Trade-offs Between Subcutaneous, Transvenous ICDs on Display in EFFORTLESS

April 30, 2021

One of the largest and longest follow-ups of patients implanted with the subcutaneous-lead implantable cardioverter defibrillator (S-ICD), now marketed as the EMBLEM S-ICD (Boston Scientific), paints a picture of solid performance that arguably has improved over the years.

The S-ICD has held on to its reputation as a niche device after more than a decade on the market, probably in large part owing to its limited capabilities compared with more fully featured implantable cardioverter defibrillators (ICDs) with transvenous leads, which remain the gold standard.

But experience and design and software upgrades over the years may have diminished the clinical impact of some of those S-ICD limitations, which include a lack of antitachycardia pacing (ATP) and — at least in earlier years — a propensity for t-wave oversensing, researchers say. Both issues can promote shocks that might be avoidable with most conventional ICDs. The device also lacks chronic pacing for bradycardia, which rules out its use in many patients with an ICD indication.

Still, the prevalence of appropriate shocks — that is, for potentially life-threatening ventricular tachycardia or fibrillation (VT/VF) — was about 16%, in this "final report" from the EFFORTLESS European postmarket S-ICD registry study. About half the nearly 17% prevalence of inappropriate shocks over 5 years was seen in the first 12 months after implantation.

On the other hand, cardioversion was successful in 98% of cases, and on the first shock in 90%, Pier D. Lambiase, PhD, reported April 24 at the the virtual European Heart Rhythm Association (EHRA) 2021 congress.

"Only" 2% of the cohort were switched from their implanted S-ICD to transvenous systems to take advantage of their pacing capability for bradycardia in four patients, ATP in seven patients, or cardiac resynchronization therapy (CRT) in nine patients, he said.

The S-ICD was approved in Europe in 2008 and by the US Food and Drug Administration (FDA) in 2012. Patients in EFFORTLESS were implanted with the device from 2011 to 2014.

Although one suspected case of lead failure was investigated and discounted, "Interestingly, there were no definite electrode failures in this population of nearly 1000 patients," said Lambiase, University College London, United Kingdom.

That observation may have referenced the recent FDA-defined Class I recall of electrode components due to a potential fracture risk. The involved devices had been manufactured from March 2016 to November 2020. It was the first of two S-ICD safety recalls in the United States to be issued in the past 6 months.

The rate of shocks delivered inappropriately in response to atrial fibrillation (AF) or supraventricular tachycardia, another perceived S-ICD drawback that is also a potential issue with transvenous-lead systems, was "low," at about 3%, Lambiase said.

Mixed Cohort

EFFORTLESS, conducted at 43 European centers, followed 984 patients for a median of 5.1 years and a mean duration of S-ICD use of 4.4 years; their initial mean age was 48 and mean ejection fraction was 43%. The patients' heart-failure etiologies were ischemic in 29%; the remainder had various nonischemic conditions including hypertrophic cardiomyopathy and an assortment of channelopathies. The ICD indication was for primary prevention in 65%.  

Inappropriate shocks delivered by the S-ICD in the later years of follow-up were correlated with prior development of sensed — but not shocked — monomorphic VT that self-terminated, Lambiase observed. These made up 45% of detected episodes of monomorphic VT, the kind of VT that responds best to ATP.

"This is important, because it means that if we can identify these inappropriate episodes early, it may be possible to reprogram the device and prevent those shocks from occurring," he said.

Complications and Inappropriate Shocks in EFFORTLESS

End Points 1 Year (%) 5 Years (%)
Overall complication rate



Inappropriate shocks 8.7 16.9
Inappropriate shocks for t-wave oversensing> 3.4 5.8


A history of valve surgery, compared with no such surgery, was the most potent measured predictor of late complications or inappropriate shock, at a hazard ratio (HR) of 5.77 (95% confidence interval [CI], 2.22 - 15.03; P = .0003) in multivariate analysis. A history of cardiac arrest almost tripled the risk, HR, 2.70 (95% CI, 1.24 - 5.91; P = .013).

Performing a cardioversion test within 30 days of S-ICD implantation and the presence of AF each slashed the risk for complications or inappropriate shock by three-fourths (P = .017 and P = .034, respectively), Lambiase reported.

There were 91 deaths (9.2%) in the study, but none were related to the S-ICD system or the implanting procedure, he said. About 40% of the deaths were noncardiac, 38% were cardiac but nonarrhythmic, and three (0.3%) were arrhythmic — although another 11% of the deaths were of unknown cause.

"A Worthy Trade-off"

"The EFFORTLESS study demonstrated efficacy comparable to studies with transvenous ICDs, avoiding the serious complications associated with transvenous leads," said Giovanni Luca Botto, MD, ASST Rhodense-Garbagnate Milanese Hospital, Italy, as invited discussant for the Lambiase presentation.

Few patients experienced more than one episode of monomorphic VT over the 5 years, he observed, and almost half self-terminated while the rest elicited shocks. Thus, a need for ATP in the study was only 0.4% per year in the cohort that, by design, did not show a need for pacing at the time of S-ICD implantation.

"I strongly believe that patients and also physicians might accept that single or infrequent episode of an ICD shock for potentially ATP-amenable VT" from an S-ICD, Botto said, "as a worthy trade-off to avoid long-term lead complications from a transvenous ICD."

Even so, the lack of pacing is probably largely behind the "lower than expected" uptake of the S-ICD in practice, he noted.

"Not all patients need pacing, but I think the great majority of patients benefit from ATP which subcutaneous ICD doesn't offer at this stage," Haran Burri, MD, University Hospital of Geneva, Switzerland, said at a panel discussion during the EHRA sessions.

There is a place for S-ICDs, "and I am a user of S-ICDs," he said. But that use hinges on appropriate, arguably narrow, patient selection. "I see advantages with patients who have a high risk of infection, problems with vascular access, and very young patients who have arrhythmias which are not [amenable] to ATP," for example.

"However, I think that this therapy really needs to improve. We're really at early stages," he said. The S-ICD has "very limited programming capacities," and "the inappropriate shock rates are still somewhat high, even though they're getting better."

Certainly, "the transvenous ICD is needed for patients who need pacing for bradycardia," observed Liesbeth Timmers, MD, University Hospital Ghent, Belgium, at the same forum. And "you also need a transvenous system" for CRT.

But, "I think the S-ICD should be considered in all patients who do not need pacing," because it avoids "lead complications and systemic infections" that are the major complications associated with transvenous leads, Timmers said.

Ironically, Botto sees a theoretically greater risk for generator- and subcutaneous-lead-associated infections with the S-ICD as a "major" downside. An oft-cited limitation of the device is its shorter projected lifespan compared with transvenous lead systems: perhaps 7 years vs 10 or more years.

"The shorter longevity of the subcutaneous ICD compared to the transvenous will lead to a higher number of device replacements in the long run." And those procedures themselves, Botto noted, are "burdened with a higher risk for infection."

In light of the 7-year estimated S-ICD generator lifetime, Lambiase said, "we're going to have to wait at least another five years, I think, to start seeing the impact of generator changes. We've only had one infection with a generator change, out of about 50 devices so far, in the cohort. But we'll need to see the long-term outcomes, really, for this, in terms of the risk of infection."

Multivariate Predictions of Death in EFFORTLESS

End Points HR (95% CI) P Value
LV ejection fraction (per 5-point increment) 0.78 (0.70 - 0.88

< .0001

Kidney disease (vs none) 5.87 (3.55 - 9.69) < .0001
Diabetes (vs none) 2.27 (1.37 - 3.76) .0015
Age at implantation (per 5 y increment) 1.14 (1.03 - 1.25) .0088
Shock of any kind (vs no shocks) 1.71 (1.05 - 2.77 .031
Not significantly associated with mortality: defibrillation threshold testing, complications, primary vs secondary prevention.


Lambiase disclosed receiving fees for speaking or serving on an advisory board for Boston Scientific, Abbott, and Medtronic, and receiving educational and research grants from all three companies.

Follow Steve Stiles on Twitter: @SteveStiles2. For more from | Medscape Cardiology, follow us on Twitter and Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.