medwireNews: Adding pembrolizumab to first-line chemotherapy does not improve progression-free survival (PFS) for patients with advanced urothelial carcinoma (UC), KEYNOTE-361 trial findings show.
The results were reported at the ESMO Virtual Congress 2020 by Ajjai Alva, from the University of Michigan in Ann Arbor, USA, who said that the study focused on unresectable or metastatic UC of the renal pelvis, ureter, bladder, or urethra, with a fifth of patients having liver metastasis and around 45% a PD-L1 combined positive score of at least 10.
The joint primary endpoints were PFS by blinded independent central review (BICR) and overall survival (OS) in the intention-to-treat (ITT) population, he explained.
Median PFS was 8.3 months for the 351 patients randomly assigned to receive pembrolizumab 200 mg every 3 weeks alongside up to six cycles of gemcitabine 1000 mg/m2 plus cisplatin 70 mg/m2 or carboplatin AUC 5, followed by up to 29 further cycles of pembrolizumab.
This compared with a median PFS of 7.1 months for the 352 patients who were given chemotherapy only. Although the confidence intervals of the hazard ratio (HR) of 0.78 did not cross 1, the p value of 0.0033 was above the prespecified boundary of 0.0019 and therefore the difference between the groups was not significant, Alva said.
Overall survival (OS) in the pembrolizumab–chemotherapy and chemotherapy-only arms was a median of 17.0 and 14.3 months, respectively, and again the HR of 0.86 did not meet the prespecified boundary for significance.
However, Alva commented that “informative early censoring for PFS and subsequent anti-PD-(L)1 therapy in almost half of patients in the chemotherapy arm may have impacted PFS by BICR and OS results, respectively.”
He highlighted an exploratory analysis of investigator-assessed PFS in the ITT group where median PFS was 8.3 months with pembrolizumab–chemotherapy versus 6.5 months with chemotherapy alone, giving a significant hazard ratio of 0.69.
“Discordance in [the] chemotherapy arm was predominantly driven by patients who had [progressive disease] by investigator assessment but not by BICR assessment,” Alva observed.
Chemotherapy-only patients were more likely to receive any subsequent treatment (61.1 vs 35.3%) than those given pembrolizumab–chemotherapy and specifically PD-(L)1-targeted therapy (48.0 vs 6.6%), he added.
After adjusting for subsequent use of atezolizumab, avelumab, durvalumab, nivolumab, or pembrolizumab, exploratory analysis suggested that first-line pembrolizumab plus chemotherapy may have extended OS compared with chemotherapy alone, with an adjusted HR of 0.71.
The lack of efficacy in the primary endpoints prevented the investigators from formally comparing the outcomes of 307 patients who were given up to 35 cycles of pembrolizumab 200 mg every 3 weeks with the other trial arms.
Nevertheless, Alva reported an exploratory analysis indicating that “pembrolizumab did not appear superior to chemotherapy.” He cited median OS durations of 15.6 and 14.3 months, respectively.
The overall response rate by BICR in the ITT population was 54.7% with pembrolizumab–chemotherapy, 44.9% for chemotherapy alone, and 30.3% with pembrolizumab alone, “illustrating that cytotoxic chemotherapy perhaps does a better job than immunotherapy in gaining initial disease control,” the presenter commented.
But the median durations of response were 8.5, 6.2, and 28.2 months, respectively, and the corresponding 12-month rates of response were 42.4%, 23.5%, and 65.1%.
“Pembrolizumab was associated with more durable responses,” Alva therefore commented.
And he concluded that “the safety profile of pembrolizumab was in line with prior studies, with a lower rate of all-cause [adverse events] than chemotherapy.”
medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2020 Springer Healthcare Ltd, part of the Springer Nature Group
This independent news story was supported by an educational grant from Pfizer and Merck Healthcare KGaA, Darmstadt, Germany