Quantitative real-time polymerase chain reaction coupled with high-resolution melting (HRM) for simultaneous detection and quantification of four isoforms of BCR::ABL1

Abstract Background The diagnosis and management of chronic myeloid leukemia (CML) critically rely on detecting and serially quantifying BCR::ABL1. The three major transcript variants are e1a2/e1a3 (p190), e13a2/e14a2 (p210), and e19a2/e19a3 (p230). Atypical transcripts have also been described, including e13a3/e14a3 (p203), which has been anecdotally linked to asciminib resistance. Currently, no clinical assay simultaneously detects multiple BCR::ABL1 isoforms; each isoform requires a separate analysis, which is labor-intensive and cost-inefficient. Consequently, baseline testing is typically limited to p210 and p190, without reflex testing for p230 or atypical isoforms if either is detected. This practice risks missing cases driven by p230 or atypical isoforms when these are co-expressed with low-level p190 or p210. Furthermore, standard monitoring focuses exclusively on baseline transcripts, potentially overlooking emerging novel isoforms during disease evolution, thereby falsely suggesting disease stability or remission and delaying timely therapeutic adjustments. Finally, the prevalence and significance of p203 remain unclear, partly because most primer sets fail to amplify the relevant ABL1 breakpoints. Methods To address these challenges, we developed a simple assay known as BloodHound to simultaneously detect and quantify four clinically relevant BCR::ABL1 transcripts (p190, p210, p230, and p203). This assay leverages real-time quantitative reverse transcription polymerase chain reaction (RT-qPCR) and high-resolution melting (HRM) technologies, achieving a limit of detection of 0.001%, 99.9% sensitivity, and > 99% specificity when validated against droplet digital PCR and Sanger sequencing. To evaluate the clinical utility of this assay, we characterized the distribution and co-expression of the four transcripts in 895 samples, including peripheral blood (n = 846) and bone marrow (n = 49), from patients with suspected (n = 566), established (n = 296), relapsed (n = 32) CML, or unknown status (n = 1). Results were confirmed by Sanger sequencing as the gold standard. Results Overall, 187/895 samples (20.9%) were BCR::ABL1-positive. Positivity rates were 3.5% (20/566) in patients evaluated for suspected CML, 33.1% (98/296) in established CML under monitoring, and 100% (32/32) in relapsed disease. As expected, p210 alone predominated (161/187, 86.1% of positives), while p190 alone (2/187, 1.1%) and p230 alone (1/187, 0.5%) were rare, and p203 was not detected. Co-expression of p190 and p210 was more frequent than historically appreciated (23/187, 12.3%), particularly in baseline diagnostic specimens (25.0%). The results showed 100% concordance with Sanger sequencing. Quantitatively, p210 levels were markedly higher than p190 when co-expressed (median p210:p190 ratio = 2,152). Quantitative values obtained with our approach correlated tightly with the standardized Qiagen ipsogen IS assay (r = 0.998; median absolute difference = 0.02%). Conclusion The BloodHound assay is sensitive and accurate, and by simultaneously quantifying four clinically significant BCR::ABL1 transcripts, it has the potential to significantly streamline clinical workflows, enhance diagnostic precision, and offer deeper insights into CML clonal dynamics. Both RT-qPCR and HRM are compatible with standard laboratory equipment, facilitating implementation and widespread clinical adoption.