Accurate assessment of therapy response in myelodysplastic neoplasm (MDS) has been challenging. Directly monitoring mutational disease burden may be useful, but is not currently included in MDS response criteria, and the correlation of mutational burden and traditional response endpoints is not completely understood. Here, we used genome-wide and targeted next-generation sequencing (NGS) to monitor clonal and subclonal molecular disease burden in 452 samples from 32 patients prospectively treated in a clinical trial. Molecular responses were compared with International Working Group (IWG) 2006-defined response assessments. We found that myeloblast percentage consistently underestimates MDS molecular disease burden and that mutational clearance patterns for marrow complete remission (mCR), which depends on myeloblast assessment, was not different than stable disease or bone marrow aplasia, underscoring a major limitation of using mCR. In contrast, achieving a complete remission (CR) was associated with the highest level of mutation clearance and lowest residual mutational burden in higher-risk MDS patients. A targeted gene panel approach was inferior to genome-wide sequencing in defining subclones and their molecular responses but may be adequate for monitoring molecular disease burden when a targeted gene is present in the founding clone. Our work supports incorporating serial NGS-based monitoring into prospective MDS clinical trials.

准确评估骨髓增生异常肿瘤（MDS）的治疗反应一直具有挑战性。直接监测突变疾病负担可能有用，但目前尚未纳入 MDS 缓解标准，并且突变负担与传统缓解终点的相关性尚未完全了解。在这里，我们使用全基因组和靶向下一代测序 (NGS) 来监测来自临床试验中前瞻性治疗的 32 名患者的 452 个样本的克隆和亚克隆分子疾病负担。将分子反应与国际工作组 (IWG) 2006 年定义的反应评估进行比较。我们发现，成髓细胞百分比始终低估了 MDS 分子疾病负担，并且取决于成髓细胞评估的骨髓完全缓解 (mCR) 的突变清除模式与疾病稳定或骨髓再生障碍没有什么不同，这强调了使用 mCR 的主要局限性。相比之下，在高风险 MDS 患者中，实现完全缓解 (CR) 与最高水平的突变清除和最低的残留突变负担相关。在定义亚克隆及其分子反应方面，靶向基因组方法不如全基因组测序，但当创始克隆中存在靶向基因时，可能足以监测分子疾病负担。我们的工作支持将基于 NGS 的连续监测纳入前瞻性 MDS 临床试验。