Patients and samples
We enrolled 34 primary and secondary patients with MF and with mutated JAK2 V617F [6] who received allo-HCT at Seoul St. Mary’s Hospital between December 2012 to November 2021. A total of 150 samples were obtained at the time of allo-HCT (n=33) and at 30 d (n=32), 100 d (n=31), 180 d (n=30), and 360 d (n=24) after allo-HCT. Bone marrow morphology and medical records were thoroughly reviewed to determine clinical courses and relapse status. This study was approved by the Institutional Review Board of Seoul St. Mary’s Hospital, The Catholic University of Korea, Seoul, South Korea (KC22RISI0120) and was conducted in accordance with the tenets of the Declaration of Helsinki.
Transplantation procedures
All of the patients received a reduced-intensity conditioning regimen, which consisted of fludarabine (30 mg/m2 for 5 days) and busulfan (3.2 mg/kg for 2 days) with total body irradiation (TBI) 200–400 cGy [7]. Graft-Versus-Host disease (GVHD) prophylaxis consisted of anti-thymocyte globulin (ATG; Thymoglobulin®)/calcineurin inhibitor/methotrexate (MTX). ATG was administered at a dose of 2.5–7.5 mg/kg according to the donor types (≥5 mg/kg in mismatched donor transplantation). MTX (5 mg/m2) was used on days +1, +3, +6, and +11, along with calcineurin inhibitor (cyclosporine for matched sibling donors and tacrolimus for unrelated donors and haploidentical familial donors). The calcineurin inhibitor dose was tapered gradually starting on day 100–120 after allo-SCT in the absence of acute GVHD. Early tapering of immunosuppressive therapy (IST) was done for the patients who were clinically suspected to relapse. The other general transplantation procedures were performed as described previously [8, 9].
Definitions for relapse
The relapse status was established based on the EBMT definition [3, 5]. The morphological and clinical criteria include the following: an increase in age-adjusted cellularity with an abnormal Myeloid:Erythroid ratio; typical megakaryocytic abnormalities; increase in grade of myelofibrosis; and/or development of myelodysplasia, monocytosis, or increased blast count, accompanied by irreversible cytopenia (hemoglobin <100 g/L, neutrophil count <1x109/L, and platelet count <100 x109/L) or an increased immature myeloid cell count in the peripheral blood. In addition to the overt relapse by morphological and clinical criteria, the prognostic relevance of cytogenetic relapse or evolution, molecular relapse, and chimerism relapse were also investigated. Cytogenetic relapse was defined as the appearing preexisting cytogenetic abnormality, and cytogenetic evolution, the new development of any abnormality, which were confirmed by repeated testing. Molecular relapse and chimerism relapse were assessed by JAK2-MRD and chimerism testing, respectively. The optimized threshold and time points for molecular relapse and chimerism relapse were subjects to be investigated.
MRD monitoring using JAK2 V617F quantification
DNA was extracted from bone marrow or from peripheral blood using QIAsymphony DSP DNA kits with QIAsymphony instrument (Qiagen, Hilden, Germany). Quantification was performed using Qubit dsDNA Broad Range Assay kits (Thermo Fisher Scientific, Waltham, MD, USA). MRD monitoring for JAK2 V617F was performed using real-time PCR (JAK2 MutaQuant kit, Ipsogen; Qiagen) according to the manufacturer’s instructions. Briefly, a short amplicon covering the JAK2 V617F region was amplified using 25 ng of purified DNA. Positive and negative calibrators at four different concentrations were included in each run to obtain standard curves. All of the samples were tested in duplicates, and the mean cycle threshold (Ct) values were transformed to copy numbers of JAK2 V617F and wild-type using the prepared standard curves. The JAK2-MRD was expressed as the variant allele frequency (VAF) as a percentage of JAK2 V617F copy numbers for total JAK2 (JAK2 V617F plus JAK2 wild-type) copy numbers. We developed the other JAK2-MRD marker, which represented the change of JAK2 V617F, through calculating the ratio of VAF at each time point to the previous VAF (JAK2-MRD ratio).
Chimerism monitoring
We monitored % donor chimerism using both next-generation sequencing-based assay (NGS chimerism) and short tandem repeat-based assay (STR chimerism). NGS chimerism was analyzed using Devyser chimerism NGS kits (Devyser, Stockholm, Sweden). In a single tube, 24 insertion–deletion mutation markers were sequenced using MiSeq (Illumina, San Diego, CA, USA). Data analysis was performed using a dedicated program.
STR chimerism was assessed using AmpFlSTR Identifier PCR Amplification (Applied Biosystems, Warrington, UK) as previously reported [10, 11]. Briefly, 16 STR markers were amplified. PCR was performed using a C1000 Touch™ Thermal Cycler (Bio-Rad laboratories Inc., Hercules, CA, USA). Amplified PCR products were analyzed via capillary electrophoresis using an ABI 3130xl genetic analyzer (Applied Biosystems, Foster City, CA, USA). GeneMapper ID Software Version 4.1 (Applied Biosystems, Foster City, CA, USA) was used for automated genotyping and the quantification of peak areas.
Statistics
The endpoints were morphological/clinical relapse (called “overt relapse” hereafter) and death. The patients’ characteristics were expressed as median and range for continuous variables and frequencies for categorical variables. Categorical data were compared by Fisher’s exact test or the χ2 test, and continuous data were compared by the Wilcoxon test. We analyzed the predictive power of clinical and molecular factors for relapse, non-relapse mortality (NRM), relapse-free survival (RFS), and overall survival (OS). We defined RFS as the time from the allo-HCT to relapse or death and OS as the time from the allo-HCT to death from any cause. The medical records were tracked until October 2022. Receiver operating characteristic (ROC) analysis was performed to determine the optimal threshold and time points of JAK2-MRD and chimerism for predicting relapse.
The obtained thresholds and time points were validated via time-dependent ROC analysis in the competing risks setting using time ROC package in R [12]. Relapse and NRM were considered competing risk events. Area-under-the-curve (AUC) values were computed at days +300 and +500 after allo-HCT. The RFS and OS were estimated using the Kaplan–Meier method. A competing risk analysis was performed to estimate the probability of a cumulative incidence of relapse (CIR) and NRM. The CIR was compared across groups using the Gray test and cmprsk module in R [13, 14]. The RFS and OS were compared using the Cox proportional hazards regression. The continuous values of JAK2-MRD and chimerism at different time points were evaluated as time-dependent covariates. Statistical analyses were performed using MedCalc version 19.1.7 (MedCalc Software; Ostend, Belgium) and R software version 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria).