Kelly Meza, MD

Really interesting #ASTRO26 abstract from MD Anderson (Hoang et al, #1044): photon VMAT CSI vs proton CSI for leptomeningeal disease, 30 Gy/10 fx with either approach. 70 patients treated in 2021–2025: 47 proton (21 IMPT, 26 passive scatter) and 23 VMAT. Mostly breast cancer (40%) and NSCLC (31%); 93% had KPS ≥70. Median OS was 17.2 months for breast/NSCLC vs 4.1 months for other histologies. Within breast/NSCLC, modality was not significantly associated with OS after adjustment. The study also reported no significant differences in LMD progression, additional LMD-directed therapy, acute toxicity, or blood counts. VMAT started sooner: median 14 vs 20 days from consultation. It was also associated with lower odds of hospitalization than passive-scatter protons (OR 0.2, 95% CI 0.03–0.95), although not compared with IMPT. Why does this matter? The MSK randomized phase II trial compared proton CSI with photon involved-field RT. It demonstrated an important benefit, but changed both treatment volume and modality. It did not compare proton CSI with photon CSI. NRG-BN014 now tests proton CSI vs photon IFRT in a phase III trial with OS as the primary endpoint. Financial clearance for proton therapy is required before step 2 registration, and there is no photon CSI arm. I appreciate the investigators taking on a randomized trial in this difficult population. A positive result would strengthen the evidence for proton CSI over photon IFRT. But it will leave an important question unanswered: can modern photon CSI provide comparable disease control with acceptable toxicity? That matters for patients who cannot readily access a proton center, travel for treatment, or wait for authorization. The randomized comparison cannot separate the contributions of treating the entire neuraxis and using protons to do so. Marrow and other normal-tissue toxicity are legitimate concerns with photon CSI. That is precisely why comparative data are useful and prospective evaluation matters. This abstract is hypothesis-generating, however. It is retrospective, includes only 23 VMAT patients and 50 breast/NSCLC patients in the key survival comparison, and adjusts for age, sex, KPS, and histology, but not systemic therapy. Failure to detect a difference does not establish equivalence. Questions for the team: What drove selection of VMAT vs protons: insurance, urgency, travel, or clinical factors? Were VMAT plans vertebral body sparing? What were the OS HR and CI for VMAT vs protons in breast/NSCLC? Looking forward to the presentation and discussion. An important question for expanding access to effective LMD treatment. #RadOnc #OncTwitter