WGS Outshines Cytogenetics in Myeloid Cancers

Whole-genome sequencing (WGS) is better at detecting chromosomal abnormalities than traditional cytogenetic analyses in patients with certain myeloid cancers, a recent study reveals (N Engl J Med 2021;384:924–35). The technique could provide a faster, more accurate option for determining risk of worsening disease and choosing appropriate treatments.

The standard workup for patients with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS) involves cytogenetic analysis, including karyotyping and FISH, along with targeted sequencing of key genes. These procedures can identify chromosomal alterations and point mutations that influence patients' prognoses and treatment. WGS is a possible alternative approach, but its usefulness had never been tested in these cancers.

Using 235 samples from patients with AML or MDS, David Spencer, MD, PhD, of Washington University School of Medicine in St. Louis, MO, and colleagues compared WGS to cytogenetic and sequencing analyses. The researchers found that WGS pinpointed all the translocations and copy-number variants revealed by cytogenetics, as well as a substantial number of changes that cytogenetics did not detect. Overall, they uncovered chromosome alterations in an additional 40 patients, including two who carried the CBFB–MYH11 gene fusion, which indicates favorable risk.

Targeted sequencing performed better than WGS at detecting mutations in individual genes. WGS uncovered 84.6% of the single-nucleotide alterations and 91.5% of the insertion–deletion mutations flagged by targeted sequencing. Spencer says that the difference stems from the higher coverage for targeted sequencing—more than 500x, versus a median of 50x for WGS.

However, WGS offered more accurate risk assessments than the other methods, the team discovered. For example, cytogenetic analyses could not provide a prognosis for 27 patients, but WGS divided these patients into adverse and not-adverse—which includes favorable and intermediate risk—categories. Overall survival data showed that none of the patients in the adverse risk group were alive at 20 months, whereas about half of the patients in the not-adverse category were.

To determine whether WGS is clinically practicable, the team analyzed 117 samples submitted for cytogenetic analysis from new patients. WGS took a median of 5.1 days, whereas the standard time for cytogenetic analysis is 10 days, with an estimated cost of $1,300 to $1,900. Although clinical WGS will likely be more expensive, the study shows that the cost will probably be comparable to the standard methods, says Spencer. “Whole-genome sequencing performed in this way is a practical, feasible, accessible, and impactful clinical assay,” he says. “Clinical laboratories that already perform sequencing should be able to replicate what we did.”

Bruno Medeiros, MD, of Stanford University in California, is enthusiastic about the possible use of WGS for routine genomic profiling in myeloid cancers. “Molecular classification is critical for selection of optimal treatments,” he says. That the researchers obtained that information in such a short time “is huge,” he says. “If this approach is readily available, physicians will be able to give more accurate information to patients, and it will result in better and more timely treatment decisions.”

Yanming Zhang, MD, a cytogeneticist at Memorial Sloan Kettering Cancer Center in New York, NY, says that colleagues have been asking him about WGS-based assays and whether they could replace cytogenetics for clinical diagnostics in leukemia. He says no and believes the lower coverage of WGS compared with standard targeted sequencing could miss important mutations. In addition, it doesn't reveal clonal evolution in the cancers, which is crucial for tracking disease progression.

Logistical issues such as cost, needed equipment, and availability of analytic software will dictate how widespread WGS becomes, says Lucy Godley, MD, PhD, of the University of Chicago in Illinois. Whether it will supersede the standard methods is “hard to predict,” she says. “The paper shows it's technologically possible—and that technology is only going to get better.” –Mitch Leslie