麻豆传媒

PD-L1 checkpoint blockade with or without anti-CTLA-4 failed to improve overall survival (OS) compared to standard chemotherapy in previously untreated metastatic bladder cancer, a randomized phase III trial found.

Durvalumab (Imfinzi) alone in PD-L1-high expressors yielded median OS of 14.4 months versus 12.1 months with platinum-based chemotherapy (HR for death 0.89, 95% CI 0.71-1.11, P=0.30).

And in the intent-to-treat population (ITT) of both high and low PD-L1 expressors, durvalumab combined with the investigational agent tremelimumab produced a median OS of 15.1 months, as compared to 12.1 months with chemotherapy (HR 0.85, 95% CI 0.72-1.02, P=0.075), reported Thomas Powles, MD, of the Barts Cancer Institute in London.

"Overall the DANUBE trial didn't meet either of its co-primary endpoints," he said during his presentation at the 2020 European Society for Medical Oncology (ESMO) virtual congress. "Secondary analyses suggested that tremelimumab added to the activity of durvalumab -- this appeared most marked in the PD-L1-positive population."

In a secondary analysis of high PD-L1 expressors, the combination improved OS over chemotherapy, reaching a median of 17.9 months (HR 0.74, 95% CI 0.59-0.93), and landmark analyses showed improvements with durvalumab-tremelimumab at both the 1 year (63% vs 51%) and 2 year (44% vs 29%) timepoints.

Findings from the study were simultaneously published in .

"Had we picked the PD-L1-biomarker-positive population for both arms, and not just the durvalumab-alone arm, the trial would have been positive for this combination," said Powles. "We didn't, and it wasn't."

In the ITT population, durvalumab alone was no better than chemotherapy, with a median OS of 13.2 months (HR 0.99, 95% CI 0.83-1.17).

ESMO-invited discussant Richard Cathomas, MD, PhD, of Kantonsspital Graubünden in Chur, Switzerland, said that the almost 6-month OS improvement with durvalumab plus tremelimumab in the PD-L1-high group is interesting, but remains hypothesis generating.

"The biomarker of PD-L1 expression does -- at this moment -- not appear to be useful to make any decisions for the first-line treatment of platinum-eligible patients," he said.

Cathomas recommended that platinum-eligible patients receive chemotherapy upfront, and those with response or stable disease should go on to a switch-maintenance approach with the anti-PD-L1 immunotherapy avelumab (Bavencio), a recently FDA-approved option.

Overall response rates (ORRs) in DANUBE were highest with chemotherapy in both the ITT (49%) and PD-L1-high (48%) populations, with complete responses in 6% and 7%, respectively. With the anti-PD-L1/CTLA-4 combination, ORRs were 36% for the ITT population and 47% in the PD-L1-high group, with complete responses in 8% and 12%. ORRs were lowest with durvalumab alone, at 26% in the ITT population and 28% in the PD-L1-high group, including complete responses in 8% and 10%.

"While chemotherapy appeared particularly good at getting initial control of disease, durvalumab, and durvalumab and tremelimumab, were associated with long-term durable remissions," said Powles.

Median duration of response (DOR) in the ITT group was 9.3 months with durvalumab alone, 11.1 months with the combination, and 5.7 months with chemotherapy. In the PD-L1-high group, median DOR was 18.5 months with durvalumab, 10.0 months with the combination, and 5.8 months with chemotherapy.

Treatment-related grade 3/4 adverse events (AEs) were highest with chemotherapy (60%), followed by the combination (28%) and durvalumab alone (14%).

Serious AEs were highest with the combination (23%), followed by chemotherapy (16%) and durvalumab alone (9%), as was treatment discontinuation, at 16%, 12%, and 6%, respectively.

AEs of special interest were higher with immunotherapy (49% with the combination, 26% with durvalumab alone) than with chemotherapy (15%), and patients on immunotherapy required more systemic corticosteroid use -- 26%, 11%, and 1%, respectively.

"Overall, the immune combination was perhaps less well tolerated than durvalumab, but both seem to be easier to give than chemotherapy," said Powles.

From 2015 to 2017, the phase III DANUBE trial randomized 1,032 patients equally to durvalumab alone (1,500 mg every 4 weeks until progression), durvalumab with tremelimumab (7 mg every 4 weeks, for up to four cycles), or standard platinum-based chemotherapy (cisplatin-gemcitabine in 57%, or carboplatin-gemcitabine in 43%, for up to six cycles). Only patients with Eastern Cooperative Oncology Group performance status of 0-1 were included. Median follow-up for the trial was 41.2 months.

Patients had a median age of 67-68, about three-fourths were men, and 60% were PD-L1 high (at least 25% positive tumor cells on the VENTANA SP263 immunohistochemistry assay). A little more than half had liver or lung metastases and about 20% had metastatic disease limited to the lymph nodes.

Patients on durvalumab alone were most likely to have progressive disease as their best response (ITT 53%; PD-L1-high 51%) compared to only 18-20% of those on chemotherapy. With the combination, 42% and 34%, respectively, had progressive disease as best response.

Progression-free survival (PFS) in the ITT group was 2.3 months with durvalumab, 3.7 months with durvalumab-tremelimumab, and 6.7 months with chemotherapy. In the PD-L1-high group, PFS was 2.4 months, 4.1 months, and 5.8 months, respectively.

Comment