Ultracentral lung SBRT produced more grade 3 toxicity than central SBRT

KEY POINTS

• This multicentre retrospective study included 133 patients from 10 centres treated with SBRT for central or ultracentral lung tumors: 103 central and 30 ultracentral. Seventy-one had early-stage NSCLC; the remainder were treated in an oligometastatic setting.
• Central tumors were located within 2 cm of major mediastinal structures, while ultracentral disease was defined by PTV overlap with the trachea, main bronchi, or esophagus. Notably, no case had the GTV directly contacting the main bronchus, so the cohort does not represent the most extreme ultracentral anatomy.
• Fractionation was individualized across centres, most commonly using 3–8 fractions with BED10 values approximately 72–115 Gy. All centres used 4DCT, an ITV approach with 3–5-mm PTV expansion, and image-guided treatment.
• Grade 2 toxicity occurred in 30.1% of central and 26.7% of ultracentral cases. Grade 3 toxicity was approximately twice as frequent in ultracentral disease: 16.7% versus 8.7%; no grade ≥4 events were reported.
• Median overall survival for the entire cohort was 24 months. Patients with a PTV <40 cm³ had longer median OS than those with PTV ≥40 cm³ (25 vs 18 months; HR 0.57, 95% CI 0.49–0.91; p=0.02).
• Concurrent immunotherapy and SBRT was the only factor independently associated with increased moderate toxicity in the multivariable model (HR 4.3, 95% CI 1.2–14.9), although systemic-therapy timing was not standardized across centres.
• In a small molecular subset, median OS was 39 months in 16 EGFR-mutant patients versus 27 months in 12 EGFR-wild-type patients (HR 0.63, 95% CI 0.43–0.88; p=0.04). This comparison is highly vulnerable to treatment-selection and systemic-therapy confounding and should remain hypothesis-generating.

CLINICAL TAKEAWAY

SOURCE

Clinical Oncology