High-dose methotrexate was not associated with a reduction in the risk for CNS relapse compared to standard interim maintenance therapy in children and young adults with acute lymphoblastic leukaemia who were enrolled in the phase 3 UKALL 2011 trial (ISRCTN64515327). Medicom interviewed Amy Kirkwood (Senior Statistician Cancer Research UK & UCL Cancer Trials Centre, University College London, UK) to talk about this finding and other results of the UKALL 2011 trial that were presented at the 64th annual meeting of the American Society of Hematology (ASH 2022), held in New Orleans, USA, from 10-13 December 2022.
The UKALL 2011 trial included children and young adults up to 25 years of age with ALL to investigate the difference of a short (14 days) versus a standard (28 days) induction course of dexamethasone, the CNS relapse risk for high-dose methotrexate compared with a standard interim maintenance regimen, and the difference between the effect of vincristine/dexamethasone pulses or no pulses on bone marrow relapse rate. At ASH 2017, the results for the dexamethasone induction courses were presented . At ASH 2022, Ms. Kirkwood presented the maintenance results.
In total, 1,570 patients receiving a backbone therapy according to the risk group they were stratified to , were randomised to 1 of 4 maintenance arms: (1) high dose methotrexate, with pulses, (2) high dose methotrexate without pulses, (3) standard interim maintenance with pulses, or (4) standard interim maintenance without pulses.
After a median follow-up of 72 months, there was no difference between high dose methotrexate or standard interim maintenance with regard to CNS relapse (HR 0.99; 95% CI 0.65-1.51; P=0.97), with 5-year rates of 5.6% for both treatment regimens. Interestingly, patients who received a short course of induction dexamethasone, followed by high-dose methotrexate and no pulses had an inferior event-free survival compared to patients who received short-duration induction dexamethasone followed by either high-dose dexamethasone with pulses or standard interim maintenance with or without pulses (EFS rate 75.9% vs 83.2%-86.0%; P-value for interaction=0.006). Furthermore, ‘no pulses’ was non-inferior to ‘pulses’ for bone marrow relapse rates, with a hazard ratio of 1.22 (95% CI 0.89-1.67).
Medicom asked Ms. Kirkwood to elaborate on these findings.
Please place the rationale of this study in context for us?
Ms. Kirkwood: Acute lymphoblastic leukaemia treatment in paediatrics has obviously improved a lot over the recent decades, but we also know that there are large numbers of patients that are being overtreated. We could heal about 50% of patients with half the current treatment load, 50 years ago. Many protocols now are looking for ways to reduce treatment for selected patients who we think can have less intense treatment to reduce toxicity and long-term side effects. The opposite also holds true, we would also like to increase treatment for the patients who we think need that. Most protocols now are stratified, focusing on de-escalation questions.
Our primary endpoint of this particular study was a reduction in toxicity during the induction phase, to better understand how we can use steroids for shorter amount time during induction in order to reduce toxicity. We randomised about 1900 patients before we did the futility analysis, where we found that there was not any difference in terms of toxicities at that primary endpoint, and there were some worrying trends for worse outcomes in terms of timed event endpoint.
We then continued recruiting patients through secondary organisations. At that point the study looked at 2 different things. Firstly, we wanted to know whether changing the interim maintenance treatment to either give standard care or a high-dose methotrexate regimen, which we hoped would reduce the risk of CNS relapse. The second randomisation was in the maintenance phase. Typically, the patients receiving standard care get these pulse doses every month, which have little evidence to support this approach, although anecdotally, we hypothesized that they merely impair children’s quality of life and cause toxicity. To obtain actual evidence about whether these pulses were needed, we wanted to look at removing these pulses and seeing whether that essentially could be done safely, and whether it would not have too much of an impact on bone marrow relapse.
What were your main findings?
Dr. Kirkwood: It’s a complex study design, but it’s also complex results. Overall we found many differences in terms of the primary endpoint, and, as I mentioned earlier, we saw these worrying trends for inferior results in terms timed event endpoint, but actually that didn’t really tell the whole story.
The study had different parts to it, and the reason we have not actually published R1 part, beyond reporting on it as an ASH abstract from 2017 , was because we were waiting for the R2 results just reported at ASH 2022 . We knew there was this interaction with the high dose methotrexate, so we could not really publish that without the R2 data.
Overall, we saw no difference in any of the timed event outcomes for high dose methotrexate treated patients, absolutely nothing in terms of CNS relapse, which is the primary endpoint. But we could see that if patients were given the standard dose upfront, it looked like actually the high dose methotrexate may have benefits in terms of some of the other endpoints. For example, we had an improvement in bone marrow relapse and we had trends toward better event-free and overall survival metrics. Nevertheless, there was still no difference in terms of what was originally our primary endpoint, the CNS relapse. However, the patients who ended in the treatment arm receiving a short course of induction dexamethasone upfront, we saw the opposite effect.
This is an unusual effect that we were not expecting. The results were very complicated, and are a challenge to interpret, but do raise questions.
Essentially our conclusions are that high dose methotrexate may be beneficial for some patients, although it is not beneficial in terms of CNS relapse, it can be beneficial in terms of some of the other outcomes when it follows standard dexamethasone. Our second conclusion, is that it may also be okay with that group to remove the pulses. We still need some longer follow-up in this study, and fully analyse quality-of-life data to properly interpret it.
How would you have changed this trial if you knew what you knew now?
Ms. Kirkwood: Oh, that’s a good question. Honestly, I do not know for sure what I would have changed because no one expected this result. Actually no one has a good idea at the moment as to why we have seen this interaction effect. The study delivered actually quite interesting results, the challenge is making sure we draw the correct conclusions. We still do need to look at this in more detail, and compare our findings carefully to other studies.
Ms. Kirkwood: In terms of this study, we will obviously still fully analyse all the toxicity data, as well as the quality-of-life data. We also have some pharmokinetic data from induction, so there is a possibility that that could shed light on the difference we observed in the steroid results.
Going forward, we are part of the ALLTogether Consortium (NCT03911128), which will be the next trial. We will be looking at various ways to both increase and decrease treatment depending on patients risk grouping. ALLToghether1 will revisit the pulses randomisation question, but it will be looking within a slightly different group, so instead of looking at all patients who are intermediate, low risk, it has a more sophisticated risk stratification system.
- Kirkwood AA, et al. High Dose Methotrexate Does Not Reduce the Risk of CNS Relapse in Children and Young Adults with Acute Lymphoblastic Leukaemia and Lymphoblastic Lymphoma. Results of the Randomised Phase III Study UKALL 2011, Session 614, Acute Lymphoblastic Leukemias: Therapies, Excluding Transplantation and Cellular Immunotherapies: Clinical Trials
- Goulden NJ, et al. Blood. 2017;130(supplement 1): 141
Photograph taken from http://www.ctc.ucl.ac.uk/
Copyright ©2022 Medicom Medical Publishers
« Should we use intensive chemotherapy prior to allo-HCT in rel/ref AML? Next Article
Metabolic tumor volume predicts response to loncastuximab tesirine in DLBCL »
Asciminib monotherapy in Ph+ CML: major molecular responses