KEY POINTS
- Investigators analyzed paired pre- and post-neoadjuvant chemotherapy RNA sequencing from two public HR+/HER2− breast cancer cohorts: 16 matched pairs in GSE191127 and 14 matched pairs in an independent GSE309004 validation cohort. RSI was derived from a published ten-gene rank-based signature in which a higher value represents greater genomic radioresistance.
- In the discovery cohort, mean RSI increased from 0.602 to 0.677; 12/16 patients (75%) had an increase and 9/16 (56.3%) exceeded a 10% relative increase. The paired change within this small HR+/HER2− subset alone was not statistically significant (p=0.118).
- The direction reproduced independently: validation-cohort RSI increased from 0.489 to 0.593 (p=0.014), with 11/14 patients (78.6%) showing an increase. Across all 30 matched pairs, RSI rose by a mean 0.089, with 23/30 (76.7%) increasing and the pooled paired comparison reaching p=0.005.
- Baseline and post-treatment RSI remained correlated, indicating that chemotherapy did not simply erase interpatient biological differences. A sensitivity model predicting post-treatment RSI directly confirmed baseline RSI as a significant predictor (β=0.404, p=0.023), while cohort identity was nonsignificant (p=0.545).
- In the 16-patient discovery cohort, inferred M2 macrophage fraction increased from 0.066 to 0.083 (p=0.038) after chemotherapy. Higher baseline M2 fraction was associated with lower baseline RSI (β=−4.933, p=0.009; R²=0.398) and with larger RSI change by linear regression (β=4.492, p=0.048), although the latter failed confirmation by Spearman correlation (p=0.222).
- The authors translated RSI into GARD for START-B 40.05 Gy/15 fractions, RTOG 1005 48 Gy/15 fractions and FAST-Forward 26 Gy/5 fractions. Individual GARD changes were heterogeneous, and mean paired GARD change in the discovery cohort was not significant (p=0.439); importantly, ΔGARD is mathematically determined by the RSI ratio and schedule dose rather than an independent biological measurement.
- No study endpoint tested locoregional recurrence, radiation benefit or the safety of increasing/decreasing adjuvant dose according to post-treatment RSI. The results therefore concern a genomic surrogate rather than clinically demonstrated radioresistance.
CLINICAL TAKEAWAY
If RSI is eventually used to personalize breast radiotherapy, measuring it on the pretreatment biopsy may not fully represent the residual disease actually being irradiated after neoadjuvant chemotherapy. The replicated directional change is interesting, but changing standard dose or fractionation on this basis would be premature.