Robust Detection Of Minimal Residual Disease In Unselected Patients With B-Cell Precursor Acute Lymphoblastic Leukemia By High-Throughput Sequencing Of IGH

High-throughput sequencing (HTS) of immunoglobulin heavy chain genes (IGH) may be useful for detecting minimal residual disease (MRD) in acute lymphoblastic leukemia. We previously demonstrated the first application of high-throughput sequencing for the detection of minimal residual disease in T-cell precursor acute lymphoblastic leukemia (TPC-ALL) (Sci. Transl. Med. 4(134):134ra63. 2012). Recently, Faham and colleagues considered deep sequencing for MRD detection in B-cell precursor acute lymphoblastic leukemia (BPC-ALL) (Blood 120(26):5173-80, 2012). As this prior analysis in BPC-ALL apparently focused only on samples known to have a clonal rearrangement in IGH, the potential applicability and wide-spread utility of sequencing of IGH in unselected clinical samples for MRD has not been tested.

Here, we consider an unselected cohort of patients enrolled in Children Oncology Group AALL0932 trial and use residual material from 99 patient samples submitted for routine multi-parametric flow cytometry (mpFC) at U. of Washington. One sample failed in the initial DNA extraction step and was not further considered. We show using high-throughput sequencing that clonal IGH rearrangements can be identified in 92 of the remaining 98 pre-treatment samples, using a definition of a V-D-J or D-J rearrangement comprising at least 10% of total nucleated cells (Fig. 1A). Similar to our prior findings in TPC-ALL, we find three subsets of patients—1) those for whom MRD is not detected by either flow cytometry or HTS; 2) those for whom MRD is detected both by flow cytometry and HTS; and 3) those for whom MRD is detected only by HTS, but not flow cytometry (Fig. 1B). There were no false negative results by HTS as compared to flow cytometry.

Figure 1

View largeDownload slide

Measurement of clonal IGH rearrangement by high-throughput sequencing (HTS) or immunphenotypically abnormal B lymphoblast population by multi-parametric flow cytometry in pre-treatment (A) or day 29 post-treatment (B) residual samples. Results are reported for both HTS (red) and mpFC (blue) as clone frequency per total nucleated cells.

Figure 1

View largeDownload slide

Measurement of clonal IGH rearrangement by high-throughput sequencing (HTS) or immunphenotypically abnormal B lymphoblast population by multi-parametric flow cytometry in pre-treatment (A) or day 29 post-treatment (B) residual samples. Results are reported for both HTS (red) and mpFC (blue) as clone frequency per total nucleated cells.

Close modal

In the third group (HTS+positive, flow cytometry-negative), a subset of these patients, (5 of 28) had MRD detectable by HTS at a level within the expected sensitivity of flow cytometry. We hypothesized that in these cases that post-treatment MRD sequences may be present within the maturing B cell compartment that is not immunophenotypically aberrant by flow cytometry. To test this hypothesis, we analyzed eight additional post-treatment samples that were negative for MRD by flow cytometry. The mature B-cell fraction was collected by triple, flow cytometry-sorting and then sequenced by HTS for IGH rearrangements to search for the index clone defined in the corresponding, paired pre-treatment samples. Although a limited finding, diagnostic index IGH sequence was indeed identified in one of eight samples, in only the mature B-cell fraction, which is consistent with the proportion of cases with high-level MRD detected by HTS but which was missed by flow cytometry. Taken together, our results provide additional support for assessment of MRD in acute lymphoblastic leukemia by high-throughput sequencing. Our findings argue that precise quantification of the level of MRD by HTS will be important, and suggest that clonal IGH rearrangement sequences may be detected in an immunophenotypically normal population of mature B cells that may not be detected by flow cytometry.