The Use of MSCs in Steroid-Refractory Acute GvHD in Europe: A Survey from the EBMT Cellular Therapy & Immunobiology Working Party

Despite novel developments in conditioning regimens and graft engineering, acute graft-versus-host disease (aGvHD) remains a life‐threatening complication of allogeneic hematopoietic stem cell transplantation (HSCT). Around 40% of patients with aGvHD do not respond to high-dose steroid therapy. Multiple therapeutic options have been explored such as ruxolitinib and mesenchymal stromal cells (MSCs). While ruxolitinib has been approved as second line therapy for aGvHD, the role of MSCs still needs to be determined.

To better understand the administration and outcomes of MSCs in steroid-refractory aGvHD (SR-GvHD), the EBMT Cellular Therapy & Immunobiology Working Party (CTIWP) conducted a survey among EBMT centers. Information on the number of administrations, MSC dose and outcomes was obtained from centers that had performed more than 20 MSC treatments between 2007 and 2020. Responses were received from 6 participating EBMT centers in 6 different countries, resulting in 313 patients. Median patient age was 55 years old (range 17.6-74.5). Patients had various hematological diagnoses, the majority being AML (33.5%) and MDS (14.4%). Most transplants were performed with unrelated donors (75.1%). 44.9% of patients received myeloablative conditioning, while 55.1% received reduced intensity conditioning. ATG was used in 50.2% of cases, PTCy was used in only 3.2% of cases. The median number of aGvHD treatments before administration of MSCs was 3 (range 1-7). Most patients were treated before approval of ruxolitinib, with only 5% of patients receiving ruxolitinib before MSCs. At the start of MSC infusion, GVHD grade was as follows: grade 1: 2.9%, grade 2: 14%, grade 3: 56% and grade 4: 27%. The primary endpoint was the proportion of patients reaching an overall response, defined as either complete response (CR) or partial response (PR) of aGvHD on day 28 after MSC administration, where CR is defined as the return of aGvHD to overall grade 0, and PR is defined as improvement in overall grade by at least one stage. Information on organ response was not included due to a large amount of missing data. Death before day 28 was considered a competing event. As a secondary endpoint the response on day 7 was calculated.

We found that on day 28, 44.5% of patients showed an overall response to MSCs. On day 7, the overall response rate was 26.5%. Overall survival in this cohort was 30.7% (95% confidence interval: 25.6-35.9%) at 12 months after MSC infusion. Using a landmark analysis, the overall survival at 12 months, conditional on being alive at day 28, was 39.2% (33.1-45.4%). The overall survival in responding patients was 48.6% (39.4-57.7%) at 12 months, compared to 24.4% (14.8-33.9%) in non-responders alive at day 28, which was a statistically significant difference (p. <0.001). The average number of MSC infusions was 4 (range 1-9) in responders and 3 (range 1-7) in non-responders. The cumulative incidence of relapse of the underlying hematological malignancy at 12 months after MSC administration was 9.5% (6-12.9%), while non-relapse mortality (NRM) was 58.8% (53-64.7%) and relapse free-survival was 31.7% (26.2-37.2%).

A previous study has shown that the response to MSC therapy on day 7 is predictive of the response on day 28 (Galleu, BJH 2019). Our study confirms these results: The vast majority (84.8%) of responders at day 7 remained responders at day 28. However, among non-responders alive at day 7, 35.1% achieved a response by day 28. When performing multivariate analysis on predictors of MSC outcome on day 28, only being a responder on day 7 was significantly associated with response on day 28 (p. <0.001), as opposed to the conditioning regimen, the interval from onset of GvHD to MSC administration, patient age, MSC dose and – interestingly - GvHD grade at first MSC administration, which were not significantly associated with response on day 28.

This is the first comprehensive survey on the use of MSCs in Europe. Limitations include that most patients were treated before the approval of ruxolitinib, the retrospective design, and the differences in products and procedures between centers. Nonetheless, our results suggest a treatment benefit of MSCs in heavily pretreated patients, and show that MSCs manufactured in academic facilities are a viable treatment option for steroid-refractory aGvHD, which can be used in third line after steroids and ruxolitinib.