To the Editor: Acute graft-versus-host disease (aGVHD) remains a major challenge for allogeneic hematopoietic stem cell transplantation (allo-HSCT).1 Despite prophylaxis, 30–50% of patients develop aGVHD.2 While approximately half of the patients achieve remission with corticosteroids, those with steroid-refractory (SR)-aGVHD require second-line therapy. Although agents like ruxolitinib and basiliximab have improved overall response rates (ORR >70%),3 outcomes for patients refractory to these therapies remain poor, with 1-year overall survival (OS) below 20%. Anti-thymocyte globulin (ATG) was historically used as second-line therapy for SR-aGVHD, with response rates lower than 60%.4,5 However, its use declined because of high non-relapse mortality (NRM) resulting from infections and post-transplant lymphoproliferative disease (PTLD).6 The mechanism of ATG involves T-cell depletion, B-cell apoptosis, and induction of regulatory cells, which is different from cytokine-targeted agents, suggesting potential utility as salvage therapy.7 Since 2014, our institution had employed rabbit ATG (rATG) as salvage therapy for SR-aGVHD in patients who experienced treatment failure after second-line therapy following unrelated cord blood transplantation (UCBT). This report presents the clinical outcomes observed in these patients. We conducted a retrospective review of 1355 patients who underwent UCBT between February 2014 and June 2023. Patients who had hematological malignancies, underwent their first single-unit UCBT, developed grade II–IV SR-aGVHD that was refractory to second-line therapy, and received rATG salvage therapy were included in the analysis Supplementary Figure 1, https://links.lww.com/CM9/C776. The study was approved by the Ethics Committee of the First Affiliated Hospital of University of Science and Technology of China (2025-RE-003), and informed consent was obtained from patients. First-line aGVHD therapy was methylprednisolone without discontinuing original GVHD prophylactic agents. Second-line therapy, selected by the physician and encompassing basiliximab with or without ruxolitinib, mesenchymal stromal cells, or methotrexate, was initiated in the case of steroid resistance, which was defined as progression within 3 days or lack of improvement by 7 days after initiating methylprednisolone. rATG was administered if ≥2 doses of basiliximab failed. The initial rATG dose was 0.9–2.0 mg/kg, repeated every 7–14 days until improvement or ≥grade 3 infection occurred. The detailed information of inclusion criteria, GVHD treatment and assessment of response, UCBT procedure, and statistical analysis is presented in Supplementary Methods, https://links.lww.com/CM9/C776. Sixty-one patients were enrolled, and all of them received basiliximab-based second-line therapy. The median doses of total rATG dose was 1.9 (range, 0.9–16.0) mg/kg. The median time from aGVHD onset to rATG was 34.5 (range, 16.0–171.0) days, and from first basiliximab to rATG was 13.0 (range, 4.0–85.0) days. Median follow-up was 602 days post-UCBT and 540 days post-rATG Supplementary Table 1, https://links.lww.com/CM9/C776. At day 56 post-rATG, ORR was 50.8% (31/61), with 37.7% (23/61) complete response (CR) and 13.1% (8/61) partial response (PR). An additional five patients achieved CR/PR after day 56, resulting in a best ORR of 59.0% (36/61; CR 42.6% 26/61, PR 16.4% 10/61). Median time to response was 27.0 (4.0–102.0) days. Methylprednisolone dosing (mg/day) was decreased over time after ATG initiated Supplementary Figure 2, https://links.lww.com/CM9/C776. Among the 36 responders, aGVHD relapsed in 3 cases, and the relapses occurred at 12, 15, and 30 days after the initial response. Multivariate analysis identified higher total rATG dose (>3 mg/kg) as an independent predictor of improved ORR. Conversely, liver involvement and >3 doses of basiliximab pre-ATG were associated with lower ORR Supplementary Figure 3 and Supplementary Table 2, https://links.lww.com/CM9/C776. Among patients with gastro-intestinal aGVHD, the day-56 ORR was 52.5% (32/61), including 39.3% (24/61) CR and 13.1% (8/61) PR. The best observed ORR was 60.7% (37/61), with 44.3% (27/61) CR and 16.4% (10/61) PR. In contrast, for hepatic aGVHD, the day-56 ORR was only 16.7% (5/30), and the best ORR was 20.0% (6/30). Viral reactivations were prevalent, but the 1-year cumulative incidence of adenovirus (AdV) viremia (n = 9) and Epstein-Barr virus (EBV) reactivation (n = 7; 3 with PTLD) was significantly higher in the high-dose rATG group (>3 mg/kg) than in low-dose rATG group (≤ 3 mg/kg) (AdV: 45.5% (5/11) vs. 6.0% (3/50), P 3 mg/kg) compared to the low-dose group (54.5% (6/11) vs. 24.0% (12/50), P = 0.07), though not statistically significant Figure 1D.Figure 1: Cumulative incidence of adenovirus viremia reactivation (A), EBV reactivation (B), CMV reactivation (C), and invasive pulmonary aspergillosis (D) were compared between patients receiving low-dose (≤3 mg/kg) and high-dose (>3 mg/kg) ATG. The estimates of 2-year overall survival (E) and disease-free survival (F) were compared between patients receiving low-dose (≤3 mg/kg) and high-dose (>3 mg/kg) ATG. ATG: Anti-thymocyte globulin; CMV: Cytomegalovirus; EBV: Epstein-Barr virus.The 2-year post-ATG NRM was 61.0% (95% CI: 47.5–72.0%), with no difference between ATG dose groups. Multivariate analysis showed >3 basiliximab doses pre-ATG and age >10 years were independent risk factors for NRM Supplementary Table 3, https://links.lww.com/CM9/C776. The 2-year OS after initiation of ATG treatment was 35.5% (95% CI: 23.6–47.5%), with no difference between ATG dose groups Figure 1E. More than three doses of basiliximab, age over 10 years, and liver involvement were predictive of inferior OS Supplementary Table 3, https://links.lww.com/CM9/C776. The 2-year disease-free survival (DFS) rate after initiation of ATG treatment was 35.6% (95% CI: 23.7–47.7%). No significant difference was observed between patients who received high (>3 mg/kg) or low (≤ 3 mg/kg) doses of ATG (P = 0.25) Figure 1F. The 2-year cGVHD-relapse free survival after ATG initiation was 29.5% (95% CI: 18.7–41.1%) Supplementary Figure 4, https://links.lww.com/CM9/C776. The 2-year cumulative incidence of cGVHD was 11.5% (95% CI: 5.0–20.8%), also without significant difference between dose groups (18.2% 95% CI: 2.9–44.2% vs. 10.0% 95% CI: 3.7–20.1%, P = 0.45). The 2-year cumulative incidence of relapse for hematologic malignancies after rATG initiation was 3.4% (95% CI: 0.6–10.5%), with no significant difference between high- and low-dose rATG groups (0% vs. 4.1% 95% CI: 0.8–12.4%, P = 0.52). By the end of follow-up, 33 patients died of NRM and 2 patients died of relapse. Pulmonary infections (37.1%) and uncontrolled aGVHD (28.6%) were the leading causes of death, other causes included hemorrhage, thrombotic microangiopathy and cGVHD Supplementary Table 4, https://links.lww.com/CM9/C776. Our study demonstrates that ATG was effective in patients with SR-aGVHD who were refractory to second-line therapy, and nearly 60% of heavily pretreated patients achieved improvement. Although the effectiveness of ATG as a second-line therapy for SR-aGVHD has been restricted by the high risks of infection and complications,6 our strategy utilized it as salvage therapy. Treatment was initiated only after the failure of second-line agents, and gradual dose escalation for non-responders. Despite its function as salvage therapy, patient outcomes in our cohort were comparable to those reported when ATG was used as second-line therapy. Our study demonstrated that higher doses of ATG were associated with an improved ORR, but they also led to increased viral reactivation and invasive fungal disease, which had an adverse impact on survival. This finding is consistent with previous reports.6 CMV and adenovirus reactivations were frequent, letermovir might reduce CMV risk even under high-dose ATG. Notably, some cases experienced late CMV reactivation beyond 100 days post-UCBT, supporting the need for extended letermovir prophylaxis in these patients. Prior studies have reported that more than four doses of basiliximab increased infection rates without improving aGVHD ORR.8,9 Our results further demonstrated that high doses of basiliximab exposure prior to ATG administration were associated with higher NRM and poorer OS, underscoring the importance of transitioning to ATG promptly in basiliximab-refractory patients. Our study has limitations inherent to its single-center, retrospective design. The small cohort size limited subgroup analyses; furthermore, the generalizability of our findings is constrained as all patients underwent UCBT with an ATG-free prophylaxis regimen. In addition, the follow-up duration may be insufficient to evaluate long-term outcomes and late effects comprehensively. In conclusion, ATG can be effective when second-line agents are ineffective for SR-aGVHD patients. Higher doses of ATG may be beneficial for aGVHD but not for survival. Meanwhile, both aGVHD and OS may improve if ATG is administered earlier upon the failure of second-line therapy. Conflicts of interest None.
Sun et al. (Fri,) studied this question.