Update on Riata ICD lead failures--Patient deaths

Apr 2, 2012 11:50 EDT

I feel compelled to update my recent Trials and Fibrillations post on the St Jude Riata implantable cardioverter-defibrillator (ICD) lead recall. New and very distressing details about patients who died as a result of lead failures were published recently in the journal Heart Rhythm. In my previous post, I emphasized the fact that no patient had died directly as a result of a Riata lead failure. That's changed, and so have some of my views on this matter.

A brief review is in order. The concept behind using ICDs is simple. They are supposed to sense the heart rhythm and deliver a high-voltage shock in the event of a malignant rhythm. In the ideal word where ICDs are implanted in a narrowly defined group of patients sick enough to have a high risk of VF, but not so sick that they are dying of competing causes, these lifesaving devices epitomize beautiful therapy. Randomized controlled trials, along with clinical experience, have shown ICDs extend life. What's more, when combined with an LV lead in those with left bundle branch block, cardiac resynchronization therapy defibrillators (CRT-Ds) greatly improve patients' quality of life. But accomplishing all this doesn't come without risk. Among other downsides, ICD therapy requires implanting complicated hardware permanently in the body, and then these devices must withstand the rigors of 100 000 daily heartbeats over the course of years.

Let's move to the recent St Jude Riata lead situation.

Before 2010, about 227 000 St Jude Riata (and the newer-generation Riata ST) ICD leads were implanted worldwide. In the US, 79 000 of these leads remain active. Most important for practicing cardiologists, St Jude's newest-generation ICD lead, Durata, which shares similar design features to the recalled Riata leads, is being implanted in patients now. It's a matter of vigorous debate exactly how similar Durata is to the recalled Riata family of leads and how durable Durata will prove to be going forward.

Prior studies and media reports on the family of Riata leads focused primarily on the more visually impressive problem with these leads—namely, that inner-conducting cables were eroding through their outer insulation and becoming exposed in the body. Indeed, externalization is surely a problem; but arguably a larger issue is the high rate of electrical failures observed with Riata leads. For if an ICD lead fails to sense, pace, or defibrillate, the patient may die.

Enter Dr Hauser's recent report in the Heart Rhythm journal. In this distressing report, he and coworkers culled the MAUDE database and found 20 cases in which patients died as a direct result of Riata lead failure. In addition, they also report on 25 other "indeterminate" deaths where lead failure was suspected but clinical data were lacking. Nearly all of the deaths occurred as a result of failure of the shocking (high-voltage) portion of the lead. Insulation abrasions resulted in internal wires touching each other, other leads, or the defibrillator can. Electrical shorts then were the dominant mechanism of shock failure.

As well outlined by my friend and fellow real-world clinician, Dr Edward Schloss (published on Cardiobrief), the details of these failures are terribly distressing. My top three concerns are:

• The first manifestation of Riata failure may be patient death. The short circuit may not manifest itself until the patient needs a shock. In some of the cases described in the Hauser report, the high-voltage shock itself caused the short; the current wasn't effectively transferred to the patient; VF persisted and the patient died.
• The act of testing for HV failure may be hazardous. At least one of the 20 patients described in the report died as a result of VF induced by testing. The details in this case aren't provided, but the possibility that the device directly induced death deeply troubles me. For the most sacred of rules in medicine holds that we first, do no harm.
• Procedure-related deaths were reported. Although Dr Hauser's study did not consider procedure-related deaths as directly lead-related, it's sobering to note that 16 deaths (eight due to perforations and eight related to extractions) were found in the database. Considering the voluntary nature of MAUDE reporting, this figure likely underestimates the number of procedure-related deaths, and it's certainly plausible that Riata-related procedures caused as many deaths as did high-voltage failures.

 

My conclusions:

I had hoped the Riata situation wasn't going to be as bad as Medtronic's Sprint Fidelis recall. Not so. These troubling new findings coupled with the inability to prospectively identify affected leads suggest Riata has indeed risen to the level of infamy reached by Fidelis. Also, the fact that the currently available Durata leads share many of the faulty design elements of Riata concerns me greatly.

The imperfectness of medical devices, sometimes realized long after implantation, needs to be added to the list of complications discussed with patients considered for ICDs. Doctors too often give short shrift to the potential harm from our interventions. Heart doctors and patients alike must know that ICDs should not be likened to insurance policies against death. An insurance policy creates no risk to the policyholder. And of course there is no reliable insurance against death.

Witnessing and being forced to work through such misadventures of the fury of modern medicine helps explain why "seasoned" doctors often evolve gradually, but steadily, toward therapeutic minimalism.

On the other hand, I don't share the "Naderesque" view of Dr Steve Nissen, who boldly told CBS News of “a consistent pattern of failure of medical devices." Cited in this CBS piece was a report that Nissen coauthored in 2006 that found more than 2800 patients died because of faulty devices. This pessimistic view is one way to look at the problem of imperfect medical devices. Dr Hauser and coworkers' first sentence in their concluding remarks of the Heart Rhythm piece echo this negative sentiment:

Our analysis of the MAUDE database suggests that Riata and Riata ST leads are prone to high-voltage failure, which caused the deaths of 20 patients in this study."

I respectfully disagree with both this conclusion and Dr Nissen's pessimism. Riata leads did not cause these patients to die. Ventricular fibrillation due to advanced heart disease caused their death; the Riata lead failed to terminate the arrhythmia. In other words, the therapy wasn't 100% effective. Although Riata (and Fidelis) lead failures pose vexing clinical problems, if one considers the small numbers of failures against the many who have benefited from ICDs, the overall results look quite favorable. For Riata, we have 20 deaths paired against 122 000 implants. That 99% nonfailure rate looks much better than cardiac-arrest therapy without an ICD, which is nearly universally fatal.

In the treatment of sudden cardiac death, there is much work to do. Clearly, we need better postimplant surveillance of medical devices. Although it's disquieting to witness human fallibility in the practice of medicine, the upside of mistakes comes from how much they teach us. In learning from our mistakes, we must strive to prevent the stifling of innovation that comes from fear, excessive premarket regulation, and pessimism.

We must not be afraid of our humanness. To slow innovation would truly be a setback for our patients.

JMM