PF160 IKZF1 DELETIONS DO NOT HAVE A STRONG IMPACT ON EVENT FREE SURVIVAL IN PATIENTS WITH B‐PRECURSOR ALL TREATED ON THE UKALL14 TRIAL (ISRCTN 66541317)

Abstract

Background:IKZF1 encodes IKAROS, a transcription factor and key regulator of lymphocyte differentiation. IKZF1 deletions (DIKZF1) are commonly detected in B‐ALL and their prognostic relevance differs between reports.Aims:We aimed to determine impact of DIKZF1 on event free survival (EFS) of patients with precursor B‐ALL aged 23–65 years recruited to UKALL14 (ISRCTN 66541317) between December 2010 and July 2018 when the trial closed to recruitment.Methods:From 655 recruits with B‐ALL, all available diagnostic DNA samples were screened by PCR (n = 497), MLPA P335 SALSA kit (n = 437) or by both (n = 419). A multiplex endpoint PCR, designed to make detection rapid and straightforward covered 4 deletions‐ the dominant negative (DN) D4–7 or the loss of function (LOF) Δ2–7, Δ4‐8, and Δ2‐8, with primer sites located close to the putative breakpoints. All PCR‐detected deletions were confirmed by Sanger sequencing.Results:In the cohort screened for both PCR and MLPA, 88/419 were not congruous. Fifty‐eight of 88 (66%) of these were explicably positive by MLPA only, due to the placement of PCR primers. Fifteen of 88 (17%) were detected by PCR but not MLPA, likely explained by the greater sensitivity of PCR. These cases are being investigated for DIKZF1 in a sub/minor clone. In order to resolve the remaining 15 disparate cases, we are in the process of mapping the breakpoint sites. By PCR, as expected, D4–7 was the most commonly detected deletion (n = 61, 12.3%) followed by D2‐7 (n = 27, 5.4%), D4–8 (n = 20, 4%) with D2‐8 being the rarest (n = 6, 1.2%). Overall, there was no significant impact of PCR‐detected DIKZF1 on EFS (see figure). In a multivariable Cox model, DIKZF1 significantly interacted with both age and Philadelphia chromosome (Ph) status, with no detrimental effect seen in Ph+ cases and any negative impact confined to Ph‐ cases which increased with increasing age. The change in effect with increasing age for the four groups (ΔIKZF1/wild type IKZF1 and Ph‐/Ph+) is shown in the figure, wherein no adverse effect is seen for ΔIKZF1 compared to wild type IKZF1 for Ph+ patients at any age, but an increased risk for ΔIKZF1 in Ph‐ ALL as age increases can be seen. Consistent with the technical approach, MLPA detected a wider range of deletions than PCR:‐ D1‐2 (n = 9, 1.8%), D1‐3 (n = 3, 0.6%), D1‐7 (n = 5, 1%), D1‐8 (n = 36, 7.2%), D2‐3 (n = 4 0.8%,) D2‐7 (n = 32, 6.4%), D2‐8 (n = 15, 3%), D4‐7 (n = 49, 9.9%), D4‐8 (n = 19, 3.8%), D6‐8 (n = 1, 0.2%). As with PCR‐detected lesions, there was no impact of MLPA‐detected DIKZF1 on EFS in univariable analysis. In multivariable analysis, the differing effect with age was less evident and did not reach significance, though the same interaction with Ph status was observed. We did not detect any differential impacts by analysing PCR or MLPA‐detected DN or LOF deletions separately.Summary/Conclusion:Both PCR and MLPA have limitations for routine detection of DIKZF1 in clinical trials. However, in this final report on the prognostic relevance of DIKZF1 on the UKALL14 trial, DIKZF1 did not have a strong impact on EFS.image