MSNPseq: An NGS-based assay for chimerism detection and risk assessment in allogeneic HSCT.

6557 Background: Chimerism (CHIM) analysis assesses the ratio of donor to recipient hematopoietic cells, confirming engraftment and detecting impending graft failure or relapse, in allogeneic hematopoietic stem cell transplantation (aHSCT). However, the conventional short tandem repeat (STR) analysis, has limited sensitivity (1–5%), often missing low-level CHIM. In this retrospective study, Multiple Single Nucleotide Polymorphism sequencing (MSNPseq), an NGS-based assay targeting DNA fragments each harboring 4–8 SNPs, demonstrated superior sensitivity and specificity compared with STR analysis, enabling reclassification of STR-negative cases and improved prognostic stratification. Methods: MSNPseq, based on Ion AmpliSeq technology, amplifies 100 fragments for sequencing on the Ion Torrent PGM platform, achieving >2000× depth for 96.8% of loci. Its sensitivity and accuracy were validated in simulated chimeric and post-transplant samples against STR analysis. Calibration models used pre-transplant donor/recipient data to generate tailored correction coefficients. CHIM was assessed via MSNPseq at three 30-day time points (TP1, 2, 3) post-transplant, in 741 aHSCT patients (AML 49.9%, ALL 27.4%, MDS 13.5%, lymphoma 4.7%, others 4.5%). PFS and OS were stratified by CHIM status and its dynamics to evaluate transplant outcomes, prognostic significance, and late relapse. Results: CHIM sensitivity was assessed using BCR-ABL1–positive standards. MSNPseq achieved a 0.01% detection limit, outperforming STR (1%) with strong correlation (R² = 0.9953). CHIM status was classified by donor fraction as complete (cCHIM, 100%), microchimerism (mCHIM, 99% to <100%), subthreshold chimerism (sCHIM, 95% to <100%; below STR sensitivity), and incomplete chimerism (iCHIM, ≤95%) across three time points. While iCHIM was consistently associated with inferior PFS and OS, sCHIM and mCHIM also showed worse outcomes compared with cCHIM, defining distinct high-risk subsets. CHIM dynamics revealed decreasing sCHIM/mCHIM and increasing cCHIM/iCHIM over time. Notably, a sustained CHIM decrease >0.5% over two consecutive time points was associated with significantly poorer OS and PFS. Interestingly, sCHIM and mCHIM status, as well as a sustained CHIM decrease >0.5% over two consecutive time points, were also associated with a high risk of acute graft-versus-host disease (aGvHD). Overall, MSNPseq enables more precise CHIM stratification than STR, improving post-transplant risk assessment. Conclusions: MSNPseq's superior sensitivity enables more precise chimerism classification and detection of clinically relevant sCHIM and mCHIM subsets that STR often misses. These refined CHIM categories, along with sustained CHIM decline, correlate with poorer survival and higher aGvHD risk, supporting MSNPseq as a superior tool for post-aHSCT risk stratification and therapeutic guidance.