Proteasome: target for acute and chronic GVHD?

In this issue of Blood, Pai et al provide exciting murine and clinical data for proteasome inhibition in the treatment of chronic graft-versus-host disease (cGVHD).¹ 

With increased utilization of unrelated donors after allogeneic hematopoietic stem cell transplantation (HSCT), cGVHD is now the leading cause of late morbidity and mortality. Most patients (≥50%) develop cGVHD requiring corticosteroids, and about half of those will require additional therapies, a major risk factor for poor survival. Several immunosuppressant agents, largely empirically tested, have shown little benefit as second-line therapy. Thus, an urgent need exists for novel approaches with improved activity and tolerability in patients that develop cGVHD that is not responsive to steroids.

Newly established National Institutes of Health criteria now provide a comprehensive classification that encompasses the diverse clinical manifestations of cGVHD; however, its underlying biology has largely remained obscure. Emerging experimental data have implicated dysregulation of humoral immunity with germinal B-cell aberrations and generation of pathologic allo- and autoantibodies as potentially key mechanisms.^2,3  Consistent with these murine data, correlative studies from patients have implicated antibodies against HY antigens in female-to-male HSCT,⁴  antibodies against platelet-derived growth factor receptor-α,⁵  and increases in B-cell activation factor (BAFF)⁶  in clinical cGVHD. These observations, together with the finding that rituxan may prevent clinical cGVHD, further support a putative role of B cells in its pathogenesis.

Bortezomib, which is Food and Drug Administration approved for 2 malignant B-cell neoplasms (multiple myeloma and mantle cell lymphoma), possesses broad immunomodulatory properties. In previous murine models and in a clinical trial, bortezomib prevented acute GVHD (aGVHD).^7,8  Whether bortezomib would have a similar effect on cGVHD was previously unknown. In a murine model of sclerodermatous cGVHD, Pai et al show that administration of bortezomib attenuates clinical and pathologic skin lesions after the onset of cGVHD. Animals treated with bortezomib showed reduced germinal center B cells and decreased expression of nuclear factor-κB and BAFF in cGVHD lesions. Because thymic epithelial damage is believed to contribute to immune dysregulation, the findings of increased thymic cellularity and regulatory T-cell/conventional T-cell ratios further suggested proteasome inhibition had broad salutary effects on the immunological aberrations of cGVHD. Bortezomib also did not impair antitumor immunity in animals. Furthermore, variations in the timing of administration had an impact on the rates of response of cGVHD but did not accelerate or increase cGVHD severity. These experiments were necessary because in previous studies, timing critically altered the response and toxicity of aGVHD. The most exciting aspect of this work by Pai et al is that they performed a pilot proof of concept clinical trial in 10 patients with steroid refractory cGVHD. Informed by their murine data, suggesting skin necrosis might occur at higher bortezomib doses, a careful approach to intrapatient dose escalation was undertaken. Preliminarily, botezomib had good tolerability and potential for inducing some striking clinical responses. These data are also supported by another recent clinical study.⁹  Thus, the study by Pai et al is exciting because, for the first time, they implicate proteasome inhibition as a pathogenic pathway for targeting and ameliorating cGVHD.

The authors should be commended for melding preclinical and human studies in such a thoughtful manner; however, several questions remain. First, much remains to be understood about the specific dysfunction of B cells and the potential complex interactions with other immune cells in cGVHD. The role played by the immune dysfunction in causing fibrosis and the impact on reversing fibrosis remain poorly understood. In the current study, although the authors demonstrate clinical responses, the specific and causative cellular and molecular mechanisms responsible for the effects of proteasome inhibition on fibrogenesis were not explored. Some of these questions have, in the past, been hampered in part due to the absence of murine models that completely mimic all aspects of clinical cGVHD (kinetics and manifestations). As such addressing these questions will require considerable and continuous efforts. Second, the nature and composition of the specific ubiquitinated proteins targeted by the proteasome that are critical for cGVHD remain unknown. It remains a possibility that targeting a limited number of the ubiquitinated proteins may allow for improved tolerability and less toxicity than global proteasome inhibition.¹⁰  Third, cGVHD has protean manifestations, and whether all or only some patients will respond, and if so, under what circumstances, has been challenging to study in a rigorous manner. Thus, the clinical observations by Pai et al, while exciting, are obviously limited by the small patient numbers. Furthermore, the authors suggest established fibrosis may be difficult to treat, and even in responders, recurrences of cGVHD were observed after bortezomib cessation. The need for prolonged treatment makes tolerability, feasibility, and infectious complications additional considerations for future studies of botezomib in cGVHD. Finally, an ounce of prevention may be worth a pound of cure; therefore, it would be interesting to explore this approach in prevention of cGVHD, as it is being studied in aGVHD.

Thus, this study by Pai et al is exciting because, for the first time, they implicate proteasome inhibition as a pathogenic pathway for targeting and ameliorating cGVHD, in addition to its previously suggested role in aGVHD. Future studies will need to explore the critical mechanisms for these experimental observations. Although the clinical efficacy of bortezomib and proteasome inhibition remain to be further explored in larger, robustly designed clinical trials, this study, along with the study by Herrera et al,⁹  suggests that it may be a welcome addition to the clinician's toolbox, even now, in severe cases.