KEY POINTS
- PRELUMEN included 11 prospective studies involving 428 patients with predominantly clinical T1–T2 N0–N1, estrogen-receptor-positive/HER2-negative breast cancer. Six were phase II or prospective single-arm studies, three phase I dose-escalation studies and two prospective multicenter cohorts.
- Preoperative radiotherapy regimens varied widely, from 15–38 Gy, predominantly delivered in a single fraction; eight studies used single-fraction treatment, one used three fractions and two used more conventional hypofractionation. Surgery followed radiation after intervals ranging from approximately 1 to 28 weeks across the main cohorts.
- Eight studies with 277 evaluable patients and 68 pathological complete responses contributed to the primary analysis. The pooled pCR rate was 21.4% (95% CI 12.1–35.1%), but heterogeneity was substantial (I²=77%).
- The strongest—and only statistically significant—meta-regression predictor was the radiotherapy-to-surgery interval: β +0.045 per week (95% CI +0.025 to +0.064; p<0.001), explaining 98.9% of between-study heterogeneity and reducing residual I² to 3.4%.
- By contrast, biological effective dose was not independently associated with pCR (p=0.175), nor were fractionation category (p=0.433) or technique (p=0.286). When interval and dose were entered together, interval remained significant while BED lost its apparent effect.
- Safety and cosmetic outcomes were encouraging but based on small nonrandomized studies. Pooled late grade ≥3 toxicity was 5.9% (95% CI 3.7–9.1%), with no grade ≥3 acute toxicity reported, while 88.3% (95% CI 79.4–93.6%) achieved good or excellent cosmesis.
- Seven studies reported local recurrence, with only four ipsilateral events across the available follow-up; however, median follow-up ranged from 12–66 months, studies were small and heterogeneous, and no randomized comparison establishes oncologic equivalence to standard surgery followed by adjuvant RT.
CLINICAL TAKEAWAY
The intriguing signal here is not simply the 21% pCR rate—it is that time between radiation and surgery appeared far more predictive of pathological response than radiation dose itself. Preoperative breast SBRT remains investigational, and the meta-regression is study-level rather than patient-level evidence, but timing deserves deliberate testing in future randomized trials.