JoVE - Journal of Visualized Experiments
JoVE Visualize
Research Domains
View All

Contact Us

Squamous cell carcinoma of unknown primary (SCCUP): a genomic landscape study

  • 1Department of Medicine, Division of Medical Oncology, University of Colorado Anschutz Medical Campus, Aurora, USA. Electronic address: hannah.r.robinson@cuanschutz.edu.
  • 2Department of Medicine, Division of Medical Oncology, University of Colorado Anschutz Medical Campus, Aurora, USA.
  • 3Foundation Medicine Inc., Cambridge, USA.
  • 4Moffitt Cancer Center, Tampa, USA.
  • 5MD Anderson Cancer Center, Houston, USA.
  • 6Fred Hutch Cancer Center & University of Washington, Seattle, USA.
  • 7State University of New York Upstate Medical University, Syracuse, USA.
  • 8San Raffaele Hospital and Scientific Institute, Milan, Italy.
  • 9Foundation Medicine Inc., Cambridge, USA; State University of New York Upstate Medical University, Syracuse, USA.
+

|

June 11, 2025

Abstract

BACKGROUND

Squamous cell carcinoma (SCC) of unknown primary (SCCUP) refers to any SCC for which the primary tumor origin cannot be identified despite guideline-directed evaluation. Although strategies employing comprehensive genomic profiling (CGP) to identify targets for non-SCC carcinomas of unknown primary (CUPs) have shown benefit, the genomic landscape in SCCUP remains poorly defined. Here we describe results of CGP in patients with SCCUP to identify potential therapy targets and characteristic biomarkers.

PATIENTS AND METHODS

Cases of advanced SCCUP were identified in the FoundationCORE® database by review of clinical history and pathology records. Samples underwent DNA extraction and sequencing to identify genomic alterations (GAs), microsatellite instability (MSI) status, tumor mutational burden (TMB), genomic mutational signatures, and viral reads. Programmed death-ligand 1 (PD-L1) expression was determined by immunohistochemistry.

RESULTS

443 SCCUP cases were identified. Common presentation sites included lymph nodes (41.1%), liver (15.6%), and soft tissue (15.1%). A mean of 6.5 GAs were observed per case. The most frequent non-targetable GAs involved TP53 (62.5%), CDKN2A (37.0%), CDKN2B (19.6%), KMT2D (18.3%), and TERT (16.0%); the most frequent GAs potentially targetable in biomarker-driven trials included PIK3CA (27.3%), PTEN (15.1%), MTAP (13.1%), KRAS (10.4%), and NOTCH1 (8.4%). Among all SCCUP cases, 2.0% had MSI-high (MSI-H) status and 33.9% had TMB ≥10 mutations/megabase. Among 204 cases with available PD-L1 testing, 39.2% were low-positive [tumor proportion score (TPS) 1%-49%] and 29.9% were high-positive (TPS ≥50%). SCCUP cases presenting with liver involvement had fewer GAs per tumor and lower TMB compared with other SCCUP cases. In contrast, cases presenting with inguinal, pelvic, or retroperitoneal involvement were more likely to demonstrate evidence of human papilloma virus (HPV) infection and apolipoprotein B mRNA editing catalytic polypeptide-like (APOBEC) genomic signatures.

CONCLUSIONS

CGP of this SCCUP cohort identified GAs that may guide consideration of molecularly targeted therapy via tumor-agnostic indications and/or treatment in biomarker-driven trials. SCCUPs frequently exhibit biomarkers associated with response to immune checkpoint inhibitors.

Related Concept Videos