Adding a FAK inhibitor to garsorasib doubled the response rate, on 36 patients

Design

Multicentre phase 1b/2 study with a single-arm cohort and an open-label randomised cohort, nine hospitals in China (NCT06166836, NCT05379946)

Population

51 adults with previously treated KRAS G12C-mutated metastatic colorectal cancer, ECOG 0-1, all Asian

Primary outcome

Investigator-assessed confirmed objective response rate (RECIST 1.1)

Effect

Single-arm 46.7% (95% CI 21.3-73.4); randomised 38.9% (17.3-64.3) combination vs 16.7% (3.6-41.4) monotherapy, difference 22.2% (-7.7 to 49.1), one-sided p=0.068

KRAS G12C inhibitors work far less well in colorectal cancer than in lung cancer, and adaptive signalling through focal adhesion kinase is one proposed reason. This phase 1b/2 study across Chinese tertiary hospitals tested the focal adhesion kinase inhibitor ifebemtinib added to the KRAS G12C inhibitor garsorasib in previously treated KRAS G12C-mutated metastatic colorectal cancer, first single-arm and then randomised.

In the 15-patient single-arm cohort, confirmed objective response was 46.7% (95% CI 21.3-73.4), enough to trigger the randomised phase. There, 36 patients were assigned to the combination or to garsorasib alone: confirmed response 38.9% (17.3-64.3) versus 16.7% (3.6-41.4), a difference of 22.2 percentage points with a confidence interval from -7.7 to 49.1 and a one-sided p of 0.068. That is a doubled response rate that did not reach significance, on eighteen patients per arm.

Toxicity was comparable. Grade 3 treatment-related events occurred in 33% on the combination and 28% on monotherapy, with diarrhoea the commonest (18%), then proteinuria and intestinal obstruction. No grade 4 treatment-related events and no treatment-related deaths in any cohort.

This is hypothesis-supporting, not practice-changing, and the honest description is a signal worth a phase 3. All patients were Asian and the arms were unbalanced in age - median 51 in the combination group against 63 on monotherapy - which in a trial this small can move a response rate on its own. Neither drug is available in Indian practice; the reason to know about it is that it tells you where the field is going for a mutation you are already testing for.

• Keep testing for KRAS G12C in metastatic colorectal cancer - the therapeutic options are moving
• Do not extrapolate lung cancer KRAS G12C response rates to colorectal disease; they are much lower
• Expect diarrhoea as the dose-limiting practical problem with these combinations
• Treat a response-rate endpoint as preliminary - it is not survival
• Neither agent is available in India; refer eligible patients to trials rather than seeking access

The statistics, in plain English

A between-group difference of 22.2 percentage points with a confidence interval from -7.7 to 49.1 includes zero, so the trial cannot rule out that the combination adds nothing - the interval is that wide because there were eighteen patients per arm. The one-sided p of 0.068 is worth noticing: a one-sided test is a more permissive standard than the two-sided one usually reported, and even by that standard it did not reach 0.05. Objective response rate is also a surrogate: it measures tumour shrinkage, not how long a patient lives or feels well.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.