Sotorasib Is a 'Triumph of Drug Discovery'
in Cancer

Neil Osterweil

September 21, 2020

KRAS, one of the most frequently mutated oncogenes in human cancer, has long been thought to be "undruggable," but early results from a clinical trial of the experimental KRAS-inhibitor sotorasib (Amgen) suggest that at least one KRAS mutation common in non-small cell lung cancers (NSCLC) has a soft underbelly.

In the phase 1 CodeBreaK 100 trial, sotorasib, an investigational first-in-class inhibitor of the KRAS p.G12C mutation, showed encouraging activity against advanced NSCLC and other solid tumors.

Among patients with NSCLC, 19 (32.2%) of 59 had a confirmed objective response to sotorasib monotherapy, and 52 (88.1%) had disease control, reported David S. Hong, MD, from the University of Texas MD Anderson Cancer Center in Houston.

"Sotorasib also demonstrated durable disease control in heavily pretreated patients with non-small cell lung cancer," said Hong.

He presented secondary efficacy endpoint results from the trial in an online presentation during the European Society of Medical Oncology (ESMO) Virtual Congress 2020.

The study was also published simultaneously online in the New England Journal of Medicine.

The trial met its primary endpoint of safety of sotorasib, with no dose-limiting toxicities or treatment-related fatal adverse events, and treatment-emergent grade 3 or higher adverse events occurring in less than 20% of patients.

"The safety profile is more favorable than that of other targeted agents, and I think the reason why you have a quite safe compound here is that sotorasib is very specific in its binding to KRAS G12C, and KRAS G12C is only present in the tumor," co-investigator Marwan G. Fakih, MD, a medical oncologist at City of Hope Comprehensive Cancer Center in Duarte, California, told Medscape Medical News. Fakih was co-lead author of the report in NEJM.

A Real "Triumph"

Sotorasib is "a triumph of drug discovery," commented Colin Lindsay, MD, from the University of Manchester, UK, the invited discussant.

"We know that KRAS, over many years, over three decades, has been very difficult to target," he said.

"The early development of KRAS G12C-targeted agents is just the beginning, lending hope that the ability to target not only other KRAS mutations but also other targets previously thought to be undruggable may be within reach," write Patricia M. LoRusso, DO, from the Yale Cancer Center in New Haven, Connecticut, and Judith S. Sebolt‑Leopold, PhD, from the University of Michigan Rogel Cancer Center in Ann Arbor, in an accompanying editorial.

The KRAS, which stands for Kristen rat sarcoma viral oncogene homologue, p.G12C mutation is a glycine-to-cysteine substitution that results in the oncogene being switched on in its active form. The mutation has been identified in approximately 13% of NSCLC tumors, in 1% to 7% of colorectal cancers, and in other solid tumors.

But the mutation has been considered too difficult to target because of KRAS' strong binding affinity for guanosine triphosphate (GTP), an essential building block of RNA synthesis, and by a lack of accessible drug binding sites.

Sotorasib is a small-molecule, specific and irreversible inhibitor of KRAS that interacts with a "pocket" on the gene's surface that is present only in an inactive conformation of KRAS. The drug inhibits oncogenic signaling and tumorigenesis by preventing cycling of the oncogene into its active form, investigator Fakih explained.

Study Details

The CodeBreaK 100 investigators enrolled patients with 13 different locally advanced or metastatic solid tumor types, all bearing the KRAS p.G12C mutation.

The trial began with a dose-escalation phase, with two to four patients per cohort assigned to receive daily oral sotorasib at doses of 180, 360, 720, or 960 mg. The 960 mg dose was selected for expansion cohorts and for planned phase 2 studies, based on the safety profile and the lack of dose-limiting toxicities.

Hong and colleagues reported results for 129 patients treated in the dose-escalation and expansion cohorts, including 59 with NSCLC, 42 with colorectal cancer and 28 with other tumor types, but focused primarily on patients with NSCLC.

After a median follow-up of 11.7 months, 59 patients with NSCLC had been treated, 3 at the 180 mg dose, 16 at 360 mg, 6 at 720 mg, and 34 at 960 mg. At the time of data cutoff in June of this year, 14 patients were still on treatment and 45 had discontinued, either from disease progression (35 patients), death (5), patient request (4) or adverse events (1).

As noted, there were no dose-limiting toxicities or treatment-related fatalities reported.

Grade 3 or 4 treatment-related adverse events were reported in 18.6% of patients. The only grade 4 treatment-related event was elevated alanine aminotransferase (ALT) in one patient (1.7%). Grade 3 events included elevated liver transaminases in 9 patients, increased alkaline phosphatase in 2, anemia in 1, and increased gamma-glutamyl transferase levels, decreased lymphocyte count, hepatitis, and hyponatremia in 1 patient each.

Fakih told Medscape Medical News that given sotorasib's high degree of specificity, it's unclear what might be causing the observed adverse events.

Responses at All Dose Levels

The confirmed partial response rate was 32.2% for patients with NSCLC treated at all dose levels, and 35.3% for patients who received the 960 mg dose.

Among all NSCLC patients, and all treated at the highest 960 mg dose level, the stable disease rates were 55.9%. The respective disease control rates were 88.1% and 91.2%.

Tumor reductions occurred across all dose levels in patients with NSCLC. The median progression-free survival was 6.3 months.

Hong reported results for one patient, a 59-year-old man with the mutation who had experienced disease progression on five prior therapies including targeted agents, chemotherapy, and a checkpoint inhibitor, and had gamma-knife surgery for brains lesions.

This patient had a complete response in target lesions and a partial response overall, which included shrinkage of central nervous system metastases; he recently had progression in non-target lesions, after 1.5 years in response, Hong said.

The median duration of response was 10.9 months for patients with partial responses, and 4 months for patients with stable disease.

Hong noted that response to sotorasib was seen across a range co-mutational profiles, including several patients with four mutations in addition to KRAS p.G12C.

Other Tumors, Possible Combinations

Among 42 patients with colorectal cancers bearing the KRAS p.G12C mutation, 3 (7.1%) had a partial response. There were also partial responses seen in one patient each with melanoma and with appendiceal, endometrial, and pancreatic tumors.

"Overall, the results of this trial are very encouraging, showing the first step in 'drugging the undruggable,' '' LoRusso and Sebolt-Leopold write in their editorial.

They suggested that therapy with sotorasib may be improved by combining it with other agents that could target resistance to KRAS inhibition.

"A recent study showed that KRAS G12C colorectal cancer cells have higher basal epidermal growth factor receptor (EGFR) activity than NSCLC cells, leading to a rapid rebound in mitogen-activated protein (MAP) kinase signaling and resistance to KRAS G12C inhibition," the editorialists write. "This observation is consistent with the weaker observed clinical activity of sotorasib in patients with colorectal cancer, a problem that may be addressed by combining it with an EGFR inhibitor (e.g., cetuximab), as seen preclinically."

"I think this drug is being positioned not only in refractory disease, but we're hoping to see it move up front in non-small cell lung cancer, and we're hoping to improve its efficacy in colorectal cancer," study author Fakih told Medscape Medical News.

The study was sponsored by Amgen and by grants from the National Institutes of Health. Hong disclosed research/grant funding and an advisory/consulting role with Amgen and others. Fakih disclosed a speaking engagement for Amgen and consulting for and grant support from others. Lindsay disclosed consulting for Amgen and institutional research funding from the company and others. LoRusso disclosed fees from multiple companies, not including Amgen. Sebolt-Leopold has disclosed no relevant financial relationships.

N Engl J Med. Published online September 20, 2020. Abstract

ESMO Virtual Annual Meeting: Abstract 1257O. Presented September 20, 2020.

For more from Medscape Oncology, follow us on Twitter:  @MedscapeOnc


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.