Veno-venous ECLS rescue for a heart transplant recipient with COVID-19, a case report

Abstract

The potential for increased rates of morbidity of SARS-CoV-2 within immunocompromised populations has been of concern since the pandemic’s onset. Transplant providers and patients can face particularly challenging situations, in the current settings as data continues to emerge for the prevention and treatment of the immunocompromised subpopulation. This case report details a patient 9-months post orthotopic heart transplant that developed SARS-CoV-2 infection despite two prior doses of the Pfizer-BioNtech COVID-19 vaccine, and had successful rescue from refractory hypoxemia with veno-venous extracorporeal membrane oxygenation (VV ECLS).

Introduction

Among those infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) roughly 14–33% progress to critical illness, and 67% of those develop acute respiratory distress syndrome (ARDS).¹ Refractory hypoxemia may develop despite mechanical positive pressure ventilation (PPV) in this setting. Extracorporeal life support (ECLS) provides ex vivo gas exchange using an artificial lung and has been utilized for more than 13,000 SARS-CoV-2 patients during the pandemic.² We report the use of ECLS for an orthotopic heart transplant patient with ARDS secondary to COVID-19.

Informed consent

Due to the retrospective nature of this case report it was classified “non-reviewable” by the Institutional Review Board. The patient did sign a disclosure statement in compliance with our corporate healthcare system.

Case

The case involves a 56-year-old male with ischemic cardiomyopathy which led to end stage systolic heart failure. Accompanying comorbidities included hypertension, chronic obstructive pulmonary disease, diabetes and ongoing tobacco abuse. The patient underwent placement of a HeartMate III left ventricular assist device (LVAD) as permanent therapy to manage refractory symptoms and the patient completely ceased tobacco use (Table 1). Two months after implantation, he developed a driveline exit site infection with staphylococcus aureus that was surgically debrided. Antibiotics, piperacillin/tazobactam and vancomycin, were prescribed as an outpatient. Three months post LVAD implant he presented to the emergency room with purulent drainage inferior to the mediastinal incision. He underwent multiple debridements of the infected durable LVAD pocket, ultimately culminating in complete explantation of the LVAD with bridging via a temporary axial LVAD.³ Impella support was continued 11 days as a bridge to orthotopic heart transplant (OHT).

Table 1.

Timeline.

Event	Duration from LVAD implant (months)
LVAD implant	0
Driveline infection	2.5
Explantation HM3 LVAD, impella 5.0 implant	5.2
Orthotopic heart transplant	5.6
1st dose Pfizer vaccine	14.3
2nd dose Pfizer vaccine	14.8
COVID + outside center	15.2
Hospital admission, negative AgFIA	15.6
Positive PCR	15.7
ECLS	15.7
ECLS wean	15.9
Discharge	16.1

AgFIA, antigen fluorescence immunoassay, HM3, Heartmate three; LVAD, left-ventricular assist device; PCR, polymerase chain reaction.

The patient was vaccinated with Pfizer-BioNtech COVID-19 vaccine at nine and 10-months post-transplant. His significant other developed COVID-19 at this time and she was the patient’s caregiver. He became symptomatic and 1-week after onset of symptoms, tested positive for SARS-CoV-2 at an outlying center. He received casirivimab and imdevimab (REGEN-COV, Regeneron Pharmaceuticals, Tarrytown, NY, USA) as an outpatient. A subsequent isolated pulse oximetry check at home resulted 79% and the patient was directed to the emergency room. Upon arrival, he was afebrile, with a dry hacking cough, and increasing shortness of breath. Sofia SARS antigen FIA test was performed with a negative result. The patient’s oxygen saturation was 87% by pulse oximetry (Table 2). He was placed on a non-rebreather mask which was later exchanged for bi-level positive airway pressure (BiPAP) ventilation. Remdesivir, dexamethasone, and sarilumab therapy were started and he was transferred to the intensive care unit. His immunosuppression regimen included: tacrolimus 1.5 mg twice a day and mycophenolate mofetil 500 mg twice a day. A chest X-ray (CXR) demonstrated bilateral diffuse airspace disease and ground glass type opacity concerning for pneumonia versus pulmonary edema.

Table 2.

Respiratory status.

Time	Ventilator or non-invasive ventilation	SaO2 or SpO2	PF ratio (mmHg:FiO2)	Dynamic lung compliance (ml/cm H2O)	Clinical events
Admission	Non-rebreather mask	95%	—	n/a	—
Pre-intubation	BiPAP FiO2 100%	88%	—	n/a	85% FiO2, SpO2 88%
Intubated, ECLS consult	Emergent intubation vent setting: Rate 24, Vt 500 peep 12, FiO2 100, PIP 25, mean 15 PRVC	88%	—	20	Hypotension secondary to hypoxia
Initial ECLS	Pressure control, FiO2 60, rate 10, peep 10	100%	—	20	—
Maintenance ECLS	HFNC 40L FiO2 40%	86–99%	—	n/a	—
ECLS decannulation	HFNC 40L FiO2 40%	91%	135	n/a	—
Discharge	6L nasal cannula	91%	135	n/a	—

BiPAP, bilevel positive airway pressure; FiO2, fraction of inspired oxygen; HFNC, high-flow nasal cannula; PEEP, positive end-expiratory pressure (cm H2O); PIP, peak inspiratory pressure (cm H2O); PRVC, pressure regulated volume control; SpO2, peripheral oxygen saturation; Vt, tidal volume (ml).

The subsequent morning, he was tachypneic with hypoxia. Repeat COVID-19 testing with PCR was reactive. Rapid intubation commenced and despite invasive ventilation his hypoxia remained refractory to intervention. Concurrent hypotension secondary to hypoxia was treated initially with 120 mcg/min phenylephrine infusion which was later switched to 0.04 units/min vasopressin infusion. A multidisciplinary decision was made to use VV ECLS as a rescue therapy. Cannulation proceeded in the catheterization lab percutaneously via the right internal jugular vein with a dual lumen ECLS cannula (31Fr Avalon, Getinge, Goteborg, Sweden). A small bolus of heparin was given prior to cannulation, but the patient was maintained only on aspirin oral 81 mg daily with 40 mg enoxaparin subcutaneous during ECLS support. From admission to cannulation anticoagulation consisted of 325 mg aspirin and 40 mg enoxaparin daily. Desmopressin acetate was given (2 mg every 6 h) on ECLS day 2. Ventilatory settings after initiation of ECLS were reduced to rest settings (mandatory rate 10⁻¹min, positive end-expiratory pressure (PEEP) 10 cm H₂O, drive 10 cm H₂O, FiO₂ 0.6). The following day he was successfully trialed on continuous positive airway pressure ventilation and was extubated to high-flow nasal cannula (HFNC). Transthoracic echocardiography revealed grossly normal cardiac function. On ECLS day-3 vasoactive drips were weaned, he was tolerating a regular diet and beginning to ambulate with assistance. The FiO₂ of the oxygenator was weaned slowly over the next 7 days. On ECLS day 9 the patient was decannulated (bedside, purse-string suture mediated closure), supported by HFNC 60% 40 LPM. Enoxaprin was discontinued after decannulation, and daily 81 mg aspirin was continued through discharge. Subsequently HFNC was replaced with nasal cannula and his ambulation continued to improve. Upon discharge the CXR findings included moderately severe diffuse peripheral infiltrates suggestive of pneumonia. Total length of stay (LOS) was 13 days. Because of the potential cardiovascular risk posed post-COVID,⁴ 1-week post discharge he underwent left heart catheterization (LHC) with intravascular ultrasound (IVUS) and echocardiography. The patient’s coronaries were free from significant stenosis by both LHC and IVUS, and a transthoracic echocardiogram confirmed normal biventricular function.

Discussion

The COVID 19 case fatality ratio (CFR) for the United States is 1.2% with 78 million cases reported.⁵ CFR for those hospitalized with COVID-19 has varied from as high as 23.2%–2.8%.⁶ Heldman et al. have reported hospitalized solid organ transplant recipients (SOTR) crude 28-days mortality 19.6% for 1 March 2020- 19 June 2020 and 13.7% for 20 June 2020-31 Dec 2020.⁷ Multiple studies have found no difference in 30-days cumulative incidence of mortality between hospitalized SOTR and non-SOTR.^8,9

However, COVID-19 infection and hospitalization in a narrower cohort consisting only of heart transplant recipients appears to carry substantial risk to survival. Within the small body of evidence regarding this population, overall mortality of approximately 15–25% and hospitalized mortality of 32% is reported.^10–12

Although our patient was in a moderately rapid spiral of decline, he was not profoundly hypoxic at the time of cannulation. Arterial saturations were approximately 87% by pulse oximetry supported by a ventilator at near-maximal settings. VV ECLS can be provided as a pre-emptive rather than salvage strategy with acceptable hazard in high-risk populations. Early intervention avoids deleterious ventilator settings that over time become unrecoverable. The trade-off for this avoidance of ventilator induced lung injury are ECLS-induced problems such as bleeding and thromboembolism. Use of a dual-lumen VV ECLS cannula approach facilitates reduction of sedation and makes physical therapy assisted ambulation possible, both of which may contribute to better outcomes.¹³

We chose an empiric dosing strategy for anticoagulation that consisted only of aspirin and enoxaparin. The most common complication associated with ECLS is bleeding.¹⁴ COVID-19 has been linked to thromboembolic events,¹⁵ but it is our opinion that the hazard of bleeding associated with ECLS outweighs more aggressive anticoagulation regimes.

This patient had a total LOS of 13 days which is substantially shorter than all of our other SARS-CoV-2 patients requiring ECLS. Interestingly a quicker rebound to pre-admission World Health Organization severity-score among SOTR has been reported.⁵ On admission the patient’s creatinine was 1.4 and it continued to rise pre-ECLS. For this reason, the patient did not have a CTA to exclude pulmonary embolism. However, there was neither RV-strain nor biventricular failure by echocardiography. This combined with the positive COVID PCR test suggests the primary etiology was neither pulmonary embolism nor myocarditis.

Although a reduction of the immunosuppression regime is commonly reported during COVID-19 infection of OHT patients,^6,7,16 we did not alter immunosuppressant dosing. The patient was maintained on mycophenolate mofetil and tacrolimus during his entire admission maintaining levels similar to pre- and post-hospital discharge. Consistent immunosuppression throughout the hospital course was verified using a T-cell immunoreactivity assay (ImmuKnow, Eurofins Viracor, Lenexa, KS, USA).

COVID-19 proved to be a terrible burden on healthcare systems and the populations they serve. The care of COVID patients, evaluation of strategies to treat the infection, and novel research all occurred in real-time during the pandemic. Placing a SOTR into such a scenario only added another layer of complexity. Since the first successful human organ transplant in 1954, healthcare providers have faced the infection/rejection challenge. If one is too immunosuppressed, they could suffer a fatal infection or paradoxically, underimmunosuppression could lead to organ rejection and/or graft loss. This report documents the successful use of an unaltered immunosuppression regime in conjunction with ECLS and minimal anticoagulation during COVID-19 ARDS.

ORCID iD

David A Baran https://orcid.org/0000-0002-7754-9953