Future AML study analyses may focus on measurable residual disease to predict OS through non-intensive modern treatment backbones.
Evaluating measurable residual disease (MRD) after induction chemotherapy among patients diagnosed with acute myeloid leukemia (AML) may help to predict overall survival (OS), according to results from a pooled analysis presented at the
Pooled data on 1858 patients undergoing treatment across 7 randomized trials showed that the achievement of MRD negativity correlated with a risk of death that was twice as low vs patients who had MRD-positive status (P <.0001). This association between MRD and OS occurred regardless of age, genetics, and treatment arm; outcomes were consistent across trials and the use of multiparameter flow cytometry (MFC) or qPCR for NPM1 in the evaluation of MRD. Overall, the data showed a strong patient-level link between survival and MRD status.
Across the evaluable trials, MRD responses showed a strong correlation with OS benefit (R2 = 0.91; 95% CI, 0.56-1.00). As far as trial-level surrogacy was concerned, presenting author Jesse Tettero, MD, PhD, noted that the confidence interval for this finding fell below what is typically used in the FDA framework.
“The lower bound of the confidence interval was 0.56. The FDA requires the lower bound to be above 0.60, so there is some caution to be noted when interpreting these data,” Tettero, a postdoctoral associate at Fralin Biomedical Research Institute (FBRI) Cancer Research Center in Washington, DC, and a visiting hematology fellow at Amsterdam University Medical Center in the Netherlands, said in the presentation.
However, findings showed an even stronger relationship between MRD responses and survival among patients who did not undergo a transplant (R2 = 0.99; 95% CI, 0.94-1.00). Tettero described how such agreement was rare for oncologic end points.
“We could identify effective therapies [with MRD] earlier than OS; that could be a benefit of 4 to 5 years. It therefore enables potential accelerated approval…only if the FDA truly regards MRD as a surrogate end point,” Tettero said. “When supported, it could lead to smarter and faster trials for [patients with] AML.”
Despite the advent of new treatment modalities such as FLT3 and IDH inhibitors, Tettero stated that AML remains a difficult-to-cure disease, as most patients will relapse following treatment. Additionally, he noted how the longer follow-up associated with OS end points in clinical trials may delay progress for therapeutic advances. Given that MRD is already used as a surrogate marker for approval in other blood cancers like acute lymphoblastic leukemia and multiple myeloma, Tettero and colleagues aimed to determine whether MRD could represent an earlier predictor of survival in AML.
As part of the largest collaboration to evaluate MRD as a surrogate end point in AML, investigators collected patient-level data from 7 randomized trials. These studies included 4 from European cooperative groups: the German-Austrian AML Study Group (AMLSG), the Dutch-Belgian Hemato-Oncology Cooperative Group and Swiss Group for Clinical Cancer Research (HOVON-SAKK), Study Alliance Leukemia (SAL), and the UK National Cancer Research Institute (UK-NCRI). The study compared MRD status with long-term OS based on FDA expectations.
Eligible studies were those that randomly assigned patients to experimental or placebo treatments in combination with standard intensive induction chemotherapy, with at least 20 patients on each arm and the inclusion of MRD subgroups.
Regarding study limitations, Tettero stated that the analysis only related to trials assessing intensive therapy in the EU, as less intensive regimens such as venetoclax (Venclexta) plus hypomethylating agents (VEN-HMA) were not represented, and the confidence intervals for trial-led surrogacy were wide, as the lower bound was below the 0.6 threshold typically used in regulatory framework. Additionally, the study did not fully standardize the use of MRD assays across clinical trials, which may have correlated with data heterogeneity. According to Tettero, several relevant MRD datasets were also inaccessible for the analysis.
Looking towards the future, Tettero and colleagues aim to expand analyses that account for modern, non-intensive treatment backbones and regimens like VEN-HMA and FLT3 or IDH inhibitors. The investigators also look to achieve global harmonization of MRD testing across these treatment modalities while promoting data sharing on an international level to expand the number of eligible trials.
Disclosures: Tettero had no relevant relationships to disclose.
