Resolution of a Clinical Dilemma with Whole Genome Sequencing, and Discovery of a New Mechanism for Generating PML-Rara: Insertional Fusion

Abstract 2755

We describe a difficult diagnostic case of t(15;17)-negative acute promyelocytic leukemia (APL). A 39 year-old woman presented with pancytopenia and low grade DIC. Bone marrow biopsy revealed AML with promyelocytic features. She was treated with Cytarabine, Daunorubicin and ATRA. However, ATRA was discontinued after FISH revealed a possible RARA-PML, but no PML-RARA fusion (one fusion, one RARA and two PML signals), and cytogenetics did not demonstrate a translocation involving chromosome 15 or 17. In fact, her cytogenetics revealed a complex pattern that predicted poor prognosis (46 XX del(9)(q12q32),del(12)(q12q21)[6]/46,idem,-6,-16,add(16)(p13.2),+2 mar[13]/46 XX[1]). Following reinduction, she entered complete remission and was empirically consolidated with arsenic. RT-PCR for PML-RARA was not performed at diagnosis, and was negative at the time of consultation in remission. Her complex cytogenetics and uncertain FISH status posed a diagnostic and therapeutic dilemma that could only be resolved by whole genome sequencing in the time frame required for a clinical decision to be made (i.e. allogeneic transplantation vs. ATRA-based consolidation).

DNA was therefore generated from bone marrow (cryopreserved at the time of diagnosis) and a skin sample (obtained in remission), and subjected to massively parallel sequencing using paired-end reads (Mardis et al, NEJM 2009). We generated 187 billion bp (tumor) and 200 billion bp (skin) of sequence, corresponding to 43.7x and 46.8x haploid coverage (99.76% and 99.74% diploid coverage) for the respective samples. Within six weeks of sample receipt, validated results were available. We confirmed the del(9)(q12q32) and del(12)(q12q21) somatic events and identified a novel oncogenic form of chromosomal rearrangement, an insertional fusion: 77 Kb of chromosome 15 (chr15:72027045–72104108 containing LOXL1 exon 6 through PML exon 3) were inserted en bloc into RARA intron 2 (chr17:35742678–35742683). This event resulted in the expression of PML-RARA (bcr3 isoform) and two novel fusion transcripts (RARA-LOXL1 and LOXL1-PML), which were all successfully amplified by RT-PCR and sequenced. The RARA-LOXL1 and LOXL1-PML fusions both created stop codons shortly after the fusion events. Re-evaluation of the FISH results revealed that the insertion generated a fusion signal, while loss of 77 Kb from the PML locus did not prevent binding of the 239 Kb commercial PML probe to chromosome 15 (thus generating 1 fusion, 1 RARA and 2 PML signals). These signals represented the PML-RARA insertional fusion event, not the RARA-PML translocation that was originally reported. We further identified and validated deletions on chromosomes 12 (60 Mb), 14 (22 Kb) and 19 (30 Kb) and non-synonymous, somatic single nucleotide variants (SNVs) in the coding regions of ZNF687, DYTN, C3orf54, CH3D19, SLC35A4, GPRC6A, ZFHX4, PTK2, PITPNM1, DEGS2, PCSK2, CDC45L, although the clinical relevance of these deletions and point mutations is not yet known. After validating PML-RARA bcr3 expression, a decision was made to consolidate the patient with an ATRA-containing regimen.

Using whole genome sequencing with paired end reads, we have identified a novel oncogenic form of chromosomal rearrangement, an insertional fusion. Similar insertional events may occur in other loci. Small structural events (under a few megabases in size) are often undetectable by conventional cytogenetics and FISH, and are expected to be invisible to standard break-apart probes (commonly used to evaluate the RARA and MYC loci). This case highlights the clinical relevance of whole genome sequencing for informing diagnostic and therapeutic decisions that must be made within weeks after sample acquisition.