Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Open Access

Peer-reviewed

Research Article

Prognostic value of PD-L1 expression in tumor infiltrating immune cells in cancers: A meta-analysis

Abstract

Programmed death-ligand 1 (PD-L1) is a promising target of cancer immune therapy. It not only expressed in tumor cells (TCs) but also up regulated in tumor infiltrating immune cells (TIICs). Although the previous meta-analysis have shown that PD-L1 expression in TCs was a valuable biomarker in predicting cancer prognosis, but few researches systematic evaluated the association between its expression in TIICs and survival of cancer patients. Thus, we performed this meta-analysis to evaluate the prognostic value of PD-L1 expression in TIICs in different types of cancers. Our results are valuable supplements when using PD-L1 expression to predict the survival of cancer patients and to select the beneficial patients from PD-L1 target therapy. PubMed, Embase, Web of Science and the Cochrane Central Search Library were used to perform our systematic literature search. Overall survival (OS) at 5th years and hazard ratios (HRs) were calculated using random effects models. Eighteen studies involving 3674 patients were included. The median positive rate of PD-L1 staining in TIICs was 36.37%. PD-L1 positive expression in TIICs related to a lower risk of death (HR = 0.784, 95%CI: 0.616–0.997, P = 0.047). Subgroup analyses found that PD-L1 positive expression in TIICs indicated a better prognosis especially in breast cancer patients (HR = 0.359, P = 0.041). When using whole tissue section slides, or using ‘any expression in TIICs’ as a cutoff value to assessing the results of IHC staining, PD-L1 expression in TIICs had a good prognostic value in cancer prognosis (HR = 0.587, P = 0.001 and HR = 0.549, P = 0.002). Our findings suggested that PD-L1 expression in TIICs was related to a better survival of cancer. The comprehensive evaluation of tumor cells and tumor infiltrating immune cells are required when evaluating the effect of PD-L1 expression on prognosis of cancer in future research.

Citation: Zhao T, Li C, Wu Y, Li B, Zhang B (2017) Prognostic value of PD-L1 expression in tumor infiltrating immune cells in cancers: A meta-analysis. PLoS ONE 12(4): e0176822. https://doi.org/10.1371/journal.pone.0176822

Editor: Tifei Yuan, Nanjing Normal University, CHINA

Received: February 18, 2017; Accepted: April 18, 2017; Published: April 28, 2017

Copyright: © 2017 Zhao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and its Supporting Information files.

Funding: This work was supported by the Finance Department of Jilin Province (No.sczsy201606) [http://czt.jl.gov.cn/jlcz/16/tindex.shtml].

Competing interests: The authors have declared that no competing interests exist.

Introduction

Cancer remains the most prominent global health-related threat[1, 2]. Traditional therapies such as tumorectomy, radiotherapy and chemotherapy are still the main treatments, but their efficacies are unsatisfactory in most cancers, especially in advanced cancers[3]. Recently, variable new cancer treatments have emerged, with immune checkpoint inhibition being one of the most promising therapies[4, 5].

Among the immune checkpoint molecules, programmed death 1 (PD-1) and its ligand, programmed death-ligand 1 (PD-L1), constitute a pair of negative co-stimulatory molecules that can suppress the functions of T cells and mediate the immune escape of cancers[6, 7]. PD-1 and PD-L1 inhibitors were developed by numerous pharmaceuticals companies and well studied in several clinical trials[8, 9]. A meta-analysis including 20 trials reported that patients with positive PD-L1 expression might have a decreased risk of mortality compared to negative cases when treated with anti PD-1/PD-L1 antibodies[10]. And the expression of PD-L1 not only linked to the response of immune checkpoint therapy but also associated with the prognosis of several types of cancer, such as non-small-cell lung cancer[11], gastric cancer[12], and breast cancer[13].

Although there has been already a lot of literatures published investigated the associations between PD-L1 expression and cancer prognosis using the method of meta-analysis[14]. However, all of them focused only on the PD-L1 expression in tumor cells. As we know, PD-1/PD-L1 pathway plays an important role in the cancer-specific immune response. PD-L1 is not only expressed in tumor cells but also up-regulated in tumor infiltrating immune cells (TIICs), including tumor infiltrating lymphocytes, mononuclear cells and other immune cells[15–17]. Current clinical trials have shown that the expression of PD-L1 in TIICs is also indicative of a higher response rate to PD-L1/PD-1 targeted therapy[18, 19]. Therefore, not only tumor cell-based but also immune cell-based PD-L1 expression appears to be clinically valuable. Recently, a number of studies have reported that the expression of PD-L1 in TIICs which was correlated with the survival of patients with tumors, but have failed to reach consistent conclusions [20–26]. In addition, there has been no research systematic evaluates the predicted value of PD-L1 positive expression in TIICs in cancer prognosis.

Thus, we performed this meta-analysis to evaluate the prognostic value of PD-L1 expression in TIICs in different types of cancers. Our results are valuable supplements when using PD-L1 expression to predict the survival of cancer patients and to select the beneficial patients from PD-L1 target therapy.

Materials and methods

Search strategy

PubMed, Embase, Web of Science and the Cochrane Central Search Library were used to perform our systematic literature search (until December 2016). Key words used included “programmed death-ligand 1 or PD-L1 or B7-H1 or CD274” and “tumor infiltrating lymphocyte or TIL or tumor infiltrating immune cells or TIIC or tumor infiltrating mononuclear cells or TIMC or tumor stroma” and “cancer or carcinoma or tumor” and “prognosis or survival”; the results were limited to human studies. In addition, we searched the reference lists of the reviews on related topics by hand to identify additional studies.

Inclusion and exclusion criteria

The eligible studies were included in this meta-analysis based on the following criteria: (1) PD-L1 expression has been measured by immunohistochemistry (IHC) stain in tumor infiltrating immune cells rather than in tumor cells; (2) studies reported 5-year OS, HR with 95% confidence interval (95% CIs), or reported original survival curves; (3) studies were published in English, and their full texts were available. Exclusion criteria for this study were as follows: (1) conference abstracts, letters, reviews and unpublished studies; and (2) insufficient data to report the hazard ratios and 95% CI, or could not extract the data from Kaplan-Meier curves. If duplicate data presented in more than one study, the largest or most recent study was included.

Data extraction and quality assessment

Two reviewers (Zhao TC and Wu YH) identified relevant articles independently. The details of these surveys included the author’s name, date of publication, type of cancer, type of pathological section, number of patients, tumor stage, age of patients, duration of follow up, antibody, staining location, and cut-off value in assessing the positive expression of PD-L1 in tumor infiltrating immune cells. Newcastle-Ottawa Scale (NOS) was used for quality assessment[27]. NOS scores no less than 6 were defined as high-quality studies.

Statistical analysis

Data were analyzed using STATA version 12.0 (STATA Corporation, College Station, TX, USA). The 3-year OS, 5-year OS, HR and 95%CI were extracted from the original studies, tables or recalculated from Kaplan—Meier curves using the program of Engauge Digitizer (http://sourceforge.net/projects/digitizer/), and applied using the Mantel—Haenszel random effect model. I2[28] was used to determine the heterogeneity of the selected studies. Subgroup analysis and sensitivity analysis were carrying out to assess the potential contributions of different type of tumors and different cutoff values for defining PD-L1 expression in TIICs. Egger’s regression asymmetry test[29] and Begg’s adjusted rank correlation[30] were performed to evaluate the potential publication bias. All tests were two-sided, and P<0.05 was considered statistically significant.

Results

Identification of eligible studies

The search results shown in Fig 1 identified 603 studies from the initial database. After careful manual selection and review of these articles, 18 studies with full text and available data according to the inclusion and exclusion criteria were included in the final analysis (Fig 1). And the PRISMA checklist was showed in S1 File.

thumbnail
Download:
Fig 1. Flow diagram of the literature search and study selection for the meta-analysis.

The flow diagram shows eligible publications at each stage of the analysis process. The database search was conducted in December 2016.

https://doi.org/10.1371/journal.pone.0176822.g001

Description of studies

The characteristics of the 18 studies summarized in Table 1. All of the studies assessed PD-L1 expression in tumor infiltrating immune cells using immunohistochemistry techniques. Seven studies evaluated cancers of the digestive system (5 gastric cancers, 1 esophageal squamous cell carcinoma and 1 colorectal cancer), two evaluated cancers of the urinary system (1 urothelial carcinoma and 1 renal cell carcinoma), two evaluated breast cancer, two evaluated lung cancer, and one each evaluated ovarian high grade serous carcinoma, extranodal NK/T-cell lymphoma, head and neck cancer, diffuse large B-cell lymphoma and spinal chordoma. In total, 3674 subjects were included in our meta-analysis.

thumbnail
Download:
Table 1. Characteristics of the studies included.

https://doi.org/10.1371/journal.pone.0176822.t001

Evaluation of PD-L1 expression in TIICs

The antibodies, cutoff values and staining locations used in the evaluation of PD-L1 expression in TIICs of the included studies are shown in Table 2. Clone E1L3N was used in four studies, and Clone SP142 was used in three studies. The cutoff values in assessing the positive expression of PD-L1 in TIICs were divided into 4 types: (1) proportion of stained cells greater than 5%, (2) proportion of stained cells greater than 1%, (3) any expression of PD-L1 in TIICs and (4) others. Most studies considered that the positive staining was located in the cell membrane (12 of 18 studies); whereas others thought both membranous and cytoplasmic staining could be considered as positive expression. The median positive rate of PD-L1 expression in TIICs was 36.37%.

thumbnail
Download:
Table 2. Detection of the PD-L1 expression in TIICs in the selected studies.

https://doi.org/10.1371/journal.pone.0176822.t002

PD-L1 expression in TIICs and five-year OS

Twelve studies reported data for 5-year OS. As shown in Fig 2, PD-L1 positive expression in TIICs seems to be associated with a better 5-year OS of cancer patients, though it did not reach statistical difference (OR = 0.778, 95%CI: 0.534–1.134, P = 0.192). Because of the significant heterogeneity among studies (I2 = 72%), subgroup analyses were conducted to assess whether the heterogeneity was due to different cancer types and cutoff values. Six studies provided the 5-year OS for digestive system cancers; others reported different types of cancer (S1 Fig). In the stratified analysis by cancer types, PD-L1 positive expression in TIICs of digestive system cancers was not associated with 5-year OS (OR = 0.862, 95%CI: 0.438–1.697, P = 0.667). Further, I2 was calculated to be 76.5%, which indicated that the heterogeneity was not due to different cancer types. We then conducted a subgroup analysis according to different cutoff values. When the cutoff value was defined as ‘proportion of stained cells greater than 5%’, PD-L1 expression in TIICs seems to be associated with a better cancer survival (OR = 0.662, 95% CI: 0.429–1.022, P = 0.062, I2 = 72.9%). When using the cutoff value of ‘proportion of stained cells greater than 1%’ to distinguish the positive and negative expression of PD-L1 in TIICs, the opposite trend has been reported (OR = 1.958, 95%CI: 0.987–1.134, P = 0.055, I2 = 0%; Fig 3).

thumbnail
Download:
Fig 2. Forest plot shows the associations between PD-L1 expression in TIICs and five year overall survival of cancer patients.

https://doi.org/10.1371/journal.pone.0176822.g002

thumbnail
Download:
Fig 3. Subgroup analysis by different cutoff values shows the associations between PD-L1 expression in TIICs and five year overall survival of cancer patients.

https://doi.org/10.1371/journal.pone.0176822.g003

PD-L1 expression in TIICs and time-to-event index

A total of 18 eligible studies were pooled to analyze the predictive value of TIICs expressed PD-L1 in cancer prognosis using HR and 95%CIs. Fig 4 shown that PD-L1 expression in TIICs indicated a decreased risk of death (HR = 0.784, 95%CI: 0.616–0.997, P = 0.047). Similar to 5-year OS, significant heterogeneity was noted (I2 = 67.7%, P<0.001). Exploratory subgroup analysis suggested that PD-L1 expression in TIICs indicted a lower risk of death in patients with breast cancer (HR = 0.359, 95%CI: 0.134–0.961, P = 0.041, I2 = 0%; S2 Fig). As shown in Fig 5, PD-L1 in TIICs was only associated with improved overall survival in those studies using cutoff value of ‘Any positive staining in immune cells’ (HR = 0.549, 95%CI = 0.378–0.798, P = 0.002, I2 = 0%). We also conducted a subgroup analysis according to different types of pathological sections (the whole tissue section slides or tissue microarrays). PD-L1 in TIICs was correlated to a favorable prognosis in those studies using whole tissue section slides in conducting the immunohistochemical stain (HR = 0.587, 95%CI: 0.425–0.810, P = 0.001, I2 = 11.9%, Fig 6). Additionally, genetic differences will contribute to the heterogeneity between individual studies. The ethnicities of included studies were divided into two parts; Asian and non-Asian. Positive expression of PD-L1 in TIICs was an indicator of a favorable prognosis, only in non-Asian cancer patients (HR = 0.709, 95%CI: 0.511–0.985, P = 0.040, I2 = 63.9%, Fig 7).

thumbnail
Download:
Fig 4. Forest plot of hazard ratios shows the associations between PD-L1 expression in TIICs and cancer prognosis.

https://doi.org/10.1371/journal.pone.0176822.g004

thumbnail
Download:
Fig 5. Forest plot of hazard ratios form subgroup analysis by different cutoff values shows the associations between PD-L1 expression in TIICs and cancer prognosis.

https://doi.org/10.1371/journal.pone.0176822.g005

thumbnail
Download:
Fig 6. Forest plot of hazard ratios form subgroup analysis by different types of pathological sections shows the associations between PD-L1 expression in TIICs and cancer prognosis.

TMAs, tissue microarrays.

https://doi.org/10.1371/journal.pone.0176822.g006

thumbnail
Download:
Fig 7. Forest plot of hazard ratios form subgroup analysis by different ethnicity shows the associations between PD-L1 expression in TIICs and cancer prognosis.

TMAs, tissue microarrays.

https://doi.org/10.1371/journal.pone.0176822.g007

Sensitivity and publication bias analyses

Omitting any individual study did not influence the combined results for 5-year OS or HR (S3 Fig). The funnel plot for the relationship between PD-L1 expression in TIICs and cancer prognosis is presenting in Fig 8. For 5-year OS the P values for Egger’s and Begg’s tests were 0.714 and 0.891. For hazard ratio, the results from Egger’s and Begg’s tests also revealed that there was no publication bias in this meta-analysis (P = 0.986 and P = 0.733, respectively).

thumbnail
Download:
Fig 8. Begg’s funnel plots show the publication bias.

(A) Begg’s funnel plot for 5-years OS (B) Begg’s funnel plot for HR.

https://doi.org/10.1371/journal.pone.0176822.g008

Discussion

Anti-PD-1/PD-L1 therapy has been discussed as a potential effective strategy for cancer treatment, and numerous studies have reported the positive expression of PD-L1 in tumor cells as a predictive biomarker for the response to PD-1/PD-L1 blocking therapy[42]. Simultaneously, the expression of PD-L1 has always been considered a predicted biomarker in cancer prognosis[14]. Not only tumor cells, but also tumor-infiltrating immune cells could express PD-L1. Previous studies have only focused on PD-L1 expression in tumor cells, but recent studies indicated that the PD-L1 expression in TIICs also played an important role in tumor immune escape and influenced tumor progression[16, 17, 31, 36]. These leading studies highlighted that PD-L1 expression in TIICs could also serve as a prognostic biomarker, and further inform the responses to anti-PD-1/PD-L1 treatment. In this meta-analysis of data from 18 studies, with a cohort of 3674 cancer patients, we firstly provided a quantitative estimate to the prognostic value of PD-L1 expression in TIICs in cancer patients.

Five-year overall survival and hazard ratio were both important indexes in cancer prognostic evaluation. Our results shown that, PD-L1 in TIICs was significantly associated with a decreased risk of death (HR = 0.784, 95%CI: 0.616–0.997, P = 0.047) compared with patients with PD-L1 negative expression in TIICs. A similar trend has observed when using 5-year OS to evaluate the predictive value of PD-L1 in TIICs in cancer prognosis, but the P value was not statistically significant. Our results were quite different from the published meta-analysis which shown that PD-L1 expression in tumor cells was associated with a worse prognosis of cancer. Different mechanisms between PD-L1 expression in TCs and TIICs might explain the inconsistent results. Transcriptome analyses indicated that, PD-L1 expression in TCs was up regulated through the tumor-intrinsic mechanisms, including the activation of endogenous oncogene and related signaling pathway[43]. However, PD-L1 expression in TIICs could be driven by adaptive mechanisms such like exogenous inflammation mediated immune attack and then reflected pre-existing immunity[44, 45]. In other words, comparing with tumor cells, the tumor infiltrating immune cells based PD-L1 expression has stronger relations with cancer immune response, and depends on tumor microenvironments. In fact, PD-L1 positive expression in TIICs was positive correlated to the quantity of multiple tumor-infiltrating immune cells, such as CD4+ T lymphocytes and CD8+T lymphocytes. Since the high expression of CD4+ T lymphocytes and CD8+ T lymphocytes was associated with better outcomes of cancer patients[43, 46], the PD-L1 expression in TIICs was possibly associated with better cancer prognosis. Although PD-L1 expression could mediate the occurrence of cancer immune escape, it also indicated an effective immune response, especially with a favorable profile of immune microenvironments in the early stage of the cancer immune response[40].

According to the results of subgroup analysis, different types of pathological sections and different definition of cutoff values when conducting and assessing the IHC stain could partly explain the large heterogeneity among individual studies. Half of the studies using tissue microarrays (TMAs) to conduct the IHC stain. TMAs usually contain limited tissue (2.0mm) and extract from the central part of tumors. Compare with the whole tissue section slides, TMAs may have less representativeness, especially in assessing the biomarkers expressed in tumor infiltrating immune cells. The subgroup analysis according to different types of pathological sections showed that when using whole section slides to investigate PD-L1 positive expression in TIICs, the heterogeneity among different studies was the lowest. It indicates that using whole section slides to conduct the IHC staining should be recommended in related clinical trials and treatments. Additionally, the appropriate cutoff value in validating the positive expression of PD-L1 remains contentious. Subgroup analysis with different cutoff values has shown that there was a contradictory trend when using the cutoff value of ‘5%’ or ‘1%’ in evaluating the correlations of PD-L1 positive expression in TIICs with survival of cancer patients. Therefore, a multi-classification of cutoff values for assessing PD-L1 expression in TIICs may be feasible and reasonable. Asian and non-Asian cancer patients exhibit distinct tumor immunity signatures. For instance, in gastric cancer, non-Asian patients show significantly higher expression level of T-cell markers, including CD3 and CD8, and lower expression level of immunosuppressive T-regulatory cell markers, such as FOXP3 compared to Asian gastric patients[47]. Immune-related biomarkers differentially expressed between Asian and non-Asian cancer patients who was related to immune function. These differences may affect the associations between PD-L1 expression and survival of cancer patients.

Several limitations should be acknowledged in our study. First, in several studies, when 5-year OS and HRs not provided in the original studies, we derived the indexes from Kaplan–Meier survival curves; six of the 18 studies did not provide 5-year OS or original Kaplan–Meier survival curves. As a result, only 12 studies were available to calculate the association of PD-L1 expression in TIICs with five-year OS, which could affect the level of evidence. Second, not all of the included studies using the multiple Cox regression to estimate the independent prognostic value of PD-L1 expression in TIICs in cancer. Thus, the results from all of the studies could not be further stratified with the same confounding factors, and further studies with more confounding factor adjustments need to be conducted.

Conclusions

Despite these limitations, we have demonstrated that PD-L1 expression in TIICs might serve as a new biomarker for prognosticating the survival of cancer patients. Thus, incorporating the expression of tumor-infiltrating immune cells into the classification of PD-L1 expression is necessary. Our results may be useful supplements when using PD-L1 expression to predict the survival of cancer patients and to select the beneficial patients from anti-PD-L1 treatment.

Supporting information

S1 Fig. Subgroup analysis by different types of cancers shows the associations between PD-L1 expression in TIICs and five year overall survival of cancer patients.

https://doi.org/10.1371/journal.pone.0176822.s001

(TIF)

S2 Fig. Forest plot of hazard ratios form subgroup analysis by types of cancers shows the associations between PD-L1 expression in TIICs and cancer prognosis.

https://doi.org/10.1371/journal.pone.0176822.s002

(TIF)

S3 Fig. Sensitivity analyses show the associations between PD-L1 expression in TIICs and cancer prognosis.

(A) Sensitivity analysis for 5-years OS (B) Sensitivity analysis for HR.

https://doi.org/10.1371/journal.pone.0176822.s003

(TIF)

S1 File. PRISMA 2009 checklist.

https://doi.org/10.1371/journal.pone.0176822.s004

(DOC)

S2 File. Full electronic search strategy in PUBMED.

https://doi.org/10.1371/journal.pone.0176822.s005

(DOCX)

Author Contributions

  1. Conceptualization: BJL BZ.
  2. Data curation: TCZ CFL YHW.
  3. Formal analysis: BZ.
  4. Funding acquisition: BZ.
  5. Investigation: TCZ.
  6. Methodology: TCZ CFL.
  7. Project administration: BZ.
  8. Resources: TCZ YHW.
  9. Software: TCZ.
  10. Supervision: BJL BZ.
  11. Validation: BJL BZ.
  12. Visualization: TCZ YHW.
  13. Writing – original draft: TCZ BJL BZ.
  14. Writing – review & editing: TCZ BZ.

References

  1. 1. Siegel RL, Miller KD, Jemal A. Cancer Statistics, 2017. CA: a cancer journal for clinicians. 2017;67(1):7–30. Epub 2017/01/06.
  2. 2. Chen W, Zheng R, Baade PD, Zhang S, Zeng H, Bray F, et al. Cancer statistics in China, 2015. CA Cancer J Clin. 2016;66(2):115–32. pmid:26808342
  3. 3. Luo D, Carter KA, Miranda D, Lovell JF. Chemophototherapy: An Emerging Treatment Option for Solid Tumors. Advanced science (Weinheim, Baden-Wurttemberg, Germany). 2017;4(1):1600106. Epub 2017/01/21.
  4. 4. Gibney GT, Weiner LM, Atkins MB. Predictive biomarkers for checkpoint inhibitor-based immunotherapy. The Lancet Oncology. 2016;17(12):e542–e51. Epub 2016/12/08. pmid:27924752
  5. 5. Teng F, Kong L, Meng X, Yang J, Yu J. Radiotherapy combined with immune checkpoint blockade immunotherapy: Achievements and challenges. Cancer letters. 2015;365(1):23–9. Epub 2015/05/20. pmid:25980820
  6. 6. Boussiotis VA. Molecular and Biochemical Aspects of the PD-1 Checkpoint Pathway. N Engl J Med. 2016;375(18):1767–78. Epub 2016/11/03. pmid:27806234
  7. 7. Sznol M, Chen L. Antagonist antibodies to PD-1 and B7-H1 (PD-L1) in the treatment of advanced human cancer. Clinical cancer research: an official journal of the American Association for Cancer Research. 2013;19(5):1021–34. Epub 2013/03/06. PubMed Central PMCID: PMCPMC3702373.
  8. 8. Costa R, Carneiro BA, Agulnik M, Rademaker AW, Pai SG, Villaflor VM, et al. Toxicity profile of approved anti-PD-1 monoclonal antibodies in solid tumors: a systematic review and meta-analysis of randomized clinical trials. Oncotarget. 2016. Epub 2016/11/17.
  9. 9. Zhang T, Xie J, Arai S, Wang L, Shi X, Shi N, et al. The efficacy and safety of anti-PD-1/PD-L1 antibodies for treatment of advanced or refractory cancers: a meta-analysis. Oncotarget. 2016;7(45):73068–79. Epub 2016/09/30. pmid:27683031
  10. 10. Gandini S, Massi D, Mandala M. PD-L1 expression in cancer patients receiving anti PD-1/PD-L1 antibodies: A systematic review and meta-analysis. Critical reviews in oncology/hematology. 2016;100:88–98. Epub 2016/02/21. pmid:26895815
  11. 11. Aguiar PN Jr., Santoro IL, Tadokoro H, de Lima Lopes G, Filardi BA, Oliveira P, et al. The role of PD-L1 expression as a predictive biomarker in advanced non-small-cell lung cancer: a network meta-analysis. Immunotherapy. 2016;8(4):479–88. Epub 2016/03/15. pmid:26973128
  12. 12. Liu YX, Wang XS, Wang YF, Hu XC, Yan JQ, Zhang YL, et al. Prognostic significance of PD-L1 expression in patients with gastric cancer in East Asia: a meta-analysis. OncoTargets and therapy. 2016;9:2649–54. Epub 2016/05/27. PubMed Central PMCID: PMCPMC4863684. pmid:27226727
  13. 13. Li X, Li M, Lian Z, Zhu H, Kong L, Wang P, et al. Prognostic Role of Programmed Death Ligand-1 Expression in Breast Cancer: A Systematic Review and Meta-Analysis. Targeted oncology. 2016;11(6):753–61. Epub 2016/07/17. pmid:27422273
  14. 14. Wu P, Wu D, Li L, Chai Y, Huang J. PD-L1 and Survival in Solid Tumors: A Meta-Analysis. PloS one. 2015;10(6):e0131403. Epub 2015/06/27. PubMed Central PMCID: PMCPMC4483169. pmid:26114883
  15. 15. Du Z, Abedalthagafi M, Aizer AA, McHenry AR, Sun HH, Bray MA, et al. Increased expression of the immune modulatory molecule PD-L1 (CD274) in anaplastic meningioma. Oncotarget. 2015;6(7):4704–16. Epub 2015/01/23. PubMed Central PMCID: PMCPMC4467109. pmid:25609200
  16. 16. Paiva B, Azpilikueta A, Puig N, Ocio EM, Sharma R, Oyajobi BO, et al. PD-L1/PD-1 presence in the tumor microenvironment and activity of PD-1 blockade in multiple myeloma. Leukemia. 2015;29(10):2110–3. Epub 2015/03/18. pmid:25778100
  17. 17. Koirala P, Roth ME, Gill J, Piperdi S, Chinai JM, Geller DS, et al. Immune infiltration and PD-L1 expression in the tumor microenvironment are prognostic in osteosarcoma. Sci Rep. 2016;6:30093. Epub 2016/07/28. PubMed Central PMCID: PMCPMC4960483. pmid:27456063
  18. 18. Powles T, Eder JP, Fine GD, Braiteh FS, Loriot Y, Cruz C, et al. MPDL3280A (anti-PD-L1) treatment leads to clinical activity in metastatic bladder cancer. Nature. 2014;515(7528):558–62. Epub 2014/11/28. pmid:25428503
  19. 19. Herbst RS, Soria JC, Kowanetz M, Fine GD, Hamid O, Gordon MS, et al. Predictive correlates of response to the anti-PD-L1 antibody MPDL3280A in cancer patients. Nature. 2014;515(7528):563–7. Epub 2014/11/28. PubMed Central PMCID: PMCPMC4836193. pmid:25428504
  20. 20. Bellmunt J, Mullane SA, Werner L, Fay AP, Callea M, Leow JJ, et al. Association of PD-L1 expression on tumor-infiltrating mononuclear cells and overall survival in patients with urothelial carcinoma. Annals of Oncology. 2015;26(4):812–7. pmid:25600565
  21. 21. Boger C, Behrens HM, Mathiak M, Kruger S, Kalthoff H, Rocken C. PD-L1 is an independent prognostic predictor in gastric cancer of Western patients. Oncotarget. 2016;7(17):24269–83. Epub 2016/03/25. PubMed Central PMCID: PMCPMC5029700. pmid:27009855
  22. 22. Choueiri TK, Fay AP, Gray KP, Callea M, Ho TH, Albiges L, et al. PD-L1 expression in nonclear-cell renal cell carcinoma. Annals of Oncology. 2014;25(11):2178–84. pmid:25193987
  23. 23. Darb-Esfahani S, Kunze CA, Kulbe H, Sehouli J, Wienert S, Lindner J, et al. Prognostic impact of programmed cell death-1 (PD-1) and PD-ligand 1 (PD-L1) expression in cancer cells and tumor-infiltrating lymphocytes in ovarian high grade serous carcinoma. Oncotarget. 2016;7(2):1486–99. Epub 2015/12/02. PubMed Central PMCID: PMCPMC4811475. pmid:26625204
  24. 24. Hatogai K, Kitano S, Fujii S, Kojima T, Daiko H, Nomura S, et al. Comprehensive immunohistochemical analysis of tumor microenvironment immune status in esophageal squamous cell carcinoma. Oncotarget. 2016;7(30):47252–64. Epub 2016/06/21. pmid:27322149
  25. 25. Hou J, Yu Z, Xiang R, Li C, Wang L, Chen S, et al. Correlation between infiltration of FOXP3+ regulatory T cells and expression of B7-H1 in the tumor tissues of gastric cancer. Experimental and Molecular Pathology. 2014;96(3):284–91. pmid:24657498
  26. 26. Jo J-C, Kim M, Choi Y, Kim H-J, Kim JE, Chae SW, et al. Expression of programmed cell death 1 and programmed cell death ligand 1 in extranodal NK/T-cell lymphoma, nasal type. Annals of Hematology. 2016;96(1):25–31. pmid:27696202
  27. 27. Newcastle–Ottawa quality assessment scale cohort studies 2013. Available from: http://www.ohri.ca/programs/clinical_epidemiology/nosgen.pdf.
    • 28. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ (Clinical research ed). 2003;327(7414):557–60. Epub 2003/09/06. PubMed Central PMCID: PMCPMC192859.
    • 29. Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ (Clinical research ed). 1997;315(7109):629–34. Epub 1997/10/06. PubMed Central PMCID: PMCPMC2127453.
    • 30. Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–101. Epub 1994/12/01. pmid:7786990
    • 31. Kawazoe A, Kuwata T, Kuboki Y, Shitara K, Nagatsuma AK, Aizawa M, et al. Clinicopathological features of programmed death ligand 1 expression with tumor-infiltrating lymphocyte, mismatch repair, and Epstein–Barr virus status in a large cohort of gastric cancer patients. Gastric Cancer. 2016.
    • 32. Kim HR, Ha S-J, Hong MH, Heo SJ, Koh YW, Choi EC, et al. PD-L1 expression on immune cells, but not on tumor cells, is a favorable prognostic factor for head and neck cancer patients. Scientific Reports. 2016;6:36956. pmid:27841362
    • 33. Kiyasu J, Miyoshi H, Hirata A, Arakawa F, Ichikawa A, Niino D, et al. Expression of programmed cell death ligand 1 is associated with poor overall survival in patients with diffuse large B-cell lymphoma. Blood. 2015;126(19):2193–201. Epub 2015/08/05. PubMed Central PMCID: PMCPMC4635115. pmid:26239088
    • 34. Li X, Wetherilt CS, Krishnamurti U, Yang J, Ma Y, Styblo TM, et al. Stromal PD-L1 Expression Is Associated With Better Disease-Free Survival in Triple-Negative Breast Cancer. American Journal of Clinical Pathology. 2016;146(4):496–502. pmid:27686176
    • 35. Paulsen E-E, Kilvaer TK, Khanehkenari MR, Al-Saad S, Hald SM, Andersen S, et al. Assessing PDL-1 and PD-1 in Non–Small Cell Lung Cancer: A Novel Immunoscore Approach. Clinical Lung Cancer. 2016.
    • 36. Saito R, Abe H, Kunita A, Yamashita H, Seto Y, Fukayama M. Overexpression and gene amplification of PD-L1 in cancer cells and PD-L1+ immune cells in Epstein–Barr virus-associated gastric cancer: the prognostic implications. Modern Pathology. 2016.
    • 37. Sun WY, Lee YK, Koo JS. Expression of PD-L1 in triple-negative breast cancer based on different immunohistochemical antibodies. Journal of Translational Medicine. 2016;14(1).
    • 38. Thompson ED, Zahurak M, Murphy A, Cornish T, Cuka N, Abdelfatah E, et al. Patterns of PD-L1 expression and CD8 T cell infiltration in gastric adenocarcinomas and associated immune stroma. Gut. 2016:gutjnl-2015-310839.
    • 39. Wang L, Ren F, Wang Q, Baldridge LA, Monn MF, Fisher KW, et al. Significance of Programmed Death Ligand 1 (PD-L1) Immunohistochemical Expression in Colorectal Cancer. Molecular Diagnosis & Therapy. 2016;20(2):175–81.
    • 40. Yang C-Y, Lin M-W, Chang Y-L, Wu C-T, Yang P-C. Programmed cell death-ligand 1 expression is associated with a favourable immune microenvironment and better overall survival in stage I pulmonary squamous cell carcinoma. European Journal of Cancer. 2016;57:91–103. pmid:26901614
    • 41. Zou MX, Peng AB, Lv GH, Wang XB, Li J, She XL, et al. Expression of programmed death-1 ligand (PD-L1) in tumor-infiltrating lymphocytes is associated with favorable spinal chordoma prognosis. American journal of translational research. 2016;8(7):3274–87. Epub 2016/08/11. PubMed Central PMCID: PMCPMC4969465. pmid:27508049
    • 42. Tie Y, Ma X, Zhu C, Mao Y, Shen K, Wei X, et al. Safety and efficacy of nivolumab in the treatment of cancers: A meta-analysis of 27 prospective clinical trials. International journal of cancer. 2017;140(4):948–58. Epub 2016/11/05. pmid:27813059
    • 43. Zeng DQ, Yu YF, Ou QY, Li XY, Zhong RZ, Xie CM, et al. Prognostic and predictive value of tumor-infiltrating lymphocytes for clinical therapeutic research in patients with non-small cell lung cancer. Oncotarget. 2016;7(12):13765–81. Epub 2016/02/13. PubMed Central PMCID: PMCPMC4924677. pmid:26871598
    • 44. Pardoll DM. The blockade of immune checkpoints in cancer immunotherapy. Nat Rev Cancer. 2012;12(4):252–64. Epub 2012/03/23. pmid:22437870
    • 45. Kim HR, Ha SJ, Hong MH, Heo SJ, Koh YW, Choi EC, et al. PD-L1 expression on immune cells, but not on tumor cells, is a favorable prognostic factor for head and neck cancer patients. Scientific reports. 2016;6:36956. Epub 2016/11/15. PubMed Central PMCID: PMCPMC5107906. pmid:27841362
    • 46. Hwang WT, Adams SF, Tahirovic E, Hagemann IS, Coukos G. Prognostic significance of tumor-infiltrating T cells in ovarian cancer: a meta-analysis. Gynecologic oncology. 2012;124(2):192–8. Epub 2011/11/02. PubMed Central PMCID: PMCPMC3298445. pmid:22040834
    • 47. Lin SJ, Gagnon-Bartsch JA, Tan IB, Earle S, Ruff L, Pettinger K, et al. Signatures of tumour immunity distinguish Asian and non-Asian gastric adenocarcinomas. Gut. 2015;64(11):1721–31. Epub 2014/11/12. PubMed Central PMCID: PMCPMC4680172. pmid:25385008