Left ventricular assist devices (LVADs) have demonstrated significant improvement in both quality of life and mortality in patients with advanced heart failure.1–3 Continuous-flow (CF) LVADs are now used for long-term support because of smaller size and increased durability. LVADs are primarily used as a bridge to cardiac transplantation or as destination therapy in the case of patients with contraindications for transplantation. In both cases, comorbid noncardiac conditions requiring surgical intervention are increasingly becoming common. This is especially true in the destination-therapy population as they tend to be older and have more comorbidities. Previous literature on the subject of noncardiac operations in LVAD patients consisted of mostly small case series and case reports.4–8 Our study reports mortality and the perioperative management of a large population of CF-LVAD patients undergoing noncardiac operation.
We retrospectively reviewed data from 110 consecutive patients at our institution who received a HeartMate II (HMII; Thoratec Corp., Pleasanton, CA), a CF-LVAD, between September 2004 and June 2010 for destination therapy. Figure 1 shows the reasons why these patients were not transplant candidates. All the patients had the outflow graft implantation to the ascending aorta. Patients with tracheotomies, thoracenteses, and mediastinal and driveline/pocket explorations (except patients requiring omental flap as part of surgical treatment for pocket infections) were excluded from the study. Preoperative laboratory data including complete blood count, blood urea nitrogen, creatinine, and international normalized ratio (INR) were collected. Data regarding type of anesthesia used, methods of intraoperative monitoring, and number of blood products transfused during the perioperative period were obtained from the anesthesia record. Details about the number of blood products transfused within the first week postoperatively were obtained from medical records. This study was approved by the institutional review board.
Clinical characteristics were compared using the Student’s t-test for continuous variables and chi-squared or Fisher exact test for categorical variables. All analyses were performed using SPSS version 11.5 (Chicago, IL). A p value less than 0.05 was considered statistically significant.
Thirty-six patients underwent 63 noncardiac surgical procedures after LVAD implantation. Of these, 81% were male patients, with a mean age of 61.4 ± 11.4 years, and 51% had nonischemic cardiomyopathy. Of the 36 patients in the study, 23 (63.8%) patients required one surgical procedure, and 13 (36.1%) patients underwent more than one procedure. Types of procedures are listed in Table 1. There were 24 (38%) total abdominal operations, 9 (14%) vascular operations, 9 (14%) urologic operations, 7 (10%) oral operations, 4 (6%) neurologic operations, 3 (5%) orthopedic operations, 2 (3%) thoracic operations, 2 (3%) head and neck operations, and 3 (5%) plastic operations. Fifty-four (84%) operations were elective and 10 were emergent. All 63 operations were performed in our LVAD implanting center.
The median number of days between the patients’ LVAD implantation and their noncardiac operation was 137 days (10–1016 days). Sixty-eight percent of the patients underwent noncardiac operation within the first year of receiving their LVAD.
In all patients undergoing intraabdominal operations, antibiotic prophylaxis was used. This typically consisted of intravenous vancomycin and cefazolin administered before operation and continued for 24 to 48 hours postoperatively. In all other operations, antibiotics were used at the discretion of the surgeon performing the operation.
Our usual LVAD anticoagulation protocol consisted of warfarin doses to achieve a target INR of 1.5–2.5 similar to that in published reports.9 Aspirin 81 mg was also administered. Warfarin and aspirin were stopped before operation. Baseline coagulation profiles, including prothrombin time and INR values, were obtained in every patient. Laboratory testing was performed the same day as the procedure or the previous day. Preoperative hemoglobin, hematocrit, white blood cell count, platelet count, blood urea nitrogen, creatinine, and INR are listed in Table 2. To reverse elevated INR levels, fresh frozen plasma (FFP) was administered; cryoprecipitate or Vitamin K was used in one case. Ideally, this was done before operation; however, in emergent cases, this was done during and after operation. Warfarin was always stopped at least 3 days before elective operations. Usually, anticoagulation was resumed within 2 weeks, depending on the underlying comorbidities and postoperative bleeding.
Of the 63 procedures, 56 (88.8%) were performed under general anesthesia. Monitored anesthesia care was used during seven oral and vascular procedures (11%). Spinal anesthesia was used once for a fulguration and transurethral resection of residual prostatic tissue. The choice of induction and maintenance anesthetic agents was similar to that used in patients without LVADs.
Intraoperative Transfusion of Blood Products
The median number of blood products transfused intraoperatively is shown in Table 3. Patients received blood-product transfusions intraoperatively based on the type of operation and associated comorbidities. Blood-product transfusions consisted of packed red blood cells (PRBCs), platelets, cryoprecipitate, or FFP.
Packed red blood cells were transfused in 24 of 63 (38%) procedures. Nine of the 24 (38%) procedures requiring PRBCs were gastrointestinal (GI) operations. More than 5 units of PRBCs were needed in three (5%) procedures. The average preoperative INR in cases requiring transfusion of blood products was 1.4 ± 0.7 compared to 1.4 ± 0.3 (p = 0.93) in cases that did not require transfusions. The maximum amount of PRBCs transfused was 7 for a single operation involving profuse bleeding during multiple tooth extraction. All transfusions were for intraoperative blood loss. The estimated blood loss (as documented in the operating notes) in operations requiring transfusion of PRBCs was 176.96 ± 292.03 ml, which was higher than the estimated blood loss of 139.64 ± 397.05 ml(p = 0.76) in patients who did not require PRBCs.
Fresh frozen plasmas were transfused in nine (14%) operations. The average preoperative INR in patients receiving FFP intraoperatively was 1.5 ± 0.5, compared to 1.2 ± 0.1 (p = 0.03) in patients not receiving FFP. Six procedures (9%) received platelet transfusions, and two (3%) procedures required cryoprecipitate as a result of bleeding. Operations requiring blood-product transfusions included abdominal, neurologic, oral, vascular, orthopedic, urologic, and thoracic operations. The average estimated overall blood loss was 160 ml per operation.
Eighteen (28%) operations were monitored with arterial lines. Ten (16%) operations used pulmonary artery right heart catheter monitoring. All 10 operations requiring pulmonary artery catheters were also monitored with arterial lines. Forty-six (72%) operations did not require invasive monitoring either by arterial lines or pulmonary artery catheters. In these procedures, measurement of blood pressure was obtained via an automated oscillometric device every 5 minutes. The mean arterial pressure was maintained above 70 mmHg during the procedures. All patients were placed on the stationary power base unit and LVAD parameters were monitored using a clinical screen that provided measurements of flow, pulsatility index (PI), power consumption, and LVAD speed (rotations per minute).
In majority of the cases, LVAD parameters were monitored either by an LVAD coordinator or perfusionist experienced with LVAD management. The role of these personnel were to adjust LVAD speeds based on flow and PI parameters; monitor blood pressure; assist anesthesia in pressor/inotrope management, placement of grounding pads for electrocautery, securing external components of the device; and advise the surgical and anesthesia team regarding management of the LVAD patient.
Postoperative Transfusion of Blood Products
Administration of blood products were also assessed within 1 week postoperation. The average number of blood products transfused postoperatively is listed in Table 3. A total of 212 units of PRBCs were transfused in 46 (72%) procedures; more than 5 units of PRBCs were needed in 11 (17%) procedures. Gastrointestinal operations had more postoperative bleeding (18/46; 39%). The procedure with the maximum units of PRBCs transfusion was a case of intractable hemorrhage after a thoracotomy that required 25 units of PRBCs.
Fresh frozen plasmas were administered within 1 week after operation in 10 (16%) operations. A total number of 55 units of FFP were transfused. Reasons for transfusions were to reverse an elevated INR caused by bleeding and coagulopathies. Platelets were given in eight (13%) procedures, amounting to a total of 42 units. Reasons for platelet transfusions were bleeding, coagulopathies, and multisystem organ failure because of sepsis. Six (9%) operations required transfusion of 15 units of cryoprecipitate because of bleeding.
For patients who required blood transfusions during the first week after operation, the average estimated intraoperative blood loss was 172.05 ± 397.14 ml compared to 149.33 ± 210.62 ml (p = 0.82) in patients not requiring postoperative transfusions. Also, for those receiving postoperative transfusions, the average preoperative INR was 1.49 ± 0.60 compared to 1.42 ± 0.30 (p = 0.65) in patients not receiving postoperative transfusions.
Intraoperative complications other than possibly unexpected excess bleeding were rare. Our study showed only one complication during operation. The patient was undergoing a craniotomy with evacuation of intracerebral hemorrhage when the patient presented with three episodes of ventricular tachycardia during operation. It was successfully reversed to sinus rhythm with the external defibrillator. The patient had history of ventricular tachycardia in the past and was receiving amiodarone. After operation, the patient presented with one more episode of ventricular tachycardia that required the use of intravenous lidocaine in addition to amiodarone. The patient was one of the six patients who died within 30 days of operation.
There were six (10%) deaths in 63 procedures during the first 30 postoperative day period. All six patients who died had undergone emergent operations (Table 4). Three (50%) of the six deaths occurred after neurosurgical procedures for intracranial hemorrhage. Three (50%) other deaths occurred in patients with the following emergent abdominal surgical issues: ischemic bowel, necrotic bowel, and multiorgan failure with bowel obstruction. The patient who suffered from multiorgan failure with bowel obstruction had a preoperative INR level of 3.7, the highest preoperative INR of any patient in the study. The preoperative INR of patients who died within30 days of operation was 1.9. This is higher than that of patients who survived past 30 days, which was 1.41. In five (83%) cases of deaths, the patients had received an average of 2.5 units of PRBCs. The average survival time for patients with 30-day mortality was 4.2 ± 1.3 (2–5 days).
Nine patients underwent 10 emergent operations. Those procedures included three intracranial hemorrhage evacuations; five abdominal operations for two ischemic bowels, necrotic bowel, bowel obstruction with multiorgan failure, and bleeding gastric ulcer; one pleurectomy; and one thoracic hematoma evacuation. Of the nine patients, one patient is living 509 days after emergent noncardiac operation for ischemic bowel. The other eight patients expired; six died within 30 days and two survived past 30 days. The median survival of the eight patients who died after emergent operations was 5 days (2–172 days).
Figure 2 shows Kaplan-Meier survival curves based on 1) timing of surgery—emergent versus elective; 2) cardiac risk of surgery—low, intermediate, and high-risk surgeries (based on the American College of Cardiology/American Heart Association guidelines for perioperative cardiovascular evaluation for noncardiac surgery)10; and 3) patients with no surgery versus those undergoing elective surgery.
Among the 37 patients in this study, 20 (54%) are currently doing well on their LVADs. The average number of days for the surviving patients is 840.6 ± 534.9 days (302–1917 days). Of the 17 patients who expired, 6 (35%) died within 30 days, 3 (18%) within 6 months, 3 (18%) within a year, and 3 (18%) survived for more than a year, and 2 (12%) survived for more than 2 years. The average survival time of the 17 patients who expired, including the 6 who died within 30 days, was 291.4 ±339.8 (2–1159 days). Excluding 30-day mortalities, the average survival of the other 11 patients who expired was 448.1 ± 329 days (109–1159 days).
This study reports the results of patients with CF-LVAD implanted primarily for destination therapy, who underwent noncardiac surgical procedures. Unlike previous studies that evaluated noncardiac operations in various ventricular assist devices, our investigation focuses on patients with the current generation of CF-LVAD, the HMII.
Several small reports about LVADs and noncardiac operations have been published; however, these consist primarily of pulsatile LVADs and in most cases involved bridge-to-transplant patients.4–8 The largest series before our study consisted of 37 patients undergoing 59 noncardiac procedures.11 This report included a variety of ventricular assist devices of which the majority were pulsatile LVADs. The study included nine HeartMate (HM) VE, fifteen HM-XVE, four HM-IP, five Thoratec, biventricular support, two CardioWest Temporary Total Artificial Heart, one Novacor, and one HMII.11 In contrast, our study includes only patients with HMII LVADs primarily for destination therapy, who underwent noncardiac operation.
The types of procedures that required blood-product administration in our study were diverse; however, they were most often abdominal procedures. Some of the patients may have required blood products during and after operation because of ongoing coagulopathies unrelated to their LVADs. An important mechanism of GI bleed in these patients is the phenomenon of “acquired von Willebrand disease.”12,13 Like aortic stenosis (AS), CF-LVADs create an environment of high shear stress of blood flow, which results in destruction of the large multimeres that plays an important role in hemostasis in areas of high shear stress as GI arteriovenous malformations.14
Patients with CF-LVADs have a similar physiological state to the ones with AS because of the narrow pulse pressure. This relationship of AS and GI bleeding may also hold true for the continuous-flow devices, and explain the increased incidence of bleeding in abdominal/GI procedures. Boley et al.15 suggested that increased intraluminal pressure along with muscular contraction may result in dilated mucosal veins that favor the development of arteriovenous communication that may bleed when exposed to trauma/stress. Alternatively, Cappell and Lebwohl16 proposed a neurovascular cause with increased sympathetic tone resulting in smooth muscle relaxation and development of angiodysplasia. Another proposed mechanism is the lowered pulse pressure, as seen in AS, leads to intestinal hypoperfusion and resulting hypoxia can lead to vascular dilation and angiodysplasia. The abovementioned mechanisms could possibly have contributed to the increased risk of bleeding in these LVAD patients requiring intra- and post-operative transfusion of blood products, despite discontinuation of anticoagulation and antiplatelet agents before surgery. The removal of nearly all preload in the left ventricle prevents the opening of the aortic valve in most patients with CF-LVAD. The usual practice in our hospital is to reduce the speed of the VAD to decrease flow and generate pulsatility. Our study also demonstrated that anticoagulation can be discontinued safely before noncardiac operation without creating a prothrombotic state that could lead to an intra-LVAD thrombosis or systemic thromboembolism. This could be secondary to the sintered titanium blood contacting surface of the HMII, which provides enough thromboresistance.14
In our study, pulmonary artery catheters were generally not required because LVADs provide a continuous measure of cardiac output, thereby facilitating intraoperative management. Left-sided filling pressures, which reflect an adequate heart function, were also monitored by intraoperative transesophageal echocardiogram when required.
Although we studied the largest number of noncardiac operations in patients with CF-LVAD, this investigation is limited by being a retrospective, single-center study.
Our study demonstrates that patients with LVAD can safely undergo noncardiac surgical procedures with minimal intraoperative complications but may require more transfusions than nonVAD patients because of complex coagulopathies. These patients need careful monitoring as blood transfusions could transiently increase pulmonary vascular resistance and possibly reduce LVAD flows if untreated. The chances of LVAD dysfunction or device thrombosis during the noncardiac operations are rare as found by our study. To achieve these positive outcomes, it is preferable that personnel familiar with the LVADs assist in the noncardiac operation of LVAD patients at the LVAD implantation center.
1. Miller LW, Pagani FD, Russell SD, et al. HeartMate II Clinical Investigators: Use of a continuous-flow device in patients awaiting heart transplantation. N Engl J Med. 2007;357:885–896
2. Slaughter MS, Rogers JG, Milano CA, et al. HeartMate II Investigators: Advanced heart failure treated with continuous-flowleft ventricular assist device. N Engl J Med. 2009;361:2241–2251
3. Slaughter MS, Pagani FD, Rogers JG, et al. Clinical management of continuous-flow left ventricular assist devices in advanced heart failure. J Heart Lung Transplant. 2010;29:S1–S39
4. Kartha V, Gomez W, Wu B, Tremper K. Laparoscopic cholecystectomy in a patient with an implantable left ventricular assist device. Br J Anaesth. 2008;100:652–655
5. Wei B, Takayama H, Bacchetta MD. Pulmonary lobectomy in a patient with a left ventricular assist device. Ann Thorac Surg. 2009;87:1934–1936
6. Goldstein DJ, Mullis SL, Delphin ES, et al. Noncardiac surgery in long-term implantable left ventricular assist-device recipients. Ann Surg. 1995;222:203–207
7. Schmid C, Wilhelm M, Dietl KH, Schmidt C, Hammel D, Scheld HH. Noncardiac surgery in patients with left ventricular assist devices. Surgery. 2001;129:440–444
8. Garatti A, Bruschi G, Colombo T, et al. Noncardiac surgical procedures in patient supported with long-term implantable left ventricular assist device. Am J Surg. 2009;197:710–714
9. Boyle AJ, Russell SD, Teuteberg JJ, et al. Low thromboembolism and pump thrombosis with the HeartMate II left ventricular assist device: Analysis of outpatient anti-coagulation. J Heart Lung Transplant. 2009;28:881–887
10. Eagle KA, Berger PB, Calkins H, et al. ACC/AHA guideline update for perioperative cardiovascular evaluation for noncardiac surgery—Executive summary a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1996 Guidelines on Perioperative Cardiovascular Evaluation for Noncardiac Surgery). Circulation. 2002;105:1257–1267
11. Stehlik J, Nelson DM, Kfoury AG, et al. Outcome of noncardiac surgery in patients with ventricular assist devices. Am J Cardiol. 2009;103:709–712
12. Warkentin TE, Moore JC, Morgan DG. Aortic stenosis and bleeding gastrointestinal angiodysplasia: Is acquired von Willebrand’s disease the link? Lancet. 1992;340:35–37
13. Uriel N, Pak SW, Jorde UP, et al. Acquired von Willebrand syndrome after continuous-flow mechanical device support contributes to a high prevalence of bleeding during long-term support and at the time of transplantation. J Am Coll Cardiol. 2010;56:1207–1213
14. John R, Lee S. The biological basis of thrombosis and bleeding in patients with ventricular assist devices. J Cardiovasc Transl Res. 2009;2:63–70
15. Boley SJ, Sammartano R, Adams A, DiBiase A, Kleinhaus S, Sprayregen S. On the nature and etiology of vascular ectasias of the colon. Degenerative lesions of aging. Gastroenterology. 1977;72(4 Pt 1):650–660
16. Cappell MS, Lebwohl O. Cessation of recurrent bleeding from gastrointestinal angiodysplasias after aortic valve replacement. Ann Intern Med. 1986;105:54–57