<<

A Test for More Accurate Diagnosis of Pulmonary Lin Sun, PhD,a,b Xue Qi, MS,a,b Fang Liu, MS,b,c Xirong Wu, PhD,b,d Qingqin Yin, PhD,b,d Yan Guo, MS,b,d Baoping Xu, PhD,b,d Anxia Jiao, PhD,b,c Yajie Guo, MD,a,b Weiwei Jiao, PhD,a,b Chen Shen, PhD,a,b Jing Xiao, PhD,a,b Adong Shen, MSa,b

OBJECTIVES: Xpert Mycobacterium tuberculosis and rifampicin (MTB/RIF) Ultra assay has abstract increasingly been used in adult tuberculosis diagnosis, but data relating to its diagnostic accuracy in children are lacking. Because a qualified sputum specimen is difficult to obtain in children, this study evaluated the diagnostic value of Ultra in childhood tuberculosis using bronchoalveolar lavage fluid. METHODS: The accuracy of Ultra was calculated by using bacteriologic results and clinical evidence as reference standards. Concordance between Ultra and Xpert MTB/RIF assays was assessed by using к coefficients. RESULTS: In total, 93 children with pulmonary tuberculosis and 128 children with respiratory tract were enrolled. The sensitivity of Ultra, in all pulmonary tuberculosis cases and in bacteriologically confirmed tuberculosis cases, was 70% and 91%, respectively. Ultra could detect Mycobacterium tuberculosis in 58% of cases with negative culture or acid-fast–staining results. The specificity of Ultra was 98%. There was no significant difference in sensitivity between samples with a volume #1 and .1 mL (66% vs 73%; P = .50; odds ratio [OR] = 0.71). Among 164 children for which Ultra and Xpert were simultaneously performed, the sensitivity was 80% and 67%, respectively, indicating good agreement (к = 0.84). An additional 6 children were identified as Ultra-positive but Xpert-negative. The positive rate decreased from 93% to 63% after 1 month (P = .01; OR = 0.12) and to 71% after 2 months (P = .03; OR = 0.18) of antituberculosis treatment. CONCLUSIONS: Ultra using bronchoalveolar lavage fluid has good sensitivity compared with bacteriologic tests and adds clinical value by assisting the rapid and accurate diagnosis of pulmonary tuberculosis in children.

aBejing Key Laboratory of Pediatric Respiratory Infectious , Beijing Pediatric Research Institute, Beijing, WHAT’S KNOWN ON THIS SUBJECT: Early diagnosis of China; bNational Clinical Research Center for Respiratory , National Key Discipline of Pediatrics, Key childhood tuberculosis can reduce , but Laboratory of Major Diseases in Children, and National Center for Children’s Health, Ministry of Education of the it is often hindered by a lack of bacteriologic evidence. ’ ’ Peoples Republic of China, Beijing Childrens Hospital, Capital Medical University, Beijing, China; and We evaluated the diagnostic value of Xpert MTB/RIF cInterventional Pulmonary Department and dDepartment of Respiratory Medicine, Beijing Children’s Hospital, Beijing, China Ultra in childhood tuberculosis using bronchoalveolar lavage fluid (BALF). Dr Sun and Mr Shen conceptualized and designed the study, drafted the initial manuscript, and reviewed and revised the manuscript; Mr Qi and Dr Guo collected data and performed the tests; Ms WHAT THIS STUDY ADDS: Xpert MTB/RIF Ultra using Liu, Ms Guo, and Drs Wu, Yin, Xu, and Jiao enrolled the subjects and collected the samples; Drs Jiao, BALF offers higher sensitivity than bacteriologic tests. Shen, and Xiao analyzed the data; and all authors approved the final manuscript as submitted and In children with tuberculosis, Xpert MTB/RIF Ultra agree to be accountable for all aspects of the work. using BALF provides early and accurate etiologic DOI: https://doi.org/10.1542/peds.2019-0262 results. A comparison between Xpert MTB/RIF and Xpert MTB/RIF Ultra was performed in this study. Accepted for publication Aug 13, 2019

To cite: Sun L, Qi X, Liu F, et al. A Test for More Accurate Diagnosis of Pulmonary Tuberculosis. Pediatrics. 2019; 144(5):e20190262

Downloaded from www.aappublications.org/news by guest on September 27, 2021 PEDIATRICS Volume 144, number 5, November 2019:e20190262 ARTICLE Children carry nearly 10% of the a conditional test for children appears to be a good alternative global tuberculosis disease burden suspected of having tuberculosis.4,5 sample for children who have and account for 15% of the total Nowadays, the Xpert MTB/RIF Ultra difficulty producing sputum. The .1 The World Health assay has been developed as a next- sensitivity of MTB culture from BALF Organization (WHO) estimated that generation assay to overcome these is significantly higher than that from there were 99 000 new pediatric limitations. To improve sensitivity for sputum samples (39.2% vs 35.3%) in tuberculosis cases in China in 2017, the detection of MTB, another 2 adult patients with tuberculosis.12 but only 10% of them were notified multicopy amplification targets Positive rates obtained from smears and reported.1 Precise data on the (IS6110 and IS1081) and a larger DNA and MTB culture by using gastric percentage of pediatric tuberculosis reaction chamber are used in Ultra. aspirates or BALF are similar and are cases confirmed microbiologically are Furthermore, melting higher than those obtained by using not reported by many countries.1 The temperature–based analysis replaces sputum samples from children with large number of missing cases real-time polymerase chain reaction suspected complicated intrathoracic suggests poorer access to diagnosis to improve the accuracy of rifampicin tuberculosis.13 and treatment in childhood resistance detection.6 BALF shows a superior diagnostic tuberculosis than in adult Although Ultra has gradually been yield in patients with smear-negative tuberculosis. Timely and accurate introduced for diagnosing or sputum-scarce pulmonary diagnosis will be crucial to achieve tuberculosis in adults,7,8 data are still tuberculosis. However, there are little the global aim of minimizing lacking regarding its diagnostic data about its diagnostic value in childhood tuberculosis. accuracy in children. A multicenter pulmonary tuberculosis in children. However, management of childhood comparative study in children For this study, we enrolled children tuberculosis is hampered by demonstrated the higher sensitivity of with indications for fiber-optic difficulties in attaining an etiologic Ultra over Xpert using sputum bronchoscopy and had the following diagnosis because of the challenge of samples but slightly reduced aims: (1) evaluate the diagnostic obtaining samples and low bacterial specificity compared with Xpert in value of Ultra in childhood loads. In general, the diagnosis of detecting pulmonary tuberculosis.9 tuberculosis by using BALF, (2) pediatric tuberculosis in the clinic Most current research uses sputum compare the accuracy of Ultra with mainly depends on contact history, specimens or gastric aspirates to that of Xpert, and (3) confirm the clinical symptoms, and chest diagnose pulmonary tuberculosis in value of Ultra in monitoring radiography results.2 On the basis of children.10,11 However, qualified treatment effectiveness. The results their higher sensitivity, molecular sputum specimens are difficult to presented provide data for informed tests have increasingly been used to obtain in children, especially those of management of childhood confirm the presence of younger age. tuberculosis. Mycobacterium tuberculosis in Bronchoalveolar lavage is an effective patients with active tuberculosis3 as tool to control severe . Fiber- METHODS a promising alternative to traditional optic bronchoscopy is also an diagnostic methods. important intervention in children Study Population and Samples The Xpert Mycobacterium tuberculosis who present with dyspnea, This retrospective study was and rifampicin (MTB/RIF) assay is an obstructive pneumonia, pulmonary performed at the Beijing Children’s automated molecular test based on atelectasis, or localized emphysema. Hospital affiliated with Capital seminested real-time polymerase It is sometimes difficult to confirm Medical University. Children were chain reaction and molecular beacon whether pulmonary atelectasis is due defined as those ,15 years old technology targeting the rifampicin- to extrinsic compression on according to the WHO definition5 and resistance–determining region of the a bronchus from regional were enrolled in the study if (1) they rpoB gene.4 Rifampicin resistance is lymphadenopathy or from caseous had suspected symptoms of viewed as an indicator of multidrug- in an airway by using tuberculosis; (2) they had the resistant tuberculosis because most standard pulmonary imaging. In indications to undergo fiber-optic patients with rifampicin resistance children who present with these signs bronchoscopy, including those who also have isoniazid resistance. The or symptoms in addition to presented with dyspnea, obstructive WHO officially endorsed this assay as pulmonary tuberculosis, fiber-optic pneumonia, or localized emphysema, a primary diagnostic test for children bronchoscopy can be used to confirm and those needing further suspected of having multidrug- the reasons for pulmonary atelectasis microbiologic confirmation because resistant tuberculosis or HIV- and improve ventilation.7 As such, of atypical imaging changes, no associated tuberculosis and as bronchoalveolar lavage fluid (BALF) contact history with an index patient

Downloaded from www.aappublications.org/news by guest on September 27, 2021 2 SUN et al with confirmed tuberculosis, or described earlier.15 A 1:1 RESULTS negative microbiologic results by volume of sample reagent was using sputum or gastric aspirates; added to each BALF sample. If the Characteristics of the Study Participants and (3) informed consent to fiber- volume was ,1 mL, sample reagent optic bronchoscopy had been given. was added to give a total volume of Between February 2014 and BALF specimens were collected from 2 mL and mixed for 15 to 20 seconds April 2019, 580 children were February 2014 to April 2019 and before adding to the sample chamber enrolled for tuberculosis evaluation were stored at 280°C for further of the cartridge. Results were at our hospital, of whom 221 analysis. automatically read within 2 hours. children with suspected pulmonary Test specimen results included tuberculosis were enrolled in this The children enrolled were invalid, not detected (negative), study (Fig 1). Ninety-three of these categorized into 3 groups14: (1) detected (positive with children were diagnosed as having bacteriologically confirmed semiquantitation), and rifampicin- pulmonary tuberculosis, including 34 tuberculosis (positive for acid-fast resistant (detected, not detected, or (37%) with bacteriologically staining or culture of MTB), (2) indeterminate) samples. Results of confirmed tuberculosis (28 children clinically diagnosed tuberculosis (at analysis were reported as confirmed by MTB culture and 6 by least 1 symptom and sign, semiquantitative readouts on the acid-fast staining) and 59 (63%) radiographic evidence consistent with basis of the minimal cycle threshold . classified as clinically diagnosed tuberculosis, and at least 1 of the The semiquantitative scale for Xpert tuberculosis. Among the children following: positive tuberculin skin was set as very low, low, medium, or with pulmonary tuberculosis, 71 test or interferon-g release assay high. The semiquantitative scale for (76%) were ,5 years of age, and results, clinical and radiologic Ultra was set as trace, very low, low, 84 (90%) were ,10 years of age. Of improvement seen after medium, or high. the children, 128 were diagnosed antituberculosis chemotherapy, or with RTIs (Table 1). documented exposure to Demographic information and tuberculosis), and (3) patients clinical data were collected by 2 In total, 261 BALF specimens without tuberculosis with respiratory independent interviewers. Data from 221 children were collected. tract infections (RTIs) (symptomatic included age, sex, Bacillus For all 221 children, 1 BALF but not fitting the above definitions Calmette–Guérin scar, history of sample was collected within and confirmed etiologic evidence of contact with a patient with 1 week of admission, and for 29 infection with virus, mycoplasma, or tuberculosis, antituberculosis of the children, 1 or 2 more bacteria). treatment, and medical comorbidities, samples were collected during follow- such as HIV infection. up. This study was approved by the Medical Ethics Committee of Beijing Statistical Analysis Children’s Hospital. Written informed Evaluation of the Diagnostic consent was obtained from the The sensitivity, specificity, Accuracy of Ultra guardians of the patients. positive predictive value, and Diagnostic accuracy was negative predictive value were calculated when considering Procedures calculated by using bacteriologic the first BALF specimen (Table 2). results and clinical evidence as All children enrolled were tested with The sensitivity of Ultra was 70% reference standards. Concordance Ultra and simultaneously with Xpert (65 of 93), twice as high as that of between Ultra and Xpert was if the volume of the BALF sample was bacteriologic tests (37%; 34 of 93), assessed by using the percentage sufficient. If children with including both culture and acid-fast agreement and к coefficients, and endobronchial tuberculosis presented staining. The specificity of Ultra concordance of the semiquantitative with severe dyspnea due to caseous was high (98%; 125 of 128). scale of the 2 tests was presented by necrosis during treatment, fiber-optic using Spearman’s rank correlation Among 34 children with bronchoscopy was performed to coefficients. McNemar’s test was used bacteriologically confirmed remove the caseous necrosis in the to evaluate differences in sensitivity tuberculosis, the high airway once or twice during follow- and specificity between Ultra and sensitivity of Ultra was up. Xpert. P , .05 was considered demonstrated at 91% (31 of 34). Preprocessing of samples for Ultra statistically significant. SPSS 23.0 Among 59 children with clinically and Xpert (Cepheid, Sunnyvale, CA) software (IBM SPSS Statistics, diagnosed tuberculosis, the was performed in accordance with IBM Corporation) was used sensitivity of the test was 58% the manufacturer’s instructions, as for statistical analysis. (34 of 59).

Downloaded from www.aappublications.org/news by guest on September 27, 2021 PEDIATRICS Volume 144, number 5, November 2019 3 had Ultra-negative and Xpert-positive results with a very low load. The agreement between the 2 tests was moderate for the pulmonary tuberculosis group (k value 0.67) and higher for the entire group of 164 children (k value 0.84). Concordance of the semiquantitative scale for Ultra and Xpert was also FIGURE 1 analyzed in 46 children with Flowchart of the study population. pulmonary tuberculosis (Fig 2). Good agreement was observed, with a Spearman’s rank correlation coefficient of 0.81 (P , .001). The volume of the BALF Xpert. The sensitivity of obtained in each case varied from Ultra was higher than that of Usefulness of Ultra in Monitoring 0.5 to 2.5 mL. This permitted Xpert (80% vs 67%), and Ultra Antituberculosis Treatment comparison of the sensitivity of the permitted the detection of 6 out of assays regarding this variable. The 15 children with negative Xpert Two or 3 BALF samples were sensitivity of Ultra in 41 of the 93 results. The specificity of the 2 tests collected for 29 of the 93 children first BALF samples that had was 97% (115 of 118) and 99% (117 with pulmonary tuberculosis during a volume #1 mL was 66% (27 of 41), of 118), respectively (P = .61; OR = antituberculosis treatment. The fi whereas its sensitivity in the 52 0.33; 95% CI 0.03–3.20; Table 3). positive rate for the rst sample was samples that had a volume .1mL 93% (27 of 29). Sixteen samples were fi was 73% (38 of 52). There was no The agreement between Ultra collected 1 month after the rst significant difference between these 2 and Xpert was further analyzed sample, and the positive rate was groups (P = .50; odds ratio [OR] = for 46 children with pulmonary 63% (10 o f16). Twenty-four samples 0.71; 95% confidence interval [CI] tuberculosis. Concordant results were collected 2 months later, and the 0.29–1.73). were obtained for 40 children (31 positive rate was 71% (17 of 24). The with positive and 9 with negative positive rate decreased after P Comparison Between Ultra and Xpert results). Discordant results were antituberculosis treatment, with = .01 (OR = 0.12; 95% CI 0.02–0.72) In total, 164 children (46 found in 6 children with Ultra- when compared with 1 month later tuberculosis case patients and positive and Xpert-negative results. and P = .03 (OR = 0.18; 95% CI 118 case patients with RTIs) Among children with RTIs, 3 had 0.03–0.97) when compared with 2 were enrolled to be tested Ultra-positive and Xpert-negative months later. simultaneously with Ultra and results with a trace load and 1 child All 29 children showed improvement in symptoms. During 1- or 2-month follow-up appointments, TABLE 1 Main Clinical Characteristics of the Study Population 79% (23 of 29) of the children presented with alleviated tracheal Characteristics Children With Active Tuberculosis (N = 93) Children With RTIs (N = and bronchobronchial lesions, which fi 128) Total Clinically Con rmed were detected by fiber-optic Diagnosed bronchoscopy. Ultra results for 10 – – Age, mean (SD) 3.77 4.11 (1.4 5.4) 3.18 6.64 (3.9 9.1) (43%) of these 23 children (positive (0.7–5.4) (0.4–4.5) fi Sex, n (%) on the basis of the rst sample) Male 59 (63) 40 (68) 19 (56) 57 (45) converted to negative results during Female 34 (37) 19 (32) 15 (44) 71 (55) follow-up. For the other 8 children BCG scar, n (%) with persistent positive Ultra results, Yes 61 (66) 39 (66) 22 (65) 121 (95) the semiquantitative scale of MTB in No 32 (34) 20 (34) 12 (35) 7 (5) History of exposure, n the BALF samples decreased. There (%) were no differences in radiographic Yes 33 (35) 20 (34) 14 (41) 3 (2) or fiber-optic bronchoscopy findings No 60 (65) 39 (66) 20 (59) 125 (98) between those children whose follow- BCG, Bacillus Calmette–Guérin. up molecular test results remained

Downloaded from www.aappublications.org/news by guest on September 27, 2021 4 SUN et al TABLE 2 Accuracy of Xpert MTB/RIF Ultra in Children With Confirmed and Clinically Diagnosed Ultra was shown to have Tuberculosis higher sensitivity than acid-fast First BALF Sample Sensitivity, % (n of N) Specificity, % (n PPV, NPV, staining or MTB culture. Importantly, of 128) % % 58% of children with negative All children 70 (65 of 93) 98 (125) 96 82 etiologic results were Bacteriologically 91 (31 of 34) 98 (125) 91 98 detected by Ultra, which confirmed suggested that Ultra has higher Clinically diagnosed 58 (34 of 59) 98 (125) 92 83 diagnostic effectiveness in children. A Bacteriologically confirmed tuberculosis: positive for acid-fast staining or culture of MTB. NPV, negative predictive value; high sensitivity of 64.3% was PPV, positive predictive value. previously reported in children with pulmonary tuberculosis in an HIV- prevalent setting by using sputum,9 and 67.5% was reported in positive versus those who had test most studies on Ultra detection microbiologically confirmed result conversions to negative. have been conducted in adults, childhood tuberculosis.18 The positive with data lacking for children. One child was detected as rate was not affected by a limited sample volume. This improved being rifampicin resistant in both The first aim in our study was the first sample and the second sensitivity of Ultra might lead to not to evaluate the accuracy of only a larger proportion of confirmed sample collected 1 month later by Ultra for diagnosing pulmonary using both Ultra and Xpert. One child childhood tuberculosis cases, but also tuberculosis from BALF in children. rapid and accurate diagnosis. In detected as rifampicin sensitive in Various biological samples have the first sample converted to addition, both Ultra and Xpert had previously been evaluated with rifampicin resistant in the a shorter turnaround time than Ultra, including sputum, gastric second sample 2 months later conventional culture. These tests may aspirates, BALF, cerebrospinal when using Ultra while remaining prove valuable in cases in which only fluid, and feces.17 Three studies rifampicin sensitive in the second a small volume of sample can be have evaluated Ultra in pediatric sample when using Xpert. No collected. tuberculosis diagnosis to date, 2 of MTB was isolated from these 2 Our second aim was to which used sputum samples patients. compare Xpert with Ultra when the alone9,18 and 1 of which used 2 tests were run using the same sputum, gastric aspirates, and 17 samples. The sensitivity of Ultra DISCUSSION BALF. BALF samples were was higher than that of Xpert. Data from meta-analysis collected from children with the These results were consistent with fi reported that the positive indications required for ber-optic previous studies.9,15,20 A multicenter 19 fi rate of detection from sputum bronchoscopy. For example, ber- study in children reported smears in children is low, ranging optic bronchoscopy was performed sensitivities of 64.3% and 53.6% for from 0.5% in children aged 0 to if the chest radiograph results Ultra and Xpert, respectively.9 In 4 years to 14% in children aged 5 suggested the presence of obstructive adults with smear-negative sputum to 14 years.16 Xpert and Ultra are pneumonia, pulmonary atelectasis, results, the sensitivities of Ultra and rapid tests for the diagnosis of or localized emphysema or if Xpert were reported to be 63% and tuberculosis and rifampicin clinical manifestations (chronic 46%, respectively.15 The high resistance and are suitable for persistent cough, hemoptysis, or specificity of the 2 tests was also point-of-care testing, the latter dyspnea) existed but the etiology observed in our study, with good having higher sensitivity. To date, was unclear. concordance observed between the

TABLE 3 Comparison of Xpert MTB/RIF Ultra and Xpert MTB/RIF in Children With Confirmed and Clinically Diagnosed Tuberculosis First BALF Sample Sensitivity, % (n of N) OR (95% CI) P Specificity, % (n of 118) OR (95% CI) P Ultra Xpert Ultra Xpert All children 80 (37 of 46) 67 (31 of 46) 1.99 (0.77–5.12) .15 97 (115) 99 (117) 0.33 (0.03–3.20) .61 Bacteriologically confirmed 94 (17 of 18) 89 (16 of 18) 2.13 (0.18–25.78) .55 97 (115) 99 (117) 0.33 (0.03–3.20) .61 Clinically diagnosed 71 (20 of 28) 54 (15 of 28) 2.17 (0.72–6.55) .17 97 (115) 99 (117) 0.33 (0.03–3.20) .61 Bacteriologically confirmed tuberculosis: positive for acid-fast staining or culture of MTB.

Downloaded from www.aappublications.org/news by guest on September 27, 2021 PEDIATRICS Volume 144, number 5, November 2019 5 implementation of current and novel diagnostic tests for tuberculosis in children is an ongoing research focus. Because specimen processing, DNA amplification, and detection are performed under closed conditions in the Ultra test, it is possible to perform MTB detection outside of a laboratory equipped for advanced biosafety. This increases the applicability of the assay and decreases delays in the diagnosis of tuberculosis. However, the current cost of Ultra is a prohibitive factor in its implementation as a point-of-care test, especially given that higher tuberculosis burdens are commonly found in settings where resources for health care are already under great strain. The necessary equipment is also not readily available for primary medical institutions. To be financially viable in low-income countries, a large increase in tuberculosis funding and/or a further reduction in FIGURE 2 Correlation of the semiquantitative scale between Xpert MTB/RIF Ultra and Xpert MTB/RIF. The area the price of the Ultra test would be of the circles corresponds to the numbers of cases. needed. As an innovative operation, fiber-optic bronchoscopy requires tests. In children with RTIs, 3 with the duration of antituberculosis experienced physicians and Ultra-positive and 1 with Xpert- treatment in our study, which well-equipped operating rooms. positive results presented a very low suggested that the treatment was Only children with indications fi load of MTB, and on repeating the effective. However, even after should undergo ber-optic bronchoscopy. The difficulties in tests, the results remained positive. .2 months of treatment, the positive acquiring samples therefore limit the All 4 children were diagnosed rate was still 71%. Several studies use of BALF in diagnosing with Mycoplasma pneumoniae assessed the ability of Xpert to tuberculosis in children. based on detection of M monitor responses to tuberculosis – pneumoniae and positive responses treatment.22 24 Most of these studies It was highlighted by the WHO that to azithromycin treatment. It is concluded that Xpert cannot be used existing tests included in the possible that the positive result as an indicator of therapeutic efficacy diagnostic algorithms of active may have been caused by nonspecific because DNA fragments from lysed or tuberculosis can be tailored to be amplification. One study reported damaged bacteria could still be highly specific to each country’s no cross-reactivity of Ultra to 30 detected. To date, little is known settings and resources.4 Our data different isolates of nontuberculous about Ultra in this regard. As for the provide insight into the added value mycobacteria or 18 different low etiologic results in this study and of Ultra in the diagnosis of pulmonary bacteria.21 However, there are no the high percentage of first-time tuberculosis in children in China, data showing the impact of occurrence of tuberculosis in a country with a high burden of a background of M pneumoniae on children, future studies will be tuberculosis. the specificity of the assay. needed to determine the usefulness of Ultra in monitoring treatment CONCLUSIONS A third aim in our study was to assess effectiveness. the value of Ultra in monitoring The Ultra assay using BALF treatment effectiveness. The positive Identification of the most efficient offers good diagnostic value detection rate of Ultra decreased with and cost-effective strategies for the compared with acid-fast staining or

Downloaded from www.aappublications.org/news by guest on September 27, 2021 6 SUN et al MTB culture. Slightly higher monitoring of treatment ABBREVIATIONS sensitivity was detected for effectiveness. Ultra than for Xpert. Ultra added BALF: bronchoalveolar lavagefluid clinical value in assisting the CI: confidence interval rapid and accurate diagnosis of MTB: Mycobacterium tuberculosis ACKNOWLEDGMENT pulmonary tuberculosis in MTB/RIF: Mycobacterium children when fiber-optic We thank Dr Joy Fleming (Chinese tuberculosis and/or bronchoscopy is recommended Academy of Sciences Institute of rifampicin according to a chest radiograph. Biophysics in Beijing) for help with OR: oddsratio Further studies are required to assess English-language editing of the RTI: respiratory tract infection its value in early diagnosis and the article. WHO: World Health Organization

Address correspondence to Adong Shen, MS, Key Laboratory of Major Diseases in Children, National Key Discipline of Pediatrics, National Clinical Research Center for Respiratory Diseases, Beijing Key Laboratory of Pediatric Respiratory Infectious Diseases, Beijing Pediatric Research Institute, Ministry of Education of the People’s Republic of China, Beijing Children’s Hospital, Capital Medical University, No. 56 Nanlishi Rd, Xicheng District, Beijing 100045, People’s Republic of China. E-mail: [email protected] PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). Copyright © 2019 by the American Academy of Pediatrics FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose. FUNDING: Supported by the National Science and Technology Major Project of China (grants 2018ZX10103001-003 and 2018ZX10101004-002-005). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the article POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose. COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2019-2944.

REFERENCES 1. WHO. Global Tuberculosis Report. 6. WHO. WHO Meeting Report of 10. Zar HJ, Workman L, Isaacs W, et al. Geneva, Switzerland: World Health a Technical Expert Consultation: Non- Rapid diagnosis of pulmonary Organization; 2018 Inferiority Analysis of Xpert MTB/RIF tuberculosis in African children in a primary care setting by use of Xpert 2. Lamb GS, Starke JR. Tuberculosis in Ultra Compared to Xpert MTB/RIF. MTB/RIF on respiratory specimens: infants and children. Microbiol Spectr. Geneva, Switzerland: World Health a prospective study. Lancet Glob Health. 2017;5(2):541–569 Organization; 2017 2013;1(2):e97–e104 3. Eddabra R, Ait Benhassou H. Rapid 7. Bahr NC, Nuwagira E, Evans EE, et al; 11. Zar HJ, Hanslo D, Apolles P, Swingler G, molecular assays for detection of ASTRO-CM Trial Team. Diagnostic Hussey G. Induced sputum versus tuberculosis. Pneumonia (Nathan). accuracy of Xpert MTB/RIF Ultra for gastric lavage for microbiological tuberculous meningitis in HIV-infected 2018;10:4 confirmation of pulmonary tuberculosis adults: a prospective cohort study. in infants and young children: 4. WHO. Automated Real-Time Nucleic Acid Lancet Infect Dis. 2018;18(1):68–75 Amplification Technology for Rapid and a prospective study. Lancet. 2005; Simultaneous Detection of Tuberculosis 8. Berhanu RH, David A, da Silva P, et al. 365(9454):130–134 Performance of Xpert MTB/RIF, Xpert and Rifampicin Resistance: Xpert MTB/ 12. Fan L, Li D, Zhang S, et al. Parallel tests RIF Assay for the Diagnosis of Ultra, and Abbott RealTime MTB for using culture, Xpert MTB/RIF, and SAT-TB Pulmonary and Extrapulmonary TB in diagnosis of pulmonary tuberculosis in in sputum plus bronchial alveolar Adults and Children: Policy Update. a high-HIV-burden setting. J Clin lavage fluid significantly increase Geneva, Switzerland: World Health Microbiol. 2018;56(12):e00560-18 diagnostic performance of smear- Organization; 2013 9. Sabi I, Rachow A, Mapamba D, et al. negative pulmonary tuberculosis. Front 5. WHO. Guidance for National Xpert MTB/RIF Ultra assay for the Microbiol. 2018;9:1107 Tuberculosis Programmes on the diagnosis of pulmonary tuberculosis in 13. Walters E, Goussard P, Bosch C, Management of Tuberculosis in children: a multicentre comparative Hesseling AC, Gie RP. GeneXpert MTB/RIF Children. Geneva, Switzerland: World accuracy study. J Infect. 2018;77(4): on bronchoalveolar lavage samples in Health Organization; 2014 321–327 children with suspected complicated

Downloaded from www.aappublications.org/news by guest on September 27, 2021 PEDIATRICS Volume 144, number 5, November 2019 7 intrathoracic tuberculosis: a pilot study. 17. Villanueva P, Neth O, Ritz N, Tebruegge tuberculosis and resistance to rifampin Pediatr Pulmonol. 2014;49(11): M; Paediatric Tuberculosis Network in an assay suitable for point-of-care 1133–1137 European Trials Group. Use of Xpert testing. MBio. 2017;8(4):e00812-17 MTB/RIF Ultra assays among paediatric 14. Graham SM, Ahmed T, Amanullah F, tuberculosis experts in Europe. Eur 22. Friedrich SO, Rachow A, Saathoff E, et al. Evaluation of tuberculosis Respir J. 2018;51(5):1800346 et al; Pan African Consortium for the diagnostics in children: 1. Proposed Evaluation of Anti-tuberculosis clinical case definitions for 18. Nicol MP, Workman L, Prins M, et al. Accuracy of Xpert MTB/RIF Ultra for the Antibiotics (PanACEA). Assessment of classification of intrathoracic diagnosis of pulmonary tuberculosis in the sensitivity and specificity of Xpert tuberculosis disease. Consensus from children. Pediatr Infect Dis J. 2018; MTB/RIF assay as an early sputum an expert panel. J Infect Dis. 2012; 37(10):e261–e263 biomarker of response to tuberculosis 205(suppl 2):S199–S208 19. Tuberculosis Branch of Chinese Medical treatment. Lancet Respir Med. 2013; 15. Dorman SE, Schumacher SG, Alland D, Association. Guidelines for the 1(6):462–470 et al; Study Team. Xpert MTB/RIF Ultra diagnosis and treatment of bronchial 23. Metcalfe JZ, Makumbirofa S, Makamure for detection of Mycobacterium tuberculosis of trachea. Clin J Tuberc B, et al. Suboptimal specificity of Xpert tuberculosis and rifampicin resistance: Respir Dis. 2012;35:581–587 MTB/RIF among treatment-experienced a prospective multicentre diagnostic 20. Opota O, Zakham F, Mazza-Stalder J, accuracy study. Lancet Infect Dis. 2018; patients. Eur Respir J. 2015;45(5): et al. Added value of Xpert MTB/RIF – 1504–1506 18(1):76 84 Ultra for diagnosis of pulmonary 16. Kunkel A, Abel Zur Wiesch P, tuberculosis in a low-prevalence 24. Theron G, Venter R, Smith L, et al. False- Nathavitharana RR, et al. Smear setting. J Clin Microbiol. 2019;57(2): positive Xpert MTB/RIF results in positivity in paediatric and adult e01717-18 retested patients with previous tuberculosis: systematic review and 21. Chakravorty S, Simmons AM, Rowneki tuberculosis: frequency, profile, and meta-analysis. BMC Infect Dis. 2016;16: M, et al. The new Xpert MTB/RIF Ultra: prospective clinical outcomes. J Clin 282 improving detection of Mycobacterium Microbiol. 2018;56(3):e01696-17

Downloaded from www.aappublications.org/news by guest on September 27, 2021 8 SUN et al A Test for More Accurate Diagnosis of Pulmonary Tuberculosis Lin Sun, Xue Qi, Fang Liu, Xirong Wu, Qingqin Yin, Yan Guo, Baoping Xu, Anxia Jiao, Yajie Guo, Weiwei Jiao, Chen Shen, Jing Xiao and Adong Shen Pediatrics originally published online October 25, 2019;

Updated Information & including high resolution figures, can be found at: Services http://pediatrics.aappublications.org/content/early/2019/10/23/peds.2 019-0262 References This article cites 20 articles, 2 of which you can access for free at: http://pediatrics.aappublications.org/content/early/2019/10/23/peds.2 019-0262#BIBL Subspecialty Collections This article, along with others on similar topics, appears in the following collection(s): Infectious Disease http://www.aappublications.org/cgi/collection/infectious_diseases_su b Permissions & Licensing Information about reproducing this article in parts (figures, tables) or in its entirety can be found online at: http://www.aappublications.org/site/misc/Permissions.xhtml Reprints Information about ordering reprints can be found online: http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news by guest on September 27, 2021 A Test for More Accurate Diagnosis of Pulmonary Tuberculosis Lin Sun, Xue Qi, Fang Liu, Xirong Wu, Qingqin Yin, Yan Guo, Baoping Xu, Anxia Jiao, Yajie Guo, Weiwei Jiao, Chen Shen, Jing Xiao and Adong Shen Pediatrics originally published online October 25, 2019;

The online version of this article, along with updated information and services, is located on the World Wide Web at: http://pediatrics.aappublications.org/content/early/2019/10/23/peds.2019-0262

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it has been published continuously since 1948. Pediatrics is owned, published, and trademarked by the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2019 by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news by guest on September 27, 2021