Left ventricular assist devices (LVAD) provide a durable form of mechanical circulatory support (MCS) in patients with end-stage heart failure and, in the absence of device complication, can provide improvements in quality of life and survival.1 Although LVAD implantation is often a necessary, life-saving intervention, support with these devices portends a high morbidity rate. Rates of infection in LVAD patients are reported as high as 58% at 2 years with other complications, such as pump thrombosis and device failure, also a concern.2,3 Even when promptly recognized and aggressively treated, LVAD patients who develop central device infections are associated with a five-fold increase in mortality.1

Initial management of LVAD infections involves tailored antimicrobial therapy and surgical drainage; however, device explant may be required for adequate source control.1 In the bridge to transplant (BTT) population, management of central device infection is further complicated by recent changes in the United Network of Organ Sharing (UNOS) allocation system. Under current 2018 UNOS guidelines, BTT patients suffering from documented LVAD infection with bacteremia are assigned to a UNOS Status 3, which is comparatively lower status than in the 2006 system.4 The 2006 UNOS allocation system allowed reassignment to Status 1A for documented central device infection, comparable to that of temporary MCS device support (current status 2). Under current guidelines, waitlist times for patients suffering from LVAD infection with persistent bacteremia are often prohibitively long, necessitating clinicians to take novel steps to salvage the patient and prioritize their listing status.4 The clinical implications of central device infections and the coinciding increased risk in patient mortality raise concerns about the appropriateness within which the 2018 guidelines view this life-threatening condition.2,3 For BTT patients with a central LVAD infection awaiting a donor, options are limited to either subjecting the patient to the ongoing morbidity and risk of mortality associated with an active infection while waiting for a suitable donor, or as we show in this case, device explant with insertion of temporary MCS that then allows for improved mediastinal sterilization and subsequent uplisting before more definitive transplant surgery. This case report defines an inherent pitfall within the current UNOS allocation system for patients with a life-threatening central LVAD infection and highlights the use of temporary MCS to successfully salvage and mitigate ongoing complications with device infection.

Case Presentation

A 36-year-old man with nonischemic dilated cardiomyopathy supported with a Heartmate 3 (HM3) for over a year as BTT (UNOS Status 4) presented with fevers, leukocytosis, and tachycardia. Blood cultures on admission were positive for methicillin-sensitive Staphylococcus aureus. There was no evidence of driveline infection on the exam and the patient had no other indwelling lines or other identifiable sources of infection, however, positron emission tomography (PET) imaging showed diffuse and increased mediastinal uptake around the outflow graft consistent with outflow graft infection (Figure 1). The patient underwent an urgent subxiphoid window with drainage of frank pus around the outflow graft with the application of a wound vacuum dressing. Postoperatively, the patient was upgraded to UNOS Status 3 for central device infection with the hopes of finding a suitable donor without the need for an explant. Despite appropriate antibiotic coverage and wound vacuum-assisted therapy for 15 days, the patient continued to have significant amounts of purulent output from his mediastinum and remained clinically septic with ongoing tachycardia and malaise. Additionally, due to the patient’s O+ blood type and average body size, the projected time to find a suitable donor was greater than one year. Therefore, the decision was made to perform explant of his LVAD for improved source control and prevent further deterioration.

Figure 1.:

PET scan showing diffuse increased FDG avid uptake around the entire outflow graft of the HM3 device consistent with device infection. 1a: Axial view with FDG avid uptake around outflow graft. 1b: Coronal view showing involvement of the entire outflow graft and central pump. FDG, fluorodeoxyglucose; HM3, Heartmate 3; PET, positron emission tomography.

A reoperative sternotomy was performed and upon entering the mediastinum, extensive pus was encountered around the outflow graft and pump head. The patient was placed on cardiopulmonary bypass and the LVAD was explanted, leaving the inflow cuff attached to the apex and a small stump of outflow graft on the native aorta. The driveline was dissected out and explanted as well. The apex was closed with a handmade felt plug and oversewn with a bovine pericardial patch for hemostasis. This enabled use of the inflow cuff to re-insert an LVAD following sterilization of his mediastinum if a suitable donor was unable to be identified in a timely fashion. Due to the inability to wean the patient safely from cardiopulmonary bypass, an Impella 5.5 ventricular assist device was inserted through the left axillary artery and into the left ventricle under fluoroscopic guidance to provide necessary hemodynamic support. The chest was successfully closed without significant hemodynamic changes. Postoperatively, he was extubated with end-organ function intact, and blood cultures subsequently cleared. He was ambulatory while on Impella 5.5 support and remained hemodynamically stable; however, the patient remained fully dependent on this device with persistent severe right ventricular dysfunction. He was then reactivated as a Status 2 due to dependence on a peripheral temporary MCS device. Fortunately, with the patient being listed as Status 2, a suitable donor was found, and the patient underwent successful heart transplantation during the same admission. He was extubated on postoperative day 1 and was transferred out of the intensive care unit on postoperative day 6. He was discharged home postoperative day 23 and has been clinically stable with no signs of rejection or infection at his most recent follow-up 8 months posttransplantation.

Discussion

Central LVAD infection is a well-recognized and feared complication that portends a poor prognosis for patients. The mortality rate after central LVAD infection is reported in the literature to be >five-fold of that for uninfected LVAD patients.1 Device-specific infection rates are highest within the 90- to 180-day window with an overall infection rate of over 50% at 2 years.2 While infection prevention measures must be implemented for these patients, better management strategies should also be derived for those who invariably develop life-threatening infections.2 In the case of the patient above, it was not until his infected device was explanted and he was bridged with temporary MCS that appropriate organ allocation could be achieved facilitated by upgrading from UNOS Status 3 to Status 2. However, the clinical implications of this were numerous. Explant of the durable device required an additional sternotomy in the setting of mediastinitis and cardiogenic/septic hemodynamic compromise. Added to this was the significant clinical decline and deconditioning from his ongoing sepsis while more conservative measures were attempted at Status 3 as well as the additional risks and costs associated with the intermediate step requiring Impella support at Status 2. These additional steps were not only necessary for patient care, but were also needed to find a suitable donor for a patient dependent on MCS (temporary or durable) with severe right ventricular dysfunction.

While the 2018 UNOS guidelines intended to reduce waitlist mortality by preferentially allocating organs to more acutely sick patients, the implications of this revision for patients with durable support devices and the accompanying pattern practice changes are becoming increasingly evident. Numerous reports have demonstrated higher rates of pre-transplant temporary support device utilization, with rates of pre-transplant LVAD significantly lower following the 2018 revisions.5–7 This trend in clinical practice reflects the concern that patients suffering from central LVAD infection will be disadvantaged in finding a suitable donor without escalations in care. One solution to this problem would be to reclassify central device infection with septicemia to a Status 2 listing, comparable to temporary mechanical support measures. In the case of our patient, temporary MCS was able to provide sufficient support to bridge from infected LVAD to transplant. However, an additional sternotomy and days on temporary MCS with its innate risks could have been avoided if the current UNOS allocation system appropriately recognized central device infections with an equivalent acuity as it did in the 2006 system.

Conclusions

This case highlights several pitfalls within the 2018 UNOS allocation system, as BTT patients experiencing a life-threatening central LVAD infection may experience prolonged waiting as a result of the new allocation criteria to find a suitable donor. This may result in additional interventions being attempted to gain source control while simultaneously providing the necessary hemodynamic support. While our efforts proved successful in this case, the clinical implications of the additional steps needed to appropriately treat this patient as well as others like him call for efforts to redefine the 2018 UNOS allocation system to better match the needs of this critically ill population.

References 1. O’Horo JC, Abu Saleh OM, Stulak JM, Wilhelm MP, Baddour LM, Rizwan SM: Left ventricular assist device infections: A systematic review. ASAIO J 64: 287–294, 2018. 2. Zhou S, Yang G, Zhang M, et al.: Mortality following durable left ventricular assist device implantation by timing and type of first infection. J Thorac Cardiovasc Surg 2021: 2–4, 0153. 3. Patel CB, Blue L, Cagliostro B, et al.: Left ventricular assist systems and infection-related outcomes: A comprehensive analysis of the MOMENTUM 3 trial. J Heart Lung Transplant 39: 774–781, 2020. 4. Varshney AS, Hirji SA, Givertz MM: Outcomes in the 2018 UNOS donor heart allocation system: A perspective on disparate analyses. J Heart Lung Transplant 39: 1191–1194, 2020. 5. Cogswell R, John R, Estep JD, et al.: An early investigation of outcomes with the new 2018 donor heart allocation system in the United States. J Heart Lung Transplant 39: 1–4, 2020. 6. Jawitz OK, Fudim M, Raman V, et al.: Reassessing recipient mortality under the new heart allocation system: an updated UNOS registry analysis. JACC Heart Fail 8: 548–556, 2020. 7. Trivedi JR, Slaughter MS: “Unintended” consequences of changes in heart transplant allocation policy: Impact on practice patterns. ASAIO J 66: 125–127, 2020.