RAS alterations drove resistance in about one in ten oncogene-driven lung cancers progressing on targeted therapy

Newer inhibitors push resistance towards other pathways

The study starts from the observation that next-generation targeted therapies are particularly active against on-target resistance, which the authors argue favours the emergence of off-target mechanisms when patients eventually relapse.

Most patients with oncogene-driven lung cancer do relapse, despite the activity of these drugs. RAS inhibitors now in clinical development could, on the authors’ reading, have a role against resistance that runs through RAS, which is the question the study sets out to size and to test.

RAS alterations in 11.2% after first-line osimertinib, mostly KRAS G12D

Of the 312 patients with EGFR-mutant disease who progressed on first-line osimertinib, 35 carried a RAS alteration, and KRAS G12D accounted for ten of them against a single KRAS G12C mutation.

The authors report that tissue and liquid biopsies played a major complementary role in identifying resistance mechanisms in this group. The scarcity of G12C is relevant to treatment choice, because the approved KRAS inhibitors most used in lung cancer, sotorasib and adagrasib, target G12C.

In ALK-positive disease, more RAS resistance after lorlatinib

Among 148 patients with ALK-positive lung cancer, RAS alterations were present in 14.7% at resistance to the third-generation inhibitor lorlatinib, compared with 5% at resistance to second-generation ALK inhibitors (P=0.0444).

This fits the authors’ premise that a more potent on-target inhibitor leaves more room for off-target escape. The comparison is between patients who received different drugs, however, and the P value sits just under 0.05. Across MET-, ROS1- and RET-driven lung cancers at progression on targeted agents, RAS alterations were detected in 7% to 16% of cases.

Osimertinib plus a RAS inhibitor was synergistic in two patient-derived models

The investigators built the models from tumours that had acquired KRAS G12D or KRAS G12R at osimertinib progression, and tested osimertinib with the selective KRAS G12D inhibitor zoldonrasib or with the pan-RAS inhibitor daraxonrasib.

Both combinations showed marked synergistic activity in cell culture and in animal models. The evidence for the combinations is therefore preclinical and rests on two models.

A basis for clinical trials of RAS-directed combinations

The authors conclude that RAS alterations mediate resistance to targeted agents in about 10% of oncogene-driven lung cancers, and that rational combinations with the new RAS inhibitors are effective in preclinical models, providing the basis for their clinical investigation.

The distribution of alterations found here is relevant to how such trials are designed. After osimertinib, a G12D-selective agent would cover more of the patients in this series than a G12C inhibitor, and a pan-RAS inhibitor would cover alterations such as G12R that neither selective agent targets.