Clinical and Laboratory Features Associated with Flow Cytometric CD49f Expression in Pediatric B Cell Acute Lymphoblastic Leukemia

Introduction: B cell acute lymphoblastic leukemia (B-ALL) is the most common childhood malignancy and its diagnosis depends on flow cytometry. Aberrant antigen expression is common in B-ALL and can be useful in both diagnosis and minimal residual disease (MRD) assessment. CD49f (α6 integrin) is commonly overexpressed in lymphoblasts of B-ALL relative to benign B lymphoblasts, and research suggests a role for CD49f in central nervous system (CNS) invasion in B-ALL. We aimed to characterize clinical and laboratory features associated with lymphoblast CD49f expression in pediatric B-ALL.

Methods: We reviewed clinical and laboratory features, including diagnostic and day 29 (D29) MRD bone marrow aspirate flow cytometry, for patients newly diagnosed with B-ALL at our pediatric centre using current diagnostic flow cytometry panels implemented in 2013. CD49f positivity (CD49f+) was defined as greater than 20% of leukemic blasts expressing CD49f at fluorescence intensity greater than the upper limit of an internal lymphocyte population. MRD studies were considered positive with leukemic blasts ≥0.01% of mononuclear cells. Frequency of CD49f+ and CD49f median fluorescence intensity (MFI) were compared between groups defined by recurrent cytogenetic abnormalities, molecular abnormalities detected by comparative genomic hybridization, CNS involvement at diagnosis, and occurrence of relapse by Chi-square, Fisher exact, Kruskal-Wallis, and Mann-Whitney testing.

Results: 158 patients were reviewed with median age at diagnosis of 56 months (range 4 to 225 months) and median 33 months follow-up (range 1 to 67 months). 82 patients had CD49f+ B-ALL (median 62.2% CD49f+ blasts, CD49f MFI 3.4) and 76 patients had CD49f negative (CD49f-) B-ALL (median 3.7% CD49f+ blasts, CD49f MFI 0.8). CD49f+ and CD49f- cases did not differ significantly in median age at diagnosis (CD49f+, 51 months; CD49f-, 64.5 months) or presenting white blood cell count (CD49f+, 8.7x10⁹/L; CD49f-, 8.8x10⁹/L). 24 patients had CNS involvement at diagnosis, with which CD49f expression status was not significantly associated (CD49f+, 17.1%; CD49f-, 13.4%). Median CD49f MFI (1.8 CNS negative; 1.3 CNS positive) and proportion of CD49f+ leukemic blasts (27.8% CNS negative; 15.8% CNS positive) did not significantly differ between patients presenting with or without CNS involvement.

Recurrent cytogenetic abnormalities associated significantly with CD49f status (p=0.001): CD49f+ associated with higher frequency of ETV6-RUNX1 (36.6% of CD49f+ and 11.8% of CD49f-; p=0.0004) and lower frequency (without statistical significance) of KMT2A rearrangement, hypodiploidy (each 1.2% CD49f+; 6.6% CD49f-), intrachromosomal amplification of chromosome 21, and TCF3-PBX1 (neither of which occurred in patients with CD49f+). Copy number abnormalities detectable by comparative genomic hybridization in CDKN2A, CDKN2B, PAX5, IKZF1, BTG1, ERG, TP53, RB1, EBF1, and CRLF2 did not differ in frequency between patients with CD49f+ and CD49f-.

D29 MRD data were available for 154 patients, 32 of whom had positive MRD results. Frequency of D29 MRD positivity was not significantly associated with CD49f expression status at diagnosis (19.7% CD49f+; 21.9% CD49f-). Relapse occurred in ten patients and its occurrence was not associated with CD49f expression status at diagnosis (4.9% CD49f+; 7.9% CD49f-). Nine patients had evaluable relapse flow cytometry data, with three demonstrating increased proportions of CD49f+ blasts (1.5- to 3.2-fold increase from diagnosis), two demonstrating similar proportions, and three demonstrating decreased proportions of CD49f+ blasts (1.9- to 25-fold decrease from diagnosis). Two had new CNS involvement at relapse (one patient with 27% CD49f+ blasts at diagnosis and 1.1% at relapse; one patient with 93.3% CD49f+ blasts at diagnosis and 99.5% at relapse) and one had CNS involvement at diagnosis without involvement at relapse (with 12.8% CD49f+ blasts at diagnosis and 12.1% at relapse).

Conclusion: Using the aforementioned criteria for flow cytometry data analysis, CD49f positivity in pediatric B-ALL was not associated with increased frequency of CNS involvement at diagnosis, day 29 MRD positivity, or relapse. CD49f expression patterns changed inconsistently in patients experiencing relapse and positivity or negativity at diagnosis or relapse was not clearly associated with CNS involvement at relapse.