Eur opean Rev iew for Med ical and Pharmacol ogical Sci ences 2016; 20: 3980-3986 Long noncoding RNA PVT1 as a novel serum biomarker for detection of cervical cancer

J.-P. YANG, X.-J. YANG, L. XIAO, Y. WANG

Department of Laboratory, the Affiliated Hospital of Jiangxi University of Traditional Chinese Medicine, Nanchang, Jiangxi, China

Abstract. – OBJECTIVE : To investigate long There were 527,600 estimated new cases of CC noncoding RNA PVT1 expression in the serum and 265,700 estimated deaths of CC in 2012 of cervical cancer patients, and to evaluate worldwide 1. The prognosis of CC patients is serum PVT1 level as a diagnostic biomarker for closely associated with the clinical stage 2. Most cervical cancer. PATIENTS AND METHODS: Eighty-eight cervi - of CC patients with early stage could be suc - cal cancer patients, 64 cervical intraepithelial cessfully cured, but the survival rate of CC pa - neoplasia patients, 25 breast cancer patients, 25 tients diagnosed at the late stages is as low as ovarian cancer patients, and 111 healthy control 15% 3. Therefore, early diagnosis and therapy subjects were enrolled into this study. PVT1 could significantly reduce mortality of CC pa - serum level in these participants and PVT1 ex - tients 4,5 . pression in 20 pairs of cervical cancer tissues and adjacent paired normal tissues was mea - Although cervical smear cytological exami - sured by quantitative reverse -poly - nation is the mainly screening method of early merase chain reaction. The diagnostic values of diagnosis of CC, the inconveniences of exam in - serum PVT1 were evaluated by receiver operat - ing and high false-negative rate limit its applica - ing characteristic curves analysis. tion 6,7 . In addition to cervical smear cytological RESULTS: Serum PVT1 level is significantly in - examination, serum biomarkers, including carbo - creased in cervical cancer patients and correlat - hydrate antigen 125 (CA125) and squamous ed with tumor size, clinical stage, and lymph node metastasis of cervical cancer. Serum PVT1 carcinoma antigen (SCC Ag), are commonly 8,9 could accurately discriminate cervical cancer used for the diagnosis of CC . However, the low patients from cervical intraepithelial neoplasia diagnostic sensitivity and specificity limit their patients and healthy control subjects, and also utility 10 . Therefore, novel noninvasive diagnostic discriminate early stage cervical cancer patients biomarkers with high sensitivity and specificity from healthy control subjects. But serum PVT1 are urgently required. level is not changed in breast cancer and ovari - an cancer patients. Furthermore, serum PVT1 Long noncoding RNAs (lncRNAs), which are level is positively correlated with tissue PVT1 ex - longer than 200 nucleotides with limited protein - pression, and could indicate cervical cancer dy - coding potential, have multiple functions in vari - namics. ous pathophysiological processes and are dysreg - CONCLUSIONS: Long noncoding RNA PVT1 ulated in many kinds of diseases, particular in may be a novel noninvasive biomarker for early cancers 11-15 . LncRNA PVT1, CCAT2, HOTAIR, diagnosis of cervical cancer . XLOC_010588, MALAT1, lnc-CC3, and et al. Key Words: have been reported to be upregulated in CC and 16-22 Long noncoding RNA, PVT1, Cervical cancer, promote CC progression . While other lncR - Serum, Biomarker, Diagnosis . NAs, such as GAS5, LET, and MEG3 are down - regulated in CC and inhibit CC progression 23-25 . However, whether these lncRNAs are present in peripheral blood and could be used as diagnostic biomarkers for CC is still unknown. LncRNAs Introduction are really detectable and could be used as diag - nostic biomarkers for many cancers, such as Cervical cancer (CC) is the fourth most com - lncRNA H19, CUDR, LSINCT-5, PTPEN1, and mon cancer and fourth most frequent cause of UCA1 for gastric cancer 26-28 ; HOTTIP-005 and cancer mortality affecting woman in the world 1. HOTTIP-001 for pancreatic cancer 29 ; POU3F3

3980 Corresponding Author: Ying Wang, MD; e-mail: [email protected] Long noncoding RNA PVT1 as a novel serum biomarker for detection of cervical cancer for esophageal squamous cell carcinoma 30 ; XIST RNA Extraction and HIF1A-AS1 for non-small cell lung cancer 31 ; Venus blood was collected from fasting sub - UCA1, WRAP53, uc003wbd, and AF085935 for jects and centrifuged at 3000 rpm for 15 min at 4 hepatocellular carcinoma 32,33 . °C. The supernatant was immediately collected Many reports have shown that lncRNA PVT1 and stored at -80 °C until use . All CC tissues and was increased in CC tissues, associated with paired normal tissues were immediately frozen in poor prognosis, and promoted CC develop - liquid nitrogen and stored at -80 °C until use af - ment 16,17 . In this study, we further measured ter surgery. Total RNAs were extracted from tis - serum PVT1 levels in a large cohort of CC pa - sues and serum using the Trizol reagent (Invitro - tients, cervical intraepithelial neoplasia (CIN) gen, Carlsbad, CA, USA ) according to the manu - patients, and healthy control (HC) subjects, and facturer’s instructions . analyzed the diagnostic sensitivity and specifici - ty of serum PVT1 for CC . Our results showed Quantitative Reverse Transcription- that serum PVT1 is significantly increased in CC Polymerase Chain Reaction (qRT-PCR) patients and could accurately discriminate CC LncRNAs expressions in serum and tissues of patients from CIN patients and HC subjects. CC patients were measured with qRT-PCR as - However, serum PVT1 is not changed in breast says. Complementary DNA was synthesized us - cancer and ovarian cancer patients. Furthermore, ing the PrimeScript™ II 1st Strand cDNA Syn - serum PVT1 level is positively correlated with thesis Kit (Takara, Dalian, China) following the tissue PVT1 expression, and serum PVT1 level manufacturer’s instructions . Quantitative PCR is significantly decreased in post-surgery CC pa - was performed using the SYBR ® Premix Ex tients. Collectively, our data suggested that Taq™ II (Takara) on StepOnePlus™ Real-Time PVT1 may be a novel noninvasive diagnostic PCR Systems (Applied Biosystems, Foster City, biomarker for CC. CA, USA) following the manufacturer’s instruc - tions in triplicate . PVT1 expression in serum and tissues was normalized to GAPDH using the Materials and Methods comparative Ct method. The primer sequences used were as follows: for PVT1 , 5’- ATAGATC - Patients and Samples CTGCCCTGTTTGC-3’ (forward) and 5’- The Ethics Committee of the Affiliated Hospi - CATTTCCTGCTGCCGTTTTC-3’ (reverse); for tal of Jiangxi University of Traditional Chinese GAPDH, 5’- GGAGCGAGATCCCTCCAAAAT- Medicine approved the collection and usage of 3’ (forward) and 5’-GGCTGTTGTCATACTTCT - human specimens . A total of 88 CC patients, 64 CATGG-3’ (reverse) . CIN patients, 25 breast cancer patients, 25 ovari - an cancer patients, and 111 healthy control (HC) Statistical Analysis subjects were enrolled into this study at the Affil - Statistical analyses were performed with the iated Hospital of Jiangxi University of Tradition - GraphPad Prism Software. Mann-Whitney U test al Chinese Medicine . All the patients were diag - or Wilcoxon signed-rank test was used to com - nosed and confirmed by pathological examina - pare serum or tissues PVT1 level between differ - tion in our hospital. CC clinical stages were de - ent groups as indicated. Receiver operating char - termined following the International Federation acteristic (ROC) curves analysis was used to as - of Gynecology and Obstetrics (FIGO). All the sess the sensitivity and specificity of serum HC subjects enrolled in this study were age- PVT1 for CC diagnosis. Pearson correlation matched women with normal cervix examined in analysis was used to evaluate the correlation be - our hospital. Women diagnosed with other tu - tween serum and tissues PVT1 level. p-values < mors, pregnancy, or other chronic diseases are 0.05 was defined as statistically significant . excluded from our study. Twenty fresh CC tis - sues and adjacent paired normal tissues were ob - tained during the surgical resection in our hospi - Results tal. None of the patients received radiation thera - py, chemotherapy, or immunotherapy before the PVT1 is Increased in the Serum of collection of peripheral blood or surgical resec - CC Patients tion . All the patients and HC subjects signed in - To evaluate the clinical significance of serum formed consent . PVT1, we first quantified PVT1 expression by

3981 J.-P. Yang, X.-J. Yang, L. Xiao, Y. Wang qRT-PCR in the serum of 88 CC patients, 64 CIN Table I. Clinical features of the cervical cancer (CC) pa - patients, and 86 age-matched healthy control tients, cervical intraepithelial neoplasia (CIN) patients and (HC) subjects. The clinical features of CC pa - healthy control (HC) subjects. tients, CIN patients, and HC subjects are shown Variable CC CIN HC in Table I. In comparison to HC subjects, PVT1 was significantly increased in CIN patients, and All cases 88 64 86 was further increased in CC patients (Figure 1A). Age Next, we detected the correlation between serum < 50 49 36 45 ≥ 50 39 28 41 PVT1 level and clinical features in the 88 CC pa - Tumor size (cm) tients. As shown in Figure 1B, PVT1 was in - < 4 48 creased in CC patients with larger tumor size. ≥ 4 40 Furthermore, PVT1 was increased in CC patients Histology with advanced FIGO stage (Figure 1C). In com - Squamous 73 Adenocarcinoma 15 parison to CC patients without lymph node FIGO stage metastasis, PVT1 was increased in CC patients I 45 with lymph node metastasis (Figure 1D). II 39 III 4 Circulating PVT1 Could be Used as a Differentiation Well/moderate 58 Novel Noninvasive Biomarker for CC Poor 30 To evaluate whether serum PVT1 could be Lymph node metastasis used as a novel diagnostic biomarker for CC, re - Negative 66 Positive 22

Figure 1. Serum PVT1 levels in CC and CIN patients and its correlation with clinical features. A, Serum PVT1 levels in 88 CC patients, 64 CIN patients, and 86 HC subjects. B, Serum PVT1 levels in CC patients with tumor size < 4 cm or tumor size ≥ 4 cm. C, Serum PVT1 levels in CC patients at different FIGO stages. D, Serum PVT1 levels in CC patients with or without lymph node metastasis. Serum PVT1 levels were measured by qRT-PCR. Results are presented as median with interquartile range. * p < 0.05, *** p < 0.001 by Mann-Whitney U test.

3982 Long noncoding RNA PVT1 as a novel serum biomarker for detection of cervical cancer ceiver operating characteristic (ROC) curve stage I CC patients (AUC: 0.892; 95% CI: analysis were performed. ROC curve showed ac - 0.838-0.946; sensitivity: 86.7%; specificity: curate discrimination between CC patients and 73.3%), and stage I & II CC patients as well HC subjects, with an area under the ROC curve (AUC: 0. 929 ; 95% CI: 0.8 94-0.964 ; sensitivity: (AUC) of 0.932 (95% CI: 0.898-0.966), a sensi - 71.4 %; specificity: 97.7 %) (Figure 2D and E). tivity of 71.6%, and a specificity of 98.8% (Fig - These data suggested that PVT1 could be used ure 2A). ROC curve also showed that serum as a novel noninvasive biomarker for CC early PVT1 was sensitive and specific to discriminate diagnosis. CC patients from CIN patients, with an area un - der the ROC curve (AUC) of 0.817 (95% CI: 0.752-0.882), a sensitivity of 61.4%, and a speci - Diagnostic Specificity for CC of PVT1 ficity of 89.1% (Figure 2B), as well as CIN pa - To evaluate the diagnostic specificity of PVT1, tients from HC subjects, with an area under the we further measured serum PVT1 level in anoth - ROC curve (AUC) of 0.758 (95% CI: 0.683- er cohort, including 25 HC subjects, 25 breast 0.834), a sensitivity of 87.5%, and a specificity cancer (BC) patients, and 25 ovarian cancer (OC) of 52.3% (Figure 2C). patients. As shown in Figure 3, compared to HC We next evaluated the clinical values of subjects, serum PVT1 level was not significantly serum PVT1 in discriminating early stage CC different in BC and OC patients. These results patients from HC subjects. ROC curve showed implied that PVT1 might be a specific diagnostic good diagnostic sensitivity and specificity for biomarker for CC.

Figure 2. Diagnostic value of PVT1 for CC. A, ROC curve analysis of PVT1 for discrimination between CC and healthy control subjects (AUC: 0.932 (95% CI: 0.898-0.966), sensitivity: 71.6%, specificity: 98.8%). B, ROC curve analysis of PVT1 for discrimination between CC and CIN patients (AUC: 0.817 (95% CI: 0.752-0.882), sensitivity: 61.4%, specificity: 89.1%). C, ROC curve analysis of PVT1 for discrimination between CC and CIN patients (AUC: 0.758 (95% CI: 0.683- 0.834), sensitivity: 87.5%, specificity: 52.3%). D, ROC curve analysis of PVT1 for discrimination between stage I CIN pa - tients and HC subjects (AUC: 0.892 (95% CI: 0.838-0.946), sensitivity: 86.7%, specificity: 73.3%). E, ROC curve analysis of PVT1 for discrimination between stage I and II CC and CIN patients (AUC: 0.929 (95% CI: 0.894-0.964), sensitivity: 71.4%, specificity: 97.7%). p < 0.001 for all panels.

3983 J.-P. Yang, X.-J. Yang, L. Xiao, Y. Wang

CC patients and PVT1 expression in CC tissues (r = 0.838, p < 0.001) (Figure 4B). Furthermore, we measured serum PVT1 level from 20 pairs of pre- and post-surgery CC patients who under - went radical resection . As shown in Figure 4C, serum PVT1 level was significantly decreased in post-surgery CC patients (p < 0.001) . These data suggested that serum PVT1 level may indicate the expression of PVT1 in CC tissues and that PVT1 could be used for monitoring CC dynam - ics.

Discussion Figure 3. Serum PVT1 levels in breast cancer (BC) and ovarian cancer (OC) patients. Expression profiles of PVT1 Despite great progressions in CC diagnosis in the serum of 25 HC subjects, 25 BC patients, and 25 OC and therapy in the last twenty years, CC remains patients. Serum PVT1 levels were measured by qRT-PCR. to have high mortality due to the late diagnosis 34 . No statistical difference was observed between these three Therefore it is urgent to develop novel noninva - groups by Mann-Whitney U test. sive biomarkers for the early diagnosis of CC. Recently, many circulating nucleic acids have been investigated for their diagnostic values of Serum PVT1 Level Could be Used for cancers 35 . Serum microRNAs panel including Monitoring CC Dynamics miR-16-2*, miR-195, miR-2861, and miR-497, To further investigate the clinical values of has been reported to be used for the early detec - serum PVT1 in monitoring CC dynamics, we tion of CC 36 . In addition to microRNAs, lncR - measured PVT1 expression in 20 pairs of CC tis - NAs, another important type of noncoding sues and adjacent normal tissues using qRT-PCR. RNAs, have been reported to be detectable and Consistent with PVT1 expression in serum of CC used as potential serum biomarkers for early di - patients, PVT1 was also significantly increased agnosis of several cancers 32 . Although several in CC tissues compared with adjacent normal tis - microRNAs have been investigated in the early sues (Figure 4A). A significant positive correla - diagnosis of CC, the diagnostic value of lncR - tion was observed between serum PVT1 level of NAs in CC is still unknown.

Figure 4. Serum PVT1 levels indicate CC dynamics. A, PVT1 expressions in 20 pairs of CC tissues and adjacent normal tis - sues were measured by qRT-PCR. Results are presented as median with interquartile range. *** p < 0.001 by Wilcoxon signed- rank test. B, Serum PVT1 level was significantly positive correlated with PVT1 expression in CC tissues. r = 0.838, p < 0.001 by Pearson correlation analysis. C, Serum PVT1 levels in 20 pairs of pre- and post-surgery CC patients were measured by qRT-PCR. p < 0.001 by Wilcoxon signed-rank test.

3984 Long noncoding RNA PVT1 as a novel serum biomarker for detection of cervical cancer

––––––––––––––––– –– In this work, we focused on the lncRNA Acknowledgements PVT1, which has been reported to be upregulated This study was supported by a grant from the National Nat - in CC tissues and function as an in ural Science Foundation of China (No. 81206578) . CC 16, 17 . We measured serum PVT1 level in CC patients, CIN patients, and healthy control (HC) –––––––––––––––– –-– –– subjects. The results revealed that compared to Conflict of Interest HC subjects, serum PVT1 level is increased in CIN patients, and is further increased in CC pa - The Authors declare that there are no conflicts of interest. tients. Serum PVT1 level in CC is correlated with tumor size, FIGO stage, and lymph node metastasis. ROC curve analysis showed that References serum PVT1 level accurately discriminates CC 1) TORRE L A, B RAY F, S IEGEL R L, F ERLAY J, L ORTET -T IEU - patients from CIN patients and HC subjects, and LENT J, J EMAL A. Global cancer statistics, 2012. CA also discriminates CIN patients from HC sub - Cancer J Clin 2015; 65: 87-108. jects. Furthermore, serum PVT1 level could also 2) Testing for cervical cancer: new recommenda - accurately discriminate early stage CC patients tions from the American Cancer Society, Ameri - from HC subjects. These data demonstrated that can Society for Colposcopy and Cervical Patholo - serum PVT1 could be used as a novel noninva - gy, and American Society for Clinical Pathology. CA Cancer J Clin 2012; 62: 211-212. sive biomarker for the early diagnosis of CC. The 3) BURKI T K . Cervical cancer: screening and risk with combination of PVT1 with other lncRNAs , mi - age. Lancet Oncol 2014; 15: e107. croRNAs and proteins would be more effective 4) JELASTOPULU E, F AFLIORA E, P LOTA A, B ABALIS V, B ART - for the diagnosis of CC, and this hypothesis SOKAS C, P OULAS K, P LOTAS P. Knowledge, behav - needs further investigation. iours and attitudes regarding HPV infection and In addition to CC, we also measured serum its prevention in female students in West Greece. PVT1 level in breast cancer and ovarian cancer Eur Rev Med Pharmacol Sci 2016; 20: 2622- patients. Serum PVT1 level is not changed in 2629. breast cancer and ovarian cancer patients, imply - 5) HALTAS H, B AYRAK R, Y ENIDUNYA S, Y ILDIRIM U. The im - munohistochemical detection of P16 and HPV L1 ing that the diagnostic value of PVT1 may be capsid protein on cell block sections from residual specific to CC. But serum PVT1 level in more PapSpin liquid-based gynecology cytology speci - types of cancers needs to be investigated. mens as a diagnostic and prognostic tool. Eur In addition , we investigated serum PVT1 lev - Rev Med Pharmacol Sci 2012; 16: 1588-1595. els during the dynamics of CC. Our results re - 6) YILMAZ E, C ELIK O, C ELIK E, T URKCUOGLU I, S IMSEK Y, vealed that serum PVT1 level was significant KARAER A, O TLU B, G ULBAY G, Y ESILADA E. XPD and XRCC1 polymorphism in patients with nor - positive correlated with CC tissue PVT1 level. mal and abnormal cervical cytology by pap Serum PVT1 level was significantly reduced af - smear. Eur Rev Med Pharmacol Sci 2012; 16: ter surgery section of CC. These findings im - 1713-1718. plied that serum PVT1 mainly derived from CC 7) COMPARETTO C,B ORRUTO F. Cervical cancer screen - tissues and could be used for the monitor of CC ing: a never-ending developing program. World J dynamics . Clin Cases 2015; 3: 614-624. 8) ZHANG X, W ANG C, W ANG L, D U L, W ANG S, Z HENG G, L I W, Z HUANG X, Z HANG X, D ONG Z. Detection of circulating Bmi-1 mRNA in plasma and its poten - tial diagnostic and prognostic value for uterine Conclusions cervical cancer. Int J Cancer 2012; 131: 165-172. 9) SALVATICI M, A CHILARRE M T, S ANDRI M T, B OVERI S, We have observed that serum PVT1 level is VANNA Z, L ANDONI F. Squamous cell carcinoma mostly increased in CC patients , correlated with antigen (SCC-Ag) during follow-up of cervical tumor size, FIGO stage, and lymph node metas - cancer patients: role in the early diagnosis of re - tasis. Serum PVT1 level shows high accuracy in currence. Gynecol Oncol 2016; 142: 115-119. discriminating CC patients from CIN patients 10) WENTZENSEN N, S CHIFFMAN M. Filling a gap in cervi - cal cancer screening programmes. Lancet Oncol and HC subjects, and also discriminating CIN 2014; 15: 249-251. patients from HC subjects. Serum PVT1 level in - 11) MARUYAMA R, S UZUKI H, Y AMAMOTO E, I MAI K, S HINO - dicates CC dynamics. Our findings demonstrate MURA Y. Emerging links between epigenetic alter - that PVT1 may be a novel noninvasive biomarker ations and dysregulation of noncoding RNAs in for CC early diagnosis . cancer. Tumour Biol 2012; 33: 277-285.

3985 J.-P. Yang, X.-J. Yang, L. Xiao, Y. Wang

12) PONTING C P, O LIVER P L, R EIK W. Evolution and 25) ZHANG J, Y AO T, W ANG Y, Y U J, L IU Y, L IN Z. Long functions of long noncoding RNAs. Cell 2009; noncoding RNA MEG3 is downregulated in cervi - 136(4): 629-641. cal cancer and affects cell proliferation and apop - 13) MERCER T R, D INGER M E, M ATTICK J S . Long non- tosis by regulating miR-21. Cancer Biol Ther coding RNAs: insights into functions. Nat Rev 2016; 17: 104-113. Genet 2009; 10: 155-159. 26) ZHOU X, Y IN C, D ANG Y, Y E F, Z HANG G. Identification 14) YUAN J H, Y ANG F, W ANG F, M A J Z, G UO Y J, T AO Q of the long non-coding RNA H19 in plasma as a F, L IU F, P AN W, W ANG T T, Z HOU C C, W ANG S B, novel biomarker for diagnosis of gastric cancer. WANG Y Z, Y ANG Y, Y ANG N, Z HOU W P, Y ANG G S, Sci Rep 2015; 5: 11516. SUN S H . A long noncoding RNA activated by 27) DONG L, Q I P, X U M D, N I S J, H UANG D, X U Q H, TGF-beta promotes the invasion-metastasis cas - WENG W W, T AN C, S HENG W Q, Z HOU X Y, D U X. cade in hepatocellular carcinoma. Cancer Cell Circulating CUDR, LSINCT-5 and PTENP1 long 2014; 25: 666-681. noncoding RNAs in sera distinguish patients with 15) HAJJARI M, K HOSHNEVISAN A, S HIN Y K . Molecular gastric cancer from healthy controls. Int J Cancer function and regulation of long non-coding RNAs: 2015; 137: 1128-1135. paradigms with potential roles in cancer. Tumour 28) GAO J, C AO R, M U H. Long non-coding RNA UCA1 Biol 2014; 35: 10645-10663. may be a novel diagnostic and predictive bio - 16) ZHANG S, Z HANG G, L IU J. Long noncoding RNA marker in plasma for early gastric cancer. Int J PVT1 promotes cervical cancer progression Clin Exp Pathol 2015; 8: 12936-12942. through epigenetically silencing miR-200b. APMIS 29) WANG Y, L I Z, Z HENG S, Z HOU Y, Z HAO L, Y E H, Z HAO 2016; 124: 649-658. X, G AO W, F U Z, Z HOU Q, L IU Y, C HEN R. Expression 17) IDEN M, F YE S, L I K, C HOWDHURY T, R AMCHANDRAN R, profile of long non-coding RNAs in pancreatic RADER J S . The lncRNA PVT1 contributes to the cer - cancer and their clinical significance as biomark - vical cancer phenotype and associates with poor ers. Oncotarget 2015; 6: 35684-35698. patient prognosis. PLoS One 2016; 11: e0156274. 30) TONG Y S, W ANG X W, Z HOU X L, L IU Z H, Y ANG T X, 18) CHEN X, L IU L, Z HU W. Up-regulation of long non- SHI W H, X IE H W, L V J, W U Q Q, C AO X F . Identifi - coding RNA CCAT2 correlates with tumor metas - cation of the long non-coding RNA POU3F3 in tasis and poor prognosis in cervical squamous plasma as a novel biomarker for diagnosis of cell cancer patients. Int J Clin Exp Pathol 2015; 8: esophageal squamous cell carcinoma. Mol Can - 13261-13266. cer 2015; 14: 3. 19) KIM H J, L EE D W, Y IM G W, N AM E J, K IM S, K IM S W, 31) TANTAI J, H U D, Y ANG Y, G ENG J. Combined identifi - KIM Y T . Long non-coding RNA HOTAIR is associ - cation of long non-coding RNA XIST and HIF1A- ated with human cervical cancer progression. Int AS1 in serum as an effective screening for non- J Oncol 2015; 46: 521-530. small cell lung cancer. Int J Clin Exp Pathol 2015; 8: 7887-7895. 20) LIAO L M, S UN X Y, L IU A W, W U J B, C HENG X L, L IN J X, Z HENG M, H UANG L. Low expression of long 32) KAMEL MM, M ATBOLI M, S ALLAM M, M ONTASSER IF, noncoding XLOC_010588 indicates a poor prog - SAAD AS, E L-T AWDI AH . Investigation of long non - nosis and promotes proliferation through upregu - coding RNAs expression profile as potential lation of c- in cervical cancer. Gynecol Oncol serum biomarkers in patients with hepatocellular 2014; 133: 616-623. carcinoma. Transl Res 2016; 168: 134-145. 21) YANG L, B AI H S, D ENG Y, F AN L. High MALAT1 ex - 33) LU J, X IE F, G ENG L, S HEN W, S UI C, Y ANG J. Investi - pression predicts a poor prognosis of cervical gation of serum lncRNA-uc003wbd and lncRNA- cancer and promotes cancer cell growth and inva - AF085935 expression profile in patients with he - sion. Eur Rev Med Pharmacol Sci 2015; 19: patocellular carcinoma and HBV. Tumour Biol 3187-3193. 2015; 36: 3231-3236. 22) JIANG B, S UN R, F ANG S, Q IN C, P AN X, P ENG L, L I Y, 34) CROSBIE EJ, E INSTEIN MH, F RANCESCHI S, K ITCHENER LI G. Lnc-CC3 increases metastasis in cervical HC . Human papillomavirus and cervical cancer. cancer by increasing Slug expression. Oncotarget Lancet 2013; 382: 889-899. 2016 May 20. doi: 10.18632/oncotarget.9519. 35) JUAN L, T ONG HL, Z HANG P, G UO G, W ANG Z, W EN X, [Epub ahead of print] DONG Z, T IAN YP. Identification and characteriza - 23) CAO S, L IU W, L I F, Z HAO W, Q IN C. Decreased ex - tion of novel serum microRNA candidates from pression of lncRNA GAS5 predicts a poor prog - deep sequencing in cervical cancer patients. Sci nosis in cervical cancer. Int J Clin Exp Pathol Rep 2014; 4: 6277. 2014; 7: 6776-6783. 36) ZHANG Y, Z HANG D, W ANG F, X U D, G UO Y, C UI W. 24) JIANG S, W ANG HL, Y ANG J. Low expression of long Serum miRNAs panel (miR-16-2*, miR-195, miR- non-coding RNA LET inhibits of 2861, miR-497) as novel non-invasive biomarkers cervical cancer. Int J Clin Exp Pathol 2015; 8: for detection of cervical cancer. Sci Rep 2015; 5: 806-811. 17942.

3986