Quantification of Tumor-Infiltrating Lymphocytes as a Risk-Stratification Tool in Vulvovaginal Melanoma Utilizing an Open-Source Platform. Anna Sarah Erem, Gulisa Turashvili.
Sarah Erem
Scholar | Resident Pathology
Presented at: United States & Canadian Academy of Pathology (USCAP) Annual Meeting
Date:
Views: 5
Summary: Background:
Vulvovaginal melanomas (VVM) are rare with limited clinicopathologic and outcome data, mostly extrapolated from cutaneous melanomas (CM). Consequently, there is a need for more robust tools for risk stratification and therapeutic decisions in VVM. The presence and density of tumor-infiltrating lymphocytes (TILs) in vertical growth phase have been proposed as an
independent prognostic marker in CM. We have previously demonstrated a strong correlation between TILs and observed survival in The Cancer Genome Atlas (TCGA) primary CM data
Design:
We identified 40 archival primary VVMs (10 biopsies, 30 resections), each comprising 1-5 representative slides converted to Whole Slide Images (WSI), totaling 116 images analyzed. An open-source technology, QuPath, was then utilized to detect, classify and quantify WSI data with the following cell types: neoplastic VVM cells, TILs, and other non-neoplastic cells. The areas of melanoma-in-situ (MIS) were excluded from analysis. Statistical analyses were performed in RStudio and p-values <.05 were considered significant.
Results:
The median patient age was 61 years (range 16-82). The distribution of TILs in the invasive VVM ranged from 1.5% to 21.1% with median and mean values of 9.0% and 9.9%, respectively. Logistic regression analysis showed that the percentage of TILs (TIL%) in the invasive component was a significant prognostic factor (odds ratio=1.34, p<.003) and strongly correlated with both
disease-specific survival (DSS) (r=0.58, p<.001, Fig. 1A) and overall survival (OS) (r=.67, p<.001). Furthermore, VVMs with ≥14% TIL% had a longer OS and DSS than those with 6-13.9% and <6% TIL% (median 136 vs 72.6 vs 22.8 months, respectively) (Fig. 2). In clinical follow-up of 10 patients on immunotherapy, TILs showed a positive correlation with treatment response (rpb=0.85, p=.002, Fig. 1B).