Background
Phase III, double-blind, placebo-controlled RCT. 634 patients with operable stage II–III TNBC randomized 2:1:1. Designed to test whether adding carboplatin and/or veliparib (PARP inhibitor) to paclitaxel improved neoadjuvant pCR, and to disentangle their independent contributions.
Interventions and follow up
Arm A: Veliparib 50mg BID days 1–14 + carboplatin AUC 6 q3wk + paclitaxel 80mg/m² weekly × 12 → AC × 4 cycles
Arm B: Placebo + carboplatin AUC 6 q3wk + paclitaxel weekly × 12 → AC × 4 cycles
Arm C: Placebo + placebo + paclitaxel weekly × 12 → AC × 4 cycles
Primary endpoint: pCR (ypT0/is ypN0)
mFollow up: 4.5 years (EFS analysis)
Arm B: Placebo + carboplatin AUC 6 q3wk + paclitaxel weekly × 12 → AC × 4 cycles
Arm C: Placebo + placebo + paclitaxel weekly × 12 → AC × 4 cycles
Primary endpoint: pCR (ypT0/is ypN0)
mFollow up: 4.5 years (EFS analysis)
Results
pCR (Arm A vs C): 53.2% vs 31.0%, P<.0001
pCR (Arm B vs C): 58.2% vs 31.0%, P<.0001
pCR (Arm A vs B): 53.2% vs 58.2% — veliparib did not add to carboplatin
4-yr EFS: 67.6% (A) vs 71.2% (B) vs 64.5% (C)
pCR (Arm B vs C): 58.2% vs 31.0%, P<.0001
pCR (Arm A vs B): 53.2% vs 58.2% — veliparib did not add to carboplatin
4-yr EFS: 67.6% (A) vs 71.2% (B) vs 64.5% (C)
Adverse events
Hematologic: Grade ≥3 thrombocytopenia A 32.6%, B 27.5%, C 1.5%; grade ≥3 neutropenia A 68.5%, B 55.2%, C 35.9%.
GI: Grade ≥3 nausea higher in the veliparib arm.
Comment: Carboplatin substantially increased hematologic toxicity vs paclitaxel alone.
GI: Grade ≥3 nausea higher in the veliparib arm.
Comment: Carboplatin substantially increased hematologic toxicity vs paclitaxel alone.
Conclusions
Carboplatin significantly increased pCR rates in TNBC when added to paclitaxel/AC-based NACT (~27% absolute improvement). Veliparib did not add to carboplatin's efficacy. The pCR benefit from carboplatin did not translate into a statistically significant EFS improvement in the overall population at 4.5 years.
Key Limitations
Veliparib was used at a low dose (50mg BID — not a full PARP-inhibiting dose); higher doses may have different results. Carboplatin's pCR benefit was not designed to power EFS — BrighTNess was underpowered for long-term outcomes. The BRCA-mutated subgroup showed the highest pCR rates, consistent with synthetic lethality — veliparib may have insufficient potency to capture this effect. Modern KEYNOTE-522-era regimens already incorporate carboplatin, making BrighTNess primarily of historical/mechanistic interest.
Clinical Context
BrighTNess confirmed the pCR benefit of carboplatin in TNBC neoadjuvant therapy and established that PARP inhibition with veliparib does not add to carboplatin at the doses tested. Carboplatin + paclitaxel → AC is now incorporated into standard TNBC neoadjuvant regimens (including the KEYNOTE-522 backbone). Olaparib and talazoparib (more potent PARP inhibitors) have demonstrated benefit in different settings, but not yet in the combination neoadjuvant context tested in BrighTNess.