ROS1 G2032R: the solvent-front mutation that breaks crizotinib

ROS1 fusions drive about 1-2% of non-small cell lung cancers, and the first-generation inhibitors hit them hard. Then, almost on schedule, the disease comes back. The most common reason is a single amino-acid substitution at codon 2032 — glycine to arginine — sitting right at the mouth of the ATP pocket. G2032R is the ROS1 equivalent of the ALK G1202R and ABL T315I gatekeeper-class problems: a small change in a structurally critical spot that converts a frontline drug into an inert passenger.

Why a residue in the solvent does so much damage

The "solvent front" is the rim of the kinase active site where the ATP pocket opens out to bulk water. Residue 2032 sits there. In wild-type ROS1 it is a glycine — small, no side chain, plenty of room. Crizotinib and entrectinib both reach a substituent out toward that rim. When glycine becomes arginine, you bolt a long, positively charged side chain onto the edge of the pocket. It is pure steric and electrostatic interference: the bulky arginine collides with the part of the drug that pokes into solvent, and the inhibitor can no longer seat properly. Catalytic activity of the kinase is barely affected, so the fusion keeps signaling while the drug is locked out.

G2032R was the first crizotinib-resistance mechanism ever reported in a ROS1-rearranged patient, and it remains the most frequently detected secondary mutation at progression. Because it is a binding-site problem rather than a bypass-pathway problem, the fix is chemical: design an inhibitor compact enough to avoid the arginine entirely.

The drugs, in order of who survives G2032R

• Crizotinib (Xalkori) — first ROS1 TKI, FDA-approved for ROS1+ NSCLC in 2016. Potent against wild-type fusion, poor CNS penetration, and knocked out by G2032R.
• Entrectinib (Rozlytrek) — approved 2019, designed for CNS activity and active against ROS1 and NTRK fusions. Better in the brain than crizotinib, but it also reaches into the solvent front and is likewise defeated by G2032R.
• Repotrectinib (Augtyro) — approved November 2024 on the TRIDENT-1 trial. A compact macrocycle deliberately engineered to tuck inside the ATP pocket without protruding toward residue 2032, so the arginine has nothing to clash with. Retains activity against G2032R and the related D2033N, and crosses into the CNS.
• Taletrectinib (Ibtrozi) — approved June 2025 on the TRUST-I and TRUST-II trials. Built specifically to cover acquired ROS1 mutations including the G2032R solvent-front substitution. Objective response rates ran around 85-90% in TKI-naive patients and roughly 50-60% in those who had already progressed on a prior ROS1 TKI.

The pattern is the same one that played out in ALK and in BCR-ABL: the first generation is potent but fragile at the solvent front, and the next generation is drawn smaller and rounder so the resistance residue has nothing to push against.

The mutation past the mutation

Solving G2032R does not close the book. The next escape hatch for ROS1 is L2086F, a substitution in the central beta-sheet 6 (Cβ6) that sits deeper in the kinase core. L2086F can blunt even repotrectinib and taletrectinib, and the emerging strategy against it is a type-switch: moving from a type-I, DFG-in binder to a type-II inhibitor such as cabozantinib that engages the kinase in its inactive conformation. It is a useful reminder that "next-generation" is always relative to a specific resistance residue, never absolute.

Try the docking yourself

Seeing why G2032R breaks crizotinib is much clearer with a structure in front of you than from a sequence. The collision is geometric, and a pose makes it obvious.

Open Studio and pick ROS1 with the G2032R mutation, then dock crizotinib and repotrectinib side by side. With molecular docking you can watch the arginine side chain crowd the crizotinib solvent-front substituent while the more compact repotrectinib stays clear of it. Running this kind of molecular docking online — wild-type versus mutant, old drug versus new — is the fastest way to build intuition for why one inhibitor keeps working and the other stops.

Primary sources

• Drilon A, et al. Repotrectinib in ROS1 fusion-positive non-small-cell lung cancer (TRIDENT-1). Reviewed in: Cancer Med (2024). PMC11473655
• Comprehensive review of ROS1 tyrosine kinase inhibitors classified by structural design and mutation spectrum (solvent-front G2032R and Cβ6 L2086F). J Thorac Oncol (2024). S1556-0864(23)02413-9
• U.S. Food & Drug Administration. FDA approves taletrectinib for ROS1-positive non-small cell lung cancer (June 11, 2025). fda.gov