Patients with pretransplant MRD may benefit from conditioning intensification

At the 46th Annual Meeting of the European Society for Blood and Marrow Transplantation (EBMT), the AML Hub will hold a satellite symposium on how measurable residual disease (MRD) status impacts the outcome of stem cell transplantation (SCT) in patients with acute myeloid leukemia (AML). Here, we present an article demonstrating that conditioning intensification for patients with AML with pretransplant MRD can reduce relapse and improve survival.

Christopher S. Hourigan and colleagues retrieved the data from the randomized phase III MAvRIC trial and analyzed patient MRD status before assignment to either myeloablative conditioning (MAC) or reduced-intensity conditioning (RIC). Their results were selected as late-breaking in the 2019 presidential plenary of the 24th European Hematology Association (EHA) meeting and were recently published in the Journal of Clinical Oncology.¹

Study design¹

This study analyzed a total of 190 pretransplant frozen whole blood samples from patients with AML enrolled in the Blood and Marrow Transplant Clinical Trials Network (BMT CTN) 0901 trial (MAvRIC, NCT01339910). All patients were in morphologic complete response at the time of sample collection and were eligible for undergoing an allogeneic SCT (allo-SCT).

All samples were analyzed by error-corrected ultra-deep next-generation sequencing (NGS) for the presence/absence of mutations in 13 genes, which are frequent aberrations in AML: ASXL1, DNMT3A, FLT3, IDH1, IDH2, JAK2, KIT, NPM1, NRAS, RUNX1, SF3B1, TET2, and TP53.

Results¹

Patient baseline characteristics were balanced between treatment arms (Table 1), even in the proportion of patients without detectable mutations before transplantation (32% vs 37% of patients treated with MAC and RIC, respectively). Interestingly, patients with NGS positivity were significantly older than those with negative NGS (57 vs 50 median years, respectively; p < 0.001), although this did not impact the study results. 

Mutations in DNMT3A, TET2, and ASXL1 (DTA mutations) were detected in 45% of patients and frequently co-occurred in the same sample. The presence of mutations in DTA genes is associated with aging and does not seem to have a prognostic impact on the clinical outcome of patients with AML. Therefore, the analysis was also performed with the alternative definition of NGS-positive status: considering mutations in all the listed genes except the DTA group. This resulted in NGS-positive classification in 41% of all patients: 42% and 40%, treated with MAC and RIC, respectively.

Table 1. Patient clinical characteristics¹

Bu, busulfan; Cy, cyclophosphamide; Flu, fludarabine; HCT-CI, hematopoietic cell transplant comorbidity index; MAC, myeloablative conditioning; Mel, melphalan; NA, not applicable; RIC, reduced-intensity conditioning; TBI, total body irradiation
Characteristic	MAC n = 95 patients	RIC n = 95 patients
Median age, years	54.9	54.7
Median disease duration, months	4.9	4.4
Cytogenetics, %
Favorable	5.3	11.6
Intermediate	62.1	54.7
Poor	27.4	29.5
Not tested or unknown	5.3	4.2
HCT-CI, %
0	34.7	34.7
1–2	34.7	31.6
> 2	30.5	33.7
Conditioning regimen, %
Flu/Bu4	56.8	NA
Bu/Cy	39.0	NA
Cy/TBI	4.2	NA
Flu/Mel	NA	17.9
Flu/Bu2	NA	82.1

Patients who received MAC experienced higher transplant-related mortality than those who received RIC, although the latter relapsed earlier (Table 2). The median follow-up in surviving patients was > 49 months; 77% of the relapses occurred in the first year after allo-SCT.

When considering the detection of mutations pretransplant, i.e., NGS positive or negative,

• patients who were NGS positive relapsed more frequently, with highest rates reported in the RIC arm
• overall survival was similar between conditioning arms for patients who were NGS negative, but those with detectable mutations pretransplant experienced a significantly higher relapse and inferior survival when treated with RIC than with MAC
  • These results were further confirmed when considering a refined definition of NGS positivity (only non-DTA mutations) (Figure 1)

Figure 1. Overall survival reported in patients negative by NGS and with DTA mutations versus patients NGS positive with non-DTA mutations, treated with MAC or RIC. Original figure provided by the first author, Christopher S. Hourigan.¹

DTA, grouped genes DNMT3A, TET2, and ASXL1; MAC, myeloablative conditioning; NGS, next-generation sequencing; RIC, reduced-intensity conditioning

Since the diagnostic sample was not available for comparison, the investigators could not assume that the detected mutations pretransplant were from the original leukemic clone. Nevertheless, they were able to describe the different prognostic impact on relapse associated with detected mutations in each arm. Excluding patients who died due to transplant-related complications,

• for patients with detectable IDH1, IDH2, SF3B1, and/or non-ITD FLT3 mutations, none of those treated with MAC relapsed, while all those who received RIC experienced a relapse
• NPM1 and FLT3-ITD mutations were associated with higher relapse rates, irrespective of the conditioning regimen

Table 2. Clinical endpoints in patients treated with MAC or RIC, according to their NGS status¹

DTA, grouped genes DNMT3A, TET2, and ASXL1; MAC, myeloablative conditioning; NGS, next-generation sequencing; OS, overall survival; RIC, reduced-intensity conditioning; TRM, transplant-related mortality *NGS positive if mutations are detected in any of the following genes: ASXL1, DNMT3A, FLT3, IDH1, IDH2, JAK2, KIT, NPM1, NRAS, RUNX1, SF3B1, TET2, and TP53.
	MAC	RIC	p value
Median time to relapse, months	8.1	3.3
1-year relapse rate, %	15	47	< 0.001
NGS positive*	14	58	< 0.001
3-year relapse rate in NGS positive for non-DTA mutations, %	15	72	< 0.001
3-year TRM, %	27	9	0.001
3-year OS, %
NGS negative	56	63	0.96
NGS positive*	61	43	0.02
NGS positive for non-DTA mutations	59	34	0.01

Conclusion

Patients with genomic detectable disease pretransplant treated with a RIC regimen experienced an increased risk of relapse and inferior overall survival compared with patients treated with MAC. The question of whether a MAC regimen should be prioritized when feasible for these patients needs to be addressed. Furthermore, patients with MRD not eligible for MAC may benefit from additional chemotherapy or novel agents pretransplant and posttransplant, to reduce their risk of early relapse.¹

On the other hand, patients with negative NGS before allo-SCT had similar relapse and survival outcomes despite the intensity of the conditioning received.¹ Ideally, MRD and genomic analyses will be incorporated in all future trials, to study the MRD dynamics and effectively identify those patients who could benefit the most from a RIC approach and be spared of the additional toxicity.²

Future studies will be needed to overcome some limitations of this analysis, such as the availability of bone marrow samples, the possibility of correlating the NGS results with the diagnostic samples, and comparing NGS with other techniques used for MRD assessment.

During the  AML Hub Satellite Symposium, five international experts—Gert Ossenkoppele, Jacqueline Cloos, Christian Thiede, Adriano Venditti, and Charles Craddock—will discuss the latest developments and remaining challenges of MRD assessments in the context of SCT. Join us on August 30, 2020 (8:30 am CEST)!