Incorporating cytogenetics into risk stratification for the treatment of childhood ALL has contributed to increased survival rates. However 25% patients, the B-other subgroup, do not have any of the known major chromosomal abnormalities. Within this group, BCR-ABL1-like cases show a similar gene expression profile to those with BCR-ABL1 fusion and share the same high risk of relapse. BCR-ABL1-like cases are genetically heterogeneous and some harbour tyrosine kinase activating gene fusions e.g. EBF1-PDGFRB. Recent reports suggest patients with EBF1-PDGFRB ALL who are refractory to conventional therapy respond to the tyrosine kinase inhibitor (TKI), imatinib. Here we present 14 patients treated on ALL97 (n=3), UKALL2003 (n=10) or UKALL2011 (n=1) who were EBF1-PDGFRB positive including 1 patient treated with imatinib.

Involvement of PDGFRB and EBF1 was confirmed using dual colour break-apart probes (normal signal pattern=0R0G2F). Eleven patients showed signal patterns (1R0G1F) consistent with a deletion of 5q33 having breakpoints within EBF1 and PDGFRB, as seen in previously reported cases of the fusion. The presence of the deletion was confirmed by single nucleotide polymorphism (SNP) array in 6 cases. Three cases showed signal patterns (1R1G1F) consistent with a balanced rearrangement resulting in the EBF1-PDGFRB fusion and a balanced t(5;5)(q33.1;q33.3) was seen in one of these cases. SNP array analysis for 2 of these cases showed no copy number abnormalities of chromosome 5. However, in the third case a series of deletions were seen along the long arm of chromosome 5, indicating that the EBF1-PDGFRB fusion was generated by a complex rearrangement such as chromothripsis. Multiplex Ligation-dependent Probe Amplification (MLPA) using the P335 IKZF1 kit (MRC Holland) was carried out in 11 cases. Deletions of the probe for exon 16 of EBF1 were seen in all cases with the 5q33 deletion. Loss of PAX5 was detected in 4 patients, while IKZF1 and CDKN2A/B were deleted in 3 cases each.

Among the 14 patients there was a predominance of females (n=10). The median age was 12 years with 9 patients >10 years at diagnosis. The median white cell count (WCC) was 34.4 x109/L with 5 patients presenting with a WCC of >50×109/L. Two of 3 patients treated on ALL97 remain alive (one after relapse); both patients received regimen C and a transplant. The third patient, who was not treated as high risk, relapsed and died within 5 years.

An unselected cohort of 276 B-other UKALL2003 patients revealed 8 (3%) cases of EBF1-PDGFRB (~0.6% of BCP-ALL). All 8 patients achieved CR but were MRD positive at day 29. Overall 4 patients relapsed, including 2/3 patients who received standard chemotherapy (regimen A/B) and 2/5 patients who received high risk chemotherapy (regimen C). Three relapsed patients were salvaged and remain in second CR 1-7 years post relapse. Targeted screening of 13 UKALL2003 patients who failed induction revealed 2 more patients with EBF1-PDGFRB; one subsequently died while the second remains in 1st CR 7 years later. None of the UKALL2003 or ALL97 EBF1-PDGFRB patients received TKI therapy.

As a result of these findings, patients entered on UKALL2011 are screened for EBF1-PDGFRB fusion if they fail to achieve CR by day 29 or remain MRD positive (>0.5%) at week 14. A 5 year old girl who failed induction (50% blasts at day 35) tested positive for EBF1-PDGFRB. This prompted her withdrawal from UKALL2011 and she was treated according to EsPhALL with standard BFM consolidation plus imatinib (300mg/m2) followed by 3 intensive BFM HR blocks. After 4 weeks consolidation therapy her MRD levels had fallen to 6% and she was MRD negative by flow cytometry after 8 weeks. She continues in molecular remission 2 months after a transplant.

In this largest cohort of EBF1-PDGFRB patients reported to date, we demonstrate the genetic mechanisms by which the fusion occurs and the spectrum of cooperating mutations. EBF1-PDGFRB fusion is associated with female sex, older age and a mixed outcome. Although these patients had high levels of MRD and a high relapse rate there was also evidence of durable remissions; especially with intensive chemotherapy. Evidence from this study and others suggest patients with this abnormality who fail standard chemotherapy may respond to imatinib. It is possible that treatment with imatinib might avoid the need for intensive chemotherapy to achieve a cure in these patients but that would need to be tested within a prospective trial.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution