Real-World Use of Hypofractionated Radiotherapy for Primary CNS Tumors in the Elderly, and Implications on Medicare Spending

Abstract

Background: For elderly patients with high-grade gliomas, 3-week hypofractionated radiotherapy (HFRT) is noninferior to standard long-course radiotherapy (LCRT). We analyzed real-world utilization of HFRT with and without systemic therapy in Medicare beneficiaries treated with RT for primary central nervous system (CNS) tumors using Centers for Medicare & Medicaid Services data. Methods: Radiation modality, year, age (65–74, 75–84, or ≥85 years), and site of care (freestanding vs hospital-affiliated) were evaluated. Utilization of HFRT (11–20 fractions) versus LCRT (21–30 or 31–40 fractions) and systemic therapy was evaluated by multivariable logistic regression. Medicare spending over the 90-day episode after RT planning initiation was analyzed using multivariable linear regression. Results: From 2015 to 2019, a total of 10,702 RT courses (ie, episodes) were included (28% HFRT; 65% of patients aged 65–74 years). A considerable minority died within 90 days of RT planning initiation (n=1,251; 12%), and 765 (61%) of those received HFRT. HFRT utilization increased (24% in 2015 to 31% in 2019; odds ratio [OR], 1.2 per year; 95% CI, 1.1–1.2) and was associated with older age (≥85 vs 65–74 years; OR, 6.8; 95% CI, 5.5–8.4), death within 90 days of RT planning initiation (OR, 5.0; 95% CI, 4.4–5.8), hospital-affiliated sites (OR, 1.4; 95% CI, 1.3–1.6), conventional external-beam RT (vs intensity-modulated RT; OR, 2.7; 95% CI, 2.3–3.1), and no systemic therapy (OR, 1.2; 95% CI, 1.1–1.3; P<.001 for all). Increasing use of HFRT was concentrated in hospital-affiliated sites (P=.002 for interaction). Most patients (69%) received systemic therapy with no differences by site of care (P=.12). Systemic therapy utilization increased (67% in 2015 to 71% in 2019; OR, 1.1 per year; 95% CI, 1.0–1.1) and was less likely for older patients, patients who died within 90 days of RT planning initiation, those who received conventional external-beam RT, and those who received HFRT. HFRT significantly reduced spending compared with LCRT (adjusted β for LCRT = +$8,649; 95% CI, $8,544–$8,755), whereas spending modestly increased with systemic therapy (adjusted β for systemic therapy = +$270; 95% CI, $176–$365). Conclusions: Although most Medicare beneficiaries received LCRT for primary brain tumors, HFRT utilization increased in hospital-affiliated centers. Despite high-level evidence for elderly patients, discrepancy in HFRT implementation by site of care persists. Further investigation is needed to understand why patients with short survival may still receive LCRT, because this has major quality-of-life and Medicare spending implications.