Relapse in Teenage and Young Adult (TYA) Patients Treated on a Pediatric Minimal Residual Disease (MRD) Stratified Protocol Is Associated with a Poor Outcome: Results from UKALL2003

The superior outcomes of teenage and young adult (TYA) patients with acute lymphoblastic leukemia (ALL) treated on contemporary pediatric minimal residual disease (MRD) directed chemotherapy regimens compared to adult ALL protocols are well established. However, there are limited data on the outcome of relapse following such regimens and consequently no consensus on optimal salvage strategies.

UKALL 2003 was a prospective, randomised controlled trial investigating treatment modification according to MRD status at the end of induction in consecutively diagnosed children and young people with ALL recruited from the UK and Republic of Ireland between October 2003 and June 2011. The trial was initially open to children up to the age of 18 years but the upper age limit was increased to 20^th birthday in 2006 and 25^th birthday in 2007. Here we report the relapse outcomes of 16 - 24 year old TYA patients recruited to the trial.

All TYA patients recruited to UKALL 2003 who subsequently relapsed were identified using the trial database, which also provided data regarding prognostic factors at presentation, MRD risk group, treatment allocation, site of and time to relapse and overall outcome. Individual centers were then contacted for follow up information including salvage chemotherapy, disease response, use of transplantation, subsequent outcome and cause of death.

Of a total of 229 TYA patients recruited, 42 (18.3%) relapsed. The immunophenotype in those who subsequently relapsed was precursor B cell (B-ALL) in 30 patients and T cell (T-ALL) in 12. The median white cell count at diagnosis was 22x10⁹/l and 1 patient had central nervous system (CNS) involvement. In B-ALL, 4 had good risk cytogenetics (all high hyperdiploid), 13 had intermediate risk, 6 had high risk (3 with MLL rearrangements), and 7 had unknown cytogenetics. End of induction MRD risk group was low risk in 2, intermediate risk in 11 and high risk in 29.

The median time from diagnosis to relapse was 17 months (17.5 months for B-ALL and 14 months for T-ALL, p=0.1). The site of relapse was isolated bone marrow (BM) in 30 patients, combined BM and CNS in 4, combined BM and testicular in 1, and isolated CNS in 6 (T-ALL; 8 BM, 3 BM+CNS, 1 isolated CNS) (B-ALL; 23 BM, 1 BM+CNS, 1 BM + testicular, 5 isolated CNS).

36 of the 42 relapsed patients were given salvage chemotherapy, using a variety of regimens including R3 (n=9), Fludarabine/Cytarabine/Idarubicin based (n=16), and Clofarabine based (n=4). One patient received Blinatumamab and one entered the MARALL study (a Phase I/II study of the combination of Veltuzumab (anti-CD20) and Epratuzumab (anti-CD22) with intensive chemotherapy in patients with relapsed B-cell ALL). A second complete remission rate was achieved in 22 patients (61%) and a further 2 had an empty marrow without count recovery. 23 patients proceeded to an allogeneic hemopoietic stem cell transplant.

The median overall survival (OS) from relapse was 213 days (actuarial 5 year OS of 18%). 9 patients remain alive; 8 in remission (1 unconfirmed) and 1 undergoing active treatment. The cause of death in the 33 patients who died was relapsed/refractory disease in 20, complications of re-induction treatment in 4, transplant related complications in 6 and unknown in 2. 1 further patient died of sepsis post transplant in the context of pancytopenia and suspected (though unproven) relapse.

Those relapsing earlier than the median of 17 months had a shorter median survival and 5 year OS from relapse compared to those relapsing at >17 months (median 113.5 v 400 days, and OS 10% v 37%, p=0.0075). When this analysis was restricted to only the B-ALL cohort the effect was even more marked (median 151 v 595 days, OS 0% v 45%). No patient with B-ALL relapsing ≤24 months was salvageable. There were no differences in salvage approach between the early and late relapsing cohorts, nor differences in MRD risk. The late relapsing cohort did have more patients with good risk cytogenetics, and fewer with high risk cytogenetics.

TYA patients with ALL who relapse despite intensive, contemporary MRD directed chemotherapy have an extremely poor outcome. Patients with B-ALL relapsing at 24 months from diagnosis or earlier are unsalvageable. Improving the outcomes of TYA ALL patients will depend on a) more accurate early identification of those destined to relapse and b) use of innovative therapies in those at high risk of relapse or with overt relapse.