IMPACT (sipuleucel-T)

IMPACT randomised 512 men with metastatic castration-resistant prostate cancer 2:1 to sipuleucel-T (341) or placebo (171), given intravenously every 2 weeks for three infusions. The primary endpoint was overall survival, analysed by a stratified Cox model adjusted for baseline PSA and lactate dehydrogenase.

Sipuleucel-T reduced the risk of death by 22 percent, hazard ratio 0.78 (95 percent confidence interval 0.61 to 0.98, p=0.03), a 4.1-month gain in median survival (25.8 against 21.7 months). Thirty-six-month survival was 31.7 against 23.0 percent. The result held in an unadjusted model (0.77, 0.61 to 0.97, p=0.02) and after adjusting for later docetaxel use (0.78, 0.62 to 0.98, p=0.03).

The time to objective disease progression was the same in both groups. Nothing shrank. That dissociation, longer life without tumour response, is the finding that taught oncology that immunotherapy does not have to shrink a tumour to work, and it was the reason the field developed immune-related response criteria before checkpoint inhibitors arrived. Chills, fever and headache were the characteristic adverse events.

The commercial story is the cautionary one. The product required leukapheresis and a manufacturing turnaround for each man, was very expensive, and never sold in the volumes needed. Dendreon filed for bankruptcy in 2014. In England NICE TA332 on sipuleucel-T has been withdrawn because the marketing authorisation was withdrawn, so it is not available at all.