Author + information
- Received April 29, 2015
- Revision received June 23, 2015
- Accepted July 16, 2015
- Published online December 1, 2015.
- Jacob S. Koruth, MD∗,
- Marian K. Rippy, DVM, PhD†,
- Alexander Khairkhahan, MSc‡,
- David A. Ligon, MEng‡,
- Christopher A. Hubbard, MBA‡,
- Marc A. Miller, MD∗,
- Srinivas Dukkipati, MD∗,
- Petr Neuzil, MD, PhD§ and
- Vivek Y. Reddy, MD∗∗ ()
- ∗Helmsley Electrophysiology Center, Mount Sinai Medical Center, New York, New York
- †Rippy Pathology Solutions, Inc., Woodbury, Minnesota
- ‡St. Jude Medical, Inc., Sunnyvale, California
- §Na Homolce Hospital, Prague, Czech Republic
- ↵∗Reprint requests and correspondence:
Dr. Vivek Y. Reddy, Helmsley Electrophysiology Center, Mount Sinai School of Medicine, One Gustave L. Levy Place, Box 1030, New York, New York 10029.
Objectives This in vivo ovine study describes the feasibility and safety of retrieving implanted leadless pacemakers (LPs).
Background Although LPs have been shown to be removable soon after implantation, there are no data on the feasibility of removing chronically implanted LPs.
Methods This study was performed in 2 phases. In the mid-term cohort, 10 chronically (5.3 months) implanted animals underwent retrieval, followed by: 1) immediate necropsy in 5; and 2) in the remaining 5, reimplantation of a new LP followed by necropsy at 6 weeks. In the long-term cohort, 8 additional sheep underwent retrieval at 2.3 ± 0.1 years followed by necropsy. Retrieval was performed using either a single or triple loop snare. All 18 LPs (100%) were successfully retrieved. The time from retrieval catheter insertion to retrieval was 2:35 ± 01:11 and 3:04 ± 01:13 minutes in the mid-term and long-term study groups, respectively.
Results There were no significant differences in retrieval times using either snare. Intracardiac echocardiography was used pre- and post-retrieval to confirm the absence of pericardial effusion in all 8 sheep. On necropsy, there was no evidence of pericardial bleeding or perforation. Only minor tissue disruption and hemorrhage was noted at the implant site after retrieval. Histology demonstrated fibrous connective tissue at the contact sites of endocardium and LP can and at the helix. There was no evidence of pulmonary thromboembolism.
Conclusions We demonstrate the feasibility and safety of percutaneous, catheter-based retrieval in chronic LP implants of a maximum duration of approximately 2.5 years.
A novel leadless intracardiac pacemaker (LP) system implanted via a percutaneous, transvenous approach has been recently described (1). This approach has certain advantages, such as avoiding complications associated with traditional transvenous systems, including lead and surgical pocket complications (2–4). Briefly, the LP allows for bradycardia pacing via a miniature pulse generator that can be permanently implanted entirely within the right ventricle (RV). The calculated longevity (using a predictive model) with 100% pacing at 60 beats/min with an output of 2.5 V at 0.4 ms has been estimated to be between 8.8 and 9.8 years.
However, the optimal approach to addressing end of life for this device has yet to be determined. Potential options include adding a second device adjacent to the chronically implanted LP versus device retrieval and reimplantation of a new LP. Accordingly, in this in vivo ovine study of implanted LPs (up to 2.5 years), we sought to explore the retrieval option by describing: 1) a percutaneous transfemoral catheter-based retrieval procedure; 2) its efficacy and safety; and 3) the gross and histopathological changes at the RV implant site after retrieval.
The experimental protocol was approved by the Institutional Animal Care and Use Committee. The experimental protocol required an overnight fast after which sheep (60 to 80 kg) were placed under general anesthesia (details of anesthesia have been described previously ).
Salient features of the LP (Figure 1A) relevant to the retrieval procedure include: 1) proximal docking button; 2) titanium pulse generator; 3) distal nonretractable, single-turn helix (fixation mechanism); and 4) 3 radially extending nylon tines in the outer circumference of the tip of the LP (secondary fixation). The maximum depth of penetration of the fixation mechanism in tissue is 1.3 mm (details of LP have been described previously ).
Under continuous monitoring, unilateral percutaneous femoral venous access was obtained with an 18-F sheath. The implant procedure has previously been described in detail (5). Both active (capable of pacing) and inactive devices were implanted. Of the devices implanted, 10 of 10 in the mid-term arm were inactive and 3 of 8 devices in the long-term arm were inactive. The size, volume, composition (titanium), and external dimensions of both devices were identical.
A unilateral femoral vein was accessed percutaneously with an 18-F sheath. All catheters and sheaths were continually irrigated with heparinized saline. Contrast was then injected through either a pigtail catheter and/or through the protective sleeve of the retrieval catheter to identify the relative position of the LP in RV. Operators had the option of using either a single- or tri-loop retrieval catheter that was advanced through the tricuspid valve toward the RV (Figures 1B to 1D and 2A). The single-loop snare (inner diameter: 19 mm) is a tungsten coil mounted over a nitinol cable, and the triple-loop snare (inner diameter: 20 mm) is a platinum cable interlaced with nitinol cable. The protective sleeve was then pulled back, exposing the floppy portion of the retrieval catheter. The retrieval catheter was deflected to position the snare behind the docking button (Figure 2B). The snare was closed around the docking button by advancing the snare closure knob. The knob was then twisted counter-clockwise to lock the snare around the docking button (Figures 1D and 2C). The retrieval catheter was then advanced toward the LP to align the catheter docking cap with the LP docking button, and the snare control handle was pulled back to dock the retrieval catheter to the LP (Figure 2D). At this point any deflection within the system was released, and the snare control handle was turned counter-clockwise to allow the LP to complete at least 1.5 turns. Pulling back on the catheter allowed the LP to be released from the RV apex. The protective sleeve was then advanced over the LP, and then the retrieval catheter with the LP was pulled back into the LP introducer and removed from the body (Figures 2E and 2F).
Ten sheep with the LP in the RV apex were survived and underwent retrieval at a mean of 160 days (5.3 months). Retrieval was performed using the single-loop retrieval catheter. Five of the 10 sheep underwent immediate necropsy, and the remaining 5 sheep were implanted with a second LP at the RV apex, followed by survival for an additional period of 6 weeks before RV angiography and necropsy.
Eight chronically implanted sheep underwent device retrieval at a mean of 2.3 ± 0.1 years. Intracardiac echocardiographic (ICE) imaging was used pre- and post-retrieval in all animals in this cohort (Figure 3C). LP retrieval was performed using the single-loop retrieval catheter in one-half the cohort, and the triple-loop in the remaining animals. All animals underwent necropsy thereafter.
All animals were heparinized and humanely euthanized. The pericardial fluid was assessed. The hearts were explanted and inspected for the presence of pericardial adhesions. Acute changes in the right ventricle related to device retrieval and chronic changes were assessed. All valves were carefully examined.
The 10 animals in the mid-term retrieval study underwent limited necropsy without histopathological examination.
In addition to the above, the hearts of all animals in this group were immersion fixed in formalin before trimming for histopathology. The RV apex, free wall/septum, and the tricuspid valvular structures were carefully examined. Given the concern for potential pulmonary thromboemboli related to the retrieval procedure, their presence was specifically looked for. Lungs were cut at approximately 0.5- to 1.0-cm intervals, and all transverse sections of the pulmonary parenchyma were examined. Representative samples were taken from each lung lobe for histology. All peripheral organs and tissues were examined.
The implant site was excised from the RV apex and sectioned. The ventricular myocardium was cut at approximately 1.0-cm intervals and the myocardium assessed. Myocardial samples were taken from all four chambers for histological assessment. Sections were stained with hematoxylin and eosin and/or Masson’s trichrome stain.
Implant sites were assessed for the presence of fibrosis, location of the helix, and reaction surrounding the helix, connective tissue capsule surrounding the LP, inflammation, and the presence of thrombus. Nonparametric grading scales for assessing calcification and adipose tissue deposition associated with the implant site were used (Table 1). The myocardial slides were assessed for endocardial thickening and fibrosis, adherent fibrin/thrombus, edema, inflammation, infarction, and any evidence of cardiac remodeling. All lung sections were assessed for thromboemboli, infarcts, inflammation, and edema.
All continuous variables are presented as mean ± SD.
RV angiography was performed prior to retrieval in all animals and revealed no significant abnormality. The RV anatomy remained preserved, and, other than at the device tip, there was no angiographic suggestion of other adhesions along the body of the device to the endocardium.
ICE imaging (Figure 3C) was performed prior to and after retrieval in all 8 animals in the long-term group. This revealed no adherent tissue on the LP device; proximal portions of all devices were well visualized and moved freely without any adherence to the walls or the valvular apparatus. Additionally, a careful review of stored images demonstrated no evidence of restriction to tricuspid valvular motion, and there was no evidence of pericardial effusion. After device extraction, there was no evidence of pericardial effusion or hemodynamic instability until the time of euthanasia.
The LP was successfully retrieved in all 18 animals from both cohorts.
For the mid-term group, the average time from the time the single-loop retrieval catheter was introduced into the sheath until the snare was locked onto the LP docking button was 1:48 ± 1:16 min. In addition, the average time from when the retrieval catheter was inserted until the LP was fully retrieved was 2:35 ± 1:11 min (Table 2).
For the long-term group, the single-loop snare was used in 4 of 8 animals and the triple-loop snare in the remaining. The average time from when the retrieval catheter was inserted until the snare was locked onto the LP docking button was 1:53 ± 1:07 min. The average time from when the retrieval catheter was inserted until the LP was retrieved was 3:04 ± 01:13 min (Table 2).
Pathology: Mid-term group
All devices had a thin covering of fibrous tissue around the proximal docking button. In addition all helices were intact and free of tissue ingrowth. A limited examination was performed in the 5 sheep that were not reimplanted in this cohort. There was no perforation of the RV or atrium and no pericardial adhesions. Within the RV, there was mild endocardial fibrosis at the site of the distal tip of the LP implant, consistent with previous studies (5).
In the 5 sheep that underwent reimplant followed by necropsy at 6 weeks, the pericardial sac contained normal amounts of serous fluid (range of 1.0 to 7.6 ml), no evidence of perforation, and no pericardial adhesions. There was mild endocardial fibrosis surrounding the LP in the RV. All cardiac valves were normal in appearance. Each newly implanted LP was securely implanted, and the original implant site could not be identified. All LP devices were relatively free of connective tissue and thrombus material. In 4 of 5 sheep, small tags (0.5 to 1.0 cm long) of fibrous tissue were found at the implant site in the right ventricle.
There were no abnormal extracardiac findings, including no evidence of pulmonary thromboembolism or thromboembolism in any of the sheep. No device-related lesions were seen in the pulmonary parenchyma.
Pathology: Long term group
Similar to that seen in the mid-term group, there was no visible tissue on the body of the LP. There were small amounts of fibrous tissue around the proximal docking button and minimal tissue ingrowth at the helix in retrieved devices (Figures 3A and 3B). All helices were intact, demonstrating that the LP was unscrewed from the endocardium by the retrieval catheter.
At necropsy, there was no evidence of pericardial adhesions, infection or perforation of the helices into the pericardial space. Normal amounts of serous pericardial fluid (19.1 ± 12.5 ml) were noted in all animals. All hearts were normal in shape, and there was no evidence of cardiac remodeling, edema, dilatation, hypertrophy, or infarction. All cardiac valves were normal in appearance. Examination of transverse sections of ventricular myocardium did not reveal any evidence of infarction or scarring, infection, hemorrhage, edema, fibrosis, or congestion. The implant site contained small areas of tissue disruption with minimal acute hemorrhage and/or thrombus with occasional focal areas of endocardial hemorrhagic changes (Figures 4A to 4C). In 3 animals, the implant site was difficult to identify. Mild endocardial fibrosis of the free and septal right walls interfacing with the distal portion of the LP can was seen in all implanted animals; on sectioning, this did not appear to extend into the underlying myocardium (Figure 4). There was no gross evidence of valvular changes or ventricular geometric remodeling related to the implant.
A total of 11 sections from 4 chambers were examined to identify changes suggestive of cardiac remodeling. No significant abnormalities such as intercellular edema, congestion, myocardial loss, or fibrosis were found. Specifically, in the apex of the RV, small amounts of acute thrombus adherent to the endocardial surface surrounding the LP implant site was noted in 7 of 8 animals. Minor tissue disruption and hemorrhage were noted at the implant sites, likely related to device retrieval. The RV endocardium surrounding the LP implant site was mildly to moderately thickened with fibrous connective tissue in all animals. The myocardium in areas in proximity to the distal tip of the LCP was replaced by adipose tissue to varying degrees (mean score, 2.2 ± 08) (Table 1, Figure 4D). Additionally dystrophic myocardial calcification was also noted (mean score 1.1 ± 0.6) in its vicinity (5). Occasionally, mild endocardial ossification was seen directly subjacent to the tip of the can. Collagenized fibrous connective tissue was noted at the site of the helix attachment (Figure 4D). Incorporated and stabilized thrombus was present within the thickened endocardium at the interface with the LP in 1 of 8 chronic implants. Limited granulomatous inflammation, typical of a foreign body reaction, was also seen at the interface of the can and the deployed helix and on the myocardium.
There were no abnormal extracardiac findings, including no evidence of pulmonary thromboemboli, infarction, or edema in any of the 56 sections from the 8 animals.
Chronic retrievability of an implanted LP is an important capability when making decisions concerning appropriate patient selection. Although the leadless approach to single-chamber pacing offers distinct advantages, the optimal strategy for end of life for this class of pacemakers is yet to be determined. Other potential situations such as device infection may also warrant device retrieval. The small volume (1 cm3) of the LP relative to that of the RV cavity raises the possibility of adding a second device adjacent to the initial implant without attempting retrieval. Although this approach may be feasible and is being investigated, the ability to safely retrieve the LP and reimplant another LP device is more appealing for several reasons. This retrieval strategy may allow for: 1) easier reimplantation of the newer device in the RV; 2) reduced risk of potential device-device interactions; and 3) reduced concern for the long-term effects of multiple devices retained within the RV. These reasons, along with success and safety of device retrieval, may ultimately lead to increased acceptance of this mode of pacing.
This pre-clinical report demonstrates primarily the feasibility and safety of percutaneous device retrieval for LP implants in the short term (mean duration: 5.3 months) and long-term (mean duration: 2.3 ± 0.1 years). Not only were all retrievals successfully performed, they were performed with relative ease. This is evidenced by the short retrieval times: overall mean duration 2 min 48 s (range 1 min to 5 min 40 s) even in the long-term cohort. Additionally, the gross and histological analysis performed after retrieval revealed several interesting findings: 1) the local endocardial response to the implanted LP at the RV apex remained limited even at >2 years; 2) there were no significant adhesions between the LP and the RV walls; and 3) after LP retrieval, there was evidence of only small amounts of acute thrombus, very localized endocardial and subendocardial hemorrhage, and only minor tissue disruption at the original implant sites.
These findings have several implications. The relatively short retrieval times for the LP, the absence of significant adhesions between the RV and LP, and lack of significant tissue growth on the explanted device at more than 2 years suggest that the option of retrieval may remain feasible at even longer time points, such as at end of life. Furthermore, the limited occurrence of minimal tissue disruption and hemorrhage at the explant site suggests that the retrieval process is not very traumatic and therefore is likely to retain the same safety profile demonstrated in this report for LPs that have been implanted for greater durations.
On histology, the RV apex demonstrated mild fibrotic endocardial thickening, minimal myocardial fibrosis, and small amounts of incorporated fibrin thrombus at the interface of the endocardium and the LP. Histology also demonstrated the occurrence of adipose tissue at the implant site and the occurrence of (dystrophic) calcification, both of which are considered to be expected in the species (ovine) chosen for this study (5) and therefore should not result in any significant clinical consequences. In addition, the lack of extensive changes at the RV apex, which is consistent with previously published animal data (18-month follow-up) supports the feasibility of reimplanting a new LP at the RV apex after retrieval without impairment of sensing and pacing parameters.
The absence of device-related pulmonary thromboembolism in either group further supports the safe use of the LP device in the long term and indicates that there is no significant acute thrombotic risk associated with the retrieval process.
The report addresses the safety and feasibility of catheter-based retrieval for LP devices that have been implanted for a period of up to ∼2.3 years. Given that the most likely reason for retrieval is end of life, longer term studies (7–8 years) may be needed to truly assess the efficacy and safety of this approach at these longer time points. It is important to note that the tissue reaction between device and the myocardium in the ovine model may not be similar to that in humans in the long term; however, data from the ovine model at the very least serve as a guide to what may be expected in humans. The choice of the ovine model is supported by its use in the evaluation of other intracardiac devices (6,7). The success and safety of this approach as demonstrated in this report cannot necessarily be extended to the clinical realm where patients often have significant structural heart disease.
This ovine study demonstrates the feasibility and safety of percutaneous catheter-based retrieval of implanted LPs implanted in the right ventricular apex. The maximum duration of implant in this study was 2.5 years. The result of this study indicates that LP retrieval may be considered as a potential option for implanted LPs in humans in the ∼2 year time frame. However, the feasibility of retrieval in humans at longer time points requires further study.
COMPETENCY IN MEDICAL KNOWLEDGE: The approach to end of life and other situations that require removal of LPs is unknown. Results of this study indicate that LP retrieval may be considered a potential option for implanted LPs in humans in the ∼2-year time frame.
TRANSLATIONAL OUTLOOK: This ovine study demonstrates the feasibility and safety of percutaneous catheter-based retrieval of LPs implanted in the right ventricular apex. Additional clinical studies are needed to validate the safety and incremental value of this approach. This serves as a guide as to what may be expected in humans.
This study was funded by St. Jude Medical.
Dr. Koruth has received honoraria from and is a lecturer for St. Jude Medical. Dr. Rippy has received consulting fees and honoraria from St. Jude Medical. Mr. Khairkhahan, Mr. Ligon, and Mr. Hubbard have equity interests and hold stock options in St. Jude Medical. Drs. Neuzil and Reddy have received consulting fees, honoraria, and research grants from St. Jude Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- intracardiac echocardiography
- leadless pacemaker
- right ventricle
- Received April 29, 2015.
- Revision received June 23, 2015.
- Accepted July 16, 2015.
- American College of Cardiology Foundation
- Reddy V.Y.,
- Knops R.E.,
- Sperzel J.,
- et al.
- Nof E.,
- Epstein L.M.
- Bernard M.L.,
- Shotwell M.,
- Nietert P.J.,
- Gold M.R.
- Rees J.B.,
- de Bie M.K.,
- Thijssen J.,
- Borleffs C.J.,
- Schalij M.J.,
- van Erven L.
- Cooper R.A.,
- Alferness C.A.,
- Smith W.M.,
- Ideker R.E.
- Powell A.C.,
- Garan H.,
- McGovern B.A.,
- et al.