Cyclophosphamide and Thiotepa Increases Risk of Transplant-Associated Thrombotic Microangiopathy

Abstract

Transplant associated thrombotic microangiopathy (TA-TMA) is a complication of hematopoietic cell transplant (HCT) associated with endothelial injury resulting in severe end organ damage, acute and long-term morbidity, and mortality. Myeloablative conditioning is a known risk factor, though specific causative agents have not been identified. We hypothesized that the combination of cyclophosphamide and thiotepa (CY + TT) is particularly toxic to the endothelium, placing patients at elevated risk for TA-TMA. We conducted a retrospective review of pediatric and young adult patients who received conditioned autologous and allogeneic HCT between 2012 and August 2023 at UCSF Benioff Children's Hospital, San Francisco. We excluded patients undergoing gene therapy or triple tandem transplants for brain tumors. Neuroblastoma tandem transplants were classified a single transplant occurrence. High dose N-acetylcysteine (NAC) prophylaxis was incorporated into the institutional standard of care from December 2016-May 2019 and May 2022-August 2023. Defibrotide was given prophylactically to patients deemed high-risk for sinusoidal obstruction syndrome (SOS) per institutional guidelines or on clinical trial NCT#02851407 for SOS prophylaxis or NCT#03384693 for TA-TMA prophylaxis. Kaplan-Meier analysis was used to estimate the 1-year cumulative incidence of TA-TMA. Univariate analysis was performed for each of the potential risk factors of interest using log-rank tests and bivariate analysis with Cox regression models using backward selection and hazard ratios were built using all covariates with a univariate P-value < .2 for allogeneic HCT. SPSS (v29) was used to estimate all summary statistics, cumulative incidences, and uni- and bi-variate analyses. A total of 558 transplants were performed with 43 patients developing TA-TMA, for a 1-year cumulative incidence of 8.6% (95% CI, 5.9-11.3) and 7.2% (95% CI, 2.9-11.5) in allogeneic and autologous HCTs, respectively (P = .62). In allogeneic recipients (n = 417), the 1-year cumulative incidence of TA-TMA with CY + TT as part of conditioning was 35.7% (95% CI, 15.7-55.7) compared to 11.7% (95% CI, 7.2-16.2) with either CY or TT alone, and 1.2% (95% CI, 0-2.8) if neither agent was included in the conditioning regimen (P < .001). Use of either CY or TT (HR = 10.14; P = .002) or CY + TT (HR = 35.93; P < .001), viral infections (HR = 4.3; P = .017) and fungal infections (HR = 2.98; P = 0.027) were significant factors resulting in increased risk for developing TA-TMA. In subjects undergoing autologous HCT (n = 141), the 1-year cumulative incidence of TA-TMA with CY + TT was 19.6% (95% CI, 8.8-30.6) while TA-TMA did not occur in patients receiving either CY or TT alone or when neither were included (P < .001). TA-TMA occurred only in patients with neuroblastoma receiving CY + TT as part of their conditioning. For autologous patients who received CY + TT, those who were CMV seronegative at the time of HCT had an incidence of TA-TMA of 6.7% (95% CI, 0.1-15.7) compared to 38.1% (95% CI, 35-41.2) for those CMV seropositive (P = .007). These data show that CY or TT alone or in combination as part of pre-transplant conditioning prior to HCT increase the incidence of TA-TMA. Alternative conditioning excluding the combination of CY + TT should be considered whenever possible to limit the development of TA-TMA.