Decreased Chronic <scp>GvHD</scp> With Post‐Transplant Cyclophosphamide Versus <scp>ATG</scp> in Patients With Myelofibrosis Allografted With an Unrelated Donor: A Study From the Chronic Malignancies Working Party of the <scp>EBMT</scp>

The only curative option for myelofibrosis (MF) patients remains allogeneic hematopoietic cell transplantation (allo-HCT). Advances in transplant platforms, including the introduction of post-transplant cyclophosphamide (PTCY) and anti-thymocyte globulin (ATG), have reduced GvHD incidence and improved outcomes. Both strategies are effective in matched unrelated donor (MUD) transplants, with results comparable to matched related donors (MRD), and are endorsed by the latest European Society for Blood and Marrow Transplantation (EBMT) guidelines [1-3]. However, the optimal GvHD prophylaxis for MF patients undergoing allo-HCT utilizing an UD remains unresolved. We conducted a retrospective study of adult patients who underwent first allo-HCT at an EBMT center between 2012 and 2022, diagnosed with either primary (pMF) or secondary myelofibrosis (sMF), and transplanted with a BM or PB graft from an UD. All patients received ATG or PTCY as GvHD prophylaxis without Alemtuzumab or ex vivo graft manipulation (Figure S1). More information and statistical analyses are provided in the Supporting Information: Methods S1. Patients' characteristics for the series of 2607 MF patients (PTCY: 192; ATG: 2415) are summarized in the Supporting Information: Results and Table S1. Cumulative incidence of neutrophil engraftment at 28 days was significantly better with ATG compared to PTCY (87% vs. 74%, p < 0.001, Table 1). Median time to neutrophil engraftment was 23 days (95% CI 21–24) with PTCY versus 17 days (95% CI 17–18) with ATG. Platelet engraftment occurred at a median of 38 days (95% CI 31–44) with PTCY versus 22 days (95% CI 22–23) with ATG. The cumulative platelet recovery at day 100 was also significantly better with ATG compared to PTCY (80% vs. 67%, p < 0.001, Table 1). However, the incidence of primary graft failure by Day 42 did not differ significantly between PTCY and ATG (4% vs. 2%, p = 0.06, Table 1). The cumulative incidence of grade II–IV aGvHD was similar between the two groups: 27% (95% CI 20%–33%) for PTCY versus 31% (95% CI 29%–33%) for ATG, p = 0.17 (Figure 1A). This also held true for grade III–IV aGvHD (Figure 1B). When accounting for confounding factors, there was no statistically significant effect of PTCY on grade II–IV aGvHD (HR, 0.84; 95% CI 0.59–1.20, p = 0.3), or grade III-IV aGvHD (HR, 0.91; 95% CI 0.56–1.48, p = 0.7). By contrast, the 3-year cumulative incidence of cGvHD was significantly lower with PTCY than with ATG (27%, 95% CI 19%–35% vs. 43%, 95% CI 41%–45%) (p < 0.001, Table 1, Figure 1C). This finding was confirmed in multivariable analysis, with an HR for PTCY of 0.59 (95% CI 0.4–0.87, p = 0.008; Table 2). Full Cox models are reported in Table S3 and no other factors significantly influenced cGvHD. A similar pattern was observed for severe cGvHD, with an incidence of 4% (95% CI 1%–8%) with PTCY versus 10% (95% CI 9%–11%) with ATG, p = 0.028 (Table 1), consistent with multivariable analysis (Table 2). There was also an improved cGvHD-free survival (cGFS) with PTCY; the 3-year cGFS was 40% (95% CI 31%–49%) for PTCY compared to 26% (95% CI 24%–28%) for ATG (p = 0.002, Table 1). More information on cGvHD and comparison of ATG brands are provided in the Supporting Information: Results and Figures S2 and S4. After a median follow up of 3.1 years (95% CI 3–3.3), 3-year OS was 61% (95% CI 53–70) with PTCY and 60% (95% CI 58%–62%) with ATG, p = 0.6 (Table 1 and Figure S3A). After accounting for confounding factors in multivariable analysis, there was no statistically significant effect of GvHD prophylaxis on OS (PTCY HR, 0.84; 95% CI 0.6–1.18, p = 0.3). Donor type, however, was significant, with an HR of 1.36 for MMUD (95% CI 1.14–1.62), p = 0.001 (Table S4 which also reports the effects of other factors). The 3-year PFS was 57% (95% CI 48–66) with PTCY versus 52% (95% CI 50–54) with ATG, p = 0.2 (Table 1 and Figure S3B), with no significant difference by donor type (Table 2). After adjusting for confounding factors, GvHD prophylaxis had no significant effect on PFS (HR, 0.79; 95% CI 0.58–1.08, p = 0.14), while donor type was significant (Table S4). Other factors affecting PFS are detailed in Table S4. The 3-year GRFS was significantly higher with PTCY compared to ATG, 48% (95% CI 39%–57%) versus 35% (95% CI 33%–37%), p = 0.014 (Table 1 and Figure S3C). GRFS was also significantly better with MUD compared to MMUD (Table S2). In multivariable analysis PTCY remained beneficial for GRFS (HR 0.72 (95% CI 0.55–0.95), p = 0.02, Table 2), while donor type had a significant effect (HR 1.31 (95% CI 1.13–1.51), p < 0.001 Table S4). There was a tendency toward lower 3-year cumulative incidence of relapse (RI) with PTCY (14% (95% CI 8–20) vs. 21% (95% CI 19–23)) with ATG, p = 0.1 (Table 1 and Figure S3D). This trend persisted in multivariable analysis (HR, 0.65 (95% CI 0.39–1.09), p = 0.1 Table 2). No other factors significantly impacted RI in multivariable analysis (Table S4). Finally, the 3-year NRM was 28% (95% CI 20%–37%) versus 27% (95% CI 25–29) with PTCY and ATG, respectively, p > 0.99 (Table 1 and Figure S3E). To our knowledge, this study is the first focusing on MF patients undergoing allo-HCT from UD and directly comparing outcomes with PTCY or ATG for GvHD prophylaxis. Leveraging the large EBMT network, we found that PTCY was associated with a lower risk of cGvHD, as well as improved GRFS and cGFS compared to ATG. While originally developed for haploidentical transplants, the PTCY platform is increasingly used with matched related and UD with promising results [4-6]. In general, PTCY-based GvHD prophylaxis has been associated with reduced incidence of acute and/or chronic GvHD compared to standard methods [4-6], but it has not been prospectively compared to ATG, nor specifically in patients with MF. In the present study, the benefit of PTCY was seen in cGvHD, which contrasts with our recent findings in myelodysplastic neoplasms (MDS) in the UD setting, which showed a lower incidence of grade II–IV aGvHD with PTCY compared to ATG [7], but no differences in cGvHD. Both studies, however demonstrated improved GRFS with PTCY. Most large studies comparing PTCY and ATG in UD transplants have been retrospective, focusing on acute leukemia [3, 8, 9] or heterogeneous groups of hematologic malignancies [10-13], which may preclude direct comparison with the present results. Although most studies show less GvHD with PTCY—possibly explained by the down-regulation of alloreactive T-cells alongside upregulation of Treg cells [3, 4, 6, 9, 14-17]—it is difficult to understand why in some setting this concerns only aGvHD, in others only cGvHD, or both. This may be related to the disease itself, disease-specific bone marrow microenvironment, or differences in conditioning regimens (e.g., more thiotepa for MF, more TBI for ALL). One important issue in allo-HCT for MF is the lower and delayed engraftment compared to other hematologic malignancies. As in the EBMT MDS UD allo-HCT study [7], we found both lower and delayed engraftment with PTCY compared to ATG. Neutrophil recovery was delayed by 6 days, and platelet recovery by 16 days, despite the majority of stem cell source being PBSC (96%). This is also in concordance with most studies reporting delayed or lower neutrophil and platelet engraftment, except for one study showing no differences [10, 11, 13, 18]. Interestingly, the slightly higher incidence of GF with PTCY compared to ATG (4% vs. 2%, p = 0.06) did not reach statistical significance, which is encouraging and supports the feasibility of PTCY in this setting. Limitations of this study are inherent to its registry-based design, which can introduce bias. Heterogeneity in GvHD prophylaxis, conditioning regimens, and the unknown rationale for choosing ATG or PTCY further complicate the interpretation of results and underscore the need for randomized trials. Moreover, we did not analyze the impact of pre-transplant spleen size and CD34+ cell dose on outcomes, as these factors have been addressed in previous EBMT studies [19, 20], nor were quality-of-life data and other transplant complications (e.g., infections and toxicities) available. However, there are also some clear strengths and messages: this study included a large cohort of MF patients undergoing UD allo-HCT, across many centers, reflecting real-world procedural diversity, which may enhance the generalizability of the findings. In conclusion, the results of this analysis suggest that for MF patients undergoing UD allo-HCT, PTCY is a valid alternative to the standard EBMT approach using ATG. PTCY appears to decrease the incidence of cGvHD and improve GRFS, without affecting relapse incidence or NRM. While delayed engraftment is a consideration, this did not result in higher rates of GF. Given its relatively low cost and logistical simplicity, PTCY may also be suitable in resource-limited settings. Nevertheless, there is a clear need for further cooperative, prospective randomized trials to confirm these findings. Y.C., D.P.M., and J.C.H.-B. designed the study. Y.C., E.F.B., D.-J.E., L.C.d.W., L.K., J.T., J.P., J.S., U.P., I.Y.-A., U.S., M.V., M.A.d.W., J.-B.M., J.A.S., T.A.W.H., G.v.G., F.B., M.R., X.P., M.W.H.R., M.I.-R., J.Z., K.R., J.D.-S., G.B., N.P., T.C., J.C.H.-B., and D.P.M. contributed data and reviewed the manuscript. Y.C., E.F.B., and D.-J.E. analyzed the data. Y.C., D.-J.E., D.P.M., J.C.H.-B., and N.P. wrote the manuscript. We thank all the centers that report data to the EBMT contributing to that study, all the data managers involved, the health team that cared for the patients, and all the patients and their families. The authors have nothing to report. The EBMT is a non-profit, scientific society representing more than 700 transplant centers, mainly in Europe. EBMT centers commit to obtain informed consent according to the local regulations applicable at the time to report pseudonymized data stored in a central database. Y.C. has received consulting fees for advisory board from MSD, Novartis, Incyte, BMS, Pfizer, Abbvie, Roche, Jazz, Gilead, Amgen, Astra-Zeneca, Servier, Takeda, Pierre Fabre, Medac all via the institution; Travel support from MSD, Roche, Novartis, Pfizer, BMS, Gilead, Amgen, Incyte, Abbvie, Janssen, Astra-Zeneca, Jazz, Pierre Fabre, Sanofi all via the institution. N.P. served as advisor for GSK, Kite Pharma, Medac Pharma, and Novartis. Travel support from Neovii. J.D.-S. has received honoraria from AbbVie, Roche, Janssen, AstraZeneca, SOBI, Takeda, BMS, Novartis, Swixx; Advisory board meetings organized by Janssen-Cilag, Sanofi, AstraZeneca, Roche, BeiGene; Travel grants from Sanofi-Aventis, Swixx, AbbVie, Astra Zeneca. J.A.S. speaker honoraria for educational events from Gilead, Janssen, Jazz; honoraria for advisory board from Vertex, Medac, Jazz, BMS; IDMC Membership for a Kiadis Clinical Trial. J.S. served on advisory boards organized by BMS, Janssen, MSD, and Sanofi and received lecture fees from Astellas, Novartis, Jazz, Eurocept, Medac, and Janssen. U.S.: consultation: Medac. Advisory board member: Astella, Takeda, AstraZeneca, Immdica. M.R. received research support from Abbvie, Novartis, Medac, Neovii. T.A.W.H. has received honoraria from Amgen, Bristol-Myers-Squibb, GlaxoSmithKline, Jazz Pharmaceuticals; Consulting or advisory role: Amgen, Jazz Pharmaceuticals, Kite/Gilead, Novartis, Sanofi, Bristol-Myers-Squibb, Pfizer, Pierre-Fabre, GlaxoSmithKline, Otsuka, Janssen, Abbvie; Travel/accommodation/expenses: Janssen, Jazz Pharmaceuticals, Abbvie, Bristol-Myers-Squibb, Amgen, Kite/Gilead, Astellas, Neovii, GlaxoSmithKline, Sanofi, Immatics, Kyverna Therapeutics, Sobi, Medac, BeiGene. J.Z.: consulting fees for advisory board from MSD, Novartis, BMS, Pfizer, Abbvie, Roche, Jazz, Gilead, Amgen, Astra-Zeneca, Takeda, Pierre Fabre; travel support from Roche, Sobi. D.P.M. speaker honoraria from GSK, Novartis, BMS, and Abbvie. Advisory boards Novartis and GSK. All the other authors declare no conflicts of interest. As for EBMT Policy, data cannot be shared but are available on reasonable request to the Chronic Malignancies Working Party. Data S1: All EBMT centers that participated to the study are cited in the Supporting Information. Figure S1: Flow chart for patient inclusion. Figure S2: Incidence of acute and chronic graft-versus-host disease (GvHD) comparing thymoglobulin versus grafalon. Figure S3: ajh70288-sup-0001-Supinfo.docx. Figure S4: Overall survival (OS), progression-free survival (PFS) and graft-versus-host-free relapse-free survival (GRFS) with thymoglobulin versus grafalon. Table S1: Patient characteristics. Table S2: Outcome of MF patients with unrelated donor transplants by donor type (univariable). Table S3: Multivariable Cox model for acute and chronic GvHD. Patient age at transplant is in decades. Effect estimates are given with 95% confidence intervals. Corresponding p values are calculated using the Wald test. Table S4: Multivariable Cox model for OS, PFS, NRM, relapse and GRFS. Patient age at transplant is in decades. Effect estimates are given with 95% confidence intervals. Corresponding p values are calculated using the Wald test. 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.