Approximately 20–25% of children with B-cell precursor ALL harbour the t(12;21) (p13;q22)/ETV6-RUNX1 fusion which has been associated, almost universally, with a favourable outcome. However, as some ETV6-RUNX1 patients relapse additional predictors of treatment failure are required. Although additional abnormalities involving 12p and 21q (AA) are known to occur in these patients, previous studies have produced conflicting results with respect to their prognostic impact. In order to investigate the prognostic relevance of AA in this subset of patients, we screened a large cohort of children by FISH with break-apart probes targeting the ETV6 (DakoCytomation, Denmark) and RUNX1 loci (centromeric - RP11-272A03 and telomeric - RP11-396G11, Sanger Institute, UK).

Among 368 ETV6-RUNX1 patients treated on the MRC ALL97 childhood treatment trial, material was available for testing on 247 (67%). Sample availability was not related to sex or age, but those tested were more likely to have a white cell count (WCC) of >50×109/L and hence had a borderline inferior event free survival (EFS) at 5 years: 86% (SE 2.2%) v 93% (2.4%), p=0.05. FISH testing with the ETV6 and RUNX1 probes was successful in 245 (99%) and 244 (99%) patients, respectively.

In total, 202 (82%) patients harboured at least one of the following five AA:

  1. loss of the untranslocated ETV6 allele (ETV6del): 165 (67%);

  2. Gain of a normal chromosome 21 (+21): 57 (23%);

  3. Gain of a der(21)t(12;21) chromosome [+der(21)]: 38 (16%);

  4. deletions from the der(12)t(12;21) [der(12)del]: 20 (8%) and

  5. gain of the der(12) t(12;21) [+der(12)]: 18 (7%).

In 74 (30%) cases two or more of these AA were observed. Evidence that the ETV6-RUNX1 clone had doubled to a near-tetraploid clone was seen in 9 (4%) cases. In the majority of ETV6del cases (158, 96%) the deletion resulted in the loss of the entire ETV6 probe which spanned a region of ~750kb. The proportion of ETV6-RUNX1 positive cells carrying each specific AA ranged from 6% to 100% but in the majority of cases comprised more than 75% cells: 72%, 54%, 63%, 100% and 39%, respectively.

The frequency of the AA did not vary by sex, with the exception of +der(12) which was more common among females (12%) than males (4%) [p<0.03]. Patients with +der(12), +der(21) or +21 were significantly older compared to other patients: mean age 5.9 years v 4.6; 5.4 v 4.6 and 5.3 v 4.6 respectively, all p<0.03. No age association was seen for patients harbouring either of the deletions. Patients with +der(21) were significantly less likely to have a WCC >50×109/L [1/38 (3%) v 54/206 (26%), p=0.001]. Although a similar trend was seen for patients with +21 it was not significant: [8/57 (14%) v 47/187 (25%), p=0.079].

During a median follow-up period of 7 years, a total of 41 (17%) adverse events were observed, including 34 (14%) relapses and 16 (6%) deaths. The EFS and OS figures at five years were 86% (2.2%) and 94% (1.5%) respectively. None of the AA predicted a higher risk of relapse or death, whether considered separately or together. None of the 20 patients with der(12)del suffered a relapse or died; but as the total number of relapses and deaths was small, this difference was of only borderline significance (p=0.046).

We report for the first time that the spectrum of AA in ETV6-RUNX1 positive patients extends to the der(12) and that der(12) deletions may correlate with an excellent outcome. Altogether more than 80% of ETV6-RUNX1 patients harbour an AA involving 12p or 21q. This study, based on 247 patients treated on a single protocol and with 7 years follow-up, is the most comprehensive to date. Contrary to previous reports, we conclude that AA involving 12p and 21q are not predictors of relapse or death in the context of modern treatment protocols.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author