In the recently released Phase II data from AstraZeneca (AZ)’s CONDOR trial, Imfinzi (durvalumab), a programmed cell death ligand 1 (PD-L1) inhibitor, demonstrated anti-tumour responses in PD-L1 low or negative platinum-refractory recurrent/metastatic head and neck squamous cell carcinoma (rmHNSCC) patients.
However, the dual inhibition of cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) and PD-L1, which was achieved by combining Imfinzi with AZ’s in-house CTLA-4 inhibitor, tremelimumab in the combination arm of the study, failed to improve response rates in pre-treated rmHNSCC.
AZ hopes to launch Imfinzi as both a monotherapy and a combination treatment with tremelimumab in the head and neck cancer space. Two ongoing Phase III trials, KESTREL and EAGLE, are evaluating Imfinzi and Imfinzi + tremelimumab in the 1L and 2L rmHNSCC settings, respectively, and are expected to release top-line data in H1 2018. Data from the large Phase II CONDOR trial indicated how the combination Imfinzi + tremelimumab will perform in HNSCC patients, and raised some doubts about the efficacy of the combinatorial approach being pursued by AZ to increase response rates in HNSCC patients.
Trial results and follow up
Results from the CONDOR trial were presented at the Multidisciplinary Head and Neck Cancers Symposium in Scottsdale, Arizona, US, on February 15. The Phase II CONDOR trial enrolled 267 rmHNSCC patients who had progressed after receiving platinum-based chemotherapy. All patients had low or negative PD-L1 biomarker expression in their tumours (<25% of tumour cells) and were randomised 1:2:1 to three arms: Imfinzi monotherapy (67 patients), Imfinzi + tremelimumab (133 patients), and tremelimumab monotherapy (67 patients).
After a median follow-up of 5.8 months, partial responses were observed in 17 patients. In the Imfinzi monotherapy arm, an objective response rate (ORR) of 9.2% and an overall survival (OS) of 7.6 months were recorded. A lower ORR of 7.8% and OS of 6.0 months in the Imfinzi + tremelimumab arm and an ORR of 1.6% and OS of 5.5 months in the tremelimumab monotherapy arm were reported. Although Imfinzi showed positive clinical activity alone, its combination with tremelimumab did not appear to add any additional benefit in terms of OS and ORR. This is a disappointing result for AZ, especially considering that the Phase III trial KESTREL is evaluating Imfinzi + tremelimumab in the 1L rmHNSCC setting.
Immuno-oncology developers are exploiting different combinatorial strategies to increase response rates to immunotherapies in rmHNSCC. Currently, two immune checkpoint modulators, Bristol Myers Squibb (BMS)’s Opdivo (nivolumab) and Merck and Co.’s Keytruda (pembrolizumab), both of which are targeting the programmed cell death protein 1 (PD-1) receptor, are approved for platinum-refractory rmHNSCC patients. Both these agents are expected to obtain label expansions in the 1L rmHNSCC setting in combination with BMS’ Yervoy (ipilimumab)—a CTLA-4 inhibitor that is similar to tremelimumab—and chemotherapy, respectively. As such, BMS’ Opdivo + Yervoy combination is poised to compete directly against AZ’s PD-L1 + CTLA-4 combination. Unfortunately, the recent CONDOR results have cast doubt on the strategy of combining PD-1/PD-L1 and CTLA-4 inhibitors in HNSCC. This places more pressure on the soon-to-be-unveiled data from the randomised Phase III KESTREL trial, which will truly determine whether AZ’s combination stands a chance of gaining approval in 1L rmHNSCC.
GlobalData believes that there is still hope for Imfinzi + tremelimumab in head and neck cancer, as in the KESTREL trial, patients will not be stratified according to PD-L1 status and the combination may still have a chance to prove effectiveness in the all-comers group. However, in light of CONDOR’s data, AZ’s and BMS’ combinations may be positioned at a disadvantage relative to Merck and Co’s Keytruda + chemotherapy combination in PD-L1 low or negative patients in the 1L rmHNSCC setting.
GlobalData (2018). Head and Neck Squamous Cell Carcinoma - Opportunity Analysis and Forecast to 2026, to be published