Impact of a Large Hemodialysis Arteriovenous Fistula

Abstract

Arteriovenous fistulae have a substantial impact on systemic hemodynamics, but their effect on cardiopulmonary bypass is not well understood. The left-to-right shunting of blood through an arteriovenous fistula can create flow problems during cardiopulmonary bypass. We present the case of a 59-year-old man with end-stage renal disease who underwent coronary artery bypass grafting. During surgery, a large brachiocephalic arteriovenous fistula provided excessive venous return to the heart, compromising the myocardial protection offered by cardiopulmonary bypass. We modified the venous cannulation from a 2-stage single cannula to a bicaval cannula and were then able to achieve cardioplegic arrest and myocardial protection with no further problems.

Cardiac surgery is being performed with increasing frequency in patients with end-stage renal disease (ESRD) who are dependent on dialysis. The dialysis procedure requires access to the vascular system. A surgically created forearm (radiocephalic) or upper-arm (brachiocephalic) arteriovenous fistula (AVF) is preferred for access rather than a central intravenous catheter because of the need for long-term use. During cardiopulmonary bypass (CPB), however, flow through such an AVF may pose additional problems.

The effects of a surgically created AVF on cardiovascular hemodynamics have been well documented.¹ However, such effects during CPB have not been fully described. The English-language medical literature contains only a few reports regarding the effects of AVF flow during CPB.^2,3 We present an unusual case in which a large brachiocephalic AVF provided excessive venous return to the heart during CPB, preventing adequate myocardial protection.

Case Report

A 59-year-old man with ESRD was admitted to our hospital with chest pain. Coronary angiography revealed 3-vessel disease, and elective coronary artery bypass grafting was planned. The patient had been on a hemodialysis regimen through a left brachiocephalic AVF for the past 5 years. The average access flow during hemodialysis was 900 mL/min. Echocardiography showed a left ventricular ejection fraction of 0.49.

During the operation, the left internal mammary artery (IMA) and saphenous vein were harvested. Venous cannulation was accomplished with a 36F × 51F 2-stage venous cannula inserted in the right atrial appendage, as is routinely done in coronary artery bypass operations. The patient's body surface area was 1.72 m², and the pump flow was about 2.2 L/min/m². A cross-clamp was applied to the aorta. Initially, St. Thomas's II cardioplegic solution was delivered antegrade, then retrograde. However, the heart was not arrested and was overdistended despite delivery of an adequate amount of cardioplegic solution. We verified that the venous cannula was properly positioned, and we increased venting of the heart through the aortic needle to prevent distention. Nevertheless, after 1,500 cc of cardioplegic solution had been delivered, electrical activity was still significant. Repeated attempts to completely drain excessive blood flow from the AVF with use of the 2-stage venous cannula were unsuccessful. Temporary closure of the arteriovenous access by a sphygmomanometer might have enabled successful cardioplegia delivery, but we chose selective bicaval venous cannulation and avoided repositioning the patient's arm.

The aorta was declamped to prevent ischemic injury. Selective bicaval venous cannulation was established rapidly through a 26F cannula for the superior vena cava and a 28F cannula for the inferior vena cava. The cannulae were then taped and snared, and the cross-clamp was reapplied. The CPB flow was not altered. This time, the heart was not distended, and it was arrested successfully with cold-blood cardioplegic solution. Five distal anastomoses were performed, and the cross-clamp was removed. The heart started beating spontaneously. A side clamp was applied for the proximal anastomoses. Then both caval tapes were released. The patient was weaned from bypass successfully with minimal inotropic support and was transferred to the intensive care unit in stable condition. The AVF was functioning. A strong thrill was felt on palpation, and a loud bruit was audible on auscultation. The patient was discharged from the hospital without complications on the 5th postoperative day. At the 2-year follow-up visit, his clinical status was excellent.

Discussion

Hemodialysis-dependent patients who undergo cardiac surgery present many challenges to physicians, and these challenges are well documented.¹ Creation of a hemodialysis AVF has a significant impact on cardiovascular hemodynamics. The increased flow through the AVF can lead to a cascade of complications, including increased cardiac and pulmonary artery pressures, myocardial decompensation, decreased left ventricular function, and, finally, high-output cardiac failure (defined as symptoms of cardiac failure in the presence of an above-normal cardiac index [≥3 L/min/m²]).¹ However, reports about the effects of a surgically created hemodialysis AVF during CPB are few.^2,3

The impact of surgically created AVFs on bypass graft function has been examined in several reports.^4–6 An upper-extremity fistula and an ipsilateral IMA arise from the same vascular root. Therefore, the left-to-right shunting through the AVF has the potential to reduce ipsilateral IMA flow, which can lead to IMA steal and, subsequently, to reduced coronary flow after revascularization with the IMA. Using transthoracic color-Doppler ultrasound imaging, Rahbar and colleagues⁴ analyzed IMA flow changes in patients with upper-extremity AVFs and concluded that such AVFs have a modest hemodynamic effect on the in situ IMA. Kato and associates⁵ detected the steal phenomenon with use of angiography in the left IMA during diastole and decided to use the right IMA in their patient.

Nyawo and colleagues³ reported the effects of a hemodialysis AVF during CPB. They focused on compromised coronary perfusion caused by a large proximal AVF with a high flow rate, which led to failed attempts to wean their patient from CPB. They explored and tied off the fistula, and were then able to wean the patient from bypass successfully. Their case underscores the possibility of adverse effects of an AVF during CPB. However, in the absence of high-output cardiac failure, we believe that closure of the AVF should be considered only as a last option.

The possible effects of an AVF on hemodynamic values during CPB need further investigation. To our knowledge, ours is the first report to emphasize the effect of a high-flow fistula on cardioplegic solution delivery and myocardial protection during coronary artery bypass grafting in a patient who is dialysis dependent. Excessive venous return through a high-flow AVF may warm the myocardium, so that the delivered cardioplegic solution does not arrest the heart, and inadequate myocardial protection could result. We propose that selective bicaval venous cannulation is a feasible way to prevent cardiac distention caused by the shunting of blood from left to right through the AVF. This simple modification should be considered for dialysis-dependent patients undergoing cardiac surgery.

References

Editorial Commentary

As chronic renal dialysis becomes more commonly associated with myocardial ischemia, more patients with an arteriovenous fistula will be referred for surgery under cardiopulmonary bypass. The arterial flow through the fistula during nonpulsatile bypass can be substantial and may interfere with normal techniques of myocardial protection during standard aortic cross-clamping.

Cetin and colleagues found that using a customary 2-stage cannula for right atrial drainage did not prevent atrial and cardiac distention. Converting to bicaval drainage solved the problem of overdistention of the heart. I am concerned, however, that the open fistula could deprive the upper body and brain of vital circulation and perfusion.

Anticipating the problem before cannulation should provide a simple solution. Using an inflatable cuff above the fistula would prevent the excessive flow by occluding the arterial inflow. The solution that the authors used is interesting and should be remembered by surgeons and perfusionists when confronted with such cases.

Denton A. Cooley, MD,
Founder, Surgeon-in-Chief, and President Emeritus, Texas Heart Institute at St. Luke's Episcopal Hospital, Houston