FLT3-ITD MRD is here to stay

In this issue of Blood Advances, Hu et al¹ report the results of a retrospective study of a large cohort of patients with FLT3-ITD–mutated acute myeloid leukemia (AML) who underwent allogeneic hematopoietic cell transplantation (HCT) at 3 centers in China from 2016 to 2024. The study focused specifically on FLT3-ITD measurable residual disease (MRD), using an amplicon-based approach that has been used in recent large randomized trials of FLT3 inhibitors.^2,3 The patients were in complete remission at the time of HCT, and most patients received myeloablative conditioning and stem cells from haplo-identical donors, and were also treated with FLT3 inhibitors (predominantly sorafenib) after HCT.

Using the MRD assay on banked pre-HCT samples, the authors found that 32% of patients were FLT3-ITD MRD⁺, compared with only 7.8% of patients who were identified as FLT3-ITD MRD⁺ using conventional PCR (polymerase chain reaction) amplification followed by capillary electrophoresis on the same samples. As previously observed in the MORPHO study,^3,4 the presence of pre-HCT FLT3-ITD MRD was associated with an increased risk of relapse after HCT. Also as observed in the MORPHO study, the amount of MRD detected in the assay correlated with the degree of relapse risk, in that more MRD was worse compared with less MRD. The PCR-next generation sequencing (NGS) assay outperformed flow cytometry in association with relapse risk, which was in agreement with previous studies.^5,6 Multiclonal FLT3-ITD MRD was detected in 22.5% of cases and, similar to 1 other report,⁴ was worse than MRD consisting of a single FLT3-ITD clone. Finally, also consistent with the results from the recent MORPHO study, post-HCT FLT3 inhibitor maintenance was associated with beneficial outcomes, but only among patients with peri-HCT FLT3-ITD MRD.³ All these findings amount to confirmation of previous findings and provide further evidence for the utility of FLT3-ITD AML in understanding and guiding the management of AML.

The assay used in this and several other studies has been given a few different names, from PCR-NGS to amplicon-based and ultrasensitive, the recently coined term by the European LeukemiaNet (ELN) MRD Working Group.7, 8, 9 Referring to the assay simply as NGS is misleading because conventional NGS assays used for identifying hotspot mutations in AML are orders of magnitude less sensitive than this MRD assay, and simply referring to it as NGS risks sowing confusion within the field. In the United States, most centers are using a single commercial source for this assay,¹⁰ although the methodology of this FLT3-ITD MRD assay is neither mysterious nor protected by patent. The ELN MRD Working Group recently published general guidelines for performing this assay in order to start the process of achieving harmonized global results.¹¹

MRD is now increasingly being integrated into the care of patients with AML, and the use of a high-sensitivity assay for FLT3-ITD MRD should now be considered a standard of care. When sufficient DNA is used, the assay is sensitive enough to detect the presence of mutations months before relapse.⁴ In contrast, MRD monitoring for acute promyelocytic leukemia (APL) seemed like a nice idea at the time, but because hardly anyone ever relapses with APL these days, monitoring for PML::RARα transcripts now has less impact on clinical practice.¹² Likewise, quantitative PCR assays for the fusion transcripts of core-binding factor (CBF) leukemias are clinically useful for identifying potential relapse,¹³ but one cannot always be sure whether fusion transcripts arise from a cell capable of driving relapse, and we do not actually have anything to specifically target the disease (other than some semirandomly chosen chemotherapy regimen).

For years, conventional wisdom (from those who claimed to know) was that FLT3-ITD mutations were unsuitable as markers for MRD because they were unstable. Although there is no doubt that FLT3 mutations can be absent from a relapsing AML clone, or appear when not originally detected at diagnosis,^14,15 they are also usually the final mutation in leukemogenesis and therefore represent the actual disease.¹⁶ There does not appear to be any safe level of detectable FLT3-ITD MRD,^4,6 and its unique insertion length provides a reassuring specificity. For this reason, it is now incumbent upon clinicians to know (or at least try and find out) the length of the insertion at diagnosis for any case of AML with 1 of these mutations. Because we actually have drugs to target this mutation, we can use the information provided by MRD monitoring to decide when to start or stop treatment with an FLT3 inhibitor.

The field of AML is clearly moving toward an era in which MRD monitoring will be a central guide for managing the disease. At present, we have sensitive molecular MRD assays for APL,¹² CBF AML,^13,17 NPM1-mutated AML,¹⁸ FLT3-ITD AML,¹⁰ and soon, KMT2A-rearranged AML.¹⁹ These represent roughly half of all AML cases. Here’s predicting that a sensitive and specific molecular MRD assay will be readily available for the remaining AML subtypes in the not-too-distant future.²⁰

Conflict-of-interest disclosure: M.J.L. has a consulting or advisory role in Daiichi Sankyo, Astellas Pharma, Sumitomo Pharma, Kura Oncology, Menarini, Servier, and Syndax; and received research funding from Astellas Pharma.