Editor’s Note: The following guest post is published with the permission of its author, Edward J. Schloss, MD, (Twitter ID @EJSMD) the medical director of cardiac electrophysiology at Christ Hospital in Cincinnati, OH. This post is longer and far more technical than most of the content published on CardioBrief. Due to the extraordinary nature of the material, however, I believe this post will likely be of considerable interest to many cardiovascular healthcare professionals.
More Lessons From the Riata ICD Lead Recall
by Edward J. Schloss, MD
In an earlier guest post (What are the Lessons of the Riata ICD Lead Recall?) I summarized points from Robert Hauser’s important perspective in the New England Journal of Medicine regarding the St. Jude Riata ICD lead recall. In this followup, I’d like to discuss St. Jude’s responses to this piece and place the arguments into a broader context.
Last week Mark Carlson MD, Chief Medical Officer of St. Jude Medical, posted a response to Hauser’s perspective, and he made similar points in a letter subsequently published in NEJM. In the original response, Carlson writes “we believe that there were inaccuracies and omissions in the editorial that are important to recognize in order to more fully understand this issue.”
Major points of disagreement in the two pieces include:
- Timing of the initiation of a prospective trial looking at Riata externalizations and failures.
- Similarities and differences between the currently marketed Durata lead and it’s recalled predecessor Riata ST.
- Robustness of the existing St. Jude lead surveillance systems.
Having looked at this issue in some depth, I feel it would be reasonable for me to weigh in on these matters.
Regarding the timing of the recent trial, Dr. Carlson writes:
The Perspective article states, “During the past year, neither St. Jude Medical nor the FDA has instituted the clinical studies that would be required to answer critical questions…” In fact, in December 2011, St. Jude Medical began enrolling in a 500-patient, multi-center, prospective Riata Lead Evaluation Study to further evaluate the incidence of externalized conductors and more importantly to evaluate the electrical performance of the Riata silicone-only leads that have externalized conductors over time.
Carlson is correct that such a trial has, in fact, been set up within the last year. However, having heard the comments at the summit, I suspect Hauser is expressing frustration that that trial was only set up in December 2011, well after the problem was identified. First reports of externalization began in 2008 and a “Dear Doctor” letter from St. Jude was issued in December 2010. It then took another year for the trial to get started. Hauser’s wording here, while literally incorrect, was accurate in the spirit that it was probably intended.
Point number 2 above is critically important and will require some explanation. There seems to be ongoing confusion regarding differences between Riata, Riata ST and Durata ICD leads. Hauser writes in NEJM:
St. Jude Medical is marketing the Durata ICD lead, which has an outer sleeve made of silicone and polyurethane but is otherwise similar in design to the Riata ST leads.
Note that in his rebuttal to this statement, Carlson refers instead to the prior generation Riata lead.
Regarding St. Jude Medical’s newer generation Durata lead with Optim coating, the article states, “St. Jude Medical is marketing the Durata ICD lead, which has an outer sleeve made of silicone and polyurethane but is otherwise similar to the Riata ST lead.” This is incorrect as the Durata lead incorporated substantial design changes from the prior generation Riata silicone-only leads that we believe significantly reduce the risk of externalized conductors and improve overall reliability.
Note also in the NEJM letter Carlson writes:
Assertion: The Durata lead is similar in design to the Riata silicone-only lead.
Again, here Carlson exchanges Riata for Riata ST. Nowhere in Hauser’s piece is a comparison drawn between Durata and Riata, as Carlson contends in both his writings.
Let me give a brief summary of recent St. Jude ICD lead design to show why this distinction is important.
For the purposes of clarity, I will refer to three families of St. Jude ICD leads – Riata, Riata ST and Durata. It is, of course, important to recognize that there are multiple leads within each of these families.
First generation Riata leads were approved in 2001. They are often classified as 8 Fr leads based on the minimum introducer size, but are actually 6.7 Fr in diameter for most models with the largest lead at 7.6 Fr (model 1582). All Riata leads are designed with a silicone core structure with four or three lumens. A central lumen contains the distal pace/sense electrode surrounded by three or two lumens (for dual and single HV coil leads respectively). Within each of these lumens are two conductor cables. These cables serve as either the anodal low voltage electrode, high voltage electrode or ‘filler” (in the case of single coil designs). Although St. Jude refers to these leads as “silicone only,” it should be noted that the central wound filament and all of the cables are surrounded by PTFE or ETFE insulation. Riata cross sectional design is diagrammed on the left side of the figure below (reproduced with permission by St. Jude Medical):
Riata ST, the lead on the right side of this diagram, was released in 2006. Note that these are the two leads currently on FDA recall. Riata ST design includes a downsized lead body of 6.3 Fr, allowing it to fit through a 7 Fr. introducer. This was accomplished, as discussed in the diagram, by moving the cables closer to the center of the lead body. There was no change in the thickness of the outer silicone insulation. Additional design changes included the use of flat wire shocking electrodes, backfilled with silicone (not shown in the diagram above).
St. Jude has asserted that these changes from Riata to Riata ST should reduce the likelihood of inside-out lead abrasion failures. The movement of the coils centrally is said to reduce cable tension and shear stress (per diagram above). In addition, the flat wound shock cables are said to reduce likelihood of “internal shorting” from abrasion underneath the shocking coils (as discussed by Carlson at the Riata Summit).
Let’s now look at the evolution to Durata as diagrammed in the next diagram below (reproduced with permission of St. Jude Medical). Recall that the Durata lead is currently market released and not a subject of any recalls or confirmed externalizations. These are the two leads Hauser wrote were “similar in design” apart from the Optim insulation.
Durata was released in 2008. As diagrammed above, its central core structure remains unchanged from Riata ST. A layer of Optim insulation (copolymer of silicone and polyurethane) is added to the outer jacket of the lead in areas without shocking coils. This increased the lead diameter to 6.8 Fr and it still fits through a 7 Fr introducer. Two other design elements were added specifically to decrease perforation risk without expected impact on inside-out abrasion or externalization (soft tip and pre-curved RV shock coil).
Optim insulation has been shown to improve abrasion resistance as compared to silicone. St. Jude has expressed optimism that this will prevent externalization of the inner cables and at this point, no inside-out externalizations have been reported with Durata. Note, however, that the Durata lead central core still has the same silicone structure as Riata ST, and any inside-out abrasion up until breach of the external Optim jacket might occur in the same manner. This raises the possibility that Durata and Riata ST could share a similar failure mechanism of inside-out abrasion without externalization (such as might occur under the shocking coils). Inside-out abrasion lead failure has already been reported with Durata in the MAUDE database.
At the Summit and in his web statement, Dr. Carlson presented a simplified slide that condenses the two slides already cited. It is reproduced below with the permission of St. Jude:
Note that in this slide, Riata and Riata ST are grouped together on the left and all the listed design changes are diagrammed in the change to Durata. The slide certainly appears to show that all of these changes occurred only as we moved to Durata. In fact, two of these changes (movement of the conductor cables centrally and flat wire shock coils) had already occurred in the evolution to Riata ST, though you wouldn’t know this from the diagram. Seeing this diagram alone, one might form the impression that Durata is more “different” from the recalled leads than it really is.
Hauser specifically spoke in his perspective of the similarities between Riata ST and Durata. Carlson, on the other hand responded in both his pieces with a comparison between the original Riata and Durata. If one examines the design of these leads, I think it’s apparent that Hauser’s statement was accurate and Carlson’s rebuttal was misplaced.
The third point of disagreement cited above between Hauser and St. Jude pertained to the quality of St. Jude’s lead performance data.
Hauser cites the Riata/Riata ST recall as a failure of postmarketing surveillance. He’s critical of St. Jude on the lack of clinical data guiding current recommendations of how to deal with externalized leads. Detection of this problem and assessment of its clinical significance has been difficult and recommendations have been made based on educated consensus, rather than ongoing prospective clinical performance data.
Although St. Jude has cited three ongoing active registries for their Optim insulated leads including Durata, most of what we know about Riata and Riata ST comes from passive reporting.
At the Riata Summit, several studies were presented showing longitudinal performance of these leads from St. Jude and several academic centers. All of this information is available at the Minneapolis Heart Institute Foundation website (free registration required).
There were significant disagreements between St. Jude’s data and the non-industry studies. Some of the biggest divergences centered on externalization. St. Jude reported in their November 2011 communication (and on the slide above) an externalization rate of 0.10% for Riata and Riata ST based on complaints and return analysis. In stark contrast, prospective fluoro series from Lau in Ireland showed 15% externalization in a population with high numbers of single coil Riata leads. Ellis at Vanderbilt presented their series with 26% externalization. In addition, the Ellis series showed equal externalization rates in Riata and Riata ST and had a higher proportion of dual coil leads. Last month, a Swiss group published a series of 52 Riata and Riata ST leads that underwent prospective fluoroscopy with an externalization rate of 11.5%. I expect we’ll see more systematic fluoroscopic lead screening series in the near future.
Based on this data now reported, it appears that St. Jude significantly underestimated the incidence of externalization in both Riata and Riata ST, likely due to their reliance upon passive monitoring. In addition, the Vanderbilt data calls into question the effectiveness of Riata ST design mitigations to prevent externalization. Indeed, all Riata silicone leads may have a significant externalization problem.
At this point, there have been no confirmed externalizations of Durata leads due to inside out abrasion from any source. At the summit, one case of Durata externalization was presented, but St. Jude and independent engineering analysis supported a mechanism of outside-in abrasion due to a tricuspid valve ring.
One theme that emerged from the Summit was that conductor externalization and electrical performance are not clearly tied. If fact most externalized leads in all reported experiences seem to function normally. St. Jude presented bench data showing good performance of externalized ETFE conductor cables for sensing, pacing and shock delivery.
Detection of high voltage cable dysfunction, however, may be particularly challenging. Recall, that older generation St. Jude ICDs (Epic/Atlas) do not have out of clinic HV lead impedance daily measurements and therefore would not pick up dysfunction during routine follow-up. In addition, some have wondered whether ICD generator HV lead impedance testing would reliably predict response to a high voltage shock. Indeed a case report in PACE this month discussed a patient with normal Riata lead parameters exhibiting shorting of the defibrillation circuit at elective generator change. This resulted in unsuccessful defibrillation and generator battery depletion. At the summit, there was a lot of discussion about whether there is a need for performing defibrillation threshold testing to screen for HV cable dysfunction in leads with externalized cables, but no consensus was reached.
Long-term clinical performance of externalized conductors remains an unknown, but will be assessed in the prospective Riata Lead Evaluation Study started by St. Jude in December 2011. This protocol will include prospective office lead measurements, fluoroscopy and reporting of adverse events.
Given the relative lack of prospective data, it’s difficult to draw conclusions about the overall long-term electrical performance of Riata and Riata ST. St. Jude’s product performance data shows long-term survival based on passive monitoring ranging from 95.8% to 99.2% (November 11 Product Performance Report).
There are at least two published studies looking at overall performance of Riata and Riata ST. These were not discussed at the Summit nor are they linked on riatacommunications.com. Epstein et al accumulated data on over 7000 Riata leads from several St. Jude sponsored registries over a median 22 months. Porterfield et al did a retrospective review of Riata performance from chart review of over 15,000 patients at twenty-eight centers over mean 18 months. Both of these studies defined mechanical failure as those requiring revision, extraction or replacement. Results from both studies showed failure rates less than 1%.
There were two large retrospective lead performance studies presented at the Summit, by Keung of the VA Experience and Abdelhadi of the Riata Independent Multicenter Study. Riata/Riata ST performance was reported to be somewhere between high performing leads such as Medtronic Quattro and poor performing leads such as Medtronic Fidelis. Failure of these leads appears to accelerate at around 4.5 years and lead numbers at this point are low.
Durata has been followed in three St Jude active registries: OPTIMUM, SCORE, and SJ4 Post Approval Study that include over 10,000 patients. St Jude reported results have shown 0.09% all cause mechanical failure, 0.02% all cause insulation failure and 0.00% externalized leads.
I think Carlson has a valid point when he disputes Hauser’s statement:
we do not have a surveillance system in place that can detect low-frequency failures or adverse clinical events involving Durata leads or any other marketed medical device.
There are actually three such active registries currently serving that purpose. I do think however, it’s fair to say that the active and retrospective data on Riata and Riata ST until recently has been limited. It certainly was not able to predict the externalization problem, and answers about long-term performance of these leads are only now emerging.
Anyone who has been in the business of cardiac rhythm management has struggled with the problems of device failures. Because most of these only occur years after FDA approval, having a system for early detection of excess or unique failures is critical.
Industry response to these failures is also critical. Clear, honest acknowledgement of the problem is essential. We also need quick investigation to define the scope of the problem and create an action plan. These measures go a long way to maintaining the confidence of all interested parties, including physicians, hospitals, patients, and the investment community.
Every time we, the implanting physicians, implant a medical device, we are taking a measured risk that must be outweighed by its benefits. Understanding this true risk, especially in newer devices, is extremely difficult. By becoming a student of the engineering process of these devices, we can become better equipped to make good decisions for our patients.
We are only now learning the scope of the problem with Riata and Riata ST. The lead mechanical failures and externalizations of this lead may only first occur four to five years after implant. Since Durata leads are only now approaching that interval, it is not surprising that we have not seen many troubling signals with this lead. It remains to be seen what the future holds. Until we have longer experience and collect more data, it us up to individual doctors to educate themselves with available clinical and engineering data so they are equipped to make good decisions.
Dr. Schloss acknowledges the assistance of Westby Fisher MD, John Mandrola MD, and St Jude Medical Engineering in the preparation of this manuscript. Dr. Schloss has research relationships with Biotronik, Boston Scientific, Medtronic and St. Jude Medical. He has consulting relationships with Boston Scientific and Medtronic. The views expressed here are those of Dr. Schloss and do not necessarily represent those of The Christ Hospital.