Abstract：
OBJECTIVE: A randomized trial was conducted to compare neoadjuvant standard (S) anthracycline + ifosfamide (AI) regimen with histology-tailored (HT) regimen in selected localized high-risk soft tissue sarcoma (STS). The results of the trial demonstrated the superiority of S in all STS histologies except for high-grade myxoid liposarcoma (HG-MLPS) where S and HT appeared to be equivalent. To further evaluate the noninferiority of HT compared with S, the HG-MLPS cohort was expanded.
METHODS: Patients had localized high-grade (cellular component >5%; size ≥5 cm; deeply seated) MLPS of extremities or trunk wall. The primary end point was disease-free survival (DFS). The secondary end point was overall survival (OS). The trial used a noninferiority Bayesian design, wherein HT would be considered not inferior to S if the posterior probability of the true hazard ratio (HR) being >1.25 was <5%.
RESULTS: From May 2011 to June 2020, 101 patients with HG-MLPS were randomly assigned, 45 to the HT arm and 56 to the S arm. The median follow-up was 66 months (IQR, 37-89). Median size was 107 mm (IQR, 84-143), 106 mm (IQR, 75-135) in the HT arm and 108 mm (IQR, 86-150) in the S arm. At 60 months, the DFS and OS probabilities were 0.86 and 0.73 (HR, 0.60 [95% CI, 0.24 to 1.46]; log-rank P = .26 for DFS) and 0.88 and 0.90 (HR, 1.20 [95% CI, 0.37 to 3.93]; log-rank P = .77 for OS) in the HT and S arms, respectively. The posterior probability of HR being >1.25 for DFS met the Bayesian monitoring cutoff of <5% (4.93%). This result confirmed the noninferiority of trabectedin to AI suggested in the original study cohort.
CONCLUSIONS: Trabectedin may be an alternative to standard AI in HG-MLPS of the extremities or trunk when neoadjuvant treatment is a consideration.
METHODS: Patients had localized high-grade (cellular component >5%; size ≥5 cm; deeply seated) MLPS of extremities or trunk wall. The primary end point was disease-free survival (DFS). The secondary end point was overall survival (OS). The trial used a noninferiority Bayesian design, wherein HT would be considered not inferior to S if the posterior probability of the true hazard ratio (HR) being >1.25 was <5%.
RESULTS: From May 2011 to June 2020, 101 patients with HG-MLPS were randomly assigned, 45 to the HT arm and 56 to the S arm. The median follow-up was 66 months (IQR, 37-89). Median size was 107 mm (IQR, 84-143), 106 mm (IQR, 75-135) in the HT arm and 108 mm (IQR, 86-150) in the S arm. At 60 months, the DFS and OS probabilities were 0.86 and 0.73 (HR, 0.60 [95% CI, 0.24 to 1.46]; log-rank P = .26 for DFS) and 0.88 and 0.90 (HR, 1.20 [95% CI, 0.37 to 3.93]; log-rank P = .77 for OS) in the HT and S arms, respectively. The posterior probability of HR being >1.25 for DFS met the Bayesian monitoring cutoff of <5% (4.93%). This result confirmed the noninferiority of trabectedin to AI suggested in the original study cohort.
CONCLUSIONS: Trabectedin may be an alternative to standard AI in HG-MLPS of the extremities or trunk when neoadjuvant treatment is a consideration.
摘要：
目的：在选定的局部高风险软组织肉瘤（STS）中，进行了一项随机试验，以比较新辅助标准（S）蒽环类抗生素+异环磷酰胺（AI）方案与组织学定制（HT）方案。试验结果表明，除高级粘液样脂肪肉瘤（HG-MLPS）外，S和HT似乎相等，S在所有STS组织学中均具有优势。为了进一步评估HT与S相比的非劣效性，HG-MLPS队列扩大。
方法：患者四肢或躯干壁的局部高度MLPS（细胞成分>5％；大小≥5厘米；深坐姿）。主要终点是无病生存期（DFS）。次要终点是总生存期(OS)。试验采用了非劣效性贝叶斯设计,其中，如果真实风险比（HR）>1.25的后验概率<5％，则认为HT不低于S。
结果：从2011年5月至2020年6月，101例HG-MLPS患者被随机分配，45到HT臂和56到S臂。中位随访时间为66个月(IQR，37-89)。中值尺寸为107毫米(IQR，84-143），106mm(IQR,75-135)在HT臂和108毫米(IQR，86-150）在S臂中。60个月时,DFS和OS概率分别为0.86和0.73(HR，0.60[95%CI，0.24至1.46]；DFS的对数秩P=0.26）和0.88和0.90（HR，1.20[95%CI，0.37至3.93]；OS的对数秩P=0.77）在HT和S臂中，分别。对于DFS,HR>1.25的后验概率满足<5%(4.93%)的贝叶斯监测截止值。该结果证实了原始研究队列中提出的trabectedin对AI的非劣效性。
结论：当考虑新辅助治疗时，在四肢或躯干的HG-MLPS中，Trabectedin可能是标准AI的替代品。
方法：患者四肢或躯干壁的局部高度MLPS（细胞成分>5％；大小≥5厘米；深坐姿）。主要终点是无病生存期（DFS）。次要终点是总生存期(OS)。试验采用了非劣效性贝叶斯设计,其中，如果真实风险比（HR）>1.25的后验概率<5％，则认为HT不低于S。
结果：从2011年5月至2020年6月，101例HG-MLPS患者被随机分配，45到HT臂和56到S臂。中位随访时间为66个月(IQR，37-89)。中值尺寸为107毫米(IQR，84-143），106mm(IQR,75-135)在HT臂和108毫米(IQR，86-150）在S臂中。60个月时,DFS和OS概率分别为0.86和0.73(HR，0.60[95%CI，0.24至1.46]；DFS的对数秩P=0.26）和0.88和0.90（HR，1.20[95%CI，0.37至3.93]；OS的对数秩P=0.77）在HT和S臂中，分别。对于DFS,HR>1.25的后验概率满足<5%(4.93%)的贝叶斯监测截止值。该结果证实了原始研究队列中提出的trabectedin对AI的非劣效性。
结论：当考虑新辅助治疗时，在四肢或躯干的HG-MLPS中，Trabectedin可能是标准AI的替代品。
