Final results of a randomized phase II trial of second-line treatment for advanced soft tissue sarcoma comparing trabectedin, eribulin, and pazopanib: 2ND-STEP study, JCOG1802.

11508 Background: The optimal second-line treatment for advanced soft tissue sarcomas (STS) after doxorubicin therapy remains unclear. This trial aimed to identify the most promising agent among trabectedin, eribulin, and pazopanib for subsequent phase III comparison against gemcitabine plus docetaxel (GD), which has traditionally been considered a standard treatment. We previously reported the results of the primary analysis at 6 months after accrual completion. Here, we present the final analysis results with an additional year of follow-up. Methods: This multicenter, randomized, phase II selection design trial enrolled patients with unresectable or metastatic STS who were refractory to doxorubicin-based chemotherapy. Participants were randomized to receive trabectedin, eribulin, or pazopanib. The primary endpoint was progression-free survival (PFS); secondary endpoints included overall survival (OS), disease control rate (DCR), objective response rate (ORR), and safety. Results: From December 2019 to March 2023, 120 patients with advanced STS were enrolled (31 leiomyosarcomas, 26 liposarcomas, 18 translocation-related sarcomas, and 45 other types). The median PFS was 2.9 months (95% CI, 1.3-5.3) for trabectedin, 2.2 months (1.5-3.8) for eribulin, and 3.8 months (2.3-5.2) for pazopanib. The hazard ratios for eribulin and pazopanib compared to trabectedin were 1.22 (0.77-1.94) and 0.99 (0.63-1.56), respectively. When examining the PFS curves for each treatment arm, the three arms appeared to exhibit nearly identical curves; however, the PFS hazard ratio indicated that pazopanib showed the best result. The median OS was 14.7 months (10.8-18.5) for trabectedin, 13.3 months (7.6-20.1) for eribulin, and 15.7 months (9.7-18.0) for pazopanib. The DCR was 50.0% (32.4-67.6) for trabectedin, 37.1% (21.5-55.1) for eribulin, and 64.9% (47.5-79.8) for pazopanib. The ORR was 17.6% (6.8-34.5) for trabectedin, 5.7% (0.7-19.1) for eribulin, and 8.1% (1.7-21.9) for pazopanib. A histological subtype–specific subgroup analysis, demonstrated that the PFS/OS for each arm in liposarcoma were as follows: trabectedin, 3.1 months (95% CI, 1.1–7.2) / 12.9 months (3.7–36.5); eribulin, 3.8 months (1.1–6.8) / 13.3 months (4.3–20.1); and pazopanib, 5.2 months (2.0–10.2) / 18.0 months (4.1–NE). No previously unrecognized adverse events related to the investigational drugs were observed, and no treatment-related deaths occurred. Conclusions: Consistent with the primary analysis, pazopanib generally showed favorable PFS, OS, and DCR among the three agents. Based on these results, pazopanib was identified as the most appropriate candidate to be compared with GD in a phase III trial, which is currently being planned. Clinical trial information: jRCTs031190152.