Clinicopathological characteristics of cancer of unknown primary (CUP) with renal profile using gene expression profiling (GEP) based cancer classification.

456 Background: CUP is a rare and heterogenous clinicopathologic syndrome with unidentifiable origin at time of diagnosis. Based on immunophenotyping, specifically PAX8 staining, CUP with Renal profile (RCUP) is emerging as a distinct clinical subset. In addition to clinical presentation and immunohistochemistry (IHC), molecular profiling for tissue of origin may aid defining CUP subsets. Precision in diagnosis impacts therapeutics given that cytotoxic chemotherapy options for CUP do not overlap with tyrosine kinase inhibitors or checkpoint inhibitors used for renal cell carcinoma. Methods: We retrospectively (2018-2022) evaluated patients with CUP composed of cases (renal profile, N=100, clear cell or papillary) or controls (non-renal, N=200) identified using the 92-gene assay (CancerTYPE ID), a validated classifier for predicting tissue of origin. Clinicopathologic data including IHC and molecular profiling results were collected using pathology specimens and reports. Results: Baseline characteristics (age, gender, tumor grade) of cases and controls were comparable, except for histology, wherein adenocarcinoma was less common with RCUP (27% vs 48%, p<0.001). Median number of IHC stains performed was 11 (range 0-30) vs 9 (0-29) (p=0.04) in RCUP vs non-renal cases, respectively. While PAX8 was tested in 64% vs 39% (p<0.001) of cases, it was found to be strongly or focally positive in 87.5% vs 21.8% (p<0.001), respectively. In contrast to non-renal CUP, RCUP patients were more likely to have bone biopsies (24% vs 11.5%, OR 2.4, p=0.007), less likely to have high-TMB (2.8% vs 28.2%, OR 0.07, p=0.001), and had a lower proportion of TP53 mutations (18.8% vs 89.2%, OR 0.028, p<0.001), KRAS mutations (13.3% vs 77.4%, OR 0.045, p<0.001) and BRAF mutations (0% vs 42%, OR 0, p=0.015). Conclusions: RCUP appears to be a distinct subset of CUP with variable diagnostic workup. Critical IHC markers may be omitted in a considerable number of cases and despite positivity, may still confer an uncertain diagnosis. Key molecular factors appear to be different and may require tailored therapeutic approaches. This study further supports the clinical utility of GEP-based cancer classification with the 92-gene assay to predict RCUP and may be an important adjunct to IHC to improve diagnostic accuracy for CUP and to guide treatment selection for better clinical outcomes. RCUP (N=100) Non-renal CUP (N=200) IHC Cases Tested (%) Cases Positive (%) Cases Tested (%) Cases Positive (%) CK7 83 47 81 75* CK20 75 11 77 24* PAX8 64* 88* 39 22 Napsin 29 28* 25 8 RCC 26* 31 4 13 Vimentin 17 94 5.5 73 *Significantly (p<0.05) greater proportion of cases.