Safety and activity of pembrolizumab in patients with locally advanced or metastatic urothelial cancer (KEYNOTE-012): a non-randomised, open-label, phase 1b study

Background PD-1 and its ligands are expressed in urothelial cancer, and fi ndings have shown that inhibition of the PD-1 pathway has clinical benefi t. We aimed to assess the safety and activity of an anti-PD-1 antibody pembrolizumab in patients with locally advanced or metastatic urothelial cancer. Methods This study was part of the non-randomised, multi-cohort, open-label, phase 1b KEYNOTE-012 basket trial. We enrolled patients aged 18 years and older with a histologically or cytologically confi rmed diagnosis of locally advanced or metastatic urothelial cancer, including cancers of the renal pelvis, ureter, bladder, or urethra, from eight hospitals in the USA and Israel. Patients were required to have at least 1% PD-L1 expression detected on the tumour cells or in tumour stroma, as determined by immunohistochemistry. Patients were given 10 mg/kg intravenous pembrolizumab every 2 weeks until disease progression, unacceptable toxic eff ects, or the end of the study (ie, 24 months of treatment). Primary endpoints were safety and overall response (defi ned by Response Evaluation Criteria In Solid Tumors [RECIST] version 1.1), as assessed by a masked, independent central review. Safety was assessed in patients who received one or more doses of pembrolizumab (all-patients-as-treated population); activity was assessed in patients who received pembrolizumab, had measurable disease at baseline, and had one or more post-baseline scans, or discontinued because of progressive disease or treatment-related adverse events (full analysis set). This study is registered with ClinicalTrials. gov, number NCT01848834, and is no longer enrolling patients; follow-up is ongoing. Findings Between May 14, 2013, and Dec 10, 2013, 115 patients were tissue pre-screened as part of a two-part consent process. 61 (53%) patients were PD-L1 positive, of whom 33 were enrolled in this study. All enrolled patients received at least one dose of pembrolizumab and were included in the safety analyses. 27 patients comprised the full analysis set and were deemed assessable for activity. Six patients were not assessable: three discontinued study drug because of a non-treatment-related adverse event before the fi rst post-baseline scan, two withdrew before the fi rst post-baseline scan, and one had no measurable disease at baseline. The most common treatment-related adverse events were fatigue (six [18%] of 33 patients) and peripheral oedema (4 [12%]). Five (15%) patients had 11 grade 3 treatment-related adverse events; no single event occurred in more than one patient. Three (9%) patients experienced fi ve serious treatment-related adverse events. After median follow-up of 13 months (range 1-26, IQR 5-23), an overall response was achieved in seven (26% [95% CI 11-46]) of 27 assessable patients, with three (11% [2-29]) complete and four (15% [4-34]) partial responses. Of the four deaths that occurred during the study (cardiac arrest, pneumonia, sepsis, and subarachnoid haemorrhage), none were deemed treatment related. Interpretation Pembrolizumab showed anti-tumour activity and acceptable safety in patients with advanced urothelial cancer, supporting ongoing phase 2 and 3 studies of pembrolizumab in this population.