What A Mess

The single biggest problem in communication is the illusion that it has taken place.   – Apocryphal quote, likely incorrectly attributed to George Bernard Shaw

A South Carolina court recently granted a summary judgment in favor of the defendant in the case of Williams v. Quest Diagnostics, Inc., Athena Diagnostics, Inc,  ADI Diagnostics, Inc.  The ruling is particularly relevant to the genetics profession because it concerns the potential legal implications of the classification of genetic variants.

This legal odyssey began nearly 5 years ago but the clinical story began 14 years ago, and was first reported here on The DNA Exchange. Briefly, Amy Williams, the plaintiff, filed a suit in 2016 on behalf of her deceased son alleging negligence on the part of Athena (now owned by Quest) when in 2007 it classified a variant in the SCN1A gene as a Variant of Uncertain Significance (VUS) in her son who had a seizure disorder. SCN1A pathogenic variants are diagnostic for Dravet syndrome.

Multiple specialists were involved in the child’s care, but it seems from the medical records that none of them were actually aware of the genetic test results. The ordering physician could not recall having seen the report and the treating physicians never received a copy of the report or a communication from the ordering physician about the result, even though a copy of the report is in the medical records. Consequently, her son’s treating physician kept him on carbamazepine, a sodium channel blocker that is contra-indicated in children with Dravet syndrome. Sadly, he died about 6 months later, likely due to the contraindicated medication. Ms. Williams did not find out about the SCN1A result until nearly 7 years after the report was issued, and then only after a genetic counselor who was sifting through the records found a note from 2008 referring to an SCN1A VUS (for a fuller description of this saga, I refer you to the excellent articles written by Turna Ray, a journalist for Genome Web).

The lab’s defense rested on a legal technicality of the statute of limitations. Per South Carolina law, litigation cannot be brought against a healthcare provider if the offense took place more than 3 years prior to the filing. The lab’s lawyers argued, and in 2018 the South Carolina Supreme Court agreed, that a lab qualifies as a healthcare provider under state law. Subsequently, Judge Margaret Seymour, the judge who presided over the original case (and who displayed an excellent grasp of the genetic and legal issues), found that several of the plaintiff’s claims were “comprised of allegations sounding in both medical malpractice and ordinary negligence” and allowed the matter to move to discovery for the purpose of determining what caused Athena’s laboratory staff to misclassify the gene variant. Ms. Williams and her lawyers proceeded with the case based on “claims for wrongful death, survival, negligent misrepresentation, constructive fraud, and violation of the South Carolina Unfair Trade Practices Act.” Following discovery, the defendants requested a summary judgment to dismiss the case (in a summary judgment, either a plaintiff or a defendant can assert that the facts in the case are not in question and ask the judge to make a decision on the case without a full trial). 

Judge Seymour based her decision on the statute of limitations ruling by the South Carolina Supreme Court and the likely inability to prove proximate cause (i.e., that the VUS classification led to continued treatment with carbamazepine which then caused the child’s death), and dismissed the case: “The court concludes that no reasonable jury could find Defendants erred in classifying Decedent’s variant as a VUS, or that any misclassification was the result of nonmedical, administrative, ministerial, or routine care. Defendant’s motion for summary judgment is granted as to this issue.”

In an unfortunately cruel twist, Ms. Williams and her lawyers may owe Quest and its lawyers ~$140,000 in court sanctions imposed after some personal emails and other documents that, in my view likely had little bearing on the facts of the case, were deleted or improperly withheld because of less than stellar legal representation. The defendant’s lawyers offered to drop the costs stemming from sanctions if Ms. Williams agreed not to discuss the case in public forums and to discourage others from doing so. However, she did not agree to the proposal and remains firm in her belief that her son’s story needs to be discussed in public for the benefit of the public and the genetics profession, even in the face of potential financially ruin.

Was Athena’s original variant classification appropriate? Published case reports at the time the interpretation was first issued suggested that the SCN1A variant could be likely pathogenic (that was not the common terminology at the time), one of which was co-authored by Athena staff. Yet comments in the test report state there is an “absence of published studies correlating these variant(s) with clinical presentation and/or pathology.” In April of 2009, 2 years after the report was issued, Athena reclassified the variant as pathogenic, although they cited no new evidence beyond what was available when the report was first issued. Specialists in variant classification, who understand the intricacies of variant classification far better than I do, have weighed in and most have argued that Athena’s original classification of a VUS was appropriate for the knowledge available in 2007. Currently there are two entries for this variant in ClinVar, neither of which make an attempt at classification. I am not about to get into a debate with good scientists who know a heck of a lot more than I do about variant classification. I will say this, though, as someone who orders genetic testing every day, I rely heavily on labs to interpret variants and to let me know when there is in a result that might be grayer than ordinary. Especially in a case where a treatment decision with life and death implications hinges on a test result, I would expect the lab to explain their justification for the interpretation and to have made it clear in writing in the report. A phone call to the ordering provider wouldn’t hurt either, to be sure that the critical information and any uncertainty is clearly communicated.

Just as egregious, Ms. Williams should not have learned of a genetic test result almost 7 years after it was issued, and then only almost incidentally. Nor does it appear that the physicians who cared for the child were aware of the updated classification or communicated it to one another – not surprising, given that they were apparently unaware of the original report. Had she been notified in a timely manner, she may have initiated a discussion of why the variant was classified this way and if the evidence was strong enough to be the basis for treatment decisions. Of course this burden should not be on the patient but it could have offered another opportunity for further exploring treatment decisions based on the result. Many of us in the medical field, including me, have been critical of the requirement of the 21st Century CURES Act to notify patients of test results within 24 hours of when they are ready. My grumbling aside, Amy Williams and her son would have greatly benefited from being notified of her son’s result 24 hours after it were available. Does it have to take an Act of Congress to ensure that healthcare providers are responsible communicators with their patients?

In my view, nothing good came of this case in terms of the reputation of the genetics community, though of course nowhere near as bad as the devastating effects for Amy Williams and her son. To resurrect the line from the Captain’s speech in the movie “Cool Hand Luke“, what we have here is failure to communicate. Basically, a child may have died prematurely because of poor communication between the lab and care providers, between care providers, between care providers and the patient’s mother, and between the lab and the patient’s mother. Everybody lost and nobody won, even if Athena/Quest won from the perspective of not having to pay damages.

Will we now become better at communicating results to patients? Perhaps the CURES Act will help some. But as genetic testing expands well beyond the genetics community, communication about the implications of test results will likely still be deficient in many instances, in part because many non-specialists who order genetic tests are not particularly adept at interpreting them. Furthermore, although it’s hats’ off to ClinVar and other collaborative efforts for classify variants, variant classification will continue to be an Achilles’ heel of genome analysis because there is just no profit in it and it can be so damned complicated. 

I look back on this story and feel a knot in my stomach.