Childhood T-ALL: it's time to move on

The results of a randomized trial of high-dose methotrexate (HD MTX) in childhood T-cell ALL (T-ALL) reported by Asselin et al on behalf of the Children's Oncology Group (COG) in this month's issue of Blood demonstrate both the promise and the challenge of further improving the outcomes of children with high-risk ALL through intensified application of conventional chemotherapeutic agents.¹ 

Recent nonrandomized studies in childhood T-ALL report an ∼ 70% to 75% 5-year event-free survival (EFS).^2-5  The trial reported by Asselin and colleagues, POG 9404, confirms a previous nonrandomized report of the benefit of HD MTX in this high-risk group of patients.²  In POG 9404, children with newly diagnosed T-ALL or non-Hodgkin lymphoma (NHL) were randomized to receive 4 doses of 5 g/m² of HD MTX as a 24-hour infusion in the context of a modified intensive Dana-Farber Cancer Institute (DFCI) regimen.³  The benefits of HD MTX were substantial enough to warrant early closure of the randomization. That's the good news.

The not-so-good news: outcome on the HD MTX arm of POG 9404 was no better than other T-ALL regimens. The results of the POG 9404 trial demonstrate that even seemingly minor modifications to a treatment plan can significantly impact outcome. While POG 9404 closely mirrored the DFCI 87-01 regimen, modifications, including a delay in the introduction of prophylactic cranial irradiation (CXRT) and a reduction in the intensity of early intrathecal chemotherapy, were made in an effort to minimize the risk of treatment-related neurotoxicity. These changes likely contributed to a higher rate of CNS-involved relapses, many of which occurred during the first 6 months of treatment, on the POG study compared with the DFCI regimen on which it was based. For T-ALL patients treated without HD MTX on the modified control arm of the POG protocol, the overall outcome (5-year EFS of 68%) was lower than expected, while the outcome of the HD MTX arm (5-year EFS of 79%) was similar to those of other childhood T-ALL trials. Thus, the addition of HD MTX, while associated with an improvement in outcomes in those who received it compared with those who did not, did not provide a survival advantage compared with the original DFCI regimen.

So one lesson of POG 9404 is that the addition of HD MTX may allow CXRT to be delayed in patients with T-ALL without a negative impact on EFS. What we did not learn is whether the addition of HD MTX would allow CXRT to be eliminated altogether. In the trial reported by Asselin et al, as in previous BFM and DFCI trials, T-ALL patients received prophylactic CXRT. Current protocols typically use a lower dose (12 Gy instead of 18 Gy), but CXRT is still administered to the majority of patients with T-ALL on most treatment regimens. Recently, other groups have reported similar EFS results for patients with T-ALL treated without CXRT using 4 doses of postremission HD MTX and additional doses of intrathecal chemotherapy.^4,5  This strategy may be successful for some patients with T-ALL, although in general T-ALL patients treated without CXRT have a higher risk of CNS-involved relapses compared with similarly treated B-precursor patients, and for some T-ALL patients (eg, those with high-presenting leukocyte counts or CNS-3 status at diagnosis), relapse risk may exceed rates observed in trials using CXRT.⁴  In addition, sequelae of the various CNS-directed therapies in this population remains to be elucidated. Whether or not CXRT and/or HD MTX are used, the bottom line is that 20% to 25% of children with T-ALL treated with current regimens experience a relapse, and salvage after relapse remains dismal. While sorting out which subsets of patients may avoid CXRT, we need to focus on strategies to improve cure rates. Achieving this goal will require novel approaches involving more effective T-ALL–specific drugs and better prognostic factors to identify T-ALL patients who are not cured by currently available therapies.

Nelarabine, a synthetic deoxyguanosine derivative that is arguably the most promising new drug for T-ALL to emerge in more than a quarter of a century,⁶  is currently undergoing evaluation in a randomized COG study (AALL0434) of children with newly diagnosed T-ALL. Despite its significant efficacy in recurrent T-cell disease, nelarabine has been associated with substantial neurotoxicity in the relapse setting, leading to reservations about its widespread use in newly diagnosed patients. The identification of abnormalities in the Notch pathway⁷  holds the promise of new therapeutic strategies, such as γ secretase inhibition. Although early attempts to introduce such interventions have not translated into therapeutic benefit for patients with T-ALL,⁸  a number of new compounds with strong preclinical rationale and activity are now entering clinical evaluation.

In addition to new drugs, we need better prognostic factors to identify those patients most in need of more effective therapies. On the POG 9404 trial, age ≥ 10 years, high white blood cell count, and male sex were adverse prognostic factors. Other studies have not confirmed these results. The use of minimal residual disease assessments to tailor postremission therapy for patients with T-ALL is currently under investigation. A distinct subset of childhood T-ALL, termed early precursor T-cell ALL (ETP),⁹  has been associated with an especially poor outcome, and, if prospectively validated as an independent prognostic factor, may help to identify patients at diagnosis who would benefit from alternative treatment approaches.

In summary, the report by Asselin and colleagues confirms the activity of HD MTX in newly diagnosed childhood T-ALL, while also demonstrating that reductions in the intensity of CNS-directed therapy may dilute its therapeutic benefit. How, when, and in what therapeutic context to best apply HD MTX in this population remains unclear. Other strategies to optimize the use of standard antileukemic agents, such as the postinduction intensification of asparaginase,³  may also be beneficial. However, for children with high-risk T-ALL (such as ETP-ALL) and those with recurrent disease, improvement in outcome will require the introduction of new agents targeting critical pathways in T-ALL leukemogenesis, chemotherapy-resistant leukemia-initiating cells,¹⁰  or both. Rearranging the deck chairs on the Titanic will not benefit children with ETP-ALL; it is time to move on to the evaluation of molecularly targeted therapies in childhood T-ALL.