The following post was written several years ago. Although more recent developments have changed the field of clinical laboratory science since the original posting, the information contained was deemed to be of historical interest.
The regulatory landscape for clinical genomic testing is reminiscent of a twentieth century American saying, “Almost only counts in horseshoes and hand grenades”. Since there are no specific requirements in CLIA’88 for genetic testing, there has always been a wide variety in opinion for what is required for CLIA accreditation when it comes to DNA sequencing. With the clinical roll out of clinical high throughput sequencing (a.k.a Next Generation Sequencing (NGS)), what has become apparent is that not all genetic tests are regulated the same. It appears that many clinical genomic tests can be described as almost validated.
Until the advent of clinical NGS, clinical genetic sequencing was expected to follow the usual route of clinical test validation. This process is familiar to most laboratorians in that it required many of the concepts covered by CLIA’88: validation of the analytical method of detection, setting limits of the assay, defining the stability of the assay, etc. These requirements are laborious for traditional sequence based testing because few steps in the process are amenable to automation. Test development for traditional sequencing assays often requires many months of time and considerable cost for adequate clinical validation. For rare genetic diseases or newly discovered genetic diseases it has been extremely difficult to have access to clinical sequencing tests that comply with Federal law. This has led to programs such as the NIH Office of Rare Diseases CETT (Collaboration, Education and Test Translation) Program. The CETT program fosters a limited amount of funding to translate research testing into the clinical laboratory. Unfortunately, Federal funding for this program has not been ideal, and there are few laboratories financially capable of developing clinical tests for every rare or newly discovered gene. Filling in the gap between clinically validated testing and research testing has been a grey market of custom tests designed and performed in CLIA laboratories. Thus, if a researcher discovers a new genetic mutation, the result can be resequenced on a custom basis in a CLIA laboratory environment. Although CLIA does not describe the use of research or non-validated tests in a CLIA environment, this has been the standard practice for rare genetic testing.
Clinical NGS has now embraced this concept of an accredited environment for custom non-validated testing for whole exome and whole genome analysis. One may observe that it is extremely difficult to validate sequencing on all ~200,000 exons prior to testing clinical specimens, but it is technically feasible and it would be required if the target exons were much fewer (e.g., 5 exons). There are not an infinite number of nucleotides in a genome and there is not an infinite number of coding regions of DNA. There is a large but finite number of target sequences. A clinical assay should not be represented as targeting ~200,000 exons plus or minus 10% (as I have been told by clinical NGS labs!). Instead these tests should have limits defined during the test development process. If your test only detects the coding regions of 10,240 of 22,000 human genes, that is the limit of the test. And that is what the physician needs to know before they order the test for their patients.
There are few quality standards for clinical NGS; the best national effort is a workgroup established in 2011 called Nex-StoCT (Next Generation Sequencing – Standardization of Clinical Testing). This workgroup is in its infancy and still identifying information that needs to be gathered, but what is curious is the acceptance of confirmation of genomic test results by a custom non-validated test. The proposed workflow would be: 1. patient has whole exome sequencing; 2. pathogenic sequence variation is identified; 3. custom sequencing test is performed for confirmation.
As laboratorians, we know that many things can go wrong for even the best validated tests. If test validation is abandoned, how will we know the quality of the results? How do we interpret these results?
Having clinical NGS and custom sequencing is an important part of rare disease diagnosis in the United States, and perhaps tests for these diseases should have more regulatory leeway. However, any genetic test provides life-changing information that will inform critical decisions regarding diagnosis, management and frequently future reproductive planning. If any testing needs regulation, genetic tests should be among the most important to be regulated.
The issue of clinically validated testing is not just an issue for laboratorians and physicians to discuss; patients now have direct access to information on genomic tests. Typing “gene sequence testing” into Google results in >200 million hits in 0.24 seconds. Many of these are “24-hour service” and “competitively priced” and will also include your ancestry along with your DNA sequences. The terms CAP and CLIA accredited are now tossed about the clinical NGS world as if they were gold seals of approval. Even at their best, CAP and CLIA accreditation should be regarded as minimal standards that we should always seek to exceed. And we should expect laboratories that claim accreditation to exceed these standards as well. A good starting point will be for clinical NGS laboratories to validate their tests and all of their analytes prior to testing patient samples.
Genomic medicine is still at its very beginning, but hard questions need to be asked and answered. The results of genomic tests have in general far greater implications than the results of common laboratory tests for which we now do extensive validation and regulation. Almost validating a test is not good enough for a sodium, it is not good enough for a TSH and it is certainly not good enough for clinical genomic testing. Almost only counts in horseshoes and hand grenades.