Tissue microarray based analysis of prognostic markers in invasive bladder cancer: Much effort to no avail?

Abstract

Purpose

To evaluate altered protein expression with tissue microarray methodology for 15 different markers with potential prognostic significance in invasive bladder cancer.

Materials and Methods

Invasive tumor was sampled with the tissue-arraying instrument in 133 consecutive patients who underwent radical cystectomy, and at least 3, 0.6-mm tissue cores were obtained. With immunohistochemistry, the expressions of TP53, RB1, CDKN1A (p21), MKI67 (Ki67), PTGS2 (Cox-2), CTNNA1 (α-catenin), CTNNB1 (β-catenin), AKT, PTEN, RHOA, RHOC, STAT1, VEGFC, EGFR, and ERBB2 (HER2) were quantified, and correlations were made with tumor grade, pathologic stage, lymph node status, and disease-specific survival.

Results

Decreased immunohistochemical expression of CTNNA1 and of PTEN correlated with higher pathologic tumor stages (P = 0.01 and P = 0.01, respectively), whereas increased AKT1 and ERBB2 correlated with lower pathologic tumor stages (P = 0.01 and P = 0.03, respectively). Increased RHOA expression was more common in grade 3 than in grade 2 tumors (P = 0.016). There were no other correlations among the 15 factors studied and pathologic stage, lymph node status, or tumor grade. No association was found between bladder cancer death and altered marker status for any of the markers studied.

Conclusions

Currently, there are reasons to have a skeptical attitude toward the value of tissue microarray based immunohistochemistry as a method for evaluating prognostic markers in invasive bladder cancer. In this study, 15 antibodies were tested but were found to be of little clinical value. Whether this negative finding is related to the group of patients or factors studied, or the methodology is unclear.

Introduction

Cancer development and progression involve genetic alterations leading to dysregulation of different cellular functions such as growth signaling, angiogenesis, and apoptosis but also to tissue invasion and metastases [1]. In urothelial carcinoma, at least 3 different pathways of tumor development have been described, each including different genetic changes [2]. Expression profiling with oligonucleotide microarray technology has recently identified a set of genes with prognostic relevance in patients who underwent radical cystectomy [3]. Over-expression and under-expression and/or alterations of the proteins encoded by these genes involved in the different cellular functions may also predict clinical outcome. In invasive bladder cancer treated with radical cystectomy, protein alterations that have been suggested to influence independently outcome in multivariate analysis are TP53, RB1, and CDKN1A [4], [5]. However, there are also reports in which the prognostic values of TP53, RB1, and CDKN1A have not been confirmed [6], [7].

Many patients who undergo radical cystectomy will have tumor recurrence. Prognostic information obtained from pathologic tumor stage and lymph node status is insufficient to predict outcome and identify patients who will benefit from adjuvant chemotherapy. The tissue microarray is a new high-throughput technique developed by Kononen et al. [8], enabling investigation of several tumor samples simultaneously with immunohistochemical methods. For example, tissue microarray methodology has been used to identify independent prognostic markers in breast cancer and colon cancer [9], [10]. It has been suggested that several molecular markers are needed to predict outcome in invasive bladder cancer [11], and combining markers that work at least partly through different pathways would be preferable [4]. To explore potential prognostic information in the expression of several proteins involved in different cellular functions, we performed tissue microarray based immunohistochemical evaluation of tumor samples in a consecutive cystectomy series.