In a 10 to 2 vote, the FDA’s Oncologic Drug Advisory Committee (ODAC) voted that the progression-free survival (PFS) of sotorasib (Lumakras) cannot be reliably interpreted vs docetaxel in the CodeBreaK 200 (NCT04303780) study for adult patients with KRAS G12C-mutated locally advanced or metastatic non–small cell lung cancer (NSCLC), as determined by an FDA-approved test, who have received at least 1 prior systemic therapy.¹

While the applicant proposed converting the supplemental new drug application (sNDA) to full approval based on the confirmatory study, the ODAC committee considered the results, and met to discuss the benefit-risk profile of sotorasib for this patient population.

Instead of voting on whether CodeBreaK 200 should be used to convert accelerated approval to sotorasib to traditional approval, the committee voted on whether the primary end point of PFS per blinded independent central review could be reliably interpreted in CodeBreaK 200. The committee ultimately decided that the PFS could not be reliably interpreted.

“The question before the committee today was not one of the efficacy of sotorasib in lung cancer, but rather, specifically, the ability to interpret data from a relatively small clinical trial conducted with a highly anticipated agent in a hyper information age where both patients and providers had high expectations,” said Ravi Madan, MD, senior clinician at the National Cancer Institute, during the meeting.

lung cancer: © Vector Market - stock.adobe.com

Behind the FDA Approval of Sotorasib

In 2021, the FDA granted an accelerated approval to sotorasib for the treatment of adult patients with KRAS G12C-mutated locally advanced or metastatic NSCLC based on findings from the phase 2 CodeBreaK 100 trial (NCT03600883).²

Results showed that the confirmed overall response rate (ORR) was 36% (95% CI, 28-45), and the disease control rate (DCR) was 81% (95% CI, 73-87) among patients given sotorasib. For all responders, the median best reduction in tumor size was 60%, the median time to objective response was 1.4 months, and the median duration of response (DOR) was 10 months.^2,3 Additionally, at a median follow-up of 12.2 months, the median PFS was 6.8 months.

For safety, treatment-related adverse events (TRAEs) were predominantly grade 1 or 2 with sotorasib, and there were no treatment-related deaths reported in the study. TRAEs grade 3 in severity occurred in 25 (19.8%) patients, and 1 patient (0.8%) had a grade 4 TRAE.

In addition to CodeBreaK 100, sotorasib has demonstrated a favorable benefit/risk profile across multiple studies in NSCLC and other tumor types, including colorectal cancer, and several studies of sotorasib are ongoing and investigating the agent in combination with other treatments across a number of advanced solid tumors, including CodeBreaK 101 trial (NCT04185883) and CodeBreaK 202 (NCT05920356).

Applicant Stance

The applicant believes that there is data to address the FDAs questions and concerns, and analyses confirm the benefit of sotorasib.

“Treatment with sotorasib results in improved progression-free survival over docetaxel and results in rapid and durable response. Sotorasib exhibits a differentiated safety profile as compared with docetaxel, [and] CodeBreaK 200 can be reliably interpreted to confirm the clinical benefit of sotorasib,” said Jackie Kline, PhD, vice president, global regulatory affairs, Amgen Inc., during the meeting.

During the meeting, CodeBreaK 200 data were provided in order to show support for sotorasib in this space.

CodeBreaK 200

The confirmatory CodeBreaK 200 trial is a phase 3 study which enrolled patients with locally advanced/unresectable or metastatic KRAS G12C–mutated NSCLC who received 1 or more previous treatments, including platinum-based chemotherapy and a checkpoint inhibitor, and an ECOG performance status of 0 or 1.4 A total of 345 patients were randomly assigned in a 1:1 ratio to receive oral sotorasib at a daily dose of 960 mg (n = 171) or intravenous docetaxel at 75 mg/m² every 3 weeks (n = 174).

Investigators assessed the primary end point of PFS by blinded independent central review (BICR) by RECIST v1.1 criteria in the intention-to-treat (ITT) population, and key secondary end points of overall survival, ORR, duration of response, DCR, time to response, safety, patient-reported outcomes, and pharmacokinetics. Genomic co-alterations and PD-L1 protein levels served as exploratory end points.⁴

Findings from an exploratory analysis were presented at the 2023 American Society of Clinical Oncology Annual Meeting, sotorasib vs docetaxel showed a consistent benefit among the majority of key prespecified molecularly defined subsets of patients with pretreated KRAS G12C–mutated NSCLC.

At a median follow-up of 17.7 months (interquartile range, 16.4-20.1), sotorasib yielded a 34% reduction in the risk of disease progression or death compared with docetaxel (HR, 0.66; 95% CI, 0.51-0.86; P = .0017), meeting the primary end point of the trial. With sotorasib, the median PFS was 5.6 months (95% CI, 4.3-7.8) vs 4.5 months (95% CI, 3.0-5.7) with docetaxel. PFS rates at 12 months were 24.8% and 10.1%, respectively.⁴

“Sotorasib showed significant improvement in the primary end point vs docetaxel. The PFS benefit was consistent and statistically robust between central and investigator review across subgroups and in prespecified sensitivity analysis,” said Bhakti Mehta, MD, MPH, clinical research medical director at Amgen, during the meeting.

Other key findings showed that:

• PFS benefit was observed with sotorasib vs docetaxel across all key co-alteration subgroups.
• Similar overall response benefits were observed and broadly retained for sotorasib vs docetaxel, spanning all major co-alteration defined subgroups of patients.
• The PFS benefit with sotorasib was improved vs docetaxel, independent of PD-L1 expression.
• Patients with a PD-L1 protein expression of less than 1% given sotorasib (n = 57) had a median PFS of 8.3 months (95% CI, 4.1-8.6) vs 5.9 months (95% CI, 3.5-7.2) with docetaxel (n = 55; HR, 0.66; 95% CI, 0.41-1.06; P =.06).
• Patients with a PD-L1 protein expression 1% or higher but less than 50% had a median PFS with sotorasib (n = 46) and docetaxel (n = 70) of 4.6 months (95% CI, 3.4-7.8) and 3.0 months (95% CI, 2.1-4.5), respectively (HR, 0.61; 95% CI, 0.39-0.96; P =.03).
• For patients with a PD-L1 protein expression of at least 50%, the median PFS with sotorasib (n = 60) was 5.7 months (95% CI, 4.0-10.0) vs 5.4 months (95% CI, 2.0-10.2) with docetaxel (n = 40; HR, 0.75; 95% CI, 0.44-1.23; P =.14).
• BICR-assessed ORRs with sotorasib vs docetaxel were 28.1% (95%, 21.5%-35.4%) vs 13.2% (95% CI, 8.6%-19.2%; P <.001).
• Sotorasib showed a DCR of 82.5% (95% CI, 75.9%-87.8%) vs 60.3% (95% CI, 52.7%-67.7%) with docetaxel, respectively.
• No significant differences OS rates were seen between arms at 10.6 months (95% CI, 8.9-14.0) with sotorasib and 11.3 months (95% CI, 9.0-14.9) with docetaxel, respectively (HR, 1.01; 95% CI, 0.77-1.33).

“The overall response rate, disease control rate, time to response, and duration of response were all improved with sotorasib vs docetaxel. Overall survival was similar. Patient reported outcomes favored sotorasib across a variety of measures,” added Mehta.

The safety data from CodeBreaK 200 were also consistent with the known safety profile of sotorasib.

“The safety profile of sotorasib and docetaxel are characterized by different adverse events, and sotorasib patients are less frequently bothered by their [adverse events],” said Osa Eisele, MD, MPH, therapeutic area head for oncology, global patient safety at Amgen, during the meeting.

FDA Stance

The FDAs initial assessment of CodeBreaK 200 was that data from the trial were statistically significant and met its primary end point. However, the difference in PFS was small between the sotorasib arm and docetaxel arms, and there was no difference in OS.

“The patterns of behavior and steady conduct suggested a consistent bias in favor of sotorasib and created uncertainty in our ability to interpret the results of the primary efficacy end point,” said Harpreet Singh, MD, director, division of oncology II, during the meeting.

“The FDA was contacted by the applicant several months before the final CodeBreaK 200 results were submitted to discuss results of a planned PFS interim analysis which had been narrowly flipped from a negative to a positive finding, based on applicant triggered rereads of assessments between investigators and blinded radiologists,” added Singh.

At this point, the FDA advised the applicant to not submit for regulatory review, and instead, follow the data monitoring committee's advice to continue the trial as planned. It was suggested that the sotorasib arm may have underperformed in the study or that the docetaxel arm overperformed, relative to historical data.

The FDA was also concerned about the overall integrity of the study.

Upon further review, the FDA observed that there was asymmetric early dropout among patients in the study, with greater dropout seen among patients in the docetaxel arm, there were investigator assessments of progressive disease which that were biased and which favored the sotorasib arm, and there was crossover of patients from docetaxel to sotorasib prior to the assessment of disease progression by BICR.

Other issues that were found to be related to study conduct included a lack of adherence to the imaging charter as multiple imaging assessments were conducted by the BICR to resolve discrepancies between investigator and BICR assessments. According to the FDA, this is considered a protocol violation, and generates additional concern about the quality and integrity of the data.

Overall, the FDA believes that the results from the CodeBreaK 200 trial are confounded by multiple sources of systemic bias.

“Due to censoring and early dropout confounds our ability to assess the effect and importantly, the magnitude of effect of sotorasib vs docetaxel,” added Singh.

Based on these concerns, the FDA worries whether CodeBreaK 200 can be considered an adequate and well controlled trial. Furthermore, the primary end point of PFS by BICR may not be reliable for sotorasib vs docetaxel, given its magnitude relative to the imaging interval at 5 weeks vs 6 weeks.

Decision

Overall, the ODAC committee voted 10-2 that the primary end point, PFS per BICR, of the confirmatory trial, CodeBreaK 200, can not be reliably interpreted.

“The factors that contributed to the lack of certainty come from the small size, investigator conduct, and a small, 5 week PFS benefit,” said Madan. “Industry and investigators must work together to ensure clinical trials are conducted competently so we can get the best data to advise our patients.”

REFERENCES

1. Oncologic Drugs Advisory Committee (ODAC) Meeting. FDA website. October 5, 2023. Accessed October 5, 2023. https://tinyurl.com/5994v6d3

2. FDA approves LUMAKRAS™ (sotorasib), the first and only targeted treatment for patients with KRAS G12C-mutated locally advanced or metastatic non-small cell lung cancer. News release. May 28, 2021. Accessed October 5, 2021.

3. Li BT, Skoulidis F, Falchook G, et al. CodeBreaK 100: Registrational Phase 2 Trial of Sotorasib in KRAS p.G12C Mutated Non-small Cell Lung Cancer. Presented at: International Association for the Study of Lung Cancer 2020 World Conference on Lung Cancer; January 28-31, 2021; virtual. Abstract PS01.07.

4. Skoulidis F, De Langen A, Paz-Ares LG, et al. Biomarker subgroup analyses of CodeBreaK 200, a phase 3 trial of sotorasib versus (vs) docetaxel in patients (pts) with pretreated KRAS G12C-mutated advanced non-small cell lung cancer (NSCLC). J Clin Oncol. 2023;41(suppl 16):9008. doi:10.1200/JCO.2023.41.16_suppl.9008