Abstract: Purpose. Patients with heart failure supported with left ventricular assist devices (LVADs) may require coronary intervention during their support. This case series seeks to explore the indications, safety, and outcomes of percutaneous coronary intervention (PCI) in this population. Methods. Electronic medical records of patients with LVADs undergoing PCI at a large academic medical center were reviewed. Demographics, reason for PCI, procedural success, complications, and outcomes were collected. Results. From 2010-2014, a total of 6 patients underwent PCI post LVAD implantation. Three patients had PCI in the early postimplantation period (1-3 days post LVAD implantation) while the other three received it later in the LVAD support period. Three indications for PCI were found in the reviewed cases: right ventricular failure (right coronary artery stenting), bridge to left ventricular recovery, and ventricular tachycardia (VT) storm. All patients were maintained on triple blood thinning therapy (aspirin, clopidogrel, and warfarin). There were no acute complications during the interventions; however, 2 patients died in the early intervention period and 2 died much later. The 2 deaths in the early intervention period were related to fatal gastrointestinal bleeding while on dual-antiplatelet therapy and warfarin, and intractable VT that PCI did not correct. The 2 deaths in the late postintervention period occurred due to unknown causes nearly 1 and 2 years post intervention, respectively. Conclusions. PCI was performed in patients with continuous-flow LVAD with several possible indications and without acute complications. The utility of PCI in this patient population, however, is likely limited by the risk of bleeding related to combined antiplatelet and anticoagulation therapies as well as lack of immediate apparent benefit. Further studies are necessary to better characterize this risk as well as quantify any potential long-term benefits.
J INVASIVE CARDIOL 2016;28(6):238-242
Key words: heart failure, left ventricular assist devices, percutaneous coronary intervention
Continuous-flow left ventricular assist device (CF-LVAD) support has a clear survival benefit in end-stage heart failure (HF).1 With improvements in long-term outcomes, CF-LVADs are increasingly used for both bridge to transplantation and destination therapy (DT). The latest INTERMACS report shows that yearly implantation of CF-LVADs is increasing year after year, with over 2400 devices implanted in 2013 alone.2 Once on CF-LVAD support, it is unusual for patients to develop complications from coronary artery disease that would require percutaneous coronary intervention (PCI). However, in rare cases, PCI may be performed to provide better perfusion to the right ventricle (RV). Furthermore, PCI of arteries supplying the left ventricle (LV) may be considered in patients who are candidates for LVAD explantation due to ventricular recovery or in patients who experience myocardial infarction during support.3
CF-LVAD support requires the use of long-term anticoagulation therapy with both warfarin and antiplatelet drugs. The tendency toward bleeding is a consequence of both anticoagulation therapy and acquired coagulopathy. Previously, our group and others reported that device design affects coagulation factors and results in acquired von Willebrand disease in HeartMate II4, HVAD5, and other CF devices.6 The combination of anticoagulation therapy and acquired von Willebrand disease on PCI risk in this population is not well described in the current literature. In this case series, we describe our center’s experience with PCI following LVAD implantation.
From 2010-2014, a total of 6 patients underwent PCI after LVAD implantation at the University of Chicago Medical Center. Three patients had the PCI performed in the early postimplantation period (1-3 days after implantation) while the other 3 underwent PCI later in their LVAD support period.
PCI was performed in 4 patients due to RV failure. Three of those patients’ stents were placed directly to the right coronary artery (RCA) and in 1 patient to a venous graft supplying the RCA. The other 2 patients in the series received stents to the left anterior descending (LAD) artery as a bridge to LV recovery and a venous graft supplying an obtuse marginal artery (due to complete total occlusion of a LAD graft) due to refractory ventricular tachycardia (VT) storm. All patients were loaded with 600 mg of clopidogrel and were continued on 75 mg of clopidogrel daily added to their background therapy of aspirin and warfarin. There were no intraprocedural complications during PCI. Four patients underwent bare-metal stent (BMS) implantation and 2 patients underwent drug-eluting stent (DES) implantation (Tables 1 and 2).
Patient #1. A 73-year-old male with ischemic cardiomyopathy was implanted with a HeartMate II (Thoratec Corporation). LVAD implantation was combined with venous graft to the LAD. A 70% RCA ostial lesion was not bypassed. Intraoperative course was complicated by repeated failure to wean from cardiopulmonary bypass due to RV failure and hemodynamic collapse. A CentriMag (Thoratec) right ventricular assist device (RVAD) was placed. The patient was taken for left heart catheterization on postoperative day 1 via femoral approach with placement of a 2.0 x 8 mm BMS to the ostial RCA lesion (the RCA was non-dominant). On postoperative day 2, the patient was taken to the operating suite for explantation of the RVAD without complication. The patient’s postoperative course was uncomplicated and he was discharged to acute rehab on postoperative day 12. The patient was doing well for 2 years; however, 547 days after implant, he had sudden death from unknown cause.
Patient #2. A 72-year-old male with ischemic cardiomyopathy status post 4-vessel coronary artery bypass graft (CABG) surgery 21 years ago and subsequent coronary stenting after CABG was admitted for LVAD implantation. Preimplantation work-up demonstrated an 80% occlusion of his venous graft supplying the posterior descending artery (PDA), but intervention was delayed until post-VAD implantation to avoid starting dual-antiplatelet therapy (DAPT) preoperatively. He was successfully implanted with a HeartWare HVAD (HeartWare). On postoperative day 2, he was taken electively to the cath lab and received a 3.0 x 16 mm everolimus DES to the aforementioned 80% lesion of the saphenous vein graft (SVG) to PDA via femoral approach. On postoperative day 21, the patient was discharged to home and he is still alive 204 days post implantation.
Patient #3. A 65-year-old male with ischemic cardiomyopathy was implanted with HeartMate II as DT. On postoperative day 1, extubation was complicated by worsening RV failure and atrial fibrillation refractory to electrical cardioversion. The patient was re-intubated and placed on inhaled nitric oxide. On postoperative day 2, the patient was taken for left heart catheterization by femoral artery approach and a 90% lesion in the (dominant) RCA just proximal to the acute marginal branch was stented. Three overlapping BMSs (2.5 x 12 mm, 2.25 x 18 mm, and 2.25 x 14 mm) leading into the acute marginal branch were placed without complication. Despite the intervention, his RV failure persisted and on postoperative day 13 the patient developed gastrointestinal bleeding. Aspirin and heparin were held and clopidogrel was continued for 2 more weeks before being discontinued. Endoscopy revealed a gastric ulcer and a bleeding colonic polyp, both of which were cauterized. Heparin was later restarted and transitioned to warfarin. On postoperative day 67, the patient went for surveillance right heart catheterization by femoral vein access and consequently developed a hematoma requiring surgical evacuation. After a prolonged course, the patient was discharged to nursing home on postoperative day 126 on warfarin therapy. The patient did well for nearly a year post intervention; however, he passed at 231 days of unknown causes.
Patient #4. A 75-year-old male with ischemic cardiomyopathy was implanted with HeartMate II as DT. The patient presented to the hospital 30 months post implantation in VT storm with multiple implantable cardioverter defibrillator (ICD) shocks and loss of consciousness. Two-dimensional transthoracic echocardiography (TTE) on presentation demonstrated increased LV dilation and moderate RV dysfunction. The LVAD inflow cannula was correctly positioned and there was no evidence of suction events by echo or controller interrogation. Due to suspicion of ischemia causing VT, the patient underwent left heart catheterization via a femoral approach and was found to have complete total occlusion of his SVG to LAD graft and a 90% lesion of his SVG to first obtuse marginal (OM1) venous graft. A 4.0 x 12 mm BMS was placed to the OM1-supplying venous graft. Despite all efforts, the patient’s condition remained unstable with multiple infectious complications and died 45 days after admission.
Patient #5. A 59-year-old man with ischemic cardiomyopathy was implanted with HeartMate II as DT. The patient’s support period was complicated by gastrointestinal (GI) bleeding 28 months post implantation and he was stabilized with blood products. Colonoscopy revealed rectal ulcers. During this hospitalization, a follow-up TTE was obtained due to poor respiratory function and demonstrated interval worsening of RV function. Left heart catheterization was performed via femoral approach and the patient was found to have a 75% occluding lesion of the proximal (non-dominant) RCA, across which a 2.25 x 22 mm BMS was deployed. Nine days post PCI, the patient’s melanotic stools resumed. Aspirin, clopidogrel, and warfarin were held and he was supported with blood products and intravenous octreotide and esomeprazole. Intensive GI work-up including upper endoscopy, colonoscopy, and bowel angiography were performed, yet a bleeding source could not be re-identified. Despite these interventions, the patient continued to suffer from significant GI blood loss and after 10 days the patient and his family requested to transition to comfort care. The patient was discharged to hospice and expired several days later.
Patient #6. A 59-year-old male with combined nonischemic and ischemic cardiomyopathy was implanted with HeartMate II as DT; surgery included tricuspid and mitral annuloplasty ring placement. Pre-VAD work-up had revealed an 80% LAD lesion that was not intervened on in order to avoid starting DAPT prior to implantation. The patient’s immediate postoperative course was complicated by RV failure requiring inotropic support and pulmonary hypertension. Inotropes were weaned by postoperative day 16. The patient was discharged to acute rehab in stable condition on postoperative day 43. On postoperative day 96, the patient was taken electively for left heart catheterization with a 3.0 x 15 mm zotarolimus DES placed to the prior known LAD lesion via femoral approach. The patient is still alive 337 days post operation.
This is the first case series describing PCI after LVAD implantation. Our experiences highlight possible indications, safety considerations, and outcomes for PCI in this patient population as well as areas of potential future investigation.
PCI indications. In our series, we encountered three indications for PCI in patients on LVAD support: RV failure, refractory VT, and effort to achieve LV recovery.
RV failure. Four out of 6 patients in the series underwent PCI to the coronary arteries supplying the RV (RCA or PDA) to support RV failure post LVAD implantation. LVAD support has been associated with worsening RV function due to increased preload return to the RV and an ensuing hemodynamic imbalance between RV and LV function. A retrospective study by Dang et al suggests that postimplantation development of RV failure may approach 40% incidence.7 There is some evidence suggesting that in the case of ischemic injury, PCI of the RCA can lead to improved RV function in non-LVAD patients.8 As such, our patients underwent RCA revascularization in an effort to improve RV function and improve RV failure. Three of these patients underwent PCI for RV failure in the early postimplantation period, while 1 patient’s RCA stenting took place much later. There are several mechanisms other than ischemia that have been proposed for RV failure in LVAD patients that may not improve with revascularization. As discussed by Romano et al, leftward interventricular septum displacement due to LV unloading may lead to unfavorable geometry and perturb RV wall stress.9 This may also lead to tricuspid malcoaptation, leading to increased RV preload and worsening function. There is also thought that flow dyssynchrony between the RV and an LV supported by continuous flow plays a role in RV failure.10
Refractory ventricular tachycardia. End-stage heart failure puts patients at significant risk for fatal ventricular arrhythmias, particularly VT. This is in part because of the arrhythmogenicity of the scar tissue that forms in ischemic cardiomyopathy and in some cases of non-ischemic cardiomyopathy. Despite the fact that acute VT is well tolerated in CF-LVAD patients,11 sustained VT can lead to hemodynamic collapse, repeat ICD firings, and RV failure. VT storm is an established indication for PCI.12
Left ventricular recovery. Literature surrounding ischemic cardiomyopathy generally supports the notion that coronary revascularization can result in improvement of LV function with some reversal of remodeling in “hibernating” myocardium.13 We feel that there may be a role in providing additional support by revascularization to patients with a component of ischemic cardiomyopathy in order to maximize the potential recovery of hibernating myocardium.
Procedural safety. All 6 patients in the case series underwent PCI via femoral approach. Current literature favors radial approach in acute coronary syndrome (ACS) patients at high-volume centers.14 For non-ACS patients, there is no evidence to support one method versus the other and our patients received the approach judged by the PCI operator to be clinically appropriate. All patients were accessed via 6 Fr introducer sheath. Patients were either closed with the aid of collagen-based arteriotomy closure or had the sheath sewn in place and removed at a later date. Periprocedural anticoagulation was achieved with the aid of either bivalirudin or heparin with an average activated clotting time of 355 seconds. There were no arterial access-site related bleeding events.
The choice between BMS and DES was left to the PCI operator. The use of antiplatelet agents in LVAD patients is not taken lightly, as multiple authors have shown that LVAD patients have impaired platelet function and are predisposed to GI bleeding.15,16 Four of the 6 patients received BMS and 2 received DES. Support for DES versus BMS for patients on anticoagulation therapy is limited in the current literature and consists mostly of observational studies for patients anticoagulated for atrial fibrillation. In one such 2009 study by Ruiz-Nodar et al, a significantly higher rate of major bleeding was observed in patients receiving DES versus BMS.17 This study did not find a statistically significant reduction in major adverse cardiac event (MACE) rate or all-cause mortality. Similarly, in a 2014 observational study, Kivemini et al found statistically similar MACE rates between the two groups, without statistical difference in the incidence of major bleeding events.18 These studies when considered with respect to LVAD patients may suggest that BMS should be preferred over DES in order to reduce bleeding risk.
Of the study patients, 2 experienced bleeding events. Patient #3 developed GI bleeding 2 weeks after PCI, requiring the discontinuation of aspirin, holding of heparin, and interventional endoscopy. In addition to the bleeding risk inherent in triple-antithrombotic therapy, this particular patient may have been at increased risk given his history of gastroesophageal reflux disease prior to LVAD. This was supported by his upper endoscopy that revealed a gastric ulcer. This same patient later experienced a femoral vein hematoma after undergoing a routine surveillance right heart catheterization while on warfarin alone. Patient #5 was admitted with GI bleeding from rectal ulcers, was stabilized, and then later underwent PCI. The GI bleed started again 9 days later and an origin was not found. The GI bleeding sources in both of these cases are likely at least partially a manifestation of the bleeding diathesis that CF-LVAD patients develop. Proposed mechanisms include acquired von Willebrand deficiency15 as well as mucosal ischemia and angiodysplasia that occurs due to lack of physiologic pulsatile blood flow.16 To what extent the addition of DAPT exacerbated baseline bleeding risk cannot be immediately concluded. Data from WOEST, the 2013 open-label prospective study of patients on oral anticoagulation receiving PCI, suggest that bleeding risk may be reduced by using clopidogrel and oral anticoagulation without aspirin. Although not a primary endpoint of the study, this lower-risk regimen was not shown to statistically increase the risk for thrombotic and thromboembolic events.19 Whether or not functional platelet assays might help tailor periprocedural DAPT and anticoagulation in this patient population is a potential area for study that is beyond the scope of this case series. Despite the small case series we present here, the risk of bleeding in CF-LVAD patients treated with triple therapy may be substantial (33% in our patients), and requires extra caution until further studies elucidate this risk.
Outcomes. Four patients underwent intervention for RV failure. Three of these patients survived the immediate post-PCI period, while 1 patient died. Post PCI, 1 patient successfully weaned from RVAD support, which may suggest some contribution of improved coronary supply to the RV. Patient #2’s postimplantation stenting to his OM venous graft successfully achieved revascularization of the RV, yet was able to avoid the risk inherent in starting DAPT preoperatively. The utility of stenting patient #3’s RCA is less clear, as his RV failure persisted nearly 2 weeks after intervention. Patient #5 expired due to GI bleeding, although he was also admitted with GI bleeding and thus likely had high baseline bleeding risk. There is no objective evidence to qualify whether coronary revascularization improved RV function in patients #2, #3, and #5.
Catheter ablation in CF-LVAD patients, although emerging as a safe and often effective intervention,20 was not successful for patient #4 in the series. As such, he was taken for PCI in an effort to rule out ischemic etiology of VT after more common causes had been excluded. Despite the localization and stenting of a coronary lesion, his VT persisted, suggesting that ischemia was unlikely the etiology of his arrhythmia. While this intervention did not further contribute to the patient’s immediate morbidity and mortality, our experience does not support VT storm as an indication for PCI.
Patient #6 underwent stenting in an effort to minimize any ischemic contribution to his cardiomyopathy with hope for LV recovery. Although it will not be possible to distinguish the effect of PCI versus that of LVAD support on the patient’s cardiac function, the revascularization was successful and without complication. At the time of writing this manuscript, the patient is alive but has not yet undergone repeat assessment of LV function. We feel that revascularization may be worthwhile in patients with both ischemic and non-ischemic cardiomyopathy.
Study limitations. This case series is limited by several factors. Our patient sample was limited and our cases were drawn from a single center’s experience. These factors reflect the low incidence of PCI in LVAD-supported patients and the paucity of currently available supporting literature. As such, this manuscript highlights important possible indications for future clinical consideration and study. As our data were collected retrospectively, temporal correlation between intervention and benefit is not clear in all cases.
Our case series found several indications for PCI in LVAD patients, ie, RV failure, VT storm, and for recovery purposes; however, we couldn’t demonstrate benefit in those areas. There were no acute complications during the PCI procedures; however, several cases were complicated by bleeding related to combined DAPT and anticoagulation. The risk of bleeding relative to any potential benefits in LVAD patients undergoing PCI needs to be explored in larger studies.
1. Rose EA, Gelijns AC, Moskowitz AJ, et al. Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med. 2001;345:1435-1443.
2. Kirklin JK, Naftel DC, Pagani FD, et al. Sixth INTERMACS annual report: a 10,000-patient database. J Heart Lung Transplant. 2014;33:555-564.
3. Fried J, Han J, Naka Y, Jorde UP, Uriel N. Myocardial infarction after left ventricular assist device implantation: clinical course, role of aortic root thrombus, and outcomes. J Heart Lung Transplant. 2014;33:112-115.
4. Uriel N, Pak S-W, Jorde UP, et al. Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding during long-term support and at the time of transplantation. J Am Coll Cardiol. 2010;56:1207-1213.
5. Birschmann I, Dittrich M, Eller T, et al. Ambient hemolysis and activation of coagulation is different between HeartMate II and HeartWare left ventricular assist devices. J Heart Lung Transplant. 2014;33:80-87.
6. Morrison KA, Jorde UP, Garan AR, Takayama H, Naka Y, Uriel N. Acquired von Willebrand disease during CentriMag support is associated with high prevalence of bleeding during support and after transition to heart replacement therapy. ASAIO J. 2014;60:241-242.
7. Dang N, Topkara VK, Mercando M, et al. Right heart failure after left ventricular assist device implantation in patients with chronic congestive heart failure. J Heart Lung Transplant. 2006;25:1-6. Epub 2005 Dec 9.
8. Nikdoust F, Tabatabaei SA, Shafiee A, Mostafavi A, Mohamadi M, Mohammadi S. The effect of elective percutaneous coronary intervention of the right coronary artery on right ventricular function. Int Cardiovasc Res J. 2014;8:148-151.
9. Romano MA, Cowger J, Aaronson KD, Pagani FD. Diagnosis and management of right-sided heart failure in subjects supported with left ventricular assist devices. Curr Treat Options Cardiovasc Med. 2010;12:420-430.
10. Felix SEA, Martina JR, Kirkels JH, et al. Continuous-flow left ventricular assist device support in patients with advanced heart failure: points of interest for the daily management. Eur J Heart Fail. 2012;14:351-356.
11. Garan AR, Yuzefpolskaya M, Colombo PC, et al. Ventricular arrhythmias and implantable cardioverter-defibrillator therapy in patients with continuous-flow left ventricular assist devices: need for primary prevention? J Am Coll Cardiol. 2013;61:2542-2550.
12. Authors/Task Force members, Windecker S, Kolh P, et al. 2014 ESC/EACTS guidelines on myocardial revascularization: the task force on myocardial revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) Developed with the special contribution of the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur Heart J. 2014;35:2541-2619.
13. Carluccio E, Biagioli P, Alunni G, et al. Patients with hibernating myocardium show altered left ventricular volumes and shape, which revert after revascularization: evidence that dyssynergy might directly induce cardiac remodeling. J Am Coll Cardiol. 2006;47:969-977.
14. Jolly SS, Cairns J, Yusuf S, et al. Procedural volume and outcomes with radial or femoral access for coronary angiography and intervention. J Am Coll Cardiol. 2014;63:954-963.
15. Klovaite J, Gustafsson F, Mortensen SA, Sander K, Nielsen LB. Severely impaired von Willebrand factor-dependent platelet aggregation in patients with a continuous-flow left ventricular assist device (HeartMate II). J Am Coll Cardiol. 2009;53:2162-2167.
16. Demirozu ZT, Rajko R, Hochman LF, et al. Arteriovenous malformation and gastrointestinal bleeding in patients with the HeartMate II left ventricular assist device. J Heart Lung Transplant. 2011;30:849-853. Epub 2011 Apr 29.
17. Ruiz-Nodar JM, Marín F, Sánchez-Payá J, et al. Efficacy and safety of drug-eluting stent use in patients with atrial fibrillation. Eur Heart J. 2009;30:932-939.
18. Kiviniemi T, Puurunen M, Schlitt A, et al. Bare-metal vs drug-eluting stents in patients with atrial fibrillation undergoing percutaneous coronary intervention. Circ J. 2014;78:2674-2681.
19. Dewilde WJM, Oirbans T, Verheugt FWA, et al. Use of clopidogrel with or without aspirin in patients taking oral anticoagulant therapy and undergoing percutaneous coronary intervention: an open-label, randomised, controlled trial. Lancet. 2013;381:1107-1115.
20. Garan AR, Iyer V, Whang W, et al. Catheter ablation for ventricular tachyarrhythmias in patients supported by continuous-flow left ventricular assist devices. ASAIO J. 2014;60:311-316.
From the University of Chicago, Departments of 1Internal Medicine; 2Cardiac Surgery; 3Advanced Heart Failure; and 4Interventional Cardiology, Chicago, Illinois.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Nathan reports consultant fees for Medtronic and Terumo Interventional Systems; Dr Jeevanandam reports consultant fees from HeartWare, Thoratec, and ReliantHeart; Dr Shah reports consultant and speaker’s bureau fees from Medtronic and Maquet. Dr Uriel reports consultant fees and grant support from Thoratec and HeartWare; consultant fees from ABiomed and Medtronic. The remaining authors report no disclosures regarding the content herein.
Manuscript submitted November 17, 2015, provisional acceptance given February 2, 2016, final version accepted February 25, 2016.
Address for correspondence: Nir Uriel, MD, MSc, Professor of Medicine, Medical Director of Heart Failure, Heart Transplant and Mechanical Assist Device Programs, University of Chicago, Division Of Cardiology, Chicago, IL 60637. Email: email@example.com