A panoramic view of the accuracy of molecular genetic testing

Main

With the significant advancements in molecular diagnostic capabilities over the past several decades, it has become standard practice for a growing number of genetic diagnoses to be made on the basis of molecular genetic test results. Indeed, given the common occurrence of phenotypic overlap among genetic diseases, molecular testing may often be the only way to reach an accurate clinical diagnosis. Moreover, establishing an underlying molecular etiology increasingly influences how genetic conditions are managed.

Our ability to make accurate genetic diagnoses using molecular methods is only as good as the diagnostic tools that are currently available. These continue to improve and drive advances in our field. Analysis of the landscape of laboratory performance in the context of molecular genetic testing is important to inform the clinical and laboratory communities regarding what their expectations should be in terms of analytic performance of different methodologies. It can also identify areas where improvements in standardization, methodology, or interpretation guidelines are needed.

The paper by Richards et al., “Results From an External Proficiency Testing Program: Eleven Years of Molecular Genetics Testing for Myotonic Dystrophy Type 1,”¹ is the latest in a series of articles in this journal summarizing the performance of laboratories engaged in the College of American Pathologists (CAP) proficiency testing (PT) surveys.^{1,2,3,4,5,6,7,8} In each of these articles, approximately 10 years of PT data were analyzed for overall performance by molecular genetics laboratories and for global performance over time; these data were not provided to individual participating laboratories and are not available outside of these publications. It is not feasible to publish all PT survey data at once because there are so many data points to analyze and there are different aspects that are important to address for each clinical disease/test. In this commentary, we provide a summary of the overall landscape of laboratory performance for molecular genetic testing.

For clinical laboratories certified under the Clinical Laboratory Improvement Amendments of 1988 and accredited by the CAP, external PT is required. In the realm of inherited genetic diseases, the CAP offers analyte-specific PT surveys that are targeted to specific genes or mutations for many molecular genetics tests, in addition to methods-based PT for Sanger dideoxy sequencing and next-generation sequencing for germline variants. Methods-based PT is an external quality assessment approach based on methodology rather than targeted to individual analytes. It is especially helpful for large-scale genetic testing because it enables laboratories to assess their performance in detecting the range of mutation types and to closely monitor overall technical performance of evolving sequencing approaches. With the growing number of tests offered for rare genetic conditions, a combination of methods-based PT for genomic sequencing assays and, when available, analyte-specific PT offers thorough and efficient quality assessment and helps ensure optimal preanalytic, analytic, and postanalytic laboratory performance.⁹ The joint CAP/ACMG Biochemical and Molecular Genetics (BMG) committee oversees the CAP PT surveys for molecular genetic testing and has worked to summarize the longitudinal data from the PT surveys for seven of the most commonly tested genetic diseases and for methods-based PT for Sanger sequencing.^{1,2,3,4,5,6,7,8} The data from individual rounds of PT, which are offered two times per year, are not made publicly available by CAP, except to laboratories enrolled in their PT programs.

The surveys analyzed by the CAP/ACMG BMG committee encompass a variety of molecular methodologies used for single mutation testing, mutation panels, and three different trinucleotide repeat disorders. Overall, there has been excellent performance by clinical laboratories participating in the CAP molecular genetics PT programs, with >95% analytical sensitivity and >99% specificity for all tests analyzed by the CAP/ACMG BMG committee ( Table 1 ). This compares quite favorably with analytical accuracy parameters for other laboratory analytes (as summarized elsewhere⁸). This is encouraging because the educational nature of these surveys tends to include particularly challenging genotypes at a much higher frequency than would be expected to be observed in a clinical molecular diagnostic laboratory. There was more variance in the clinical interpretation of results, indicating that there may be a need for improvement or guidelines in this area, although it is noted that the limited clinical description and the multiple-choice format of the PT surveys may not completely assess interpretation by laboratories in their clinical reports. In addition, there was consistently better performance of US laboratories compared to non-US laboratories. Similarly, during the first 3 years of a methods-based survey for Sanger sequencing, in which laboratories were asked to analyze electropherograms and provide variant interpretations, 98.3% of US laboratories had acceptable performance compared to 88.9% for international participants.⁴ Finally, there was consistent demonstration of improvement over time for both analytical accuracy and clinical interpretation, which could be due to improvement in methodologies and/or to participation in the PT program, which can provide valuable feedback to laboratories regarding test accuracy and interpretation.

Table 1 Summary of analytical performance on the College of American Pathologists molecular genetics proficiency surveys

It is worth noting that virtually all of the molecular genetics assays used by laboratories participating in these surveys are laboratory-developed tests or procedures, because there are almost no US Food and Drug Administration (FDA)-approved tests available for molecular genetic conditions. The data from the CAP PT program indicate that molecular genetic testing using laboratory-developed tests/laboratory-developed procedures is quite accurate. This is critically important information in light of the recent FDA draft guidance for regulatory oversight of laboratory-developed tests.¹⁰ For most molecular genetic tests, laboratories develop their own analytical assays using standard molecular methods to test for a clinical indication that is either considered standard of care or has documented clinical validity. The molecular laboratorians who oversee this testing are well-trained medical professionals who are expert and nimble at developing and validating molecular tests for clinical use. The results of the CAP PT surveys clearly indicate that the majority of diagnostic molecular genetics laboratories have been accomplishing this quite capably for many years and that clinical molecular genetic testing is safe and accurate. This bodes well for the genetics community and, more importantly, for the patients and families it serves.

Disclosure

K.E.W. and I.S. are the present and past chairs of the CAP Biochemical and Molecular Genetics Committee.