Effects of HLA Mismatches on Cytokine Release Syndrome and Associated Non-Relapse Mortality in Allogeneic Stem Cell Transplantation with Post-Transplant Cyclophosphamide

Background Post-transplant cyclophosphamide (PTCy) is an effective method for graft versus host disease prevention in allogeneic stem cell transplantation (alloSCT). The alloreactive T-cell response in the first days after PTCy-alloSCT can induce a systemic inflammatory response known as cytokine release syndrome (CRS). The incidence of CRS appears to increase with higher degrees of HLA mismatch between patient and donor. In haplo-identical PTCy-alloSCT, CRS occurs frequently (73-94%) and has a negative impact on non-relapse mortality (NRM) and survival. In 10/10 HLA matched related and unrelated allogeneic stem cell transplantation (MRD/MUD PTCy-alloSCT), CRS is less frequently observed (14-23%) without effects on NRM and survival. The specific contributions of different HLA mismatches in CRS development are not well known.

Aim To study the effects of HLA mismatching in MRD/MUD PTCy-alloSCT transplantation on development of CRS and CRS-associated NRM and survival.

Methods We retrospectively studied CRS, survival and NRM in a group of 97 MRD/MUD PTCy-alloSCT patients with different degrees of HLA matching (20% of patients were 12/12, 50% 10/10 and 30% 9/10 HLA-matched).

Results CRS developed in 69% of patients, mostly grade 1 (85%), 12% grade 2 and 3% grade 3. CRS developed significantly more in patients transplanted with a 9/10 or 10/10 match compared to a 12/12 match (90% and 75% versus 26%). By focusing on the maximum grade of fever developing during CRS, CRS severity grading was refined, revealing effects of specific HLA mismatches on CRS. In 10/10 matched patients, HLA-DPB1 non-permissiveness significantly increased grade 2-3 fever CRS incidence from 10 to 62%. Although regarded to be unimportant in HLA matching, single HLA-DRB3/4/5 mismatches could induce grade 3 fever CRS. Strongest effects on fever developed in 9/10 matched patients (35% grade 3 fever CRS), showing differences between various HLA mismatch groups. HLA-DRB1 mismatches induced significantly higher fever than HLA-C mismatches. The total number of HLA mismatches (combined HLA-A, B, C, DRB1, DRB3/4/5, DQ and DP non-permissive mismatches) was found to correlate with the height of fever developing. Importantly, patients with grade 3 fever CRS showed significantly increased NRM and significantly inferior survival compared to patients with grade 0-2 fever CRS.

Conclusions In contrast to earlier publications in which CRS was demonstrated mainly in haplo-identical PTCy-alloSCT, we show that also in HLA matched related and unrelated PTCy-alloSCT incidence of CRS can be high and influence transplantation results. Development of CRS is dependent on the degree of HLA matching (90% in 9/10, 75% in 10/10 and 26% in 12/12 matched patients). By focusing on the maximum grade of fever developing during CRS, CRS severity grading can be refined, revealing effects of specific HLA mismatches on CRS. The height of fever during CRS is dependent on the number and types of HLA mismatches. CRS with grade 3 fever increases NRM and decreases survival. HLA matching decisions may be used to prevent grade 3 fever CRS, and thereby possibly also NRM. Alternatively, anti-inflammatory interventions may be considered in patients developing grade 3 fever early after alloSCT.

No relevant conflicts of interest to declare.