In recent weeks, company representatives began informing physicians at cancer centers and hospitals about a problem with a specific variant in the MSH2 gene

Invitae Corporation (NYSE:NVTA), a genetic testing company in San Francisco, has told some physicians and clinicians in recent weeks that it has reported false-negative results. Clinical laboratory professionals with knowledge of the facts in this case believe the cause of the false negative results may have gone undetected for months and that genetic tests for a large number of patients may be involved.

For several weeks, Invitae has reported to its ordering physicians that it knows about a small number of false-negative reports that affect an estimated two to 10 patients who have a rare genetic variant in the MSH2 gene. The variant is known as the Boland Inversion and the gene is associated with Lynch syndrome (AKA, hereditary nonpolyposis colorectal cancer).

This episode may be a watershed event in the evolution of the genetic testing industry. Evidence indicates that genetic tests for a large number of patients were done incorrectly, and that the problem was systemic and went undetected by the lab company’s staff for as long as 11 months. Because these genetic tests were laboratory-developed tests (LDTs), the problem at Invitae could be used by some to argue that FDA regulation of LDTs is needed.

Invitae provided two written statements to The Dark Report, Dark Daily’s sister publication. The full statements can be read at the end of this story. The Dark Report is preparing a detailed intelligence briefing about this matter in its upcoming August 28 issue.

False Negative Reports for Some Genetic Tests

In one statement, the company wrote, “For the past several weeks Invitae has been working with clinicians to address an issue related to our analysis of a rare genetic variant in the MSH2 gene associated with Lynch syndrome (0.007% of inherited cancer tests), also known as the Boland inversion, which we believe could have led to a false negative report for a small number of patients (estimated 2-10 patients impacted).”

Invitae Corporation, founded in 2010, is a clinical laboratory company based in San Francisco that provides genetic testing services and has used aggressive pricing to fuel fast growth in specimen volume in recent years. According to the company’s 2017 second quarter earnings report, for the first six months of 2017 Invitae reported revenue of $27.7 million. Its net loss before taxes for the first six months of 2017 was $57.3 million. (Photo copyright: Yelp.)

After detecting the problem, the company began a root-cause analysis to determine the extent of the problem. “We would expect to observe the MSH2 Boland inversion event in 0.007% of patients undergoing hereditary cancer testing and approximately one in every 1,250 in patients with Lynch syndrome-spectrum cancers,” the company stated. “Based on these estimates, we expect this to impact a very small number of patients.”

Limited Number of Patients with False Negatives, but How Many Tests Involved?

The number of patients whose test results may have been affected is the subject of speculation among medical laboratory professionals who refer genetic tests to Invitae. Two medical directors at genetic testing laboratories pointed out that—based on the lab company’s estimate that false negatives were reported on just two to 10 patients—the problem could involve 3,000 to 12,500 patients.

The photo above taken Feb. 12, 2015, is of the Invitae management team at a happier time when the company’s shares began trading on the New York Stock Exchange (NYSE). (Photo copyright: Invitae.)

The company’s internal quality systems did not identify this problem. They learned about the problem from an outside source. Invitae said, “A client recently reported a discrepancy between an Invitae report and a report issued by another laboratory for the presence of a single rare mutation in MSH2, known as the Boland inversion. As soon as we learned of the discrepancy, we quickly identified and rectified the issue.”

The company confirmed that the problem with the Boland Inversion had gone undiscovered for 11 months, stating “We have identified all samples impacted by this issue, which were processed between September 2016 and July 2017. We are reanalyzing all previous results over the next several weeks to ensure their accuracy.”

Quality Control Checks for Omission of Assay Components

In its statement, Invitae referenced the quality-control issue, saying, “Because of the unique characteristics of how we were testing for the MSH2 Boland inversion, our quality control checks did not catch omission of the components of the assay. As soon as the omission was recognized and relevant components returned to the assay, it once again performed properly. We have added two separate quality controls to ensure this issue will not reoccur.”

The statement continued, saying, “We have identified all samples that could have been impacted by this error and are in the process of reanalyzing them free-of-charge.” The company also said, “Our ability to detect this specific MSH2 mutation has been fully restored and is functioning properly …. Moving forward, the new assay incorporates a quality check for successful capture and sequencing of the region around both ends of the Boland inversion so that the absence of the inversion gives a positive signal and the presence of the inversion gives a separate and different signal, while failure or absence of successful capture of these regions gives a third and different signal.”

To comply with federal and state clinical laboratory regulations, Invitae confirmed that it had conducted a root cause analysis (RCA) and was addressing the problem, notifying physicians and patients as necessary.

“We have been reaching out to all customers with patients who could have been impacted by this issue. We have samples to conduct reanalysis for all patients and will reach out to individual clinicians if any of those samples are deemed ‘quantity not sufficient’ (QNS) and new ones are required. However, the assay developed and validated for reanalysis is designed to use very small amounts of DNA, so we anticipate the number of new samples needed will be small.”

Retesting may be a greater challenge for Invitae than is indicated by its statement. Several pathologists told The Dark Report that such retesting comes with several problems. For example, certain states require patient samples used in genetic testing to be destroyed within a set time period. In such cases, the lab would need to work with the physician to have the patient provide a new sample for the retest. Also, it is common for genetic testing to use so much of a sample that the amount remaining is inadequate for a retest. In these cases, a new sample must be collected.

Another issue for Invitae will involve both the time to retest and the cost to retest. One lab executive pointed out that Invitae’s lab accessions had almost tripled from 12,500 in the second quarter of 2016 to 30,500 in second quarter of 2017. “Their lab is already straining to stay up with that volume increase. If Invitae must retest, say, 10,000 or more patient samples because of the MSH2 Boland inversion problem, this can seriously overload their lab and cause significant delays in turnaround time for all samples,” he explained.

A point of interest for pathologists and laboratory directors is whether any clinicians have filed a complaint or notified Invitae’s laboratory accreditation organizations, and the federal and state lab regulators, about the problems they had with this specific MSH2 mutation in their genetic tests and genetic test panels. Invitae has not commented on that situation.

Finally, the consequences for the patients whose genetic tests were performed by Invitae during this 11-month period should be considered. One executive from a large genetic testing lab in the Northeast said it best: “Every lab that performs genetic testing is in the rare mutation business!” he declared. “The mission is specifically to test for rare mutations and accurately identify the 1-in-1,000,000 mutation to enable that patient to get the right treatment that is invariably life-changing.

“Thus, for any lab like Invitae to tell its physicians that ‘only a few patients’ may have been given a false negative result from their genetic test betrays the quality and accuracy that all physicians, patients, and their families expect of our labs,” he continued. “Remember that what physicians and patients do with these results is very drastic! I consider it a massive failure anytime a genetic lab—whether large or small—misses rare mutations in even a small number of patients because of problems at the bench.”

—The Dark Report Editorial Team

Invitae provided two statements to The Dark Report. Here is the first statement, in full, dated Aug. 24 from Invitae, titled, “Statement on Boland Inversion Testing.”

“For the past several weeks Invitae has been working with clinicians to address an issue related to our analysis of a rare genetic variant in the MSH2 gene associated with Lynch syndrome (0.007% of inherited cancer tests), also known as the Boland inversion, which we believe could have led to a false negative report for a small number of patients (estimated 2-10 patients impacted).

“Our immediate priority has been getting accurate and actionable information to patients and clinicians about what happened and the steps we are taking to address the situation. We have identified all samples that could have been impacted by this error and are in the process of reanalyzing them free-of-charge. We have been personally reaching out to clinicians whose patients may have been impacted to discuss the issue, outlining what we have done to correct it and explaining the timeframe for receiving updated information. We are also offering no-cost family variant testing for the immediate families of any patient who tests positive for the Boland inversion variant, something we do for all our patients who test positive for a pathogenic variant.

“Our ability to detect this specific MSH2 mutation has been fully restored and is functioning properly. The corrected assay has been revalidated and shown to have regained its previous high sensitivity for the Boland inversion for all samples currently in the lab. Dual quality control checks specific to this issue have been added and are performing properly. We have also reviewed all of our other testing and are confident that this was an isolated error. Our ability to detect other MSH2 mutations or mutations in any other genes in our testing panels was not affected.

“Because of the unique characteristics of how we were testing for the MSH2 Boland inversion, our quality control checks did not catch omission of the components of the assay. As soon as the omission was recognized and relevant components returned to the assay, it once again performed properly. We have added two separate quality controls to ensure this issue will not reoccur.

“We take the reliability and validity of our test results extremely seriously. Nothing is more important to our company than ensuring that the information we provide is accurate. It is extremely rare that we find an error, but when we do we will quickly to correct it and share information with clinicians, in keeping with the medical community’s standards for error reporting.”

This is the second written statement, in full, provided to The Dark Report on Aug. 24. Invitae said:  

“A client recently reported a discrepancy between an Invitae report and a report issued by another laboratory for the presence of a single rare mutation in MSH2, known as the Boland inversion. As soon as we learned of the discrepancy, we quickly identified and rectified the issue.

“We have identified all samples impacted by this issue, which were processed between September 2016 and July 2017. We are reanalyzing all previous results over the next several weeks to ensure their accuracy.

“We would expect to observe the MSH2 Boland inversion event in 0.007% of patients undergoing hereditary cancer testing, and approximately one in every 1,250 in patients with Lynch syndrome-spectrum cancers. Based on these estimates, we expect this to impact a very small number of patients.

“Moving forward, the new assay incorporates a quality check for successful capture and sequencing of the region around both ends of the Boland inversion so that the absence of the inversion gives a positive signal and the presence of the inversion gives a separate and different signal, while failure or absence of successful capture of these regions gives a third and different signal.

“We have been reaching out to all customers with patients who could have been impacted by this issue. We have samples to conduct reanalysis for all patients and will reach out to individual clinicians if any of those samples are deemed ‘quantity not sufficient’ (QNS) and new ones are required. However, the assay developed and validated for reanalysis is designed to use very small amounts of DNA, so we anticipate the number of new samples needed will be small.”