Thrombotic microangiopathy after GVHD prophylaxis with tacrolimus/sirolimus: a call for use of consensus definition in reporting

To the editor:

We read with great interest the manuscript published by Rodriguez et al describing outcomes of a phase II pilot study of a tacrolimus/sirolimus regimen for graft-versus-host disease (GVHD) prophylaxis in 85 recipients of matched-sibling donor hematopoietic cell transplantation (HCT) using 3 different regimens (fludarabine-melphalan [Flu-Mel] = 46, total body irradiation [TBI] plus etoposide = 28, and busulfan-cyclophosphamide [Bu/Cy] = 11).¹  The authors describe an overall incidence of posttransplantation thrombotic microangiopathy (TMA) of 19% occurring at a significantly higher rate (55%, P = .005) in patients who received conditioning with Bu/Cy.

The reported high incidence of TMA in 19% of cases is particularly noteworthy and certainly higher than previously reported (10.8%) with the concurrent use of tacrolimus and sirolimus in the transplant setting.²  There are several limitations, however, to this finding. First, the definition of posttransplantation TMA used by Rodriguez et al “consisting of simultaneous occurrence of schistocytosis, increased lactate dehydrogenase (LDH), and persistent thrombocytopenia (below 50 000/uL)” is particularly nonspecific.^1p1099 In contrast, the Blood and Marrow Transplant Clinical Trials Network (BMT-CTN) operational definition of TMA requires the following: microangiopathic hemolysis (red blood cell fragmentation and ≥ 2 schistocytes per high-power field on a peripheral blood smear), concurrent increased serum LDH above institutional baseline, concurrent renal (defined as doubling of serum creatinine from baseline or 50% decrease in creatinine clearance from baseline) and/or neurologic dysfunction without other explanations, and negative direct and indirect Coombs tests results.³  The BMT-CTN consensus definition of TMA, as well as the proposed adaptation of the CTC criteria⁴  (Version 3.0) for TMA severity, requires renal or neurologic compromise. Importantly, these are not included in the criteria used by Rodriguez et al, which may have resulted in overestimating the true incidence of TMA. Therefore, comparisons to the incidence of TMA previously reported with the concurrent use of tacrolimus and sirolimus are limited. We believe that it would be of more value to the transplant community if the authors report the incidence of TMA by grade according to proposed BMT-CTN consensus definition. Second, while the association between TMA and myeloablative conditioning (with Bu/Cy or TBI-etoposide) is intriguing, one should also consider the impact of elevated tacrolimus levels in the Bu/Cy group: the median and the interquartile range (IQR) for serum tacrolimus levels (nanograms per milliliter) were significantly higher (P = .003) in patients conditioned with Bu/Cy compared with those receiving TBI-etoposide or Flu-Mel. As well, univariate analysis demonstrated a non-significantly increased hazard for TMA with advancing quartile of median tacrolimus level. The authors acknowledge the limitations imposed by relatively small number of patients and TMA events in this analysis. This may in part explain the higher incidence of TMA in this group and suggests that a narrower therapeutic range for tacrolimus be considered when used in combination with sirolimus in the setting of Bu/Cy.