Severe Cytokine Release Syndrome Associated with Worse Outcomes in Recipients of a Haploidentical Stem Cell Transplant

Introduction: The utilization of haploidentical (haplo) donors for allogeneic hematopoietic stem cell transplant (HCT) recipients has allowed more donor options to treat a multitude of hematopoietic malignancies and disorders. However, there are substantial complications and morbidities associated with haplo HCT. Cytokine release syndrome (CRS), a manifestation of alloreactive T cell expansion, has posed an obstacle early after stem cell infusion. CRS is well described among chimeric antigen receptor (CAR) T cell therapy recipients, where anti-interleukin (IL) 6 therapy and glucocorticoids to abrogate cytokine activity depends on objective physiologic markers and neurologic changes. In 2019, the American Society of Transplantation and Cellular Therapy (ASTCT) developed a consensus CRS grading scale for CAR T cell therapy that has since been widely adopted for these recipients (Lee et al, Biol Blood Marrow Transpl 2019). While the etiology of CRS differs between haplo HCT and CAR T cell recipients, severe CRS (grade 3 or 4) is seen in both patient populations, defined by ASTCT CRS grading as non-fluid responsive hypotension and/or capillary leak necessitating increased ventilatory support. We investigated whether the newly adopted ASTCT CRS grading could apply to a haplo HCT population in predicting overall survival (OS) and progression-free survival (PFS) at 1 year as well as the 2-year probability of graft-versus-host disease (GVHD).

Methods: This IRB-approved retrospective analysis assessed 64 patients who underwent haplo HCT at Colorado Blood Cancer Institute from July 2015 to December 2018. All but one patient received peripheral blood-derived stem cells, and all patients received post-transplant cyclophosphamide, tacrolimus, and mycophenolate mofetil for GVHD prophylaxis. Variables of birth sex, primary disease, disease risk index (DRI), blood type, degree of HLA-mismatching, presence of donor-specific antibodies, Hematopoietic Cell Transplant Comorbidity Index (HCT-CI), conditioning regimen, and cell dose were assessed. In all patients, fever, blood pressure, and oxygen supplementation were used to identify the maximum CRS grade between 0-5 days after HCT using ASTCT CRS grading. Patients with other etiologies (i.e. infection, cardiomyopathy) were excluded from the analysis. CRS grading was then dichotomized to mild (CRS grades 0-2) or severe (grades 3-4), reflecting the physiologic changes that align with minimal compared to intensive cardiopulmonary support. As all patients had complete follow-up to 2 years, logistic regression was used to estimate association between severe CRS and outcome.

Results: When adjusted for age, DRI and HCT-CI, development of severe CRS was associated with an inferior PFS at 1 year (odds ratio (OR) 0.22, (95% Confidence Interval (CI) 0.07, 0.69), p = 0.01). Inferior OS was also seen in those who developed severe CRS but the magnitude of the effect was smaller and not considered definitive (OR 0.38, (95%CI 0.13, 1.13), p=0.082). Increased non-relapse mortality (NRM) was also seen with severe CRS. Due to the low number of events, adjustment for only one factor at a time was necessary. When adjusted for age (OR 2.94, (95%CI 0.82, 11.1), p=0.101), DRI (OR 3.23, (95%CI 0.91, 11.1), p=0.069), HCT-CI (OR 2.94, (95%CI 0.85, 10), p=0.089), and conditioning regimen (OR 3.23, (95%CI 0.91, 11.1), p=0.071), consistent results were demonstrated, although not definitive. This was also observed with the risk of relapse, adjusted for conditioning regimen (OR 2.56, (95%CI 0.76, 9.09), p=0.126) and DRI (OR 2.56, (95%CI 0.75, 8.33), p=0.132). There was no demonstrable association between CRS and the probability of acute GVHD. However, severe CRS was associated with a reduction in the development of any chronic GVHD at 2 years (OR 0.24, (95%CI 0.08, 0.72), p=0.011).

Conclusion: This analysis suggests that an adapted ASTCT CRS grading, dichotomizing into those with mild (grades 0-2) or severe (grades 3-4) CRS may identify haplo HCT recipients at risk for inferior outcomes. Haplo HCT recipients with severe CRS demonstrated an inferior PFS. The data also showed numerical decreases in overall survival, higher relapse, and higher NRM. On the other hand, severe CRS was associated with a decreased probability of chronic GVHD. As the analysis was limited to a single site, validation of adapted ASTCT grading in a larger haplo HCT population is necessary.