Int J Clin Exp Med 2016;9(6):11859-11864 www.ijcem.com /ISSN:1940-5901/IJCEM0025330 Originl Article Serum tumor mrkers used for predicting esophgogstric junction denocrcinom in esophgel mlignncy Yongkng Chen 1, Xin Qi 2, Yue Xu 3, Feng Chen 1, Mingrong Wng 3, Jun Qi 1 1 Deprtment of Clinicl Lbortory, Cncer Hospitl, Chinese Acdemy of Medicl Sciences & Peking Union Medicl College, Beijing 100021, Chin; 2 Deprtment of Clinicl Lbortory, Chin-Jpn Union Hospitl of Jilin University, Jilin Chngchun 130031, Chin; 3 Stte Key Lbortory of Moleculr Oncology, Cncer Hospitl, Chinese Acdemy of Medicl Sciences & Peking Union Medicl College, Beijing 100021, Chin Received Jnury 31, 2016; Accepted My 1, 2016; Epub June 15, 2016; Published June 30, 2016 Abstrct: Bckground/ims: To evlute pnel of complementry biomrkers for differentil prediction of esophgogstric junction denocrcinom (EGAC) from other esophgel mlignncy. Methods: Preopertive CA199, CEA, SCC nd CA72-4 were evluted mong 410 ptients (164 EGAC, 192 esophgel squmous cell crcinom (ESCC) nd 54 benign esophgel diseses). Kruskl-Wllis test, Wilcoxon rnk-sum test, receiver operting chrcteristic curve nd logistic regression nlysis were pplied. Results: The medin CA19-9, CEA nd CA72-4 levels were sttisticlly higher in ptients with EGAC thn in those with ESCC (Wilcoxon rnk sum test; P<0.001, respectively). Being djusted to optiml cut-off vlue, positive CA19-9, positive CEA, negtive SCC nd positive CA72-4 were significntly correlted with high risk of EGAC (Logistics regression; CA19-9 OR=9.600 P<0.001; CEA OR=7.368 P<0.001; SCC OR=2.754 P<0.001; CA72-4 OR=2.487 P=0.001). When pplied s mrkers pnel, serum tumor mrkers show medium sensitivity (65.2%) nd high specificity (79.7%) (AUC=0.791 P<0.001) in EGAC prediction. Conclusions: Synchronous exertion of CA72-4, CEA, SCC, nd CA19-9 could be vluble method to distinguish esophgogstric junction denocrcinom from esophgel squmous cell crcinom. Keywords: Serum tumor mrkers, prediction, esophgogstric junction denocrcinom, esophgel squmous cell crcinom Introduction Growing rpidly in Western countries [1], the incidence of the esophgogstric junction denocrcinom (EGAC) hs been discussed in Chin. Study from Chin reported tht the prevlence of EGAC, which hd yerly increse, ws higher thn esophgel squmous cell crcinom (ESCC) nd gstric crcinom (GC) in some high risk res [2, 3]. The helthy problems of EGAC drw the public ttentions. Siewert, et l clssified the EGAC into three types bsed purely on the ntomic loction of the epicenter of the tumor or the loction of the tumor mss [4]. Chnging the clssifiction slightly, NCCN 2015 defines these three subtypes of EGAC s: type I, the denocrcinom of the distl esophgus with the tumor center locted within 1 to 5 cm bove the ntomic esophgogstric junction (EGJ); type II, the true crcinom of the crdi with the tumor cnter within 1 cm bove nd 2 cm below the EGJ; type III, the subcrdil crcinom with the tumor center between 2 to 5 cm below the EGJ, infiltrting the EGJ nd the distl esophgus from below [5]. Attributing to the complicted ntomicl structure of EGJ, opertive pthwy for surgicl opertion requires differentil dignosis of EGAC from esophgel msses [6, 7]. The dignosis of EGAC minly depends on invsive exmintion, such s endoscopy nd biopsy [7]. Serum biomrker is useful indictor in dignosis, surveillnce nd prognosis. However, there is few rticle focusing on the sensitive nd specific tumor biomrkers for distinguishing EGAC from esophgel msses. In this study, we hypothesized tht the combined use of severl serum tumor mrkers (CA72-4, CEA,
Serum mrker for esophgogstric junction denocrcinom Tble 1. Chrcteristics of the study popultion Age (yer) Sex Benign disese (n=54) Esophgel squmous cell crcinom (n=192) Esophgogstric junction denocrcinom (n=164) Medin 56 60 62.5 Rnge (20-76) (40-82) (37-84) Mle 36 158 143 Femle 18 34 21 Histologicl types of benign disese Dysplsi 10 Chronic inflmmtion 24 Hitl herni 5 Benign tumor 15 CA19-9 (U/mL) Men 10.71 12.94 301.36 Rnge 0.6-27.59 0.6-160.5 0.6-38060 CEA (ng/ml) Men 2.55 4.78 18.84 Rnge 0.718-8.59 0.406-293.4 0.2-538.2 SCC (U/mL) Men 1.11 1.77 0.98 Rnge 0.3-4.9 0.3-31.3 0.3-4.5 CA72-4 (ku/ml) Men 2.58 3.31 14.24 Rnge 0.736-73.240 0.579-32.7 0.327-629.7 SCC nd CA19-9) my increse the sensitivity nd ccurcy for the dignosis of EGAC. Thus, employing the pnel of tumor mrkers, our reserch ims to distinguish EGAC from other mlignnt esophgel diseses, especilly ESCC. Methods Study popultion A totl of 410 inptients with preopertive serum vilble for protein nlysis were enrolled from Cncer Hospitl, Chinese Acdemy of Medicl Sciences (CAMS) & Peking Union Medicl College (PUMC), Beijing, Chin from July 2012 to My 2013. All ptients were newly dignosed nd previously untreted. The pthologicl dt for these ptients ws obtined by endoscopy nd confirmed by biopsy. This study ws pproved by the Institute Review Bord of the Cncer Hospitl, CAMS & PUMC. Ech ptient ws informed bout the study nd gve his/her consent. Mesurement of tumor mrkers Four serum tumor mrkers, CEA, CA19-9, CA72-4 nd SCC, were mesured before tretment. CEA, CA72-4 nd CA19-9 levels were investigted using electrochemiluminescence immunossy (ECLIA) (Roche, Germny), while SCC levels were investigted using chemiluminescent enzyme immunossy (CLEIA) (Abbott, USA). These ssys were done t the Deprtment of Clinicl Lbortory, Cncer Hospitl, Chinese Acdemy of Medicl Science. The cut-off vlues for CEA, CA19-9, SCC, nd CA72-4 were defined s 5.0 ng/ml, 37 U/mL, 1.5 U/mL nd 9.8 ku/ml, respectively. All cutoff vlues were recommended by mnufcturer, nd confirmed by the Deprtment of Clinicl Lbortory. The sensitivity of the mrkers ws clculted s the number of ptients who showed elevted levels of ech mrker bove the cut-off vlue divided by the totl number of ptients. The sensitivity of the combintion of multiple mrkers ws clculted s the percentge of ptients who showed elevtion in ny of the combined mrkers. 11860 Int J Clin Exp Med 2016;9(6):11859-11864
Serum mrker for esophgogstric junction denocrcinom Tble 2. Higher CA19-9, CEA nd CA72-4 levels re ssocited with EGAC rther thn ESCC Benign disese (n=54) Esophgel squmous cell crcinom (n=192) Esophgogstric junction denocrcinom (n=164) CA19-9 (U/mL) 8.59 9.26 b 11.435 b CEA (ng/ml) 2.565 2.435 b 3.19 b SCC (U/mL) 0.9 1.1 b 0.8 b CA72-4 (ku/ml) 1.675 1.875 b 2.795 b Kruskl-Wllis test ws used to compre the difference of tumor mrkers between ptients with different esophgel disese; b Wilcoxon rnk sum test ws used to compre the difference of tumor mrkers between ptients with esophgogstric junction denocrcinom, nd esophgel squmous cell crcinom, P<0.05 ws considered sttisticlly significnt. confidence intervls (CIs). All sttisticl nlyses were conducted using SPSS 19.0 for Windows (SPSS Inc., USA). P vlues below 0.05 were considered sttisticlly significnt. Results Ptient chrcteristics Ptient chrcteristics re shown in Tble 1. A totl of 410 ptients enrolled in the study. Of the ptients enrolled, 54 ptients (13.2%) hd dignosis of benign esophgel disese. One hundred nd ninety-two ptients (46.8%) were dignosed with esophgel squmous cell crcinom, nd 164 ptients (40.0%) were dignosed with esophgogstric junction denocrcinom. Higher CA19-9, CEA nd CA72-4 levels re ssocited with EGAC rther thn ESCC Figure 1. Are under the receiver operting chrcteristic curve vlues of mrkers pnel used for predicting EGAC. Mrkers pnel: AUC, 0.791 (95% CI, 0.745-0.832), P<0.001. Sttisticl nlysis All dt ordered in n bnorml distribution re expressed s the Medin (Rnge). The nonprmetric Kruskl-Wllis test nd Wilcoxon rnk-sum test were used for the comprison of ech originl vlue of tumor mrker between groups. To investigte the dignostic efficcy of tumor mrkers, receiver operting chrcteristic (ROC) curve ws constructed using the serum concentrtions in both EGAC nd ESCC to ssess the predicting efficcy. Additionlly, An optiml cut-off vlue, hving the highest Youden Index, would be point nerest to the top of verticl xis in ROC. Logistic regression models were used to detect ssocitions of tumor dignosis with ech of serum tumor mrker nd provided estimtes of odds rtio (ORs) nd The medin of tumor mrker levels between ptients with different types of underlying esophgel disese is shown in Tble 2. There ws significnt difference between tumor mrker levels in those ptients with benign lung disese, esophgogstric junction denocrcinom nd esophgel squmous cell crcinom (Kruskl-Wllis test, CA19-9: P= 0.0019; CEA: P<0.001; SCC: P<0.001; CA72-4: P<0.001). There ws sttisticlly significnt difference between medin tumor mrker levels in ptients with EGAC compred with those ptients with ESCC (Wilcoxon rnk sum test, CA19-9: P<0.001; CEA: P<0.001; SCC: P< 0.001, CA72-4: P<0.001). The medin vlues of CA19-9, CEA, CA72-4 re sttisticlly higher in EGAC ptients thn in ESCC ptients, while SCC ws lower in EGAC. Serum mrkers with djusted cut-off vlue contribute to distinguishing EGAC from ESCC A receiver operting chrcteristic (ROC) curve for tumor mrkers in distinguishing different mlignnce involvement is shown in Figure 1. Significnce ws observed regrding the re under curve (AUC) of serum mrkers (CA19-9 AUC=0.597 P=0.0017; CEA AUC=0.643 P< 0.001; SCC AUC=0.634 P<0.001; CA72-4 AUC 11861 Int J Clin Exp Med 2016;9(6):11859-11864
Serum mrker for esophgogstric junction denocrcinom Tble 3. The optiml cut-off vlue of tumor mrkers to distinguish EGAC from ESCC AUC 95% confidence intervl P<0.001) (Tble 3). To sttisticlly develop the effective differentil vlue of tumor mrkers, the study identified the point nerest to the top of verticl xis in ROC s the optiml cut-off vlue. Logistic regression models identified the tumor mrkers s sttisticlly significnt between EGAC nd ESCC (Tble 4). In prticulrly, positive CA19-9 beyond 34.61 U/ml hd the highest risk of EGAC (OR=9.600, P<0.001). Ptients with CEA >4.97 or SCC <1.3 or CA72-4 my more likely EGAC (CEA OR=7.368 P<0.001; SCC OR=2.754 P<0.001; CA72-4 OR=2.487 P= 0.001). This result indictes CA199, CEA, SCC P Optiml cutoff vlue CA19-9 (U/mL) 0.597 0.544-0.648 0.0017 34.61 CEA (ng/ml) 0.643 0.591-0.693 <0.001 4.97 SCC (U/mL) 0.634 0.582-0.685 <0.001 1.3 CA72-4 (ku/ml) 0.637 0.584-0.687 <0.001 4.1 AUC: Are under the receiver operting chrcteristic curve; Optiml cut-off vlue: the cut-off vlue representing highest Youden Index, the point nerest to the top of verticl xis in receiver operting chrcteristic curve. P<0.05 ws considered sttisticlly significnt. Tble 4. Serum mrkers with djusted cut-off vlue contribute to distinguishing EGAC from ESCC Indictors Odds rtio (OR) 95% confidence intervl CA19-9 >34.61 9.600 3.427-26.888 0.000 CEA >4.97 7.368 3.599-15.084 0.000 SCC <1.3 2.754 1.563-4.853 0.000 CA72-4 >4.1 2.487 1.437-4.304 0.001 Logistics regression ws used to nlysis the ttribution of different serum tumor mrkers to predicting EGAC. Attribution expresses s odds rtio. P<0.05 ws considered sttisticlly significnt. Tble 5. Sensitivity nd specificity of CA19-9, CEA, SCC, CA72-4 nd the mrkers pnel for predicting the dignosis of EGAC Tumor mrker Sensitivity (%) Specificity (%) Youden Index CA19-9 >34.61 24.39 97.40 0.2179 CEA >4.97 38.41 94.74 0.3315 SCC <1.3 79.88 40.10 0.1998 CA72-4 >4.1 42.07 82.29 0.2436 Mrkers pnel 65.24 79.69 0.4493 Youden Index: single sttistic tht cptures the performnce of dignostic test. Youden Index = sensitivity + specificity - 1. P (0.4493) would be more effective thn single mrkers in serum predicting of esophgogstric junction denocrcinom. Discussion nd CA72-4 would be the effective rick indictors to EGAC. Mrkers pnel cn be helpful serum predictor for esophgogstric junction denocrcinom mong esophgel mlignncy Medin tumor mrkers vlues were mrkedly incresed in ptients with esophgogstric junction denocrcinom compred with ptients with ESCC, s shown in Tble 2. And positive CA19-9, CEA nd CA72-4 showed the highest correltion with underlying esophgogstric junction denocrcinom (Tble 4). Therefore, we hypothesized tht pnel of serum mrkers my be used to predict the type of esophgel mlignncy. Are under the receiver operting chrcteristic curve vlue ws used for evlute the efficiency of tumor mrkers fter combined ppliction. The re under the curve (AUC) ws 0.791 (95% CI= 0.745-0.832 P<0.0001) (Figure 1). Under the optiml cut-off vlue set, sensitivity nd specificity from single predictive mrkers were compred with those of combintion (Tble 5). Mrkers pnel conferred higher sensitivity of 65.24% nd lower specificity of 79.69% fter joint ppliction. Considering the blnce between sensitivity nd specificity, Youden Index suggested tht mrkers pnel Serum biomrkers, with less trum nd inconvenience, hve been speedy pproch for ssistnt dignosis, surveillnce nd prognosis [8-10]. As definition-renewed tumor in the dignosis of gstrointestinl msses [5], EGAC hs been discussed in the clinicl prctice only for short period of time. Severl indictors hve been found s prognostic vlue for EGAC ptients [11, 12]. However, these studies were minly focused on the reltionship between the 11862 Int J Clin Exp Med 2016;9(6):11859-11864
Serum mrker for esophgogstric junction denocrcinom serum biomrkers nd the survivl rte of ptients with EGAC. Our study firstly discusses the indicting competence of serum biomrkers to distinguish EGAC from other esophgel msses. Our results exhibited tht difference of serum biomrkers between ptients with EGAC nd ptients with ESCC is significnt. Therefore, elevted serum biomrkers vlue is helpful predictor of existence of EGAC. Surgicl strtegy options for ptients with esophgel mlignncy were determined by extent of their lymphtic metstsis [13]. Considering the underlying lymph node involvement round gstric re in EGAC compred with ESCC, gstrectomy or semigstrectomy is recommended to ssist tumor excision [14, 15]. Since the metstsis route is vried with the histopthologicl type, serologicl prediction before dissection fcilittes the tretment options. As our results show, indictive risk on EGAC from elevted CA19-9 vlue ws 9.6 times s high s vlue under cut-off (Tble 4). Gstrointestinl endoscope nd endoscopic ultrsonogrphy re the best preopertive choice in dignosis of gstrointestinl msses [16]. Since the concept regrding esophgogstric junction is generlized recently, endoscopic definition of esophgogstric junction endoscopic detection still need some systemtic trining in the ntionwide [17, 18]. Lck of stndrdiztion in EGAC differentil dignosis from esophgel crcinom, especilly ESCC, my cuse misleding nd decrese the specificity in primry hospitl. Our dt showed tht optiml cutoff vlue of severl common tumor mrkers resulted in high specificity of 0.80 nd medium sensitivity of 0.65 (Tble 5) for elevted tumor mrkers pnel predicting EGAC. Although mrkers pnel cn not identify the precise loction, the combintion of endoscope nd mrkers pnel my increse the sensitivity nd specificity of esophgel mlignncy predictions. Here, the limittions of this study hve to be mentioned. Limittion of cses my obstruct the representtiveness of the receiver groups becuse the cses merely come from single center. In ddition, our reserch pertins to retrospective study rther thn perspective one. It is promising tht prospective study could be designed to confirm the predictive vlue of biomrkers in clinics. Conclusions With the help of the ROC curve nlysis nd logistic regression nlysis, our results suggest tht the combined ppliction of CA72-4, CEA, SCC, nd CA19-9 could be vluble method to distinguish ptients with EGAC from those with ESCC. Further studies re recommended to confirm these results. Acknowledgements We thnk Liyn Xue (Deprtment of Clinicl Pthology, Cncer Hospitl, Chinese Acdemy of Medicl Sciences & Peking Union Medicl College) for review of pthologicl dt, Huiyo Hung (Ntionl Office for Cncer Control, Cncer Hospitl, Chinese Acdemy of Medicl Sciences & Peking Union Medicl College) for suggestion on sttisticl nlysis. This work ws supported by grnts from the Ntionl High-tech Reserch nd Development Progrm (NO. 2012AA02A503) nd Beijing Hope Mrthon specil project of Chinese Cncer Foundtion (NO. LC2012A05). Disclosure of conflict of interest None. Address correspondence to: Jun Qi, Deprtment of Clinicl Lbortory, Cncer Hospitl, Chinese Acdemy of Medicl Sciences & Peking Union Medicl College, No. 17 PnjiyunNnli, Choyng District, Beijing 100021, P. R. Chin. Tel: +86-10- 87788250; E-mil: 445293629@qq.com References [1] Hsegw S, Yoshikw T. Adenocrcinom of the esophgogstric junction: incidence, chrcteristics, nd tretment strtegies. Gstric Cncer 2010; 13: 63-73. [2] Zhou Y, Zhng Z, Zhng Z, Wu J, Ren D, Yn X, Wng Q, Wng Y, Wng H, Zhng J, Zhu X, Yng Y, Luo C, Guo X, Tng C, Qio L. A rising trend of gstric crdi cncer in Gnsu Province of Chin. Cncer Lett 2008; 269: 18-25. [3] Niu X, Wei WQ, Ho CQ, Song GH, Li J, Hu ZL, Li YW, Chng J, Wng XZ, Zho DL, Wng GQ, Hsieh E, Qio YL. Evlution of routine biopsies in endoscopic screening for esophgogstric junction cncer. World J Gstroenterol 2014; 20: 5074-81. [4] Slewert JR. Crcinom of the crdi: crcinom of the gstroesophgel juction clssifiction, pthology, nd extent of resection. Dis Esophgus 1996; 9: 173-182. 11863 Int J Clin Exp Med 2016;9(6):11859-11864
Serum mrker for esophgogstric junction denocrcinom [5] Rüdiger Siewert J, Feith M, Werner M, Stein HJ. Adenocrcinom of the esophgogstric junction: results of surgicl therpy bsed on ntomicl/topogrphic clssifiction in 1,002 consecutive ptients. Ann Surg 2000; 232: 353-61. [6] Orditur M, Glizi G, Lieto E, De Vit F, Cirdiello F. Tretment of esophgogstric junction crcinom: An unsolved debte. World J Gstroenterol 2015; 21: 4427-31. [7] Al-Hddd S, Chng AC, De Hertogh G, Grin A, Lnger R, Sgert X, Slemme M, Streutker CJ, Soucy G, Tripthi M, Upton MP, Vieth M, Villncci V. Adenocrcinom t the gstroesophgel junction. Ann N Y Acd Sci 2014; 1325: 211-25. [8] Wng RF, Song BR, Peng JJ, Ci GX, Liu FQ, Wng MH, Ci SJ, Ye X. The Prognostic Vlue of Preopertive Serum CEA nd CA19-9 Vlues in Stge I-III Colorectl Cncer. Heptogstroenterology 2014; 61: 994-9. [9] Grunnet M, Christensen IJ, Lssen U, Jensen LH, Lydolph M, Knox JJ, McNmr MG, Jitll M, Wsn H, Bridgewter J, Vlle JW, Mu-Sorensen M. Decline in CA19-9 during chemotherpy predicts survivl in four independent cohorts of ptients with inoperble bile duct cncer. Eur J Cncer 2015; 51: 1381-8. [10] Zhng Y, Yng J, Li H, Wu Y, Zhng H, Chen W. Tumor mrkers CA19-9, CA242 nd CEA in the dignosis of pncretic cncer: met-nlysis. Int J Clin Exp Med 2015; 8: 11683-91. [11] Tokung R, Immur Y, Nkmur K, Uchihr T, Ishimoto T, Nkgw S, Iwtsuki M, Bb Y, Skmoto Y, Miymoto Y, Yoshid N, Oym S, Shono T, Noe H, Seki H, Oki E, Wtnbe M, Sski Y, Mehr Y, Bb H. Crbohydrte ntigen 19-9 is useful prognostic mrker in esophgogstric junction denocrcinom. Cncer Med 2015; 4: 1659-66. [12] Zhng H, Wng W, Dio D, Cheng Y, Song Y, Zhu K, Dng C. Rtio of metsttic to exmined lymph nodes, helpful stging system nd independent prognostic fctor of esophgogstric junction cncer. PLoS One 2013; 8: e73238. [13] Uzunoglu FG, Reeh M, Kutup A, Izbicki JR. Surgery of esophgel cncer. Lngenbecks Arch Surg 2013; 398: 189-193. [14] Peng J, Wng WP, Yun Y, Hu Y, Wng Y, Chen LQ. Optiml Extent of Lymph Node Dissection for Siewert Type II Esophgogstric Junction Adenocrcinom. Ann Thorc Surg 2015; 100: 263-9. [15] Orditur M, Glizi G, Lieto E, De Vit F, Cirdiello F. Tretment of esophgogstric junction crcinom: Anunsolved debte. World J Gstroenterol 2015; 21: 4427-4431. [16] Grhm DY, Schwrtz JT, Cin GD, Gyorkey F. Prospective evlution of biopsy number in the dignosis of esophgel nd gstric crcinom. Gstroenterology 1982; 82: 228-31. [17] Ishimur N, Amno Y, Kinoshit Y. Endoscopic definition of esophgogstric junction for dignosis of Brrett s esophgus: importnce of systemtic eduction nd trining. Dig Endosc 2009; 21: 213-8. [18] Ege B, Dinç T, Yildiz BD, Blci Z, Bozky H. Utility of endoscopy for dignosis of brrett in non-western society: endoscopic nd histopthologic correltion. Int Surg 2015; 100: 720-5. 11864 Int J Clin Exp Med 2016;9(6):11859-11864