In this study we examined the association of GVL and GVHD in two separate cohorts based on the regimen used for GVHD prophylaxis in adult patients with de-novo AML given a first allogeneic SCT from HLA-matched donors in CR1/2. In the first large group of more than 12,000 patients given standard calcineurin-based GVHD prophylaxis we observed a strong association between these immune effects. Patients who had acute GVHD or extensive chronic GVHD had a reduced risk of relapse (HR 0.85 and 0.76, respectively). However, since this was also associated with a higher risk of NRM, the survival of these patients was reduced. This historical association reported in the early era of SCT [1–3] remains in the more recent era with modern SCT techniques. It was even more significant than previously reported in patients with AML [11, 12]. Similar observations were observed more recently in large registry cohorts of patients with secondary AML [22] and patients with acute lymphoblastic leukemia [23] also transplanted in the modern era.
PTCy is increasingly being used for GVHD prophylaxis. It was shown to nullify the effect of HLA mismatch between recipients and donors allowing the safe expansion of the donor pool to haploidentical and HLA-mismatched unrelated donor transplants 5–7, 24,25]. PTCy is also used to reduce the risk of GVHD in other high- risk groups such as older patients, patients intended for earlier tapering of immune suppression to allow maintenance therapies, patients with a higher GVHD risk due to prior treatment with check point inhibitors, and in non-malignant disorders [25]. The Blood and Marrow Transplantation Clinical Trial Network (BMT-CTN) conducted a phase II trial (BMT CTN 1203) comparing three novel GVHD prophylaxis regimens with the more standard tacrolimus/ methotrexate combination in patients given SCT with RIC [9]. Among these three regimens the combination of PTCy–tacrolimus– mycophenolate mofetil was the best promising in terms of GVHD-free, relapse-free survival. More recently, the phase III randomized BMT CTN 1703 trial reported that following RIC the PTCy–tacrolimus– mycophenolate mofetil regimen was associated with better GVHD-free, relapse-free survival than tacrolimus/ methotrexate, due to reduction of severe acute GVHD and of chronic GVHD that required treatment with no difference in relapse and survival [10]. This may soon lead to this regimen becoming standard of care after RIC. The BMT CTN 1301 trial compared single agent PTCy to other regimens following MAC [26]. Severe chronic GVHD was reduced but this did not translate to better survival. There was a suggestion of lower relapse in the PTCy arm. It seems PTCy is the first regimen to show that more intensive GVHD prophylaxis is not associated with reduced GVL.
In the second group of patients investigated in the current study given PTCy- based GVHD prophylaxis regimens, we observed a different pattern of GVL / GVHD association than that following the calcineurin- based regimens. We showed that acute GVHD grade II-IV did not reduce relapse risk but there was in fact a borderline statistically significant increased risk of relapse, possibly due to additional immune suppressive therapy given to these patients. NRM was increased and survival was reduced in patients with acute GVHD. Chronic GVHD did not reduce relapse, but was relatively well tolerated in the PTCy setting with no increase in NRM or reduced survival. The sample size did not allow defining the effect of different GVHD grades. We have previously reported similar observations after haploidentical SCT with PTCy [15, 16]. The similar report in the HLA-matched setting suggests that it is the use of PTCy rather than the haploidentical donor source that is responsible for the separation of GVHD and GVL. Similar results following haploidentical SCT were reported by Japanese and Italian registries [27, 28]. The Baltimore group reported different outcomes in a group of 340 patients with various hematological malignancies following haploidentical SCT with non-myeloablative conditioning and PTCy [29]. They showed that acute GVHD grade II was associated with reduced relapse rates, similar NRM and improved OS compared to no GVHD. Acute GVHD grade III-IV did not reduce relapse, probably due the high immune suppression burden. It was associated with a markedly increased NRM and reduced OS. Chronic GVHD showed a trend towards reduction of relapse but no effect on survival. Similarly, a French study showed improved outcome after haploidentical SCT with acute GVHD grade II [30]. The Baltimore group also reported similar observations with the use of PTCy after HLA-matched transplants [31]. These discrepancies may be explained by the use of different transplantation platforms, different conditioning regimens, stem cell source (PBSC versus BM), use of ATG, as well as different patient characteristics. Clearly, more studies with biological correlates are required.
There are initial data showing that GVHD after haploidentical SCT with PTCy shows some different clinical characteristics than GVHD after unrelated donor transplant. It involves the gastro-intestinal tract less often and associates with better response and lower NRM [32–34]. However, there are no similar comparisons of GVHD with PTCy and no PTCy in the HLA- matched setting.
Both GVHD and GVL are mediated by donor T-cells and natural killer (NK) cells and therefore occur in parallel. They require the presence of mismatched alloantigens between the donor and recipient and activation of the immune response. However, there are differences between the two effects that can allow targeting of GVL without GVHD. There are different immune signatures and different T-cell subset reconstitutions related to the mechanism of action of PTCy that may allow the separation between the effects. PTCy impairs the proliferation and cytokine production of alloreactive T-cells but does not completely eradicate them [5, 35]. Thus, it may reduce the progression to a severe form of GVHD and persistence of the GVL effect. PTCy promotes the recovery of regulatory T-cells (T- regs). T-regs, similarly to hematopoietic stem cells are protected from PTCy by a relatively high expression of aldehyde dehydrogenase [36]. They contribute to long term post-transplant tolerance and prevention of progression to chronic GVHD by limiting T-cell proliferation and down regulation of pro-inflammatory cytokines, while maintaining CD8 + T-cell anti-leukemia activity. These mechanisms are distinct from the mechanism of standard GVHD prophylaxis. The threshold level of T-cells necessary to trigger GVL is lower than that required to trigger GVHD and a lower level of GVHD may be sufficient to reduce the risk of relapse [37]. McCurdy et al. used machine learning techniques to define clinically relevant signatures from multiple immunophenotypic, proteomic and clinical factors in patients given PTCy after both haploidentical and HLA-matched transplants [38]. Conventional CD4 + T-cell recovery, their activation status and metabolic signature were associated with acute GVHD and in particular the CXCL9-CXCR3 axis. CD8 + T-cell hypo-responsiveness was less important for protection from acute GVHD. However, NK and CD8 + cells were important in preventing relapse and a loss of inflammatory gene signature in NK cells and transcriptional exhaustion phenotype in CD8 + T-cells predicted relapse. Interestingly, this was similar for PTCy after matched or haploidentical transplants. Similarly, Zhao et al. also showed a distinct T-cell immune signature after PTCy [39]. In particular there was remarkable enhancement of multiple inhibitory receptors both on CD4 + and CD8 + cells with reduced response to stimulation, while loss of granzyme and perforin expression on CD8 + cells was associated with relapse.
The current analysis has several limitations. The study was not designed to compare standard GVHD prophylaxis with PTCy but rather to analyze the GVHD/ GVL association separately. Although overall outcome seems similar, the two groups may not have had similar patient and transplant characteristics and randomized studies would be more appropriate for this comparison [9]. The PTCy group was relatively small, not allowing definition of the role of different grades of acute and chronic GVHD in this group. The database did not allow characterization of the different clinical pictures, organ involvement, and response to therapy in the two groups.
In conclusion, GVHD of any type or grade is not associated with an improved relapse rate after HLA-matched SCT with PTCy and offers no survival advantage. Severe forms are associated with higher NRM and lower survival. Future novel strategies for further reducing significant GVHD while preserving GVL are warranted.