With 4 Guideline-Preferred Options, CNS Activity Is Deciding Frontline ROS1+ NSCLC Therapy

ROS1 gene fusions drive an estimated 1% to 2% of advanced non–small cell lung cancer (NSCLC) cases, occurring more often in younger patients, women, and never or light smokers with adenocarcinoma histology. Although rare, the fusion is among the most actionable drivers in lung cancer: 4 tyrosine kinase inhibitors (TKIs) now carry preferred status for frontline use. With no head-to-head trial comparing crizotinib (Xalkori), entrectinib (Rozlytrek), repotrectinib (Augtyro), and taletrectinib (Ibtrozi), oncologists are left weighing central nervous system (CNS) penetration and daily tolerability against one another rather than a clear efficacy leader.

In a virtual Case-Based Roundtable event for oncologists across the Rocky Mountain and Southwest regions, Eric K. Singhi, MD, a thoracic medical oncologist at UT MD Anderson Cancer Center, reviewed efficacy, safety, and CNS data across the 4 approved ROS1 TKIs. Singhi noted that as systemic response rates have converged, CNS activity and cumulative low-grade toxicity have become the more practical differentiators in frontline selection.

CASE SUMMARY

• A 58-year-old never-smoker, presents with persistent cough, mild dyspnea, and new right-sided chest discomfort.
• CT chest: solid, 7 cm, right mid-lobe mass; + ipsilateral, 20 mm, mediastinal lymph node
• PET/CT: avid uptake detected from L3 to L5; bilateral iliac crests and ischial spine
• Brain MRI: Negative
• Biopsy: Reveals adenocarcinoma
• IHC of FFPE Tissue Specimen: + elevated ROS1 protein level (H-score >150); TTF-1+; PD-L1 TPS: 95%
• Molecular Profiling: CD74-ROS1 gene fusion; negative for other actionable mutations
• ECOG performance status: 1

Before turning to the case, polling showed most participants' patients were tested in-house, but one-fifth had been referred with no molecular testing at all. Participants agreed RNA-based next-generation sequencing remains the preferred detection method, since it catches fusion partners older platforms miss.

For the case patient, nearly half of participants (46.7%) said they would offer taletrectinib, followed by repotrectinib (26.7%), entrectinib (20%), and crizotinib (6.7%). All 4 carry preferred National Comprehensive Cancer Network (NCCN) status frontline, though the guideline favors entrectinib, repotrectinib, or taletrectinib with brain metastases, and lists crizotinib only as category 2B in its separate CNS-specific guideline. Singhi cited PROFILE 1001 (NCT00585195), which established crizotinib as the first targetable option, with a confirmed objective response rate (cORR) of 72% and a median progression-free survival (PFS) of 19.3 months (95% CI, 15.2–39.1), though limited CNS penetration has left it largely supplanted.¹

An integrated analysis of the ALKA-372-001, STARTRK-1, and STARTRK-2 trials (NCT02568267) of entrectinib reported a cORR of 68.2%—the first ROS1 TKI to show meaningful intracranial activity.² The phase 2 TRIDENT-1 trial (NCT03093116) of repotrectinib built on that CNS signal, with a cORR of 79% and a median PFS of 31.1 months (95% CI, 21.9-not evaluable).³ James Ewing, MD, of Ironwood in Tempe, Arizona, said his patient tolerated repotrectinib's efficacy well but struggled with cumulative low-grade toxicity. "It's that kind of low-grade stuff that keeps nagging my patient," Ewing said, describing dysgeusia and lightheadedness that required a dose hold. Dose reductions occurred in 38% of the TRIDENT-1 population, most often tied to dizziness and ataxia.

Taletrectinib, the newest agent, reported the highest numbers in the TKI-naive setting, with a pooled cORR of 89.8% (95% CI, 84.0–94.1) across TRUST-I and TRUST-II (NCT04395677; NCT04919811). "There's no cross-trial comparison, but numerically speaking, PFS seems to be longer with taletrectinib," said Robert Yoo, DO, of Ironwood Cancer & Research in Scottsdale, Arizona. Its toxicity profile diverged from repotrectinib's, with transaminase elevations and gastrointestinal effects more prominent than dizziness. Still, Khusroo Qureshi, MD, of Texas Oncology in Richardson, Texas, asked whether dizziness on the more CNS-penetrant agents could mask a new brain metastasis. "I think you have to have your guard up," Singhi answered, adding that he screens patients with actionable fusions for CNS disease at least every 6 to 12 months.

Asked which treatment-emergent adverse event was hardest to manage across ROS1 TKIs, 80% of participants pointed to CNS/neurologic effects, well ahead of vision disorders, hepatotoxicity, and fatigue combined. Singhi agreed low-grade dizziness, more than any single severe toxicity, most often complicates long-term management, since patients often remain on therapy for years if it controls disease.

With efficacy differences narrowing across all 4 agents, the panel's consensus was that day-to-day tolerability, not response rate, is now the more practical differentiator in frontline ROS1-positive NSCLC.

DISCLOSURES: Dr Singhi reports receiving research support from Rexanna’s Foundation for Fighting Lung Cancer; and receiving advisory board fees from AstraZeneca, Novocure, Amgen, Janssen, Regeneron, Bristol Myers Squibb, Bayer, Boehringer Ingelheim, Eli Lilly, CARIS, and Nuvation Bio.

REFERENCES

1. Shaw AT, Riely GJ, Bang YJ, et al. Crizotinib in ROS1-rearranged advanced non-small-cell lung cancer (NSCLC): updated results, including overall survival, from PROFILE 1001. Ann Oncol. 2019 Jul 1;30(7):1121-1126. doi: 10.1093/annonc/mdz131.

2. Drilon A, Siena S, Dziadziuszko R, et al. Entrectinib in ROS1 fusion-positive non-small-cell lung cancer: integrated analysis of three phase 1-2 trials. Lancet Oncol. 2020 Feb;21(2):261-270. doi: 10.1016/S1470-2045(19)30690-4.

3. Drilon A, Camidge DR, Lin JJ, et al. Repotrectinib in ROS1 Fusion-Positive Non-Small-Cell Lung Cancer. N Engl J Med. 2024 Jan 11;390(2):118-131. doi: 10.1056/NEJMoa2302299. PMID: 38197815; PMCID: PMC11702311.

4. Bazhenova L, et al. TRUST I/II: updated pooled efficacy outcomes of taletrectinib in TKI-naive ROS1+ non-small cell lung cancer. Presented at: American Association for Cancer Research (AACR) Annual Meeting; 2026. Abstract CT300.