Urothelial carcinoma of bladder is one of the most prevailing malignancies across the globe. It is comprised of both benign and malignant subtypes, namely non-muscle invasive bladder cancer (NMIBC) and muscle invasive bladder cancer (MIBC), which present distinct prognosis after diagnosis and treatment(45). Despite recent years’ great advances in tumor detection techniques, reliable prognostic biomarkers are still lacking. According to several clinical guidelines, UCB prognosis is mainly dependent on tumor numbers, prior recurrence rate, pathological T stage (depth of invasion), WHO grade, invasion of lymph nodes and remote metastasis to predict disease recurrence and survival(46, 47). UCB is in fact considered to be an “immunogenic” tumor since the famous Bacillus Calmette-Guerin (BCG) therapy, which harnesses immune responses, is well established and widely used in low risk NMIBC context(48-50). Thus the purpose of our study was to investigate the relationship between existing tumor-infiltrated lymphocytes, including the presence of TILs and different subsets, with the prognosis of UCB patients.
Firstly, we investigated the prognostic role of CD3+ T cells in UCB patients. Our meta-analysis showed that high CD3+ TILs were associated with a favorable prognosis in bladder cancer. This result is consistent with other studies which investigated CD3+ TILs in various tumors and ended up with similar outcomes, reflecting that general T cells in the tumor microenvironment are of great importance for clinical outcomes (51-53).
The prognostic value of CD8+ TILs has already been explored widely in most cancer types and many of them concluded that high CD8+ TILs indicated better outcomes(54, 55). Similarly, this meta-analysis suggested that high CD8+ T cell infiltration was significantly correlated with better overall survival. In fact, CD8+ TILs, functionally known as cytotoxic T lymphocytes, take the main adaptive immune responses against tumor by directly contacting and killing tumor cells(56). Thus, it could be easily understood that high tumor-infiltrated CD8+ markers could possibly serve as a favorable prognostic maker.
We failed to conduct a meta-analysis on the relationship between CD4+ TILs and the prognosis of UCB patients due to only one research on CD4+ TILs met our inclusion criteria. Notably, the one study on CD4+ lymphocytes indicated that high infiltration of CD4+ TILs were significantly associated with poor outcome (57). Nevertheless, the importance of CD4+ TILs should be noted. CD4+ T lymphocytes are traditionally known as T helper cells which assist cytotoxic immune cells to recognize and eliminate tumor cells(58). However, various studies have illuminated their two types of polarizations which exhibiting reciprocal antagonism(18, 20). Type 1 CD4+ T lymphocytes promote CD8+ T cell or NK cell-mediated tumor elimination by secreting cytokines such as IL-12, IFNγ and TNF, while type 2 CD4+ T lymphocytes promote tumor growth by secreting cytokines such as IL-4 and IL-5 (18, 20). This contradictory polarization could possibly explain the conflicting clinical outcomes in other malignancies. Studies of CD4+ TILs with clear discrimination of subpopulation on UCB patients are warranted.
The prognostic value of FoxP3+ TILs was also explored. Of five studies, two considering recurrence-free survival showed that high FoxP3+ TILs were associated with unfavorable clinical outcome. As for overall survival, progression-free survival and cancer-specific survival categories, there were no significant propensions whether high infiltration of FoxP3+ T lymphocytes were correlated with worse prognosis. This could be interpreted with two reasons. On the one hand, the included studies under FoxP3+ category investigated clinical outcomes with different time-to survival events, which resulted in small subgroups and attenuated statistical power. On the other hand, Tregs are also a conflicting subset of T lymphocytes which exhibit internal heterogeneity. In most cancers such as non-small -cell lung cancer and ovarian cancer, Tregs were associated with a poor prognosis, however in colorectal cancer, it can be correlated with a good prognosis(8, 59, 60). This contradiction might due to the combined effects of TIL subsets’ interaction in tumor immune microenvironment.
Finally, we conducted a meta-analysis on whole TILs’ prognostic role in UCB patients. In the five included studies, TILs were presented as a whole biomarker investigated by staining mononuclear inflammatory cells or combining CD3+ and CD8+ T lymphocytes, which means no subtypes of TILs were divided and analyzed respectively. However, our meta-analysis showed that TILs were not a representative biomarker concerning overall survival of UCB patients. This result might be due to the fact that TILs are consisted of various types of T cell subsets and the internal heterogeneity contributes to the statistical insignificance.
Certain limitations of this review must be noted. The main limitation of our meta-analysis is the heterogeneity regarding to tumor clinical and pathological characteristics among these included studies. The prognostic role of TILs may vary according to different tumor subsites and tumor stages. However, the relatively small sample sizes of included studies prevent us from stratification of different conditions for subgroup analysis. Furthermore, there is also heterogeneity in treatment modalities. UCB patients are generally divided into NMIBC or MIBC subtype according to their histopathologic analysis, which usually require distinct treatment plans. Although these included studies accounted for treatment modality in multivariate analysis, they still didn’t combine it to TILs with survival analysis. Notably, it is crucial to determine standardized and validated cut-offs for TIL quantification since clinical practice requires operable parameters. But this meta-analysis couldn’t suggest general cutoffs because the methods of quantifying TILs varied among the included studies. Thus in order to incorporate tumor-infiltrated T lymphocyte markers with clinical prognosis, more prospective studies using homogeneous patients, subgrouping tumor subsites, tumor stages and treatment modalities as well as adopting reliable cut-offs for quantification are needed.