Triple‐marker immunohistochemical assessment of muscle‐invasive bladder cancer: Is there prognostic significance?

Abstract

Background Bladder cancer is the ninth most common cancer worldwide, and the third most common cancer in Lebanon. Immunohistochemistry (IHC) has been used to stratify muscle‐invasive bladder cancer (MIBC) into different subtypes. However, to our knowledge, there exists no study that investigates the use of this low‐cost technique to predict prognosis in bladder cancer patients in our region. Aim To examine the feasibility of low‐cost triple‐marker IHC assessment for MIBC subtyping in order to predict patients' survival and cisplatin sensitivity. Methods and results We collected the specimens of deceased patients diagnosed with MIBC on pathology at our institution. For each case, tumor tissue blocks were retrieved and stained for hematoxylin and eosin in addition to three molecular markers by IHC: cytokeratin 5/6, cytokeratin 14 staining basal BC, and GATA3 staining luminal BC. A cut‐off of ≥20% was set as positive. Kaplan‐Meier curves were built, factored by BC subtype, to predict overall survival (OS), disease‐specific survival (DSS), and progression‐free survival (PFS). Hazard ratios in Cox regression were also created accounting for oncological factors and BC subtype. We categorized specimens as either luminal (GATA3 positive only) (n = 21; 56.7%) or as double‐positive (GATA3 and basal cytokeratin 5/6 or cytokeratin 14 positive) (n = 16; 43.3%). The overall median survival was similar between the two categories (27.0 ± 4.82 months). Numbers favored luminal disease for PFS (Breslow P = .032). After adjusting for covariates, luminal molecular expression predicted PFS (0.28; [0.09‐0.94]). Yet, the Cox model was not able to identify any predictors of OS or DSS. Conclusion Specimens enriched with only a luminal molecular profile were more likely to exhibit cisplatin sensitivity. Despite the absence of guidelines recommending the utilization of molecular profiling in clinic practice, triple‐marker IHC could serve as a potential low‐cost prognostic indicator to identify patients at high risk of progression.

Clinicopathological Features and Survival Analysis in Molecular Subtypes of Muscle-Invasive Bladder Cancer

“…In our study, we did not observe consistent differences in terms of OS and CSS between the molecular subtypes obtained through surrogate IHC markers of any kind, in keeping with previous studies [ 16 , 24 , 29 , 30 , 31 , 32 , 33 ]. Serag-Eldien et al [ 34 ] applied a CK5/6/GATA3 classifier to their cohort of 80 BCs of any grade and stage and reported a trend of better OS and PFS for luminal MIBCs, although no statistical significance was achieved, in line with our results.…”

“…Conversely, in the study by Olkhov-Mitsel et al, basal (CK5/6+/GATA3-) MIBCs were independent predictors of worse disease-specific survival as compared to urothelial-like (CK5/6-/GATA3+/p16-) tumors ( p = 0.033) [ 31 ]. In keeping with this, specimens enriched with a luminal (CK5/6- and/or CK14-/GATA3+) phenotype showed higher PFS ( p = 0.032) in a cohort of MIBCs mostly treated with adjuvant cisplatin-based chemotherapy [ 30 ].…”

Are We Ready to Implement Molecular Subtyping of Bladder Cancer in Clinical Practice? Part 2: Subtypes and Divergent Differentiation

“…In a recent study, Labban et al classified their cohort of 64 MIBCs into Lum (GATA3+/CK5/6−/CK14−, 56.7%) and DP (GATA3+/CK5/6 and/or CK14+, 43.3%), with the former being significantly associated with a better PFS ( p = 0.039) [ 77 ]; the authors speculated that their DP tumors could be a surrogate marker for more aggressive biology, possibly due to an undisclosed P53-like signature [ 78 ]. Conversely, the CK5/6+/GATA3+ DP subtype identified in the MIBC cohort assessed by Bejrananda et al showed a significantly higher OS of 42.8% as compared to other subgroups [ 30 ].…”

“…In the study by Olkhov-Mitsel et al, SD-UCs yielded the highest levels of heterogeneity with 47.7% cases classified as Bas, 31.8% as Uro, and 20.5% as GU [ 26 ], in keeping with the findings from a previous study by Sjodahl et al, with most SD-UCs clustering into the SCCL (Bas) (44%) and UroB (25%) subtypes [ 22 ]. The only SD-UC case in the cohort analyzed by Labban et al expressed both luminal and basal markers [ 77 ].…”