Antihuman T Lymphocyte Globulin Fresenius in Graft‐Versus‐Host Disease Prophylaxis for Unrelated Hematopoietic Stem Cell Transplantation After Myeloablative Conditioning: A Long‐Term Real‐Life Retrospective Study

This is the authors' abstract. We don't add key points for this paper.

Graft-versus-host disease (GvHD) is a frequent complication of hematopoietic stem cell transplantation (HSCT) and remains among the leading causes of post-transplant morbidity and mortality. Acute GvHD (aGvHD) affects 30%–50% of HSCT patients, while chronic GvHD (cGvHD) affects 30%–70%. In vivo T depletion using rabbit antithymocyte globulins (ATG) during conditioning has been shown to reduce the occurrence of both aGvHD and cGvHD, with no impact on overall survival (OS) or relapse [1]. Among available antihuman lymphocytes serums, ATG (Thymoglobulin; Sanofi-Genzyme, Saint-Germain-en-Laye, France) and ATG Fresenius (ATLG) (Grafalon; Neovii, Rapperswil-Jona, Switzerland) can be used as GvHD prophylaxis. In myeloablative conditioning (MAC), ATG infusion is correlated with a significant reduction in aGvHD and cGvHD, with no impact on OS, relapse, disease-free survival (DFS), or non-relapse mortality (NRM). However, initial trials using a high dose of ATG reported an increased rate of lethal viral infections [2], such as cytomegalovirus (CMV) or Epstein–Barr virus (EBV). Since little real-life data has been reported so far, we aimed to study the impact of ATLG on a long-term real-life perspective in unrelated transplantation after MAC. Patients were included if they received an HSCT at Saint-Louis Hospital from an unrelated donor after MAC containing ATLG between December 2012 and October 2018. Data collection was based on the European Group for Blood and Marrow Transplantation registry via the central data processing system “Project Manager Internet Server” (ProMISe). HLA genotypes were determined at four-digit alleles (A, B, C, DRB1, and DQB1). Donor types included matched (10/10) and mismatched (9/10) unrelated donors. Stem cell sources were peripheral blood and bone marrow. MAC included busulfan with cyclophosphamide or fludarabine, TBI with cyclophosphamide and TBI with fludarabine. Standard protocols of immunosuppression, including cyclosporine and methotrexate or cyclosporine and mycophenolate mofetil, were used for GvHD prophylaxis. All recipients received ATLG (20 mg/kg/day from Day (D)3 to D1) and antibiotic, antiviral, antiparasitic prophylaxes. No patient received letermovir prophylaxis. Evaluated endpoints included aGvHD and cGvHD (according to the National Institute of Health criteria), NRM, relapse, OS, immunosuppressive therapy (IST) withdrawal, GvHD-free relapse-free survival (GRFS), and DFS. GRFS was defined as survival without Grades III and IV aGvHD, moderate/severe cGvHD, relapse or death. IST withdrawal was defined as the cumulative incidence (CI) of IST withdrawal for alive patients. DFS was defined as survival without relapse or progression. Infection was defined as bacterial, fungal or viral infection requiring systemic treatment. Bacterial infections were reported only if they occurred after transplantation hospitalization. Ninety-one patients were included between December 2012 and October 2018 (Table S1). GvHD prophylaxis was cyclosporine and methotrexate in 98%. Most patients had acute lymphoblastic leukemia (ALL) (62%) and the median age was 30 years. The source of stem cells was mainly peripheral blood (95%). The median follow-up was of 56 months. At D100, the CI of Grades II–IV aGvHD was 46% and Grades III and IV aGvHD 15% (Figure 1a,b), with five patients (5.5%) resistant to corticosteroids. In univariate analysis, factors associated with Grade II–IV aGvHD were recipient age older than 40, AML diagnosis and TBI (Table S2). In multivariate analysis, TBI was associated with an increased risk of Grades II–IV aGvHD. At 2 years, the CI of cGvHD was 36% with 19% moderate/severe (Figure 1c,d). In univariate analysis, the most significant factor associated with an increased risk of moderate/severe cGvHD was the number of CD34+ infused cells (Table S2). The CI of infections was 26% at D30 and 82% at D100 (Figure 1f). During follow-up, 23 patients (25%) presented fungal infection, including 14 aspergillosis, 6 candidosis, two mucormycosis, and one fusariosis. Sixty-five percent of these infections occurred before D30. Seventy-nine patients (87%) presented viral infections requiring treatment. Thirty-six patients (40%) presented CMV viremia requiring preemptive therapy and 67% reactivated EBV requiring rituximab infusion. Six patients (7%) presented PTLD. Six patients (7%) presented bacterial infection after HSCT hospitalization. Infection was the second cause of death. In multivariate analysis, no risk factor for infections was identified. Immune reconstitution was evaluated at D100, 1 and 3 years for NK, CD3+ T, CD3+ CD4+ T, CD3+ CD8+ T, and B lymphocytes (Figure S1). Sixteen patients progressed or relapsed during follow-up, resulting in a five-year CI of relapse of 20%. Relapse-related mortality was 11% (10/91). In multivariate analysis, recipients older than 40 years had an increased risk of relapse (Table S2). At 5 years, OS was 72% (Figure 1e). DFS, GRFS, and CI of IST withdrawal were 65%, 29%, and 62%, respectively (Figure S2a,b,d). During follow-up, 27 patients died. Main causes of death were relapse (37%), infection (30%, 7 patients) and GvHD (19%, 5 patients). At 5 years, NRM was 15% (Figure S2c). The deaths by infection without concomitant cGvHD included four viral, two fungal (one mucormycosis and one aspergillosis), and one bacterial infections. We conducted a long-term single-center retrospective study to evaluate ATLG in HSCT from an unrelated donor after MAC. aGvHD CI at D100 was 46% for Grades II–IV and 15% for Grades III and IV. At 2 years, cGvHD CI was 35% for all grades and 17% for moderate/severe. CI of Infection was high, including fungal and viral infections in 21% and 87%, respectively. Infection was the second cause of death after relapse. The 5-year OS was 72%. In vivo T-cell depletion for HSCT remains controversial. ATG was associated with a lower risk of extensive cGvHD, but a higher risk of CMV reactivation. In our population, the D100-CI of aGvHD was similar than in the no-ATLG or placebo arms of prospective trials evaluating ATLG in matched unrelated transplantation with MAC [3, 4]. Because of the young median age and the significant proportion of ALL, our population included 67% of TBI, increasing aGvHD, which could at least partly explain the higher-than-expected rate of cGvHD. The CI of cGvHD at 2 years in our population was higher than in the ATLG versus placebo trial [3]. However, our population was particularly at risk, with only 63.7% of patients in first complete remission after first-line therapy, 95% peripheral HSCT graft, and 21% HLA-mismatched donor. Moreover, HLA matching was determined at 10 alleles in our population versus eight alleles in the no-ATLG study. Despite the absence of control group, our high incidence of infections confirms that ATLG is associated with an increased risk of infections, especially viral and fungal. No major delay in immune reconstitution was observed. Most fungal infections occurred here before D30, suggesting considering ATLG as a high-risk factor and raising the question of systematic azole antifungal agent, especially for early post-transplant follow-up. This study was also conducted before routine use of letermovir as CMV infection prophylaxis and it would be interesting to perform further analyses with letermovir in the ATLG context. In the placebo versus ATLG study [3], OS was better in the placebo arm, but it was confirmed only for patients receiving TBI with a severe lymphopenia at D3. Kennedy et al. [5] reported a correlation between absolute lymphocyte count (ALC) at first day of ATG administration and ATG clearance. Low ALC leads to lower clearance and higher infections and mortality. In haploidentical HSTC, the dose monitoring, and ATG adjustment was associated with a reduction of EBV/CMV reactivation and improved survival without increasing GvHD [6]. The CI of relapse at 5 years was low in our population (15%). OS was encouraging (72% at 5 years), seemingly higher than in other studies using ATLG (3). It could be explained by the youngest median age in our population. In the study versus placebo [3], GRFS at 2 years was 44% in the ATLG arm, versus 31% at 3 years in our population. The higher cGvHD rate in our study could explain this difference, even if the OS was better. Our results on aGvHD, relapse, and OS are consistent with the previous randomized trials with ATG or ATLG [1], which support the use of ATG/ATLG in MAC, but highlight an infectious issue. The retrospective setting of our study and the absence of a control group limit our interpretation. The young age and the significant proportion of ALL induced a significant bias. However, the reasonable number of patients (91) and the 5 years long-term follow-up strengthen our results. Other more recent strategies included post-transplant cyclophosphamide (PTCy) in non-haploidentical setting [2] with no more infections for reduced intensity conditioning especially. To our knowledge, no study compared yet PTCy and ATG/ATLG in MAC. In conclusion, in the present long-term real-life study, OS appears encouraging with the use of ATLG in unrelated transplant after MAC. Nevertheless, the infection rate is very high, especially for viral and fungal infections. Active and prolonged infectious prophylaxis, both for fungal and viral agents, is of particular importance in the setting of unrelated HSCT after MAC regimen. Optimizing the dose of ATLG might be explored in future prospective clinical trials to properly balance the risk between GvHD and infectious complications and should then be considered in new strategies comparisons. The review of medical records was approved by the Institutional Ethical Committee in agreement with the Helsinki Declaration of 1975, revised in 2008. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. Data S1 Supporting Information. 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.