Conference Coverage

MRD better measure of ALL remission than morphology



– In children with acute lymphoblastic leukemia, minimal residual disease findings appear to be better at defining remission than morphology, Children’s Oncology Group investigators reported.

A study of outcomes of more than 9,000 children and young adults with B-lineage or T-lineage acute lymphoblastic leukemia (ALL) showed that patients who would be defined as being in remission by morphology but have minimal residual disease (MRD) of 5% or greater have survival outcomes similar to those of patients who do get a morphologic remission. Additionally, patients with discordant morphologic and MRD findings have significantly worse outcomes than do patients who were in morphologic remission and had concordant MRD findings, said Sumit Gupta, MD, PhD, from the Hospital for Sick Children in Toronto.

Dr. Sumit Gupta from the Hospital for Sick Children in Toronto

Dr. Sumit Gupta

“Minimal residual disease assessment after initial therapy is now integral to modern risk stratification in ALL, and there have been a few studies that have now shown fairly convincingly that intensification of therapy in patients with high MRD actually improves their outcomes,” he said at the annual meeting of the American Society of Pediatric Hematology/Oncology annual meeting.

“Given that, however, although MRD is used to measure the depth of remission either using flow cytometry or PCR [polymerase chain reaction]-based methods, remission itself continues to be defined by basic morphological assessment, whether that’s in clinical practice or clinical trials,” he added.

To see whether the practice of declaring remissions by morphology still makes sense, Dr. Gupta and his colleagues in the Children’s Oncology Group looked at outcomes for children and young adults with discordant ALL remissions as assessed by morphology, compared with MRD.

They looked at data on 9,350 patients from the ages of 1 to 31 years who were enrolled in one of three Children’s Oncology Group trials for patients with newly diagnosed ALL. Two of the trials (AALL0331 and AALL0232) were for patients with B-lineage ALL, and one (AALL0434) was for patients with T-lineage ALL.

They looked at morphologic responses as assessed by local centers, with M1 responses defined as less than 5% leukemic blasts (remission), M2 defined as 5% to less than 25% blasts, and M3 as 25% or more blasts. MRD was measured by flow cytometry at one of two central labs.

They found that discordant results (M1 morphology but MRD of 5% or greater) occurred in only 0.9% of patients with B-ALL, but in 6.9% of patients with T-ALL (P less than .0001).

In multivariate analysis, significant predictors of discordance in patients with B-ALL were patients age 10 years or older (P = .03), white blood cell counts of 50,000/mcL or greater (P = .005), and neutral or unfavorable cytogenetics vs. favorable (P less than .0001 for each).

Among patients with T-ALL, the only significant predictor of discordant results was the early T-precursor phenotype, with an odds ratio of 4.7 (P less than .0001).

Comparing event-free survival (EFS) between patients with concordant remission findings (M1/MRD less than 5%), they investigators saw that for patients with B-ALL, the 5-year EFS was 87%, compared with 59% for patients with discordant findings (M1/MRD 5% or greater, P less than .0001 vs. concordant remissions), and 39% for patients with concordant results showing a lack of remission (P = .009 vs. discordant findings).

Similarly, respective EFS rates for patients with T-ALL were 88%, 80% (P = .011) and 63% (not significant).

In a subanalysis of EFS by risk category, they found no differences according to concordance/discordance among patients with standard-risk B-ALL but a significant difference among patients with high-risk disease.

Attempting to determine what was driving the intermediate outcomes of patients with discordant findings, “we hypothesized that maybe it’s a difference in their actual MRD levels.” Specifically, they found that while both discordant and concordant not-in-remission patients had MRD levels of 5% or higher, the MRD levels were higher among those patients who were conclusively not in remission, Dr. Gupta said.

Finally, they found that for those patients with known overall survival data, concordant in remission patients with B-ALL had a 94% rate out to 12 years, compared with 73% for those with discordant results (P less than .0001). There was no significant difference in OS among patients with T-ALL, however.

“Should MRD assessment actually replace morphology in defining remission in subjects with ALL? I think these data strongly support that,” Dr. Gupta said.

The study was supported by the National Institutes of Health. Dr. Gupta reported having no conflicts of interest.

Next Article: