Acute lymphoblastic leukaemia (ALL) is the most common childhood cancer. Matching therapies to patients based on a combination of clinical and molecular profiles has led to high survival rates, yet ALL remains one of the top causes of death in paediatric cancer due to high-risk subtypes that are associated with treatment failure and relapse. Both DNA and RNA sequencing have been used to generate these molecular profiles, with RNA sequencing (RNA-seq) particularly useful due to the ability to detect multiple hallmarks of ALL simultaneously, including gene fusions, clinically relevant gene expression profiles, and intragenic deletions.
Deletions in IKAROS Family Zinc Finger 1 (IKZF1) in B-cell ALL (B-ALL) have been a particular focus of our previous work [1,2]. These methods however have been limited to the detection of internal exon deletion events where exons excluding the first or last are deleted at the gene transcript level. While our methods could reliably detect the clinically relevant IK6 isoform of IKZF1 (exon 4-7 deletion / e3-e8 junctions) in paediatric ALL patients, detection of other likely clinically relevant IKZF1 deletion transcripts was limited. This limitation was a source of frustration, and much like the literary 'white whale,' these deletions in our cohort eluded capture with our method despite numerous attempts.
The pursuit was reignited recently by work from Tsai et al. [3] that identified transcripts for deletions of IKZF1 including exon 8 in targeted RNA-seq data, and confirmed the occurrence of these in an independent total RNA-seq data set. By modifying the algorithm in our Toblerone [2] program, we revisited a paediatric cohort of B-ALL patients, and detected known deletions in exons 4-8 and exons 2-8 from total RNA-seq. In this work we outline updates to Toblerone to accommodate terminal exon deletion and apply it to a cohort of 99 B-ALL patients that includes validated deletion samples. We also establish background distributions and set thresholds for the proportion of sequencing reads needed to support IKZF1 gene deletions. With the pursuit of the ‘white whale’ now complete, we also discuss navigating uncharted waters—extending the Toblerone method beyond IKZF1 to other relevant oncogenes.