Sacituzumab govitecan versus datopotamab deruxtecan for chemotherapy-pretreated HR-positive, HER2-negative metastatic breast cancer: a systematic review and indirect treatment comparison

In chemotherapy-pretreated hormone receptor–positive, HER2-negative metastatic breast cancer (HR+/HER2 − mBC), two Trop-2–directed antibody–drug conjugates (ADCs)—sacituzumab govitecan (SG) and datopotamab deruxtecan (Dato-DXd)—have been approved for overlapping indications, yet clinicians lack guidance on which to prioritize: the two have never been compared head-to-head, and no such trial is currently registered. Their pivotal trials share a common chemotherapy comparator, making indirect comparison formally feasible, but published syntheses have not reported between-drug effect estimates; moreover, the final analysis of TROPION-Breast01 reversed the direction of its overall survival (OS) conclusion relative to the interim estimate. Following PRISMA 2020, we searched PubMed and Cochrane CENTRAL (January 2012 to December 2025) for phase III randomized trials comparing SG or Dato-DXd with investigator’s choice chemotherapy in this population. Eligibility was restricted to chemotherapy-pretreated patients; trials in chemotherapy-naïve cohorts were excluded because prior chemotherapy exposure may modify treatment effects and violate exchangeability. Only trials with peer-reviewed full reports were included, ensuring every value entering the analysis could be traced to a source stating denominator, analysis population, and ascertainment method. Bias risk was assessed with Cochrane RoB 2, indirect estimates derived via the Bucher method through the common chemotherapy comparator, and certainty rated pairwise using GRADE. Sensitivity analyses addressed three analytic decisions capable of changing the conclusion: use of crossover-adjusted Dato-DXd OS, investigator-assessed SG PFS, and fixed-effect pooling. All analyses used trial-level summary data. Three trials (1606 patients) met inclusion criteria. Both ADCs improved progression-free survival (PFS) versus chemotherapy: pooled SG HR 0.66 (95% CI 0.56–0.79; GRADE high), Dato-DXd 0.63 (0.52–0.76; moderate); indirect comparison showed no difference (HR 1.05; 0.82–1.36; p = 0.68), and substitution of investigator-assessed PFS did not alter the conclusion (HR 1.08). For OS the two diverged: SG showed benefit (pooled HR 0.74; 0.61–0.90; p = 0.002; GRADE moderate), whereas the final TROPION-Breast01 analysis showed no Dato-DXd benefit (HR 1.01; 0.83–1.22; p = 0.94). Indirect OS comparison hinged on a single analytic decision—nominally favoring SG with the unadjusted estimate (HR 0.73; 0.56–0.96; p = 0.024), but nonsignificant with the crossover-adjusted estimate (HR 0.86; 0.65–1.14; p = 0.29)—with GRADE certainty rated very low. Toxicity profiles largely did not overlap: SG had grade ≥ 3 neutropenia 58.2% and grade ≥ 3 diarrhea 8.8% versus 1.1% and 0% for Dato-DXd; Dato-DXd had grade ≥ 3 stomatitis 6.4% and any-grade ocular-surface events 43.9%. Because source reports differed in adverse-event scope and ophthalmology monitoring requirements, these safety comparisons represent upper bounds on differences. The two drugs are comparable in progression control. Their OS difference is an evidence-level difference, not a demonstrated ranking: SG has direct randomized evidence supporting an OS benefit, Dato-DXd does not, and the indirect OS comparison is insufficient to rank the two. Accordingly, until head-to-head data become available, choice should be driven by toxicity susceptibility: select Dato-DXd when myelosuppression risk is dominant and G-CSF prophylaxis is unavailable or contraindicated; select SG when mucosal or ocular vulnerability dominates, noting that SG trials did not systematically assess these two event classes and its “safer” designation applies only to the measured range. This study was not prospectively registered.