In July, AstraZeneca revealed that its Phase III trial of selumetinab – a drug being developed to treat a rare form of cancer (Uveal Melanoma) – failed to meet its primary endpoint of progression-free survival (PFS). The result has come as a blow to the pharmaceutical company after results from its Phase II study showed promising signs, causing experts to ponder whether a more defined clinical trial design would lead to a better analysis of the drug’s efficacy. The trial, named SUMIT, compared a combined form of selumetinab and dacarbazine (DTIC) with DTIC on its own. As it stands, analysis of SUMIT’s data is still ongoing.

Selumetinib, whose exclusive rights were purchased by AstraZeneca in 2003, is a MEK inhibitor that slows down RAS/RAF/MEK/MAPK signalling pathways pivotal to tumour growth in most cancers. The first positive signs of selumetinib’s effect on tumours were showcased in preclinical models of the cancer, uveal melanoma. The cancer harboured gene mutations that encode subunits for the G-protein (GNAQ and GDNA11). The mutated subunits drive the RAS/RAF/MEK/MAPK signalling pathway therefore making it vulnerable to the drug’s action. The trial results demonstrated that Selumetinib monotherapy increased PFS by two months (15.9 versus 7 weeks, hazard ratio 0.46, 95% CI 0.30-0.71; p<0.001) when compared with chemotherapy using DTIC or temozolomide.

Based on the Phase II results, which suggested that the drug could be applicable to a broader patient population, AZ chose not to restrict the Phase III SUMIT trial to only those exhibiting GNAQ/GNA11 mutation status, leading to the trial’s failure to meet its primary endpoint. Epidermal growth factor receptor (EGFR)-targeting agents Erbitux (cetuximab) and Vectibix (panitumumab) both also failed Phase III trials in gastric cancer going after enrolling all patients, rather than only those with EGFR overexpression, who would have been more scientifically suitable for treatment. However, approximately 80% of uveal melanoma patients’ tumors have mutations, so the exclusion of the minority wild-type patients could be viewed as a beneficial move if it allows sufficient efficacy to be proven.

Experts from GlobalData expect AZ to perform a retrospective subgroup analysis on uveal melanoma patients with GNAQ/GNA11 mutations in order to understand if selumetinib works better in that subset, or whether some additional pathological features are coming into play. It is notable that the selumetinib regimens investigated in the trials differed; in the Phase II study, selumetinib was tested as a monotherapy, but the drug was combined with DTIC in the SUMIT trial. This combination had not been tested in earlier trials, and it is possible that the addition of DTIC could interact with pathways in a way that negatively affects the efficacy of selumetinib. While it is true that Phase II trials are exploratory in nature, changes in design raise the risk that pivotal Phase III results will not confirm earlier successful findings.

What this means for the wider selumetinib development program still remains to be seen. A Phase III trial of selumetinib is ongoing in thyroid cancer, again without any stratification of patients for particular biomarkers. However, a further Phase III trial in non-small cell lung cancer specifically enrolling patients with KRAS mutations is also ongoing; this follows earlier successful trials in patients positive for KRAS mutation. Further, several additional studies are ongoing to assess groups with other specific molecular profiles. Biomarker assessment has potentially raised the complexity of drug development planning, especially where a range of sub-group possibilities exist. However, companion diagnostics such as biomarker assessment should also increase the chances of cost-benefit profiles being viewed more positively during pricing and reimbursement decisions.

*This article first appeared on GlobalData on 27 July 2015