Next generation sequencing from dried blood spots: “Spot on CML” results from low- and middle-income countries

Abstract Objectives. Most chronic myeloid leukemia (CML) cases occur in low- and middle- income countries (LMICs) with limited or no access to essential tools for CML management and published studies demonstrate a higher CML disease-related morbidity and mortality, especially in younger patients, as compared to patients with higher socioeconomic status. Access to diagnostic testing and medication can radically improve the outcomes of CML patients in LMICs. The Max Foundation has provided tyrosine kinase inhibitors (TKIs) through the Glivec International Patient Assistance Program and Max Access Solutions programs at no cost to eligible patients. Dried blood spots (DBS) are a practical solution that can be carried out with minimal training and readily implemented in settings with fragile health infrastructure. DBS mitigate potential biohazard issues and costs of shipping fresh blood samples to distant laboratories. Cold storage is not required. Previously, we have shown the feasibility of BCR::ABL1 transcript monitoring using DBS including the use of point-of-care assays. We have now developed methods for next-generation sequencing (NGS) from DBS of targets implicated in treatment resistance (e.g., ABL1 tyrosine kinase domain (TKD) mutations), and associated with treatment response (e.g., ASXL1 variants) and disease progression (e.g., ASXL1, RUNX1, GATA2, IKZF1, andBCORL1). Methods. Blood was spotted locally on WhatmanTM 903 Proteinsaver Snap-Apart Cards (Thermo Fisher Scientific Inc., Waltham, MA) with 200 µL blood per card (4 spots, each 50 µL) and then dried. DBS samples from 177 patients from 9 countries were received between November 2019 and June 2024. The QIAamp® DNA Blood Mini Kit (QIAGEN, Hilden, Germany) was used to extract DNA. Two sequencing platforms were used during the study: the Archer VariantPlex® Myeloid panel (75 genes, N=50) and the Genexus OncomineTM Myeloid Assay GX V2 (40 genes, N=133). OEFC rules were set up to filter for any non-synonymous somatic mutations with < 106 population minor allele frequency across three genome population databases (dbSNP, 5000Exome global, and ExAC), as well as variants annotated as pathogenic (Tier I) or likely pathogenic (Tier II) by the VarSome database. The call threshold for pathogenic ABL1 mutations was set at 2%. Otherwise, the call threshold for hotspot variants was set at 5% except for ASXL1 G646Wfs*12 which was set at 10%. Results. The median age was 42 years (range, 18–70 years); 67 patients (38%) were from Africa, 97 (55%) from Southeast Asia, and 13 (7.3%) from Central and South America. ABL1 TKD mutations were identified in 34% of patients (74 ABL1 TKD mutations in 61 patients). The T315I mutation was the most common ABL1 TKD mutation identified (N= 28, 37.8% of mutations). The next most common mutations were E255K/V (N=8), F317L (N=5), and Q252H (N=5). Among 11 patients with multiple mutations, 5 had combinations with T315I (Y253H, M244V, E255K, F317L, and M351T). We identified 89 tier I and tier II gene variants (i.e., strong or potential clinical significance) by NGS in 69 patients including 14 patients with multiple variants. The most frequent gene variants detected were ASXL1 variants (52 variants in 49 (29%) patients). Among 61 patients with ABL1 mutations 27 (44%) had variants in ASXL1 in contrast to 116 patients without ABL1 mutations where 22 (19%) had gene variants in ASXL1. The next most common variants were DNMT3A (6), NF1 (6), RUNX1 (4), and BCOR (3). In univariate analysis the detection of ABL1 mutations, ASXL1 variants, or other gene variants did not correlate with sequencing method, patient age, or region of origin. The presence of pathogenic ASXL1 variants correlated strongly with the presence of any ABL1 TKD mutation and with T315I (P < 0.001). Patients with ASXL1 variants were twice as likely to harbor ABL1 TKD mutations. Conclusions. We have now demonstrated the value of DBS combined with sequencing technologies to detect ABL1 TKD mutations and other leukemia-associated gene variants. These new investigations expand our prior work and demonstrate the use of DBS across the CML clinical care spectrum including diagnostics, monitoring and resistance mutation detection. These methods offer hope of a cost-effective, practical, and sustainable approach for CML care globally. These approaches are also amenable to delineating molecular targets in other hematological malignancies with the goal of improving healthcare accessibility.