Correlations between Pre-Transplant Treatment and Bone Marrow Response with Post-Transplant Outcomes in Patients with Myelodysplastic Syndrome Undergoing Allogeneic Stem Cell Transplantation

Background: Allogeneic stem cell transplant remains the only curative modality for myelodysplastic syndrome (MDS), yet there is significant controversy and individual practice variation in offering pre-transplant therapy with the goal of reducing disease burden and/or improving marrow function. In this study, we analyze the impact of bone marrow response prior to transplant on the post-transplant overall survival and relapse.

Methods: We collected data from 2007 to 2018 from patients undergoing allogeneic transplant for MDS at Weill Cornell Medicine. Patients with MDS transformed to AML were excluded. P values for survival analyses were derived from Cox models. All statistical tests were 2 sided with 0.05 alpha levels.

Results: We analyzed 85 patients who underwent an allogeneic stem cell transplant for MDS and had all pertinent data available. Response to pre-transplant treatment was not a pre-requisite for proceeding to transplant. The donor source was MUD, MRD and cord blood in 40%, 37.65%, and 22.35% of patients, respectively. Median age at the time of transplant was 61 years (32.94% females). IPSS-R at diagnosis was low, intermediate, high and very high-risk in 12.94 %, 31.76 %, 24.71% and 30.59% of patients, respectively. Most (80%) patients were treated with hypomethylating agents (HMA) prior to transplant.

Within 2 months of transplant, 27.06% of patients achieved CR, with 8% cytogenetic response. Hematological improvement (HI) only (without marrow CR), marrow CR and stable disease/no response were seen in 9.41%, 18.83% and 42.35% of patients, respectively. Bone marrow blasts were 0-5%, 5-10% and 11-19% in 77.5%, 18.75% and 3.75% of patients pre-transplant, respectively. Median overall survival after transplant was 28.6 months (range 14.9-48.5). 90-day transplant related mortality was 10.7 %. One-year survival after transplant was 65.7%; with a relapse rate of 30.6% at median follow up of 12 months. The incidence of grade 3 and 4 GVH of any kind was 33.33%.

Using univariate models, age at transplant (HR 1.049, 95% C.I. 1.017-1.083, p value 0.002) and a high IPSS-R (p value.004) at diagnosis were associated with inferior post-transplant survival, while gender and use of HMA prior to transplant did not impact post-transplant survival. In multivariate cox regression analyses adjusting for other confounding variables, achieving a CR or HI was significantly associated with higher post-transplant survival, compared to stable disease/no response prior to transplant (HR 0.46, 95% C.I. 0.23-0.90, p value 0.02). Achievement of a marrow CR resulted in a trend toward higher post-transplant OS (HR 0.45, 95% C.I. 0.18-1.09, p value 0.07). Pre- transplant blast percentage (<5, 5-10 or >10%) or cytogenetic response only (without CR) did not impact survival outside the presence of a CR. In the multivariate models IPSS-R very high risk at baseline (HR 4.75, 95% C.I. 1.81-12.43, p value 0.001) was associated with reduced post-transplant OS compared to low risk. There was no association between the types of marrow response on RFS post-transplant. Of note, among 34 patients with baseline mutations in ASXL1, ETV6, EZH2, RUNX1 and TP53, mutation clearance at the time of transplant was observed in only 6/34 (17.6%) patients. The presence of a mutation in any of these 5 genes prior to transplant did not impact RFS (HR 1.12, 95% C.I. 0.38-3.26, p value 0.04) or OS (HR 3.09, 95% C.I. 0.86-11.11, p value 0.08). However, we were not able to assess the effect of individual mutations (e.g. TP53) due to small numbers. We did not see an effect of GVHD type (acute or chronic) or grade on RFS or OS.

Conclusions: Achievement of CR or hematological improvement (HI) prior to allogeneic transplant was associated with superior survival post-transplant, while marrow CR and cytogenetic response did not impact post-transplant survival any different than going into allogeneic transplant with stable disease/no response. IPSS R score at diagnosis was associated with inferior post-transplant survival. The presence of mutations in ASXL1, ETV6, EZH2, RUNX1 or TP53 genes prior to transplant did not appear to impact OS or RFS post-transplant, but the total numbers were small. Although CR or HI after pre-transplant treatment improved OS post-transplant, only one third of patients achieved these responses prior to transplant. Better MDS directed treatments and conditioning regimens are needed to improve outcomes after allogeneic transplantation.

Figure.

View largeDownload slide

Figure.

View largeDownload slide

Close modal