Introduction

Endomyocardial biopsy has long been used in patients with cardiovascular disease to diagnose disease, guide treatments, and assist with prognostication. To better understand the role of this invasive diagnostic procedure, reviewing the history of the procedure, approach to patients, safety, and its utility are important. Assessment of the risk/benefit ratio of the procedure is dependent on understanding the nuances of our experience with its application. Clinicians, investigators, and professional societies have all addressed this issue, and guideline recommendations to help with expert clinical practice have been developed. Nonetheless, one will sometimes encounter a pejorative comment that endomyocardial biopsy is “a procedure looking for an indication.” That is not the case.

Development of the Procedure—A Historic Perspective

Linking pathologic information obtained during post mortem examination of the heart or tissue specimens removed at the time of cardiac surgery has given insight into myocardial pathology. Linking these findings to clinical scenarios is a foundation of medical practice. Histologic, histobiochemical, immune histochemical, and nuclear protein analysis of cardiac tissue can more precisely characterize disease in many patients premortem, particularly when a clinical diagnosis is challenging to make. However, examination of biopsy specimens could only be performed after cardiac tissue was obtained at the time of thoracic or beating heart surgery. In the 1950s, a limited thoracotomy approach was sometimes used to obtain samples of myocardium. The first percutaneous, nonsurgical, transthoracic, needle biopsy approaches to heart biopsy were reported in the late 1950s. These adventures were associated with an incidence of major complications including cardiac tamponade, coronary artery laceration, and pneumothorax in almost 10% of procedures. Direct percutaneous needle biopsy of the heart was limited by these problems as well as the difficulty of obtaining adequate tissue samples using relatively small needle puncturing and cutting devices as were used for percutaneous biopsy of the liver and kidney. The beating heart with a high pressure blood-filled system proved a difficult target. Multiple passes into the myocardium to obtain adequate tissue increased risk. Sutton’s report of a percutaneous transthoracic biopsy approach noted that insufficient tissue for examination was the outcome in almost a quarter of the patients. Often, no endomyocardial tissue could be obtained during these procedures. It was the development of a bioptome that could be utilized with a transvenous, or transarterial, approach to the right or left ventricle via the left basilica, and either femoral vein or from the left axillary, or either femoral artery and then into the heart that first allowed more reliable examination of adequate endomyocardial tissue. This facilitated diagnoses that were made after autopsy or direct surgical recovery of specimens. Sakakibara and Konno used a flexible bioptome with retractable and then grasping terminus cutting cusps to pinch off pieces of myocardium during an endovascular approach to the left ventricle. Subsequently, Caves, utilizing a modified Konno biopsy forceps (subsequently coming to be known as the Stanford Caves-Shulz bioptome, which evolved to the Scholten apparatus), to obtain tissue after passing the instrument through the right internal jugular vein with only local anesthesia used, and rapid tissue. This promoted procedure safety and success. The motivation for development of this device was diagnosing cardiac transplant allograft rejection so that immunosuppression could be modified in order to optimize transplant heart function and prevent or treat potentially catastrophic rejection. This tool was reusable and underwent subsequent modification that improved tissue biopsy while becoming the standard management system for heart transplant patients that subsequently morphed into multiple disposable and reusable instruments used today. The Stanford Caves-Shulz system had the advantage over the Konno bioptome in that it did not require a cutdown to facilitate entry into the saphenous or basilic vein (or the femoral or brachial artery if left ventricular tissue was the target). With fluoroscopic or echocardiographic (subsequently) imaging, the bioptome could be safely advanced to the endocardial surface with the jaws closed, the catheter slightly withdrawn and the jaws opened, followed by the catheter being re-advanced onto the endocardial surface (usually signaled by premature ventricular contractions, ectopy couplets, or short runs of nonsustained ventricular tachycardia). After closure of the clasps and then gentle tugging back the closed-jaw bioptome, withdrawal of tissue was usually successful and a few cubic millimeters in size. Crush artifact was not often a problem and, unless significant fibrotic endocardium was at the biopsy site, adequate samples of heart tissue were obtained. The Stanford Caves-Shulz bioptome could be utilized for many procedures (often more than 50) without reconstructive maintenance or sharpening. Subsequently, the King bioptome became utilized as well and was a modification of biopsy forceps used for bronchoscopic transbronchial biopsy. This flexible shaft bioptome could be introduced through a catheter into the right or left ventricle and, utilizing the same technique with the jaw-opening mechanism, allowed retrieval of several samples from slightly varying locations in the ventricle during several reinsertions and passes at the myocardium. Subsequent modification of biopsy catheters allowed disposal of single-use bioptomes and improved flexibility, which allowed greater percutaneous positioning and mobility of the catheter.

Technique of the Procedure

Today, the right internal jugular vein is most commonly accessed with a puncture and sheath-based catheter system for right ventricular endomyocardial biopsy. Some experts routinely approach the patient by accessing the femoral vein using longer bioptomes and sheath systems. This approach has the disadvantage of requiring the patient to remain supine for some time after the procedure to ensure that the puncture site does not bleed. Usually when doing the procedure from the neck or subclavian site, the patient can sit up and walk immediately post biopsy. Also, this approach, because the catheter system is quite long, makes manipulation and proper placement of the bioptome in the ventricle more challenging. Utilizing sonographic imaging of the internal jugular vein has remarkably decreased the challenges of doing endomyocardial biopsy. The number of attempts required for successful venipuncture, complications, and duration of the procedure have been improved with the ability to image the great veins of the neck and mediastinum by facilitating determination of vessel size, patency, and phasic respiratory variation of the target size and location. Of course, ultrasound guidance for femoral vein puncture can also be utilized and might be particularly helpful in patients who have unusual femoral triangle anatomy or scarring in this region from prior catheterization procedures. For central venous puncture some operators place the patient in a reverse Trendelenburg position during the internal jugular vein approach to facilitate venous engorgement, or simply place a wedging device under the patient’s legs. This should be removed after successful venous access has been achieved, particularly if right heart hemodynamic measurements are to be assessed. Having the patient do a Valsalva maneuver during the puncture in order to engorge and, thus, expand the internal jugular vein to a more sizeable dimension may facilitate safe and successful vascular access. During the procedure, the patient is monitored from an electrocardiographic, systemic blood pressure, and pulse oximetric perspective. If the internal jugular venous system is atretic or occluded for one reason or another, the subclavian vein is an option. With a natural “C” curvature formed with the bioptome, utilization of the left subclavian vein seems preferable to the right, although both approaches are relatively straightforward and feasible. If left ventricular tissue, for one reason or another, is desirable, the approach usually employs a sheath placed into a femoral artery. Constant positive pressure is maintained during this approach to avoid blood stasis and clot formation within the sheath with its risk of subsequent embolization, particularly when the system end-hole is in the left ventricle or aortic arch. The sheath system can be pushed across the aortic valve and positioned near the endocardium with the bioptome then passed repeatedly through it to obtain samples, again with care taken to avoid air or thrombus embolization.

Guidance of the biopsy catheter into the heart is generally done under fluoroscopic surveillance in a cardiac catheterization laboratory or radiologic procedure room but two-dimensional echocardiography has also been employed. Portable echocardiographic and fluoroscopic units can facilitate the biopsy procedures being done at the bedside, usually in an intensive care unit for critically ill, hospitalized patients. Arguably, fluoroscopy generally provides more information than echocardiography (some bioptomes commercially available are not very echogenic). With the patient in the PA (posterior-anterior) fluoroscopic view the course of the bioptome through the great thoracic vessels as it crosses the tricuspid, or aortic, valve and then enters the right or left ventricular cavity can be monitored ( Figure 34-1 ,Video 34-1A ). During fluoroscopy the image can be rotated into the LAO (left anterior oblique) position to assist with confirmation that bioptome jaws are facing the interventricular septum rather than the free left ventricular wall ( Figure 34-2 ,Video 34-1B ). The main challenge of echocardiographic-guided endomyocardial biopsy is imaging the bioptome, and particularly the jaws of the device. This can be a function of the device echogenicity and the ease of obtaining echocardiographic images in any particular patient. When successful images are obtained, however, endomyocardial biopsy can be made safer and provide better target images. An apical four-transthoracic echocardiographic view can allow a panoramic image of the right ventricle and help with target identification for the bioptome jaws. It is particularly important to avoid the free right ventricular wall, if at all possible, to decrease the likelihood of ventricular perforation. This is particularly important in the nonheart transplant patient (Video 34-1C ). Also helpful is the fact that echocardiographic images can identify the papillary muscles and give insight into where chordae tendinae might lay such that biopsy of these structures (particularly the tip of the papillary muscle) is avoided, specifically in the cardiac allograft where multiple biopsies over time are required. Another approach to lessening the risk of biopsy-induced tricuspid insufficiency is to utilize a longer sheath system, one which can be placed as a guiding sheath across the triscupid valve. This would likely prevent damage to the valve itself from the rigid and relatively unforgiving bioptome as it is pushed through the tricuspid valve. Perhaps this is more important when repeated biopsies are done on a cardiac allograft, but some would argue that it is more important when doing a biopsy in a native heart. Unfortunately, sometimes when the sheath is left across the tricuspid valve during the biopsy procedure it limits mobility of the bioptome when it is trying to be placed against the right ventricular target. With either fluoroscopic or echocardiographic guidance, after confirmation that the bioptome jaws are opposing the interventricular septum, the device is advanced into the septum, premature ventricular are induced (usually), with the jaws then quickly and firmly closed and specimen removed with a gentle tug (Video 34-1D ).

The posterior-anterior projection is used to position the bioptome in the right ventricle.

The left anterior oblique projection is used to confirm the bioptome is facing the right ventricular septum.

Employing both fluoroscopic and echocardiographic monitoring when a specific location in the ventricle is the target can be helpful ( Figure 34-3 ,Video 34-2 ). Prior to the biopsy procedure, computed tomographic (CT) or cardiac magnetic resonance (CMR) imaging of the heart might be helpful with making a pathologic diagnosis and give the operator insight into precise cardiac anatomy, particularly the angle of the intraventricular septum in relationship to the superior or inferior vena cava, and specific location of target areas for biopsy. It has been noted that knowledge of anatomic relationships characterized by multi-modal imaging may lessen the risk of inadvertent free right ventricular wall biopsy during a fluoroscopic-directed procedure, which could decrease the likelihood of right ventricular perforation and subsequent problematic pericardial effusion and cardiac tamponade. It should be noted that using MRI to help target focal disease location that might allow bioptome navigation into areas of interest is challenging. Three-dimensional echocardiography has also been utilized to assist with the procedure and obviate the need for fluoroscopy but limited data are available comparing this approach to the others.

Echocardiography shows that the bioptome is pointing toward the right ventricular free wall and needs to be repositioned toward the right ventricular septum.

Safety of Endomyocardial Biopsy

Today, the vast majority of endomyocardial biopsies are done to monitor cardiac allograft rejection after heart transplant. ^. This relates to the importance of knowing allograft rejection status, relative safety of the procedure, and reasonable diagnostic accuracy. As with any invasive procedure, the likelihood of an adverse experience or event seems dependent on the experience of the operator and team doing the procedure. This is further influenced by the clinical status or stability of the patient, bioptome utilized, vascular access approach chosen, presence or absence of left bundle branch block, location in the ventricle biopsied, and underlying disease. Table 34-1 summarizes potential difficulties occurring during endomyocardial biopsy. Risks that develop acutely include problems associated with vascular access such as inadvertent puncture of central or peripheral arteries, biopsy site hematoma formation, creation of peripheral arterial-venous fistula, or arterial-venous fistulae within the heart itself related to coronary artery injury. One usually transient and generally minor difficulty is nerve paralysis related to anesthesia infiltration of tissue overlying the vascular access site with dissection down the arterial-venous tissue sheath running into the mediastinum where recurrent laryngeal nerve paralysis can develop. Ventricular or supraventricular arrhythmias can be seen, including paroxysmal supraventricular tachycardia and atrial fibrillation (usually transient) as well as isolated ventricular arrhythmias, sustained ventricular tachycardia, and ventricular fibrillation with frank cardiac arrest (which is rare). Vasovagal reactions can occur in the nonheart transplant population where the hearts remain innervated, though rarely has been noted in patients with orthotopic cardiac allografts if re-innervation of the graft is present. Sudden heart block can develop and may be more frequent in patients with left bundle branch block, or trifascicular right bundle branch block. Pneumothorax and hemothorax are substantial risks but lessened by utilizing sonographic imaging of the neck and mediastinal great vessels to help guide venous access and catheter journey. Pulmonary embolization can occur, including air embolization. The operator must be ever mindful of this potential difficulty and take efforts to not allow air into the venous system, particular during deep inspiration by the patient. Long term, perhaps one of the more devastating difficulties can be disruption of tricuspid valve integrity, particularly after repeated right ventricular endomyocardial biopsies during follow-up of heart transplant patients. Indeed, tricuspid valve leaflets can be punctured or torn by the sheath, bioptome, chordae tendinae biopsy, or tissue removal from the tip of a papillary muscle. This can produce substantial tricuspid insufficiency. Utilizing a long sheath that can be placed into the right ventricle across the tricuspid valve may decrease the risk of tricuspid valve trauma; however, biopsy of tricuspid valve chordae tendineae may still occur even when a sheath is utilized. The operator must be ever alert to the possibility of bleeding into the pericardium, particularly when it is vigorous enough to cause cardiac tamponade after right ventricular free wall puncture. Fortunately, most right ventricular perforations seal off without sequelae after the culprit catheter is pulled back into the ventricle because of myocardial elasticity. However, the thin free wall of the right ventricle or areas of substantive scar tissue can produce challenges in some patients when perforation occurs. One must be vigilant postprocedure and remember that delayed complications can be seen and include access site bleeding (particularly when an artery is inadvertently punctured during the procedure, or the patient is anticoagulated or on antiplatelet agents), late cardiac tamponade, access site venous thrombosis, and, as mentioned, tricuspid insufficiency severe enough to cause hemodynamic alterations. The precise frequency of long-term adverse events, such as tricuspid regurgitation, is not well characterized.

Data are available to give insight into the complications noted during right ventricular endomyocardial biopsy. Overall, when large registry-based data are reviewed, the complication rate for access site acquisition is in the 2% to 3% range and with actual biopsy procedure 3% to 4% ( Table 34-1 ). The most common complication during access site approach is unplanned arterial puncture during infusion of local anesthesia or inadvertent insertion of access catheters, or bioptomes, into an artery (around 2%), vasovagal reaction (less than 1%), and prolonged venous oozing after bioptome and sheath removal (well under 1%). The most common adverse events associated with obtaining tissue is arrhythmia excluding isolated premature contractions (around 2% when conduction abnormalities are included), undiagnosed perforation manifesting as chest pain (less than 1% though more frequently patients report sharp pain or pleuritic pain when specimens are removed), and definite perforation manifest by pericardial fluid and rarely cardiac tamponade (also well under 1%). Despite these low numbers it should be noted that deaths can occur after perforation of the ventricle with cardiac tamponade or development of malignant, hemodynamically unstable arrhythmias. Patients having a higher risk of ventricular perforation include those with increased right ventricular systolic pressure, blood clotting abnormalities, on anticoagulants or antiplatelet agents, or having right ventricular enlargement, which can be associated with right ventricular wall thinning. Whenever the operator is concerned about myocardial perforation because of a pain syndrome, hypotension, or tachycardia, before central venous access is removed and the patient leaves the diagnostic laboratory, echocardiography should be performed to confirm, or refute, presence of pericardial fluid and document any imaging evidence of hemodynamic compromise. The capability to surgically address substantive pericardial effusions, particularly those with hemodynamic compromise, from a surgical or pericardiocentesis approach should be present and immediately available at any center performing endomyocardial biopsy.

Many risks can be diminished by utilizing ultrasound imaging for access site and catheter guidance, as noted, but also by avoiding the immediate supraclavicular approach with a higher internal jugular puncture technique. Rarely is heart block a permanent problem, but in patients with left bundle branch block, when the bioptome or guiding catheter is placed into the right ventricle and pressed against the intraventricular septum, it can develop. Most of the time simply pulling back on the sheath, or bioptome, is enough to allow normal conduction to reappear and only on rare occasions will a patient require temporary pacing. Horner syndrome with vocal paresis and sometimes diaphragmatic excursion limitation can be noted with large volume lidocaine infusion into the jugular venous and carotid sheath but does not generally result in permanent damage and resolves quickly. Trauma from the puncture needle itself can also cause this difficulty and, perhaps, more permanent impairment. Again, one of the troubling problems in heart transplant patients is development of tricuspid insufficiency, or worsening of this problem after exposure to multiple serial biopsies.