Abstract
Pembrolizumab, an anti-programmed death 1 monoclonal antibody, has revolutionized the treatment of metastatic urothelial carcinoma. However, the optimal treatment duration for treatment responders has not been established. To address this, we retrospectively assess the treatment outcomes and duration of pembrolizumab for patients whose best response was complete response (CR) or partial response (PR) in a Japanese nationwide cohort of platinum-refractory metastatic urothelial carcinoma. Of 203 patients whose best response was CR or PR, 83 patients discontinued pembrolizumab before progression. The median pembrolizumab treatment duration was 6.9 months. The 2-year relapse-free survival (RFS), treatment-free survival, and OS rates after discontinuation were 49.0%, 57.4%, and 74.5%, respectively. CR, higher hemoglobin levels, and a better Eastern Cooperative Oncology Group performance status at the time of discontinuation were associated with significantly better RFS. Pembrolizumab was re-administered to 12 patients. Pembrolizumab re-challenge resulted in CR, PR, stable disease, and progressive disease in six, three, two, and one patient, respectively. Propensity score-matched landmark analysis revealed no significant OS difference between patients who continued or discontinued pembrolizumab at 6, 12, and 18 months (p = 0.91, 0.99, and 0.25, respectively). Our findings demonstrated that patients with objective responses had favorable survival outcomes and suggested that pembrolizumab could be discontinued safely in this population. This study should drive further efforts to optimize the treatment duration for pembrolizumab responders.