CUPISCO

CUPISCO opened in July 2018 across more than 30 countries. Patients with newly diagnosed, unfavourable-subset cancer of unknown primary received three cycles of platinum-based chemotherapy; those whose disease was controlled were randomised three to one between molecularly guided therapy and continued chemotherapy. Tumour tissue and plasma were profiled with Foundation Medicine's assays and a molecular tumour board assigned each patient in the experimental arm to one of the trial's targeted or immunotherapy options: atezolizumab (alone or with chemotherapy), alectinib, entrectinib, erlotinib, vemurafenib with cobimetinib, vismodegib, trastuzumab emtansine, pertuzumab with trastuzumab, olaparib or ipatasertib, or to atezolizumab-based therapy where no alteration was targetable.

Alwin Krämer and colleagues reported the primary analysis in the Lancet in 2024. Progression-free survival, the primary endpoint, was longer with molecularly guided therapy: median 6.1 months against 4.4 months with continued chemotherapy, a hazard ratio of 0.72, and the benefit was clearest in patients with an alteration that matched one of the targeted drugs. Overall survival data were immature, and the experimental arm was heterogeneous, so the trial establishes the strategy rather than any single drug.

CUPISCO's practical effect is to give comprehensive genomic profiling a randomised evidence base in a disease where guidelines had recommended it on faith, and to shift the standard first question in unfavourable cancer of unknown primary from tissue of origin to targetable alteration. What remains open is whether the modest progression-free gain translates into survival, how to sequence profiling so that results arrive before chemotherapy stops working, and whether the approach holds outside a trial with a dedicated molecular tumour board.