MINIMALLY invasive cardiac surgery for coronary artery bypass grafting (CABG) and cardiac valve repair or replacement is a rapidly expanding field in cardiothoracic surgery. [1–3]Although it offers potential advantages to the patient and the health-care system, including improved cosmetic results, reduced pain, shorter hospital stays, and an earlier return to normal activity, it is not without its own risks.
We report a case of a temporary reduction in cerebral perfusion, detected by transcranial Doppler (TCD), resulting from endovascular aortic balloon clamp malposition during minimally invasive mitral valve surgery. Monitoring strategies to detect endoaortic clamp malposition are discussed.
A 51-yr-old man, with preserved left ventricular function and severe mitral regurgitation resulting from myxomatous degeneration, was scheduled to undergo mitral valve repair using a minimally invasive port-access system.
The use of a catheter-based endovascular cardiopulmonary bypass (CPB) system (EndoCPB system; Heartport Inc., Redwood City, CA) was planned for closed-chest CPB. This system (Figure 1) consists of a set of catheters, including femoral arterial (endoarterial return cannula), venous, coronary sinus (endosinus catheter), pulmonary artery vent (endopulmonary vent), and aortic balloon occlusion clamp (endoaortic clamp), and allows complete femoral bypass, aortic clamping, and antegrade (via a port on the aortic balloon clamp) and retrograde (via the coronary sinus catheter) cardioplegia. Venous drainage is increased using a centrifugal pump and additionally assisted with a multiorificed pulmonary artery vent.
After establishment of intravenous access and placement of a right radial arterial cannula, anesthesia was induced, and endotracheal intubation was achieved using a No. 39 left-sided, double-lumen endotracheal tube. Under fluoroscopic and transesophageal echocardiographic (TEE) guidance, the coronary sinus catheter and pulmonary artery vents were positioned. Bilateral TCD monitoring (Neurogard; Medasonics Inc., Fremont, CA) was established using two 2-MHz, pulsed-wave TCD probes with 18-mm sample lengths, gated at constant depths (range, 45–55 mm) secured to the head with a circumferential harness. TCD insonation of the left and right middle cerebral arteries (MCA), which is routinely used on our port-access heart surgery patients, demonstrated pulsatile flow of equivalent magnitude.
Surgery proceeded with a right anterolateral minithoracotomy to expose the heart and cutdowns for the placement of the femoral CPB cannula. After CPB commenced, the endoaortic clamp was advanced into the ascending aorta and positioned using fluoroscopic and TEE guidance. Non-pulsatile hypothermic (28 [degree sign] Celsius) CPB was used with flows of 1.9–2.4 l [center dot] min-1[center dot] m-2. Arterial blood gases were managed using alpha-stat methodology (PaO2, 148–187 mmHg; PaCO2, 37–41 mmHg).
Before cardioplegic arrest, the CPB flow was briefly decreased, and the endoaortic clamp was inflated. On restoration of normal CPB flow, the right radial arterial pressure remained low (less than 30 mmHg), and the TCD monitor demonstrated an asymmetric pattern (Figure 2) with a marked reduction in right middle cerebral artery blood flow velocity (Vmca). TEE imaging confirmed a malposition of the endovascular aortic balloon clamp with obstruction of the innominate artery. CPB flow was again briefly decreased, and the balloon clamp deflated, repositioned (advanced slightly toward the aortic valve), and reinflated, with restoration of a normal right Vmcawhen full CPB flow resumed (Figure 3). The duration of reduced right Vmcawas approximately 2 min. No further problems were encountered during surgery, and a successful mitral valve quadrangular resection with sliding ring annuloplasty was performed. The total CPB time was 196 min with an aortic endovascular clamp time of 134 min. Cardiopulmonary bypass was easily discontinued, and the patient was transferred to the intensive care unit after surgery. The patient awakened 4 h after surgery with no apparent neurologic deficit and recovered uneventfully.
This case outlines a potentially serious complication of endovascular aortic balloon clamp use for port-access heart surgery. Other reported complications from minimal access cardiac surgery have included wound infection, pulmonary complications, early recurrent ischemia caused by anastomotic stenosis or other vascular obstruction, sinus node dysfunction, 4* and low postoperative cardiac output. This is the first full description of endovascular aortic balloon clamp migration leading to temporary obstruction of cerebral blood vessels (although it is briefly mentioned as a complication in a recent article describing this procedure.)
Prompt detection with TCD (by a sudden reduction in Vmca) alerted us to this potentially serious problem. TEE imaging helped confirm the catheter malposition. TEE, however, may have difficulty identifying the position of the catheter in relationship to the aortic arch vessels because of imaging artifacts resulting from the interposition of the left mainstem bronchus between the esophagus and aortic arch. The use of combined monitoring modalities (TCD and TEE) allowed quick correction and rapid assessment of cerebral perfusion. Although it is unclear in this case whether temporary obstruction of the aortic arch vessels resulted in any cerebral ischemia, one may speculate that this risk was possible, particularly if more prolonged occlusion had occurred.
The relative risk of endovascular compared with traditional externally applied aortic clamping has yet to be defined. The risks of the endovascular balloon clamp may not only be limited to vascular malposition, as the potential for embolization of aortic plaque dislodged during balloon placement, inflation, and deflation is also possible.
As minimally invasive cardiac surgery continues to increase in popularity, monitoring practice guidelines will be developed. Several monitoring combinations could allow diagnosis of potential problems. The use of a right radial arterial cannula should facilitate detection of innominate artery obstruction. This may not always be the case, however, as a reduction in overall CPB flow may also cause a reduction in right radial artery mean arterial pressure (MAP), despite otherwise unobstructed flow. Bilateral radial artery cannulation may have a theoretical benefit in this situation as a difference in blood pressure between the two may indicate partial obstruction of the innominate artery. However, there are other reasons for blood pressure discrepancies between the two sites, particularly in patients with atherosclerosis. A baseline difference would make any changes more difficult to interpret. TCD monitoring, identifying a sudden asymmetrical velocity profile, proved useful in this case. However, bilateral TCD may not available in all centers, is expensive, and is operator-dependent with a significant failure rate in obtaining a good quality TCD signal in all patients. As an alternative to bilateral TCD, unilateral right MCA TCD or Doppler ultrasonography of the carotid arteries (unilateral right or bilateral) could also be performed.
Aortic balloon placement can be confirmed with intermittent TEE or fluoroscopy. The combined use of fluoroscopy, TEE, TCD/carotid Doppler, and right (or bilateral) radial arterial cannula may allow early detection of malposition or migration with possible prevention of cerebral ischemia. Electroencephalography (EEG) might also be used to identify arterial obstruction with the onset of an cerebral ischemic EEG changes. However, availability and the need for qualified interpretation limit its usefulness. Although several monitoring modalities can be used, alone or in combination, firm recommendations on the best strategies must await prospective evaluations.