NOGA: Neoadjuvant treatment optimization via MRI-guided de-escalation in stage II–III TNBC—A phase 2 study.

TPS654 Background: Pembrolizumab combined with neoadjuvant chemotherapy improves pathological complete response (pCR) rates and survival in stage II–III triple-negative breast cancer (TNBC), establishing the KEYNOTE-522 regimen as the standard of care. This regimen consists of two phases: four cycles of paclitaxel/carboplatin with pembrolizumab (TCa+P), followed by four cycles of adriamycin/cyclophosphamide with pembrolizumab (AC+P). However, anthracycline-containing chemotherapy contributes substantial short- and long-term toxicity. A subset of patients may achieve pCR after the initial TCa+P phase, raising the possibility that treatment intensity may be reduced in selected responders. Breast MRI demonstrates high concordance with pCR in hormone receptor–negative breast cancer and represents a promising, noninvasive biomarker to guide response-adapted de-escalation strategies. Methods: This is a prospective, investigator-initiated, multicenter, single-arm phase II trial evaluating MRI-guided de-escalation of neoadjuvant chemotherapy plus immunotherapy in patients with early TNBC. Eligible patients have Stage II-III (T2-T3N0 or T1-T3N1) TNBC and are candidates for the KEYNOTE-522 regimen. All patients receive four cycles (12 weeks) of TCa+P, followed by mid-treatment breast MRI. Patients achieving radiologic complete response (MRI-CR) proceed directly to surgery, omitting anthracycline-containing chemotherapy. Patients with radiologic residual disease (MRI-RD) complete the full neoadjuvant regimen including AC+P. The primary endpoint is pCR rate among patients with MRI-CR undergoing early surgery. Key secondary endpoints include recurrence-free survival, overall survival, toxicity, and patient-reported outcomes. Exploratory analyses include longitudinal circulating tumor DNA (ctDNA) dynamics to assess early molecular response, minimal residual disease, and correlates with MRI-CR, pCR, and long-term outcomes. The expected pCR rate among patients achieving MRI-CR is ≥87%, compared with a benchmark rate of 65%. The study uses a Simon’s optimal two-stage design (one-sided α=0.05, 80% power). In stage 1, 10 MRI-CR patients proceeding directly to surgery will be evaluated, with early stopping for futility if the efficacy threshold is not met. If successful, stage 2 will enroll an additional 17 MRI-CR patients, for a total of 27 evaluable patients. Assuming an MRI-CR rate of ~50%, 54 patients will be enrolled overall. The study is open for recruitment at two tertiary academic hospitals in Israel, and 9 of the planned 54 patients have been enrolled. The NOGA trial represents a novel precision oncology approach in early TNBC, with the potential to minimize toxicity without compromising efficacy in selected responders. Clinical trial information: NCT07327021 .