A comparison of qPCR and ddPCR used for quantification of the JAK2 V617F allele burden in Ph negative MPNs

Abstract

Philadelphia-negative myeloproliferative neoplasms (MPNs) are a diverse group of diseases whose common feature is the presence of V617F mutation of the JAK2 gene. In the era of novel therapeutic strategies in MPNs, such as JAK-inhibitor therapy, there is a growing need for establishing high sensitive quantitative methods, which can be useful not only at diagnosis but also for monitoring therapeutic outcomes, such as minimal residual disease (MRD). In this study, we compared the qPCR and ddPCR methods and their clinical utility for diagnosis, prognostication, and treatment monitoring of MPNs with JAK2 V617F mutation in 63 MPN patients of which 6 were subjected to ruxolitinib treatment. We show a high conformance between the two methods (correlation coefficient r = 0.998 (p < 0.0001)). Our experiments revealed high analytical sensitivity for both tests, suggesting that they are capable of detecting the JAK2 V617F mutation at diagnosis of MPN with a limit of detection (LoD) of 0.12% for qPCR and 0.01% for ddPCR. The alterations of JAK2 V617F allele burden in patients treated with ruxolitinib were measured by both methods with equal accuracy. The results suggest an advantage of ddPCR in monitoring MRD because of allele burdens below the LoD of qPCR. Overall, the clinical utility of qPCR and ddPCR is very high, and both methods could be recommended for the routine detection of the V617F mutation at diagnosis, though ddPCR will probably supersede qPCR in the future due to cost-effectiveness.

Keywords: JAK-inhibitor; Limit of detection; Minimal residual disease; Myeloproliferative neoplasms; ddPCR; qPCR.

Fig. 1

Estimation of limit of detection (LoD) for qPCR and ddPCR methods. For qPCR, fractions of positive reads were fitted to sigmoidal curve and the LoD was estimated at 95% confidence level. For ddPCR, LoD calculation was based on limit of blank (LoB) value and standard deviation (SD) of measurements

Fig. 2

Correlation between qPCR and ddPCR (r = spearman correlation coefficient)

Fig. 3

The differences between measurements by qPCR and ddPCR (y axis) vs. the mean of the sample quantification performed by qPCR and ddPCR (x axis) (the Bland-Altman plot). Dotted lines indicate bias (mean difference) and 95% limits of agreement (upper limit + 2 SD; lower limit − 2 SD) between the two given methods. A 95.12% of presented measurements fitted to the limit of agreement

Fig. 4

Comparison of median JAK2 V617F allele burden measured by qPCR and ddPCR. Mann-Whitney U test shows no significant differences between the two methods. PV+: patients diagnosed with polycythemia vera, ET+: patients diagnosed with essential thrombocythemia, PMF+: patients diagnosed with primary myelofibrosis

Fig. 5

Distribution of JAK2 V617F-positive MPN patients and their allele burden within subgroups on the basis of ddPCR results. Among the patients diagnosed with ET, the JAK2 V617F mutation was detected in 12 patients (60%) with qPCR and in 13 patients (65%) with ddPCR. One patient had a low mutation burden of 0.011% measured by ddPCR with a corresponding mutation burden of 0.02% measured in qPCR (this value was calculated below qPCR LoD)