Optical Mapping, a Promising Alternative to Gold Standard Cytogenetic Approaches in Acute Lymphoblastic Leukemias: A Blind Comparison on 10 Patients

Acute lymphoblastic leukemias (ALL) are characterized by a large number of genetic abnormalities of prognostic and theranostic interest. Their detection requires the use of several conventional and molecular cytogenetic techniques in order to establish the most exhaustive genetic profile: conventional karyotype, FISH, SNP-Array and RT-MLPA for detection of fusion transcripts. Optical genome mapping technique is based on analysis of ultra-high molecular weight DNA (UHMW: 150 kbp-2.5 Mbp). It allows an automatic, high resolution, genome-wide detection of structural anomalies, including balanced translocations and copy number anomalies, making this technology a promising tool in ALL. However, very few data are available in ALL so far. Our objective was to compare optical mapping (Bionano Genomics) to standard techniques performed routinely (combination of karyotype, FISH, SNP-array and RT-MLPA) in 10 selected B or T-ALL patients carrying at least one major genetic abnormality including t(10;14) TLX1-TRA, t(9;22) BCR-ABL1 (e1a2 and e14a2 transcripts), t(10;11) MLLT10-PICALM, t(11;19) KMT2A-MLLT1, typical hyperdiploid profile, iAMP21, SET-NUP214 fusion, various deletions involving CDKN2A/CDKN2B, IKZF1, PAX5, PTEN, ERG and FBXW7. Abnormalities on SNP-array were retained when allele fraction was > 10% and larger than 1Mbp or > 20 kbp and involving one of the 186 genes defined as potentially relevant in leukemogenesis. A total of 79 anomalies meeting these criteria in SNP-array, and/or found by other standard techniques were identified (45 deletions, 16 whole chromosome gains, 10 duplications or partial gains, and 8 translocations).

Bone marrow DMSO pellets from the 10 patients, all treated at the CHU Amiens-Picardie, were used for this study. The UHMW DNA extraction technique provided molecules with an average N50 (≥ 150 kbp) of 286 kbp, and an average coverage of 339x. Direct imaging of the DNA molecules was performed on the Bionano Saphyr system and structural variants were automatically detected using the Rare Variant Pipeline. Abnormalities analysis was blindly performed by two independent operators on the Bionano Access software (v 1.5.2.), one cytogeneticist with no training in optical mapping (operator 1) and the other optical mapping technician without training in cytogenetics (operator 2).

Among the 79 relevant anomalies, 72 (91.1%) were found by operator 1 and 71 (89.9%) by operator 2. All (8/8) translocations, 43/45 (95.6%) of deletions, 8/10 (80%) of gains were found by both operators whereas 13/16 (81.25%) and 12/16 (75%) of chromosomal gains were found by operator 1 and 2 respectively. Concerning the 8 discrepancies between optical mapping and standard techniques, 4 of them were filtering issues. Three were low allelic frequency large gains clearly visible in raw data but rejected by software's CNV filters, totally, or partially for a trisomy 3 that caused the discrepancy between the two observers. The fourth was a deletion of VPREB1 gene, erroneously eliminated by operators because included in recurrent deletion zones of IGL locus. The 4 others discrepancies were non detected anomalies. Two of them were copy number variations in Xp22.33 (pseudoautosomal region), a non-covered area with the optical mapping technology, one of which was a P2RY8-CRLF2 fusion. Another sub-clonal gain (trisomy 21) found by karyotype and FISH (quantified at 7% of all nuclei) but not by SNP-array, was missed, even in raw data. The last was a near-tetraploid subclone, found only on karyotype, that might be strictly tetraploid and by definition undetectable by quantitative techniques. Interestingly, new anomalies, not detected by standard techniques, were revealed, including a translocation t(11;14)(p13;q11.2) LMO2-TRA, a translocation with inversion t(7;8)(q34;q24.1) TRB-MYC and a translocation t(14;20)(q32.33;q13.13) IGH-CEBPB. At last, in 3 cases, unidentified complex rearrangements by cytogenetics could be resolved using optical mapping.

Thus, optical mapping represents an innovative, time and cost effective alternative to the different cytogenetic techniques currently performed routinely for ALL genetic characterization. It enables identification of complex cytogenetic events, including those currently inaccessible to standard techniques such as chromotripsis or certain complex translocations.