Identification of diagnostic and predictive morphological and molecular markers in renal tumors

Objectives: Ongoing advances in tumor biology keep expanding the spectrum of renal neoplasms, with more than twenty renal cell carcinoma histotypes recognized in the last WHO classification. Thus, the discovery of novel diagnostic and predictive biomarkers for rarer tumors, such as those belonging to the “molecularly defined category”, is warranted to improve physicians’ ability to recognize and treat these uncommon diseases. Furthermore, even in the most common entities, a noteworthy percentage of the patients treated for an organ-confined or locally advanced neoplasm lately develop distant metastases. In the present project, we aimed to identify novel pathological, immunohistochemical, and molecular features bearing a diagnostic and predictive potential in a broad spectrum of common and rare epithelial or mesenchymal renal neoplasms. Material and methods: The following cases were retrieved, including samples from both primary neoplasms and metastases when available: four high-grade eosinophilic renal cell tumors, twenty-seven TFE3-rearranged renal cell carcinomas, ten TFEB-rearranged renal cell carcinomas, one TFE3-rearranged PEComa, thirteen FH-deficient renal cell carcinomas, and one hundred forty-six clear cell renal cell carcinomas. The specimens were extensively studied by immunohistochemistry (IHC), FISH, and their molecular analyses, including diagnostic and predictive markers, such as STING and PD-L1, and targetable molecular alterations, like mTOR mutations. Results: Concerning eosinophilic tumors, the cases herein described belong to a novel entity, provisionally named mTOR-mutated eosinophilic renal cell carcinoma, sharing metastatic potential, mTOR mutations, loss of chromosome 1, cathepsin-K expression, and response to target therapy. Cathepsin K, CA9, CK7, and parvalbumin are the most reliable markers in distinguishing TFE3/TFEB-rearranged renal cell carcinomas from common renal cell tumors; the positivity threshold depends on the marker considered. Renal TFE3-rearranged PEComas are usually composed of nested clear to eosinophilic cells and harbor SFPQ/PSF::TFE3 fusions in more than half of the cases. By IHC, they express cathepsin K, HMB45, MART-1, and CD68 (PG-M1) while labeled negative for PAX8, CD10, and CKs. Conversely, TFE3-rearranged renal cell carcinomas are positive for PAX8, CD10, and, less frequently, cathepsin K, HMB45, and MART-1, whereas they commonly stain negative for CD68 (PG-M1) and CKs. About FH-deficient renal cell carcinoma, 75% of our cases (8/12) developed metastases, mainly involving the abdominal lymph nodes (5/8, 67%), and the peritoneum (4/8, 50%). Metastases revealed morphological heterogeneous architecture often differing from the primary renal tumor, closely resembling a primary mesothelial process when occurring in the peritoneum. In all primary and secondary samples, the IHC combination of FH loss and positive 2SC staining was diagnostically helpful. STING staining was detected in most primary tumors (9/11, 82%), including all but two aggressive renal neoplasms and 86% (6/7) of tumors significantly expressing PD-L1. Finally, in clear cell renal cell carcinoma, STING immunolabeling was detected in 36% (53/146) of the samples, more frequently in high-grade (G3-G4) tumors (48%, 43/90) and recurrent/metastatic ones (75%, 24/32). Moreover, STING correlated with aggressiveness parameters, like granular necrosis (p = 0.001), stage (p < 0.001), and metastases (p < 0.001), and reached an independent statistical significance (p = 0.029) in multivariate analysis. Conclusions: The identification of reliable morphological, immunohistochemical, and molecular findings in kidney cancer is crucial to address the needs raised by evolving classifications. The diagnostic and predictive markers described in the present project will provide novel insight to improve the global management of kidney cancer patients, filling the gap between basic research and daily practice.