Optical genome mapping changed the diagnostic classification or risk assessment for 35 of the first 200 adults tested after a Canadian tertiary care center added the method to its acute leukemia workflow. The result shows meaningful added diagnostic yield, but it does not establish that the test improves survival or should replace every existing assay.
The study was published in The Journal of Molecular Diagnostics. Vancouver General Hospital introduced optical genome mapping, or OGM, as a first-line test in November 2023 while continuing karyotyping, fluorescence in situ hybridization and panel-based next-generation sequencing.
OGM Affected Diagnosis or Risk in 35 Cases
Researchers prospectively reviewed the first 200 clinical cases after implementation. OGM detected 640 reportable variants. Of those, 25% were classified as tier 1A, 3% as tier 1B, 2% as tier 2 and 70% as tier 3A under the study's clinical reporting system.
The added findings affected 35 cases, or 18% of the cohort. Diagnostic classification changed in 12 cases, while risk stratification changed in 31. Those groups overlap, so the numbers should not be added together.
OGM found 64 variants that karyotype and FISH had missed. It also detected three KMT2A partial tandem duplications not identified by the panel-based sequencing workflow and provided results in nine cases where karyotyping failed.
Validation Showed High Accuracy but Not Perfect Detection
Before clinical deployment, the laboratory validated OGM using 83 bone marrow and peripheral blood samples representing several hematologic malignancies. The comparison produced 100% specificity, 96.1% sensitivity and 98% accuracy for the assessed variants.
Those figures describe analytical performance under the study's validation conditions. They do not mean that OGM detects every leukemia-related change. The validation recorded five false negatives or failures, and different methods remain suited to different types of genetic abnormality.
OGM works by imaging long labelled DNA molecules and comparing their patterns with a reference genome. Its genome-wide view can reveal balanced and unbalanced structural changes and copy-number abnormalities without requiring a separate probe for each suspected target. That helps explain why it added information beyond tests with lower resolution or narrower target lists.
The Study Center Kept Other Leukemia Tests
The researchers described OGM as a first-tier test at their institution, but not as a stand-alone replacement for the full workflow. Karyotyping, FISH and next-generation sequencing continued in parallel during the study period.
Turnaround time was one reason. The authors said OGM took longer than karyotyping at their center, preventing it from fully replacing that method. The study also came from one tertiary referral center, and lead investigator Tara Spence cautioned that the yield may reflect its particular patient population.
The paper assessed laboratory performance and changes to diagnostic or prognostic classification. It did not compare patient survival, treatment response, cost effectiveness or insurance coverage between testing strategies. A changed risk category can influence clinical reasoning, but that is not the same as evidence that OGM itself produced a better outcome.
Diagnostic Yield Is Stronger Than an Adoption Mandate
Finding clinically important information in 18% of cases is a serious result. It supports using OGM as part of acute leukemia testing where laboratories have the expertise, quality controls and reporting framework to interpret structural variants.
It does not justify claiming that hospitals using older workflows will have worse survival rates, that payers must cover one particular platform or that a single test can make every treatment choice. Those conclusions would require multicenter outcome and economic evidence that this study did not provide.
The real pressure is on diagnostic completeness, not on producing a universal technology mandate. The Vancouver experience shows that OGM can rescue failed karyotypes and expose abnormalities missed by parallel tests. It also shows why responsible adoption keeps complementary methods in view: higher resolution is valuable only when its limits, turnaround time and clinical interpretation are reported just as clearly as its extra findings.