Ahead of a meeting of the Oncologic Drugs Advisory Committee on Thursday, FDA reviewers raised a host of concerns about the confirmatory trial data for sotorasib (Lumakras) in previously treated non-small cell lung cancer (NSCLC).
Chief among them were whether “multiple sources of systemic bias” during the conduct of the CodeBreaK 200 trial may have tilted the results toward the KRAS inhibitor, and whether the progression-free survival (PFS) benefit over docetaxel was truly meaningful, given the lack of difference in overall survival (OS) and sensitivity analyses that revealed a potential median PFS benefit of less than a week.
The phase III study met its primary endpoint, showing a superior PFS per blinded independent central review (BICR) with sotorasib versus docetaxel, a standard second-line option in NSCLC (5.6 vs 4.5 months; HR 0.66, 95% CI 0.51-0.86). However, OS was no different between the two arms (10.6 vs 11.3 months, respectively; HR 1.01, 95% CI 0.77-1.33), with some docetaxel patients crossing over to the KRAS inhibitor.
“Upon initial receipt of this topline data, FDA noted the median PFS benefit of approximately 5 weeks was less than the standard imaging interval of 6 weeks,” FDA reviewers wrote in briefing documents released ahead of the advisory meeting. “Both the FDA and Applicant [Amgen] report that the median PFS benefit in CodeBreaK 200 could be as little as 5 days. This coupled with equivocal OS results raised questions about the clinical significance” of the confirmatory trial.
The FDA granted accelerated approval to sotorasib in 2021 based on objective response data from the single-arm CodeBreaK 100 trial, making it the first targeted agent for treating tumors with KRAS 512C mutations, which account for roughly 13% of all mutations in nonsquamous NSCLC.
As a condition of the accelerated approval, a confirmatory trial was required to confirm the benefit in this patient population.
At Thursday’s meeting, the Oncologic Drugs Advisory Committee will be asked whether the primary PFS endpoint per BICR can be “reliably interpreted” given that “multiple features of CodeBreaK 200 do not appear consistent with an adequate and well-controlled trial,” according to the agency’s reviewers.
CodeBreaK 200 randomized 345 patients with unresectable or metastatic NSCLC with KRAS 512C mutations to either sotorasib or docetaxel. All participants had to have received at least one prior line of systemic therapy. The study was conducted open-label, given that sotorasib is an oral drug while docetaxel is administered intravenously, along with the differing toxicity profiles of the two agents. After a protocol amendment, about a fourth of patients in the docetaxel arm ultimately crossed over to receive sotorasib at progression.
The initial results of the single-arm CodeBreaK 100 trial (36% response rate, median duration 10 months), which led to the accelerate approval, were met with considerable excitement in the oncology community, given that KRAS mutations were long considered undruggable. And that’s where the problems appeared to start for the phase III study, given that patients were likely hoping to receive the much-hyped drug.
During the conduct of the confirmatory trial, a high dropout rate was seen in the docetaxel arm following randomization, with 23 patients assigned to the chemotherapy never treated as compared with just two patients in the sotorasib arm.
“This asymmetric early withdrawal of patients in the docetaxel group was one of the first indications of possible systemic bias in CodeBreaK 200,” wrote FDA staff.
On further review, other signs of bias showed up: investigators appeared more likely to diagnose progressive disease earlier for patients on docetaxel when compared with BICR (69% vs 58% for the sotorasib arm), patients on docetaxel at times were switched to sotorasib before BICR assessment of progression, and violations of the imaging charter and protocol were observed.
“Adequate measures were either not put in place, or not adequately followed to minimize bias on the part of investigators, given the rates of discrepancy between investigator and BICR calls for progression,” according to the briefing document. “Investigators, and likely patients as well, were eager to access sotorasib given its early success and differing route of administration and toxicity profile.”
While the FDA is not bound to follow the advice of its advisory committees, it usually does.