A worldwide survey of labs performing BRCA1 and BRCA2 testing found that while nearly all labs use massively parallel sequencing (MPS) platforms to identify genetic alterations in breast and ovar-ian cancer genes, their procedures and practices “differed widely” on six other aspects of testing (npj Genomic Medicine 2018;3: doi:10.1038/s41525-018-0046-7). 

Responding labs varied in their quality criteria for average and minimum read depths, and report-ed variants of unknown significance (VUS) at different rates and calculated VUS rates differently. They had starkly different testing volumes, and followed different guidelines and involved different types of professionals to classify variants. Participants also varied in whether they shared sequence results with public databases.

Based on these findings, the authors called for international efforts to set quality standards for BRCA1/2 testing. “Global best-practice guidelines for BRCA testing are needed to ensure consisten-cy in testing for patients, regardless of where they obtain their testing,” said first author Amanda To-land, PhD, an associate professor of cancer biology and genetics at The Ohio State University, in a separate statement.

In all, 86 laboratories responded to the 65-question online survey, eight based in the U.S., and 78 from other countries. Nearly all the labs (93%) reported using MPS to identify variants; six reported using Sanger sequencing as their only method, while 27 (35%) indicated that they use Sanger most-ly to confirm variants identified by other methods.

All respondents covered coding exons fully, but few—none in the U.S. and 9% in other coun-tries—reported performing full intronic sequencing. Slightly more than a quarter of U.S.- and non-U.S.-based labs reported confirming by another method only pathogenic or likely pathogenic vari-ants; whereas 71% of U.S.- and 56% of non-U.S.-based labs indicated that they confirm by another method all VUS, pathogenic, or likely pathogenic variants. Slightly more than half of non-U.S. labs had not specifically calculated their BRCA1/2 VUS rates; among those that had, the rate varied from 3%-50%. Five U.S. labs had calculated BRCA1/2 VUS rates, ranging from less than 2% to 6%.

Not all labs reported their sequence read depths, but among the 37 non-U.S. labs that did so, the median and average BRCA1/2 read depths were 100 and 484, respectively. In contrast, U.S. labs re-ported median and average read depths of 500 and 820, respectively.