JAK2-V617F–triggered preemptive and salvage adoptive immunotherapy with donor-lymphocyte infusion in patients with myelofibrosis after allogeneic stem cell transplantation

To the editor:

Primary myelofibrosis is a myeloproliferative disease, and results of conventional treatment remain unsatisfactory.^1,2  Allogeneic stem cell transplantation after dose-reduced conditioning has become a reasonable, curative treatment option.^3,4  Single case reports about successful donor lymphocyte infusion (DLI) for relapsed patients provided evidence of a graft-versus-myelofibrosis effect.^5-7  Here, we report on 17 patients with either myelofibrosis (n = 16) or secondary AML post myelofibrosis (n = 1) and a median age of 52 years (range, 32-63 years) who received DLI from related (n = 5) or unrelated (n = 12) donor, either for clinical relapse (salvage DLI; n = 9) or residual disease monitored by JAK2 mutation level in peripheral blood (preemptive DLI; n = 8). Details are summarized in Table 1. One patient (no. 9) received DLI twice: once for molecular residual disease and once for reappearance of molecular disease. Sixteen patients were JAK2-V617F–positive. The median time from transplantation to first DLI was 269 days (range, 127-1570 days). The median percentage of JAK2V617 mutation level in peripheral blood before first DLI was 6.2% (range, 0.2%-72.8%) and significantly higher in patients with clinical relapse than with molecular relapse (24.7% vs 0.37%; P = .03). The median cell dose of the first DLI was 10⁶ CD3⁺ cells per kilogram of body weight (BW; range, 0.5-9 × 10⁶ cells/kg BW). Two patients received as first DLI CD4-selected T cells (5 × 10⁶ CD4⁺ cells/kg BW). Second and subsequent half-log–increased DLI were given if no graft-versus-host disease (GVHD) and no significant response were observed. The median interval between the first and the second DLI was 103 days (range, 43-661 days). The response to DLI for relapsed patients was determined by the response criteria of the International Working Group,⁸  and for residual disease by quantitative real time JAK2-V617F polymerase chain reaction (PCR) performed from genomic DNA from peripheral blood as recently described.⁹  The sensitivity of this quantitative PCR is 0.01%. For one patient without JAKV617F mutation molecular remission was determined by 100% donor chimerism with quantitative PCR using genetic polymorphisms.

No treatment-related mortality after DLI was observed. Acute GVHD grade II through IV was seen in 3 patients (18%), which occurred only in the salvage DLI cohort. In this cohort, complete remission was seen in 4 patients (44%), which was complete on molecular level in 3 of them. In 2 of them complete remission was associated with acute GVHD (grade III). All other patients with clinical relapse (n = 5) developed no GVHD, but best response was only clinical improvement or stable disease. In contrast to salvage DLI, all patients who received preemptive DLI responded with complete molecular remission (100%), and none of these patients developed any signs of grade II through IV acute GVHD. Only in one patient was mild chronic GVHD of the liver noted. The median time to achieve complete remission was 79 days (range, 31-495 days). The overall rate of complete molecular remission rate was 68%, which was higher in the preemptive group than in the salvage group (100% vs 44%; P = .04).

This report provided evidence for a strong donor T cell–mediated graft-versus-myelofibrosis effect. Adoptive immunotherapy for molecular residual disease monitored with highly sensitive PCR for JAK2-V617F mutation seems to be more effective and less toxic than using donor lymphocyte infusion for clinical relapse and should be implemented in further clinical trials.