A Standardized Approach to Orthotopic (Life-supporting) Porcine Cardiac Xenotransplantation in a Nonhuman Primate Model

Cardiac xenotransplantation from swine has been proposed to “bridge the gap” in supply for heart failure patients requiring transplantation. Recent preclinical success using genetically modified pig donors in baboon recipients has demonstrated survival greater than 6 mo, with a modern understanding of xenotransplantation immunobiology and continued experience with large animal models of cardiac xenotransplantation. As a direct result of this expertise, the Food and Drug Administration approved the first in-human transplantation of a genetically engineered cardiac xenograft through an expanded access application for a single patient. This clinical case demonstrated the feasibility of xenotransplantation. Although this human study demonstrated proof-of-principle application of cardiac xenotransplantation, further regulatory oversight by the Food and Drug Administration may be required with preclinical trials in large animal models of xenotransplantation with long-term survival before approval of a more formalized clinical trial. Here we detail our surgical approach to pig-to-primate large animal models of orthotopic cardiac xenotransplantation, and the postoperative care of the primate recipient, both in the immediate postoperative period and in the months thereafter. We also detail xenograft surveillance methods and common issues that arise in the postoperative period specific to this model and ways to overcome them. These studies require multidisciplinary teams and expertise in orthotopic transplantation (cardiac surgery, anesthesia, and cardiopulmonary bypass), immunology, genetic engineering, and experience in handling large animal donors and recipients, which are described here. This article serves to reduce the barriers to entry into a field with ever-growing enthusiasm, but demands expertise knowledge and experience to be successful.


INTRODUCTION
The use of cardiac xenotransplantation has been proposed for patients with end-stage heart failure requiring transplantation, but without the ability to obtain an organ due to a shortage of supply. Life-supporting xenograft function in the preclinical setting has been obtained for up to 9 mo in a model of cardiac xenotransplantation, achieving the preclinical efficacy required for human clinical trials. [1][2][3][4][5][6] As a direct result of this expertise, the Food and Drug Administration (FDA) approved the first in-human transplantation of a genetically engineered cardiac xenograft through an expanded access application for a single patient. [7][8][9][10] Although this human study demonstrated proof-of-principle application of cardiac xenotransplantation, further FDA-monitored preclinical work in large animal models of xenotransplantation may be required using Good Laboratory Practices to support an FDA-supported human clinical trial. 11 Moreover, patient selection and surveillance are critical. [12][13][14] Preclinical studies in pig-to-baboon cardiac xenotrans-and postoperative care nuances that are critical to the success of this model. We have developed a protocolized approach to this model, incorporating our experience of failures but also successes in obtaining the longest surviving preclinical model of orthotopic, life-supporting cardiac xenotransplantation. In this article, we describe (1) the surgical procurement and implantation of pig hearts into nonhuman primates (NHPs) in the orthotopic position, considering nuances in pig and NHP anatomy, anesthesia, and cardiopulmonary bypass (CPB), (2) the medical and physical management of these animals in the immediate postoperative period and in the months thereafter, (3) postoperative surveillance of the cardiac xenograft, and (4) common problems and associated solutions. Our goal is to disseminate this information in the transplantation field, empower other laboratories to learn from our experience, and highlight potential management principles relevant to the clinical translation that are unique to caring for NHP recipients of a transplanted genetically modified porcine heart xenograft. All procedures described here have been approved by the Institutional Animal Care and Use Committee at the University of Maryland School of Medicine. All methods were carried out in accordance with relevant guidelines and regulations. The reporting in the article follows the recommendations in the ARRIVE guidelines. 15

Selection of Donor and Recipient for Surgery
Recipients are selected for low preoperative recipient titers for preformed antibodies 16 toward xenografts and adequate organ function as measured by a complete metabolic profile. Animals are also acclimated to the jacket, tether, and caging systems and removed as potential recipients if they fail acclimatization. Donors are weight-matched to the recipient. Weight-matching is performed in human allotransplantation by direct weight and sex-matching. Although no comprehensive analysis has been done in xenotransplantation, body weight is readily accepted as an index for heart size-matching, 17 as heart weight matches body weight across mammalian species independent of sex. 18 Donors are further selected for normal biventricular function without valvular abnormalities, septal defects, or pulmonary hypertension by screening echocardiography. 19,20 Selection for normal function without anatomic abnormalities is a standard screening method by echocardiography in heart allotransplantation, which is also employed here in our xenotransplantation program with success. [19][20][21] This ensures that genetically modified source animals do not have any off-target congenital abnormalities related to the heart. Donors within 10% of the recipient's body weight are typically acceptable.

Preparation of the Donor or Recipient for Surgery
The animal is first sedated to facilitate transport from the holding facility to the operating suite. The sedation is dependent on the animal. A pig is sedated with ketamine 10 mg/kg and xylazine 2 mg/kg intramuscularly, whereas an NHP is sedated using just 10 mg/kg ketamine. Before entering the operating room, the animal is shaved liberally with veterinarian hair clippers on the chest, neck, abdomen, back, and legs. The animal is then scrubbed with chlorohexidine, 4.0%, diluted 1:4 in all shaved areas and additionally on the hooves or hands and feet. The animal is then transferred to the operating room, where it is prepped and draped widely over the relevant operative field.

Preoperative Placement of the Tunneled Central Line in the Baboon Recipient
The placement of a tunneled central line is required to administer medications in the perioperative period, including induction immunosuppression. The recipient is positioned in a partial left lateral decubitus position with a shoulder roll to maximize the right lateral neck operative field. The right internal jugular vein is the optimal placement for the tunneled central line; however, the left internal jugular vein can also be used. The arms are fastened inferiorly to move the shoulders inferiorly and expose the lateral neck further.
The baboon is prepped and draped in a sterile fashion, and a 5-cm incision is made 1 cm medial and parallel to the medial border of the sternocleidomastoid. The platysma is exposed and divided with blunt dissection followed by retraction of the sternocleidomastoid laterally. At this layer, the external jugular vein can sometimes be seen. This is distinguished from the internal jugular vein, as it is more superficial and is not contained within the carotid sheath. The fascial layers are carefully divided, and small bridging veins are ligated down to the carotid sheath. The carotid sheath is incised, the internal jugular vein is separated from the vagus nerve and carotid sheath, and proximal and distal control is obtained with 2-0 silk ties.
Attention is then turned to tunneling the line and sizing the catheter. We prefer a 10Fr triple-lumen Hickman central line (Bard Access Systems, Inc., Salt Lake City, UT, Ref # 0606560) that comes with a plastic "tunneler," but any rigid narrow rod that can be manipulated under the subcutaneous plane of the animal will suffice. The tunneler is pushed from medial to lateral, starting from the subcutaneous plane in the lateral edge of the neck incision, and tunneled toward an exit site midpoint between the scapulae. The central line is then attached to the tunneler sterilely and pulled through the freshly made tract. If the line is cuffed, the line should pull through the skin and the cuff should lie subcutaneously. The distal portion of the line with infusion ports is now exiting the back and the proximal portion of the line is in the surgical neck incision, with a subcutaneous track connecting the 2 points. The central line is trimmed to allow the tip of the catheter to end at the cavoatrial junction, which is about the right fifth intercostal space at the costosternal junction. Fluoroscopy has been performed by us in the past to trim the catheter more accurately to length in heterotopic experiments, but in orthotopic cases, the catheter is trimmed to size during recipient cardiectomy.
When the catheter is appropriately tunneled and sized, the distal (superior) internal jugular vein is ligated, a generous venotomy is performed and the catheter is introduced proximally. The central line is then fastened proximally with the proximal 2-0 silk tie to ensure hemostasis. Several 2-0 silk sutures are tied along the length of the catheter in the neck to create a general inverted "U" loop to prevent line kinking with turning of the neck. Three silicon beads are put on the catheter before its insertion into the jugular vein. These beads are then secured to the fascia with silk sutures. This prevents lateral movement and kinking of the catheter with neck movement. The incision is then closed in several layers, with an emphasis on a fascial closure over the length of the catheter where possible to prevent catheter infection.

Procurement of the Donor Heart
Donor cardiac xenografts are procured from genetically modified pigs with multiple carbohydrate knockouts and knockins for human thromboregulatory proteins, human complement regulatory proteins, and human anti-inflammatory proteins to render xenografts more immunocompatible for cross-species transplantation. 22 Donor heart procurement has been described elsewhere, with international consensus guidelines on this topic. 23 Therefore, the principles of heart procurement will not be repeated here, but the emphasis on heart procurement to the extent that minimizes ventricular distention and ischemia cannot be understated. In cardiac xenotransplantation, primary graft dysfunction (ie, perioperative cardiac xenograft dysfunction [PCXD]) can occur in 40%-60% of transplantations and can be reduced with ischemia mitigation during this process. 24,25 The donor is prepped and draped from the edge of the mandible superiorly, 2-4 cm below the xiphoid process inferiorly, and from the anterior axillary line laterally. A procurement checklist is provided in Table 1.
Several things should be noted with emphasis for our approach to protection of the donor heart in cardiac xenotransplantation with pig donors. Induction of cardioplegia should be performed with a blood-based solution, ideally 30 cc/kg, 25% of which is fresh blood procured from the donor's aortic root. 26 Delivery into the clamped ascending aorta is via a 9Fr cardioplegia catheter connected to bladder irrigation IV tubing, pressurized with an infusion bag set to 150 mmHg. Perfusion pressures are titrated with the bladder irrigation tubing to 20-30 mmHg measured at the root by palpation ( Table 2). In our experience, this gives the best cardioplegia induction with appropriate arrest. Moreover, the aortic cross-clamp is best performed across the arch, from left to right, between the right and left innominate arteries, as the donor pig has a bovine arch (carotid and subclavian arteries come off brachiocephalic arteries bilaterally). The right brachiocephalic artery can be simply ligated before cross-clamping. Cardiectomy should be planned to minimize ischemic back table preparation before implantation. Therefore, we elect to vent the left side, only venting the right side if left or right ventricular distention is observed. If the right ventricle (RV) begins to distend, the superior vena cava (SVC) should be quickly opened superior to the azygos vein. If the left ventricle (LV) distends, the right pulmonary vein should be opened, cutting widely toward the hilum of the right lung.

Nonischemic Continuous Perfusion Preservation
Nonischemic continuous perfusion (NICP) has been shown to improve outcomes in cardiac xenotransplantation. 1,24,26 If electing to preserve the cardiac xenograft with NICP, a perfusion cannula is inserted into the aortic arch at the level of the right innominate artery. During donor explant, the aorta is transected at the level of the right innominate artery so that if the NICP cannula is too large, the transected aortotomy can be enlarged by incorporating the transected right innominate artery into a common orifice. The tip of the NICP cannula should be just above, but not through, the aortic valve to prevent incompetence. We use an XVIVO dual lumen cannula (central inflow, axially oriented outflow and quick connects for machine circuit connection) for NICP. Silicone tubing is fastened onto the left atrial cuff and placed through the mitral valve to render the valve incompetent and prevent LV distention during NICP. During implantation of the xenograft, we gently perfuse the heart with 50 cc of NICP perfusate every 15 min, or after each anastomosis, whichever comes sooner.

CPB Circuit
CPB circuits are configured based on the weight of the recipient. In 15-25 kg baboons, we use 1/4-inch (6.4 mm) and 3/8-inch (9.5 mm) tubing, for the arterial inflow and venous outflow ends of the circuit, respectively. In baboons <15 kg, we elect to use ¼-inch (6.4 mm) tubing for both the arterial and venous ends of the circuit to minimize the dilutional effects of CPB. We use centrifugal pump heads along with leukocyte filters and oxygenators as listed in Table 2. Priming volumes are based on the length and size of the circuit tubing and the oxygenator used, but generally, around 600 cc of crystalloid is required and is supplemented with 3000 units of heparin, 0.5 g/kg of mannitol, and 50 mEq of sodium bicarbonate. We elect to use retrograde autologous priming of the circuit instead of priming the circuit with blood products. When the heart is reperfused and before weaning from CPB, 1 mg/kg of lidocaine and 1 g of magnesium are administered.

Recipient Heart Explantation and Donor Implantation
Heart implantation has been described extensively elsewhere. 27,28 Here, we will emphasize the implantation process in the context of a pig xenograft donor biatrial implantation into an NHP recipient model. In clinical allotransplantation, studies have shown an advantage of the bicaval technique of implantation due to reduced tricuspid and mitral valvular insufficiency, arrhythmias, and improved atrial geometry. 29 However, we have seen an improved survival with the biatrial technique performed to avoid caval obstruction from anastomoses twisting and caval suture line narrowing. We have not encountered arrhythmias and valvular insufficiencies. Although the survival benefit is anecdotal, there are differences between humans' (NHPs) and pigs' superior and inferior vena cavae, which support our experience (in humans, the angle at which the cavae drain into the right atrium is 180°, whereas in pigs, this angle is about 90°). 30 The baboon sternum is particularly thin (close to 0.25 cm in 10-30 kg animals). It is important to use fingers to palpate at every rib space and define the midline superiorly and inferiorly to minimize malunion upon chest closure. The sternum is highly malleable, which makes it difficult to perform a midline sternotomy with an oscillating saw. We prefer a bicaval CPB cannulation technique with a right-angled metal-tipped cannula in the SVC and a straight venous cannula for the inferior vena cava (IVC). Approximate cannula sizes are listed in Table 3. The IVC cannula is placed to minimize the risk of IVC stenosis closer to the lateral border of the right atrium, just superior to the cavoatrial junction. We tighten caval umbilical tapes on Rummel sliders at the SVC and IVC cannulation sites.
We do not vent the left side of the xenografts during explantation or implantation. The recipient's cardiectomy is standard. The recipient atria are left as cuffs for the anastomoses, beginning with the right atrium. The RV is removed at the atrioventricular groove, just above the right coronary artery at the base of the tricuspid valve. The left atrial cuff is created by removing the LV from the left atrium, beginning at the dome of the left atrium posterior to the aortic root. The great vessels are separated close to their outlet valves. We prefer to leave the recipient's distal pulmonary artery (PA) long and shorten the donor's instead. This makes the PA-PA anastomosis implant easier to expose and less likely to injure the early branching right main PA.
We fashion a pericardial well from the anteriorly divided sac. The left edge is fastened to silk ligatures placed on hemostats. During the left atrial anastomosis, the pleural space can be opened, and the heart can be dropped into the left chest. This setup makes the initial posterior left atrial intima-to-intima suture line to be performed efficiently. The heart is rested on a cold saline-soaked sponge and can be relocated within the pericardium on subsequent anastomoses. The implantation sequence is performed with the left atrial anastomosis, followed by the right atrial septal anastomosis, and then the PA and aortic anastomoses. The cross-clamp is removed, and the aorta is de-aired with an 18-gauge needle followed by the remainder of the right atrial anastomosis to minimize warm ischemia time. We perform the atrial anastomoses with 4-0 double-armed prolene and the PA and aorta  with 5-0 double-armed prolene. The xenograft is defibrillated and temporarily paced as needed upon reperfusion and weaning from CPB. When the heart is full, the coronary sinus is examined, as an azygos vein drains into the coronary sinus in pigs. 31 This can be a major source of bleeding upon cardiac reperfusion if not dealt with at the time of procurement or during implantation. A telemetry device (DSI L21, New Brighton, MN) is placed in the subxiphoid preperitoneal space. One pressure transducer is placed in the aorta within the aortic cannula site after decannulation, and the other pressure transducer is placed in the right atrial appendage. The electrocardiogram lead is placed on the anterior surface of the right atrial appendage and the ground lead is placed on the LV near the apex. Up to 8 simultaneous implants can be transduced through 1 receiver (Figure 1). Chest tubes are placed in the bilateral pleural spaces and the pericardial well, connected to a Jackson-Pratt bulb ( Table 2). The sternum is closed with a 0-ethibond suture in an interrupted fashion. Fascial and skin closure is performed in several layers, taking special care to enclose the telemetry hardware in a separate fascial pouch in the subxiphoid space. A postoperative chest X-ray is performed to ensure there are no large undrained pleural effusions or pneumothoraces.

Immediate Postoperative Management (First 48 h)
The life-supporting model of cardiac xenotransplantation is difficult to manage. Even with perfect surgical technique and immune modulation, many physiologic perturbations occur postoperatively that require careful attention in an intensive care setting with frequent blood draw sampling hourly, and continuous monitoring by veterinary and clinical staff in the first 48 h. We have developed a postoperative protocol after extensive anecdotal experience with this model (Table 4). Following chest closure, we make every attempt at extubation within 2-3 h. If prolonged intubation is required, the recipient should remain sedated with fentanyl 2-4 µg/kg/h and propofol 1-4 mg/kg/h. Caution should be used with fentanyl boluses, as even at small doses, they can cause hypotension. Sevoflurane with a minimum alveolar concentration of 1.0-1.2 works well with fentanyl/propofol but may have some cardiac depressive effects. Once the recipient is extubated, it is transferred to our intensive care unit (ICU), which contains adjunctive medications, resuscitative fluids, and a working space for veterinary and clinical staff to care for the animal.
The ICU cages are climate-controlled and can administer a fractional-inspired oxygen concentration of up to 40% with dedicated high-efficiency particulate air filtration ( Figure 2A). It also has a tether system that allows for tunneling of the central line through a jacket and the cage wall onto a manifold that can be connected to medication administration pumps ( Figure 2B-D). Any postoperative drips and fluids can be administered, and frequent blood sampling can be employed.
Our preference is to begin fluids at 1.5 times the basal rate by ideal body weight and titrate up and down judiciously in the first few hours after transplantation based on fluid and acid/base status and the amount of resuscitation needed. We also empirically load the recipient with 1 mg/  kg of intravenous lidocaine intraoperatively and start an amiodarone drip at 1 µg/kg/min, since cardiac xenografts are prone to arrhythmias. Antiarrhythmics are weaned in a staged fashion if the animal demonstrates minimal to no ectopy in the first 12 h postoperatively. If ventricular arrhythmias occur despite these measures, lidocaine is preferred, as the porcine xenograft responds to this better than amiodarone in this setting. It can be given as an additional 1 mg/kg bolus, followed by a 20 µg/kg/min infusion. Defibrillator paddles are kept on the recipient animal, underneath the jacket, for the first 48 h after transplantation, with a defibrillator setup cage-side for swift access in the event of ventricular fibrillation ( Figure 2E). In our experience, these paddles were rarely used but save time if defibrillation is required in jacketed animals. Recipients of xenografts from swine are particularly susceptible to hypocalcemia, for reasons that are not understood. Therefore, calcium should be expected to be repleted at a rate of about 500 mg to 1 g of calcium gluconate every hour, driven by ionized calcium levels (goal ionized calcium levels are 1.0-1.2 mg/dL; Table 4). Potassium is repleted for levels below 3.2 mEq/dL at 1 mEq/kg infused in 50 cc of saline over 1 h, up to 20 mEq. Bicarbonate is repleted only in severe cases of metabolic acidosis in the absence of hypercarbic respiratory insufficiency. The bicarbonate deficit is calculated by the following formula: 1/3 × weight (kg) × base deficit. The deficit is repleted by 50% and a blood gas is repeated in 1 h to measure response. It should be noted that the cardiac xenograft can exhibit marked sinus tachycardia, but is left untreated, with heart rates of 148 ± 14 beats per minute in our most recent cohort. 2 Moreover, common electrocardiographic morphologies are noted in Figure 1 and should be considered "normal variants" of a sinus rhythm.
We have protocolized sedating and examining recipients each morning in the first 48 h postoperatively to assess cardiopulmonary and volume status more appropriately and to manage chest tubes. Although there is a telemetry pressure FIGURE 2. ICU setup. A, Climate-controlled intensive care unit cage with high-efficiency particulate air filtration, ability to provide ambient oxygen saturation of up to 40% fractional inspiration of inspired oxygen, and a urine collection system. B, Example of IV catheter setup with numerous pumps delivering medications. C, Example of a manifold on the external wall of the cage, which connects IV medication delivery pumps to the tether system. D, Recipient jacket connected to the tether with a swivel allowing delivery of medication without supercoiling. E, ICU cart allowing rapid access to frequently used medications, supplies, and external defibrillator. ICU, intensive care unit; IV, intravenous. transducer placed in the right atrial appendage, central venous pressure is somewhat unreliable in an awake, nonsedated, and ambulatory baboon transplant recipient. Therefore, volume status is assessed based on a chest X-ray, an ultrasound, and a physical exam. We examine the xenograft by echocardiography and the lungs by portable chest X-ray and strip the chest tubes and let them drain the pleural spaces entirely. The chest tubes remain until the total pleural fluid from each drain is <50 cc in a 24-h period. Generally, if >100 cc of fluid is collected in total, the drains will stay collectively for another 24 h, until the tubes can be reexamined and a repeat X-ray can be performed. Generally, our recipients can be transferred out of the ICU the morning of postoperative day 2 and by that time they are typically off any supplemental medications, drips, or fluids and are receiving only maintenance immunosuppression. Our detailed immunosuppression regimen has been described previously. 32 Anticoagulation is started at 50 units per hour and then titrated slowly to an ACT of twice the baseline when surgically appropriate, usually by postoperative day 5.

Long-term Postoperative Management and Graft Surveillance
During the first 2 wk postoperatively, animals are being monitored continuously by cameras and telemetry. The telemetry readings sometimes raise false alarms due to loss of connection with the receiver and the baboon status is confirmed by observing them on the camera. Both cameras and telemetry readings can be accessed remotely and are also recorded. The laboratory examination is performed 1-2 times per week and then twice monthly for the first month. Thereafter, laboratory examination is done every month and reserved for clinical status changes. Standard laboratory examination includes a complete blood count, a complete metabolic panel, and an activated clotting time. Donor-specific assays to measure anti-xenograft antibodies and longitudinally sampled troponins are also employed to guide the early detection of potential antibody-mediated rejection. 3,16 More novel xenograft-derived cell-free DNA assays have also been developed as a sensitive assay to monitor for rejection. 33 Recipients are visually examined daily for behavior, appetite, abdominal distention, respiratory effort, global and dependent edema, and normal stool and urine production. Cage and jacket changes are done every 2 wk and in doing so, the animal is sedated and a full physical exam is performed. Special attention is given to the animal's weight, body condition, surgical incision, central line exit site, mucosal membranes, and abdomen. An echocardiographic examination is also performed in parasternal short, parasternal long, and 4-chamber views ( Figure 3A-D).
In addition to telemetric and echocardiographic data, we have started to employ endomyocardial biopsies, right heart catheterizations, and coronary angiograms to our surveillance regimen through femoral venous and arterial access sites. Endomyocardial biopsies of the RV and right heart catheterization for pressures are also planned if concerned for rejection, based on clinical status, donor-specific antibodies, and troponin elevations. We perform histology with hematoxylin & eosin stains and immunohistochemistry (CD3, CD20, CD68, CD3d, and CD4d) on biopsy specimens and analyze for rejection based on allotransplantation and xenotransplantation criteria. 34,35 Although the specifics on how to perform these procedures are not detailed here, the supplies required to perform these procedures specifically for cardiac xenotransplantation are detailed in Table 5. Our preference is to perform these surveillance adjuncts every 2 mo starting 30 d after transplantation.

Common Problems in Pig-to-baboon Cardiac Xenotransplantation
The tunneled central line, jacket, and associated tether system are the most common problems in this baboon model. All animals are acclimated to the system 1 wk before the placement of a central line, and animals that show signs of stress, intolerance to the system, or are otherwise incapable of keeping their jacket on and intact during this time frame are excluded from the study. Despite this, however, these animals have insatiable curiosity and are continuously pulling and jerking the tether. For the tether to function properly, several joints have to be properly secured. The transplanted baboons, due to continuous activity and manipulation of the tether, make these joints malfunction, which can result in supercoiling of the tether, damaging of the central line, and repeated manipulation of the line, which can result in infection. Each of the following is addressed below: 1. Supercoiling tethers: The tether system is depicted in Figure 2B and C, which consists of a tunneled catheter, protected by a jacket worn by the recipient, threaded through a metal tether system that houses the catheter and 4Fr silicon intravenous (IV) tubing. The IV tubing connects through a swivel, to the outside of the cage with a multichannel pump system for drug administration and blood draws. The tether system is predicated on the ability to have one fixed end (closest to the jacket) and one end attached to a swivel. Therefore, any rotation of the IV line within the tether is prevented by the axial tension created by the beginning of a supercoil. However, when this system fails, axial rotation of the IV line within the tether is created, and a supercoil develops with the recipient's free movement around the cage system ( Figure 4A). This supercoil can cause an obstruction or can lead to the line being pulled off from its attachment at either end. The most common indication that is unique to supercoiling is the inability to either flush the line or draw back blood. This can largely be prevented by daily visual inspection of the swivel and ensuring that it can freely rotate around its axis with minimal interference. If a supercoil does occur, the animal must be sedated and the entire line and tether system must be replaced in a sterile fashion. Replacement of the tunneled central line in the contralateral jugular vein is required to ensure timely delivery of the maintenance immunosuppression. 2. Damage to the central line: NHPs can gain access to the central line (or other devices such as power or pump cords) by means of either damaging the pouch on the back of the jacket, which houses the infusion ports of the central line, or by disconnecting the metal tether from the swivel and exposing the silicon IV tubing. In animals that have demonstrated the ability and desire to disconnect the tether, we reinforce the distal and proximal ends of the tether with steel ( Figure 4B). Although performed on a per-animal basis, we have had good success with reinforcement of the tether base when required. 3. Central line infections: central lines are the most common sources of infection with our recipients and are difficult to treat. Moreover, these animals are all severely immunosuppressed and require daily medication administration, precluding any ability to perform line holidays or line removal without an immediate replacement. Therefore, line manipulation should be minimized and if required, sterility is paramount. We also have developed a protocol for prophylaxis, which consists of an antibiotic and heparin lock for any line manipulations, including animal transport, cage changes, or catheter/tether repair. Formulations are based on historic bacterial susceptibility and previous literature on this topic. 36,37 Any recipient with a fever (indicated by elevated temperature readings on the telemetry) and outside of the immediate postoperative phase has their blood cultured for bacterial and fungal pathogens, started on broad-spectrum antibiotics, and narrowed based on the presence or absence of the cultured pathogen.  4. Blockage of the central line due to precipitation of infused drugs: Some of the infused drugs in our regimen, such as mycophenolate mofetil and ganciclovir, are incompatible with heparinized saline. This incompatibility leads to the precipitation of these drugs in the IV tubing when infusing concurrently. We routinely reconstitute all of our drugs in a 5% dextrose solution, which prevents drug precipitation. 5. Central line dislodgement: A very rare complication is when the central catheter is either dislodged from the internal jugular vein or fluid begins extravasating at the insertion site. In this case, the animal demonstrates an accumulation of fluids in the cheek pouch. A definitive diagnosis is made via fluoroscopy when contrast is visualized leaking around the insertion site of the catheter. This catheter requires immediate replacement to ensure proper medication infusion and to maintain hemostasis.

Nutritional Supplementation
Weight loss does occur in most recipients after transplantation. Our Institutional Animal Care and Use Committee regulations necessitate a 20% weight loss from the time of animal study enrollment as an endpoint requiring euthanasia, regardless of the clinical status of the animal. Therefore, we have protocolized our care of these animals to minimize weight loss.
All animals are placed on a high-protein diet with Teklad Global high-protein biscuits (Inotiv, Inc., West Lafayette, IN) and given diet enrichment of high-calorie snacks, to maintain body weight after transplantation. Animals are monitored for biscuit intake daily, with a goal target caloric intake of 25 kcal/kg/d and total protein intake of 1 g/d. If an animal reaches the threshold of 10% weight loss, appetite stimulants are administered, and potential causes of weight loss and decreased appetite are investigated. If an animal reaches a threshold of 15% weight loss, intermittent bolus tube feeds twice weekly are administered by gavage feeding to supplement nutrition.

RESULTS AND DISCUSSION
We have developed a protocolized approach to performing cardiac xenotransplantation of genetically modified xenografts in an NHP recipient model culminating from years of experience. It requires a multidisciplinary team of scientists, immunologists, clinical staff, surgeons, and perfusionists with expertise and experience working with NHPs and heart transplant recipients. 38 However, this is a robust preclinical model of cardiac xenotransplantation with immediate translational potential even if appropriate FDA approval is required. [39][40][41][42][43] Moreover, postoperative protocolization, although specifically optimized for cardiac xenotransplantation, can be applied to cardiac allotransplantation and other disciplines of transplantation using NHP models.
This serves as a model for life-supporting, orthotopic transplantation that emulates clinical heart allotransplantation and has afforded the opportunity to examine several outstanding questions in cardiac xenotransplantation. Längin et al elegantly demonstrated that nonischemic continuous preservation (NICP) can overcome PCXD almost universally, which was previously an insurmountable barrier to clinical translation. 1,24,25 We have demonstrated that alternative strategies to NICP that minimize ischemia-reperfusion injury may warrant further investigation and signal toward a potential underlying mechanism of PCXD. 26 NICP has also afforded our group the opportunity to use this model to assess the progressive and iterative genetic modifications of cardiac xenograft donors and their effect on improving overall xenograft survival. 3 We have demonstrated that select "multi-gene" cardiac xenografts, with multiple knockouts and human transgene expression of complement regulatory, thromboregulatory, and anti-inflammatory proteins have preserved life-supporting function up to 9 mo. 3 We have also demonstrated that physiologic mismatch is present in xenografts after transplantation and that posttransplantation xenograft growth can be reduced with growth hormone receptor knockout donors. 2 Current studies are focused on understanding the contributors of life-limiting posttransplantation xenograft growth in this model.
Principles of cardiac xenotransplantation in NHPs include procurement and implantation that minimizes ischemia, early extubation of the recipient, and postoperative care that emphasizes repletion of calcium. Antiarrhythmics, frequent physical exams, and noninvasive diagnostic imaging are important adjuncts in the first 48 h. Long-term surveillance should include a multimodal approach, including monitoring of donor-specific antibodies, troponin I, echocardiography, and endomyocardial biopsies. Any elevation in troponin or donor-specific antibodies should prompt swift treatment for presumed rejection and an endomyocardial biopsy for a definitive diagnosis. Although rejection can be difficult to treat in this model, efforts include pulse-dose steroids, reinduction with rituximab, and an extra dose of costimulation blockade to target presumed antibody-mediated rejection.
The technical approach and postoperative care of NHPs are cumbersome but necessary for demonstrating the preclinical efficacy of cardiac xenotransplantation. The International Society of Heart and Lung Transplantation has recommended that a survival of 90 d in 6 of 10 baboons in a study of multitransgenic source animals is required. 4 Therefore, the FDA will likely require continued preclinical research despite the first in-human experience of cardiac xenotransplantation. 44 This article serves as a guide for ensuring quality studies with optimal success to achieve this goal.

CONCLUSIONS
The pig-to-baboon model of cardiac xenotransplantation continues to be a challenging, but prototypic model for demonstrating preclinical efficacy for the use of cardiac xenografts from swine for end-stage heart failure, without the ability to obtain an allograft or durable mechanical circulatory support. Specific keys to success include a protocolized approach to the procurement of the xenograft, implantation, and postoperative care of the animal in the acute and long-term phases of care. Many questions remain to characterize cardiac xenografts in the foreign environment of a xenogeneic recipient, and surveillance measures will also be a key to demonstrating preclinical efficacy before clinical translation, which have been summarized here as well. Moreover, once the clinical translation is realized, the principles of cardiac xenograft care and surveillance described here can be used as a template for the eventual human application of this novel source of solid organs for transplantation in patients with failing hearts and no alternatives.