Document Type : Review Article


1 Assistant Professor of Urology, Research Center for Non. Communicable Diseases, Jahrom University of Medical Sciences, Jahrom, Iran

2 Assistant Professor of Surgery, Research Center for Non.Communicable Diseases, Jahrom University of Medical Sciences, Jahrom, Iran

3 Associate Professor of Anesthesiology, Research Center for Non. Communicable Diseases, Jahrom University of Medical Sciences, Jahrom, Iran

4 Anesthesiology, Critical Care and Pain Management Research Center, Jahrom University of Medical Sciences, Jahrom, Iran


Introduction: Various molecular pathways are proposed to predict the prognosis of prostate cancer. However, due to previous achievements in prostate cancer management by immune checkpoint (IC) inhibitors, multiple studies have addressed the ICs role in the prognosis of prostate cancer and PD-L1 is known as one of the most cited ICs in this era.
Objective: To systematically review the prognostic value of the immune checkpoints (ICs) in prostate cancer.
Methods: This was a systematic review and meta-analysis study on online databases for studies reporting hazard ratios (HRs) of different survival outcomes of prostate cancer based on the dichotomized gene expression data of immune checkpoints. Studies presenting Kaplan Meier (KM) curves were included and time to event data was extracted for the KM reconstruction and HRs calculations for meta-analysis. Newcastle-Ottawa scale for cohort studies was used to assess the quality of studies.
Results: In the qualitative review, among the relevant selected ICs, 7 studies were available for PD-L1. Six studies were included in the meta-analysis for PD-L1. Concerning the high PD-L1 status as the reference, the chance of Biochemical Recurrence Free Survival was statistically lower in cases with low PD-L1 status (HR= 0.69, 95%CI of 0.57 to 0.58), under a fixed effect model with no heterogeneity (I2= 8%). There was a low possibility of publication bias based on the funnel plot.
Conclusion: High PD-L1 expression role in the prognosis of prostate cancer was shown to be supported by good levels of high-quality cohort studies, supported by pooled quantitative results of a meta-analysis.

Graphical Abstract

Prognostic Value of Immune Checkpoints in Prostate Cancer: A Systematic Review and Meta-Analysis of Reconstructed Time to Event Data


Main Subjects


Prostate cancer is the most common cancer in men [1], and despite the decrease in the incidence and death rate of prostate cancer in the United States and some other western countries [2], the incidence and death rate of this cancer is increasing in less developed and developing countries [3]. The rate of prostate cancer in Asian countries is much lower than the cases reported in the western population [3]. Prostate cancer is a disease related to the elderly [4] because the incidence and mortality rate of this disease is directly related to age [4]. Radical prostatectomy is a procedure in which the entire prostate gland and surrounding lymph nodes are removed to treat localized prostate cancer [5]. Prognosis and treatment options for prostate cancer depend on the cancer stage (Prostate-Specific Antigen (PSA) levels, Gleason score, tumor grade, and spread of cancer to nearby organs, and other parts of the body, age of the patient, and whether the cancer was newly diagnosed or diagnosed at advanced stages. Most men with prostate cancer do not die from prostate cancer [6]. In the context of genetic prognostic factors of prostate cancer, environmental suppressors of the tumor immune system are one of the distinct biological features that cause tumor growth and metastasis [7-8]. Currently, immunotherapy with checkpoint inhibitors has shown unprecedented clinical activity in a wide range of tumor types, like melanoma [9] and lung cancer [10]. Cells and tissues are constantly monitored by the immune system, and cancer has evolved in a way that can disrupt the immune checkpoints (ICs) and escape the host's immune system [11]. Sometimes cancer cells use immune checkpoint proteins as a shield to avoid being recognized and attacked by the immune system. Controlling molecules of the immune system including Cytotoxic T lymphocyte antigen-4 (CTLA-4), programmed cell death protein 1 (PD-1), and T-cell immunoglobulin mucin-3 (TIM-3), galectins are immune checkpoints that are proposed as therapeutic goals and prognostic factors in prostate cancer [12-14]. Considering the importance of ICs molecules such as PD-1/PD-L1, along with lesser-known ICs such as TIM3 and galectins, in suppressing immune responses in prostate cancer, prompted us to investigate the relationship between the intensity of the expression of the mentioned ICs with the prognosis of prostate cancer.


This was a systematic review (SR) and meta-analysis (MA) study based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines. PI (E) CO model was used to design the study question. Our study patients are prostate cancer patients. Exposure in our study was the expression or presence of inhibitory immune checkpoints genes. Based on our initial search, a list of inhibitory immune checkpoints was extracted; the most studied ICs include B7-H [15], BTLA [16], TIM-3, galectin 9, and PD-1/PD-L1. A comparison was made to examine the presence or absence of high or low expression of these genes among patients with prostate cancer, taking into account survival variables of biochemical recurrence (BCR) free survival or overall survival. There were no restrictions on the year of the publication.

The search strategy was a combination of selected ICs with keywords of “prostate cancer”, “prostatic neoplasms”, “prostate carcinoma”, and “immune checkpoints” in databases of PubMed, Scopus, Elsevier, Web of Science, and Google Scholar Search Engine. Two independent authors carried out the searches. As there were not enough studies for other ICs than the PD-L1, we excluded other ICs from the study. 

The inclusion condition in the present study was to investigate the protein expression of inhibitory ICs in a population of prostate cancer patients by separating patients with high and low expression of inhibitory ICs. Studies that presented Kaplan Meier (KM) curves were considered. RCTs reporting data of particular interventions were not included.

A piloted form was used to record the study information was filled by two independent reviewers. Newcastle-Ottawa score was used to evaluate the quality of studies. A pooled estimate of Hazard ratios (HRs) and the related 95 % confidence intervals (CIs) were estimated based on the reconstructed Kaplan Meier data. KM ‎data were pooled by extracting data using Engauge Digitizer software. Modena et al. proposed a method by using their spreadsheet of HR calculation based on the time to event data [17]. In the meta-analysis, for non-heterogeneous or heterogeneous data, a fixed-effects model (Peto technique) or a random-effects model (Mantel-Haenszel method) was utilized. The z-test was used to determine the overall impact, and the statistical significance threshold would be set to P< 0.05. ‎ I2 was used to address the amount of heterogeneity, where higher than 50% values were considered non-heterogenic. The possibility of publication bias was grossly assessed by a funnel plot.


The process of study selection is displayed in Figure 1. As two independent authors carried out the searches for PD-L1, 734 potentially relevant study titles were queried. As two independent authors carried out the queries, duplicated records were deleted. Ninety-four studies were evaluated based on the abstract. 39 studies were full text reviewed. The reason for the exclusion of studies is reported in Table 1. Eleven studies were included in SR (Table 1). Six studies were included in MA. Data were simulated based on the recurrence rate calculations and was descriptively reported. MA of PD-L1 low versus high expression was made based on the 7 studies.

In Li et al. study, the KM curves were stratified by PD-1 positive or negative and HR was calculated based on the reported univariate HR of 7.295 (2.981-17.852) high PD-L1 vs. low PD-L1 status, as listed in Table S2. We were not able to perform any Interrater agreement calculation between the reconstructed data and the reported HR as only comparisons were available in 3 studies. In Ness et al. study HR of 1.34, 95% CI of 0.97 to 1.85 was reported that matched our calculation of 1.32, 95% CI of 0.949 to 1.837. Finally in the MA, with no heterogeneity (I2= 8%), considering the high PD-L1 status as the reference, the BCR chance was statistically lower in cases with low PD-L1 status (HR= 0.69, 95%CI of 0.57 to 0.58), as indicated in Table 2. There was a low possibility of publication bias, based on the funnel plot, as displayed in Figure 2.

Possibility of bias

Newcastle-Ottawa scale for cohort study’s quality was used to evaluate the bias possibility in the included studies. The scales are summarized in Figure 3. A minimum score of 6 was approved to be included in the study. Most studies had an acceptable quality, while there were not adequate follow-ups performed in some studies.


Table 1: Characteristics of included studies






Mean follow-up


N1 Nodal status

ICS method of assessment


Gevensleben et al., 2016 [test cohort] [18]


Aggressive Primary Prostate Cancer


BCR-free survival

49.5 months


9 (1.5)

Cytometric  analysis, Pd-L1 cut-off: median expression of PD-L1

Sharma et al., 2019 [19]


Radical prostatectomy patient cohort


BCR-free survival

48.2 months

60.4 vs 58.8, in low and high PD-L1


Strong staining in at least 1% of cells.

Vicier et al., 2021 [20]


Radical prostatectomy patient cohort

Retrospective cohort

BCR-free survival

Metastasis free survival

8.1 years




Mo et al., 2019 [21]


Localized and castration-resistant prostate cancer


BCR-free survival

11.9 years



Immunohistochemistry in tumor-associated nerves*

Li et al., 2019 [22]


Received adjuvant hormonal therapy (AHT) after radical prostatectomy


BCR-free survival

40 months



Immunohistochemical PD-L1 assessment; Pd-L1 cut-off: median expression of PD-L0

Goltz et al., 2016 [23]

247 (only 86 entered MA)

Radical prostatectomy patient cohort


BCR-free survival

61.3 months


14 (5.7%)

Immunohistochemical PD-L1 assessment; Pd-L1 cut-off: median expression of PD-L1

CXCL12 gene was also evaluated a cohort of low CXCL12 was entered this MA to decrease the confounding effect.

Ness et al., 2017[24]


Radical prostatectomy patient cohort


BCR-free survival

150 months

357 >65 years old


Immunohistochemistry, dichotomization based on the staining extent

*There was a confounding effect of other evaluated genes.

Table 2: Comparison of the reconstructed time to event data and main data



Reported in the main text


Lower 95% CI

Upper 95% CI


Lower 95% CI

Upper 95% CI

Gevensleben et al., 2016 [test cohort] [18]





Sharma et al., 2019 [19]





Vicier et al., 2021 [20]







Mo et al., 2019 [21]







Li et al., 2019 [4]







Goltz et al., 2016 [23]





Ness et al., 2017 [24]








Table 3: MA of log(HR) for PD-L1 status effect on BCR-free survival

Study ID

Log (HR)

Standard Error




95% CI



Goltz et al., 2016 [23]







Vicier et al., 2021 [20]







Li et al., 2019







Sharma et al., 2019 [19]







Ness et al., 2017 [24]







Mo et al., 2019 [21]







Gevensleben et al., 2016 [18]














Figure 1: Funnel plot of MA for PD-L1 status effect on BCR-free survival

Figure 2: Newcastle-Ottawa scale of cohort study quality


Figure 3: PRISMA flow chart of study



Our review showed that while most studies were not individually conclusive for the prognostic role of the high PD-L1 expression in BCR-free survival of prostate cancer, pooled results revealed significant findings in line with poor prognosis of high PD-L1 status patients with a high chance of recurrence.

The previous similar MA study included only four studies [25], but the results were similar and found PD-L1 expression (HR = 1.78; 95% CI 1.39 to 2.27; p < 0.00001) associated with poor BCR free survival. While this statement of their MA was not true as they only had included 2 cohorts of the same study by Gevensleben et al. actually, they only represented the study results of Gevensleben et al.

PD-1 and its ligand PD-L1 are part of the inhibitory pathways that act to maintain environmental tolerance by limiting the lytic activity of immune cells [30]. As soon as the T cell receptor is activated, it is expressed on its surface and its connection to PD-L1 and PD-L2 ligands reduces the messages sent from the T cell receptor, and finally suppresses the immune system by inducing apoptosis in T cells [26]. It is notable that the PD-1/PD-L1 pathway is an important factor related to tumor-induced immunosuppression [27]. Several shreds of evidence point to the failure and exhaustion of T cells by the PD-1/PD-L1 pathway. Among this evidence is that PD-1 is expressed on the surface of TCD8+ cells infiltrating many solid tumors and antigen-specific CD8+ T cells in non-solid tumors [28]. Dysfunction of PD1+ T cells and high expression of PD-L1 has been observed in many cancers, which are associated with poor disease prognosis [29]. Disrupting the PD-1/PD-L1 pathway, either through a blocking antibody or through creating a gene defect in PD-1, leads to better clinical results and recovery of T cell function [30].

Our study was not conclusive for TIM-3 as we were not able to data extraction for MA. However, the results obtained from a meta-analysis study conducted by Yang Zhang et al. in 2016 showed that the high expression of TIM3 in patients with solid tumors decreased the overall survival of the patients is associated and the increased expression of TIM3 is also related to the advanced stages of the tumor [31].

PD-L1 roles have also been evaluated in meta-analysis studies for other cancers that we would mention for comparison of the meta-analysis methods and research methods. Li et al. investigated its role in colorectal cancer and showed that it is a poor prognostic marker of colorectal cancer [32]. Wang et al. conducted a similar analysis on solid tumors and included 61 studies in their review as distinct types of the cancers evaluated in their study. They had a similar approach to extracting HRs from the KM curves [36].

Limitations and suggestions

We were not able to generate evidence for other ICs as there were not enough studies. Likewise, in the case of PD-1, the same problem exists and still many studies are needed for any pooled analysis of HRs. Another limitation of this study was that some studies had evaluated the PD-L1 gene expression along with some other genes that we tried to use cohorts of patients with the lowest expression of secondary evaluated genes to decrease the chance of heterogeneity.

In the study of Li et al. [35] in addition to the pooled HR analysis, they compared the rate of PD-L1 positivity in different stages of the tumor and different pathological stratifications. In further analyses, we suggest considering this issue for PDl-L1. Furthermore, Wang et al. analyzed alternative survival indicators, the overall survival (OS), and disease-free survival (DFS) that we suggest such analysis for further meta-analysis study, while there were not enough reported OSs to be analyzed in our study.

Re-analyzing the KM data might be associated with some sources of human error as it was extracted by hand-drawn paths based on the original KM. As mentioned in supplementary Table S2, in some circumstances, the difference between re-calculated HR and published one was so significant that further studies should aid decreasing this error of calculations.


Our review showed that while most studies were not individually conclusive for the prognostic role of the high PD-L1 expression in BCR-free survival of prostate cancer, pooled results revealed significant findings in line with poor prognosis of high PD-L1 status patients with a high chance of recurrence.


We would like to thank the Clinical Research Development Unit of Peymanieh Educational and Research and Therapeutic Center of Jahrom University of Medical Sciences for providing facilities to this work.


This research did not receive any specific grant from fundig agencies in the public, commercial, or not-for-profit sectors.

Authors' contributions

All authors contributed to data analysis, drafting, and revising of the paper and agreed to be responsible for all the aspects of this work.

Conflict of Interest

There are no conflicts of interest in this study.



Hossein Hakimelahi

Mohammad Sadegh Sanie Jahromi


Hossein Hakimelahi, Alireza Yousefi, Marzieh Haghbeen, Mohammad Sadegh Sanie Jahromi, Hasan Zabetian, Reza Inaloo, Prognostic Value of Immune Checkpoints in Prostate Cancer: A Systematic Review and Meta-Analysis of Reconstructed Time to Event Data. J. Med. Chem. Sci., 2023, 6(1) 142-151   


[1]. Pernar C.H., Ebot E.M., Wilson K.M., Mucci L.A., The epidemiology of prostate cancer, Cold Spring Harbor perspectives in medicine, 2018, 8:a030361 [Crossref], [Google Scholar]
[2]. Culp M.B., Soerjomataram I., Efstathiou J.A., Bray F., Jemal A., Recent global patterns in prostate cancer incidence and mortality rates, European urology, 2020, 77:38 [Crossref], [Google Scholar], [Publisher]
[3]. Kimura T., Egawa S., Epidemiology of prostate cancer in Asian countries, International Journal of Urology, 2018, 25:524 [Crossref], [Google Scholar], [Publisher]
[4]. Liu X., Yu C., Bi Y., Zhang Z.J., Trends and age-period-cohort effect on incidence and mortality of prostate cancer from 1990 to 2017 in China, Public health, 2019, 172:70 [Crossref], [Google Scholar], [Publisher]
[5]. Costello A.J., Considering the role of radical prostatectomy in 21st century prostate cancer care, Nature Reviews Urology, 2020, 17:177 [Crossref], [Google Scholar], [Publisher]
[6]. Mcneal J., Kindrachuk R., Freiha F., Bostwick D., Redwine E., Stamey T., Patterns of progression in prostate cancer, The Lancet, 1986, 327:60 [Crossref], [Google Scholar], [Publisher]
[7]. Drake C.G., Prostate cancer as a model for tumour immunotherapy, Nature Reviews Immunology, 2010, 10:580 [Crossref], [Google Scholar], [Publisher]
[8]. Janiczek, Ł. Szylberg, A. Kasperska, A. Kowalewski, M. Parol, P. Antosik, Radecka B., Marszałek A., Immunotherapy as a promising treatment for prostate cancer: a systematic review, Journal of immunology research, 2017, 2017:4861570 [Crossref], [Google Scholar], [Publisher]
[9]. Lugowska I., Teterycz P., Rutkowski P., Immunotherapy of melanoma, Contemporary Oncology/Współczesna Onkologia, 2018, 2018:61 [Crossref], [Google Scholar], [Publisher]
[10]. Raez L.E., Fein S., Podack E.R., Lung cancer immunotherapy, Clinical medicine & research, 2005, 3:221 [Google Scholar], [Publisher]
[11]. Dyck L., Mills K.H., Immune checkpoints and their inhibition in cancer and infectious diseases, European journal of immunology, 2017, 47:765 [Crossref], [Google Scholar], [Publisher]
[12]. Topalian S.L., Targeting immune checkpoints in cancer therapy, Jama, 2017, 318:1647 [Crossref], [Google Scholar], [Publisher]
[13]. Venkatachalam S., McFarland T.R., Agarwal N., Swami U., Immune checkpoint inhibitors in prostate cancer., Cancers, 2021, 13:2187 [Crossref], [Google Scholar], [Publisher]
[14]. Modena A., Ciccarese C., Iacovelli R., Brunelli M., Montironi R., Fiorentino M., Tortora G., Massari F., Immune checkpoint inhibitors and prostate cancer: a new frontier?, Oncology reviews, 2016, 10:293 [Crossref], [Google Scholar], [Publisher]
[15]. Flem-Karlsen K., Fodstad Ø., Nunes-Xavier C.E., B7-H3 immune checkpoint protein in human cancer, Current medicinal chemistry, 2020, 27:4062 [Crossref], [Google Scholar], [Publisher]
[16]. Leone R.D., Lo Y.C., Powell J.D., A2aR antagonists: Next generation checkpoint blockade for cancer immunotherapy, Computational and structural biotechnology journal, 2015, 13:265 [Crossref], [Google Scholar], [Publisher]
[17]. Tierney J.F., Stewart L.A., Ghersi D., Burdett S., Sydes M.R., Practical methods for incorporating summary time-to-event data into meta-analysis, Trials, 2007, 8:1 [Crossref], [Google Scholar], [Publisher]
[18]. Gevensleben H., Dietrich D., Golletz C., Steiner S., Jung M., Thiesler T., Majores M., Stein J., Uhl B., Müller S., Ellinger J., Canc. Res., The Immune Checkpoint Regulator PD-L1 Is Highly Expressed in Aggressive Primary Prostate CancerPD-L1 Is Highly Expressed in Primary Prostate Cancer, Clinical Cancer Research, 2016, 22:1969 [Crossref], [Google Scholar], [Publisher]
[19]. Sharma M., Yang Z., Miyamoto H., Immunohistochemistry of immune checkpoint markers PD-1 and PD-L1 in prostate cancer, Medicine, 2019, 98:e17257 [Crossref], [Google Scholar], [Publisher]
[20]. Vicier C., Ravi P., Kwak L., Werner L., Huang Y., Evan C., Loda M., Hamid A.A., Sweeney C.J., Association between CD8 and PD‐L1 expression and outcomes after radical prostatectomy for localized prostate cancer, The Prostate, 2021, 81:50 [Crossref], [Google Scholar], [Publisher]
[21]. ‎ Mo R.J., Han Z.D., Liang Y.K., Ye J.H., Wu S.L., Lin S.X., Zhang Y.Q., Song S.D., Jiang F.N., Zhong W.D., ‎Wu C.L., Expression of PD‐L1 in tumor‐associated nerves correlates with reduced CD8+ tumor‐associated lymphocytes and poor prognosis in prostate cancer, International Journal of Cancer, 2019, 144:3099 [Crossref], [Google Scholar], [Publisher]
[22]. Li H., Wang Z., Zhang Y., Sun G., Ding B., Yan L., Liu H., Guan W., Hu Z., Wang S., The immune checkpoint regulator PDL1 is an independent prognostic biomarker for biochemical recurrence in prostate cancer patients following adjuvant hormonal therapy, Journal of Cancer, 2019, 10:3102 [Crossref], [Google Scholar], [Publisher]
[23]. ‎ Goltz, EE. Holmes, H. Gevensleben, V. Sailer, J. Dietrich, M. Jung, M. Röhler, S. Meller, J. ‎Ellinger, G. Kristiansen, D. Dietrich, CXCL12 promoter methylation and PD-L1 expression as prognostic biomarkers in prostate cancer patients, Oncotarget, 2016, 7:53309 [Crossref], [Google Scholar], [Publisher]
[24]. Ness N., Andersen S., Khanehkenari M.R., Nordbakken C.V., Valkov A., Paulsen E.E., Nordby Y., Bremnes R.M., Donnem T., Busund L.T., Richardsen, E., The prognostic role of immune checkpoint markers programmed cell death protein 1 (PD-1) and programmed death ligand 1 (PD-L1) in a large, multicenter prostate cancer cohort, Oncotarget, 2017, 8:26789 [Crossref], [Google Scholar], [Publisher]
[25]. Li Y., Huang Q., Zhou Y., He M., Chen J., Gao Y., Wang X., The clinicopathologic and prognostic significance of programmed cell death ligand 1 (PD-L1) expression in patients with prostate cancer: a systematic review and meta-analysis, Frontiers in pharmacology, 2019, 9:1494 [Crossref], [Google Scholar], [Publisher]
[26]. Zak K.M., Grudnik P., Magiera K., Dömling A., Dubin G., Holak T.A., Structural biology of the immune checkpoint receptor PD-1 and its ligands PD-L1/PD-L2, Structure, 2017, 25:1163 [Crossref], [Google Scholar], [Publisher]
[27]. Jiang X., Wang J., Deng X., Xiong F., Ge J., Xiang B., Wu X., Ma J., Zhou M., Li X., Li Y.,  Role of the tumor microenvironment in PD-L1/PD-1-mediated tumor immune escape, Molecular cancer, 2019, 18:1 [Google Scholar]
[28]. Nowicki T.S., Akiyama R., Huang R.R., Shintaku I.P., Wang X., Tumeh P.C., Singh A., Chmielowski B., Denny C., Federman N., Ribas A., Infiltration of CD8 T Cells and Expression of PD-1 and PD-L1 in Synovial SarcomaCD8, PD-1, and PD-L1 in Synovial Sarcoma, Cancer immunology research, 2017, 5:118 [Crossref], [Google Scholar], [Publisher]
[29]. Sakuishi K., Apetoh L., Sullivan J.M., Blazar B.R., Kuchroo V.K., Anderson A.C., Targeting Tim-3 and PD-1 pathways to reverse T cell exhaustion and restore anti-tumor immunity, Journal of Experimental Medicine, 2010, 207:2187 [Crossref], [Google Scholar], [Publisher]
[30]. Arasanz H., Gato-Cañas M., Zuazo M., Ibañez-Vea M., Breckpot K., Kochan G., Escors D., PD1 signal transduction pathways in T cells, Oncotarget, 2017, 8:51936 [Crossref], [Google Scholar], [Publisher]
[31]. Zhang Y., Cai P., Liang T., Wang L., Hu L., TIM-3 is a potential prognostic marker for patients with solid tumors: A systematic review and meta-analysis, Oncotarget, 2017, 8:31705 [Crossref], [Google Scholar], [Publisher]
[32]. Li Y., He M., Zhou Y., Yang C., Wei S., Bian X., Christopher O., Xie L., The prognostic and clinicopathological roles of PD-L1 expression in colorectal cancer: a systematic review and meta-analysis, Frontiers in pharmacology, 2019, 10:139 [Crossref], [Google Scholar], [Publisher]
[33]. Wang Q., Liu F., Liu L., Prognostic significance of PD-L1 in solid tumor: an updated meta-analysis, Medcine, 2017, 96:e6369 [Crossref], [Google Scholar], [Publisher]