No OS benefit with first-line durvalumab ± tremelimumab in advanced urothelial carcinoma

Durvalumab with or without tremelimumab in the first-line setting does not improve overall survival (OS) over standard-of-care (SoC) chemotherapy in patients with unresectable, locally advanced or metastatic urothelial carcinoma, according to the phase III DANUBE trial.

“The results of the DANUBE trial … do not show that durvalumab is superior to chemotherapy as a first-line treatment, questioning the approach of immune checkpoint inhibitor monotherapy in this setting,” said the researchers.

Participants in this open-label, multinational study were 1,032 treatment-naïve adults with unresectable, locally advanced or metastatic urothelial carcinoma and ECOG performance status 0–1. They were randomized 1:1:1 to receive intravenous doses of the PD-L1 inhibitor durvalumab (1,500 mg Q4W) alone, ≤4 doses of durvalumab plus the CTLA-4 inhibitor tremelimumab (75 mg Q4W) followed by durvalumab maintenance (1,500 mg Q4W), or ≤6 cycles of SoC chemotherapy (gemcitabine plus cisplatin or carboplatin). They were followed up for a median 41.2 months. Baseline characteristics were well-balanced between groups. Sixty percent of patients from each group had high PD-L1 expression.

In the high PD-L1 population, OS was comparable between the durvalumab monotherapy and chemotherapy groups (median 14.4 vs 12.1 months; hazard ratio [HR], 0.89, 95 percent confidence interval [CI], 0.71–1.11; p=0.3039). [ESMO 2020, abstract 697O; Lancet Oncol 2020;doi:10.1016/S1470-2045(20)30541-6]

OS was also similar between the durvalumab-tremelimumab and chemotherapy groups in the intention-to-treat (ITT) population (regardless of PD-L1 expression; median 15.1 vs 12.1 months; HR, 0.85, 95 percent CI, 0.72–1.02; p=0.0751).

The OS results were generally consistent across subgroups, and between cisplatin eligible and ineligible patients.

In secondary analyses, median OS was 13.2 months with durvalumab in the ITT population and 17.9 months with durvalumab-tremelimumab in the high PD-L1 population (HR, 0.99 and 0.74 vs chemotherapy, respectively).

Median progression-free survival (PFS) was 2.3, 3.7, and 6.7 months in the durvalumab, durvalumab-tremelimumab, and chemotherapy groups, respectively, in the ITT population, and 2.4, 4.1, and 5.8 months, respectively, in the high PD-L1 population.

Disease progression occurred in 53 and 51 percent of durvalumab monotherapy recipients in the ITT and PD-L1 high populations, respectively, and in 18 and 20 percent of chemotherapy recipients, respectively.

Grade 3–4 treatment-related adverse events (TRAEs) occurred in fewer durvalumab (14 percent) and durvalumab-tremelimumab (27 percent) than chemotherapy (60 percent) recipients. The most common grade 3–4 TRAE was elevated lipase in durvalumab and durvalumab-tremelimumab recipients (2 and 5 percent, respectively), and neutropenia and anaemia in chemotherapy recipients (21 and 20 percent, respectively). AEs led to study drug discontinuation in 12, 24, and 17 percent of durvalumab, durvalumab-tremelimumab, and chemotherapy recipients, respectively. Nine, 23, and 16 percent of durvalumab, durvalumab-tremelimumab, and chemotherapy recipients, respectively, experienced serious TRAEs, the most common being pneumonia, diarrhoea, and anaemia in those respective groups. Study drug toxicity led to two deaths each in the durvalumab and durvalumab-tremelimumab groups and one in the chemotherapy group.

Further research warranted

“While responses [to SoC platinum-based chemotherapy] are initially high, patients relapse and long-term durable remissions are rare,” said Professor Thomas Powles from Barts Cancer Institute, Queen Mary University of London, UK, at ESMO 2020.

“[I]n our study, chemotherapy appeared better than durvalumab at achieving initial control of disease [in PD-L1–positive patients], showing higher response rates and longer PFS,” said the researchers.

Of note was the higher objective response rate in chemotherapy vs durvalumab recipients in both the ITT (49 percent vs 26 percent) and high PD-L1 populations (48 percent vs 28 percent). Similar results were observed with durvalumab-tremelimumab vs chemotherapy in the ITT population, though the duration of response was longer with the former, suggesting more durable remission.

“Secondary analyses suggested that tremelimumab added to the activity of durvalumab. This appeared most marked in the PD-L1–positive population,” said Powles. However, this may come at the risk of greater toxicity.

Research is necessary to identify the role of CTLA-4 inhibitors in metastatic urothelial cancer, particularly in patients with PD-L1–positive disease, said the researchers.

“[The results also suggest that] PD-L1 expression alone might not be sufficient to identify patients who benefit from PD-1 or PD-L1 inhibitors,” they said. “More specific patient selection might be required for biomarker-targeted monotherapy to outperform effective chemotherapy.”

They acknowledged that the “substantial proportion” of patients who received subsequent therapy following disease progression could have affected the results.