随着FLT3抑制剂、Menin抑制剂等靶向药物不断进入急性髓系白血病（AML）治疗领域，如何优化治疗策略、克服继发性耐药并实现精准治疗，成为临床关注的重要课题。同时，MRD监测在疗效评价、移植决策及维持治疗中的价值也日益凸显。在接受《肿瘤瞭望-血液时讯》的采访中，美国杜克大学医学院Harry Paul Erba教授结合最新研究进展与临床实践，围绕FLT3突变AML的治疗人群筛选、靶向药物联合与序贯应用、耐药后的治疗选择，以及MRD指导下的个体化治疗展开深入分享，为优化AML全程管理提供了重要借鉴。
《肿瘤瞭望-血液时讯》：Erba 教授您好，将 FLT3 与 Menin 抑制剂纳入 AML 标准治疗前景广阔。请问临床上应如何精准筛选获益人群，并平衡多药联合叠加的毒性？
Harry Paul Erba教授：在强化化疗方案中联合 FLT3 抑制剂，已显著改善伴 FLT3 突变AML患者的生存结局，米哚妥林、奎扎替尼的临床数据均印证了这一点；相较于米哚妥林，吉瑞替尼亦可进一步改善患者的无复发生存。由此可见，筛选能够耐受强化化疗的患者群体，具有重要临床意义。
然而，随着去甲基化药物联合BCL-2抑制剂维奈克拉等低强度方案的应用日益广泛，许多临床医生——尤其是在美国——倾向于在本地直接采用该方案，即使患者实际适合强化疗，也不再转诊至学术中心。必须明确的是，去甲基化药物联合维奈克拉（无论是否加用奎扎替尼或吉瑞替尼）的治疗目标并非治愈，而是延长生存；该方案需作为维持治疗持续进行，除非患者后续接受异基因造血干细胞移植（allo-HSCT）。
移植决策面临多重挑战。初诊时看似适合移植的患者，可能因移植相关毒性及并发症而拒绝该方案。部分患者将移植视为“最后手段”，但临床实践表明，移植必须在缓解期实施，待疾病复发后再行移植往往难以奏效。此外，即便患者经低强度方案诱导缓解后接受移植，现有小样本数据尚未明确显示移植能为该人群带来显著的生存获益。
因此，选择低强度方案的患者群体需要被审慎评估。对于拒绝移植的患者，强化疗联合FLT3抑制剂后序贯维持治疗可能更具优势。QuANTUM-First研究显示，与单纯化疗相比，奎扎替尼维持治疗可将死亡风险降低约60%，其长期生存获益明确优于持续使用维奈克拉联合阿扎胞苷的方案。
当前血液肿瘤医生面临的核心难点，是精准甄别真正不适合强化化疗的患者。参考Ferrara 立场声明中划定的不适合强化化疗人群：年龄≥75 岁、基线体能状态不佳、合并器官功能损伤、合并其他转移性肿瘤、存在活动性感染，这类患者优先选择阿扎胞苷联合维奈克拉（联合或不联合 FLT3 抑制剂）方案。
此外，特定分子亚型患者从强化疗中获益尤为显著。携带NPM1、DNMT3A及FLT3-ITD三重突变的患者，即使在老年群体中，亦可通过"7+3"方案联合奎扎替尼获得长期生存；该策略为限时强化疗后序贯口服维持治疗，兼具疗效与可行性。
综上，临床决策应围绕两个核心问题展开：患者是否真正不适合强化疗？基于其突变谱特征，能否通过强化疗实现治愈？
Oncology Frontier-Hematology Frontier：Professor Erba, incorporating FLT3 and Menin inhibitors into the standard of care for AML holds great promise. How should clinicians precisely select the patient populations most likely to benefit, and how can the overlapping toxicities of multi-drug combinations be effectively balanced?
Professor Harry Paul Erba：Incorporating FLT3 inhibitors into intensive chemotherapy regimens has significantly improved survival outcomes in patients with FLT3-mutated AML. Clinical data from midostaurin and quizartinib have demonstrated these benefits. Compared with midostaurin, gilteritinib has also shown potential to further improve relapse-free survival. These findings underscore the clinical importance of identifying patients who are fit enough to tolerate intensive chemotherapy.
However, with the increasing use of lower-intensity regimens, such as hypomethylating agents combined with the BCL-2 inhibitor venetoclax, many clinicians—particularly in the United States—are inclined to administer these regimens locally, even when patients may actually be eligible for intensive chemotherapy. As a result, such patients are no longer referred to academic centers for further evaluation. It is important to recognize that the goal of hypomethylating agent plus venetoclax therapy, with or without quizartinib or gilteritinib, is generally to prolong survival rather than achieve a cure. These regimens typically require ongoing treatment unless patients subsequently proceed to allogeneic hematopoietic stem cell transplantation (allo-HSCT).
Transplantation decisions involve multiple challenges. Some patients who initially appear eligible for transplantation may decline it because of concerns about transplant-related toxicity and complications. Others regard transplantation as a “last resort.” However, clinical experience indicates that transplantation should ideally be performed while the disease is in remission; proceeding to transplantation after relapse is often much less successful. Furthermore, in patients who achieve remission following lower-intensity induction therapy and subsequently undergo transplantation, existing small-sample studies have not yet clearly demonstrated a significant survival benefit.
Therefore, patients being considered for lower-intensity therapy require careful evaluation. For patients who are eligible for intensive chemotherapy but decline transplantation, intensive chemotherapy combined with an FLT3 inhibitor, followed by maintenance therapy, may be a treatment strategy worth considering. The QUANTUM-First study showed that, compared with chemotherapy alone, the addition of quizartinib to intensive chemotherapy, followed by maintenance therapy, significantly reduced the risk of death. Long-term follow-up data further support the survival benefit of this treatment strategy.
One of the key challenges facing hematologists today is accurately identifying patients who are genuinely unfit for intensive chemotherapy. The Ferrara position statement outlines several clinical characteristics associated with ineligibility for intensive chemotherapy, including age ≥75 years, poor baseline performance status, impaired organ function, the presence of another metastatic malignancy, and active infection. For these patients, azacitidine plus venetoclax is generally a preferred option, with the addition of an FLT3 inhibitor considered according to the patient's molecular profile and clinical circumstances.
In addition, certain molecular subgroups may derive substantial benefit from intensive chemotherapy. For example, patients harboring concurrent NPM1, DNMT3A, and FLT3-ITD mutations, including older patients, may achieve long-term survival with a “7+3” regimen combined with quizartinib. This strategy involves a time-limited course of intensive chemotherapy followed by oral maintenance therapy, offering a potentially effective and feasible treatment approach.
In summary, AML treatment decisions should focus on two key questions: Is the patient truly unfit for intensive chemotherapy? And, given the patient's molecular mutation profile, can intensive chemotherapy offer a realistic opportunity for cure?
《肿瘤瞭望-血液时讯》：靶向治疗中继发性耐药是难点。从目前研究看，联合或顺序使用 FLT3 与 Menin 抑制剂能否有效遏制耐药克隆出现？出现复发后后续策略如何选择？
Harry Paul Erba教授：肿瘤学领域长期以来药物开发的核心原则，即将不同作用机制的药物联合使用，以通过同时靶向多个通路来降低耐药性的发生。然而，在AML中，这一策略面临特殊挑战——联合用药往往加重骨髓抑制，甚至可能导致早期死亡。因此，对于不适合强化疗的患者，盲目叠加药物可能弊大于利。
当前亟待解答的核心问题是：这些靶向药物究竟应联合使用，还是采用序贯或交替方案？遗憾的是，制药企业缺乏动力开展此类研究——若某药物仅在交替方案的某一周期中使用，很难获得监管批准。因此，这类关键问题的解答只能依赖学术机构主导的临床试验设计。
近年的研究提示，多数AML耐药来源于新克隆的出现，该克隆对原有靶向抑制剂不再敏感。我们在 QuANTUM-First研究中观察到该现象（相关数据尚未发表）；在本中心临床实践中，部分接受 7+3 联合奎扎替尼治疗后复发的患者，复查已检测不到FLT3-ITD突变。这也说明，单纯依靠现有手段阻断耐药、预防复发存在较大困难。
期待随着耐药机制研究的深入，我们可以针对性开发后续治疗手段。例如吉瑞替尼治疗后复发患者常出现 RAS 通路突变，而阿扎胞苷联合维奈克拉方案对 RAS 突变 AML 疗效有限；未来若 RAS 抑制剂研发成熟，有望用于这部分人群。总而言之，针对这些让AML逃逸靶向抑制的继发靶点，还有大量研发工作需要完成。
Oncology Frontier-Hematology Frontier：Secondary resistance remains a major challenge in targeted therapy. Based on current research, can the combination or sequential use of FLT3 and Menin inhibitors effectively prevent the emergence of resistant clones? What subsequent strategies should be chosen if relapse occurs?
Professor Harry Paul Erba：An established principle in oncology drug development is to combine agents with different mechanisms of action, thereby targeting multiple pathways simultaneously to reduce the emergence of drug resistance. However, this strategy presents unique challenges in AML, where combination therapy often exacerbates myelosuppression and may even lead to early mortality. Therefore, in patients who are unfit for intensive chemotherapy, indiscriminately adding more agents may do more harm than good.
A key question that remains to be answered is whether these targeted agents should be administered in combination or through sequential or alternating regimens. Unfortunately, pharmaceutical companies have limited incentives to conduct such studies. If a drug is used in only one cycle of an alternating regimen, it may be difficult to obtain regulatory approval. Consequently, addressing these critical questions will depend largely on clinical trials designed and led by academic institutions.
Recent studies suggest that resistance in most cases of AML arises from the emergence of new clones that are no longer sensitive to the original targeted inhibitors. We observed this phenomenon in the QUANTUM-First study, although the relevant data have not yet been published. In our own clinical practice, some patients who relapsed after receiving 7+3 chemotherapy combined with quizartinib no longer had detectable FLT3-ITD mutations upon reassessment. This highlights the considerable challenges involved in preventing resistance and relapse through currently available approaches alone.
As our understanding of resistance mechanisms deepens, we hope to develop subsequent therapies that specifically target these mechanisms. For example, relapse following gilteritinib treatment is often associated with mutations in the RAS pathway, whereas azacitidine combined with venetoclax has limited efficacy in RAS-mutated AML. In the future, as RAS inhibitors advance through development, they may offer a treatment option for this patient population. Overall, substantial research and drug development efforts are still needed to target the secondary mutations and pathways that enable AML to escape the effects of targeted inhibitors.
《肿瘤瞭望-血液时讯》：随着靶向药进入一线及维持治疗，MRD 监测尤为关键。您认为 MRD 阴性能否作为评价疗效和指导停药的标准？获得深层缓解后最佳维持治疗多长？
Harry Paul Erba教授：目前已有两项大型随机试验系统评估了基于二代测序（NGS）的FLT3-ITD特异性MRD检测，其灵敏度可达10-5。QuANTUM-First研究与MORPHO研究均证实，无论干预措施如何，MRD 阳性患者的预后与生存结局均差于 MRD 阴性患者。因此，MRD有望成为FLT3突变AML的替代终点，从而加速新药研发进程，而不必等待长期生存数据。
生存终点本身存在局限性。例如欧洲开展的吉瑞替尼对比米哚妥林的研究中，吉瑞替尼可显著改善无复发生存，但两组总生存无差异，原因在于米哚妥林组患者复发后可接受吉瑞替尼挽救治疗。由此可见，在两种药物缓解率相近的情况下，预防复发至关重要。一旦疾病复发，后续治疗需要重新启动；在毒性可接受的前提下，很少有患者会更倾向于先接受一线方案、复发后再更换挽救治疗，这一点也适用于该项吉瑞替尼对照研究。
然而，MRD作为替代终点的确立，并不意味着其已能直接指导所有治疗决策。目前的关键问题是：当患者在特定阈值以上持续MRD阳性或阴性时，临床应如何调整策略？
在NPM1突变AML中，英国AML17/AML19研究的汇总分析显示，接受两周期强化疗后达到MRD阴性的患者，在首次完全缓解期（CR1）接受异基因造血干细胞移植（allo-HSCT）并未带来OS获益。尽管该结论基于回顾性分析且诱导方案存在差异，但对于MRD阴性的NPM1突变患者，我倾向于暂缓提供移植——尤其考虑到当前针对NPM1突变复发疾病的挽救手段日益丰富，包括已获FDA批准的Menin抑制剂、维奈克拉联合方案及吉妥珠单抗奥唑米星（CD33抗体偶联药物）等。
相比之下，FLT3-ITD突变患者接受"7+3"联合奎扎替尼治疗后，MRD指导决策的难度更大。QuANTUM-First研究证实奎扎替尼可显著提高MRD阴性率，且MRD阴性与更优预后相关，但研究并未确定某一MRD阈值可作为停止治疗的依据。
MRD在移植后维持治疗中的指导价值最为明确。MORPHO研究证实，移植前 MRD 阳性的患者，接受吉瑞替尼维持对比安慰剂，无复发生存与总生存均可获益。这提示，若移植后患者经FLT3抑制剂维持治疗并持续MRD阴性达一年，考虑到移植物抗白血病（GVL）效应已发挥作用，此时复发的可能性极低，可考虑提前停药；反之，若持续MRD阳性，则应继续治疗。QuANTUM-First研究中维持治疗时长被设定为144周（36个28天周期），但临床实践中可依据MRD动态监测做出更个体化的决策。目前欧洲白血病网（ELN）正在制定相关指南，强烈推荐以MRD检测指导移植后FLT3抑制剂的使用。
Oncology Frontier-Hematology Frontier：As targeted agents enter first-line and maintenance therapies, MRD monitoring becomes increasingly critical. Do you think MRD negativity can serve as a benchmark for evaluating efficacy and guiding treatment discontinuation? What is the optimal duration of maintenance therapy after achieving a deep remission?
Professor Harry Paul Erba：Two large randomized trials have systematically evaluated next-generation sequencing (NGS)-based, FLT3-ITD-specific measurable residual disease (MRD) assays, with a sensitivity of up to 10⁻⁵. Both the QUANTUM-First and MORPHO trials demonstrated that, regardless of the intervention received, patients who were MRD-positive had worse prognoses and survival outcomes than those who were MRD-negative. Therefore, MRD has the potential to serve as a surrogate endpoint in FLT3-mutated AML, potentially accelerating drug development without the need to wait for long-term survival data.
Survival endpoints themselves have limitations. For example, in a European study comparing gilteritinib with midostaurin, gilteritinib significantly improved relapse-free survival (RFS), but no difference in overall survival (OS) was observed between the two groups. This was likely because patients in the midostaurin arm could receive gilteritinib as salvage therapy following relapse. This highlights the importance of preventing relapse, particularly when two drugs achieve similar remission rates. Once the disease relapses, subsequent treatment must be initiated, and few patients would prefer to receive one first-line regimen only to switch to salvage therapy after relapse, provided the toxicity profile is acceptable. This consideration also applies to the gilteritinib-versus-midostaurin study.
However, establishing MRD as a surrogate endpoint does not mean that it can already directly guide every treatment decision. The key question now is how clinicians should adjust treatment strategies when patients remain persistently MRD-positive or MRD-negative above or below a defined threshold.
- In NPM1-mutated AML: A pooled analysis of the UK AML17 and AML19 trials showed that patients who achieved MRD negativity after two cycles of intensive chemotherapy did not derive an OS benefit from allogeneic hematopoietic stem cell transplantation (allo-HSCT) in first complete remission (CR1). Although this conclusion was based on a retrospective analysis and the induction regimens varied, I would be inclined to defer transplantation in patients with NPM1-mutated AML who are MRD-negative, particularly given the expanding range of salvage treatment options for relapsed NPM1-mutated disease. These include the FDA-approved menin inhibitors, venetoclax-based combinations, and gemtuzumab ozogamicin, an anti-CD33 antibody-drug conjugate (ADC).
- In FLT3-ITD-mutated AML: MRD-guided decision-making is more challenging in patients treated with “7+3” chemotherapy combined with quizartinib. The QUANTUM-First trial demonstrated that quizartinib significantly increased MRD negativity rates and that MRD negativity was associated with better outcomes. However, the study did not establish a specific MRD threshold that could be used as a basis for discontinuing treatment.
The role of MRD in guiding post-transplant maintenance therapy is currently the clearest. The MORPHO trial demonstrated that, among patients who were MRD-positive before transplantation, gilteritinib maintenance was associated with improvements in both RFS and OS compared with placebo. This suggests that if a patient remains MRD-negative for one year while receiving post-transplant FLT3 inhibitor maintenance, treatment discontinuation may be considered, given that the graft-versus-leukemia (GVL) effect is likely to have taken hold and the risk of relapse is very low. Conversely, if MRD remains detectable, treatment should be continued. In the QUANTUM-First trial, maintenance therapy was prescribed for 144 weeks (36 cycles of 28 days each). In clinical practice, however, treatment duration may be individualized based on serial MRD monitoring. The European LeukemiaNet (ELN) is currently developing relevant guidelines and strongly recommends using MRD testing to guide post-transplant FLT3 inhibitor therapy.
专家简介
Harry Paul Erba教授
美国杜克大学医学院
美国杜克大学医学院医学系血液恶性肿瘤与细胞治疗部医学教授
杜克癌症研究所白血病项目主任
自2012年起担任SWOG白血病委员会主席
自2012年起担任美国国家癌症研究所（NCI）白血病指导委员会委员
自2019年起担任NCI资助的AML和MDS精准医学计划MyeloMATCH高级科学委员会联合主席
已在同行评审期刊发表论文170余篇