The IG/TR Next Generation Marker Screening Developed within Euroclonality-NGS Consortium Is Successful in 94% of Acute Lymphoblastic Leukemia Samples

Introduction

The current gold standard molecular method for minimal residual disease (MRD) assessment is RQ-PCR for detection of leukemia-associated immunoglobulin (IG) and T cell receptor (TR) rearrangements, which must be identified beforehand in the diagnostic material. Conventional screening of IG/TR rearrangements in the diagnostic sample consists of several multiplex PCRs followed by Sanger sequencing.

Recently, the EuroClonality-NGS Consortium (www.euroclonalityngs.org, coordinated by AW Langerak) developed a set of assays for next-generation sequencing-based marker identification in lymphoid malignancies (Brüggemann, Haematologica, 2017). We have been using this approach in routine diagnostics since 2016 and herein we provide an overview of markers detected in acute lymphoblastic leukemia (ALL) patients.

Material & Methods

Between 02/2017 and 06/2018, 471 ALL samples were screened employing the EuroClonality-NGS assay. The assay consists of 5-8 multiplex PCRs, depending on the type of malignancy [IGH-VJ, IGH-DJ, TRB-VJ, TRB-DJ, TRG, TRD, and IGK-VJ-Kde (and intron-Kde) tubes] followed by a 2nd-step PCR in which barcodes and sequencing adaptors are introduced. The resulting libraries were sequenced on the Illumina MiSeq, producing 2×250bp reads, aiming to reach at least 3000 reads per sequencing library. Rearrangements with an abundance ≥ 5% were considered to be leukemia-associated.

Out of the 471 investigated samples, 453 were diagnostic (dx) and 18 were relapse (rel) samples. If available, bone marrow samples (BM) were investigated (n=370), otherwise peripheral blood (PB) samples were used (n=101). For B-ALL patients (n=344) all PCRs were performed, whereas for T-ALL (n=127) the IGH-VJ and both IGK tubes were omitted.

Results

At least one leukemia-associated marker was detected in 244/253 (96%) B-ALL dx BM samples (Fig. 1A), 237 (94%) of patients carried two or more rearrangements. On average we detected 5.1 markers per patient (range 0-10). The majority of patients carried at least one complete IGH (168; 66%), TRG (155; 61%), TRD (154; 61%), or IGK (142; 56%) rearrangement (Fig. 1B).

In dx BM T-ALL samples at least one leukemic marker was detected in 88/100 (88%) cases, and in 80 cases (80%) two or more were detected (Fig. 1C). On average 3.6 markers per patient were detected (range 0-10). The majority of patients carried TRG (67; 67%), TRD (65; 65%), or complete TRB (52; 52%) rearrangements (Fig. 1D).

In dx PB samples at least 1 marker was detected in 74/75 (98%) B-ALL (on average 5.2 markers per patient), and 23/25 (92%) T-ALL samples (on average 3.7 markers per patient).

Out of 18 rel samples (17 BM, 1 PB; 16 B-ALL, 2 T-ALL), 14 (78%) carried at least 1 rearrangement and 13 (72%) at least two rearrangements. On average 3.7 markers per patient were detected.

Conclusions

The EuroClonality-NGS assays detected at least 1 leukemic marker in 443/471 (94%) patients in our cohort. The number of patients with no marker was higher in the T-ALL cohort (15/127; 12%) compared to the B-ALL cohort (13/344; 4%). Only 3/100 (3%) dx PB samples had no marker detected, which is surprising considering the usually lower leukemic infiltration of PB in B-ALL, but can be attributed to the higher sensitivity of NGS.

Besides higher sensitivity, NGS-based marker identification brings along other benefits: the turn-around time is shorter and it is possible to detect multiple markers per tube without laborious cloning, or splitting the multiplex PCRs into singleplex.

Overall, EuroClonality-NGS assays have been shown to be robust, sensitive, and routinely applicable in a clinical diagnostic setting.

View largeDownload slide

View largeDownload slide

Close modal