Children’s Oncology Group Study AALL-0434 was a phase III study looking at outcome in patients with T-ALL. Following the initial consenting, a standard four drug induction regimen was used which included the use of the corticosteroid prednisone. The response to that initial therapy was assessed, as we’ll discuss, using flow cytometric minimal residual disease detection and patients were assigned to their respective risk groups.
Following a second consent, post-injection therapy was given using a two by two randomisation protocol. The first randomisation was to Capizzi methotrexate plus peg asparaginase versus high dose methotrexate. The second randomisation was plus or minus nelarabine. Neither of the results of these randomisations are currently available as we’re waiting for the data to mature and are not part of the discussion for today. Patients also received cranial radiation if they fell into the intermediate or high risk groups which represented about 90% of the patients on this trial.
In the end about 1,144 patients with T-ALL representing children and young adults between the ages of 1 and 30 were enrolled on this trail, so by far the largest trial of T-ALL ever conducted. This occurred from the time period beginning in January of 2007 through about June or July of 2014.
The ETP sub-classification or early thymic precursor subtype was assessed at the time of enrolment onto the study because it had been shown that early thymic precursor T-ALL was a particularly poor outcome subset of patients with T-ALL. Following therapy, minimal residual disease was assessed at a single central laboratory using eight or nine colour flow cytometry. We assessed initial response at day 8 in the peripheral blood and response at end of induction at day 29 in the bone marrow.
This table describes the results of this trial. About 11.3% of the patients were identified as having this poor outcome subset of ETP, or early thymic precursor. 17.0% of the patients had a related immunophenotype that we called near-ETP, a phenotype that was similar to ETP but with an unusually high level of CD5 expression that disqualified it from standard ETP classification. The remaining 71.6% were in a category we called not-ETP.
There was a notable statistical difference in the level of residual disease that was detected at the end of induction, being a level of disease below 0.1% with only 18.6% of patients receiving this MRD negative status if they were of ETP type, 35.2% if they were of near-ETP type and the remainder having a much more frequent absence of MRD at 69.5%. Induction failure was also notably different between these groups, ranging between 7-8% for the ETP and near-ETP types while being noticeably lower in the not-ETP category at 1.1%. Despite these differences in both MRD and induction failure, the five year event free survivals and five year overall survivals were essentially the same between these groups being respectively about 85% and about 92%.
These outcome data are shown here on this slide. One can see that following after about the first one to two years there are essentially no events on this trial and there’s really no difference at all between any of the ETP subsets that were described. So the poor ETP status that had been previously described for this disease appears to have been eradicated by the therapy used on this trial. Also note that these events occur all during the first year or so after therapy.
Minimal residual disease analysis that was performed at the end of induction was able to stratify patients based on outcome. In particular, patients having MRD greater than 0.1% in the bone marrow had a worse outcome than those that had MRD at a lower frequency. In particular, there was a bad outcome subset of patients having MRD above 10% that were essentially closely related to induction failure and this is a particularly poor outcome subset of patients with this disease.
If one goes further out from therapy to the end of consolidation many of the patients that were MRD positive early on eventually became MRD negative at the end of consolidation and had a generally good outcome as well. However, a subset of patients remained MRD positive above a level of 0.1% and these also had a particularly poor outcome.
So, in conclusion, what we can say is that T-ALL when treated with the regimen that we used on AALL-0434 has an excellent outcome. There really is no difference in outcome that we were able to identify between ETP, near-ETP or the not-ETP subsets; they all had a similar outcome. This is despite the presence of increased levels of minimal residual disease at the end of induction as well as a higher incidence of induction failure for the ETP and near-ETP subsets.
This study also demonstrates the utility of post-therapeutic monitoring using minimal residual disease detection that was able to stratify patients in terms of risk at both end of induction at day 29 as well as the end of consolidation.