Login to Access Video or Poster Abstract: MP88-03
Sources of Funding: AMA Foundation, H&H Lee Research Program

Introduction

Latent recurrences despite favorable pathology has been a longstanding conundrum in surgical oncology. In muscle-invasive urothelial carcinoma of the bladder, patients with clinically localized disease (pT2 N0) have a recurrence rate of 11-35% following radical cystectomy, with recurrence rates in higher stages of localized disease up to 50-69% (pT4 N0). Recurrence with N0 disease may be attributed to lymph node micrometastases, therefore improved pathologic staging of lymph nodes may add precision to selecting adjuvant therapy as well as stratifying patients with metastatic disease. Here, we describe a diagnostic approach combining fluorescence activated cell sorting and next-generation sequencing that identifies micrometastases in a patient with muscle-invasive urothelial carcinoma of the bladder. _x000D_

Methods

Tumor specimens from multiple regions of patient’s bladder tumor as well as lymph nodes were obtained and were dissociated into a single cell suspension. Fluorescence activated cell sorting using the surface markers CD44 and CD49f was utilized to isolate a cancer cell subpopulation associated with stemness. Whole-exome sequencing and RNA sequencing of total cells and the CD44+CD49f+ subpopulation in multiple tumor specimens and regional lymph nodes were performed in order to identify micrometastases and evaluate gene expression among the different regions and populations._x000D_

Results

Pathology of the lymph nodes resected demonstrated N0 disease. Mean allele frequency analysis demonstrated a significant correlation between tumor cancer cells and cancer cells isolated from the lymph nodes. RNA-sequencing revealed intratumoral heterogeneity as well as enrichment for immune system and lipid metabolism gene sets in the micrometastatic cancer cell subpopulations. _x000D_

Conclusions

Although pathology demonstrated no lymph node disease, we were able to isolate a cancer cell subpopulation from the pathologically negative lymph nodes. Whole-exome sequencing verified the cells isolated from the lymph nodes were indeed similar to the primary tumor and RNA-seq demonstrated differences in gene expression among the total population and subpopulation. Our analysis illustrates how next-generation sequencing of cancer cell subpopulations can be utilized to improve pathologic staging, and provide insight into cancer stem cell biology._x000D_

Funding

AMA Foundation, H&H Lee Research Program