Background: Murine double minute 2 (MDM2) is a negative regulator of tumor suppressor protein p53. MDM2 induces degradation of p53 and promotes tumorigenesis. MDM2 amplification occurs in many cancers but is documented in up to 100% of well-differentiated or dedifferentiated liposarcomas (WD/DDLPS) [Cancer Genome Atlas Research Network. Cell 2017]. Inhibition of the MDM2-p53 interaction is a promising therapeutic approach to restore p53 tumor suppressor activity in WD/DDLPS. Milademetan (RAIN-32) is a small-molecule MDM2 inhibitor that inhibits the MDM2-p53 interaction and restores p53 function at nanomolar concentrations. In a phase 1 study, milademetan showed promising efficacy in 53 patients with WD/DDLPS when administered on an intermittent schedule (260 mg QD on Days 1–3 and 15–17 on a 28-day cycle), with a median progression-free survival (PFS) of 7.4 months [Gounder et al. AACR-NCI-EORTC 2020]. WD/DDLS are relatively resistant to chemotherapy, and systemic treatment options for patients with advanced disease are limited. MANTRA (RAIN-3201) is a randomized, multicenter, open-label, phase 3 registration study designed to evaluate the efficacy and safety of milademetan versus trabectedin in patients with unresectable or metastatic DDLPS with disease progression on ≥ 1 prior systemic therapies. Methods: Eligible patients are ≥ 18 years of age with histologically confirmed unresectable and/or metastatic DDLPS, with or without a WD component, who have received ≥ 1 prior systemic therapies, including ≥ 1 anthracycline-based regimen, with radiographic evidence of progression by RECIST v1.1 within 6 months before study entry. Prior treatment with trabectedin or an MDM2 inhibitor is not permitted. Patients will be randomly assigned (1:1) to receive milademetan (260 mg once daily orally Days 1–3 and 15–17 on a 28-day cycle) or trabectedin (1.5 mg/m² as a 24-hour intravenous infusion every 3 weeks). Randomization is stratified by Eastern Cooperative Oncology Group performance status (0 or 1) and number of prior treatments for WD/DDLPS (≤ 2 or > 2). Tumor response will be evaluated by RECIST v1.1 at Weeks 8, 16, 24, and 32, and then every 12 weeks. Primary endpoint: PFS by blinded independent central review. Secondary endpoints: overall survival; disease control rate; objective response rate; duration of response; PFS by investigator assessment; safety; health-related quality of life. Exploratory endpoints: molecular markers in peripheral blood and/or tumor tissue; milademetan pharmacokinetics. To demonstrate a 3-month increase in PFS (from 3 to 6 months) corresponding to a hazard ratio of 0.5, approximately 160 patients will be required to observe 105 events with 93.9% power and 2-sided significance level of 5%. Clinical trial information: NCT04979442.