Conclusions
For muscle-invasive bladder cancer patients, the proliferative activity as determined by the nuclear staining of survivin or RT-qPCR on the basis of molecular subtype characteristics outperforms single marker detections and single technology approaches. Infiltration by macrophages detected by IHC or PCR is associated with worse outcome in defined subsets of tumors. The limitations of this study are the retrospective nature and the limited number of patients. However, the number of molecular markers has been restricted and based on predefined assumptions, which resulted in the dissection of muscle-invasive disease into tumor-biological axes of high prognostic relevance, which warrant further investigation and validation.
Methods
We used tissue microarrays (TMA) from n = 50 patients (38 males, 12 female) with muscle-invasive bladder cancer. All patients had been treated with radical cystectomy followed by adjuvant triple chemotherapy. Median follow-up time was 60.5 months. CD68, CLEVER-1, MAC387, and survivin protein were detected by immunostaining and subsequent visual inspection. BIRC5, KRT5, KRT20, ERBB2, and CD68 mRNAs were detected by standardized RT-qPCR after tissue dot RNA extraction using a novel stamp technology. All these markers were evaluated in three different centers of excellence.
Results
Nuclear staining rather than cytoplasmic staining of survivin predicted DSS as a single marker with high levels of survivin being associated with better PFS and DSS upon adjuvant chemotherapy (p = 0.0138 and p = 0.001, respectively). These results were validated by the quantitation of BIRC5 mRNA by PCR (p = 0.0004 and p = 0.0508, respectively). Interestingly, nuclear staining of survivin protein was positively associated with BIRC5 mRNA, while cytoplasmic staining was inversely related, indicating that the translocation of survivin protein into the nucleus occurred at a discrete, higher level of its mRNA. Combining survivin/BIRC5 levels based on molecular subtype being assessed by KRT20 expression improved the predictive value, with tumors having low survivin/BIRC5 and KRT20 mRNA levels having the best survival (75% vs. 20% vs. 10% 5-year DSS, p = 0.0005), and these values were independent of grading, node status, and tumor stage in multivariate analysis (p = 0.0167). Macrophage infiltration dominated in basal tumors and was inversely related with the luminal subtype marker gene expression. The presence of macrophages in survivin-positive or ERBB2-positive tumors was associated with worse DSS. Conclusions: For muscle-invasive bladder cancer patients, the proliferative activity as determined by the nuclear staining of survivin or RT-qPCR on the basis of molecular subtype characteristics outperforms single marker detections and single technology approaches. Infiltration by macrophages detected by IHC or PCR is associated with worse outcome in defined subsets of tumors. The limitations of this study are the retrospective nature and the limited number of patients. However, the number of molecular markers has been restricted and based on predefined assumptions, which resulted in the dissection of muscle-invasive disease into tumor-biological axes of high prognostic relevance, which warrant further investigation and validation.