Low-dose post-transplant cyclophosphamide and anti-thymocyte globulin as an effective strategy for GVHD prevention in haploidentical patients

During the past 20 years, wider application of easily available haploidentical donor hematopoietic cell transplant (haplo-HCT) has been made possible through the T cell-replete (TCR) regimens including T cell regulation with anti-thymocyte globulin (ATG)/granulocyte colony-stimulating factor (GCSF) and post-transplant cyclophosphamide (PTCy) [ 1– 9]. To achieve decreased non-relapse mortality (NRM) and improved long-term outcomes in haploidentical transplant, the joint use of ATG and PTCy might effectively reduce graft-versus-host disease (GVHD) and mortality associated with severe forms of GVHD [ 10– 14]. There are multiple papers using ATG/PTCy (high-dose) in patients with Fanconi anemia, aplastic anemia, and sickle cell disease [ 10]. To date, the combination of conditioning regimens with low-dose ATG and high-dose PTCy after haplo-HCT [ 11– 14] or unrelated donor (URD) HCT [ 15, 16] for hematological malignancies has been documented in several reports with reduced rates of GVHD and acceptable relapse rate albeit with somewhat high rates of graft failure (8%) [ 11] or delayed engraftment [ 12]. Recently, results of a prospective trial have been available which analyze the efficacy of combined use of low-dose ATG and 1-day high-dose PTCy (50 mg/kg at + 3 days) in preventing GVHD among haplo patients [ 17].

Recently, we established a regimen using low-dose PTCy in conjunction with standard-dose ATG in order to lower the risk of GVHD without compromising engraftment and disease relapse. This protocol consisted of a myeloablative conditioning (MAC) regimen containing cytarabine, busulfan, Cy combined with standard-dose ATG/G-CSF (the so-called Beijing protocol), and followed by low-dose PTCy (14.5 mg/kg on days 3 and 4 after HCT) for haplo-HCT recipients. We previously reported this strategy in preliminary prospective study wherein reduced incidences of both acute and chronic GVHD were observed in comparison with non-combined regimens [ 18]. In this prospective trial, results indicated that low-dose PTCy is sufficient to lower acute GVHD in mouse model, partly due to the boosting of fast regulatory T cell (Treg) reconstitution. In addition, low-dose PTCy could augment the protective effect of ATG on GVHD both in mouse and human sets while ensuring high rates of engraftment and keeping disease control [ 18]. These results seem to highlight the feasibility of this novel procedure with low-dose PTCy in ATG-based MAC for GVHD prevention and perfect outcomes post-haplo-HCT. However, the potentially effective regimen needs to be validated on a larger population with hematological neoplasms.

With the aim to assess the efficacy of this regimen in haplo-HCT, we extended our prospective study in consecutive patients treated with ATG/GCSF-PTCy (low-dose) regimen and initiated a multicenter analysis on comparing the results to the contemporary cohort of patients who received ATG/GCSF without low-dose PTCy. In the current study, patients in both study group and contemporary control group are all transplanted from maternal donor (mother donors, MDs) [ 3] or collateral relative donors (CRDs, e.g., uncle, aunt, nephew, niece, and cousins) [ 19, 20] who had especially high risk of GVHD occurrence under “Beijing protocol.” Herein, we tested the efficacy of ATG/PTCy (low-dose) as compared with standard ATG-based prophylaxis using maternal donor or collateral relatives. Thus, this new approach in unmanipulated haplo-HCT with low-dose PTCy in ATG/GCSF-based MAC would expand the donor pool for patients eligible for allogeneic HCT, by allowing the safe and effective use of specific haploidentical donors associated with especially high risk of GVHD occurrence.

We compared our novel GVHD prevention strategy (low-dose PTCy in ATG/GCSF-based regimen) with contemporary control patients all transplanted from maternal donors or collateral relatives at particular high GVHD risk [ 3, 19, 20]. Cumulative incidence of acute GVHD, grade III–IV aGVHD, and chronic GVHD was significantly lower without GVHD-related death in the ATG-PTCy cohort as compared with the control cohort. Reliable engraftment and disease control have also been achieved. As a result, significantly decreased NRM and improved GRFS have been accomplished. Our observations reveal that low-dose PTCy combined with ATG/GCSF-represented conditioning could be a promising new regimen for GVHD prophylaxis and improve outcomes after haplo-HCT.

The current results suggest that the administration of low-dose PTCy along with ATG/GCSF significantly decreased the rate of both acute and chronic GVHD as compared to contemporary controls who were also transplanted from MDs or CRDs proved to be at particular high GVHD risk under “Beijing protocol” in both single-center and multi-center studies [ 3, 5, 6, 19, 20]. Furthermore, grade III–IV aGvHD or severe cGVHD GVHD was the main cause of death in 7 out of 27 patients in the control cohort, in contrast to none in the combined treatment cohort (Table  3). It should be stated that despite the significantly decreased rate of grade III–IV aGvHD and total chronic GVHD in the ATG-PTCy cohort compared to control cohort, the incidence for moderate to severe chronic GVHD remains higher. And the rate of moderate to severe chronic GVHD was also higher or comparable than that in the high-dose PTCy plus ATG protocols reported to be 0–18.8% with smaller study population or shorter follow-up [ 12– 14, 17]. MAC and predominantly female donors especially female-to-male pairs [ 3] in the current study may be the attributable reasons as compared with previous reports on high-dose PTCy plus low-dose ATG regimens [ 12– 14]. More recently, Wachsmuth et al. [ 29] showed that the dose of PTCy is important, with tested doses between 10 and 50 mg/kg/day at days + 3 and + 4 effectively prevented fatal GVHD and 25 mg/kg/day being the optimal dose in their model while doses outside this range (≤ 5 or ≥ 100 mg/kg/day) proved ineffective in preventing mortality. Overall, our current results verify the feasibility of low-dose PTCy adding to the backbone of ATG/GCSF-based regimen, and this combination for GVHD prevention may represent a promising strategy.

Understanding the mechanism of low-dose PTCy in combination with ATG is vital to explore its maneuver in innovative procedure. In our previous reported prospective trial, data suggested that low-dose PTCy is sufficient to reduce acute GVHD in mouse model, partly due to the enhancement of quick Treg reconstitution. Besides, low-dose PTCy could promote the protective effect of ATG/G-CSF on GVHD both in mouse and human sets [ 18]. Additionally, the timing and sequential order of given drugs may be crucial to the synergistic effect. It is conceivable that in spite of slow donor T cell depletion by ATG, donor T cells escape blockade leading to aGVHD occurrence. These escaped donor T cells exponentially expand to give rise to aGVHD when exposed to recipient antigens. Wachsmuth et al. [ 29] proved that PTCy induced alloreactive T cell dysfunction and suppression and supported the increasingly important role of Tregs over time post-HCT in the amelioration of chronic GVHD. Nonetheless, centers using PBSC in haploidentical transplants with PTCy have shown rates of extensive chronic GVHD as high as 38% [ 30]. Even with the addition of 4.5–5 mg/kg ATG to a backbone of high-dose PTCy, the rate of moderate to severe chronic GVHD was reported to be 10–13% with 2-day PTCy [ 13, 14] or 18.8% with 1-day PTCy [ 17]. This may partly explain why only total cGVHD was reduced with the current combined treatment while moderate to severe cGVHD was not. Future randomized studies including data on immune reconstitution and Treg levels will aid to further elucidate the biological aspects.

In the previous report on the combination of ATG and high-dose PTCy, the primary rejection rate was 8% [ 11]. In the current study, although the myeloid engraftment was 100%, hematopoietic recovery using low-dose PTCy along with ATG was slower than that in the control group. It may partly due to the immediate allo-response in which alloreactive T cells were indicated to dominate early after PTCy. It was documented that the inclusion of ATG in conditioning regimens yields quicker achievement of donor chimerism [ 31], and ATG was suggested to be capable to aid reliable engraftment by decreasing the aforementioned immediate allo-response early after PTCy [ 29]. As for disease control, it is speculated that low-dose PTCy had little impact on relapse, which is evidenced by our current data. Previous studies also demonstrated that high-dose PTCy with ATG is effective in alleviating GVHD without increasing relapse [ 13, 14] with the relapse incidences reported to be 16–37% [ 11– 14, 17].

It was taken in the context of the fact that the combination of ATG and PTCy appears to increase infectious complications particularly regarding herpes virus control and may lose some of the protective effects of PTCy alone. Higher rate of CMV reactivation was noted in the combined treatment cohort as compared to the control cohort although the rate was similar to the previous report on the joint use of high-dose PTCy and ATG (74%). As compared with the control cohort, T cell impairment between donor and recipient secondary to higher extent of HLA mismatch and the dual T cell depletion with PTCy and ATG may contribute to the higher incidence of CMV reactivation. Nonetheless, only one CMV-related death occurred (Table  3). We have implemented a preemptive CMV-CTL strategy with the aim to alleviate the CMV infection risk. As for the observation that low-dose PTCy increased the risk of CMV reactivation but not the risk of relapse, although it looks in contrast to the hypothesis that PTCy acts partly by eliminating activated alloreactive T cells, it may be partly explained by recent report that CMV reactivation and expansion of CD56 ^brightCD16 ^dim/−DNAM1+ natural killer cells are associated with antileukemia effect after haploidentical HCT in acute leukemia [ 32, 33]. In addition, Wachsmuth et al. [ 29] revealed that 25 mg/kg PTCy did not significantly reduce CD8+ T cell proliferation while the T cell reconstitution after CMV reactivation under the combined treatment strategy needs to be elucidated. One of the characteristics of the PTCy studies has been a reduced incidence or even absence of PTLD [ 34]. Though we found EBV reactivation incidence was 21%, it was remarkably less in comparison with other studies using high-dose PTCy and ATG (32–64%) [ 11– 14, 17]. In order to reduce the infectious complications, the next step with testing ATG/PTCy (low-dose) strategy may be decreasing the ATG dose also.

The limitation of the study is the non-randomized feature. We did matched-pair analysis to balance characteristics of the two populations, and the matched-pair analysis confirmed the main outcome results that we found in the standard analysis (data not shown due to the small population). Another limitation is that the comparison here is across institutions. Although the identical ATG/GCSF-based “Beijing protocol” is adopted in all the three centers (the largest and the most experienced transplant centers in China as described in the “Patients and methods” section), one cannot completely rule out that there exist small differences of medical practices between participating centers which may partly contribute to the transplant outcomes. Nevertheless, since we extended the prospective trial with the combined treatment for all HCT from MDs or CRDs in Peking University, the contemporaneous external control also transplanted from MDs or CRDs can only come from other centers. Multi-center randomized trials in different transplant settings and dose-finding studies including decreasing the ATG dose are needed to extend the practical use of the new combined treatment strategy.

In conclusion, our results showed that T cell-replete haplo-HCT after ATG/GCSF-based intensified conditioning combined with low-dose PTCy is a feasible and effective protocol, especially for patients at particular high GVHD risk. This strategy produced reliable donor cell engraftment with low rates of GVHD while keeping disease control.