differences in the origin of philosophy between Chinese Medicine (no. K15070) and by NRF (National Research Foun- medicine and western medicine: exploration of the holistic dation of Korea) grant funded by the Korean Government advantages of chinese medicine,” Chinese Journal of Integrative (NRF-2015-Global Ph.D. Fellowship Program). Medicine, vol. 19, no. 9, pp. 706–711, 2013. [19] D. Hinton and S. Hinton, “Panic disorder, somatization, and the new cross-cultural psychiatry: the seven bodies of a medical References anthropology of panic,” Culture, Medicine and Psychiatry,vol. 26,no.2,pp.155–178,2002. [1] S. Schachter and J. E. Singer, “Cognitive, social, and physiolog- [20] N. Yu, The Chinese Heart in aCognitive Perspective: Culture, ical determinants of emotional state,” Psychological Review,vol. Body, and Language, Walter de Gruyter, Berlin, Germany, 2009. 69,no.5,pp.379–399,1962. [21] W.-S. Cha, J.-H. Oh, H.-J. Park, S.-W. Ahn, S.-Y. Hong, and [2] A.BecharaandN.Naqvi,“Listeningtoyourheart:interoceptive N.-I. Kim, “Historical difference between traditional Korean awareness as a gateway to feeling,” Nature Neuroscience,vol.7, medicine and traditional Chinese medicine,” Neurological no. 2, pp. 102-103, 2004. Research,vol.29,no.1,pp.S5–S9,2007. [3]J.T.Cacioppo,G.G.Berntson,J.F.Sheridan,andM.K. [22] T. Lee, W.M. Jung, I. S. Lee, and etal., “Data mining of acupoint McClintock, “Multilevel integrative analyses of human behav- characteristics from the classical medical text: dongUiBoGam ior: social neuroscience and the complementing nature of social of Korean medicine,” Evidence-Based Complementary and Alter- and biological approaches,” Psychological Bulletin,vol.126,no. native Medicine, vol. 2014, Article ID 329563, 10 pages, 2014. 6, pp. 829–843, 2000. [23]W.-M.Jung,T.Lee,I.-S.Leeetal.,“Spatialpatternsofthe [4] A. Damasio and G. B. Carvalho, “The nature of feelings: indications of acupoints using data mining in classic medical evolutionary and neurobiological origins,” Nature Reviews Neu- text: a possible visualization of the meridian system,” Evidence- roscience,vol.14,no.2,pp.143–152,2013. based Complementary and Alternative Medicine,vol.2015, [5] W. James, “Mind II. What is an Emotion?, 1884”. ArticleID457071,7pages,2015. [6] C. G. Lange, Ueber Gemuethsbewegungen, 1887. [24]T.D.Wager,J.Kang,T.D.Johnson,T.E.Nichols,A.B. [7]A.Bechara,A.R.Damasio,H.Damasio,andS.W.Anderson, Satpute,andL.F.Barrett,“Abayesianmodelofcategory-specific “Insensitivity to future consequences following damage to emotional brain responses,” PLoS Computational Biology, vol. 11, human prefrontal cortex,” Cognition,vol.50,no.1-3,pp.7–15, no. 4, Article ID e1004066, 2015. 1994. [25]A.B.Satpute,J.Kang,K.C.Bickart,H.Yardley,T.D.Wager, [8]R.W.Levenson,“Blood,sweat,andfears:theautonomic andL.F.Barrett,“Involvementofsensoryregionsinaffective architecture of emotion,” Annals of the New York Academy of experience: a meta-analysis,” Frontiers in Psychology,vol.6, Sciences, vol. 1000, pp. 348–366, 2003. Article ID 01860, 2015. [9] L. Nummenmaa, E. Glerean, R. Hari, and J. K. Hietanen, “Bodily [26] M. Porkert and C. Ullmann, Chinese Medicine,HenryHoltand maps of emotions,” Proceedings of the National Academy of Company, NY, USA, 1982. Sciences of the United States of America, vol. 111, no. 2, pp. 646– [27] M. Porkert, The theoretical foundations of chinese medicine: 651, 2014. systems of correspondence, MIT Press, Cambridge, MA, USA, [10]T.Ots,“Theangryliver,theanxiousheartandthemelancholy 1974. spleen. The phenomenology of perceptions in chinese culture,” [28] M. Merleau-Ponty, “Phenomenology of perception, 1962”. Culture, Medicine and Psychiatry,vol.14,no.1,pp.21–58,1990. [29] A. Kleinman, Patients and Healers in the Context of Culture: An [11] V. Scheid, “Constraint as a window on approaches to emotion- Exploration of the Borderland between Anthropology, Medicine related disorders in east asian medicine,” Culture, Medicine and and Psychiatry, University of California Press, Berkeley, Calif, Psychiatry,vol.37,no.1,pp.2–7,2013. USA, 1980. [12] S. Suh, “Stories to Be Told: Korean doctors between hwa-byung [30] V. Scheid, “Depression, constraint, and the liver: (dis)assem- (fire-illness) and depression, 1970–2011,” Culture, Medicine and bling the treatment of emotion-related disorders in Chinese Psychiatry,vol.37,no.1,pp.81–104,2013. medicine,” Culture, Medicine and Psychiatry,vol.37,no.1,pp. [13] K. Daidoji, “Treating emotion-related disorders in Japanese 30–58, 2013. traditional medicine: language, patients and doctors,” Culture, [31] Y. S. Lee, W.M. Jung, H. Jang, S. Kim, S. Y. Chung, and Y. Chae, Medicine and Psychiatry,vol.37,no.1,pp.59–80,2013. “The dynamic relationship between emotional and physical [14] C. Tan, W. Chen, Y. Wu, and S. Chen, “Chinese medicine for states: an observational study of personal health records,” mental disorder and its applications in psychosomatic diseases,” Neuropsychiatric Disease and Treatment,vol.13,pp.411–419, Alternative Therapies In Health And Medicine,vol.19,no.1,pp. 2017. 59–69, 2013. [32] I.-S. Lee, S.-S. Yoon, S.-H. Lee et al., “An amplification of feed- [15] L. Park and D. Hinton, “Dizziness and panic in China: asso- back from facial muscles strengthened sympathetic activations ciated sensations of zang fu organ disequilibrium,” Culture, to emotional facial cues,” Autonomic Neuroscience: Basic and Medicine and Psychiatry,vol.26,no.2,pp.225–257,2002. Clinical,vol.179,no.1-2,pp.37–42,2013. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 1947315, 7 pages https://doi.org/10.1155/2017/1947315

Research Article Effects of Combined Far-Infrared Radiation and Acupuncture at ST36 on Peripheral Blood Perfusion and Autonomic Activities

Cheng-Chan Yang,1,2 Gen-Min Lin,1,3,4 Jen-Hung Wang,5 Hsiao-Chiang Chu,1 Hsien-Tsai Wu,1 Jian-Jung Chen,6,7 and Cheuk-Kwan Sun8

1 Department of Electrical Engineering, National Dong Hwa University, Hualien 97401, Taiwan 2Department of Chinese Medicine, Buddhist Tzu Chi General Hospital, Hualien 97002, Taiwan 3Department of Medicine, Hualien-Armed Forces General Hospital, Hualien 97144, Taiwan 4Department of Medicine, Tri-Service General Hospital, National Defense Medical Center, Taipei 114, Taiwan 5DepartmentofMedicalResearch,BuddhistTzuChiGeneralHospital,Hualien97002,Taiwan 6Department of Chinese Medicine, Taichung Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Taichung 42743, Taiwan 7School of Chinese Medicine, Tzu Chi University, Hualien 97002, Taiwan 8Department of Emergency Medicine, E-Da Hospital, I-Shou University, Kaohsiung 82445, Taiwan

Correspondence should be addressed to Cheuk-Kwan Sun; [email protected]

Received 7 April 2017; Revised 5 July 2017; Accepted 6 July 2017; Published 13 August 2017

Academic Editor: Woojin Kim

Copyright © 2017 Cheng-Chan Yang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Using four-channel photoplethysmography (PPG) for acquiring peripheral arterial waveforms, this study investigated vascular and autonomic impacts of combined acupuncture-far infrared radiation (FIR) in improving peripheral circulation. Twenty healthy young adults aged 25.5±4.6 were enrolled for 30-minute measurement. Each subject underwent four treatment strategies, including acupuncture at ST36 (Zusanli), pseudoacupuncture, FIR, and combined acupuncture-FIR at different time points. Response was assessed at 5-minute intervals. Area under arterial waveform at baseline was defined as AreaBaseline, whereas AreaStim referred to area at each 5-minute substage during and after treatment. AreaStim/AreaBaseline was compared at different stages and among different strategies. Autonomic activity at different stages was assessed using low-to-high frequency power ratio (LHR). Theresults demonstrated increased perfusion for each therapeutic strategy from stage 1 to stage 2 (all 𝑝 < 0.02). Elevated perfusion was noted for all treatment strategies at stage 3 compared to stage 1 except pseudoacupuncture. Increased LHR was noted only in subjects undergoing pseudoacupuncture at stage 3 compared to stage 1 (𝑝 = 0.045). Reduced LHR at stage 2 compared to stage 1 was found only in combined treatment group (𝑝 = 0.041). In conclusion, the results support clinical benefits of combined acupuncture-FIR treatment in enhancing peripheral perfusion and parasympathetic activity.

1. Introduction FIR has also been shown to suppress vascular endothelial proliferation and enhance endothelial repair through sup- Far-infrared radiation (FIR), which comprises electromag- pressing the action of vascular endothelial growth factor [4, netic waves of wavelength 3–1000 𝜇m [1], possesses physiological actions because not only of its high penetrating power 7, 8]. Besides, improvement of wound healing has also been inhumantissuebutalsoofitsabilitytoelicitbothheat- reported through FIR-induced fibroblast recruitment and related [2] and nonheat-related [3] biological effects. It has collagen disposition [9, 10]. Other demonstrated therapeutic been demonstrated that FIR causes vasodilatation, thereby actions of FIR also include the suppression of tumor cell improving human tissue perfusion [2, 4, 5] and skin microp- proliferation and spreading [11–14], reduction of intravas- erfusion in rats through enhancing the action of endothelial cular lipid deposition and risk of arteriosclerosis [15], and nitric oxide synthase, eNOS, in vascular endothelium [6]. improving sleep quality [16]. 2 Evidence-Based Complementary and Alternative Medicine

On the other hand, acupuncture is a traditional Chinese therapeutic approach [17] that has been proposed to exert its actions through neurovascular modulation [18]. It has Gallbladder meridian been accepted as a valid therapeutic option by the World Health Organization (WHO) which listed 64 indications for 4 cm 4 cm the procedure including neurological and vascular diseases Stomach such as seizure, headache, Parkinson’s disease, and stroke [3]. meridian The present study aimed at investigating whether improved therapeutic benefits are achievable through a combination of invasive acupuncture and noninvasive FIR by using the noninvasive tools of photoplethysmography (PPG) and heart rate variability for the assessment of peripheral perfusion and autonomic nervous activities, respectively.

2. Materials and Methods 2.1. Subject Population. Twenty healthy young adult volunteers (17 males and 3 females) were recruited. Individuals withknownsystemicdiseasesaswellashabitsofsmoking Meridian Zusanli (ST36) and drinking were excluded from this study. All participants Acupoint Pseudoacupoint were required to have breakfast but refrained from beverages with alcohol or caffeine within 24 hours of the examinations. Figure 1: Locations of acupoints along the stomach and gallbladder Besides, basic demographic (i.e., age and gender), anthropo- meridians over the lower leg. Acupuncture performed at acupoint metric (i.e., body weight, body height, and body-mass index), Zusanli (ST36) located 4 cm anterior to the gallbladder meridian. Pseudoacupoint chosen on the same horizontal plane 4 cm posterior and hemodynamic (i.e., systolic and diastolic pressure) infor- to gallbladder meridian. mation was recorded from all testing subjects. No acupuncture was performed on all 20 testing subjects one month prior to the present study. The study protocol was reviewed and approved by the Institutional Review Board on Ethics of Bud- received four different interventions during the experimental dhist Tzu Chi General Hospital (IRB number IRB103-152-A). sessions: Session 1 (acupuncture), Session 2 (pseudoacupuncture), Session 3 (far-infrared radiation (FIR)), and Session 4 (acupuncture combined with FIR). During each session, 2.2. Equipment and Data Acquisition. Four-channel photo- the initial 5 minutes are defined as baseline stage (stage 1), plethysmography (PPG) (Lite-On Electronics Co., Tianjin, the following 15 minutes as intervention stage (stage 2), and China) and electrocardiography (ECG) (Clamp Electrode the last 10 minutes as postintervention (i.e., recovery) stage 0102014, Qingdao Bright Medical Manufacturing Co., Ltd., (stage 3). The second stage of sessions 1 and 4 involved the China) were used for acquisition of data on pulse volume insertion of a 4 cm acupuncture needle at Zusanli following andheartratevariability(HRV),respectively.Signalsofblood 5 minutes of baseline recording. The identification of Zusanli flow from PPG were filtered with second-order high-pass and the procedure of acupuncture were in accordance with filter and low-pass filter with cut-off frequencies of 0.48- the standard of traditional Chinese medical practice. With the 10 Hz, while signals from Lead II of ECG went through notch kneeflexed,theacupoint“Dubi”(ST-35)canbeidentifiedas filter (59–61 Hz) and band-pass filter (0.98–19.4 Hz) before the depression lateral to the patellar ligament. Zusanli (ST36) being processed with a analog-to-digital converter (USB-6210 was located on horizontal plane 4 fingerbreadths below DAQ, National Instruments, TX, USA) with a sampling rate “Dubi” and is one fingerbreadth lateral to the anterior border of 500 Hz to 16-bit digital signals which were then stored in a of the tibia on vertical plane. The second stage of session computer with appropriate software (Labview Signal Express 2 (pseudoacupuncture) involved the insertion of a 4 cm 2012, National Instruments, TX, USA) for waveform moni- acupuncture needle at pseudoacupoint following 5 minutes toring and analysis [11]. WS TY-101N emitter (Far IR Medical of baseline recording. The pseudoacupoint was located on the Technology Co., Ltd., Taipei, Taiwan) was used for sessions 3 same horizontal plane as ST36 at a point 8 cm posterior to and 4. The wavelength of the light generated from the electri- 𝜇 ST36 so that ST36 and the pseudoacupoint were each 4 cm fied ceramic plates was in the range between 3 and 25 mwith away from and on both sides of the gallbladder meridian 𝜇 2 apeakat5 m. Intensity is 20 mW/cm at 20 cm distance. (Figure 1). The pseudoacupoint was chosen because it was not located on any meridian to ensure the absence of acupunc- 2.3. Procedures of Examinations. Testing subjects were tural effect. The needle was inserted perpendicularly through allowed to rest in supine position for over 5 minutes theskinwithatwistingmotionofthethumbandindexfinger before data acquisition during which 4 PPG detectors were, with repeated clockwise and anticlockwise rotation of the respectively, attached to bilateral index finger as well as the needle between 90 to 180 degrees till the state of “Deqi” (i.e., a second toe, while the three ECG detectors were, respectively, numbness sensation signifying correct position of the needle) attached to the right wrist and bilateral ankles. Each subject was achieved. The depth of penetration was the full length Evidence-Based Complementary and Alternative Medicine 3

Stage 1 Stage 2-1 Stage 2-2 Stage 2-3 Stage 3-1 Stage 3-2

Baseline Intervention Recovery (needle retention/FIR)

Needle insertion/FIR Needle withdrawal/FIR

0 min 5 min 20 min 30 min Figure 2: Recording lasting for 30 minutes with the first 5 minutes being the baseline (stage 1), 5th to 20th minute being intervention phase (stage 2), and 20th to 30th minute being the recovery phase (stage 3). FIR: far-infrared radiation.

(4 cm) of a 32 G acupuncture needle (Toba acupuncture (HFP) represents parasympathetic activities. The activity needle, Seoul, South Korea). The needle was allowed to stay of the sympathetic nervous system relative to that of the in place for 15 minutes before being removed. An experienced parasympathetic can then be given by the LFP/HFP ratio physician of traditional Chinese medicine was responsible (LHR). for acupuncture in all testing subjects. During the first stage of data acquisition, six sets of 5-minute signals were 2.5. Statistical Analysis. Data are expressed as mean ± SD. obtained. At stage 2 of the study, data were acquired for 15 All statistical analyses were performed using SPSS software minutes after placing the needle in the appropriate position version 17.0 (SPSS Inc., Chicago, IL, USA). Repeated measures through confirming the tingling and mild soreness sensation ANOVA was used for determining the significance of fluctu- of the punctured site (i.e., Deqi)withthetestingsubject.The ation throughout the course of treatment for each therapeutic recording then continued for 10 more minutes after the needle strategy, while paired 𝑡-test was adopted for analyzing the was withdrawn. Therefore, data acquisition lasted for totally significance of difference in measurement parameters for the 30 minutes (Figure 2). same testing subject at different stages of treatment. A 𝑝 value < 0.05 was considered statistically significant. 2.4. Assessment of Peripheral Blood Flow and Autonomic Ner- vous Activities. Physiological signals before, during, and after 3. Results acupuncture were recorded, including area under arterial waveforms(i.e.,volumeofbloodflow)andelectrocardiogram 3.1. Characteristics of Testing Subjects. The mean age of the 25.5 ± 4.6 from which the time periods between two successive R 20 volunteers (17 males and 3 females) was 24.2 ± 4.0 waves (i.e., R-R intervals, RRI) were obtained and processed (range, 19–37). The mean body-mass index was with Fast Fourier Transform (FFT) for determination of (range, 17.6–32.0). The mean systolic and diastolic pressure 124.2 ± 14.5 75.3 ± heart rate variability (HRV). The computation then yielded was mmHg (range, 100–163 mmHg) and 10.6 low frequency (LF) power, high frequency (HF) power, and mmHg (range, 62–98 mmHg), respectively. The mean 76.1 ± 10.2 low-to-high frequency ratio (LHR). The areas under the heart rate was beats per minute (range, 63–106). arterial waveforms within 5 minutes of baseline recording 3.2. Changes in Peripheral Perfusion. AsshowninFigure3,all were summated (AreaBaseline). The areas under the waveforms during 15 minutes of intervention (i.e., acupuncture or FIR) subjects showed a significant increase in perfusion for each and 10 minutes postintervention (i.e., recovery phase) were therapeutic strategy on entering from stage 1 to stage 2 (all divided into 5 parts each of which represented 5 minutes of 𝑝 < 0.02) (Figure 3, Table 1). On the other hand, significant elevation in perfusion was noted for all treatment strategies recordings (Figure 2). AreaStim is defined as the area under the waveforms for each five-minute recording at each substage at stage 3 compared to that at stage 1 with the exception after baseline. The change in blood flow every 5 minutes after of pseudoacupuncture. Considering the overall differences intervention compared with the baseline was then expressed among all treatment strategies at stage 2 and stage 3, no as a fraction: AreaStim/AreaBaseline. Therefore, five values were remarkable difference in perfusion was noted at both stages. obtained from each testing subject after baseline recording. On the other hand, the present study also noninva- 3.3. Changes in LFP/HFP Ratio (LHR). In terms of the impact sively assessed the changes in autonomic nervous activities of treatment strategy on LHR, significant increase in LHR through electrocardiograms by investigating the variations was noted only in subjects undergoing pseudoacupuncture in RRI which, after Fast Fourier Transform (FFT), reflects at stage 3 compared to that at stage 1 (𝑝 = 0.045) the HRV. The power spectrum (0–0.04 Hz) after FFT can but not in the same subjects undertaking other treatment be divided into 4 parts, namely, ultralow frequency power strategies (Figure 4(a), Table 1). Notable drop in LHR at (0–0.003 Hz), very low frequency power (0.003–0.04 Hz), stage 2 compared to that at stage 1 was demonstrated only low frequency power (0.04–0.15 Hz), and high frequency in subjects receiving combined treatment (𝑝 = 0.041). At power (0.15–0.4 Hz). While the low frequency power (LFP) stage 2, subjects undertaking pseudoacupuncture showed reflects sympathetic activities, the high frequency power significantly higher LHR𝑝 ( = 0.013)comparedtothat 4 Evidence-Based Complementary and Alternative Medicine

Table 1: Changes in peripheral perfusion and LHR for four different treatments at different stages in the same group of testing subjects (𝑛=20). ∗ a ∗ b Parameter Treatment Stage 1 Stage 2 𝑝 value Stage 3 𝑝 value Acupuncture 2.20 ± 0.35 2.33 ± 0.36 <0.001 2.29 ± 0.30 0.001 Pseudoacupuncture 2.33 ± 0.30 2.41 ± 0.29 <0.001 2.38 ± 0.26 0.069 Perfusion (aU) FIR 2.30 ± 0.31 2.40 ± 0.35 0.020 2.42 ± 0.37 0.032 Acupuncture-FIR 2.26 ± 0.45 2.37 ± 0.44 <0.001 2.38 ± 0.41 0.004 Acupuncture 1.84 ± 1.14 1.64 ± 0.83 0.442 1.62 ± 0.80 0.306 Pseudoacupuncture 1.58 ± 0.89 1.80 ± 0.85 0.206 2.00 ± 1.17 0.045 LHR FIR 2.07 ± 0.97 1.72 ± 0.88 0.108 2.07 ± 1.89 0.601 Acupuncture-FIR 2.42 ± 1.58 1.74 ± 0.93 0.041 2.15 ± 1.48 0.691 a b aU: arbitrary unit; FIR: far-infrared radiation; LHR: low frequency power to high frequency power ratio. Comparison between stages 1 and 2. Comparison ∗ between stages 3 and 1. Significance of difference determined by paired 푡-test.

1.07 acupuncture and combined acupuncture-FIR groups showed 1.06 adecreaseinLHRfromstage1tostage2,thereductionfailed 1.05 †

＜；Ｍ？ＦＣＨ？ ∗ to reach statistical significance. On the contrary, the pseu- 1.04 doacupuncture group demonstrated an increase in LHR on

/！Ｌ？； 1.03 entering stage 2 in spite of being insignificant. The elevation

ＭＮＣＧ 1.02 in LHR, which signifies decreased parasympathetic activity, wasnotedinthesubjectsundergoingpseudoacupuncture ！Ｌ？； 1.01 𝑝 = 0.045 1 at stage 3 of treatment ( ). The increase in LHR was probably due to pain-induced stimulation of sympathetic 1 2 2 22 3 3 -1 - -3 -1 -2 tone. The findings, therefore, support the clinical benefits Stage of adopting acupuncture and FIR as both monotherapeutic Acupuncture FIR ！＝ＯＪＯＨ＝ＮＯＬ？ + ＆）２ regimen and combined treatment in terms of augmenting Pseudoacupuncture peripheral perfusion and parasympathetic activity. Figure 3: Changes in peripheral perfusion in testing subjects (𝑛= A number of experimental and clinical studies have 20 ) on receiving four different treatment strategies. AreaBaseline: shown that acupuncture enhances the generation of nitric summation of areas under arterial waveforms within 5 minutes oxide (NO) and increases local circulation [19–21]. Several of baseline recording. AreaStim: summation of areas under arterial theories have been proposed to explain the phenomenon, waveforms for each five-minute recording at each substage after including reduction of neuronal apoptosis [22–24], aug- ∗𝑝 baseline. FIR: far-infrared radiation. < 0.02 at stage 2 versus stage mentation of antioxidative activity [25–27], and enhance- †𝑝 < 0.04 1 for each treatment strategy; at stage 3 versus Stage 1 for ment of neurotransmission [28–30]. A previous study in acupuncture, FIR, and combined treatment. a rat model showed that acupuncture at ST36 can trigger the generation of nitric oxide in the gracile nucleus that modulates the control of blood pressure and heart rate when they received FIR (Figure 4(b)) or combined treatment [31]. Besides, acupuncture at ST36 has been found to boost (Figure 4(c)). Both time course and choice of therapeutic intestinal motility in rats by elevating parasympathetic tone strategyhadnosignificantimpactonLHRatstage3. through vagus nerve stimulation [32]. Consistently, it has been reported that ST36 acupuncture in a rat model can 4. Discussion enhance parasympathetic but suppress sympathetic activities [33]. Clinically, a study on the application of transcutaneous The current study represents the first investigation into the electrical nerve stimulation (TENS) at ST36 in patients with therapeutic benefits of a combination of two alternative scleroderma demonstrated an alleviation of gastrointestinal medical approaches to enhancing neurovascular activities symptoms through achieving sympathovagal balance [34]. including peripheral perfusion (i.e., increase in PI) and Another clinical study applying TENS at ST36 in patients parasympathetictone(i.e.,decreaseinLHR).Theresults withfunctionaldyspepsiaalsoshowedimproveddyspeptic showed that FIR, acupuncture, and the combination of both symptoms and increased high frequency heart rate variability could cause an increase in perfusion that persisted through possibly related to an increase in plasma neuropeptide Y stage 2 (i.e., intervention period) and stage 3 (i.e., recovery level [35]. Furthermore, a previous study in healthy subjects period) of treatment, despite the lack of significant difference revealed an increase in skin and muscle blood flow up to 50% among the four treatment strategies. On the other hand, after acupuncture at ST36 for 20 minutes [21]. significant drop in LHR, which represents an increase in Consistently, the current study showed that acupuncture parasympathetic activity, was noted only in subjects receiving could boost peripheral perfusion as previously reported [36, combined treatment at stage 2 (𝑝 = 0.041). Although both 37]. Nevertheless, combination of acupuncture and FIR had Evidence-Based Complementary and Alternative Medicine 5

2 3 † 1.5 2 1 LHR LHR 1 ∗ 0.5 ∗

0 0 1 2-1 2-2 2-3 3-1 3-2 5 101520 Stage Time course (min) Acupuncture FIR Pseudoacupuncture Pseudoacupuncture ！＝ＯＪＯＨ＝ＮＯＬ？ + ＆）２ FIR (a) (b) 3

2 LHR 1 ∗

0 5 10 15 20 Time course (min) Pseudoacupuncture ！＝ＯＪＯＨ＝ＮＯＬ？ + ＆）２

(c)

Figure 4: (a) Changes in LHR at different stages of treatment for subjects undergoing four kinds of therapies. LHR: low frequency power to ∗ high frequency power ratio; FIR: far-infrared radiation. 𝑝 = 0.041 at stage 2 versus stage 1 for subjects receiving acupuncture-FIR combined † treatment; 𝑝 = 0.045 at stage 3 versus stage 1 for subjects undergoing pseudoacupuncture. Significance of difference determined by paired ∗ 𝑡-test. (b) Comparison of changes in LHR between pseudoacupuncture and acupuncture treatment. 𝑝 = 0.012 determined with post hoc † test; (c) comparison of changes in LHR between pseudoacupuncture and combined acupuncture-FIR treatment. 𝑝 = 0.019 determined with post hoc test.

no additional benefit in terms of increasing peripheral blood perfusion enhancement effect during the recovery phase (i.e., flow in this study. stage 3) in subjects receiving acupuncture, FIR, or combined Similar to the vasodilatory action of acupuncture, not treatment. The findings, therefore, are consistent with the only has FIR been found to exhibit heat-related effect that previously reported carryover effects of acupuncture [38] involves resonance in tissue, but it has also been reported and FIR [6]. Nevertheless, further enhancement in perfusion to trigger nonheat-related biochemical effects. Since human during the recovery phase after FIR treatment described in body consists of close to 70% of water, FIR-induced resonance a previous animal study [6] was not noted in the current of water molecules leads to severance of the hydrogen bonds, investigation. causing a cascade of thermally induced reactions [2]. Besides, Heart rate variability (HRV), which is an easily available using a FIR transmission model, it has been demonstrated physiological parameter reflecting autonomic neural activi- that the energy can be absorbed by protein molecules of ties, has been widely used in medical research [39]. While organisms to produce energy of quantity about that from low frequency (LF) power represents sympathetic tone, high hydrolysis of adenosine triphosphate [3]. The energy pro- frequency (HF) power reflects parasympathetic activity [40]. ducedistransmittedamonglargemoleculesinanorganism Since sympathetic and parasympathetic activities are antag- without raising the temperature or altering molecular struc- onistic, LHR signifies the net sympathetic activity relative tures, accounting for the nonheat-related biological effect to that of the parasympathetic system. Chronic elevation which is the predominant contributor to FIR-related physio- in sympathetic tone has been reported to be a detrimental logical action. The results of the present study demonstrated factor that contributes to the development of cardiovascular that FIR could increase peripheral perfusion. On the other diseases such as hypertension [41], coronary artery disease hand, significant elevation in parasympathetic activity was [42], and stroke [43]. The present study demonstrated sig- noted only when combining FIR and acupuncture. One of the nificant reduction in LHR after combined acupuncture-FIR interesting findings in the present study is the persistence of treatment, highlighting the beneficial role of combining the 6 Evidence-Based Complementary and Alternative Medicine two alternative therapies in neuromodulation. Moreover, the microcirculation in rats,” Photodermatology, Photoimmunology elevation in LHR after pseudoacupuncture further illustrates & Photomedicine,vol.22,no.2,pp.78–86,2006. the positive therapeutic influence of acupuncture on the [7] M. Imamura, S. Biro, T. Kihara et al., “Repeated thermal therapy enhancement of parasympathetic activity. improves impaired vascular endothelial function in patients There are several limitations in the present study. First, with coronary risk factors,” JournaloftheAmericanCollegeof the relatively small number of testing subjects precluded Cardiology, vol. 38, no. 4, pp. 1083–1088, 2001. the drawing of a strong conclusion from the findings. On [8] Y.-H. Hsu, Y.-C. Chen, T.-H. Chen et al., “Far-infrared therapy the other hand, the same subject was enrolled to undergo induces the nuclear translocation of PLZF which inhibits different kinds of therapy with an interval of at least one VEGF-induced proliferation in human umbilical vein endothe- week between two strategies not only to eliminate individual lial cells,” PLoS ONE,vol.7,no.1,ArticleIDe30674,2012. variation but also to minimize the residual effect from previ- [9]Y.UdagawaandH.Nagasawa,“Effectsoffar-infraredray ous treatment. Second, since the real-world clinical practice on reproduction, growth, behaviour and some physiological involves several sessions of acupuncture and FIR to achieve parameters in mice,” In Vivo,vol.14,no.2,pp.321–326,2000. the anticipated outcomes, the utilization of acupuncture and [10] H. Toyokawa, Y. Matsui, J. Uhara et al., “Promotive effects of FIR only once in the present study may have underesti- far-infrared ray on full-thickness skin wound healing in rats,” Experimental Biology and Medicine,vol.228,no.6,pp.724–729, mated the therapeutic effects of the two approaches both as 2003. monotherapy or as combined treatment. Third, only healthy [11] H. Nagasawa, Y. Udagawa, and S. Kiyokawa, “Evidence that young adults were enrolled in this study so that the impact of irradiation of far-infrared rays inhibits mammary tumour ageanddiseaseontreatmentoutcomeswasnotevaluated. growth in SHN mice,” Anticancer Research,vol.19,no.3A,pp. In conclusion, the present study demonstrated the clinical 1797–1800, 1999. benefits of utilizing acupuncture and FIR as both monothera- [12] Y. Udagawa, H. Ishigame, and H. Nagasawa, “Effects of hydrox- peutic strategies and combined treatment in terms of enhanc- yapatite in combination with far-infrared rays on spontaneous ing peripheral perfusion and parasympathetic activity, under- mammary tumorigenesis in SHN mice,” American Journal of scoring the positive roles of the two alternative therapies in Chinese Medicine,vol.30,no.4,pp.495–505,2002. neurovascular modulation. [13] Y. Hamada, F. Teraoka, T. Matsumoto et al., “Effects of far infrared ray on Hela cells and WI-38 cells,” International Conflicts of Interest Congress Series,vol.1255,pp.339–341,2003. [14] J. Ishibashi, K. Yamashita, T. Ishikawa et al., “The effects inhibit- The authors declare that they have no conflicts of interest. ing the proliferation of cancer cells by far-infrared radiation (FIR) are controlled by the basal expression level of heat shock protein (HSP) 70A,” Medical Oncology,vol.25,no.2,pp.229– Acknowledgments 237, 2008. [15]C.Tei,Y.Horikiri,J.-C.Parketal.,“Effectsofhotwater Financial support of the study by the Ministry of Science and bath or sauna on patients with congestive heart failure: Acute Technology, Taiwan (Grants nos. MOST 105-2221-E-259-007 hemodynamic improvements by thermal vasodilation,” Journal and MOST 104-2221-E-259-014), is gratefully acknowledged. of Cardiology,vol.24,no.3,pp.175–183,1994. [16] K. Honda and S. Inoue,´ “Sleep-enhancing effects of far-infrared References radiation in rats,” International Journal of Biometeorology,vol. 32, no. 2, pp. 92–94, 1988. [1] V. J. Robertson, A. Ward, J. Low, and A. Reed, Electrother- [17] A. White and E. Ernst, “A brief history of acupuncture,” apy Explained: Principles and Practice, Elsevier Butterworth- Rheumatology, vol. 43, no. 5, pp. 662-663, 2004. Heinemann, 2006. [18] S.-H. Hsiao and L.-J. Tsai, “A neurovascular transmission [2] S. Inoue and M. Kabaya, “Biological activities caused by far- model for acupuncture-induced nitric oxide,” JAMS Journal of infrared radiation,” International Journal of Biometeorology,vol. Acupuncture and Meridian Studies,vol.1,no.1,pp.42–50,2008. 33,no.3,pp.145–150,1989. [19] S.-Y. Wang, D. Zhang, L.-M. Tang, S.-Y. Li, M. Wen, and X.- [3] X.-F. Pang, “Vibrational energy-spectra of protein molecules J. Song, “Effects of electroacupuncture stimulation at ‘Zusanli’ and non-thermally biological effect of infrared light,” Interna- acupoint on hepatic no release and blood perfusion in mice,” tional Journal of Infrared and Millimeter Waves,vol.22,no.2, Evidence-Based Complementary and Alternative Medicine,vol. pp. 291–307, 2001. 2015, Article ID 826805, 6 pages, 2015. [4]Y.Akasaki,M.Miyata,H.Etoetal.,“Repeatedthermaltherapy [20]S.Ma,P.C.Lee,T.L.Anderson,X.Li,andI.Z.Jiang, up-regulates endothelial nitric oxide synthase and augments “Response of local nitric oxide release to manual acupuncture angiogenesis in a mouse model of hindlimb ischemia,” Circu- and electrical heat in humans: effects of reinforcement meth- lation Journal,vol.70,no.4,pp.463–470,2006. ods,” Evidence-Based Complementary and Alternative Medicine, [5] C.-C.Lin,C.-F.Chang,M.-Y.Lai,T.-W.Chen,P.-C.Lee,andW.- vol. 2017, pp. 1–8, 2017. C. Yang, “Far-infrared therapy: a novel treatment to improve [21] M.Sandberg,T.Lundeberg,L.Lindberg,andB.Gerdle,“Effects access blood flow and unassisted patency of arteriovenous of acupuncture on skin and muscle blood flow in healthy fistula in hemodialysis patients,” Journal of the American Society subjects,” European Journal of Applied Physiology,vol.90,no. of Nephrology,vol.18,no.3,pp.985–992,2007. 1-2,pp.114–119,2003. [6]S.Y.Yu,J.H.Chiu,S.D.Yang,Y.C.Hsu,W.Y.Lui,andC.W. [22] M.-F. Wu, S.-Q. Zhang, J.-B. Liu, Y. Li, Q.-S. Zhu, and R. Gu, Wu, “Biological effect of far-infrared therapy on increasing skin “Neuroprotective effects of electroacupuncture on early- and Evidence-Based Complementary and Alternative Medicine 7

late-stage spinal cord injury,” Neural Regeneration Research,vol. blood flow and autonomic activities in normal and overweight 10, no. 10, pp. 1628–1634, 2015. subjects,” Journal of Medical and Biological Engineering,vol.36, [23] H.-D. Guo, J.-X. Tian, J. Zhu et al., “Electroacupuncture no.3,pp.386–395,2016. suppressed neuronal apoptosis and improved cognitive impair- [38] T.-J. Ho, J.-R. Duann, C.-M. Chen et al., “Carryover effects alter ment in the ad model rats possibly via downregulation of fMRI statistical analysis in an acupuncture study,” American notch signaling pathway,” Evidence-based Complementary and Journal of Chinese Medicine,vol.36,no.1,pp.55–70,2008. Alternative Medicine,vol.2015,ArticleID393569,2015. [39]J.Song,M.Ogawa,A.Y.Tan,P.-S.Chen,andS.-F.Lin,“Heart [24] F. Liu, Y.-J. Jiang, H.-J. Zhao, L.-Q. Yao, and L.-D. Chen, “Elec- rate variability and autonomic nerve activities in ambulatory troacupuncture ameliorates cognitive impairment and regulates dogs,” in Proceedings of the 28th Annual International Confer- the expression of apoptosis-related genes Bcl-2 and Bax in ence of the IEEE Engineering in Medicine and Biology Society, rats with cerebral ischaemia-reperfusion injury,” Acupuncture in EMBS’06, pp. 1780–1783, usa, September 2006. Medicine,vol.33,no.6,pp.478–484,2015. [40] H. T. Wu, P. C. Hsu, C. K. Sun et al., “Assessment of autonomic [25] Y.Cao,L.-W.Zhang,J.Wangetal.,“Mechanismsofacupuncture dysfunction in patients with type 2 diabetes using reactive effect on Alzheimer’s disease in animal-based researches,” Cur- hyperemia,” Journal of Theoretical Biology,vol.330,pp.9–17, rent Topics in Medicinal Chemistry,vol.16,no.5,pp.574–578, 2013. 2016. [41] A. Ylitalo, K. E. J. Airaksinen, L. Sellin, and H. V. Huikuri, [26] E. Lv, J. Deng, Y. Yu et al., “Nrf2-ARE signals mediated the “Effects of combination antihypertensive therapy on baroreflex anti-oxidative action of electroacupuncture in an MPTP mouse sensitivity and heart rate variability in systemic hypertension,” model of Parkinson’s disease,” Free Radical Research,vol.49,no. American Journal of Cardiology,vol.83,no.6,pp.885–889,1999. 11, pp. 1296–1307, 2015. [42] M. Esler and D. Kaye, “Increased sympathetic nervous system [27]T.H.Joh,H.Park,S.Kim,andH.Lee,“Recentdevelopmentof activity and its therapeutic reduction in arterial hypertension, acupuncture on Parkinson’s disease,” Neurological Research,vol. portal hypertension and heart failure,” JournaloftheAutonomic 32, no. sup1, pp. 5–9, 2013. Nervous System, vol. 72, no. 2-3, pp. 210–219, 1998. [28] B.-Y. Zeng, S. Salvage, and P. Jenner, “Effect and mechanism [43] A. M. Dorrance and G. Fink, “Effects of stroke on the autonomic of acupuncture on alzheimer’s disease,” International Review of nervous system,” Comprehensive Physiology,vol.5,no.3,pp. Neurobiology, vol. 111, pp. 181–195, 2013. 1241–1263, 2015. [29] Q. Xu, J.-W. Yang, Y. Cao et al., “Acupuncture improves locomotor function by enhancing GABA receptor expression in transient focal cerebral ischemia rats,” Neuroscience Letters,vol. 588, pp. 88–94, 2015. [30] M. Xu, L. Ge, and D. Zhao, “Electro-acupuncture regulation of central monoamine neurotransmitters in ischaemia- reperfusion,” Current Research in Acupuncture,vol. 9781461433576, pp. 401–430, 2013. [31] S. Chen and S.-X. Ma, “Nitric oxide in the gracile nucleus mediates depressor response to acupuncture (ST36),” Journal of Neurophysiology,vol.90,no.2,pp.780–785,2003. [32]K.Imai,H.Ariga,C.Chen,C.Mantyh,T.N.Pappas,andT. Takahashi, “Effects of electroacupuncture on gastric motility and heart rate variability in conscious rats,” Autonomic Neuro- science: Basic and Clinical,vol.138,no.1-2,pp.91–98,2008. [33] K. Imai, H. Ariga, and T. Takahashi, “Electroacupuncture improves imbalance of autonomic function under restraint stress in conscious rats,” American Journal of Chinese Medicine, vol. 37, no. 1, pp. 45–55, 2009. [34]H.Sallam,T.A.McNearney,D.Doshi,andJ.D.Z.Chen, “Transcutaneous electrical nerve stimulation (TENS) improves upper GI symptoms and balances the sympathovagal activity in scleroderma patients,” Digestive Diseases and Sciences,vol.52, no. 5, pp. 1329–1337, 2007. [35]S.Liu,S.Peng,J.D.Chen,X.Hou,andM.Ke,“Transcuta- neous electroacupuncture improves dyspeptic symptoms and increases high frequency heart rate variability in patients with functional dyspepsia,” Neurogastroenterology & Motility,vol.20, no. 11, pp. 1204–1211, 2008. [36] A. Bar,¨ Y. Li, R. Eichlisberger, F. Angst, and A. Aeschlimann, “Acupuncture improves peripheral perfusion in patients with reflex sympathetic dystrophy,” Journal of Clinical Rheumatology, vol. 8, no. 1, pp. 6–12, 2002. [37] C. Yang, H. Wu, W. Chuang et al., “Application of short- time mse in assessing impact of acupuncture on peripheral Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 7483639, 10 pages https://doi.org/10.1155/2017/7483639

Research Article Effective Heart Disease Detection Based on Quantitative Computerized Traditional Chinese Medicine Using Representation Based Classifiers

Ting Shu, Bob Zhang, and Yuan Yan Tang

Department of Computer and Information Science, University of Macau, Taipa, Macau

Correspondence should be addressed to Bob Zhang; [email protected]

Received 11 May 2017; Revised 14 June 2017; Accepted 28 June 2017; Published 13 August 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Ting Shu et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

At present, heart disease is the number one cause of death worldwide. Traditionally, heart disease is commonly detected using blood tests, electrocardiogram, cardiac computerized tomography scan, cardiac magnetic resonance imaging, and so on. However, these traditional diagnostic methods are time consuming and/or invasive. In this paper, we propose an effective noninvasive computerized method based on facial images to quantitatively detect heart disease. Specifically, facial key block color features are extracted from facial images and analyzed using the Probabilistic Collaborative Representation Based Classifier. The idea of facial key block color analysis is founded in Traditional Chinese Medicine. A new dataset consisting of 581 heart disease and 581 healthy samples was experimented by the proposed method. In order to optimize the Probabilistic Collaborative Representation Based Classifier, an analysis of its parameters was performed. According to the experimental results, the proposed method obtains the highest accuracy compared with other classifiers and is proven to be effective at heart disease detection.

1. Introduction contents in the blood, respectively, while overall the results can help create a clear picture of a person’s heart health [12]. Heart disease (HD) is actually a broad term used for a An ECG records electrical signals, while a Holter monitor is a wide variety of diseases of the heart and blood vessels such portabledevicethepatientwearstorecordacontinuousECG, as coronary artery disease (CAD) [1] and heart rhythm usually for 24 to 72 hours. An echocardiogram uses sound disorders called arrhythmias (ARR) [2]. According to the waves to produce images of a person’s heart, while a stress test World Health Organization (WHO), HD is the number one records a person’s signs and symptoms during exercise using cause of death globally [3]. In 2012, it was estimated that an ECG or echocardiogram. For cardiac catheterization, a HD caused about 17.5 million deaths, which means a person special dye needs to be injected into a person’s coronary died from HD every 2 seconds [4]. There are many tests to arteries through a long, thin, and flexible tube (catheter) diagnose HD; the main traditional diagnostic methods of usually in the leg. The dye then outlines narrow spots and HD are [5] blood tests, Electrocardiogram (ECG) [6], Holter blockages that appear in X-ray images. A CT scan and MRI monitoring [7], echocardiogram [8], cardiac catheterization can also help doctors detect calcium deposits in the patient’s [9], cardiac computerized tomography (CT) scan [10], and arteries that can narrow it. cardiac magnetic resonance imaging (MRI) [11]. Blood tests performed on individuals with HD are con- Manycluesaboutthehealthofaperson’sheartcanbe sidered invasive as bodily fluids are removed and can take discovered in his/her blood. However, a single blood test time for the laboratory technician to reach a result. ECG on cannot reflect the risk of heart disease. Two common blood the other hand might not be as invasive as a blood test, but tests for heart disease are a cholesterol test and a C-reactive in the case of Holter monitoring, it is time consuming. As protein (CRP) test. These tests analyze cholesterol and CRP for cardiac catheterization, the injection of a special dye is 2 Evidence-Based Complementary and Alternative Medicine

171 214

404 295

Male Male Female Female No record No record (a) H (b) HD

Figure 1: Gender distribution of the dataset. the definition of invasive. Therefore, given these issues, there in this paper has two distinctive classes: (1)HDwith581 is a need to develop a noninvasive computerized method to samples and (2) healthy (H) consisting of 581 samples. Based detect HD. on the seven facial key block permutations, ProCRC with its In 2008, Kim et al. proposed one such method to conduct optimal parameters is applied to classify HD versus H. To the color compensation of a facial image based on the analysis the best of our knowledge, this is the first time noninvasive of facial color [13] rooted in Traditional Chinese Medicine computerized heart disease detection has been proposed in (TCM). In [13], they extracted the center forehead and lips the literature. of a person and analyzed the red color value distribution of The organization of this paper is given as follows. The the center forehead and lips. The authors wanted to survey details about the dataset are represented in Section 2. Feature real clinical data of HD patients and group them into different extraction of the facial key blocks is given in Section 3, cases based on the analysis that facial color can help doctors succeeded by a description of our proposed method in diagnose HD. However, the authors just proposed a method Section 4 using ProCRC. Section 5 describes and discusses and did not experiment on a real dataset. the experimental results and Section 6 concludes this paper. Recently, Zhang et al. [14] used facial block color features to detect diabetes in a noninvasive manner with the Sparse 2. Dataset Representation Based Classifier (SRC). Even though their detection results are relatively high, further analyses using The dataset we collected and used in this work consists of 581 other representation algorithms have not been studied nor H and 581 HD samples from the Guangdong Provincial TCM have these algorithms been applied to detect other nondia- Hospital, Guangdong, China, in 2015. Individuals were diag- betic diseases. To resolve these issues, we propose an effective nosed as healthy by medical professional practicing Western noninvasive computerized method to detect HD through medicine, while heart disease patients were determined using facial image analysis via the Probabilistic Collaborative Rep- the methods described in Section 1. Please note the handling resentation Based Classifier (ProCRC) and apply our pro- of human subjects was done according to the principles posed method on a real dataset. ProCRC was first proposed in outlined in the Declaration of Helsinki and each individual [15] and applied in pattern recognition, being developed from gavetheirconsenttobeapartofthisstudy.Ethicalapproval the Collaborative Representation Based Classifier (CRC) of was obtained from the Science and Technology Development [16]. Zhang et al. [16] proved that Collaborative Representa- Fund (FDCT) of Macao for this study with the project tion played a more important role than sparsity in pattern number FDCT 124/2014/A3. recognition and proposed CRC, which outperformed the The gender and age distributions of H and HD are SRC [17] and also runs much faster. In our work, the ProCRC described in this section. During data collection, it is was modified to be applied for HD detection based on facial sometimes difficult to record the information of everyone key block color features. The ProCRC combines CRC and the due to many circumstances. Therefore, in gender and age probabilistic theory. distributions, there are cases of no record (NR). The following For the proposed method, facial images are first captured pie charts (Figure 1) are used to show the dataset gender through a specially designed facial image capture device distribution. In the pie chart, blue represents males, yellow and four facial key blocks are extracted from each image. is for females, and NR is illustrated in gray. In Figure 1, A color gamut with six-facial-color centroids is employed there are two pie charts describing the gender distributions to extract color features from each block. The dataset used of the dataset: (1)H(Figure1(a))and(2) HD (Figure 1(b)). Evidence-Based Complementary and Alternative Medicine 3

Table 1: Age distribution of the dataset.

Class 1–17 18–24 25–60 61–80 ≥81 NR Sum 13271794070581 H 0.17% 56.28% 30.81% 0.69% 0% 12.05% 100% 6 38 398 115 18 6 581 HD 1.03% 6.54% 68.5% 19.79% 3.1% 1.03% 100%

According to Figure 1(a), 72 people are missing their gender information in H and about half of the healthy dataset is female (295), while the number of males is 214. Different from the H dataset, the HD dataset has only 6 NR cases. About 1/3 of the HD patients are female (171) with 404 male HD patients 2/3 (see Figure 1(b)). F The age distribution is given through a table (see Table 1). Toshowtheagedistribution(inyears)clearly,theageis split into 5 parts: [1–17], [18–24], [25–60], [61–80], and [≥81]. From this table, the first column is the class name, where each class has two rows: the first row is the number of the N people belonging to the age range and the second row is the RL corresponding percentage of people out of the total. For the Hdataset,theageofmostpeopleisfrom18to60(56.28% + 30.81% = 87.09%) with no healthy person above 80 and it contains only 4 people above 60. As for the HD dataset consisting of 581 samples, the majority of HD patients are C aged from 25 to 80 (68.5% + 19.79% = 88.29%). It should be noted that the missing gender and age Figure 2: Different facial regions according to TCM. information does not affect our study since we are only interested in each individual’s health status. using four facial key blocks instead of the whole facial image is more appropriate and efficient. Figure 3 depicts an example of 3. Facial Key Block Feature Extraction a facial image with its four marked facial key blocks. The four facial key blocks are forehead block (FHB) on the forehead, In order to decrease the effects of the capture environment, left and right cheek blocks (LCB and RCB) below the left a specially designed facial image capture device was applied. andrighteyeswhicharesymmetrical,andnosebridgeblock Using the device, the individual just needs to place his/her (NBB) on the nose, the midpoint of LCB and RCB. The four head on the chin rest and the device operator clicks the facial key block sizes are the same at 64 × 64 pixels. capture button. More details about the device can be found In the automatic key blocks extraction procedure, the in [14]. A color correction procedure [18] was also performed pupils are first detected and marked. The positions of the to portray the facial images in an accurate way after image two pupils are denoted as 𝐿𝑙𝑝 =(𝑥𝑙𝑝 ,𝑦𝑙𝑝 ) (left) and 𝐿𝑟𝑝 = capture. (𝑥𝑟𝑝,𝑦𝑟𝑝) (right). Based on 𝐿𝑙𝑝 and 𝐿𝑟𝑝,thefourfacialkey In Traditional Chinese Medicine (TCM), it is believed blocks are located through that the status of the internal organs can be determined from different regions of the face [19–21]. Figure 2 shows 1 𝐿 =(𝑥 ,𝑦 − 𝐻) , a human face partitioned into various regions according to LCB 𝑙𝑝 𝑙𝑝 4 TCM [22]. Facial blocks were previously defined in [23] 1 to detect hepatitis from digital facial images. The authors 𝐿RCB =(𝑥𝑟𝑝,𝑦𝑟𝑝 − 𝐻) , extracted 5 facial blocks, one between the eyebrows, two 4 below the eyes, one under the bridge of the nose, and one 𝑥 +𝑥 𝑦 +𝑦 1 (1) 𝐿 =( 𝑙𝑝 𝑟𝑝 , 𝑙𝑝 𝑟𝑝 + 𝐻) , underneath the lower lip. Applying this idea to our proposed FHB 2 2 3 method, four facial key blocks are automatically extracted from each facial image representing the main regions. No 𝑥𝑙𝑝 +𝑥𝑟𝑝 𝑦𝑙𝑝 +𝑦𝑟𝑝 2 𝐿 =( , − 𝐻) , facial block is used to represent region C in Figure 2 due to NBB 2 2 9 the existence of facial hair. Hence, according to the five facial regions, four facial key where 𝐿𝑖th key block name means the position of 𝑖th key block, blocks are automatically extracted from each calibrated facial such as 𝐿FHB is the position of FHB, and 𝑊 and 𝐻 are the image. Furthermore, the dimensionality of the whole facial width and height of the facial image, respectively. Figure 3 image is much larger than four facial key blocks. Therefore, depicts the locations of the four facial key blocks based on 4 Evidence-Based Complementary and Alternative Medicine

0.8

FHB 0.7 Image 1 height H 0.6 (H) 3 0.5 2 1 H 1 H 9 H 0.4

4 4 CIE xyY - y NBB LCB RCB 0.3

0.2

0.1 Image width (W) −0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Figure 3: Four facial key block positions. CIExyY-x

R (red) GL (gloss) [253 203 174] [241 224 198] Class FHB LCB NBB RCB Y (yellow) DR (deep red) [226 186 137] [239 154 108] LY (light yellow) BK (black) [236 211 173] [223 188 191]

Figure 5: Facial color gamut with its six-color centroids marked by red crosses.

color gamut as a solid colored square, whose label is on top and correspondingly RGB value is below. Each pixel in a facial block is compared to one of the six-color centroids and assigned to its nearest centroid. After evaluating all pixels of a facial block, the total of each color (based on the six-color centroids) is summed and divided by the total number of pixels. This ratio forms the facial color feature vector 𝑘,where𝑘=[𝑟1,𝑟2,𝑟3,𝑟4,𝑟5,𝑟6] and 𝑟𝑖 represents the sequence of the six-color centroids in Figure 5. By comparing the four facial color feature vectors (per facial image) in groups of two (using all images in the dataset), andcalculatingthemeanabsolutedifferenceofeachgroup, HD LCBandRCBareshowntohavethesmallestdifference[14]. ThisisnotsurprisinggivenLCBandRCBaresymmetrical and located on either side of the face. Therefore, in the following experiments, RCB is removed.

4. Representation Based Classifiers 4.1. Sparse Representation Based Classifier (SRC). The SRC Figure 4: Three typical examples of four facial key blocks from the was first proposed by Wright et al. [17] and used for face two classes. recognition. Since then, this classifier has been applied in numerous fields such as pattern recognition [14, 24], object detection [25], image restoration [26], image denoising [27], the left and right pupil positions. Three typical examples from video restoration [28], image super-resolution [29]. For the each class are illustrated in Figure 4. following, 𝐷 represents a dataset; 𝑠 donates a sample; 𝑋, 𝑌,or The color features are extracted from each facial key 𝑍 stands for a coefficient; and 𝛼 or 𝛽 is a positive scalar. block. A color gamut (see Figure 5) with six-facial-color The principle of the SRC is using the linear combination centroids are applied for color feature extraction, where of the training data (𝐷)torepresentthequerytestingsample 6 color values are extracted from each facial key block. (𝑠) while keeping the coefficients (𝑌) sparse enough. The Figure 5 illustrates the six-color centroids from the facial coefficients of the class that the testing samples belong to have Evidence-Based Complementary and Alternative Medicine 5

󸀠 󸀠 (1) Input: 𝐷, 𝐷𝑙 , 𝐷𝑙 , 𝐿, 𝛼, 𝛽,and𝑠 (2) Output: id(y) (3) Code 𝑠 with 𝐷 via 𝑙2-norm: ̂ 2 2 𝐿 2 (4) (𝑋) = arg min𝑋{‖𝑠 − 𝐷𝑋‖2 +𝛼‖𝑋‖2 +(𝛽/𝐿)∑𝑙 ‖𝐷𝑋𝑙 −𝐷 𝑋𝑙‖2} (5) Calculate the solution of the coefficient: ̂ 𝑇 𝐿 󸀠 𝑇 󸀠 −1 𝑇 (6) 𝑋=(𝐷 𝐷+(𝛽/𝐿)∑𝑙=1(𝐷𝑙 ) 𝐷𝑙 +𝛼𝐼) 𝐷 𝑠 (7) Compute the residual for each class: (8) for 𝑙=1; 𝑙≤𝐿; 𝑙++do ̂ ̂ 2 (9) 𝑟𝑙(𝑠) = ‖𝐷𝑋−𝐷𝑙𝑋𝑙‖2 (10) end for (11) With the residuals, determine the class label of 𝑠: (12) 𝑖𝑑(𝑠) = arg min𝑙{𝑟𝑙(𝑠)}

Algorithm 1: ProCRC algorithm procedure.

𝑀×1 significant values, while the other coefficients are nearly zero. coefficient 𝑋 of 𝐷 representing a test sample 𝑠∈R via The SRC is defined as ProCRC is solved with the following: ̂ 2 (𝑌) = arg min {‖𝑠−𝐷𝑌‖ +𝛼SRC ‖𝑌‖1} , ̂ 2 2 𝑌 2 (2) (𝑋) = arg min {‖𝑠−𝐷𝑋‖ +𝛼‖𝑋‖ 𝑋 2 2 (5) where 𝛼SRC can be set to obtain the real sparse coding vector 𝐿 𝛽 󵄩 󵄩2 𝑌̂ of 𝑠 over 𝐷. + ∑ 󵄩𝐷𝑋 −𝐷 𝑋 󵄩 }, 𝐿 󵄩 𝑙 𝑙󵄩2 𝑙 4.2. Collaborative Representation Based Classifier (CRC). where 𝛼 and 𝛽 are regularization parameters. In [16], Zhang et al. established the Collaborative Repre- Using ProCRC, the class label of the test sample is sentation (CR) mechanism, but not the 𝑙1-norm sparsity determined through locating the minimum value of the constraint, that truly improved the method’s effectiveness residual error for each class: and further proposed a Collaborative Representation Based 󵄩 󵄩2 󵄩 ̂ ̂ 󵄩 𝑖𝑑 (𝑠) = arg min 󵄩𝐷𝑋−𝐷𝑙𝑋𝑙󵄩 , (6) Classifier (CRC). 𝑙 2 The authors of [16] proposed CRC by modifying the 𝑙1- ̂ norm of the SRC (2) to a 𝑙2-norm: where 𝑋𝑙 represents the coefficients of the test sample 𝑠 in the 𝑙th class. Algorithm 1 shows the procedure of ProCRC. ̂ 2 2 𝐷󸀠 = (𝑍) = arg min {‖𝑠−𝐷𝑍‖2 +𝛼CRC ‖𝑍‖2} , In order to show the ProCRC procedure clearly, let 𝑙 𝑍 (3) 𝑀×𝑁 󸀠 󸀠 [0,...,𝐷𝑙,...,0]∈ R and 𝐷𝑙 =𝐷−𝐷𝑙 have the same size of 𝐷. More details about ProCRC can be found in [15]. where 𝛼CRC is the regularization parameter. The solution of (3) can be easily and analytically derived as 5. Experimental Results ̂ 𝑇 −1 𝑇 𝑍 = (𝐷 𝐷+𝛼CRC ⋅𝐼) 𝐷 𝑠. (4) The experimental results are represented in this section. The settings for HD detection are first given followed by 𝑇 −1 𝑇 The first part ((𝐷 𝐷+𝛼CRC ⋅𝐼) 𝐷 ) of (4) is independent of the detection results using 10 classifiers to compare and 𝑠. Therefore, it can be precalculated and once a query sample contrast with the ProCRC. Finally, the analysis of the ProCRC ̂ 𝛼 𝛽 𝑠 is available, it is projected to get 𝑍.Thismakescalculating parameters and is represented in Section 5.3. ̂ 𝑍 faster than 𝑌̂ in (2). More details about CRC can be found in [16]. 5.1. Experimental Setting. We randomly selected close to half (580) of the data for training and the remaining data (582) for testing, where accuracy (which is the proportion 4.3. Probabilistic Collaborative Representation Based Classifier of the correctly classified samples divided by all samples) (ProCRC). Cai et al. [15] proposed the Probabilistic Collab- is the performance measurement used. To overcome the orative Representation Based Classifier (ProCRC) algorithm 𝑀×𝑁 shortcoming of different results for different data partitions for pattern classification. Let 𝐷=[𝐷1,𝐷2,...,𝐷𝐿]∈R [30], 5 random partitions were applied, where the final 𝑀×𝑁푙 denote the training samples, where 𝐷𝑙 ∈ R represents accuracy is its mean. The following experimental results were the training samples from the 𝑙th class with 𝑁𝑙 samples conducted on a PC with 8 i7-6700 CPU @3.40 GHz processor, 𝐿 (𝑁=∑𝑙=1 𝑁𝑙), and the dimension of each sample is 𝑀.The 16.0 GB RAM, and a 64-bit OS. 6 Evidence-Based Complementary and Alternative Medicine

Eleven classifiers accuracy with various block combinations Table 2: Comprehensive HD detection results using 11 classifiers. 88 86 Block combination ACC SEN SPC 84 82 𝑘-NN 80 FHB 80.45% 67.56% 93.33% 78

Accuracy (%) Accuracy 76 LCB 81.27% 68.11% 94.43% 74 72 NBB 76.80% 63.99% 89.62% FHB + LCB 81.89% 69.69% 94.09% LCB FHB NBB FHB + NBB 81.72% 71.27% 92.16%

NBB LCB + NBB 80.89% 69.69% 92.10% FHB + LCB LCB + NBB LCB FHB + NBB

FHB + LCB + FHB + LCB FHB + LCB + NBB 83.13% 72.30% 93.95% Block combination SVM  -NN Softmax FHB 83.71% 75.74% 91.68% SVM Decision Tree SRC AdaBoost LCB 84.36% 76.01% 92.71% DL with SRC LogitBoost NBB 79.11% 69.83% 88.38% ProCRC Gentle Boost FHB + LCB 87.32% 82.47% 92.16% CRC FHB + NBB 85.26% 80.07% 90.45% Figure 6: HD detection accuracies of all 11 classifiers including LCB + NBB 85.33% 78.28% 92.37% ProCRC. FHB + LCB + NBB 87.66% 83.78% 91.55% SRC FHB 79.76% 77.18% 82.34% The dataset we collected and used in this work consists of LCB 81.24% 77.73% 84.74% 581 H and 581 HD samples from the Guangdong Provincial TCM Hospital, Guangdong, China, in 2015. NBB 72.44% 69.62% 75.26% Based on Section 3, three facial key blocks (FHB, LCB, FHB + LCB 84.12% 81.44% 86.80% andNBB)areusedinsteadofthewholefacialimage. FHB + NBB 82.37% 78.35% 86.39% Therefore, there are seven combinations for the three facial LCB + NBB 82.92% 79.31% 86.53% key blocks and all seven combinations were applied separately FHB + LCB + NBB 85.09% 79.52% 90.65% for each classifier. The seven block combinations consist of3 DL with SRC cases with one block (FHB, LCB, and NBB), 3 cases with 2 blocks (FHB + LCB, FHB + NBB, and LCB + NBB), and all FHB 75.74% 64.88% 86.60% blocks combined together (FHB + LCB + NBB). LCB 76.22% 62.75% 89.69% NBB 78.63% 68.45% 88.80% FHB + LCB 76.77% 66.39% 87.15% 5.2. HD Detection Results. Other than the ProCRC, 10 other classifiers were applied to detect HD representing an array of FHB + NBB 82.54% 73.13% 91.96% traditional and the state of the art. The 10 classifiers are (i) 𝑘- LCB + NBB 82.85% 73.47% 92.23% Nearest Neighbor (𝑘-NN) [31] with 𝑘=1, (ii) Support Vector FHB + LCB + NBB 83.54% 76.49% 90.58% Machines (SVM) [31] with linear kernel function, (iii) SRC ProCRC 𝜆 = 0.1 [17] with , (iv) Dictionary Learning (DL) with SRC FHB 83.57% 73.61% 93.54% [32] using 𝜆SRC = 0.1, 𝜆DL = 0.1, and a dictionary size equal LCB 84.30% 73.75% 94.85% to half of the feature dimensionality, such as 3 for one key block, (v) CRC [16] with 𝜆 = 0.01, (vi) Softmax [33], (vii) NBB 78.08% 63.71% 92.44% Decision Tree [34], (viii) AdaBoost [35] with Tree Leaner, FHB + LCB 87.11% 82.06% 92.16% (ix) LogitBoost [36] with Tree Leaner, and (x) Gentle Boost FHB + NBB 85.74% 79.73% 91.75% [37]. The classifier parameters were fine-tuned based onits LCB + NBB 85.43% 78.42% 92.44% best performance and for the ProCRC its two parameters are FHB + LCB + NBB 88.01% 84.95% 91.07% analyzed in Section 5.3. CRC Figure 6 illustrates the best accuracies of all 11 classifiers basedonfacialkeyblockcolorfeaturesforallsevenblock FHB 78.76% 59.04% 98.49% combinations. From this bar chart, it is obvious that the LCB 78.97% 59.86% 98.08% ProCRC results (in red) outperformed or came close to NBB 76.19% 59.86% 92.51% achieving the highest accuracy for almost each combination. FHB + LCB 82.51% 67.42% 97.59% To be thorough, the complete set of results including FHB + NBB 82.65% 69.48% 95.81% accuracy, sensitivity, and specificity [38] of the 11 classifiers LCB + NBB 81.68% 68.45% 94.91% using seven block combinations is shown in Table 2. In the table, ACC, SEN, and SPC represent accuracy, sensitivity, FHB + LCB + NBB 84.43% 72.23% 96.63% Evidence-Based Complementary and Alternative Medicine 7

Table 2: Continued. Block combination ACC SEN SPC HD Softmax FHB 83.68% 90.09% 79.15% LCB 84.30% 91.48% 79.40% NBB 78.97% 85.71% 74.48% FHB + LCB 87.39% 91.14% 84.41% FHB + NBB 85.15% 89.25% 81.92% H LCB + NBB 84.88% 89.67% 81.22% FHB + LCB + NBB 87.39% 90.30% 85.00% Decision Tree Figure 7: Three examples of FHB from HD and H that cannot be FHB 78.21% 76.63% 79.79% recognized with the naked eye. LCB 79.97% 78.76% 81.17% NBB 73.61% 72.65% 74.57% FHB + LCB 83.30% 81.44% 85.15% To further demonstrate the effectiveness of the proposed FHB + NBB 80.86% 78.56% 83.16% method, Figure 7 shows three examples of FHB for HD LCB + NBB 81.65% 79.79% 83.51% and H, respectively. In this figure, the top row is FHB FHB + LCB + NBB 81.58% 80.34% 82.82% from HD and the bottom row is from H. Looking at the figure,itisdifficulttodistinguishtheblockswiththenaked AdaBoost eye.However,theproposedmethodcanclassifyeachblock FHB 83.40% 76.29% 90.52% correctly. LCB 83.81% 76.70% 90.93% NBB 79.97% 74.23% 85.70% 5.3. ProCRC Parameters Analysis. Based on Section 4.3, both FHB + LCB 86.53% 84.81% 88.25% of the two parameters range from [0.001, 0.01, 0.1 : 0.1 : 1.0]. FHB + NBB 84.40% 81.92% 86.87% In order to find the optimal values of 𝛼 and 𝛽 for HD detec- LCB + NBB 84.57% 80.27% 88.87% tion, experiments using each of the seven block combinations FHB + LCB + NBB 86.56% 84.60% 88.52% were analyzed. These results are shown in Figure 8. In each LogitBoost subfigure, the red line represents the accuracies of a fixed 𝛽 FHB 83.40% 76.29% 90.52% with 𝛼 changing its values, while in the blue line it is the 𝛼 𝛽 LCB 84.40% 79.59% 89.21% opposite with being equal to a constant and changing. 𝛼 𝛽 NBB 79.90% 74.23% 85.57% and results based on FHB are shown in Figure 8(a). After 𝛼 = 0.7 and 𝛽 = 0.4, the red and blue lines remained FHB + LCB 87.08% 84.47% 89.69% constant, respectively. The best accuracy of FHB was 83.71%, FHB + NBB 84.78% 83.92% 85.64% where 𝛼 = 0.3 and 𝛽=0.2. Figure 8(b) depicts the LCB + NBB 84.91% 81.10% 88.73% ProCRC parameter results for LCB. Except for 𝛽 = 0.001, FHB + LCB + NBB 87.70% 85.29% 90.10% the accuracies of the other 𝛽 values were the same. For 𝛼, Gentle Boost the accuracies also had only two values, which were the FHB 83.26% 76.43% 90.10% same with 𝛽, where 0.4 caused a change. The ProCRC with LCB 83.81% 79.52% 88.11% 𝛼 = 0.4 and 𝛽 = 0.001 based on LCB obtained the best NBB 79.97% 73.75% 86.19% accuracy of 84.33%. The results for NBB are represented 𝛽 FHB + LCB 87.04% 85.02% 89.07% in Figure 8(c). For , the top accuracy was achieved at the 𝛽 = 0.001 𝛼 FHB + NBB 84.67% 84.12% 85.22% initial point ( ). The result of from 0.01 to 0.3 did not change and the highest accuracy was 78.08%. LCB + NBB 85.02% 81.37% 88.66% Figure 8(d) illustrates 𝛼 and 𝛽 for FHB + LCB. The best FHB + LCB + NBB 87.08% 84.26% 89.90% accuracies of 𝛼 werethesame(87.18%)from0.3to0.7.The two parameters of ProCRC based on FHB + NBB are depicted in Figure 8(e). The 𝛽 results decreased with the increasing and specificity, respectively. As can be seen in Table 2, the of 𝛽. In contrast, the accuracies of 𝛼 increased with the ProCRC using FHB + LCB + NBB (highlighted) achieved increasing of 𝛼. The highest accuracy of FHB + NBB was the highest result (88.01%) amongst all classifiers. Using this 85.77%, where 𝛼=0.3and 𝛽 = 0.001.Figure8(f)shows grouping, the second highest result was 87.7% obtained by the result of 𝛼 and 𝛽 for LCB + NBB. With the increasing of LogitBoost. The biggest difference between the ProCRC and 𝛼 and 𝛽, its accuracies increased and decreased, respectively. the 10 other classifiers with FHB + LCB + NBB was 6.43%, The best result of 85.53% was obtained from LCB + NBB with where the classifier was Decision Tree. When compared to 𝛼 = 0.8 and 𝛽 = 0.001. The final subfigure (Figure 8(g)) the representation based algorithms (SRC, DL with SRC, and represents the two parameters for FHB + LCB + NBB. Similar CRC), the ProCRC achieved on average a 3.65% increase in to Figure 8(f), the results decreased with an increasing 𝛽. accuracy using FHB + LCB + NBB. From 0.001to0.1,the 𝛼 accuracies increased with 𝛼 increasing 8 Evidence-Based Complementary and Alternative Medicine

83.8 84.4 83.75

83.7 84.35 83.65

83.6 84.3 Accuracy (%) Accuracy 83.55 (%) Accuracy

83.5 84.25

83.45 84.2 0.001 0.01 0.1 0.2 0.3 0.4 0.001 0.01 0.1 0.2 0.3 0.4 Parameter values Parameter values

 changing with  = 0.2  changing with =0.001  changing with  = 0.3  changing with =0.4 (a) FHB (b) LCB

78.15 87.25

78.1 87.2

78.05 87.15

78 87.1 Accuracy (%) Accuracy Accuracy (%) Accuracy 77.95 87.05

77.9 87

0.001 0.01 0.1 0.2 0.3 0.4 0.001 0.01 0.1 0.2 0.3 0.4 Parameter values Parameter values

 changing with =0.001  changing with =0.01  changing with =0.01  changing with =0.3 (c) NBB (d) FHB + LCB

85.85 85.6 85.8 85.55 85.5 85.75 85.45 85.7 85.4 85.65 85.35 85.6 85.3 Accuracy (%) Accuracy Accuracy (%) Accuracy 85.55 85.25 85.2 85.5 85.15 85.45 85.1 0.001 0.01 0.1 0.2 0.3 0.4 0.001 0.01 0.1 0.2 0.3 0.4 Parameter values Parameter values

 changing with =0.001  changing with =0.001  changing with =0.3  changing with =0.8 (e) FHB + NBB (f) LCB + NBB

Figure 8: Continued. Evidence-Based Complementary and Alternative Medicine 9

88.1

87.9

87.8

87.7 Accuracy (%) Accuracy 87.6

87.5

87.4 0.001 0.01 0.1 0.2 0.3 0.4 Parameter values

 changing with =0.001  changing with =0.1 (g) FHB + LCB + NBB

Figure 8: ProCRC accuracy with 𝛼 and 𝛽 changing.

and had small fluctuations after 0.2. The best accuracy in this Acknowledgments case, which was also the highest accuracy in all 11 classifier, was 88.01%, where 𝛼 = 0.1 and 𝛽 = 0.001. This work was supported by the Research Grants of Univer- sity of Macau [MYRG2015-00049-FST, MYRG2015-00050- FST]; the Science and Technology Development Fund 6. Conclusions (FDCT) of Macau [128/2013/A, 124/2014/A3]; and Macau- China Join Project [008-2014-AMJ]. This research project was This paper proposed a noninvasive computerized method to also supported by the National Natural Science Foundation of detect HD based on facial key block color analysis classified China [61273244] and [61602540]. using the ProCRC. The experiments were conducted on a new dataset consisting of 581 HD samples and 581 H samples. The facial images are first captured through a specially designed References device, where four facial key blocks are extracted to represent [1] P. A. McCullough, “Coronary artery disease,” Clinical Journal one sample. For each facial key block, color features are of the American Society of Nephrology,vol.2,no.3,pp.611–616, extracted using a facial color gamut with six-color centroids. 2007. To obtain optimal HD detection, three facial key blocks [2] T. H. Foundation, “What is heart disease?” 2016, https:// arepermutedandappliedforclassification.Theproposed www.theheartfoundation.org/heart-disease-facts/about-heart- method used the ProCRC which was developed from CRC disease/. andanalyzedCRCbasedontheprobabilistictheory[15]. [3] W. H. Organization, “The top 10 causes of death,” 2016, http:// Compared with 10 other classifiers, the best accuracy of www.who.int/mediacentre/factsheets/fs310/en/. HD detection was 88.01% with a sensitivity of 84.95% and [4] W. H. Organization, “Cardiovascular diseases (cvds),” 2016, a specificity of 91.07% (using the ProCRC with 𝛼 = 0.1 http:// www.who.int/mediacentre/factsheets/fs317/en/. and 𝛽 = 0.001 with FHB + LCB + NBB). This proves the [5] M. Clinic, “Heart disease - tests and diagnosis,” 2016, http:// effectiveness of the ProCRC based on facial key block color www.mayoclinic.org/diseases-conditions/heart-disease/basics/ feature analysis to detect HD and potentially provides a new tests-diagnosis/con-20034056. innovative noninvasive way to detect this disease. [6]A.P.Davie,C.M.Francis,M.P.Loveetal.,“Valueof As part of the future work, more features from the facial the electrocardiogram in identifying heart failure due to left key blocks will be explored and extracted. In addition, other ventricular systolic dysfunction,” British Medical Journal,vol. representation learning algorithms will be developed and 312, no. 7025, pp. 222-223, 1996. applied to HD detection. [7] D. L. Kuchar, C. W.Thorburn, and N. L. Sammel, “Prediction of serious arrhythmic events after myocardial infarction: Signal- averaged electrocardiogram, Holter monitoring and radionu- Conflicts of Interest clide ventriculography,” Journal of the American College of Cardiology,vol.9,no.3,pp.531–538,1987. The authors declare that there are no conflicts of interest [8]D.Y.Leung,P.M.Davidson,G.B.Cranney,andW.F.Walsh, regarding the publication of this article. “Thromboembolic risks of left atrial thrombus detected by 10 Evidence-Based Complementary and Alternative Medicine

transesophageal echocardiogram,” American Journal of Cardi- [26] J. Mairal, M. Elad, and G. Sapiro, “Sparse representation for ology,vol.79,no.5,pp.626–629,1997. color image restoration,” IEEE Transactions on Image Processing, [9]R.M.Wyman,R.D.Safian,V.Portway,J.J.Skillman,R.G. vol.17,no.1,pp.53–69,2008. McKay, and D. S. Baim, “Current complications of diagnostic [27]T.Lu,S.Li,L.Fang,Y.Ma,andJ.A.Benediktsson,“Spectral- and therapeutic cardiac catheterization,” Journal of the Ameri- Spatial Adaptive Sparse Representation for Hyperspectral can College of Cardiology, vol. 12, no. 6, pp. 1400–1406, 1988. Image Denoising,” IEEE Transactions on Geoscience and Remote [10] D. J. Brenner and E. J. Hall, “Computed tomography, an Sensing,vol.54,no.1,pp.373–385,2016. increasing source of radiation exposure,” NEnglJMed,vol.357, [28] L. Kang, C. Yu, C. Lin, and C. Yeh, “Image and Video pp. 2277–2284, 2007. Restoration and Enhancement via Sparse Representation,” in [11] K. R. Nandalur, B. A. Dwamena, A. F. Choudhri, M. R. Emerging Technologies in Intelligent Applications for Image and Nandalur, and R. C. Carlos, “Diagnostic performance of stress Video Processing, Advances in Computational Intelligence and cardiac magnetic resonance imaging in the detection of coro- Robotics,pp.1–28,IGIGlobal,2016. nary artery disease: a meta-analysis,” Journal of the American [29] W. Dong, F. Fu, G. Shi et al., “Hyperspectral image super- College of Cardiology, vol. 50, no. 14, pp. 1343–1353, 2007. resolution via non-negative structured sparse representation,” [12] M. Clinic, “Boold tests for heart disease,” 2016, http://www IEEE Transactions on Image Processing,vol.25,no.5,pp.2337– .mayoclinic.org/diseases-conditions/heart-disease/in-depth/ 2352, 2016. heart-disease/art-20049357. [30] A. K. Jain, R. P. W. Duin, and J. Mao, “Statistical pattern recog- [13]B.H.Kim,S.H.Lee,D.U.Cho,andS.Y.Oh,“Aproposalof nition: a review,” IEEE Transactions on Pattern Analysis and heart diseases diagnosis method using analysis of face color,” Machine Intelligence,vol.22,no.1,pp.4–37,2000. in Proceedings of the 7th International Conference on Advanced [31]R.O.Duda,P.E.Hart,andD.G.Stork,Pattern classification, Language Processing and Web Information Technology (ALPIT John Wiley and Sons, 2012. ’08), pp. 220–225, July 2008. [32] K. Kreutz-Delgado, J. F. Murray, B. D. Rao, K. Engan, T.-W.Lee, [14] B. Zhang, B. V. K. V. Kumar, and D. Zhang, “Noninvasive and T. J. Sejnowski, “Dictionary learning algorithms for sparse diabetes mellitus detection using facial block color with a sparse representation,” Neural Computation,vol.15,no.2,pp.349–396, representation classifier,” IEEE Transactions on Biomedical Engi- 2003. neering,vol.61,no.4,pp.1027–1033,2014. [33]R.Memisevic,C.Zach,G.Hinton,andM.Pollefeys,“Gated [15]S.Cai,L.Zhang,W.Zuo,andX.Feng,“Aprobabilisticcollabo- softmax classification,”in Proceedings of the 24th Annual Confer- rative representation based approach for pattern classification,” ence on Neural Information Processing Systems 2010 (NIPS ’10), in Proceedings of the IEEE Conference on Computer Vision and December 2010. Pattern Recognition (CVPR ’16),pp.2950–2959,July2016. [34] C. E. Brodley and M. A. Friedl, “Decision tree classification [16] L. Zhang, M. Yang, and X. Feng, “Sparse representation or of land cover from remotely sensed data,” Remote Sensing of collaborative representation: Which helps face recognition?” in Environment,vol.61,no.3,pp.399–409,1997. Proceedings of the IEEE International Conference on Computer [35] A. Khammari, F. Nashashibi, Y. Abramson, and C. Laurgeau, Vision (ICCV ’11), pp. 471–478, Barcelona, Spain, November “Vehicle detection combining gradient analysis and AdaBoost 2011. classification,” in Proceedings of the 8th International IEEE [17] J.Wright,A.Y.Yang,A.Ganesh,S.S.Sastry,andY.Ma,“Robust Conference on Intelligent Transportation Systems,pp.1084–1089, face recognition via sparse representation,” IEEE Transactions aut, September 2005. on Pattern Analysis and Machine Intelligence,vol.31,no.2,pp. [36]Y.-D.Cai,K.-Y.Feng,W.-C.Lu,andK.-C.Chou,“UsingLogit- 210–227, 2009. Boost classifier to predict protein structural classes,” Journal of [18] X. Wang and D. Zhang, “An optimized tongue image color cor- Theoretical Biology,vol.238,no.1,pp.172–176,2006. rection scheme,” IEEE Transactions on Information Technology [37] U. Lipowezky, “Indoor-outdoor detector for mobile phone in Biomedicine,vol.14,no.6,pp.1355–1364,2010. cameras using gentle boosting,” in Proceedings of the Computer [19] H. Wang, Huangdi Neijing, New World Publish, 1999. Society Conference on Computer Vision and Pattern Recognition [20] Z. Bing and W. Hongcai, “Basic theories of traditional Chinese - Workshops (CVPRW ’10),pp.31–38,June2010. medicine,” Singing Dragon,2010. [38] A. G. Glaros and R. B. Kline, “Understanding the accuracy [21] Z. Bing and W. Hongcai, “Diagnostics of traditional Chinese of tests with cutting scores: The sensitivity, specificity, and medicine,” Singing Dragon,2010. predictive value model,” Journal of Clinical Psychology,vol.44, [22] S.-W. Youn, E.-S. Park, D.-H. Lee, C.-H. Huh, and K.-C. Park, no. 6, pp. 1013–1023, 1988. “Does facial sebum excretion really affect the development of acne?” British Journal of Dermatology,vol.153,no.5,pp.919– 924, 2005. [23] M. Liu and Z. Guo, “Hepatitis diagnosis using facial color image,” in Medical Biometrics,vol.4901ofLecture Notes in Computer Science, pp. 160–167, Springer, Berlin, Germany, 2007. [24] H. Zhang and V. M. Patel, “Sparse Representation-Based Open Set Recognition,” IEEE Transactions on Pattern Analysis and Machine Intelligence,vol.39,no.8,pp.1690–1696,2017. [25] S. Agarwal and D. Roth, “Learning a Sparse Representation for Object Detection,” in Computer Vision — ECCV 2002,vol. 2353 of Lecture Notes in Computer Science, pp. 113–127, Springer Berlin Heidelberg, Berlin, Heidelberg, 2002. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 3140403, 7 pages https://doi.org/10.1155/2017/3140403

Research Article Clinical Effects of Laser Acupuncture plus Chinese Cupping on the Pain and Plasma Cortisol Levels in Patients with Chronic Nonspecific Lower Back Pain: A Randomized Controlled Trial

Mu-Lien Lin,1,2 Jih-Huah Wu,3 Chi-Wan Lin,1 Chuan-Tsung Su,1 Hung-Chien Wu,4 Yong-Sheng Shih,3 I-Ting Chiu,3 Chao-Yi Chen,3 and Wen-Dien Chang5

1 Institute of Biomedical Engineering, National Taiwan University, Taipei, Taiwan 2Department of Pain Management, Taipei City Hospital Zhongxing Branch, Taipei, Taiwan 3Department of Biomedical Engineering, Ming Chuan University, Taoyuan, Taiwan 4Yi Sheng Chinese Medicine Clinic, Taipei, Taiwan 5Department of Sports Medicine, China Medical University, Taichung City, Taiwan

Correspondence should be addressed to Wen-Dien Chang; [email protected]

Received 6 April 2017; Revised 22 June 2017; Accepted 9 July 2017; Published 7 August 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Mu-Lien Lin et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Objectives. Chronic nonspecific lower back pain (LBP) is a common disease. Insufficient data is currently available to conclusively confirm the analgesic effects of laser acupuncture on LBP. This study evaluated the effectiveness of laser acupuncture plus Chinese cupping in LBP treatment. Methods. Patients with chronic nonspecific LBP were enrolled for a randomized controlled trial and assigned to the laser acupuncture group (laser acupuncture plus Chinese cupping) and control group (sham laser plus Chinese 2 cupping). Laser acupuncture (808 nm; 40 mW; 20 Hz; 15 J/cm ) and Chinese cupping were applied on the Weizhong (BL40) and Ashi acupoints for 5 consecutive days. Plasma cortisol levels were assessed before and after the 5-day treatment session. The visual analog scale (VAS) scores were recorded at baseline and throughout the 5-day treatment session. Results. After the treatment session, the plasma cortisol levels and VAS scores decreased significantly in both groups. In the laser acupuncture group, the VAS scores decreased significantly on days 4 and 5, and an enhanced reduction in VAS scores was observed. Conclusion. Laser acupuncture plus Chinese cupping at the Weizhong (BL40) and Ashi acupoints effectively reduced pain and inflammation in chronic nonspecific LBP. This therapy could be a suitable option for LBP treatment in clinical settings.

1. Introduction System in Taiwan. A systematic review and meta-analysis revealed that acupuncture treatment for LBP has moderate Chronic nonspecific lower back pain (LBP) is commonly < effects on pain relief and results in decreased functional observed in people aged 45 years in the United States. limitations at short-term follow-up [4]. Chinese cupping is Furthermore, LBP is the second most common reason for another traditional Chinese medicine treatment and uses clinic visits and the fifth most common cause of admission to hospitals for surgical procedures [1]. LBP causes spinal insta- open side of a cup against the skin, causing air pumping bility and eventually becomes chronic symptom [2]. How- todecreasetheairpressureinthecup.Itcouldcauselocal ever, conservative treatments are the first choice for patients stasis of blood, qi activation, and muscle microcirculation withchronicnonspecificLBP.Acupunctureisoneofthemost [5]. Yuan et al. indicated that Chinese cupping had good popular treatments applied in traditional Chinese medicine effects on reducing muscle pain and revealed that it was more and has been used for relieving pain in musculoskeletal effective at decreasing pain than analgesic in LBP patients [6]. diseases [3]. Acupuncture has been adopted as a clinical Combing acupuncture with Chinese cupping, the therapeutic practice guideline for LBP by the National Health Insurance effects can be enhanced [5]. Therefore, acupuncture plus 2 Evidence-Based Complementary and Alternative Medicine

Chinesecuppingmaybeusedasanimmediatetreatmentfor patients, who had other physical conditions, were excluded reducing LBP. because they did not satisfy the inclusion criteria. Random- In traditional Chinese medicine treatment, the clinical ization was conducted after 48 patients completed signing an value of acupuncture for musculoskeletal disease has been informed consent form. The sequence generation process was sufficiently established [7, 8]. Laser acupuncture is a type of based on a computer-generated random number table using low-level laser therapy (LLLT) on acupoints and is similar to Microsoft Excel (Microsoft Corp., Redmond, WA, USA). needle acupuncture. Karu explored the biostimulation effect Blinding of patients was achieved using a random number, of LLLT on cells [9]. In vitro, adenosine triphosphate can be where odd digit in ones is assigned to laser acupuncture group activated after He–Ne low-level laser irradiated on cells [10]. (laser instrument number 1 plus Chinese cupping) and even The photobiomodulation process involves light absorption digitinonesisassignedtocontrolgroup(laserinstrument by a photoreceptor and leads to signal transduction and number 2 plus Chinese cupping). There are 25 patients in molecular modulations in DNA and RNA syntheses caused laser acupuncture group and 23 patients in control group. by low-level laser radiation [11]. A systematic review and This study was approved by the Ethical Committee of Taipei meta-analysis of in vivo studies revealed that low-level Municipal Chung-Hsin Hospital (number TCHIRB-990304). laser acupuncture progressively alleviates temporomandibu- lar disorders, rheumatoid arthritis and gingival healing, and 2.2. Procedure. This study is a randomized controlled trial. musculoskeletal pain [12]. Moderate evidence supports that The patients with chronic nonspecific LBP who were assigned laser acupuncture exerts positive effects on musculoskeletal to the laser acupuncture and control groups received laser diseases [12]. Furthermore, LLLT exerts additional effects acupuncture plus Chinese cupping and sham laser acupunc- on LBP, including the inhibition of central nerve synaptic ture plus Chinese cupping, respectively. All the procedures activity, peripheral nerve blockade, neurotransmitter mod- were done in Taipei Municipal Chung-Hsin Hospital. Real ulation, and muscle spasm reduction [13]. However, studies laser acupuncture was applied by the laser instrument (num- on laser acupuncture for LBP are rare. Lin et al. used laser ber 1), and sham laser acupuncture was applied by the acupuncture plus soft cupping on the Weizhong (BL40) laser instrument (number 2). Sham laser acupuncture was acupoints for treating patients with LBP and demonstrated performed to appear like real laser instrument, applying the that laser acupuncture can increase the relative meridian val- same treatment procedure but without energy output. All ues in Ryodoraku analysis and alleviate pain [14]. Although patients were blinded to receiving a true laser (number 1) or evidence from traditional Chinese medicine has provided sham laser (number 2) acupuncture by a physician (Lin M. an explanation for these effects, the analgesic effects of laser L.). Plasma cortisol levels were recorded before and after the acupuncture are yet to be conclusively confirmed. Therefore, treatment session. The VAS scores for LBP were recorded at we administered laser acupuncture plus Chinese cupping to baseline and during days 1–5 of the treatment session. The patients with chronic nonspecific LBP and recorded their outcomes data were collected and analyzed by an assessor plasma cortisol levels and visual analog scale (VAS) scores to (Wu J. H.). The physician and assessors were blinded to assess the changes after the treatment. laser instrument and group assignments until the data were analyzed. Allocation concealment was achieved using opaque 2. Methods sealed envelopes containing information of the assignments. 2.1. Participant Selection. Patients from the Orthopedic Department of Taipei Municipal Chung-Hsin Hospital were 2.3. Treatment Protocol. A 4-channel 808 nm LLLT instru- included as study participants. Patients experiencing chronic ment LA400 (United Integrated Services Co., Ltd., Taiwan; output power, 40 mW; frequency, 20 Hz; duty cycle, 50%; nonspecific LBP for at least 3 months between the 12th rib 2 and gluteal fold were included in this study. Patients who were dosage, 15 J/cm ) with near infrared light was used. Laser pregnant and had other conditions, such as chronic obstruc- acupuncture was applied on the Weizhong (BL40) acupoints tive pulmonary disease, were excluded. They were requested (Figure 1) of the popliteal fossa and the Ashi acupoints on the not to intake drug and use ointment containing steroid. Low back muscles for 10 minutes. After laser acupuncture, four dosage (500 mg) of acetaminophen (one time a day) was 6 cm DongBang cups (DongBang Acupuncture, Kyunggi- administered to the laser acupuncture and control groups do, Korea) were used for Chinese cupping. After the cups during the study, if the pain was absolutely unbearable. were placed on the lower back muscles at the level of the ∗ ∗ Sample size was estimated using G Power software (G Power L2–5 lumbar spinal disks, suction of each cup was applied 3.1.9.2, Heinrich-Heine-Universitat¨ Dusseldorf,¨ Germany) until 1 cm of the skin was drawn up and then held for 5 and by referring to the patient data reported by Lin et al. minutes. The same treatment protocol was followed for the [14].ThemeansandstandarddeviationsoftheassessedVAS control group; however, patients in this group received sham scores in the experimental and control groups of the previous LLLT (laser instrument number 2) during laser acupuncture. study results were calculated, and the mean and standard The physician (Lin M. L.) administered the treatment to all deviation of the difference were 0.08 and 0.17, respectively. patients between 3 and 6 pm (exuberant flowing time of the The estimated sample size (𝛼 =0.05andpower=0.80)in bladder meridian) for 5 continuous days. our study was 20–25 patients per group. Finally, 50 patients who received a diagnosis of chronic nonspecific LBP were 2.3.1. Acupoint Selection. The patients were positioned on a recruited in our study. When we enrolled the participants, 2 treatment table with their knees maintained in slight flexion. Evidence-Based Complementary and Alternative Medicine 3

Table 1: Demographic data of two groups.

Laser acupuncture group Control group 𝑝 (𝑛=20) (𝑛=20) Sex 5/15 6/14 0.50 (male/female) Age (years) 63.09 ± 16.19 63.70 ± 15.69 0.76 Height (cm) 159.66 ± 9.10 158.60 ± 7.13 0.64 Weizhong (BL40) ± ± acupoint Weight (kg) 61.84 10.69 54.30 6.97 0.06 2 BMI (kg/m ) 24.26 ± 10.10 21.59 ± 7.05 0.78 Onset duration 4.13 ± 3.59 3.98 ± 2.83 0.67 (years)

Figure 1: Weizhong (BL40) acupoint. the demographic data, VAS, and cortisol level were analyzed using Mann–Whitney 𝑈 test and chi-square test. The VAS scores and plasma cortisol levels before and after the treat- The Weizhong (BL40) acupoint was palpated at the midpoint ment were compared using the paired-sample 𝑡-test. The ofthetransversecreaseandlocatedbetweenthebiceps differences between the two groups were analyzed using the femoris and semitendinosus tendons on the popliteal fossa. independent 𝑡-test. All statistical tests were two-tailed, and Subsequently, the lower back muscles of the patients were 𝑝 < 0.05 was considered statistically significant. palpated, and the tender point was then considered the Ashi acupoint. The acupoints were selected according to the acupoint theory of the traditional Chinese medicine and were 3. Results treated by the physician (Lin M. L.). In this study, 48 patients with chronic nonspecific LBP participated, and 40 patients completed the study procedure (Fig- 2.4. Outcome Measures ure2).Ofthe25patientsinthelaseracupuncturegroup,20 patients completed the treatment and 5 patients dropped out. 2.4.1. Pain. The VAS scores were recorded to assess the Of the 23 patients in the control group, 20 patients completed intensity of LBP at baseline and during days 1–5 of the the treatment and 3 patients dropped out. In two groups, treatment session. After the treatment session, the patients thesamereasonofdroppingoutwasthatthepatientswere were allowed to rest for approximately 15 minutes, and their too busy to complete the treatment session. There were no VAS scores were subsequently recorded. The tender points of adverse events in the both groups. Significant differences were their lower back muscles were palpated, and pain intensity of not observed in the age, weight, height, and body mass index tenderness was recorded by drawing a marker on a 100-mm of the two groups in per-protocol analysis (p > 0.05, Table 1). VAS. The same assessor performed the VAS assessment and The baseline plasma cortisol levels did not significantly was blinded to the group assignment. differ between the laser acupuncture and control groups (p = 0.65), and plasma cortisol levels for each group were 2.4.2. Blood Test. Plasma cortisol levels were analyzed before no significantly different after the treatment (p = 0.93; and after the treatment session. Blood samples were obtained Table 2). A comparison between the cortisol levels in the to measure cortisol levels before and after the treatment laser acupuncture group before and after the treatment session. Two 5-mL blood samples were obtained two times demonstrated increased plasma cortisol levels in 6 patients, from the cubital vein of each patient. These blood samples significantly decreased levels in 13 patients, and unchanged were allowed to stand for 30 minutes, followed by sample levels in 1 patient. In the control group, 6 and 14 patients centrifugation to obtain the plasma samples. Finally, the showed increased and decreased cortisol levels, respectively. blood supernatants were collected in two 1-mL tubes. All ∘ Moreover, plasma cortisol levels did not significantly decrease plasma cortisol samples were frozen at −70 C. One blood in both groups, and the effect size was 0.14 (95% CI = samplewasanalyzedbeforethetreatmentonday1,and −0.50∼0.78). the other sample was obtained after treatment on day 5. There was no significant difference in VAS between The cortisol analysis was performed at the Taipei Municipal groups before the treatment (p =0.44),buttheVASscorewas Chung-Hsin Hospital. significantlylowerinthelaseracupuncturegroupthaninthe control group after the treatment (p = 0.005). The VAS scores 2.5. Statistical Analyses. Data of the patients were ana- of all patients before and after the treatment were significantly lyzed by using the SPSS statistical software (SPSS Ver- decreasedinbothgroups(Table2),andtheeffectsizewas sion 17.0, Chicago, Illinois, USA). The per-protocol analysis −0.94 (95% CI = −1.62∼−0.27). Compared with the control was used to analyze the outcomes of laser acupuncture group, the laser acupuncture group exhibited significantly group and control group. Comparing between two groups, decreased VAS scores at days 4 and 5 (p < 0.05; Figure 3). 4 Evidence-Based Complementary and Alternative Medicine

Assessed for eligibility (n = 50)

Excluded (n = 2) Enrollment Not meeting inclusion criteria

Randomized (n = 48)

Allocation

Laser acupuncture group Control group (n = 25) (n = 23) Not attending all the Not attending all the treatment sessions treatment sessions (n = 5) (n = 3)

Analysis

Analyzed Analyzed (n = 20) (n = 20)

Figure 2: Consort flow diagram.

Table 2: Effects on VAS and cortisol level before and after the treatment session. Laser acupuncture group Control group (𝑛=20) (𝑛=20) a a Before After 𝑝 Before After 𝑝 ∗ VAS 6.75 ± 1.46 4.20 ± 1.88 0.001 6.84 ± 1.41 5.80 ± 1.41 0.001 Cortisol level 11.41 ± 4.57 8.90 ± 3.78 0.01 10.57 ± 4.74 8.38 ± 3.64 0.03 a ∗ 𝑝 meant the significant difference before and after treatment session. 𝑝 <0.05, laser acupuncture versus control group.

Furthermore, the laser acupuncture group tended to have medicine, qi and blood stagnation leading to yang deficiency lower VAS scores than the control group. are the causes of LBP. Liu et al. indicated that massage, Chinese cupping, and acupuncture on the Weizhong (BL40) 4. Discussion acupoint can improve qi and blood circulation and reduce LBP[4].Ourfindingsareconsistentwiththeresultsof LBPisthemostcommondiseaseworldwide,andtotal previous studies [4, 13], which showed that laser acupuncture treatment costs for LBP exceed $100 billion per year in plus Chinese cupping alleviates chronic nonspecific LBP. the United States [15]. Medical expenditure of patients with A pulse semiconductor laser (wavelength, 808 nm; output 2 LBP has increased substantially. Therefore, we explored the power, 40 mW; energy density, 15 J/cm )wasusedinthis laseracupunctureplusChinesecuppingtoalleviatepainand study. Of the 50 patients with chronic nonspecific LBP plasma cortisol levels in patients with chronic nonspecific recruited in this study, 40 completed the treatment. The LBP. The Weizhong (BL40) acupoint is an acupuncture point plasma cortisol levels of each patient were recorded before of the bladder meridian with two branches traveling down and after the treatment. Furthermore, all patients received from the upper back to the sacrum near the spine. The qi and equal attention; low dose acetaminophen was administered blood circulation of the bladder meridian converge on the to the laser acupuncture and control groups during the Weizhong (BL40) acupoint. According to traditional Chinese study. Moreover, the patients were allowed to withdraw from Evidence-Based Complementary and Alternative Medicine 5

9 analgesic effects of laser acupuncture with adequate energy 8 power were confirmed. The wavelength and energy density of laser acupuncture 7 arerelatedtotheenergyabsorbedfromthelaserradiation 6 by superficial tissues [21]. A previous study revealed that ∗ 5 maximum penetration of the deeper tissue was achieved at ∗ 770–850 nm wavelength of the LLLT [22]. Hudson et al. 4 measured the energy density across different thicknesses of

Visual analog scale 3 thesampletissueandrevealedthat808nmwavelengthof 2 the LLLT had improved penetration [23]. Anders and Wu indicated that the depth of penetration is a crucial factor 1 andreportedthatapulsedlaseroutputcanachieveimproved Baseline Day 1 Day 2 Day 3 Day 4 Day 5 penetration effects compared with a continuous wave laser output [24]. Lin et al. applied laser acupuncture with a pulsed 2 Laser acupuncture group laser output and 15 J/cm of energy density in patients with Control group LBP and demonstrated that suitable parameters can be used on the acupoint [14]. Therefore, the parameters (wavelength, Figure 3: The change of VAS (mean and 95% confidence intervals) 2 ∗ 808 nm; energy density, 15 J/cm ) of the pulsed laser output in two groups during the treatment session. 𝑝 < 0.05. could be used for laser acupuncture treatment and to exert positive effects following the Arndt–Schulz law [12]. Laser acupuncture is a new noninvasive treatment and the study if they did not want to follow the treatment is easy to use. Compared with needle acupuncture, laser protocol. The results showed most significant changes in the acupuncture is advantageous because of decreased instances cortisol levels and VAS scores of the two groups after the of infection and lack of pain from needles [18]. According treatment, with significant changes on days 4 and 5. This to traditional Chinese medicine, qi is a type of body energy study demonstrated significantly decreased plasma cortisol that can be obtained through receiving needle acupuncture levelsinthetwogroups.Theplasmacortisollevelsweremore treatment [3]. Laser acupuncture does not provide physical affected in the laser acupuncture group than in the control stimulation to obtain qi but exerts photostimulation effects group. Although the plasma cortisol levels were significantly through LLLT to cause biomodulation. The physiological decreasedinthetwogroups,somepatientshadincreased mechanism of LLLT has been adequately established in cellu- plasma cortisol levels. In both groups, increased plasma lar experiments. LLLT has been used in culture experiments, cortisol levels of some patients after the 5-day treatment in which it affects the survival of plasmids in Escherichia mightindicatethattheaccumulativedosagewasinadequate coli cells, and DNA repair was produced by pulsed emission andcouldnotreducetheirplasmacortisollevels.The modes through low-level infrared laser action [25]. Further- previous study showed that serum cortisol levels increased more, AlGhamdi et al. demonstrated that LLLT is useful in significantly after needling on traditional acupuncture loci enhancing the proliferation rate of cell lines [26]. LLLT alters [16]. Furthermore, in our study, Chinese cupping reduced photoreceptor functions and subsequent cellular signaling the plasma cortisol levels of the control group despite the and cellular functions [27]. Five hypotheses describe the inactive laser. Kim et al. suggested that compared with usual primary reactions after laser light absorption, namely, the care (p < 0.01)andanalgesia(p < 0.001), Chinese cupping single-oxygen hypothesis, redox property alteration hypothe- significantly reduced LBP in randomized clinical trials [17]. sis, nitric oxide hypothesis, transient local heating hypothesis, Figure 2 indicates that the plasma cortisol levels were more and superoxide anion hypothesis [28]. Secondary reactions decreasedinthelaseracupuncturegroupthaninthecontrol of photobiomodulation involved a cellular signaling pathway, group. According to our study data, patients with chronic which involves the cell between the photoacceptor and a nonspecific LBP experienced pain relief on days 4and5. nucleus[28,29].Thechemicalenergywithinthecellcan Laser beams are monochromatic, coherent, and colli- be converted in the form of adenosine triphosphate, which mated. Energy power, wavelength, and energy density are leads to pain relief and cell proliferation after the absorption the three important parameters of laser acupuncture. Energy of photonic energy. The tissues on acupoints were excited power of the LLLT was considered as the treatment dose through LLLT, thus obtaining qi, similar to the process [13]. A systematic review by Baxter et al. demonstrated the involved in the needle acupuncture technique [12]. Therefore, moderate effects of laser acupuncture (energy power, 10 mW) laser acupuncture can exert analgesia effects in accordance in reducing muscular pain [18]. Ceylan et al. applied laser with traditional Chinese medicine acupuncture theory. acupuncture with 8 mW of energy power on myofascial pain, The hypothalamopituitary-adrenal axis is involved in which had an effect size of 0.81 for pain relief [19]. Laser stressors that play a role in chronic pain [30]. It is the acupuncture with 11.20 mW of energy power has been used principal neuroendocrine system that activates the central for alleviating myofascial pain and fibromyalgia, with effect nervous system of mammals and produces cortisol as end sizesof0.97and1.31,respectively,forpainrelief[20,21]. product of being affected by stressors. The present results We used laser acupuncture with 40 mW of energy power to revealed that pain is alleviated by decreased stress hormone decrease the VAS scores of LBP with an effect size of 1.21. The (i.e., cortisol) levels after a 5-day treatment session. The pain 6 Evidence-Based Complementary and Alternative Medicine was significantly reduced as depicted by the VAS scores [4]L.Liu,M.Skinner,S.McDonough,L.Mabire,andG.D. after laser acupuncture plus Chinese cupping. However, the Baxter, “Acupuncture for low back pain: An overview of system- changes in pain severity were not apparent when patients atic reviews,” Evidence-based Complementary and Alternative received only Chinese cupping five times. The patients with Medicine,vol.2015,ArticleID328196,2015. chronic nonspecific LBP experienced pain relief after five [5] T. Li, Y. Li, Y. Lin, and K. Li, “Significant and sustaining elevation of blood oxygen induced by Chinese cupping therapy treatment sessions. Therefore, we proposed that the improve- as assessed by near-infrared spectroscopy,” Biomedical Optics ments after laser acupuncture and Chinese cupping at the Express,vol.8,no.1,pp.223–229,2017. Weizhong (BL40) acupoints are similar to the effects exerted [6] Q. L. Yuan, T. M. Guo, L. Liu, F. Sun, and Y. G. Zhang, by traditional acupuncture. In Taylor’s study, cortisol levels “Traditional Chinese medicine for neck pain and low back pain: decreased when 𝛽-endorphin was below the basal level in asystematicreviewandmeta-analysis,”PLOS ONE,vol.10,no. human participants [31]. In future studies, we suggested 2, Article ID e0117146, 2015. using an enzyme-linked immunosorbent assay to evaluate the [7]D.WongLitWan,Y.Wang,C.C.L.Xue,L.P.Wang,F.R. variations in 𝛽-endorphin and cortisol levels in patients with Liang, and Z. Zheng, “Local and distant acupuncture points chronic nonspecific LBP. It is an in-depth exploration where stimulation for chronic musculoskeletal pain: a systematic laser acupuncture plus Chinese cupping for consecutive review on the comparative effects,” European Journal of Pain (United Kingdom),vol.19,no.9,pp.1232–1247,2015. periodsmayalleviateLBPandinfluencetherelative𝛽- [8] M. Lam, R. Galvin, and P. Curry, “Effectiveness of acupuncture endorphin and cortisol levels. for nonspecific chronic low back pain: a systematic review and There are several limitations in this study. First, VAS scale meta-analysis,” Spine,vol.38,no.24,pp.2124–2138,2013. is a clinical tool for self-perception pain assessment. However, [9] T. Karu, “Primary and secondary mechanisms of action of usageofpainpressurealgometercanbeadvantageousfor visible to near-IR radiation on cells,” Journal of Photochemistry quantifying the pain threshold on affected muscle in LBP and Photobiology B: Biology,vol.49,no.1,pp.1–17,1999. patients. Second, LBP could affect the patients’ activities [10] P. Rola, A. Doroszko, and A. Derkacz, “The use of low-level of daily living and functional ability, so comprehensive energy laser radiation in basic and clinical research,” Advances evaluations for LBP patients, that is, assessments of disability in Clinical and Experimental Medicine,vol.23,no.5,pp.835– or functional limitation, could get more outcomes of the 842, 2014. treatment. Third, our study also lacks follow-up, and it cannot [11] A. D. Liebert, B. T. Bicknell, and R. D. Adams, “Protein prove the long-term effects of laser acupuncture plus Chinese conformational modulation by photons: A mechanism for laser treatment effects,” Medical Hypotheses,vol.82,no.3,pp.275– cupping. 281, 2014. [12] D.Law,S.McDonough,C.Bleakley,G.D.Baxter,andS.Tumilty, 5. Conclusion “Laser acupuncture for treating musculoskeletal pain: a systematic review with meta-analysis,” JAMS Journal of Acupuncture The present results indicated that laser acupuncture com- and Meridian Studies,vol.8,no.1,pp.2–16,2015. bined with Chinese cupping at the Weizhong (BL40) and [13] G. Glazov, M. Yelland, and J. Emery, “Low-level laser therapy Ashi acupoints can alleviate LBP symptoms. The variations for chronic non-specific low back pain: A meta-analysis of in plasma cortisol levels indicated that laser acupuncture randomised controlled trials,” Acupuncture in Medicine,vol.34, plus Chinese cupping is an effective pain relief treatment. no.5,pp.328–341,2016. Moreover, laser acupuncture plus Chinese cupping at the [14] M.-L. Lin, H.-C. Wu, Y.-H. Hsieh et al., “Evaluation of the effect of laser acupuncture and cupping with ryodoraku and visual Weizhong (BL40) and Ashi acupoints effectively reduces LBP. analog scale on low back pain,” Evidence-Based Complementary This therapy could be a suitable treatment option for patients and Alternative Medicine, vol. 2012, Article ID 521612, 7 pages, with chronic nonspecific LBP in clinical settings. 2012. [15]L.Andronis,P.Kinghorn,S.Qiao,D.G.T.Whitehurst,S. Conflicts of Interest Durrell, and H. McLeod, “Cost-effectiveness of non-invasive and non-pharmacological interventions for low back pain: a The authors declare that they have no conflicts of interest. systematic literature review,” AppliedHealthEconomicsand Health Policy,vol.15,no.2,pp.173–201,2017. [16] S. Lee, S. Yin, M. Lee, W. Tsai, and C. Sim, “Effects of acupunc- References ture on serum cortisol level and dopamine beta-hydroxylase activity in normal chinese,” The American Journal of Chinese [1] C. E. Ladeira, “Evidence based practice guidelines for anage- Medicine,vol.10,no.1–4,pp.62–69,1982. ment of low back pain: Physical herapy implications,” Revista [17] J.-I. Kim, M. S. Lee, D.-H. Lee, K. Boddy, and E. Ernst, Brasileira de Fisioterapia,vol.15,no.3,pp.190–199,2011. “Cupping for treating pain: a systematic review,” Evidence-Based [2] M. M. Panjabi, “Clinical spinal instability and low back pain,” Complementary and Alternative Medicine, vol. 2011, Article ID Journal of Electromyography and Kinesiology,vol.13,no.4,pp. 467014, 7 pages, 2011. 371–379, 2003. [18] G. D. Baxter, C. Bleakley, and S. McDonough, “Clinical effec- [3] Y.-T. Liu, C.-W. Chiu, C.-F. Chang et al., “Efficacy and safety of tiveness of laser acupuncture: a systematic review,” Journal of acupuncture for acute low back pain in emergency department: Acupuncture and Meridian Studies,vol.1,no.2,pp.65–82,2008. a pilot cohort study,” Evidence-Based Complementary and Alter- [19]Y.Ceylan,S.Hizmetli,andY.Silig,ˇ “The effects of infrared native Medicine,vol.2015,ArticleID179731,8pages,2015. laser and medical treatments on pain and serotonin degradation Evidence-Based Complementary and Alternative Medicine 7

products in patients with myofascial pain syndrome. A controlled trial,” Rheumatology International,vol.24,no.5,pp.260– 263, 2004. [20] A. Gur, A. J. Sarac, R. Cevik, O. Altindag, and S. Sarac, “Efficacy of 904 nm Gallium Arsenide low level laser therapy in the management of chronic myofascial pain in the neck: a double- blind and randomize-controlled trial,” Lasers in Surgery and Medicine,vol.35,no.3,pp.229–235,2004. [21] A. Gur,¨ M. Karakoç, K. Nas, R. Ç evik, J. Saraç, and E. Demir, “Efficacy of low power laser therapy in fibromyalgia: a single- blind, placebo-controlled trial,” Lasers in Medical Science,vol. 17,no.1,pp.57–61,2002. [22]K.R.Byrnes,R.W.Waynant,I.K.Ilevetal.,“Lightpro- motes regeneration and functional recovery and alters the immune response after spinal cord injury,” Lasers in Surgery and Medicine,vol.36,no.3,pp.171–185,2005. [23] D. E. Hudson, D. O. Hudson, J. M. Wininger, and B. D. Richardson,“Penetrationoflaserlightat808and980nmin bovine tissue samples,” Photomedicine and Laser Surgery,vol. 31,no.4,pp.163–168,2013. [24] J. J. Anders and X. Wu, “Comparison of light penetration of continuous wave 810 nm and superpulsed 904 nm wavelength light in anesthetized rats,” Photomedicine and Laser Surgery,vol. 34,no.9,pp.418–424,2016. [25] L. L. Barboza, V. M. A. Campos, L. A. G. Magalhaes,˜ F. Paoli, and A. S. Fonseca, “Low-intensity red and infrared laser effects at high fluences on Escherichia coli cultures,” Brazilian Journal of Medical and Biological Research,vol.48,no.10,pp.945–952, 2015. [26] K. M. AlGhamdi, A. Kumar, and N. A. Moussa, “Low-level laser therapy: a useful technique for enhancing the proliferation of various cultured cells,” Lasers in Medical Science,vol.27,no.1, pp. 237–249, 2012. [27]X.Xu,X.Zhao,T.C.-Y.Liu,andH.Pan,“Low-intensitylaser irradiation improves the mitochondrial dysfunction of C2C12 induced by electrical stimulation,” Photomedicine and Laser Surgery,vol.26,no.3,pp.197–202,2008. [28] S. Farivar, T. Malekshahabi, and R. Shiari, “Biological effects of low level laser therapy,” JournalofLasersinMedicalSciences,vol. 5, no. 2, pp. 58–62, 2014. [29] T. I. Karu, “Mitochondrial signaling in mammalian cells activated by red and near-IR radiation,” Photochemistry and Photobiology,vol.84,no.5,pp.1091–1099,2008. [30] E. A. Young, J. Abelson, and S. L. Lightman, “Cortisol pulsatility and its role in stress regulation and health,” Frontiers in Neuroendocrinology,vol.25,no.2,pp.69–76,2004. [31]T.Taylor,R.G.Dluhy,andG.H.Williams,“𝛽-endorphin suppresses adrenocorticotropin and cortisol levels in normal human subjects,” Journal of Clinical Endocrinology and Meta- bolism,vol.57,no.3,pp.592–596,1983. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 4218468, 8 pages https://doi.org/10.1155/2017/4218468

Research Article Ethanol Extract of Mylabris phalerata Inhibits M2 Polarization Induced by Recombinant IL-4 and IL-13 in Murine Macrophages

Hwan-Suck Chung, Bong-Seon Lee, and Jin Yeul Ma

Korean Medicine (KM) Application Center, Korea Institute of Oriental Medicine (KIOM), 70 Cheomdan-ro, Dong-gu, Daegu 41062, Republic of Korea

Correspondence should be addressed to Jin Yeul Ma; [email protected]

Received 26 April 2017; Accepted 21 June 2017; Published 25 July 2017

Academic Editor: Woojin Kim

Copyright © 2017 Hwan-Suck Chung et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Mylabris phalerata (MP) is an insect used in oriental herbal treatments for tumor, tinea infections, and stroke. Recent studies have shown that tumor-associated macrophages (TAM) have detrimental roles such as tumor progression, angiogenesis, and metastasis. Although TAM has phenotypes and characteristics in common with M2-polarized macrophages, M1 macrophages have tumor suppression and immune stimulation effects. Medicines polarizing macrophages to M1 have been suggested to have anticancer effects via the modulation of the tumor microenvironment. In this line, we screened oriental medicines to find M1 polarizing medicines in M2-polarized macrophages. Among approximately 400 types of oriental medicine, the ethanol extract of M. phalerata (EMP) was the most proficient in increasing TNF-𝛼 secretion in M2-polarized macrophages and TAM. Although EMP enhanced the levels of an M1 cytokine (TNF-𝛼) and a marker (CD86), it significantly reduced the levels of an M2 marker (arginase-1) in M2- polarized macrophages. In addition, EMP-treated macrophages increased the levels of M1 markers (Inos and Tnf-𝛼)andreduced those of the enhanced M2 markers (Fizz-1, Ym-1, and arginase-1). EMP-treated macrophages significantly reduced Lewis lung carcinoma cell migration in a transwell migration assay and inhibited EL4-luc2 lymphoma proliferation. In our mechanism study, EMP was found to inhibit STAT3 phosphorylation in M2-polarized macrophages. These results suggest that EMP is effective in treating TAM-mediated tumor progression and metastasis.

1. Introduction many researchers have studied the control of TAM and have shownthatswitchingTAMorM2withM1significantly Tumor burden comprises a group of heterogeneous cells, inhibits tumor progression and metastasis [5–7]. Therefore, including T cells, neutrophils, and macrophages. The major switching TAM or M2 with M1 is a potential target for cancer cells comprising the tumor burden are macrophages, ac- treatment. counting for approximately 50% of the burden. Tumor- Mylabris phalerata (MP) is an insect used in the prepa- associated macrophages (TAM) are involved in tumor pro- ration of Mylabris, a Korean medicine listed in the Korean gression and metastasis, and the number of TAM in the Herbal Pharmacopoeia that is used to treat tumors. MP tumor burden is positively correlated with poor prognosis [1]. has antitumor effects and a proliferative effect on leucocytes Macrophages can be M1 polarized by stimulation with IFN- 𝛾 or LPS, and these M1-polarized macrophages secrete IL-12, [8, 9]. The major component of MP, cantharidin, also has TNF-𝛼,andIL-1𝛽, which kill cancer cells [2]. However, M2 anticancer and apoptotic effects on cancer cells [10]. Norcan- macrophages are polarized by stimulation with IL-4, IL-13, tharidin, a demethylated form of cantharidin, is used as an or M-CSF and release IL-10, CCL17, and CCL22, which help anticancer drug in China [11]. in tumor progression and metastasis. M2 macrophages have We screened 400 herbal ethanol extracts to examine phenotypes and functions similar to TAM [3, 4]. Because the effect of M1 polarization on M2-polarized macrophages TAM play a critical role in tumor progression and metastasis, induced by IL-4 and IL-13. We found that the ethanol extract 2 Evidence-Based Complementary and Alternative Medicine of MP (EMP) polarized M2 into M1 and that this effect was were harvested. Total RNA was extracted using the EasyBlue not mediated by endotoxins. RNA extraction kit (iNtRON Biotechnology, Inc., Seongnam, Korea). The quality and concentration of the RNA were 2. Materials and Methods assayed using the ND-1000 spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA). cDNA was syn- 2.1. Bone Marrow Macrophage Culture. Animal procedures thesized using CycleScript Reverse Transcriptase (Bioneer, ∘ were approved by the IACUC in the Korea Institute of Seoul, Korea) and stored at −20 C. Real-time PCR was con- Oriental Medicine. Bone marrow cells (BMC) were isolated ducted using the CFX96 Touch Real-Time PCR System (Bio- from the tibia and femur of 6-week-old male C57BL/6 Rad, CA, USA) employing the AccuPower GreenStar qPCR mice (Samtako Bio Korea, Gyeonggi-do, South Korea). Bone Master Mix (Bioneer, Daejeon, Korea). The PCR protocol ∘ ∘ marrow macrophages (BMM) generated using BMC were comprised 10 min at 95 Cfollowedby45cyclesof10sat95 C, ∘ ∘ differentiated in the RPMI1640 medium supplemented with 10 s at 60 C, and 10 s at 72 C. After the cycles were completed, ∘ ∘ 10% FBS and macrophage colony-stimulating factors (M- the signal at each temperature between 65 Cand95Cwas CSF, 60 ng/mL, Peprotech, Rocky Hill, NJ, USA) for 1 week. recorded to generate a dissociation curve. The sequences of The medium was replaced with a fresh M-CSF-containing the murine primers were listed in Table 1. The target mRNA medium 3 days after seeding the cells. levels were compared by calculating the crossing point (Cp) value and normalized to the reference gene GAPDH. 2.2. TAM Culture. To prepare TAM, mice were subcutaneously implanted with Lewis lung carcinoma (LLC) cells (2 × 2.6. Preparation of Nuclear and Cytoplasmic Extracts. 5 6 10 /mouse). They were sacrificed after 3 weeks, and tumor RAW264.7 cells (5 × 10 )werepretreatedwithIL-4(20 tissues were isolated. Single cells were dissociated from tumor ng/mL) + IL-13 (20 ng/mL) for 6 h to polarize to M2 and then tissues using a tumor dissociation kit (cat. 130-096-730, treated with EMP for 1 h. Nuclear and cytoplasmic extracts Miltenyi Biotec, Bergisch Gladbach, Germany) following the were prepared using NE-PER Nuclear and Cytoplasmic manufacturer’s instructions. To separate the macrophages, Extraction Reagents (Thermo Fisher Scientific, Rockford, IL, the cells were labelled with CD11b microbeads (cat. 130-049- USA) according to the manufacturer’s protocol. 601, Miltenyi Biotec), and the CD11b+ cells (macrophages) were isolated with MACS columns. Approximately 10%–20% 2.7. Western Blotting. Cells were washed with phosphate- of the tumor-dissociated cells were CD11b+. When we ana- buffered saline (PBS) and lysed using the radioimmuno- lyzed the purity of TAM, over 90% were CD11b+. precipitation assay buffer (Millipore, MA, USA) containing protease and phosphatase inhibitors. Total protein 2.3. Preparation of EMP. MP was purchased from an herbal (15–20 𝜇g) was separated by 10% SDS-PAGE gel electrophore- supplier (Yeongcheon herb, Yeongcheon, Korea), and a sis, transferred to polyvinylidene fluoride (PVDF) mem- voucher specimen (number E233) was deposited in the brane, and immunoblotted with specific antibody. Antibod- herbal bank of the Korea Medicine Application Center, Korea ies for arginase-1, 𝛽-actin (Santa Cruz Biotechnology, CA, Institute of Oriental Medicine. To prepare EMP, dried MP USA), phosphorylated signal transducer and activator of (30g)wasgroundintoafinepowder,soakedin300mL transcription 3 (p-STAT3) (Tyr705), p-STAT6 (Tyr641), P65, of 70% ethanol, and extracted in a shaking incubator at p-i𝜅B-𝛼, and proliferating cell nuclear antigen (PCNA) (Cell ∘ 40 C for 24 h. The extract was filtered through a testing Singling Technology, MA, USA) were used in this study. sieve (150 𝜇m; Retsch, Haan, Germany), evaporated on a Chemiluminescent signals were detected using the Chemi- rotary evaporator, concentrated by lyophilization, and then Doc imaging system (Bio-Rad Laboratories, CA, USA) and ∘ stored at −20 C. EMP powder (50 mg) was dissolved in a chemiluminescence reagent (Thermo Scientific, Rockford, 10mLof50%ethanol(v/v)andfilteredthrougha0.22𝜇m IL, USA). disk filter. Endotoxin was examined using the Pierce LAL chromogenic endotoxin quantitation kit (Thermo Scientific, 2.8. LLC Tumor Cell Migration Assay. Cell migration was Bonn, Germany) according to the manufacturer’s protocol. assayed using a 24-transwell chamber with a diameter of 6.5 mm and an 8 𝜇m pore polyethylene terephthalate (PET) 2.4. TNF-𝛼 and TGF-𝛽 Analysis. To polarize BMM to M2, membrane (SPL Lifesciences, Seoul, Korea) as described by they were treated with recombinant IL-4 (20 ng/mL) and Kimetal.[12]. IL-13 (20 ng/mL) for 6 h. EMP was added for 18 h, and the ∘ supernatants were harvested and kept at −80 Cuntiluse. 2.9. Tumor Cell Proliferation Assay. EL4-luc2 cells, a lym- TNF-𝛼 and TGF-𝛽 were analyzed by OptEIA ELISA kit (BD phoma cell line from C57BL/6 mice expressing the firefly Biosciences Pharmingen, San Diego, CA, USA) and eBio- luciferase gene (Caliper Life Science, MA, USA), were used science6 Human/Mouse TGF beta 1 ELISA Ready-SET-Go!6 to evaluate drug efficacy on macrophage tumoricidal activity Kit, 2nd Generation (cat. 88-8350-76, eBioscience, San Diego, in coculture conditions. BMM were pretreated with IL-4 CA, USA), respectively, following manufacturer’s instruction. (20 ng/mL) + IL-13 (20 ng/mL) for 6 h to polarize to M2 and then with EMP for 18 h. The cells were washed with 4 2.5. Real-Time PCR. To polarize BMM to M2, they were DPBS and 3 × 10 cells in 200 𝜇l of media were seeded 4 treated with recombinant IL-4 (20 ng/mL) and IL-13 (20 in 96-well white plates. EL4-luc2 cells (1 × 10 cells) were ng/mL) for 6 h. EMP was added for 18 h, and the cells cocultured with BMM for 48 h. Luciferin (150 𝜇g/mL) was Evidence-Based Complementary and Alternative Medicine 3

Table 1: Primers used for real-time RT-PCR. Target gene Primer sequence 󸀠 󸀠 𝛼 F: 5 -TTCTGTCTACTGAACTTCGGGGTGATCGGTCC-3 TNF- 󸀠 󸀠 R: 5 -GTATGAGATAGCAAATCGGCTGACGGTGTGGG-3 󸀠 󸀠 F: 5 -TGGAGAATAAGGTCAAGGAAC-3 Fizz1 󸀠 󸀠 R: 5 -GTCAACGAGTAAGCACAGG-3 󸀠 󸀠 F: 5 -CATTCAGTCAGTTAT CAGATTCC-3 YM1 󸀠 󸀠 R: 5 -AGTGAGTAGCAGCCTTGG-3 󸀠 󸀠 F: 5 -GGCAGCCTGTGAGACCTTTG-3 iNOS 󸀠 󸀠 R: 5 -GCATTGGAAGTGAAGCGTTTC-3 󸀠 󸀠 F: 5 -AGACAGCAGAGGAGGTGAAGAG-3 Arginase 1 󸀠 󸀠 R: 5 -CGAAGCAAGCCAAGGTTAAAGC-3 󸀠 󸀠 F: 5 -AGTGGCAGGTGGCTTATG-3 MMR 󸀠 󸀠 R: 5 -GGTTCAGGAGTTGTTGTG-3 󸀠 󸀠 F: 5 -ACCCAGAAGACTGTGGATGG-3 GAPDH 󸀠 󸀠 R: 5 -CACATTGGGGGTAGGAACAC-3 F, for ward; R, reverse.

added and luminescence was detected using the SpectraMax flow cytometry after EMP treatment in M2 macrophages, L microplate reader (Molecular Devices, Sunnyvale, CA, EMP significantly increased CD86 expression but did not USA). affect CD68 expression (Figure 2(b)). Arginase-1 catalyzes L- arginine as a substrate and produces L-ornithine and urea. 2.10. Statistical Analysis. All values are expressed as means ± It is known that the depletion of L-arginine by arginase-1 SEM. The statistical significance (𝑝 < 0.05 for all analyses) could inhibit the L-arginine-dependent immune functions was assessed by one-way ANOVA followed by Tukey’s post [13]. For instance, L-arginine depletion suppresses T-cell hoc test for multiple comparisons using the Prism 5.01 proliferation [14]. Although macrophages polarized to M2 by software (GraphPad Software Inc., San Diego, CA, USA). IL-4 and IL-13 displayed a significantly increased expression of arginase-1, EMP alleviated the increased expression of arginase-1 in a dose-dependent manner. Intriguingly, LPS did 3. Results not significantly alter the increased expression of arginase-1 3.1. Increased TNF-𝛼 and Decreased TGF-𝛽 Release by EMP in (Figure 2(c)). M2 Macrophages and TAM. Because TNF-𝛼 is a prominent M1 marker, we screened 400 types of herbal extracts for 3.3. EMP-Treated Macrophages Attenuate LLC Tumor Migra- their effect on TNF-𝛼 release in M2 macrophages. M2 tion and EL4-luc2 Lymphoma Proliferation. We studied the macrophages were induced by treating BMM with mouse effect of macrophages polarized by EMP on LLC tumor recombinant IL-4 (20 ng/mL) and IL-13 (20 ng/mL) for 6 h, cell migration. To exclude a direct effect of EMP on LLC, after which EMP was added for 18 h. Among the 400 herbal EMP-treated macrophages were washed out with DPBS and extracts, EMP showed the strongest effect on TNF-𝛼 release the macrophages were seeded in the lower compartment in M2 macrophages. TNF-𝛼 induction in M2 macrophages by andthenLLCwereculturedintheuppercompartment EMP was also shown in TAM (Figure 1). On the other hand, of the transwell chamber. As shown in Figure 3(a), EMP TGF-𝛽 is a typical M2 marker. TGF-𝛽 release was reduced treatment in M2 macrophages attenuated the migration of by EMP treatment in TAM and BMM (Figure 1). To exclude LLC tumor cells in a dose-dependent manner. We also eval- the possibility of TNF-𝛼 release by endotoxin contamination uated the tumoricidal activity of EMP-treated macrophages in EMP, we also examined the endotoxin level in EMP and in coculture conditions with EL4-luc2 lymphoma. LPS- and foundthatitwaslessthan0.1EU/mL(datanotshown). EMP-treated macrophages significantly reduced EL4-luc2 proliferation compared with the control group (Figure 3(b)). 3.2. Enhanced M1 and Reduced M2 Markers after EMP These data show that EMP-treated M2 macrophages can TreatmentinM2Macrophages. We analyzed the expression inhibit tumor metastasis and progression. of M1 and M2 genes after EMP treatment in M2 macrophages. Although the increased M2 markers (Fizz1, Ym1,andArg1) 3.4. EMP Inhibits STAT3 Phosphorylation in M2 Macrophages. were significantly inhibited by EMP,Tnfa M1( and Inos) To study the mechanism of EMP in M2 macrophages, markers were significantly increased by EMP based on the we analyzed STAT6 phosphorylation, which is a critical real-time RT-PCR analysis (Figure 2(a)). When we ana- transcription factor in the M2 polarization induced by IL- lyzed M1 (CD86) and M2 (CD68) phenotype changes using 4 and IL-13. Although M2 macrophages increased STAT6 4 Evidence-Based Complementary and Alternative Medicine

1500 1500 BMM TAM ###

∗∗∗ 1000 1000 (pg/ml) (pg/ml) ∗∗∗  TNF- 500  TNF- 500

0 0 B102550 B 0 10 25 50  EMP ( g/ml) IL-4 (20 ng/ml)+IL-13 (20 ng/ml) + EMP (g/ml) 150 150

100 100 ∗ ∗ # ∗∗ level (%) level level (%) level

 TGF- 50  TGF-

0 0 B102550 B010 25 50 EMP (g/ml) IL-4 (20 ng/ml)+IL-13 (20 ng/ml) + EMP (g/ml) (a) (b) Figure 1: Increased TNF-𝛼 and decreased TGF-𝛽 in M2 macrophages and TAM by treatment with EMP. (a) TAM were isolated with CD11b + microbeads after dissociation of tumor tissue into single cells. CD11b TAM were treated with various doses of EMP for 24 h. (b) BMM were pretreated with IL-4 (20 ng/mL) + IL-13 (20 ng/mL) for 6 h to polarize to M2 and then EMP was added for 18 h. TNF-𝛼 and TGF-𝛽 release ∗ ∗∗ ∗∗∗ were assessed by ELISA. Values are indicated as mean ± SEM. 𝑝 < 0.05, 𝑝 < 0.01,and 𝑝 < 0.001, compared with the B (Blank) samples; # ### 𝑝 < 0.05 and 𝑝 < 0.001, compared with the 0 (EMP nontreated) samples.

phosphorylation, there were no significant differences in Fizz1, and arginase-1 in M2 polarized macrophages by IL-4 STAT6 phosphorylation upon EMP treatment (Figure 4(a)). and IL-13. In addition, EMP increases M1 phenotype CD86 NF-𝜅B is a critical transcription factor for proinflammatory expression and reduces M2 phenotype arginase-1 expres- cytokinessuchasTNF-𝛼, IL-6, and IL-1𝛽.AlthoughLPS sion in M2-polarized macrophages. These data indicate that treatment increased the translocation of p65 (NF-𝜅Bsubunit) EMP can repolarize the M2-polarized macrophages into M1 intothenucleusandthephosphorylationofI-𝜅B 𝛼 in macrophages. M1-polarized macrophages induced by EMP cytoplasm, EMP did not show any significant changes in NF- inhibited LLC cancer-cell migration and EL-4 lymphoma 𝜅B translocation or I-𝜅B 𝛼 phosphorylation (Figure 4(b)). It proliferation. Because STAT6 is a critical signal pathway has been reported that STAT3 suppression can convert TAM’s of IL-4 and IL-13 [17], we studied the phosphorylation of phenotype from M2 to M1 [15, 16]. We also explored whether STAT6 in M2 macrophages. Although the phosphorylation STAT3 was involved in the effect of EMP on macrophage of STAT6 is remarkably increased in M2 macrophages, LPS polarization. The phosphorylation of STAT3 induced by IL- and EMP did not affect the phosphorylation of STAT6 in M2 4 and IL-13 was diminished by EMP treatment (Figure 4(a)). macrophages. In addition, we also explored the translocation of NF-𝜅B, a key transcription factor for inflammation, into 4. Discussion the nucleus by EMP.Although the phosphorylation of I-𝜅Bin cytoplasm and the translocation of p65 into the nucleus were EMP enhances M1 cytokine (TNF-𝛼) release and inhibits increased by LPS, EMP did not affect the phosphorylation M2 cytokine (TGF-𝛽)andM2markerswsuchasYM1, and translocation. Inhibition of STAT3 phosphorylation in Evidence-Based Complementary and Alternative Medicine 5

M1 M2

15 TNF- ∗ 1.5 Fizz1 1.5 MMR

10 1.0 1.0

5 0.5 0.5 mRNA folds mRNA mRNA folds mRNA mRNA folds mRNA ∗∗∗ ∗∗∗ ∗∗∗ 0 0.0 0.0 B20 10 550B2010 550B2010 550

IL-4 + IL-13 + EMP (g/ml) IL-4 + IL-13 + EMP (g/ml) IL-4 + IL-13 + EMP (g/ml)

iNOS Ym1 Arginase-1

5 1.5 1.5 ∗ 4 1.0 1.0 3

2 ∗∗∗ 0.5 0.5 mRNA folds mRNA mRNA folds mRNA 1 folds mRNA ∗∗∗ ∗∗∗ ∗∗ ∗∗ 0 0.0 0.0 B2010 550B2010 550B 01025 50 + 13 +  IL-4 IL- EMP ( g/ml) IL-4 + IL-13 + EMP (g/ml) IL-4 + IL-13 + EMP (g/ml) (a) CON LPS 10 103 103 103

102 102 102

101 101 101

100 100 100

100 101 102 103 100 101 102 103 100 101 102 103

25 50 103 103 80 ∗∗ 102 102 ∗ ∗∗ 60

1 1 10 10 40 CD68

0 0

10 10 CD86+ cells (%) 20 CD68 0 CON LPS 10 25 50 100 101 102 103 100 101 102 103 IL-4 + IL-13 + EMP (g/ml) CD86 (b)

Figure 2: Continued. 6 Evidence-Based Complementary and Alternative Medicine

1.5 EMP (g/ml)

IL-4 + IL-13 −+ + + + + 1.0 −−LPS 10 25 50

Arginase-1 0.5 ∗∗ ∗∗ Arginase-1/actin

-Actin ND 0.0 B2CON LPS 10 550 IL-4 + IL-13 + EMP (g/ml) (c)

Figure 2: Repolarization to M1 by EMP in M2 macrophages. (a) Effect of EMP on mRNA expression in M2 macrophages. BMM were pretreated with IL-4 (20 ng/mL) + IL-13 (20 ng/mL) for 6 h to polarize to M2 and then EMP was added for 18 h. The amounts of mRNA were quantified by real-time RT-PCR. The expression levels of mRNA were normalized by dividing the values by the GAPDH intensity. (b) CD86+and CD68+ BMM after EMP treatment were analyzed by flow cytometry. (C) Arginase-1 expression was determined by Western blotting. ND= ∗ ∗∗ ∗∗∗ not detected. Values are indicated as the mean ± SEM. 𝑝 < 0.05, 𝑝 < 0.01,and 𝑝 < 0.001 compared with the 0 (CON, EMP nontreated) samples.

1.5 B IL-4 + IL-13 IL-4 + IL-13 + LPS

1.0 ∗ ∗ ∗ IL-4 + IL-13 IL-4 + IL-13 IL-4 + IL-13 + EMP (10) + EMP (25) + EMP (50) 0.5 Relative migrated cells migrated Relative

0.0 B CON LPS 10 25 50 IL-4+IL-13 + EMP (g/ml) (a)

1.5

∗∗ ∗ 1.0 ∗∗

0.5 Luminescence folds Luminescence

0.0 BCONLPS102550 IL-4+IL-13 + EMP (g/ml) (b)

Figure 3: EMP-treatedmacrophagesreducemigrationandproliferationoftumorcells. (a) A transwell migration assay was performed to determine the migration of LLC tumor cells by EMP-treated macrophages. LLC tumor cells on the lower surface of the transwell membrane were stained with crystal violet solution and observed under a phase contrast microscope with 50x magnification. (b) EL4-luc2 lymphoma ∗ ∗∗ was cocultured with EMP-treated macrophages for 48 h. Values are indicated as the mean ± SEM. 𝑝 < 0.05 and 𝑝 < 0.01 compared with the CON samples. Evidence-Based Complementary and Alternative Medicine 7

1.5 2.5 EMP (g/ml) 2.0 + − +++++ IL-4 IL-13 1.0 −−LPS 10 25 50 p-STAT6 1.5 1.0 p-STAT3 0.5 ∗ p-STAT3/actin p-STAT6/actin 0.5  ND -Actin 0.0 0.0 B CON LPS10 25 50 BCONLPS10 25 50 IL-4 + IL-13+ EMP (g/ml) IL-4 + IL-13+ EMP (g/ml) (a)

EMP (g/ml) IL-4 + IL-13 − +++++ 6 − − LPS 10 25 50 2.0 p-IB C 4 1.5 /actin -Actin C B  1.0 2

P65 p-I  N P65/PCNA 0.5 PCNA 0 N BCONLPS10 25 50 0.0 BCONLPS10 25 50 IL-4 + IL-13 + EMP (g/ml) IL-4 + IL-13 + EMP (g/ml) (b) Figure 4: Inhibition of STAT3 phosphorylation by EMP in M2 macrophages. (a) BMM were pretreated with IL-4 (20 ng/mL) + IL-13 (20 ng/mL) for 6 h to polarize to M2 and then EMP was added for 1 h. The phosphorylation of STAT3 and STAT6 was analyzed by immunoblot analysis. ND = not detected. (b) RAW264.7 cells were pretreated with IL-4 (20 ng/mL) + IL-13 (20 ng/mL) for 6 h to polarize to M2 and then EMP was added for 1 h. p-I𝜅B𝛼 and 𝛽-actin were analyzed in the cytosolic fraction and P65 and PCNA were analyzed in the nuclear fraction. C = ∗ cytosol; N = nucleus. Values are indicated as the mean ± SEM. 𝑝 < 0.05 compared with the CON samples.

ashiftfromM2toM1polarizationhasbeenreportedto is supposed that the anticancer effects of MP in animal are suppress the growth and metastasis of tumor cells [15, 16]. mediated by M1 polarization and not just by the apoptosis of In the present study, we show that EMP inhibits STAT3 tumor cells. phosphorylation in M2 macrophages. These findings suggest that treatment with EMP polarizes When we screened the TNF-𝛼 secretion in M2 macro- M2/TAM into M1. Because of these effects of EMP, it may be phages using 400 species of herb, Mylabris phalerata,Genkwa used as an adjuvant for anticancer drugs to boost anticancer Flos, Solani Nigri Herba, Pinelliae Tuber, Sambuci Lignum, immunotherapy. Sanguisorbae Radix, Euphorbiae Kansui Radix, Phaseoli Radiati Semen, Poria, and Melandrii Herba were the most Conflicts of Interest potent top 10 herbs without endotoxin. Although there are some reports on anti-inflammatory effects of these 10 herbs, The authors declare that they have no conflicts of interest macrophage polarization by these herbs has not been studied. related to this study. It has been reported that norcantharidin, a biosynthesized demethylated cantharidin, has anticancer effects by the regu- Acknowledgments lation of M1 macrophage polarization via miR-214 expression [16]. Because norcantharidin is a synthetic compound and is This work has been supported by Ministry of Science, ICT notacomponentofMP,theeffectsofEMPonM1polarization and Future Planning (MSIP), Republic of Korea, Grant may not be mediated by norcantharidin. K17281 awarded to Korea Institute of Oriental Medicine There is a lot of evidence showing that M2-polarized ma- (KIOM). crophages can be converted to M1 macrophages and the con- References verted M1 macrophages exert anticancer and antimetastatic properties [15, 18–20]. Although we did not perform an in [1] C. E. Lewis and J. W. Pollard, “Distinct role of macrophages in vivo study, there are many reports on the anticancer effects different tumor microenvironments,” Cancer Research,vol.66, of MP [8, 21, 22]. Because MP per se has anticancer effects, no. 2, pp. 605–612, 2006. it may not be easy to differentiate its anticancer effects by [2]A.H.Klimp,E.G.E.DeVries,G.L.Scherphof,andT.Daemen, tumor killing from those by M1 polarization. Conversely, it “A potential role of macrophage activation in the treatment of 8 Evidence-Based Complementary and Alternative Medicine

cancer,” Critical Reviews in Oncology/Hematology,vol.44,no.2, [19]Y.Li,W.Qi,X.Song,S.Lv,H.Zhang,andQ.Yang,“Huaier pp. 143–161, 2002. extract suppresses breast cancer via regulating tumor-Associ- [3]A.Mantovani,S.Sozzani,M.Locati,P.Allavena,andA.Sica, ated macrophages,” Scientific Reports,vol.6,ArticleID20049, “Macrophage polarization: tumor-associated macrophages as a 2016. paradigm for polarized M2 mononuclear phagocytes,” Trends in [20]S.Chatterjee,A.Mookerjee,J.M.Basuetal.,“Anovelcopper Immunology,vol.23,no.11,pp.549–555,2002. chelate modulates tumor associated macrophages to promote [4] J. W.Pollard, “Tumour-educated macrophages promote tumour anti-tumor response of T cells,” PLoS ONE,vol.4,no.9,Article progression and metastasis,” Nature Reviews Cancer,vol.4,no. ID e7048, 2009. 1,pp.71–78,2004. [21] T.-C. Hsia, C.-C. Yu, Y.-T. Hsiao et al., “Cantharidin impairs cell [5] C. Guiducci, A. P. Vicari, S. Sangaletti, G. Trinchieri, and M. migration and invasion of human lung cancer NCI-H460 cells P. Colombo, “Redirecting in vivo elicited tumor infiltrating via UPA and MAPK signaling pathways,” Anticancer Research, macrophages and dendritic cells towards tumor rejection,” vol. 36, no. 11, pp. 5989–5997, 2016. Cancer Research,vol.65,no.8,pp.3437–3446,2005. [22] D. Liu and Z. Chen, “The effects of cantharidin and cantharidin [6] A. Olsson, J. Nakhle,´ A. Sundstedt et al., “Tasquinimod trig- derivates on tumour cells,” Anti-Cancer Agents in Medicinal gers an early change in the polarization of tumor associated Chemistry,vol.9,no.4,pp.392–396,2009. macrophages in the tumor microenvironment,” Journal for Im- munoTherapy of Cancer,vol.3,no.1,article53,2015. [7] S.K.Jeong,K.Yang,Y.S.Parketal.,“Interferongammainduced by resveratrol analog, HS-1793, reverses the properties of tumor associated macrophages,” International Immunopharmacology, vol. 22, no. 2, pp. 303–310, 2014. [8] J.-E. Huh, K.-S. Kang, K.-S. Ahn, D.-H. Kim, I. Saiki, and S.-H. Kim, “Mylabris phalerlata induces apoptosis by caspase activation following cytochrome c release and Bid cleavage,” Life Sciences,vol.73,no.17,pp.2249–2262,2003. [9] G.-S. Wang, “Medical uses of mylabris in ancient China and recent studies,” Journal of Ethnopharmacology,vol.26,no.2,pp. 147–162, 1989. [10] C.-C. Wang, C.-H. Wu, K.-J. Hsieh, K.-Y. Yen, and L.-L. Yang, “Cytotoxic effects of cantharidin on the growth of normal and carcinoma cells,” Toxicology,vol.147,no.2,pp.77–87,2000. [11] Y.-N. Chen, J.-C. Chen, S.-C. Yin et al., “Effector mechanisms of norcantharidin-induced mitotic arrest and apoptosis in human hepatoma cells,” International Journal of Cancer,vol.100,no.2, pp.158–165,2002. [12] A. Kim, M. Im, N.-H. Yim, Y. P. Jung, and J. Y. Ma, “Aqueous Extract of Bambusae Caulis in Taeniam Inhibits PMA-Induced Tumor Cell Invasion and Pulmonary Metastasis: Suppression of NF-𝜅B Activation through ROS Signaling,” PLoS ONE,vol. 8, no. 10, Article ID e78061, 2013. [13] M. Munder, H. Schneider, C. Luckner et al., “Suppression of T- cell functions by human granulocyte arginase,” Blood,vol.108, no. 5, pp. 1627–1634, 2006. [14] M. Munder, “Arginase: an emerging key player in the mammalian immune system,” British Journal of Pharmacology,vol. 158, no. 3, pp. 638–651, 2009. [15] X. Zhang, W. Tian, X. Cai et al., “Hydrazinocurcumin Encap- suled Nanoparticles “Re-Educate” Tumor-Associated Macro- phages and Exhibit Anti-Tumor Effects on Breast Cancer Fol- lowing STAT3 Suppression,” PLoS ONE,vol.8,no.6,ArticleID e65896, 2013. [16] S. Lu, Y. Gao, X. Huang, and X. Wang, “Cantharidin exerts anti- Hepatocellular carcinoma by mir-214 modulating macrophage polarization,” International Journal of Biological Sciences,vol.10, no. 4, pp. 415–425, 2014. [17] S. Goenka and M. H. Kaplan, “Transcriptional regulation by STAT6,” Immunologic Research,vol.50,no.1,pp.87–96,2011. [18] M. Liu, F. Luo, C. Ding et al., “Dectin-1 activation by a natural product 𝛽-glucan converts immunosuppressive macrophages into an M1-like phenotype,” JournalofImmunology,vol.195,no. 10, pp. 5055–5065, 2015. Hindawi Evidence-Based Complementary and Alternative Medicine Volume 2017, Article ID 2147408, 10 pages https://doi.org/10.1155/2017/2147408

Research Article Long-Term Course to Lumbar Disc Resorption Patients and Predictive Factors Associated with Disc Resorption

Jinho Lee, Joowon Kim, Joon-Shik Shin, Yoon Jae Lee, Me-riong Kim, Seon-Yeong Jeong, Young-jun Choi, Tae Kyung Yoon, Byung-heon Moon, Su-bin Yoo, Jungsoo Hong, and In-Hyuk Ha Jaseng Spine and Joint Research Institute, Jaseng Medical Foundation, Seoul, Republic of Korea

Correspondence should be addressed to In-Hyuk Ha; [email protected]

Received 18 April 2017; Accepted 5 June 2017; Published 9 July 2017

Academic Editor: Gihyun Lee

Copyright © 2017 Jinho Lee et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The long-term course to lumbar intervertebral disc herniation (LDH) patients receiving integrative Korean medicine treatment and predictive factors associated with disc resorption were investigated. LDH patients who received integrative Korean medicine treatment from February 2012 to December 2015 and were registered in the “longitudinal project for LDH on MRI” were included. Disc resorption amount was measured 3-dimensionally with disc degeneration and modic change levels on baseline and follow-up MRIs. Patient characteristics, Korean medicine use, pain, symptom recurrence, and satisfaction were assessed through medical records and phone surveys. Of 505 participants, 19 did not show disc resorption, while 486 did. A total of 220 displayed resorption 3 3 rates of ≥50%. LDH volume at baseline was 1399.82±594.96 mm , and that on follow-up MRI was 734.37±303.33 mm , indicating a 47.5% decrease (𝑝 < 0.0001). Predictive factors for disc resorption were disc extrusion, Komori migration classification, and LDH amount. Approximately 68.4% did not experience symptom recurrence over the 51.86 ± 19.07-month follow-up, and 90.3% were satisfied with Korean medicine treatment. The majority of LDH patients who improved after integrative Korean medicine treatment showed disc resorption within 1 year with favorable long-term outcomes. Predictive factors for disc resorption should be duly considered for informed decision-making. This trial is registered with ClinicalTrials.gov NCT02841163.

1. Introduction work and disability of all medical conditions [2], and LBP from spinal diseases such as spondylosis and intervertebral Recent developments in the modernization of traditional disc disorders were the 2nd highest ranking reason for Oriental medicine include greater integration with medical hospitalized care in the working population (age 18–44) in the technology and devices toward greater efficacy, safety, and US in 2005 [3]. Approximately 317,000 lumbar surgeries were diagnostic and prognostic accuracy. A dual medical system performed in the US in 1997 [4], which has steadily risen to 1 of conventional medicine and traditional Korean medicine million spinal procedures in 2002 [5]. is employed in Korea, and conventional diagnostic imaging However, the natural history of lumbar IDH is favorable, such as X-rays and magnetic resonance imaging (MRI) and and 70% of patients recover from sciatica without surgery advances in technology may be used to further promote the within 6 weeks [6]. The 10-year observation results of surgical modernization of traditional Oriental medicine by means of and nonsurgical treatment show no significant difference an integrative and collaborative approach. after 4 years [7], and a recent large-scale study on acute Prevalence estimates for sciatica have been reported to IDH reported that while early surgical intervention may range from 1.6% in the general population to 43% in select be more beneficial economically through swift recovery working populations, and the most commonly cited cause and return to work, 1-year pain and function results did of sciatica is lumbar intervertebral disc herniation (IDH) not differ significantly from nonsurgical treatment [8]. An [1]. Low back pain (LBP) and sciatica incur most time off estimated 10% of lumbar IDH patients consider surgery due 2 Evidence-Based Complementary and Alternative Medicine to continuous pain and/or neurological deficit, and accurate recurrence of symptoms, or underwent lumbar surgery, prediction of development into chronic pain is therefore of awareness of spontaneous disc resorption, and satisfaction importance [6]. Advances in radiological examination have with Korean medicine treatment. revealed how IDH may be spontaneously resorbed in time with resolution of neurological symptoms [9], and various 2.3. Analysis of Imaging Results. This study used three- factors predictive of disc resorption are being studied as they dimensional measurements on MRI to determine IDH hold valuable information on chronicization and need for amount according to previous measurement methods [10]. surgical intervention [10–16]. The authors selected a main IDH level on sagittal sections In Korea, many IDH patients opt for primary care at of T2-weighted MRIs best correlating with symptoms. A Korean medicine institutions to receive Korean medicine line connecting 2 points indicating the posterior edges of treatment and avoid surgery. The authors have previously the superior and inferior endplates was drawn as reference. reported the long-term outcomes and risk factors for poor The IDH area and amount outside the reference line were prognosis in lumbar IDH patients receiving integrative demarcated and measured using the “measure area freehand” Korean medicine treatment, excluding conventional treat- function in the picture archiving and communication system ment [17, 18]. The objectives of this study were to observe (PACS) on sagittal sections. The volume was calculated by the long-term course to IDH patients with spontaneous disc multiplying the IDH area on each sagittal section with the resorption after integrative Korean medicine treatment and MRIscanthicknessplustheintersliceintervals. identify predictive factors associated with resorption amount. The IDH level was classified into bulging, protrusion, extrusion, migration, and sequestration and by the migration 2. Materials and Methods criteria suggested by Komori et al. [20] where migrating disc amountiscategorizedincomparisontotheposteriorheight Jaseng Hospital of Korean Medicine, certified by the Korean of the adjacent vertebrae. IDH < 1/3oftheposteriorheight Ministry of Health and Welfare as a spine-specialty Korean of the vertebrae immediately superior or inferior to the IDH medicine hospital, operates 19 branches which treat 900,000 level was designated as Grade 2, that between 1/3 and 2/3 spinal disorder patients per year [19]. Electronic medical height as Grade 3, >2/3heightasGrade4,andnoIDHas records (EMRs) and radiological assessments of participants Grade 1. The IDH degeneration categorization employed the who were diagnosed with lumbar IDH on MRI, received classification criteria put forth by Pfirrmann et al., which integrative Korean medicine treatment at these medical classifies degeneration by structure, signal intensity, and disc institutions from February 2012 to December 2015, and were height on MRI into 5 levels from Grades I to V.Higher grades included in the “longitudinal project for IDH on MRI” were imply greater degeneration, and generally Grades IV and V assessed. The aim of the project was to assess differences in are taken to indicate disc degeneration [21]. Modic types were diagnostic imaging before and after treatment by allowing classifiedinto0(nomodicchange),andmodictypes1,2, physicians to select 1-2 patients annually for complimentary and 3. The location of modic type change in the vertebral follow-up MRIs with written consent for use of data for aca- body was recorded as (i) above, (ii) below, or (iii) both above demic means. Survey assessments were conducted by phone and below the IDH level, or (iv) no modic type change. from April to October 2016. This study received approval Two Korean medicine doctors (KMDs) experienced in spinal from the Institutional Review Board of Jaseng Hospital of imaging received training for standardized reading for higher Korean Medicine (JASENG 2016-06-003). reliability. Interrater reliability was assessed through measurement and classification of 20 random pairs of imaging 2.1. Participants. The inclusion criteria were as follows: (1) results (total of 40 MRIs before and after treatment) (see patients diagnosed with lumbar IDH on MRI, (2) patients Supplementary Table 1 in Supplementary Material available who experienced improvement of LBP and/or sciatica from online at https://doi.org/10.1155/2017/2147408). integrative Korean medicine treatment, and (3) written consent to participate in the “longitudinal project for IDH on 2.4. Statistical Analysis. All continuous variables are ex- MRI” and a follow-up MRI. The exclusion criteria were (1) pressed as mean ± SD, and categorical variables as 𝑛 (%). patients who received lumbar spinal surgery during inte- Consistency of IDH imaging measurement was determined grative Korean medicine treatment, (2) patients with main through intraclass correlation coefficient (ICC), and that of complaint of cervical IDH, and (3) patients with MRI errors IDH and degeneration classification was analyzed by Cohen’s or low resolution. kappa. Predictive factors for ≥50% IDH regression on follow- up MRI were determined through logistic regression analysis 2.2. EMR Assessment and Phone Survey. EMRs were reviewed considering for major influence variables (age, gender, level to assess participant characteristics including sex, age, LBP of disc degeneration, IDH type, Komori migration classifica- and/or sciatica, and details of integrative Korean medicine tion, modic type, modic change area, time interval between treatment (whether the patient underwent hospitalization, baseline and follow-up MRIs, and total treatment duration). days in inpatient and ambulatory care, and frequency of treat- Univariate analysis was performed, and factors that were ment). Phone surveys were also conducted to investigate significant in univariate analysis were included in additional initial and current levels of pain associated with lumbar IDH multivariate analyses with age and baseline IDH amount, as assessed by the numeric rating scale (NRS), whether the adjusted for. Comparison of IDH amounts before and after patient received recommendation for surgery, experienced treatment was analyzed using paired 𝑡-test and IDH type Evidence-Based Complementary and Alternative Medicine 3

Patients registered for the “longitudinal project for intervertebral disc herniation on MRI” by physicians as eligible candidates for disc resorption (n = 660) Excluded (n = 150) (i) Chief complaint related to cervical intervertebral disc herniation (n=53) (ii) MR imaging errors or low resolution (n=97) Candidates for lumbar intervertebral disc resorption (n = 510) Excluded (n=5) (i) EMR errors (n=5) Participants analyzed by MRI and EMR (n = 505) Did not complete follow-up (n = 195) (i) Unavailable for phone interview (n = 195) Survey participants analyzed by phone interview (n = 310)

Figure 1: Flow diagram of the study. and level of degeneration were analyzed with chi-square test. volume decrease of 47.5% (𝑝 < 0.0001). In Komori migration Associations between IDH volume on baseline and follow- classification, 38.6% of patients were classified as Grade 2 or3 upMRIs,changeinvolume,resorptionrate,andagewere at baseline, which decreased to 3.4% at follow-up. Regarding assessed through regression analysis. All analyses comparing IDH type, the percentage of patients with extrusion or greater measurements before and after treatment were performed herniation level (i.e., extrusion, migration, or sequestration) using STATA 14.0 (StataCorp, College Station, Texas, USA). at baseline was 87.9%, which declined to 22.8% at follow- up. Disc degeneration and modic type grade did not show 3. Results significant difference (Table 2). Of the 505 participants, 186 (38.6%) were hospitalized A total of 660 patients were registered in the “longitudinal while the others received integrative Korean medicine treat- project for IDH on MRI” and underwent baseline and follow- ment in the outpatient department. The number of days in up MRIs. Of these eligible participants, patients with chief inpatient care was 34.34 ± 29.53 days, that of outpatient visits complaint diagnosis of cervical IDH, imaging or EMR errors, was 32.88 ± 20.72 days, and the total number of days in care and low imaging resolution were excluded, and 505 lumbar was 45.53 ± 28.61 days. The vast majority of patients were IDH patients participated. Phone surveys were performed in given herbal medicine (96.4%), acupuncture (96.4%), and all participants except those that the researchers were unable Chuna manipulation (87.9%), followed by bee venom phar- to reach, yielding a total of 310 survey participants (61%) macopuncture (65.0%), pharmacopuncture (53.3%), electro- (Figure 1). acupuncture (46.5%), and cupping (22.6%). Of herbal medi- Baseline demographic characteristics of the 505 partici- cine and pharmacopuncture, Chungpa-jun and variations, of pants were as follows: Average age was 39.08 ± 10.19 years, which GCSB-5 is the main ingredient [22–25], and Shinbaro with higher male percentage (60.6%), and the most common pharmacopuncture [26] were used most frequently, respec- IDH level was L4/5 (53.3%) and L5/S1 (38.8%). The majority tively (Table 3). did not have LBP (90.1%), and most presented with sciatica The mean time interval between baseline and follow-up (83.6%) which was mainly unilateral (76.0%) (Table 1). phone interviews in the 303 survey participants was 51.86 ± The average time interval between baseline and follow- 19.07 months. Baseline pain NRS was reported at 8.34 ± 1.42, up MRI was 341.38 ± 306.83 days, and the difference which was put at 1.27±1.57 for current pain. Lumbar surgery between MRIs was analyzed. Evidence of disc resorption had previously been recommended for around half of all was not seen in 19 patients and seen in 486 of whom 220 patients at baseline (54.2%), and 73.2% were not aware of exhibited a resorption rate of ≥50%. The IDH volume at spontaneous disc resorption. Ninety-four patients reported 3 baseline was 1399.82 ± 594.96 mm , and that at follow-up pain recurrence(s) during the follow-up period (30.3%), 3 was 734.37 ± 303.33 mm , indicating a statistically significant andthemajoritychoseKoreanmedicineforitstreatment. 4 Evidence-Based Complementary and Alternative Medicine

Most patients were satisfied with Korean medicine (90.1%) Table 1: Baseline demographic characteristics of participants of the and nonsurgical methods for lumbar IDH treatment were “longitudinal project for IDH on MRI”. recommended for them (94.2%) (Table 4). 𝑛 = 505 Analysis of predictive factors for ≥50% resorption (dicho- Variables 𝑛 tomous variable) in IDH resorption amount revealed that %Mean(SD) IDH type and amount (Komori migration classification) were Age (years) 39.08 ± 10.19 significantly associated. When Komori migration classifica- <25 23 4.6 tion Grade 1 was set as reference, the odds ratios (ORs) for ≥25–<55 436 86.3 Grades 2 and 3 were 1.7 (95% CI 1.04–2.76) and 2.46 (95% ≥55 46 9.1 CI 1.37–4.42), respectively. When protrusion was designated Sex as the reference in IDH type, the ORs for extrusion and Male 306 60.6 migrationwere2.49(95%CI1.10–5.60)and6.3(95%CI Female 199 39.4 2.58–15.42), respectively (Table 5). Baseline IDH amount Herniated disc level was shown to be positively related to disc resorption rate L1/2 2 0.4 (continuous variable), whereas age was negatively associated (Figure 2). L2/3 9 1.8 L3/4 29 5.7 4. Discussion L4/5 269 53.3 L5/S1 196 38.8 Of 505 participants who were mainly middle-aged males Low back pain presenting with unilateral sciatica symptoms due to IDH No 454 90.1 at L4/5 and L5/S1 levels, 486 displayed spontaneous disc Yes 4 8 9.5 resorption (96.2%), and 220 showed resorption rates of Radiating leg pain ≥ 50% (43.6%). In Komori migration classification, 38.6% of No 80 15.8 patients were classified as Grades 2 or 3 at baseline which Yes, radiating to thigh 127 25.2 decreased to 3.4% at follow-up, and 87.9% were classified as disc extrusion or higher herniation level on baseline MRI, Yes, radiating to calf 239 47.3 compared to 22.8% at follow-up regarding IDH type. Major Yes, but with no indication of area 56 11.1 factors determined to predict disc resorption were baseline Unknown 3 0.6 IDH type, Komori migration classification, and age. Over Bilateral radiating leg pain a 51.86 ± 19.07-month follow-up period, 68.4% did not No 432 85.5 experience LBP recurrence, and 90.3% replied that they were Yes 6 6 13.1 satisfiedwithKoreanmedicinetreatment. Unknown 7 1.4 The underlying mechanism of disc resorption has been Unilateral radiating leg pain suggested to be enzymatic degradation and phagocytosis of No 115 22.8 IDH matter through inflammatory reaction and vasculariza- Yes 384 76.0 tion. As IDH material enters the vascularized epidural space, Unknown 6 1.2 it is identified by the body as a foreign substance, leading to immune and inflammatory response, and thus induces neo- IDH, intervertebral disc herniation; MRI, magnetic resonance imaging; SD, standard deviation. vascularization, enzymatic degradation, and macrophage phagocytosis. This consequently leads to production of matrix proteinases and increased cytokine levels, resulting in disc resorption [27, 28]. went on to publish that whereas the complete resolution The results of this study are consistent with other studies rate of migrated discs was 41%, that of nonmigrated discs wheresequestrateddiscmatterwasshowntobebetter was 0% [31]. In a study by Ahn et al., 25 out of 36 discs resorbedthanprotrudeddisc.Thisisspeculatedtobedue showed reduction in size, where 56% of subligamentous, 79% to the degree of penetration (tear) of the annulus fibrosus of transligamentous, and 100% of sequestered disc herniation and posterior longitudinal ligament and increased exposure resulted in IDH size reduction, and the study concluded that to systemic circulation within the epidural space [29, 30]. transligamentous extension and posterior longitudinal liga- Arecentsystematicreviewontheprobabilityofspon- ment rupture were the most influential factors in IDH reduc- taneous disc resorption by IDH type including 9 papers tion [29]. Moreover, in a 2014 observational study comparable covering 361 patients reported that 96% of disc sequestra- to the current study on 102 patients receiving traditional Chi- tion (resorption in 52/54 patients), 70% of disc extrusion nese medicine treatment, 81.37% reported symptom improve- (108/154), 41% of disc protrusion (38/93), and 13% of disc ment and 18.63% were considered to require surgery. IDH 3 bulging types (8/60) were resorbed, indicating high resorp- volume decreased from 1433.89 ± 525.49 mm to 1002.01 ± 3 tion rates in sequestration and extrusion types [12]. Similarly, 593.95 mm , resulting in an average resorption rate of 27.25± Komori et al. reported that although 78% of migrated discs 32.97%, and 20 patients presented resorption rates of ≥50% resulted in resorption (resorption in 36/64 patients), only [10]. While modic change was not recognized as a predictive 17% (7/48) were resorbed in nonmigrated discs [20], and factor of resorption in this study, a 2014 study by Shan et al. Evidence-Based Complementary and Alternative Medicine 5

Table 2: Baseline and follow-up MRI results.

Baseline MRI Follow-up MRI Variables 𝑝 value 𝑛 % 𝑛 % 3 Volume of herniated disc (mm )a 1399.82 ± 594.96 734.37 ± 303.33 0.000 Time interval between baseline and follow-up MRIs (days) 341.38 ± 306.83 Resorption rate (%) Aggravation 19 3.76 >0–≤25 75 14.85 >25–≤50 191 37.82 >50–≤75 205 40.59 >75–≤100 15 2.97 b Komori migration classification 1 310 61.4 488 96.6 210621.0163.20.000 3 89 17.6 1 0.2 b Disc herniation type Bulging 1 0.2 87 17.2 Protrusion 60 11.9 303 60.0 Extrusion 328 65.0 100 19.8 0.000 Migration 93 18.4 12 2.4 Sequestration 23 4.6 3 0.6 b Disc degeneration grade I 6 1.2 4 0.8 II 47 9.3 39 7.7 III 215 42.6 202 40.0 0.582 IV 224 44.4 243 48.1 V132.6173.4 b Modic change type of vertebral body 0 358 70.9 353 69.9 1163.2142.8 0.930 2 128 25.3 134 26.5 3 3 0.6 4 0.8 b Modic change of vertebral body area Above the herniated disc 22 4.4 20 4.0 Below the herniated disc 20 4.0 20 4.0 0.981 Both above and below the herniated disc 105 20.8 109 21.6 None 358 70.9 356 70.5 aPaired 푡-test was used in analysis of change in amount of herniated disc matter from baseline; bChi-square test was used in analysis of change in classification of herniated disc from baseline; MRI, magnetic resonance imaging. reported that whereas 35 of 85 patients in the modic change as it is difficult for small sample sized studies to secure group (of whom the majority were of Type 2) did not show sufficient statistical power. Furthermore, while predictive much difference in IDH size, the group with no modic change factor studies should consider various factors through pre- displayed significant decrease [32]. Iwabuchi et al. purported diction models using multivariate analysis, only univariate thatresorptionfactorscouldbeidentifiedonT1-andT2- analysismaybepossibleinstudieswithsmallsamplesize. weighted MRIs, with high signal intensity of IDH areas in T1- Another major strength of this study is that IDH was and T2-weighted MRIs in the nonregression group [33]. measured 3-dimensionally on MRI. Two-dimensional cross- The largest strength of this study is that, to the best sectional images may differ greatly from positioning, and of our knowledge, it covers the largest sample size (𝑛= various limitations and errors arise in measurement from 505) of studies addressing disc resorption. This is especially the sectional directions in image acquisition. This study also important in research on predictive factors for disc resorption includes long-term observation results at an average of 51.86 6 Evidence-Based Complementary and Alternative Medicine

Table 3: Use of integrative Korean medicine treatment. 𝑛 = 505 Variables 𝑛 %Days(mean± SD) Hospitalized care 186 36.8 Number of hospitalized days 34.34 ± 29.53 Number of outpatient visit days 32.88 ± 20.72 Total number of treatment days 45.53 ± 28.61 Frequency of integrative Korean medicine treatment Herbal medicine 487 96.4 136.35 ± 90.81 Pharmacopuncture 269 53.3 31.15 ± 28.62 Bee venom pharmacopuncture 328 65.0 26.78 ± 20.38 Chuna manipulation 444 87.9 39.06 ± 35.08 Acupuncture 487 96.4 45.42 ± 32.79 Electroacupuncture 235 46.5 32.80 ± 22.42 Cupping 114 22.6 20.00 ± 17.81 SD, standard deviation.

1 2500 ) 3 Ｇ 2000 .5 1500

1000 0 Resorption rate Resorption 500

−.5 MRI (m follow-up on Volume 0 0 1000 2000 3000 4000 5000 0 20 40 60 80 3 Volume on baseline MRI (mＧ ) Age (year) Fitted values Fitted values (a) (b) 4000 ) 3 3000 Ｇ

2000

1000

Change in volume (m in volume Change 0

0 20 40 60 80 Age (year) Fitted values (c)

Figure 2: Associations between herniated disc volume on baseline MRI, volume on follow-up MRI, change in volume, disc herniation resorption rate, and age. (a) Association between herniated disc volume on baseline MRI and resorption rate. 𝑦 = 0.19 + 0.0001703𝑥, 2 2 𝑅 = 0.2148, 𝑝 ≤ 0.001. (b) Association between herniated disc volume on follow-up MRI and age. 𝑦 = 944.39 − 5.37𝑥, 𝑅 = 0.0326, 2 𝑝 ≤ 0.001. (c) Association between change in herniated disc volume and age. 𝑦 = 744.62 − 2.03𝑥, 𝑅 = 0.0016, 𝑝 = 0.362. Evidence-Based Complementary and Alternative Medicine 7

Table 4: Outcome measures assessed through phone interview. 𝑛 = 310 Variables 𝑛 %Mean± SD Time interval between initial visit and phone interview (months) 51.86 ± 19.07 NRSofpainatbaselinea 8.34 ± 1.42 NRS of current pain 1.27 ± 1.57 Recommendation of surgerya Recommended for surgery 168 54.2 Not recommended for surgery 112 36.1 Do not know 6 1.9 Did not visit conventional medicine institution 24 7.7 Awareness of possibility of disc resorptiona Aware 83 26.8 Unaware 227 73.2 Experience of symptom recurrencea Yes 94 30.3 No 212 68.4 Do not know 4 1.3 Treatment type used for recurrencea Korean medicine treatment 104 36.1 Conventional nonsurgical treatment 14 4.9 Surgery 7 2.4 No recurrence requiring treatment 163 56.6 Satisfaction with Korean medicine treatment Very satisfied 159 51.5 Satisfied 120 38.8 Slightly satisfied 28 9.1 Unsatisfied 1 0.3 Very unsatisfied 1 0.3 Recommendation of treatment to others Surgical treatment 4 1.3 Nonsurgical treatment 292 94.2 Do not know 14 4.5 Effective Korean medicine treatment type Herbal medicine 64 20.7 Bee venom pharmacopuncture, pharmacopuncture 135 43.6 Acupuncture 26 8.4 Chuna manipulation 60 19.4 Do not know 61 19.7 aBaseline timepoint was assessed retrospectively at the time of follow-up survey; SD, standard deviation; NRS, numeric rating scale.

months (approximately 4.3 years) in addition to imaging as a result. The fact that MRI scans were performed using analyses to better portray current state, symptom recurrence, different imaging apparatus and under different conditions at and satisfaction with treatment. various in-hospital and external sites in this multicenter study Despite these strengths, this study also suffers from the mayalsobeviewedasalimitation. following weaknesses. The largest limitation is probably the WhilethenaturalcourseoflumbarIDHiswidelyconsid- fact that although the “longitudinal project for IDH on MRI” ered to be favorable, surgical interventions are still common itself was conducted prospectively, outcome measures were [4, 5]. International consensus recommends consideration partly investigated in a retrospective manner through EMR of surgery if symptoms persist after a certain period of examination and include limited clinical information. More- conservative treatment [34]; however, consensus has not been over, although follow-up MRIs were conducted allowing for reached regarding its duration [35, 36]. Although severe IDH sufficient time for potential disc resorption as clinical settings potentiallyincursneurologicaldisabilityandhigherpain and timeframes permitted, the time intervals are inconsistent levels, it also entails greater possibility of disc resorption. 8 Evidence-Based Complementary and Alternative Medicine

Table 5: Assessment of predictive factors at baseline associated with herniated disc resorption in participants.

Univariate Multivariatea OR 95% CI OR 95% CI Age (continuous) 1.01 (0.99, 1.03) Sex, male (ref. female) 0.97 (0.67, 1.37) Disc degeneration grade (ref. I) II 3.39 (0.37, 31.38) III 3.14 (0.36, 27.38) IV 4.57 (0.53, 39.77) V 16.67 (1.36, 204.03) Disc herniation type (ref. protrusion) Bulging — Extrusion 4.66 (2.15, 10.13) 2.49 (1.10, 5.60) Migration 11.82 (5.02, 27.85) 6.3 (2.58, 15.42) Sequestration 12.190 (3.91, 37.95) 3 (0.84, 10.68) Komori migration classification (ref. 1) 2 2.48 (1.58, 3.89) 1.7 (1.04, 2.76) 3 5.46 (3.24, 9.18) 2.46 (1.37, 4.42) Modic change type of vertebral body (ref. 0) 1 0.84 (0.30, 2.37) 2 4.4 (0.94, 2.10) 3 0.7 (0.06, 7.81) Modic change of vertebral body area (ref. none) Above the herniated disc 0.94 (0.37, 2.35) Below the herniated disc 1.410 (0.57, 3.46) Both above and below the herniated disc 1.38 (0.90, 2.12) Time interval between MRIs (continuous) 1 (0.99, 1.00) Total treatment duration (continuous) 1 (0.99, 1.00) aAdjusted for age and amount of disc herniation; OR, odds ratio; CI, confidence interval; MRI, magnetic resonance imaging.

A recent trial on recovery of motor deficit which is widely be factored into prognosis and informed decision-making in considered to require surgical attention reported that dif- treatment selection as most patients were unaware of the fact ferences between early surgery and prolonged conservative that disc resorption may occur spontaneously. treatment at 1 year were nonsignificant [37]. Early surgery in lumbar IDH patients should be approached carefully and Conflicts of Interest relevant information shared with patients in the decision- making process. Despite the fact that most participants were JinhoLeeandIn-HyukHaareemployeesofJasengHospital recommended for surgery in this study, not many were aware of Korean Medicine and Jaseng Medical Foundation. Joon- of spontaneous disc resorption. Shik Shin is the chairman of Jaseng Hospital of Korean Prospective large-scale studies further examining the disc Medicine.YoonJaeLeeisanemployeeofJasengHospitalof resorption prediction process are warranted. Risk scores such Korean Medicine. Joowon Kim, Me-riong Kim, Seon-Yeong as the Framingham Risk Score used to estimate 10-year car- Jeong, Young-jun Choi, Tae Kyung Yoon, Byung-heon Moon, diovascular risk may be similarly proposed in IDH through Su-bin Yoo, and Jungsoo Hong are employees of Jaseng development of a prediction model to provide patients, Medical Foundation. physicians, and healthcare givers with valuable data necessary for informed decision-making in selecting IDH treatment. Authors’ Contributions 5. Conclusion JinhoLeeandJoowonKimcontributedequallytothiswork as first authors. In conclusion, the majority of patients who received conservative integrative Korean medicine treatment indicated a References disc resorption volume of nearly 50% within a 1-year average, and the long-term course at 4.3 years was also favorable. [1] K. Konstantinou and K. M. Dunn, “Sciatica: review of epidemi- IDH amount and type were identified as predictive factors ological studies and prevalence estimates,” Spine,vol.33,no.22, associated with disc resorption and this information should pp. 2464–2472, 2008. Evidence-Based Complementary and Alternative Medicine 9

[2]M.W.vanTulder,B.W.Koes,andL.M.Bouter,“Acost-of- factors for surgery: a prospective observational 5-year follow- illness study of back pain in The Netherlands,” Pain,vol.62,no. up study,” Spine,vol.41,no.16,pp.E955–E963,2016. 2, pp. 233–240, 1995. [19] N. Robinson and J. Liu, “Oriental and traditional medi- [3]C.T.MerrillandA.Elixhauser,Hospitalization in the United cine—supporting the vision for integrated health,” European States, 2002, Agency for Healthcare Research and Quality, JournalofIntegrativeMedicine,vol.4,no.4,pp.e363–e365,2012. Rockville, Md, USA, 2005. [20] H. Komori, K. Shinomiya, O. Nakai, I. Yamaura, S. Takeda, and [4]B.S.Gillum,E.Graves,andE.Wood,“NationalHospitalDis- K. Furuya, “The natural history of herniated nucleus pulposus charge Survey,” Vital and Health Statistics,vol.13,no.133,pp. with radiculopathy,” Spine,vol.21,no.2,pp.225–229,1996. 1–51, 1998. [21] C. W. A. Pfirrmann, A. Metzdorf, M. Zanetti, J. Hodler, and [5] R. A. Deyo and S. K. Mirza, “Trends and variations in the use of N. Boos, “Magnetic resonance classification of lumbar inter- spine surgery,” Clinical Orthopaedics and Related Research,no. vertebral disc degeneration,” Spine,vol.26,no.17,pp.1873–1878, 443, pp. 139–146, 2006. 2001. [6] R. A. Deyo and J. N. Weinstein, “Low back pain,” The New [22] H.-J. Chung, H.-S. Lee, J.-S. Shin et al., “Modulation of acute England Journal of Medicine,vol.344,no.5,pp.363–370,2001. and chronic inflammatory processes by a traditional medicine [7]H.Weber,“Lumbardischerniation:acontrolled,prospective preparation GCSB-5 both in vitro and in vivo animal models,” study with ten years of observation,” Spine,vol.8,no.2,pp.131– Journal of Ethnopharmacology,vol.130,no.3,pp.450–459,2010. 140, 1983. [23] T.-H. Kim, S.-J. Yoon, W.-C. Lee et al., “Protective effect of [8] W.C.Peul,W.B.VanDenHout,R.Brand,R.T.W.M.Thomeer, GCSB-5, an herbal preparation, against peripheral nerve injury and B. W. Koes, “Prolonged conservative care versus early in rats,” Journal of Ethnopharmacology,vol.136,no.2,pp.297– surgery in patients with sciatica caused by lumbar disc hernia- 304, 2011. tion: two year results of a randomised controlled trial,” BMJ,vol. [24] J.-K. Kim, S.-W. Park, J.-W. Kang et al., “Effect of GCSB-5, 336, no. 7657, pp. 1355–1358, 2008. a herbal formulation, on monosodium iodoacetate-induced [9] S.Keskil,G.Ayberk,C.Evlıyaoglu,˘ T. Kizartici, E. Yucel,¨ and H. osteoarthritis in rats,” Evidence-based Complementary and Anbarci, “Spontaneous resolution of ‘protruded’ lumbar discs,” Alternative Medicine,vol.2012,ArticleID730907,11pages,2012. Minimally Invasive Neurosurgery,vol.47,no.4,pp.226–229, [25] Y.-G. Park, C.-W. Ha, C.-D. Han et al., “A prospective, random- 2004. ized, double-blind, multicenter comparative study on the safety [10] J.-Y. Seo, Y.-H. Roh, Y.-H. Kim, and K.-Y. Ha, “Three-dimen- and efficacy of Celecoxib and GCSB-5, dried extracts ofsix sional analysis of volumetric changes in herniated discs of the herbs, for the treatment of osteoarthritis of knee joint,” Journal lumbar spine: does spontaneous resorption of herniated discs of Ethnopharmacology,vol.149,no.3,pp.816–824,2013. always occur?” European Spine Journal,vol.25,no.5,pp.1393– [26]W.K.Kim,H.-J.Chung,Y.Pyeeetal.,“Effectsofintra-articular 1402, 2016. SHINBARO treatment on monosodium iodoacetate-induced [11] G. L. Cvetanovich, A. R. Hsu, R. M. Frank, H. S. An, and osteoarthritis in rats,” Chinese Medicine, vol. 11, article 17, 2016. G. B. Andersson, “Spontaneous resorption of a large cervical [27] M. Doita, T. Kanatani, T. Ozaki, N. Matsui, M. Kurosaka, and S. herniated nucleus pulposus,” American journal of orthopedics Yoshiya, “Influence of macrophage infiltration of herniated disc (Belle Mead, N.J.),vol.43,no.7,pp.E140–E145,2014. tissue on the production of matrix metalloproteinases leading [12] C.-C. Chiu, T.-Y. Chuang, K.-H. Chang, C.-H. Wu, P.-W. Lin, to disc resorption,” Spine,vol.26,no.14,pp.1522–1527,2001. and W.-Y. Hsu, “The probability of spontaneous regression of [28] T. Henmi, H. Endo, S. Nakano, and K. Sairyo, “Natural history lumbar herniated disc: a systematic review,” Clinical Rehabilita- of extruded lumbar intervertebral disc herniation,” The Journal tion,vol.29,no.2,pp.184–195,2015. of Medical Investigation,vol.49,no.1/2,pp.40–43,2002. [13]J.T.Oh,K.S.Park,S.S.Jungetal.,“Surgicalresultsand [29] S.-H. Ahn, M.-W. Ahn, and W.-M. Byun, “Effect of the risk factors for recurrence of lumbar disc herniation,” Korean transligamentous extension of lumbar disc herniations on their Journal of Spine,vol.9,no.3,pp.170–175,2012. regression and the clinical outcome of sciatica,” Spine,vol.25, [14]W.K.Erly,D.Munoz,andR.Beaton,“CanMRIsignalchar- no. 4, pp. 475–480, 2000. acteristics of lumbar disk herniations predict disk regression?” [30] J.-Y. Maigne, B. Rime, and B. Deligne, “Computed tomographic Journal of Computer Assisted Tomography,vol.30,no.3,pp. follow-up study of forty-eight cases of nonoperatively treated 486–489, 2006. lumbar intervertebral disc herniation,” Spine,vol.17,no.9,pp. [15] M. J. McGirt, S. Eustacchio, P.Varga et al., “Aprospective cohort 1071–1074, 1992. study of close interval computed tomography and magnetic [31] H. Komori, A. Okawa, H. Haro, T. Muneta, H. Yamamoto, and resonance imaging after primary lumbar discectomy: factors K. Shinomiya, “Contrast-enhanced magnetic resonance imag- associated with recurrent disc herniation and disc height loss,” ing in conservative management of lumbar disc herniation,” Spine,vol.34,no.19,pp.2044–2051,2009. Spine,vol.23,no.1,pp.67–73,1998. [16] C. Dora, M. R. Schmid, A. Elfering, M. Zanetti, J. Hodler, and N. [32] Z. Shan, S. Fan, Q. Xie et al., “Spontaneous resorption of lumbar Boos, “Lumbar disk herniation: do MR imaging findings predict disc herniation is less likely when modic changes are present,” recurrence after surgical diskectomy?” Radiology,vol.235,no.2, Spine,vol.39,no.9,pp.736–744,2014. pp. 562–567, 2005. [33] M. Iwabuchi, K. Murakami, F. Ara, K. Otani, and S.-I. Kikuchi, [17] J. J. Park, J. Shin, Y. Choi et al., “Integrative package for low “The predictive factors for the resorption of a lumbar disc back pain with leg pain in Korea: a prospective cohort study,” herniation on plain MRI,” Fukushima Journal of Medical Science, Complementary Therapies in Medicine,vol.18,no.2,pp.78–86, vol.56,no.2,pp.91–97,2010. 2010. [34] G. B. Andersson, M. D. Brown, J. Dvorak et al., “Consensus [18] J.-S. Shin, J. Lee, Y. J. Lee et al., “Long-term course of alternative summary on the diagnosis and treatment of lumbar disc and integrative therapy for lumbar disc herniation and risk herniation,” Spine,vol.21,no.24,pp.75S–78S,1996. 10 Evidence-Based Complementary and Alternative Medicine

[35] P. A. J. Luijsterburg, A. P. Verhagen, S. Braak, C. J. J. Avezaat, and B. W. Koes, “Do neurosurgeons subscribe to the guideline lumbosacral radicular syndrome?” Clinical Neurology and Neu- rosurgery,vol.106,no.4,pp.313–317,2004. [36] J.-P. Vader, F. Porchet, T. Larequi-Lauber, R. W. Dubois, and B. Burnand, “Appropriateness of surgery for sciatica: reliability of guidelines from expert panels,” Spine,vol.25,no.14,pp.1831– 1836, 2000. [37]G.M.Overdevest,C.L.A.M.Vleggeert-Lankamp,W.C.H. Jacobs,R.Brand,B.W.Koes,andW.C.Peul,“Recoveryofmotor deficit accompanying sciatica—subgroup analysis of a randomized controlled trial,” Spine Journal,vol.14,no.9,pp.1817– 1824, 2014.