The ACCESS trial was a prospective multicentre phase II study which was sponsored by the National Marrow Donor Program. We evaluated PTCy-based GVHD prophylaxis in adults and children undergoing unrelated donor transplantation with a mismatch unrelated donor graft.

What was the study design?

The trial included three predefined strata. Stratum 1 were adults receiving peripheral blood after myeloablative conditioning; stratum 2 adults receiving peripheral blood following reduced intensity or non-ablative conditioning; and stratum 3 a paediatric ablative bone marrow cohort which will be analysed separately.

Our original adult target was 70 patients per stratum and those 145 patients were already published in JCO. But because the accrual to the reduced intensity conditioning was substantially higher than we anticipated, we amended the protocol to expand that cohort to 193 patients. That was also reported separately at ASH.

But the analysis we are discussing is focussed specifically on adult recipients of <7/8 mismatch unrelated donor grafts enrolled across stratum 1 and 2. We compared that with the cohort of patients who received a single antigen mismatch or 7/8. So we enrolled patients with haematologic malignancies requiring a transplant. These were patients with adequate performance status and [??] and they had to have an available mismatch unrelated donor between 4/8 and 7/8 at HLA-A, -B, -C and -DRB1. All donors have to be young, less than 36 years old, and peripheral blood was a required graft source. We excluded patients with an available and suitable matched donor, those with donor-specific antibodies with an [??] over 3,000, patients who received a prior allo, with a diagnosis of myelofibrosis or participating in another innovational GVHD trial.

All these patients received peripheral blood from a mismatched unrelated donor on day zero. We gave standard dose PTCy at 50mg/kg on days 3 and 4 in addition to tacrolimus and MMF. We followed patients through prespecified timepoints up to one year after transplant.

So in total we included 183 patients recipients of 7/8 and 85 patients received <7/8. Those who received <7/8 were modestly younger – the median age was 57 versus 63. Sex distribution was nearly identical. Cryopreservation was more common in the <7/8 – close to 77% versus 61% in the 7/8. As one would expect, the mismatched cohort reflected greater racial and ethnic diversity. So non-Hispanic white made up over half of the 7/8 group but only one third of the <7/8 group. Very importantly, socioeconomic indicators such as cost and SVI were similar across cohorts which suggests no major differences in baseline social vulnerability.

Both groups received a range of conditioning intensities although most patients in each cohort were treated with RIC regimens, FluMel being the most common. The diagnostic spectrum was very similar – AML was the leading transplant indication for both groups. Performance status was excellent overall, with over 80% of patients in each arm having a KPS over 80, and the same was true for comorbidity. These were well balanced – roughly two thirds of patients in both groups had an [??] of 0-2.

Donor characteristics were also very similar. These donors were uniformly young with a median age of 25 years. Sex distribution was balanced, C834 cells were also comparable. It’s important to state that within the <7/8 group the majority of these grafts were 6/8 with much smaller numbers of 5/8 and only a few 4/8 grafts and this distribution was consistent across ablative and reduced intensity conditioning.

What were the results of this study?

Survival at one year was 79% in the 7/8 group and 86% in the <7/8 group with overlapping confidence intervals and no real signal of worse survival with deeper HLA mismatching. This pattern was true for patients receiving ablation – survival at one year was 81% for 7/8 and 91% for <7/8 – and also true in the larger RIC cohort – one year survival was 78% with 7/8 donors and 84% with <7/8 donors. Again, no evidence that deeper mismatch worsened outcomes.

Looking at secondary endpoints the pattern was similar. GRFS, so GVHD-free relapse free survival, was nearly identical between 7/8 and <7/8 at 51% and 53% for <7/8. Rates of grade 3/4 acute GVHD and rates of moderate to severe chronic GVHD were also very similar. Primary graft failure was low in both cohorts and non-relapse mortality does not appear to be higher with deeper mismatch – 14% for 7/8, 8% for <7/8. Relapse was also comparable – 17% for 7/8, 23% for <7/8 with overlapping confidence intervals and no clear signal of increased relapse risk.

Primary graft failure was infrequent in both cohorts. We observed four events in both the 7/8 and <7/8 groups and all of these events occurred after reduced intensity or non-ablative conditioning with nearly all of these involving cryopreserved grafts. Many of these cases represented late neutrophil recovery rather than [??] and only a few patients ultimately required subsequent transplantation.

What do you think is the importance of these results?

We also felt it was very important to place these clinical results in the context of donor availability today. So we did a registry analysis looking at over 2,400 patients with no potential 8/8 match. Essentially, you’re looking at the very unlikely search prognostic group from 1702. We restricted this search to donors who were younger than 36 and required a 75% likelihood of matching.

What the analysis showed was that every patient had multiple donor options at less than 7/8. Even among ethnically diverse patients where the probability of a full match is lower, the global registry consistently provided several young MMUD options. Very importantly, the number of potential donors increased dramatically as we moved from 7/8 to 6/8, 5/8 and 4/8, highlighting that being permissive to these higher degrees of mismatch could have major implications for equity and access.

It's also important to understand that a larger donor pool allows optimisation of important non-HLA donor factors such as age, which we know can influence post-transplant outcomes. So the study clinically showed that outcomes with <7/8 graft using PTCy comparable and operationally this registry data showed that these transplants are possible today.

Is there anything else you would like to add?

Among these recipients of <7/8 MMUD graft treated with PTCy, survival exceeded 80% and it was comparable to that of 7/8 recipients. Secondary endpoints were also comparable. We have two ongoing trials at NMDP, OPTIMIZE and ACCELERATE, that are building on this platform to improve other outcomes such as GVHD infection and relapse. Taken together, the data support broadening acceptable MMUD match levels within the PTCy platform.

The idea here is that this may enable near universal donor access across race and ethnicities while allowing optimisation of key non-HLA factors that can influence transplant outcomes.