Alwin Krämer, Tilmann Bochtler, Chantal Pauli, Kai-Keen Shiu, Natalie Cook, Juliana Janoski de Menezes, Roberto A Pazo-Cid, Ferran Losa, Debbie G J Robbrecht, Jiří Tomášek, Cagatay Arslan, Mustafa Özgüroğlu, Panoraia Arni, Frederic Bigot, Sun Young Kim, Yoichi Naito, Antoine Italiano, Nasséra Chalabi, Gonzalo Durán-Pacheco, Ning Ma, Jeremy Scarato, Jeffrey S Ross, Holger Moch, Linda Mileshkin
CUPISCO (ClinicalTrials.gov identifier: NCT03498521) demonstrated longer progression-free survival (PFS) with comprehensive genomic profiling (CGP) and subsequent molecularly guided therapies (MGTs), versus standard platinum-based chemotherapy, in patients with previously untreated, unfavorable cancer of unknown primary (CUP) who reached disease control after induction chemotherapy (three cycles). We report efficacy and safety after >1 year of additional follow-up. Eligible patients were randomly assigned (3:1) to MGT (investigator-chosen after discussion in a molecular tumor board) or three further cycles of chemotherapy. The primary end point was PFS. Secondary end points included overall survival (OS) and safety. At data cutoff (December 6, 2024), 436 patients were randomly assigned (326 to MGT; 110 to chemotherapy). Median follow-up was 37.0 months (range, 0.0-67.8). Updated median PFS was 6.1 months (95% CI, 4.7 to 6.5) with MGT and 4.4 months (95% CI, 4.2 to 6.4) with chemotherapy (hazard ratio [HR], 0.75 [95% CI, 0.59 to 0.95]; P = .017); median OS was 15.2 months (95% CI, 13.9 to 18.4) and 12.8 months (95% CI, 9.8 to 15.4), respectively (HR, 0.79 [95% CI, 0.61 to 1.02]; P = .0689). No new safety signals were identified. These updated results aligned with the primary analysis, demonstrating the benefit of CGP with subsequent MGT and highlighting the importance of incorporating CGP at initial diagnosis to guide treatment decisions for patients with unfavorable CUP.