Key points are not available for this paper at this time.
Allogeneic hematopoietic stem cell transplantation (allo-HSCT) remains the only potentially curative treatment for patients with high-risk myeloid neoplasms (HR-MN) 1. Despite the recent progress in donor selection, graft versus host disease (GVHD) prophylaxis, and supportive care, disease relapse remains a major cause of death post-transplant. Currently, several prospective ongoing clinical trials (i.e., AMADEUS, VIALE-T) are assessing the role of maintenance therapy following HSCT. In 2020, Gao et al. evaluated the role of low-dose granulocyte colony-stimulating factor (G-CSF) with decitabine (LD-GD) as a post-transplant maintenance strategy. This demonstrated a significant improvement in the cumulative incidence of 2-year relapse (HR 0.32, p 10% blasts. All relapsed patients had either complex karyotype (CK), monosomal karyotype (MK), or both complex and monosomal karyotype (CK + MK) cytogenetic profiles. Twenty-five patients (50%) had grade 3–4 adverse events; 22 (88%) related to cytopenia and seven (28%) related to infection. Following maintenance, grade III-IV acute graft versus host disease (aGVHD) occurred in five (10%) patients and steroid-requiring chronic GVHD (cGVHD) occurred in 11 patients (22%), with severe cGVHD occurring in three patients (6%). A total of 10 (20%) deaths occurred at the time of last follow-up, four related to disease progression, while six were attributed to nonrelapse mortality (4 sepsis, 1 stroke, 1 hemorrhage). A total of 219 patients were analyzed for propensity-score matching (173 control, 46 LD-GD); four LD-GD patients were excluded due to relapse before day +100 or diagnosis of CMML/MPAL (Table S6). The median age at allo-HSCT was 64 years (IQR: 56–67 years) and the median follow-up after transplant was 33.6 months (95% CI 28.8–37.2). The median time to maintenance initiation was 109 days after transplant (IQR: 87–147 days). A total of 138 patients were included after propensity score matching (Table 1, Figures S3, S4). Patients receiving maintenance had a significantly lower relapse rate at 1 year after transplant (4.3% vs. 23.5%, p = 0.007). At 2 years post-transplant, relapse rates favored maintenance LD-GD (18.7% vs. 26.2%, p = 0.11) (Figure S5). The 2-year RFS was superior in patients who received maintenance therapy (74.5% vs. 56.9%, p = 0.022) (Figure 1a, S6a) and a strong trend towards better RFS was also observed after evaluating maintenance therapy as a time-dependent covariate (73.4% vs. 58.2%, p = 0.069) (Figure S6b). The 2-year OS was also higher in patients receiving maintenance therapy (89.0% vs. 65.6%, p = 0.01, Figures 1b, S7a). Evaluating post-transplant maintenance therapy as a time-dependent covariate showed that LD-GD was associated with an improved 2-year OS (2-year OS rate 88.7% vs. 66.2%, p = 0.026, Figure S7b). Median OS of the entire cohort was not reached. Forty-two patients (30.4%) received post-transplant cyclophosphamide (PTCy) as GVHD prophylaxis; 23 (55.8%) of whom received LD-GD maintenance. The median follow-up of these patients was 12.8 months (95% CI: 11.3–18.5). In patients receiving PTCy, LD-GD was associated with a significantly lower 1-year cumulative incidence of relapse (8.7 vs. 40.7%, p = 0.049, Figure S8). In multivariate competing risks analysis for relapse, LD-GD was associated with a significantly decreased risk of relapse (HR 0.28, 95% CI: 0.0996–0.78, p = 0.015). Other prognostic factors included: complex karyotype (HR 3.72, 95% CI: 1.51–9.14, p = 0.004), CR at transplant (HR 0.33, 95% CI: 0.14–0.80, p = 0.014), and high/very high DRI (HR 4.45, 95% CI: 1.03–19.26, p = 0.046) (Table S4). In multivariate overall survival analysis, LD-GD maintenance was associated with a significantly improved 2-year OS (HR 0.28, 95% CI 0.08–0.95, p = 0.04), while high/very high DRI was associated with worse 2-year OS (HR 2.42, 95% CI: 1.03–5.72, p = 0.04) (Table S5). Here, we report our experience using LD-GD maintenance in HR-MN following allo-HSCT. This marks the first real-world experience of LD-GD maintenance therapy. Our findings build upon previous literature by including an older patient population receiving a reduced intensity conditioning regimen, including both AML and MDS. Overall, LD-GD was feasible and well tolerated; patients received a median of 4 cycles. The grade 3–4 adverse event rate was 50%, primarily related to infection and cytopenias. Although adverse event rates were higher than reported in Gao et al., our rates of severe aGVHD and steroid-requiring cGVHD were low at 10% and 22%, respectively. When compared to propensity-matched control patients, patients with LD-GD demonstrated significantly improved survival and decreased relapse rates. These findings remained consistent in patients receiving PTCy. Our study is limited by the retrospective design, lacking data on intent to treat, especially for those who did not receive maintenance due to early relapse, provider decisions, or patient comorbidities. Although we attempted to compare similar patient populations and isolate the impact of LD-GD maintenance using propensity matching on HCT-CI and DRI, the retrospective design precludes drawing causal conclusions. There was also a statistically significant difference between our maintenance and control cohorts' representation of primary diagnosis. The control cohort was enriched with MDS compared to the maintenance cohort (41% vs. 17.4%, p = 0.005). Due to limited control sample size, we were unable to propensity match to diagnosis and utilized DRI to counterbalance this discrepancy in diagnosis. MDS carries a worse prognosis compared to AML, with increased incidence of relapse 3 which was supported by the increased hazard ratio of MDS (HR 2.57, p = 0.009) in univariate analysis. Therefore, our results could be partially driven by the higher proportion of AML in the maintenance cohort compared to the control cohort. Additional differences between cohorts included younger patients (61.4 vs. 65.9 years, p = 0.008), higher rates of myeloablative conditioning (37% vs. 19.6%), and decreased rates of RAS mutations (32.6% vs. 14%, p = 0.04) in the control cohort compared to the LD-GD cohort. Previous literature has demonstrated that both age and RAS mutations negatively impact transplant outcomes 4. Our data demonstrated that despite advanced age, the maintenance cohort had better outcomes. GVHD prophylaxis regimens differed between the two groups, with significantly higher PTCy in the maintenance cohort compared to the control (50% vs. 20.7%, p = 0.0015). This may have contributed to the modest rates of GVHD in our maintenance cohort. However, in a dedicated subgroup analysis of patients who received PTCy, LD-GD maintenance remained associated with improved outcomes. The mechanism of LD-GD is not fully understood, but decitabine has been shown to enhance both NK cell and effector T cell activity, which may act synergistically with the expansion of NK cell populations from rhG-CSF 5 to improve the graft versus leukemia effect. Recently, several targeted therapies have demonstrated improvements in relapse rates as post-HSCT maintenance therapies 6, patients who lack targetable mutations may benefit from LD-GD maintenance. Our findings would support the use of LD-GD in these high-risk patients, and further prospective randomized studies would be beneficial to further quantify the benefit of LD-GD maintenance. Mathias Palmer: data collection, cohort statistical analysis, initial manuscript drafting. Anmol Baranwal: data collection, matched pair analysis, figure generation, manuscript editing. Rami Basmaci and Khalil Hassan: data collection and manuscript review. Jade Kutzke, Gabriel Bartoo: manuscript review. Hemant Murthy, James Foran, Abhishek A. Mangaonkar, Mehrdad Hefazi, Aasiya Matin, William J. Hogan, Mark Litzow and Mithun Vinod Shah: contributed patients and manuscript review. Hassan B. Alkhateeb: study conceptualization, contributed patients, data collection, data analysis, manuscript drafting, editing, and review. The authors have nothing to report. There are no conflicts of interest to disclose. This study was approved by the Mayo Clinic Institutional Review Board. Some of the data was presented at the European Hematology Association 2024 conference. The data that support the findings of this study are available from the corresponding author upon reasonable request. Data S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
Palmer et al. (Tue,) studied this question.