Improved Outcome for Children and Young People with T-Acute Lymphoblastic Leukaemia: Results of the UKALL 2003 Trial

T-ALL accounts for around 15% of paediatric ALL and carries a less favourable prognosis than B-ALL due to a higher relapse rate (RR) and poor response to salvage therapy. We sought to determine if the outcome for patients with T-ALL had improved on a contemporary, response based treatment protocol, UKALL 2003, which did not include pre-emptive cranial radiotherapy or high dose methotrexate. NCI standard risk patients received a 3 drug induction with dexamethasone, vincristine and pegylated asparaginase (1000 units/m²) (regimen A). NCI high risk (HR) patients also received daunorubicin (regimen B). Early morphological marrow response was measured on day 8 for NCI HR patients and day 15 for NCI standard risk patients except patients > 16 years. Rapid early responders were stratified by minimal residual disease (MRD) at day 29 which was measured by RQ-PCR. MRD negative patients were randomised to standard or reduced intensity therapy. MRD high risk (>0.01%) patients were randomised to standard therapy (regimen A/B) or an intensified regime consisting of augmented BFM consolidation, escalating Capizzi methotrexate and 2 delayed intensifications (regimen C). Patients with high risk cytogenetics or a slow early morphological response received regimen C, regardless of MRD. Regimen A and B patients received 4 doses of pegylated asparaginase (2 in induction and 1 in each DI) and Regimen C patients received an additional 8 doses.

Of 3126 patients in the trial, 388 (12.4%) had T-ALL. Of those, 73% were male with a median age of 10 yrs (range 1-24). Median WCC was 94x10⁹/L (range 1-881) and 85% were HR by NCI criteria. 76% had a rapid early response (RER), defined as <25% bone marrow blasts. MRD was evaluable for 72% of patients. MRD at day 29 was ≥0.1% in 37%, detectable at a level of <0.1% in 22% and undetectable in 41%. Compared to the trial population as a whole, patients with T-ALL were more often stratified to receive the higher intensity regimens. T-ALL 11%, 52% and 37% treated on Regimens A, B and C compared with 51%, 26% and 23% for trial overall.

With a median follow up of 5 yrs 7 mths (range 1 yr 4 mths – 9 yrs 11 mths), patients with T-ALL had an inferior outcome compared to those with B-ALL due to a higher risk of relapse (p=0.0003) and death in remission (p=0.04); 5 year event free survival (EFS) 81.2% (95% CI 77.3-85.1) vs 88.1% (86.6-89.5), overall survival (OS) 86.4% (82.9-89.9) vs 92.4% (91.4-93.4), RR 13.5% (10.0-17.0) vs 8.2% (7.0-9.4) and death in remission 4.3% (2.3-6.3) vs 2.4% (1.8-3.0). The site of excess relapses in T-ALL was both bone marrow (10.1% (7.0-13.2) vs 5.9% (4.9-6.9)) and isolated CNS (3.5% (1.5-5.5) vs 1.9% (1.3-2.5)). However, the CNS relapse rate is comparable to other protocols containing cranial radiotherapy and high dose methotrexate. The excess deaths in remission are a reflection of the higher intensity regimens that a majority of T-ALL patients received. In contrast to B-ALL (data not shown), increasing WCC was not a risk factor for relapse in T-ALL. 5 year EFS for T-ALL patients with a WCC <100 was 85% vs 81% for those with a WCC ≥100. CNS relapse rate was not different between the two groups, 5% vs 8% (p=0.3). There was no difference in outcome between those with a WCC of 100-300 and ≥300. Response to induction was the strongest predictor of relapse. Patients with a slow early response had a high relapse risk (Hazard Ratio=2.31 (1.11-4.84), p = 0.03 in multivariate analysis) despite receiving Regimen C. Patients with MRD ≥0.1% had a 5 year EFS of 71.5% compared to 89.6% for the MRD <0.1% group and 90.3% for the MRD negative group (log rank p(trend)=0.001), RR 23.2% vs 7.1% vs 8% (p=0.003). There was a trend for a worse outcome in the patients aged ≥16 although this was not statistically significant, 5 year EFS 74.1% vs 82.8% (log rank p=0.7), RR 20.8 vs 11.9%, (p=0.09). Compared with our previous trial, UKALL97/99, outcome for T-ALL patients has improved with a higher EFS, 81.7% vs 72.8% at 5 years (log rank p=0.08), due mainly to a non-significantly lower relapse rate 13.5% vs 20.6% (p=0.09).

T-ALL outcomes have improved with better risk stratification and the use of dexamethasone and pegylated asparaginase. Pre-emptive cranial radiotherapy or high dose methotrexate are not essential to preventing CNS relapse. The outcome for patients with T-ALL, however, remains inferior to those with B-ALL highlighting the need for novel therapies for patients with high risk T-lineage disease.