Are MRD Relapses in AML Associated with Different OS Outcomes than Morphological Relapses?

There were no significant differences in overall survival (OS) between a measurable residual disease (MRD) relapse or a morphologic relapse after an initial MRD-negative complete response (CR) in patients with acute myeloid leukemia (AML) or a high-grade myeloid neoplasm, according to a recent study.

Lauren Shih, MD, of the University of Washington, and colleagues conducted the study because the “uncertain” prognosis of MRD-only relapse makes “management of MRD challenging.”

“If [OS] after MRD relapse resembled OS after morphologic relapse, strategies to eliminate MRD may be justified; however, if the prognosis of MRD relapse was considerably better than that of morphologic relapse, elimination of MRD may be pursued more cautiously,” Dr. Shih and colleagues wrote.

They compared OS outcomes after a relapse detected first as an MRD-only relapse with outcomes after a morphologic relapse in 432 adults who initially achieved an MRD-negative CR. All patients received initial treatment at the University of Washington Fred Hutchinson Cancer Center from 2008 to 2017. The patients included those who were diagnosed with AML, a high-grade myeloid neoplasm, or chronic myelomonocytic leukemia-two.

Around two-thirds (66.6%) of patients remained in MRD-negative CR at last contact, while 10.2% developed an MRD-only relapse and 23.1% developed a morphological relapse. The median follow-up was 8.5 years among patients who were alive at the last contact.

The researchers conducted a one-year landmark analysis to stratify outcomes for patients by relapse category, including no relapse, MRD-only relapse, and morphologic relapse.

Of the 432 patients, 79 died prior to the landmark date, including 40 who were in MRD-negative CR, 12 who had an MRD-only relapse, and 27 who had a morphological relapse at the time of death. The researchers censored data from 35 patients because the last contact was prior to the landmark date. All those patients were in ongoing MRD-negative CR.

Of the 318 patients who were alive at the one-year landmark date, 83% remained MRD negative, 7% had an MRD-only relapse, and 9% had a morphological relapse. The researchers did not find any significant difference in age, gender, secondary disease, blasts <20% at diagnosis, or initial treatment intensity among these groups of patients.

An MRD-negative CR was associated with a longer OS than an MRD-only relapse or a morphological relapse in the patients who remained alive at the one-year landmark date. An MRD-negative CR remained associated with a lower risk of death than morphological relapse in a multivariable regression model (hazard ratio [HR], 0.17; 95% CI, 0.1–0.29; P<.001).

Furthermore, comparing OS outcomes among relapse types “suggests there is a decreased hazard for death” with an MRD-only relapse compared to morphological relapse “although the small sample size precludes a narrow enough confidence interval for more definitive conclusion,” the study’s authors wrote (HR, 0.52; 95% CI, 0.25–1.06; P=.072).

“In conclusion, following an initial MRD-[negative] CR in AML or high-grade myeloid neoplasm, development of either [an MRD relapse] or [a morphological relapse] were both associated with decreased OS,” Dr. Shih and colleagues concluded. “In this cohort, notably, no significant differences in terms of OS were seen between patients who presented with MRD as opposed to morphologic relapse. There were no clear predictors identified for development of MRD or morphologic relapse based on baseline characteristics.”

Reference

Shih L, Othus M, Schonhoff K, et al. Comparison between measurable residual disease relapse and morphologic relapse in acute myeloid leukemia and high-grade myeloid neoplasms. Leukemia. 2023. doi:10.1038/s41375-023-01981-2