External validation of FXYD3 and KRT20 as predictive biomarkers for the presence of micrometastasis in muscle invasive bladder cancer lymph nodes.

INTRODUCTION: Recent studies have proposed that FXYD3 and KRT20 mRNA quantified by quantitative reverse transcription polymerase chain reaction (qRT-PCR) in paraffin could be biomarkers to detect lymph nodes with micrometastases that avoid detection by conventional analysis with hematoxylin-eosin (HE). A validation study was conducted on the lymph nodes of patients who underwent radical cystectomy. OBJECTIVE: To classify the adenopathic state of a sample of patients who underwent cystectomy, based on the lymph node expression of FXYD3 and KRT20. The secondary objective was to assess whether there is a differential oncologic evolution for the patients, depending on the lymph node expression of these proteins. MATERIAL AND METHOD: The study included lymph nodes from 64 patients who underwent cystectomy for infiltrating bladder tumor: The model was developed using metastatic lymph nodes from 15 patients and lymph nodes from 4 patients with no known tumor. Genetic expression was measured using real-time qRT-PCR. We calculated (using qRT-PCR) the median expression of FXYD3 and KRT20 mRNA in the lymph node tissue. We then analyzed the receiver operating characteristic (ROC) curves, according to the function y=0.1400+0.250FXYD3-2.532. The cutoff was established using an ROC curve. The formula was applied to the remaining lymph node tissue, based on the previously established cutoff. The sample was classified into 4 subgroups: HE- qRT-PCR-, HE- qRT-PCR+, HE+ qRT-PCR+ y HE+, qRT-PCR-. A descriptive, comparative analysis was performed, as well as a metastatic progression-free survival analysis, calculating the Kaplan and Meyer curves for the 3 established subgroups. The test results were considered statistically significant at P<.05. RESULTS: Using qRT-PCR, we verified that there were differences in the median expression of FXYD3 (P=.05) and KRT20 (P=.009) between the lymph node tissues of patients with benign prostate hyperplasia and those of patients with lymph node metastasis. A cutoff was assigned to 0.377. The sample was classified as follows: 37.5% of the patients were pN0 by HE and pN0 by qRT-PCR (-HE -qRT-PCR), 39.1% were pN0 by HE but metastatic by qRT-PCR (-HE +qRT-PCR), and 15 patients (23.4%) were metastatic by both techniques (+HE +qRT-PCR). The Kaplan and Meyer curves showed poorer metastatic progression-free survival for the patients who were +HE and +qRT-PCR than for the other subgroups, with no significant differences between -HE +qRT-PCR and -HE -qRT-PCR. CONCLUSIONS: According to our results, 39.1% of the patients with infiltrating vesical tumors overexpressed the FXYD3 and KRT20 biomarkers and were N0 by HE. We observed no differential clinical behavior among the patients who underwent cystectomy according to their expression of FXYD3 and KRT20 when they were N0 by HE.