New data on multiple targeted agents and evolving diagnostic frameworks highlight progress in classifying and treating EGFR PACC–mutated NSCLC.
EGFR P-loop and αC-helix compressing (PACC) mutations represent a distinct, increasingly well-defined subclass of uncommon EGFR alterations in non–small cell lung cancer (NSCLC), accounting for approximately 12% of all EGFR mutations across several large genomic databases, according to a presentation by Xiuning Le, MD, PhD, during the
“We know that [patients with] PACC-[mutated disease] are not a trivial population,” Le an associate professor of thoracic/head and neck medical oncology at The University of Texas MD Anderson Cancer Center in Houston, said during the presentation. “We are starting to understand this population a little better, both on the biology front and on the therapeutic front. We look forward to seeing more drugs come through, and we are hoping that we will see more molecular diagnostic platforms adapt to this concept.”
During the presentation, Le reviewed the characteristics that distinguish patients with PACC mutations from the classical-like, T790M-like, and exon 20 loop insertion subgroups. She also highlighted data across several EGFR-directed agents that are active in this population, including a phase 2 readout (NCT05256290) of silevertinib (BDTX-1535) presented during the
PACC mutations were first classified as a distinct structural subgroup of EGFR alterations in 2021, based on their location proximal to the drug-binding pocket, where they exert a direct or indirect effect on the P-loop and/or αC-helix. Reported incidence estimates for PACC mutations across large NSCLC genomic databases include 11.6% in a Guardant Health dataset of 22,501 EGFR variants, 13.7% in a Foundation Medicine dataset of roughly 20,000 EGFR cases, and 11% to 11.7% in Guardant Inform and aggregated multi-database cohorts, with PACC mutations generally 1.5- to 2-fold more common than exon 20 insertions across these same datasets.
PACC mutations also occur more frequently as compound mutations than classical EGFR alterations do, and when co-mutated, PACC variants such as E709X and G719X are found on the same allele essentially universally in phased sequencing data. Preclinical modeling in EGFR-mutant Ba/F3 cells indicates that compound PACC/classical mutations behave similarly to compound PACC/PACC or single PACC mutations, with generally reduced sensitivity to first- and third-generation TKIs relative to classical EGFR mutations. The co-mutation landscape of PACC-mutated tumors also closely mirrors that of classical/classical-like and exon 20 insertion–mutant tumors, suggesting shared underlying tumor biology across these EGFR subgroups.
Silevertinib (BDTX-1535), a covalent, central nervous system (CNS)-penetrant fourth-generation EGFR TKI with broad EGFR mutation coverage, was evaluated in a phase 2 dose-expansion cohort of treatment-naive patients with EGFR non-classical mutation (NCM)–positive NSCLC (cohort 3; n = 43).² Patients received silevertinib at 200 mg once daily, and all patients underwent CNS MRI surveillance every 6 weeks during therapy.
At a median follow-up of 11.2 months, the confirmed objective response rate (ORR) was 60%, including a complete response rate of 2%, with a disease control rate (DCR) of 91%. ORR varied by mutation subgroup, at 56% for PACC mutations (n = 25), 73% for classical-like mutations (n = 15), and 69% for compound mutations (n = 16).
The preliminary median progression-free survival (PFS) was 15.2 months (95% CI, 10.8-not estimable [NE]), median duration of response (DOR) had not been reached (95% CI, 7.0-NE) with 23 patients remaining on therapy, and no patients developed de novo brain metastases during the study.
The primary end point of the ongoing cohort is ORR by blinded independent central review (BICR), with PFS, DOR, overall survival (OS), safety, and pharmacokinetics as key secondary end points.
Firmonertinib is being evaluated in the prospective phase 1b FURTHER dose-expansion study (NCT05364073), reported as the first prospective trial to use PACC as a formal classifier for enrollment.3 The study is randomly assigning treatment-naive patients with EGFR PACC–mutated NSCLC to 160-mg or 240-mg daily doses, with ORR by BICR as the primary end point.
Reported data with the 240-mg dose (n = 22) showed a confirmed ORR of 68.2% (95% CI, 40.7%-82.8%) and a DCR of 100% (95% CI, 84.6%-100%), compared with a confirmed ORR of 34.8% (95% CI, 16.4%-57.3%) with the 160-mg dose (n = 23).3 The median PFS was 16.0 months (95% CI, 12.5-not available [NA]) with the 240-mg dose vs 11.1 months (95% CI, 5.5-NA) with the 160-mg dose (HR, 0.728; 95% CI, 0.305-1.736; P = .4672).
Enozertinib (ORIC-114) was evaluated in a phase 1/2 study (NCT05315700) of patients with PACC-mutated NSCLC receiving an 80-mg effective dose (n = 22), with a best ORR of 36% (95% CI, 17%-59%) and DCR of 91% (95% CI, 71%-99%). Among the subset of patients with baseline CNS disease (n = 13), the best ORR was 31% (95% CI, 9%-61%).4
Longer-term follow-up of the CHRYSALIS-2 study (n = 49) of amivantamab plus lazertinib showed a median OS of 41.0 months (95% CI, 27.7-NE) at a median follow-up of 31.3 months.5 Among treatment-naive patients (n = 49), the ORR was 57% with a median PFS of 19.5 months; among previously treated patients (n = 56), the ORR was 48% with a median PFS of 7.8 months.
In the prospective phase 3 ACHILLES/TORG1834 study (jRCTs031180175), which compared afatinib with platinum-doublet chemotherapy in patients with uncommon EGFR mutations (n = 109), afatinib improved median PFS (HR, 0.421; 95% CI, 0.251-0.706; P = .0010) and ORR vs chemotherapy.6
Pooled retrospective data across mutation subtypes, including S768I, G719X, and L861Q, also suggest second-generation TKIs, such as afatinib, achieve higher ORRs and DCRs than first-generation or third-generation TKIs in this population.1
A side-by-side testing report comparison illustrated how structure-function classification is beginning to be incorporated into clinical molecular pathology output: an EGFR p.A702T classical-like alteration and an EGFR p.E709A PACC alteration, both detected via a tissue-based mutation analysis panel, were annotated with distinct structure-function classifications and differing predicted sensitivities to second- and third-generation TKIs.
Looking ahead, Le identified several open questions the field is prioritizing, including differences in outcomes between single and compound PACC mutations, the co-mutation landscape across classical, exon 20 insertion, and PACC subgroups, potential differences in tumor microenvironment and responses to immunotherapy or antibody-drug conjugates, and mechanisms of TKI resistance across these EGFR subgroups.
