First Analysis of the UKALL14 Phase 3 Randomised Trial to Determine If the Addition of Rituximab to Standard Induction Chemotherapy Improves EFS in Adults with Precursor B-ALL (CRUK/09/006)

UKALL14 (NCT01085617) randomised 655 patients aged 25-65 years with B-precursor ALL, irrespective of Philadelphia chromosome (Ph) status or cell surface CD20 expression to determine if the addition of four doses of rituximab to standard induction chemotherapy (SOC+R) resulted in improved event free survival (EFS). Patients were recruited between Dec 2010 - Jul 2017 and the primary analysis population comprises 577 patients recruited after an April 2012 amendment in which the SOC therapy was altered. The trial was powered with an 84% chance to detect a 12% improvement in EFS (Hazard ratio (HR): 0.71). Secondary endpoints included complete remission (CR), OS, non-relapse mortality, levels of minimal residual disease (MRD) after induction and relationship of response to CD20 expression. SOC consisted of daunorubicin 30mg/m², vincristine 1.4mg/m², dexamethasone 4 day blocks of 10mg/m² starting d1, 8, 15 and 22. Pegylated asparaginase 1000IU/m² was added on d4 and 18 (d18 only if >40 years) for Ph- ALL and continuous daily imatinib 600mg for Ph+ ALL. Intrathecal MTX was given on d14. Rituximab 375mg/m² was given on d3,10,17 and 24. Two patients did not start trial treatment, both were in the SOC+R arm. Analysis is by intention-to-treat. There were 288 patients in the SOC arm and 289 in the SOC+R arm, 273 (95.5%) of whom received all 4 doses of rituximab. The arms were well-balanced for risk characteristics. Of note, 63.9% (SOC) and 62.6% (SOC+R) were aged over 40 years at randomisation and 86 patients in each arm (29.9% and 29.8%, respectively) had Ph+ ALL. CR rate, 92.7% SOC and 94.8% SOC+R, did not differ between the arms. There was no difference in the MRD response between the arms, whether assessed as positive vs. negative or as a continuous variable; 121 (42.2%) of patients were MRD negative at induction completion in the SOC arm vs. 120 (41.8%) in the SOC+R arm. Likewise, the rate of severe/adverse events and non-relapse mortality did not differ between arms. At a median follow-up of 50.5 months (7 days - 83.6 months), 3 year EFS for SOC is 41.9% (95% CI: 35.8 - 48.0) versus SOC+R 48.7% (42.4 - 54.8), Hazard Ratio(HR) 0.88 (0.71 - 1.11), p=0.28, see figure1a. Pre-planned subgroup analyses by cytogenetic ("high risk" was defined as t(9;22), t(4;11), low hypodiploidy/near triploidy and complex karyotype) and other risk groups (age, presenting WBC) as well as by cell surface CD20 expression did not reveal any significant interactions. However, we did find that % blasts expressing CD20 was an independent poor prognostic factor; Youden's cut-off, to determine the best cut-off for a continuous variable, suggested a % CD20 expression of 11.7% as the optimal cut-off. EFS HR for 10-20% CD20 was 1.74 (0.98-3.10) and >20% was 2.20 (1.27 - 3.81), compared to <10% CD20. Outcomes were also analysed by post-induction treatment assignment; myeloablative allogeneic stem cell transplant (MAalloSCT) was assigned for patients with high risk ALL aged 40 and under with suitable donor, reduced intensity-conditioned RICalloSCT was evaluated, as part of the trial, in all patients over 40 years, with suitable donor and consolidation/maintenance was assigned to standard risk/no suitable donor. Of note, there was a large and statistically significant benefit to SOC+R among those who received MAalloSCT, SOC N= 53, 3 year EFS 50.7% (35.9 - 63.7) vs. SOC+R, N=40, 3-year EFS 72.2% (54.9 - 84.2), HR 0.47 (0.23 - 0.95), p=0.03 (figure 1b) which was not evident in the RICalloSCT or maintenance groups and related to a reduction in relapse risk. Among those who received MAalloSCT, SOC and SOC+R, the arms were not totally balanced, with a greater preponderance of high-risk genetics in the SOC+R arm: 29 (64.4%) SOC versus 28 (77.8%) SOC+R, suggesting that the benefit of SOC+R combined with MAalloSCT was able to outweigh some of the adverse genetic risk. By contrast to the GRAALL-2005/R study, in which rituximab was given during all treatment phases, for a total of 16 to 18 infusions, we did not find an overall statistically significant benefit for 4 infusions of rituximab during induction. The non-significant trend to a better outcome with SOC+R underscores the necessity for including 16-18 doses of rituximab to obtain the full benefit. However, our trial shows no treatment interaction of SOC+R with Ph+ status or CD20+ expression level, suggesting that the GRAALL-2005/R results may possibly be generalizable to all patients with B-precursor ALL.

View largeDownload slide

View largeDownload slide

Close modal