Digital PCR-based dynamic monitoring of JAK2V617F burden: A decade-long study unravels its prognostic value in MPN treatment response

Abstract Background: The JAK2V617F mutation is a key driver mutation in myeloproliferative neoplasms (MPNs), and its dynamic changes may reflect disease progression and treatment response. This study aimed to evaluate the clinical utility of digital PCR (dPCR)-based longitudinal monitoring of JAK2V617F variant allele frequency (VAF) in therapeutic assessment and prognostic prediction for MPNs. Methods: The cohort comprised 445 JAK2V617F-positive MPN patients diagnosed at Peking Union Medical College Hospital between January 2016 and January 2025. All patients provided informed consent, and the study adhered to Helsinki Declaration guidelines. According to WHO 2016 criteria, diagnoses included: ET (n=115; 36 males, 79 females), PV (n=230; 112 males, 118 females), prePMF/PMF/secondary MF (n=87; 35 males, 52 females), and other MPNs (n=13; 9 males, 4 females) including MPN-AML, MPN-U, and MDS/MPN. Retrospective data collection encompassed laboratory parameters, non-driver mutations, fibrotic/leukemic transformation, and treatment regimens. We used dPCR to quantify mutation burden and analyze its association with disease subtypes, treatment regimens, and molecular kinetic characteristics. Results: This study establishes digital PCR (dPCR) as a precise methodology for JAK2V617F variant allele frequency (VAF) quantification in myeloproliferative neoplasms (MPNs). Analytical validation demonstrated excellent linearity (R² = 0.9994) across clinically relevant VAF ranges (0.0039%-12.20%), with optimal reproducibility at VAF ≥0.020% (CV<5%). Disease-specific patterns of molecular response were observed. Polycythemia vera (PV) patients achieved significantly higher molecular response rates to interferon-alpha (IFN-α) therapy (71% overall response) compared to essential thrombocythemia (ET) patients (43% overall response; p<0.01). The ELN 2009 criteria demonstrated superior sensitivity for molecular response assessment, particularly in patients with baseline VAF 10-20%. Early molecular response kinetics showed strong predictive value. For ET, a >2.84% monthly decline (≥34% over 12 months) distinguished responders (77% PR or CMR). In PV, thresholds were >2.6% monthly (≥31.2% over 12 months) for response (88% PR or CMR) and >5.4% monthly (≥65.3% over 12 months) for CMR (54%). These kinetic patterns may serve as early indicators for treatment efficacy assessment. Non-driver mutation burden inversely correlated with molecular response quality. Complete molecular responders (CMR) exhibited the lowest mutation load (0.21 mutations/patient), while non-responders carried significantly higher mutational complexity (0.83 mutations/patient; p<0.001). However, this association was attenuated in IFN-treated patients (p=0.12 for Non-driver mutation burden difference between responders and non-responders), suggesting IFN may partially overcome the negative impact of non-driver mutations. This molecular stratification provides a framework for personalized treatment approaches. Conclusion: dPCR enables highly precise dynamic monitoring of JAK2V617F, and VAF decline rates serve as reliable indicators for predicting MPN treatment efficacy. PV patients respond best to IFN therapy, while higher non-driver mutation burden may indicate adverse prognosis, though this effect could be partially mitigated by IFN treatment. Our findings provide critical evidence for precision medicine in MPN management.