Varying Stereotactic Body Radiotherapy dose-fraction schema in non-small cell lung cancer and the effect on outcomes

Purpose: Stereotactic body radiotherapy (SBRT) dose-fraction (fx) schema for non-small cell lung cancer (NSCLC) varies at different centers and commonly involves 3 to 5fx. The frequency of treatment per week, or treatment schedules (TS), can also differ. We evaluated and compared the effect of different dose-fx schemas on overall survival (OS).

Methodology: We retrieved data from the National Cancer Database. Eligibility included early stage NSCLC with cT1-2 N0M0 disease treated with SBRT using 3-5f x . Characteristics were identified. Biologically effective dose (BED) and total dose (TD) were calculated. TD included: 45-60Gy for 3fx, 48-60Gy for 4fx, and 50-60Gy for 5fx. TS was classified as one of the following: daily, every other day, twice per week, or prolonged (less frequently than twice weekly). Non-prolonged TS was defined as daily, every other day, and twice per week. Kaplan Meier method was used to estimate OS and log-rank tests were used to compare the difference between groups. TS was compared pairwisely within each fx-group. Univariable Cox proportional models with robust standard errors were used to investigate the association between BED, TD, TS, fx-group and OS. Within each fx-group, BED were also assessed and compared.

Results: 3244 patients were identified, with median age 73. 48% were males and 52% females. 601 (19%) patients received 3 fx, 884 (27%) received 4fx, and 1759 (54%) received 5fx treatments. For the overall cohort, median and mean BEDs were 112 and 119, respectively. TS did not show a difference in OS among each individual fx group. Overall, prolonged vs non-prolonged TS also did not show a difference in OS (p=0.64). OS was superior in the 3fx group compared to 4 and 5fx groups, with median survival times of 61.5 vs 53.7 vs 52.6 months, respectively (p=0.046). Overall, increase in BED reduced risk of death (HR=0.97 per 10-units, CI: 0.94-0.99, p=0.04). When stratified by fx, each 10 unit rise in BED for those treated with 5 fx reduced risk of death by 8% (HR=0.92, CI: 0.87-0.97, p< 0.01). However, increasing BED was not associated with change in OS for 3 fx (HR=1.00, CI: 0.93-1.08, p=0.93) and 4 fx (HR=1.08, p=0.08) regimens. Patients with BED greater than the median (BED>112) had improved OS compared to those receiving BED < 112 (HR=0.86, p=0.047). When evaluating the 5 fx-group individually, BED >100 (median of 5fx group) had superior OS compared to < 100 (HR=0.80, p=0.02).

Conclusions: Three-fraction regimens have superior OS compared to 4 and 5 fx regimens. Higher BED improves OS in 5 fx regimens. TS variations among individual fx-groups does not impact OS.

References: