Comparison of Acute Myeloid Leukemia Measurable Residual Disease Detection By Flow Cytometry in Peripheral Blood and Bone Marrow

BACKGROUND:

The current recommendation against the need for bone marrow aspiration (BMA) in routine follow-of persons with acute myeloid leukemia (AML) in remission preceded the recognition that multiparameter flow cytometry (MFC) is a sensitive and specific means to detect imminent morphologic relapse. Given this recognition, we wondered whether BMA is now necessary, or if concordance between MFC results in peripheral blood (PB) and BMA is such as to make BMA unnecessary, at least for evaluation of measurable residual disease (MRD) by MFC. Previous studies have demonstrated a strong correlation between disease detection by MFC in PB and BMA. Here we examined 724 paired PB and BMA samples from 482 patients to further examine the concordance between PB and BMA blast detection by MFC, particularly among patients in morphologic remission.

PATIENTS AND METHODS:

We included adults in our institutional AML database, covering 2008-2018. Our Hematopathology database was queried to identify PB and BMA MFC sample pairs with samples considered "paired" if measured within one week of each other. If an individual had multiple pairs, all were included unless otherwise specified. Ten-color MFC was performed routinely on BMA aspirates with a panel of three antibody combinations, with the same antibody combinations applied to PB samples. When identified, the abnormal population was quantified as a percentage of the total CD45⁺ white cell events. Any level of residual disease was considered positive. Complete remission (CR) and relapse were defined according to the European LeukemiaNet 2017 classification. Relationship between PB and BMA blast % was measured using Spearman's Rank-Order Correlation. Relationship between PB and BMA samples identified as positive or negative is illustrated using 2 X 2 tables (Table 1).

RESULTS:

Considering all 724 sample pairs, the Spearman correlation coefficient between PB and BMA blast percentage was 0.93, and was 0.91 considering only the first sample pair for each individual patient (n= 482). 315 sample pairs were positive by PB, 97% of which were also positive by BMA while 95% of 409 pairs negative by PB were also negative by BMA. Similar results were seen considering only a patient's first pair. Restricting analysis to patients with pairs obtained between the dates of CR and relapse, the Spearman correlation coefficient was 0.82 with 91% of 35 cases positive in PB also positive in BMA; 93% of 114 pairs negative in PB were also negative in marrow. As a complementary means to compare pairs when AML burden was low, we examined only pairs where the BMA MFC showed <5% blasts. Here, the Spearman correlation coefficient between PB and BMA blasts was 0.83. 90% of 70 positive PB cases were also positive by BMA while 95% of 295 negative PB cases were also negative by BMA. Examining pairs taken from patients in morphologic remission immediately prior to undergoing hematopoietic cell transplant yielded a Spearman correlation coefficient of 0.92, with all 9 PB positive cases also being positive in BMA and 96% of PB negative cases being negative in BMA.

CONCLUSIONS:

This is the largest cohort of AML PB and BMA sample pairs analyzed by MFC to-date. The percentages of blasts measured in PB and BMA are strongly correlated. In the 365 pairs from patients with MRD-level disease, the predictive value of PB MFC positivity for BMA positivity was 90% (63/70) while the predictive value PB MFC negativity for BMA negativity was 95%.

View largeDownload slide

View largeDownload slide

Close modal