Survival Impact of Early Versus Late Discontinuation of Immunosuppression after Allogeneic Stem Cell Transplantation

The optimal duration of Graft-Versus-Host-Disease (GVHD) prevention following allogeneic stem cell transplantation (alloSCT) has not been well studied. Although practice varies among transplant centers, immunosuppression (IS) is typically tapered or stopped between 3 and 6 months post alloSCT. Moreover, disease risk, adverse IS reactions, and donor-recipient HLA disparity may influence the duration of GVHD prophylaxis. Therefore, we conducted a single center retrospective analysis to determine the effects of stopping IS prior to 180 days versus continuing for 180 days or more on overall survival (OS), relapse free survival (RFS), chronic GVHD free relapse free survival (CGFRFS) and non-relapse mortality (NRM).

We evaluated patients that received an alloSCT between 1/1/13 and 12/31/23. We excluded patients that were receiving a second alloSCT, and those that had GVHD or died prior to day 100 after alloSCT. Three hundred and fifty two patients met these criteria. Median age was 55 (range 19-78) and 55% were male. The most common diagnoses were acute myelogenous leukemia (36%), myelodysplastic syndrome (21%), acute lymphocytic leukemia (18%) and non-Hodgkin's lymphoma (7%). IS was discontinued prior to 180 days in 185 patients (53%). Donor types were matched-related (42%), haploidentical (31%) or matched-unrelated (27%). Eighty-nine percent of the patients were given GVHD prevention with either a post-cyclophosphamide (P-Cy) based regimen (41%) or a tacrolimus/methotrexate regimen (48%). Patient demographics and disease-related characteristics were similar in each group except for the Dana Farber disease risk index (DF-DRI) and more patients on IS < 180 days received post-transplant cyclophosphamide based GVHD prevention. Acute GVHD occurred in 30% of the IS < 180 day group versus 19% in the > 180 day group. Chronic GVHD incidence was similar between the groups (35% vs. 41%). The 3-year survival estimates for IS < 180 vs. IS > 180 were OS=77% vs. 91% (p=0.009), RFS=76% vs. 90% (p=0.026), CGFRFS=72% vs. 80% (p=0.36), and NRM=12% vs. 3% (p=0.08). Using a 1-year landmark analysis, OS and RFS were significantly better in the group taking IS > 180 days. A multivariate Cox analysis revealed the significant variables for OS with a hazard ratio (HR) > 1 were stopping IS < 180 days (HR 8.6, p<0.001), patient age > 55 (HR 2.42, p<0.001), high/very high DF-DRI (HR 1.72, p<0.010), and non-myeloablative versus reduced intensity conditioning (HR 2.26, p=0.005). P-Cy containing GVHD prevention had the only significant positive survival benefit of HR< 1 (HR .60, p=0.033). Interestingly, the significant variables for RFS with a HR > 1 were patient age > 55, high/very high DF-DRI, and non-myeloablative versus reduced intensity conditioning. Stopping IS <180 days had no benefit on in reducing relapse risk (p=0.27).

This analysis highlights how complicated the post-transplant period is with several variables effecting GVHD, disease-related, and survival outcomes. More systematic prospective studies are needed to determine the ideal duration of IS. Moreover, studies that incorporate newer GVHD prevention strategies like cyclophosphamide and abatacept are needed to evaluate their effects on IS duration, survival, GVHD, and other long-term effects.

Solh:GlaxoSmithKline: Speakers Bureau; Sanofi: Consultancy; Bristol Myers Squibb: Consultancy, Speakers Bureau.