The clinicopathological and prognostic signi ficance of PD-L1 expression in pancreatic cancer:A meta-analysis

Introduction

Pancreatic ductal adenocarcinoma(PDAC)is the fourth leading cause of cancer-related deaths worldwide[1].It is reported that 60%−80%of newly diagnosed PDAC patients cannot be cured by surgical intervention and the 5-year survival rate is less than 5%[2].Although systemic treatments have improved the overall prognosis of PDAC patients,chemotherapy tolerance is still the main problem[3,4].

Programmed cell death 1(PD-1)and one of its major ligands,programmed cell death ligand 1(PD-L1,also known as B7-H1),constitute a major tolerance mechanism in tumors[5].The binding of PD-L1 to PD-1 in T cells leads to T-cell apoptosis or exhaustion.PD-L1 expression induced by tumor cells also helps tumor cells escape from immune surveillance[6].High PD-L1 expression has been correlated with poor prognosis in several solid cancers,including melanoma[7],lung cancer[8],and renal clear cell cancer[9].The expression of the PD-1/PD-L1 axis has been previously characterized in PDAC[10].The development of PDAC is associated with alterations in in filtration of immunosuppressive cells,such as regulatory T cells(Tregs)and myeloid-derived suppressive cells(MDSCs),along with pro-cancerous in flammatory signals[11].

Despite some studies correlating PD-L1 expression with a poor prognosis in some cancers,information on the role of PD-L1 and its pathway in PDAC remains relatively limited.A number of recent publications are focused on the association between PD-L1 expression and the prognosis of PDAC patients.This meta-analysis aimed to clarify this relationship between PD-L1 positive status and the prognosis of PDAC patients.

Methods

Search strategy

We searched the PubMed,Cochrane Library and Web of Science electronic databases for articles published up to December 12,2016.The keywords used in our searches are “PD-L1”,“B7-H1”,“CD274”,“programmed cell death ligand 1 protein”and “pancreatic neoplasm”.The language of the studies was limited to English.The reference lists of the selected articles were searched to ensure that no studies were overlooked.

Selection criteria

The inclusion criteria were as follows:(1)all patients were diagnosed with PDAC by histopathology;(2)the study reported the PD-L1 expression levels in PDAC patients,whether or not the overall survival(OS)or disease-speci fic survival(DSS)were included;(3)the results were part of an original analysis;and(4)if the same patient population was used in several publications,only the most complete study was included.The exclusion criteria were as follows:(1)studies about the relationship between the co-expression of PD-L1 and other factors and PDAC prognosis and(2)studies with no data concerning hazard ratios(HR)or 95%con fidence intervals(CI)and without Kaplan–Meier curves to calculate these data.

Data extraction

Each article was reviewed independently by two authors.If there was a difference in extracted data,the article was then reviewed by the third author.The included articles were assessed by the methodology presented in“The Newcastle-Ottawa Scale for assessing the quality of non-randomized studies in meta-analyses”[12].The general data collected from each study included the first author’s name,country where study took place,publication year,patient number,disease stage and grade,methodology of PD-L1 analysis,and HR with 95%CI,which were used for measuring the effective value.If the data were not given directly,the available data were obtained from Kaplan–Meier curves(using Engauge Digitizer version 4.1)and the HR with 95%CI were calculated using the methods reported by Tierney et al.[13].

A flow chart of the study selection criteria is shown in Fig.1.The patient characteristics and study quality scores are listed in Table 1.Nine articles with 989 PDAC patients were evaluated.Among these articles,one had no direct data of PD-L1 positive rate[14].Three articles had no available data for prognostic calculation[15–17].Five studies with 688 PDAC patients were included in the prognostic meta-analysis[18–22].One study had the HR and 95%CI for the reported OS directly;Kaplan–Meier curves were used to estimate the HRs and 95%CIs for the remaining studies.

Statistical analysis

The expression of PD-L1 was de fined as “positive/up”or “negative/low”according to the immunohistochemistry(IHC)cut-off or polymerase chain reaction(PCR)values used in each study.The end-point was OS.The association between PD-L1 and the clinical outcomes was evaluated using the HR of patients with high PDL1 expression level compared to patients with low PD-L1 expression level and the 95%CI.When we utilized the HRs of positive/up versus negative/low levels of PD-L1 expression,an HR＞1 and 95%CI that did not overlap with 1 implied a good prognosis for patients with a negative/low PD-L1 level,while an HR＜1 and 95%CI that did not overlap with 1 signi fied a poor prognosis.An HR of 1 indicates a lack of association between PD-L1 expression levels and the clinical prognosis.All of the data were synthesized by Review Manager(version 5.2).The Mantel–Haenszel test was used to test signi ficance,and a P value＜0.05 was considered statistically signi ficant.Publication bias was not tested as the study number was limited,heterogeneity was tested by performing Chi-square tests,and by calculating I2 values.If P＜0.1 or I2＞50%,this indicated that there was signi ficant heterogeneity across the studies,and a random-effects model was used for calculating the pooled estimate;otherwise,the fixed-effects model was used.

在预防方面，大家都会结合养殖场实际情况或多或少使用消毒药进行杀毒灭菌，切断传播途径，保护养殖畜禽。特别是常见疫病高发季节，流行地区，新病流行阶段（如非洲猪瘟），日常消毒工作更加凸显重要性。但是大家对消毒药的使用真的用对了吗？下面给大家介绍一下常用消毒药的性质及正确使用方法。

Results

Selected studies and study characteristics

脑型脂肪栓塞综合征可表现为头痛、烦躁不安、失眠、易怒、谵妄甚至昏迷、复视、去大脑强直、偏瘫，也可伴有呕吐、尿失禁及自主神经功能紊乱，常早期出现病理反射，脑脂肪栓塞呈弥漫性时较少出现定位体征［2,3,5］。累及中枢神经系统的FES病例常伴有瘀点性皮疹、发热、心动过速及低血压，但肺动脉高压的症状或体征反而较少。FES可能成为术后患者出现中枢神经系统症状（如意识障碍等）的原因之一。

PD-L1 expression in PDAC patients

The details of the PD-L1 expression analyses are shown in Table 1.Seven articles detected PD-L1 expression using an IHC method[14–18,20,21];three detected mRNA expression via PCR[15,19,22].PD-L1 expression levels by IHC were classi fied as low or high based on the cut-off value of 10%of positive tumor cells.Birnbaum et al.[22]de fined “PD-L1 up”as a T/NP ratio(PDL1 expression in tumors/PD-L1 mean expression in normal pancreatic samples)≥2 and “not-up”(no up regulation)by a T/NP ratio＜2.Loos et al.[19]chose the median B7-H1 mRNA expression level as the cut-off point.A total of 247 patients had positive/up PD-L1 expression levels and 519 patients had negative/low PD-L1 expression levels.The PD-L1 positive rate was 32.2%with a range from 19%to 62.5%in the whole group.The average PD-L1 positive rate in the IHC subgroup is 54.5%with a range from 39.2%to 62.5%,which was higher than in the PCR method subgroup(P＜0.001).

PD-L1 as a prognostic factor for PDAC

Several studies have examined the correlation between PD-L1 expression and PDAC clinicopathological features,with incongruous results.Chen et al.[21]found that PD-L1 expression correlated with tumor size.Geng et al.[15]and Wang et al.[20]found that B7-H1 was correlated with the pathological grade and stage.While Nomi et al.[18]and Birnbaum et al.[22]did not find any correla-tion between PD-L1 expression and patient age or gender,or with pathological tumor size,lymph node status or grade.Although an unfavorable prognostic value for PD-L1 expression was reported in a few studies[20,21],our meta-analysis results showed that PDL1 expression may correlate with T stage because the positive PDL1 expression rate was higher in the T3-4 group than in the T1-2 group of PDAC patients.The relationship between PD-L1 expression and pathological grade and TNM stage requires further investigation.

2)场景裁切。大范围数字城市场景浏览过程中，许多用户往往只关心局部地区或者热点地区，这时直接裁切掉裁切区以外的数据，并直接返回裁切结果给场景，从而进一步提升场景显示的整体性能。

Correlation between PD-L1 expression and clinicopathological features

Four articles contained data about PD-L1 expression levels and clinicopathological factors including AJCC T stage,lymph node metastasis,distant metastasis and grade.Of the PDAC patients in stages T1–2,30.7%were PD-L1 positive and in stages T3-4,35.1%were PD-L1 positive.The combined odds ratio(OR)for the T1-2 group versus the T3-4 group was 0.37(95%CI:0.20–0.67;P=0.001;Fig.3).The pooled OR indicated no signi ficant correlations between PD-L1 expression levels and lymph node metastasis,distant metastasis and/or grade.

政府是创新体系的主导者而非独断者，因此需要不断推进体制改革，发挥组织者的优势，优化资源配置，充分利用社会资源，为政府公共服务的创新提供最有力的支持。

Fig.1.PRISMA flow chart of the selection process for identifying eligible studies.

Table 1 Patients’characteristics from the selected studies.

IHC:immunohistochemistry;NA:not available.

Studies Year Country Patients number PD-L1 detection IHC cut-off PD-L1 positive Hutcheson et al.[14] 2016 USA 223 Protein 10% NA Geng et al.[15] 2008 China 40 Protein&mRNA 10% 55%Song et al.[16] 2014 China 30 Protein 10% 62.5%Winograd et al.[17] 2015 USA 8 Protein NA 50%Nomi et al.[18] 2007 Japan 51 Protein 10% 39.2%Loos et al.[19]2008 Germany 40 mRNA–50%Wang et al.[20] 2010 China 81 Protein 5% 49.4%Chen et al.[21] 2014 China 63 Protein 10% 57.1%Birnbaum et al.[22]2016 France 453 mRNA–19%

Discussion

The clinical signi ficance of PD-1 and PD-L1 expression in solid cancers has been widely studied.Overexpression of PD-L1 has been noted in several types of human cancers[23].Blockade of the PD-1/PD-L1 pathway is a promising treatment in several tumors,such as esophageal cancer[24],breast cancer[25],and gastroenteric tumors[26].There are also some studies examining PD-L1 expression in PDAC.Some studies have indicated that positive/up PD-L1 expression is associated with a signi ficantly poorer OS[22];however,others did not con firm these findings[21].The present meta-analysis demonstrates a correlation between PD-L1 expression levels and the prognosis of PDAC patients.

In Shakespeare’s time，no women acted（表演）in plays.Men and boys played all the parts（部分）.

GHL and LL contributed to the literature searching.WWQ and WCT reviewed the searching results.GHL and ZSR wrote the manuscript.GHL,LL and XHX performed the statistical analyses.NQX and YXJ participated in study design and coordination and helped to draft the manuscript.All authors read and approved the final manuscript.GHL and LL contributed equally to this article.YXJ is the guarantor.

This meta-analysis examined data from 9 studies on tumor PDL1 expression in pancreatic cancer patients.Approximately one third of PDAC patients had high or up-regulated PD-L1 expression.The PD-L1 positive rate ranges from 19%to 62.5%.The PD-L1 positive rate measured by IHC method was higher than that by PCR.A positive relationship between PD-L1 protein and mRNA expression levels was reported in breast carcinoma[27].The differences found in this analysis may be partially due to different cut-off values in each group.We also analyzed the relationship between PD-L1 expression levels and the prognosis of PDAC patients.Our findings suggested that up-regulation of PD-L1 expression correlates to the poor prognosis of PDAC patients.These results are consistent with PD-L1 expression levels in several other cancers.

Fig.2.Forest plot showing the fixed-effects model of the hazard ratio(HR)estimates of PD-L1 expression levels and prognosis in PDAC patients.A:The whole cohort;B:IHC method cohort.

Fig.3.Forest plot showing the fixed-effects model of the odds ratio(OR)estimates of PD-L1 expression levels and T stage in PDAC patients.

Table 2 Studies reporting PD-L1 expression levels and outcomes in PDAC patients.

OS:overall survival;DSS:disease-speci fic survival;HR:hazard ratio;95%CI:95%con fidence interval.

Studies PD-L1 detection OS/DSS HR(95%CI)Nomi et al.[18] Protein 2.66(1.21–5.85)Loos et al.[19] mRNA 4.67(1.97–11.06)Wang et al.[20] Protein 2.08(1.17–3.72)Chen et al.[21] Protein 1.60(0.65–3.93)Birnbaum et al.[22] mRNA 2.22(1.48–3.33)

The details of the survival data are shown in Table 2.Based on the 5 articles with OS/DSS data,we evaluated the correlation between PD-L1 expression and prognosis among the 688 patients.The meta-analysis of PD-L1 expression levels and prognosis of PDAC patients were shown in Fig.2A.The PDAC patients with high PD-L1 expression levels had signi ficantly poorer OS than those with low PD-L1 expression levels(HR=2.34;95%CI:1.78–3.08).Furthermore,analysis of PDAC patients in the IHC method subgroup indicated that PD-L1 expression was associated with poor outcomes(Fig.2B)(HR=2.11;95%CI:1.39–3.19).Heterogeneity between reports was not signi ficant(P=0.48 and 0.71 respectively,I2=0%).

The results of this meta-analysis may lead to improvements in the outcomes of anti-PD-1/PD-L1 therapies.Several clinical trials have focused on using PD-L1 inhibitors in PDAC patients;however,the outcome is not promising.Unlike melanoma,PDAC is not commonly present with a robust in filtration of CD8+T cells in the tumor stroma[29].In genetically engineered mouse models of PDAC,a prominent immune-suppressive network is present even in the early stages of PDAC[30].Winograd et al.[17]demonstrated that induction of T-cell immunity overcomes the resistance toαPD-1 andαCTLA-4 blockade in PDAC to signi ficantly improve mice survival.PD-L1 expression is not an adaptive response to immune pressure,as PD-L1 expression in murine PDAC is neither dependent on T cells nor IFNγexpression[31].Thus,the combined regimen induces the establishment of antitumor immune memory,which may contain curative potential.

However,there are some limitations to our study.Firstly,the cut-off values for PD-L1 expression measured by the IHC method varied between studies,which may cause heterogeneity in the overall results.Most of the included articles used 10%of total tumor cells as the cut-off value when using the IHC method,but one study used only 5%as the cut-off value.We did not determine if there were prognostic differences when using different PD-L1 cutoff values.Secondly,race may affect prognostic values of PD-L1 in PDAC.The patients included in the studies for this meta-analysis were from Asia,Europe and America.Different races or different levels of medical development in different areas may also in fluence the outcome of PDAC patients,which may affect the results.Lastly,not all HRs and 95%CIs were directly extracted from the studies,as we had to extract some data using Kaplan–Meier curves,which may in fluence the precision of the data.

涉外的经济贸易合作在给国内企业带来机遇的同时，也加大了征管的难度。我国可以针对不同的国别设立相应的境外税收机构，该机构应该做好税收征管服务，提供税收情报等，加强税收的征管可靠性；根据不同国家的国情和最新的经济法律等制度与政策，修改相关协议与法律法规等规范性文件。

To our knowledge,this is the first meta-analysis focused on the relationship between PD-L1 status and PDAC prognosis.This study indicates that there is a signi ficant association between PD-L1 expression levels and the OS of PDAC patients.Our meta-analysis strongly supports that PDAC patients with high PD-L1 expression levels have poor prognosis.Additionally,PD-L1 expression strongly correlates with the T stage of PDAC.

Contributors

PD-L1 expression levels in cancers are typically detected via IHC.However,there are multiple non-standardized techniques and scoring systems that impede drawing conclusion[28].Two studies compared the PD-L1 expression level between a PCR method and an IHC method[19,22].Although the studies used different de finitions to determine the “up”and “not-up”PD-L1 groups,both studies made the same conclusion,that elevated PD-L1 mRNA expression correlated with poor outcomes for PDAC patients.In our meta-analysis,both method subgroups(IHC and PCR)showed the prognostic value of PD-L1 expression.

Funding

This work was supported by grants from the National Natural Science Foundation of China(81472670,81172005,81402397,81402398 and 81172276),the National Natural Science Foundation of Shanghai(14ZR1407600),the “Yang-Fan”Plan for Young Scientists of Shanghai(14YF1401100),and the PhD Programs Foundation of the Ministry of Education of China(20110071120096).

Ethical approval

Not needed.

Competing interest

No bene fits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

References

[1]Siegel RL,Miller KD,Jemal A.Cancer statistics,2016.CA Cancer J Clin 2016;66:7–30.

[2]Ma J,Siegel R,Jemal A.Pancreatic cancer death rates by race among US men and women,1970–2009.J Natl Cancer Inst 2013;105:1694–1700.

[3]Conroy T,Desseigne F,Ychou M,Bouché O,Guimbaud R,Bécouarn Y,et al.FOLFIRINOX versus gemcitabine for metastatic pancreatic cancer.N Engl J Med 2011;364:1817–1825.

[4]Jin SF,Fan ZK,Pan L,Jin LM.Gemcitabine-based combination therapy compared with gemcitabine alone for advanced pancreatic cancer:a meta-analysis of nine randomized controlled trials.Hepatobiliary Pancreat Dis Int 2017;16:236–244.

[5]Boussiotis VA.Molecular and biochemical aspects of the PD-1 checkpoint pathway.N Engl J Med 2016;375:1767–1778.

[6]Sanmamed MF,Chen L.Inducible expression of B7-H1(PD-L1)and its selective role in tumor site immune modulation.Cancer J 2014;20:256–261.

[7]Gadiot J,Hooijkaas AI,Kaiser AD,van Tinteren H,van Boven H,Blank C.Overall survival and PD-L1 expression in metastasized malignant melanoma.Cancer 2011;117:2192–2201.

[8]Konishi J,Yamazaki K,Azuma M,Kinoshita I,Dosaka-Akita H,Nishimura M.B7-H1 expression on non-small cell lung cancer cells and its relationship with tumor-in filtrating lymphocytes and their PD-1 expression.Clin Cancer Res 2004;10:5094–5100.

[9]Thompson RH,Dong H,Lohse CM,Leibovich BC,Blute ML,Cheville JC,et al.PD-1 is expressed by tumor-in filtrating immune cells and is associated with poor outcome for patients with renal cell carcinoma.Clin Cancer Res 2007;13:1757–1761.

[10]Seo YD,Pillarisetty VG.T-cell programming in pancreatic adenocarcinoma:a review.Cancer Gene Ther 2017;24:106–113.

[11]Komura T,Sakai Y,Harada K,Kawaguchi K,Takabatake H,Kitagawa H,et al.In flammatory features of pancreatic cancer highlighted by monocytes/macrophages and CD4+ T cells with clinical impact.Cancer Sci 2015;106:672–686.

[12]Wells GA,Shea BJ,O’Connell D,Peterson J,Welch V,Losos M,et al.The Newcastle–Ottawa Scale(NOS)for assessing the quality of nonrandomized studies in meta-analysis,http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp;2011.

[13]Tierney JF,Stewart LA,Ghersi D,Burdett S,Sydes MR.Practical methods for incorporating summary time-to-event data into meta-analysis.Trials 2007;8:16.

[14]Hutcheson J,Balaji U,Porembka MR,Wachsmann MB,McCue PA,Knudsen ES,et al.Immunologic and metabolic features of pancreatic ductal adenocarcinoma de fine prognostic subtypes of disease.Clin Cancer Res 2016;22:3606–3617.

[15]Geng L,Huang D,Liu J,Qian Y,Deng J,Li D,et al.B7-H1 up-regulated expression in human pancreatic carcinoma tissue associates with tumor progression.J Cancer Res Clin Oncol 2008;134:1021–1027.

[16]Song X,Liu J,Lu Y,Jin H,Huang D.Overexpression of B7-H1 correlates with malignant cell proliferation in pancreatic cancer.Oncol Rep 2014;31:1191–1198.

[17]Winograd R,Byrne KT,Evans RA,Odorizzi PM,Meyer AR,Bajor DL,et al.Induction of T-cell immunity overcomes complete resistance to PD-1 and CTLA-4 blockade and improves survival in pancreatic carcinoma.Cancer Immunol Res 2015;3:399–411.

[18]Nomi T,Sho M,Akahori T,Hamada K,Kubo A,Kanehiro H,et al.Clinical signi ficance and therapeutic potential of the programmed death-1 ligand/programmed death-1 pathway in human pancreatic cancer.Clin Cancer Res 2007;13:2151–2157.

[19]Loos M,Giese NA,Kleeff J,Giese T,Gaida MM,Bergmann F,et al.Clinical signi ficance and regulation of the costimulatory molecule B7-H1 in pancreatic cancer.Cancer Lett 2008;268:98–109.

[20]Wang L,Ma Q,Chen X,Guo K,Li J,Zhang M.Clinical signi ficance of B7-H1 and B7-1 expressions in pancreatic carcinoma.World J Surg 2010;34:1059–1065.

[21]Chen Y,Sun J,Zhao H,Zhu D,Zhi Q,Song S,et al.The coexpression and clinical signi ficance of costimulatory molecules B7-H1,B7-H3,and B7-H4 in human pancreatic cancer.Onco Targets Ther 2014;7:1465–1472.

[22]Birnbaum DJ,Finetti P,Lopresti A,Gilabert M,Poizat F,Turrini O,et al.Prognostic value of PDL1 expression in pancreatic cancer.Oncotarget 2016;7:71198–71210.

[23]Wang X,Teng F,Kong L,Yu J.PD-L1 expression in human cancers and its association with clinical outcomes.Onco Targets Ther 2016;9:5023–5039.

[24]Ohigashi Y,Sho M,Yamada Y,Tsurui Y,Hamada K,Ikeda N,et al.Clinical signi ficance of programmed death-1 ligand-1 and programmed death-1 ligand-2 expression in human esophageal cancer.Clin Cancer Res 2005;11:2947–2953.

[25]Bertucci F,Finetti P,Colpaert C,Mamessier E,Parizel M,Dirix L,et al.PDL1 expression in in flammatory breast cancer is frequent and predicts for the pathological response to chemotherapy.Oncotarget 2015;6:13506–13519.

[26]Bertucci F,Finetti P,Mamessier E,Pantaleo MA,Astol fiA,Ostrowski J,et al.PDL1 expression is an independent prognostic factor in localized GIST.Oncoimmunology 2015;4:e1002729.

[27]Schalper KA,Velcheti V,Carvajal D,Wimberly H,Brown J,Pusztai L,et al.In situ tumor PD-L1 mRNA expression is associated with increased TILs and better outcome in breast carcinomas.Clin Cancer Res 2014;20:2773–2782.

[28]Rimm D,Schalper K,Pusztai L.Unvalidated antibodies and misleading results.Breast Cancer Res Treat 2014;147:457–458.

[29]Fukunaga A,Miyamoto M,Cho Y,Murakami S,Kawarada Y,Oshikiri T,et al.CD8+tumor-in filtrating lymphocytes together with CD4+tumor-in filtrating lymphocytes and dendritic cells improve the prognosis of patients with pancreatic adenocarcinoma.Pancreas 2004;28:e26–e31.

[30]Beatty GL,Chiorean EG,Fishman MP,Saboury B,Teitelbaum UR,Sun W,et al.CD40 agonists alter tumor stroma and show efficacy against pancreatic carcinoma in mice and humans.Science 2011;331:1612–1616.

[31]Spranger S,Spaapen RM,Zha Y,Williams J,Meng Y,Ha TT,et al.Up-regulation of PD-L1,IDO,and T(regs)in the melanoma tumor microenvironment is driven by CD8(+)T cells.Sci Transl Med 2013;5.