Contributed Commentary by John Milligan, PhD, and Ninad Pendse, Asuragen
November 21, 2023 | Cystic fibrosis (CF) is one of the few genetic disorders to be considered both serious and common enough to merit universal carrier or prenatal screening. Unfortunately, though, current testing approaches are caught between evolution and revolution even as they continue to produce suboptimal results for people of non-European ancestry.
The American College of Medical Genetics and Genomics (ACMG) has worked hard to overcome these problems, but their changing guidance can itself be a burden for clinical laboratory teams. Most recently, ACMG updated its recommendation of which variants should be covered in routine CF screening protocols in the hopes of reaching 95% detection of carriers in all ancestries. The shift from a 23-variant panel to a 100-variant panel is well intentioned, to be sure, but it increases the burden on clinical labs to develop and validate such a large panel test, particularly since it limits the use of more accessible, streamlined technologies.
Worse still is that expanding to 100 variants—or even the larger panels offered by some labs—does not achieve the equitable results ACMG and the entire clinical laboratory community are aiming for. Variants that are more prevalent in non-European populations are still missed with this approach. That’s because these expanded panels continue to be based on the well-established CFTR2 database, which is populated almost entirely with data from people of European descent, the group in which CF is most prevalent. It’s also because variants of varying clinical consequence (VVCC) were excluded from ACMG recommendations categorically without consideration of available clinical evidence in CFTR2, functional studies, or other sources that could inform the association of VVCC with classical or nonclassical CF phenotypes. These VVCC tend to be more common in non-European populations, probably in part because genotype-phenotype associations have been less studied in those groups.
The clinical laboratory community can and should do better. Fortunately, there is a clear path to achieving our shared goals of more equitable and accurate testing for all patients. We need to go beyond CFTR2 and improve coverage for all ancestries by embracing the vast amounts of data generated through large-scale sequencing studies.
Variants discovered through population-scale studies have been summarized in ClinVar and gnomAD, two invaluable repositories that now contain more reported CF variants for a broader range of ancestries than traditional sources such as CFTR2. ClinVar is particularly useful for adjudicating pathogenicity for variants—especially the VVCC omitted from the ACMG recommendations—while gnomAD can help scientists estimate carrier frequencies for each variant among various genetically-defined ancestries without relying on self-reported ethnicity.
There is substantial evidence to support many of these variants for specific populations, and there is no reason to continue excluding them from CF testing panels. By integrating this information from ClinVar and gnomAD, it is possible to identify the variants that are most important for each ancestral group so they can be included in a CF testing panel that’s more effective for patients of all ancestries.
In an internal analysis recently presented at the annual meeting for the Association for Molecular Pathology, for example, we found that VVCC with strong pathogenic evidence represent around 24% of CF-associated allele frequency for people of African descent, 39% to 45% for people of Asian descent, and 16% for Americans who describe themselves as having Latino or admixed heritage. The ideal screening panel may include a number of the VVCC that are currently not represented in ACMG guidelines, but capture a significant amount of CF carrier risk among global populations.
While selecting these variants might seem just as onerous as developing and validating the 100-variant panel recommended by ACMG, in fact, it offers a much more straightforward approach. In our deep dive into CF-relevant variants reported in ClinVar and gnomAD, we found that incorporating just a few common VVCC for each ancestral group is sufficient to produce more reliable and equitable results for all patients compared to ACMG’s list of 100 variants. Our study identified individual variants that represented a substantial portion of CF risk in any population. Among East Asians and South Asians, for instance, a single VVCC in each population captures about 40% of that group’s risk for CF. Among people of African descent, the most prevalent VVCC covers 12% of CF-associated allele frequency.
Indeed, it appears that testing between 60 and 70 carefully selected variants can improve coverage and results for all patients while giving labs more options to manage the needs of their testing population, their operations, and their costs. We applaud the progress ACMG has made in pushing CF testing forward, but now it is time for the clinical lab community to broaden our view of pathogenic CF variants and establish new recommendations that serve all populations. Adding more variants is not always the answer; choosing the most representative variants could improve patient results while easing the burden on clinical labs.
John Milligan, PhD, is the director of diagnostics development at Asuragen, a Bio-Techne brand. He can be reached at John.Milligan@bio-techne.com. Ninad Pendse serves as a senior product manager at Asuragen. He can be reached at Ninad.Pendse@bio-techne.com.