Can JAK inhibition prevent GVHD?

In this issue of Blood, Pidala et al report on the results of a multicenter phase 2 trial, in which a graft-versus-host disease (GVHD) prevention regimen of pacritinib, an oral JAK2 inhibitor, in combination with sirolimus and tacrolimus successfully suppressed pSTAT3 signaling in CD4⁺ T cells but failed to decrease the incidence of acute GVHD compared with historical controls.¹ 

Historically, acute GVHD has been a fundamental barrier to successful allogeneic hematopoietic cell transplantation (HCT). After years of inadequate efficacy, GVHD prophylaxis regimens have advanced and resulted in significant reductions in severe acute GVHD. Most notably, posttransplant cyclophosphamide has been widely adopted due to its ease of administration and effective protection against acute and chronic GVHD.² Still, novel approaches to GVHD prevention remain an unmet need, appreciating that one size never fits all. Optimal GVHD prophylaxis regimens are effective, tolerable, and logistically simple.

The current trial builds upon preclinical data demonstrating that JAK2/STAT3 activity is integral to differentiation of alloreactive Th1 and Th17 cells, and as such, JAK2 inhibition could prevent the development of GVHD.³ Furthermore, the phase 1 trial from the authors identified pacritinib 100 mg twice daily as the minimum biologically active dose for further investigation.⁴ In the current study, the combination of pacritinib, sirolimus, and tacrolimus successfully replicated the previously demonstrated biological activity, by suppressing %pSTAT3⁺ CD4⁺ T cells to ≤10% (goal ≤ 35%). Disappointingly, the pharmacodynamic effect failed to translate clinically, with rates of acute GVHD (grade 2-4: 46%; grade 3-4: 18%) that mirror historical experiences with calcineurin inhibitor-based approaches.

The key question from this study is why did this promising regimen fail to prevent acute GVHD? The authors’ meticulous approach to the development and investigation of this pacritinib-based combination is refreshing, anchoring the trial design on immunological activity. One consideration is whether GVHD solely developed in participants who did not achieve sufficient suppression of the targeted pathway. However, the authors demonstrate no correlation between acute GVHD and depth of pSTAT3 inhibition or T-cell subset ratios. Another possibility questions the timing of the biological assessment: is an evaluation on day +21 representative of the immunological effect that seemingly needs to be maintained over several months to prevent acute GVHD? The authors show some loss of effect at day 100, although participants were no longer on pacritinib at that time. Another factor may have been adherence to pacritinib. Although only 1 dose-limiting toxicity occurred in the phase 1 trial, only 9 of the 28 participants in the current study completed the intended 100 days of administration without interruption, with gastrointestinal toxicities (regardless of cause) being a main barrier. However, the authors again show no correlation between duration of pacritinib therapy and acute GVHD, noting that most GVHD cases occurred while receiving pacritinib. Although it is possible that the phase 1 trial was underpowered to assess the clinical activity of the regimen, the current study is a reminder of the trials and tribulations of translational research. Despite our advances in knowledge, we likely underappreciate the entire spectrum of GVHD biology, as well as the full range of on- and off-target effects of our interventions.

Does this trial close the book on JAK inhibition in GVHD prevention? Not at all. There are multiple JAK inhibitors in clinical practice, each with slightly different targets and activity.⁵ Ruxolitinib, which inhibits JAK 1 and 2, is the JAK inhibitor with the most well-established clinical activity in GVHD.^6,7 Preliminary data suggest that ruxolitinib is associated with low rates of both acute and chronic GVHD when administered during the peri- and posttransplant setting.^8,9 Although larger prophylaxis trials are planned with ruxolitinib (BMT CTN 2203), the authors should be congratulated on their work in elucidating the mechanistic effects of JAK inhibition during allogeneic HCT, as the transplant community continues to strive for targeted, safe, and effective approaches for our patients.

Conflict-of-interest disclosure: Z.D. receives research support from Incyte, Taiho Oncology, and Kura Oncology and has received consulting fees from Sanofi, Incyte, MorphoSys AG, Inhibrx, PharmaBiome AG, and Ono Pharmaceutical.