LAPAROSCOPIC RENAL-ADRENAL SURGERY IN PATIENTS ON ORAL ANTICOAGULANT THERAPY

0022-5347/05/1743-1020/0 THE JOURNAL OF UROLOGY® Copyright © 2005 by AMERICAN UROLOGICAL ASSOCIATION

Vol. 174, 1020 –1023, September 2005 Printed in U.S.A.

DOI: 10.1097/01.ju.0000169461.35421.f6

Physiology/Medical Disease LAPAROSCOPIC RENAL-ADRENAL SURGERY IN PATIENTS ON ORAL ANTICOAGULANT THERAPY IOANNIS M. VARKARAKIS, SOROUSH RAIS-BAHRAMI, MOHAMAD E. ALLAF, GUILHERME C. LIMA, SOMPOL PERMPONGKOSOL, PRAVIN RAO, THOMAS W. JARRETT AND LOUIS R. KAVOUSSI* From The James Buchanan Brady Urological Institute, The Johns Hopkins Medical Institutions, Baltimore, Maryland

ABSTRACT

Purpose: Patients requiring chronic anticoagulation are theoretically at increased risk for hemorrhage or thromboembolism perioperatively. Experience with laparoscopic renal/adrenal surgery in patients on chronic warfarin is limited. We assessed hemorrhagic/thromboembolic complications in this group of patients. Materials and Methods: The records of 787 patients undergoing laparoscopic renal/adrenal surgery were retrospectively reviewed. A total of 25 patients on chronic oral anticoagulation with warfarin were identified. The indications for warfarin therapy as well as perioperative management were reviewed. Clinical parameters, including operative time, estimated blood loss, hemorrhagic/thromboembolic complications and transfusions, were documented and compared with those in patients not receiving chronic anticoagulation. Results: Atrial fibrillation (56% of cases) and a prosthetic mitral valve (28%) were the most frequent indications for chronic anticoagulation. Bridging anticoagulation with unfractionated heparin was the most frequent management method (68% of cases). Patients with anticoagulation were older (p ⬍0.001) and hospitalized longer (⬍0.001) than those without anticoagulation. Operative time, estimated blood loss and the conversion rate were not significantly different between the groups. However patients on chronic warfarin significantly more often required transfusion (24% vs 5.2%, p ⬍0.005) and had more postoperative bleeding episodes (8% vs 0.9%, p ⬍0.05) than patients not on chronic anticoagulation. No thromboembolic events occurred in the anticoagulated group, while 3 occurred in the nonanticoagulated group (p ⫽ 1). Conclusions: Laparoscopic renal/adrenal surgery in patients requiring chronic anticoagulation therapy can be performed safely. The risk of intraoperative bleeding is not increased, although the incidence of postoperative bleeding as well as transfusions is higher. KEY WORDS: laparoscopy, kidney, adrenal glands, hemorrhage, blood transfusion

Many patients undergoing elective surgery may be on chronic oral anticoagulant therapy, eg warfarin, because of underlying atrial fibrillation (AF), mechanical prosthetic heart valves, or previous venous thromboembolism (VT) or pulmonary thromboembolism. These patients are at increased risk for perioperative complications. The continuation of full anticoagulation throughout surgery increases bleeding during the procedure, while interrupting of anticoagulant therapy increases the risk of thromboembolism.1, 2 Appropriate management with temporary perioperative interruption of oral anticoagulant therapy and the institution of appropriate bridging anticoagulation therapy, eg unfractionated heparin (UFH) or low molecular weight (LMW) heparin, can decrease complications in patients at high risk.3 Hemorrhagic complications in patients on chronic anticoagulation undergoing open noncardiac surgery have been evaluated to aid in establishing management strategies.4 Similar evaluation in patients undergoing laparoscopy is limited and it has been reported only for procedures with a low

risk of bleeding, eg cholecystectomy.5, 6 Surgery such as laparoscopic renal/adrenal surgery requires more extensive dissection, thus, conveying a potentially higher risk of hemorrhage. We evaluated bleeding complications associated with laparoscopic renal/adrenal surgery in patients on chronic anticoagulant therapy. METHODS

The records of 787 patients undergoing laparoscopic renal/ adrenal surgery, including partial nephrectomy in 216, nephrectomy in 329 (radical, simple and nephroureterectomy), pyeloplasty in 219 and adrenalectomy in 23, were retrospectively reviewed to identify patients receiving chronic oral anticoagulation with warfarin for high risk medical conditions. A total of 25 such patients, including 6 with partial nephrectomy, 16 with radical nephrectomy, and 1 each with nephroureterectomy, pyeloplasty and adrenalectomy, were identified. The reasons for chronic anticoagulation, the warfarin dose, the baseline and preoperative international normalized ratio (INR), and perioperative anticoagulant management were recorded in these patients. Insufflation pressure and surgical technique were similar

Submitted for publication December 22, 2004. * Correspondence: 600 North Wolfe St., Suite 161 Jefferson Street Building, Baltimore, Maryland 21287-8915 (telephone: 410 –5027710; FAX: 410 –5027711; e-mail: [email protected]). 1020

LAPAROSCOPIC RENAL/ADRENAL SURGERY AND ORAL ANTICOAGULANT THERAPY

in all patients undergoing the same procedure and in the 2 groups. Meticulous hemostasis and adjuvant hemostatic measures when needed was the rule in all patients regardless of coagulation status. Clinical parameters, including patient age, American Society of Anesthesiologists (ASA) score, operative time, estimated blood loss (EBL), the conversion rate, length of hospitalization, thromboembolic and bleeding complications, and the overall transfusion rate were documented and compared between patients on chronic warfarin and individuals not receiving chronic anticoagulation. Statistical analysis was performed using the Wilcoxon rank sum and Fisher exact test using SPSS software (SPSS, Chicago, Illinois).

1021

TABLE 2. Patients with and without chronic anticoagulant therapy No. pts Mean age ⫾ SD Mean ASA score (range) Mean operative time ⫾ SD Mean EBL ⫾ SD (ml) Mean hospital stay ⫾ SD (days) No. overall transfusion (%) No. complications (%): Postop bleeding Thromboembolic No. conversion (%)

Warfarin

No Warfarin

p Value

25 68.4 ⫾ 9.4 2.8 (2–3) 205 ⫾ 64 303 ⫾ 402 5.6 ⫾ 5.3

762 50.1 ⫾ 16.7 1.9 (1–4) 211.7 ⫾ 77.7 348.5 ⫾ 343.11 3.1 ⫾ 1.49

⬍0.001 ⬍0.0005 0.976 0.716 ⬍0.001

6 (24) 2/25 (8) 0/25 1 (4)

40 (5.2)

⬍0.005

7/762 (0.9) 3/762 (0.4) 17 (2.2)

⬍0.05 1 0.444

RESULTS

A total of 25 patients on chronic anticoagulant therapy with warfarin were identified who required maintenance of a baseline INR of greater than 2.0 in the majority. Table 1 lists the indications for chronic warfarin administration, mean warfarin dose, and INR at baseline and preoperatively. The most frequent reason for chronic anticoagulation was atrial fibrillation (14 of 25 patients or 56%) and a prosthetic mitral valve (7 of 25 or 28%). Preoperative INR was always decreased to 1.5 just prior to surgery according to the recommendations of the medical physician of the patients. Perioperative management included preoperative cessation of warfarin 3 to 4 days before surgery in all patients. Of the 25 patients 22 (88%) continued with bridging anticoagulation using intravenous UFH (17 of 25 or 68%) or subcutaneous LMW heparin (5 of 25 or 20%). Bridging anticoagulation with intravenous UFH was discontinued 6 hours before surgery and restarted 12 hours postoperatively if no evidence of bleeding was detected. The UFH dose varied according to activated partial thromboplastin time and it was continued until INR achieved therapeutic levels after the reinstitution of warfarin, usually 24 to 36 hours postoperatively. In 2 patients no bridging prophylaxis was deemed necessary. Table 2 lists clinical parameters, including patient age, ASA, operative time, EBL, length of hospitalization, hemorrhagic/thromboembolic complications and overall transfusion rates. Patients on chronic warfarin were older (p ⬍0.001) and more ill (p ⬍0.0005), and had a longer hospital stay (p ⬍0.001). However, intraoperative blood loss (p ⫽ 0.716), operative time (p ⫽ 0.976) and conversion rates (p ⫽ 0.444) were similar to those in patients not on any anticoagulant therapy, indicating similar surgical difficulty between the 2 groups. Thromboembolic complications were not statistically different between the 2 groups. No such complication occurred in patients with anticoagulation. Three thromboembolic events occurred in the nonchronically anticoagulated group during partial nephrectomy, radical nephrectomy and pyeloplasty, respectively.

TABLE 1. Characteristics of patients on chronic anticoagulation with warfarin No. chronic anticoagulation indication: Cardiac arrhythmias (atrial fibrillation/other) Prosthetic valve (mitral/aortic/aortic ⫹ mitral) Previous deep venous thrombosis or pulmonary embolism Other (antiphospholipid syndrome, factor V Leiden deficiency) Mean mg immediate preop period (range) Mean INR (range): Baseline Preop No. bridging anticoagulation management: Warfarin cessation, no other prophylaxis Warfarin cessation, UFH bridging therapy Warfarin cessation, LMW heparin bridging therapy Not specified

16 (14/2) 8 (5/1/2) 6 2 4.6 (1.5–10) 2.4 (1.7–3.4) 1.2 (0.9–1.5) 2 17 5 1

Active postoperative bleeding requiring transfusion and/or intervention was more frequent in patients receiving chronic anticoagulation (p ⬍0.05). These patients also required more frequent transfusions overall (p ⬍0.005). In the anticoagulation group 2 patients had postoperative bleeding, were hemodynamic unstable and required serial transfusions. In the first patient observation was adequate for managing significant hematuria beginning on postoperative day 2 after laparoscopic nephroureterectomy. In other patient emergent reexploration was required on postoperative 1 for intraabdominal bleeding, as confirmed by computerized tomography after laparoscopic pyeloplasty. No definitive site of bleeding was found at exploration. The remaining 4 patients, including 3 undergoing radical nephrectomy and 1 with partial nephrectomy, received transfusion on postoperative days 1 and 2 since postoperative hemoglobin was below 10 gm/dl and transfusion was advised due to cardiac comorbidities. In the nonanticoagulated group 7 patients had active postoperative bleeding, requiring management during partial nephrectomy in 4, radical nephrectomy in 1 and pyeloplasty in 2. DISCUSSION

Laparoscopic surgery has evolved in the last decade, expanding its indications to treat patients with complex pathological as well as medical conditions. Patients on chronic oral anticoagulation represent a challenging subset when surgical intervention is required. Reports indicate that these patients can undergo minor operative procedures without discontinuing oral anticoagulation.7⫺10 Continuing oral anticoagulation is not possible with more invasive surgery due to the risk of serious bleeding complications. Reported experience with patients on chronic anticoagulation undergoing laparoscopic surgery is limited with small studies in those undergoing low risk procedures.5, 6 Renal/adrenal laparoscopic surgery is a higher risk procedure and, as such, anticoagulation must be discontinued perioperatively. Halting anticoagulation can place these patients at risk for thromboembolic events. If the risk is low, the procedure can proceed without any perioperative prophylaxis, thus, minimizing the need for unnecessary anticoagulation. However, close collaboration with the medical service is necessary to determine the risk in each patient when discontinuing oral warfarin (see Appendix). Patients with a recent thromboembolic event (less than 3 months), a history of multiple episodes, or recurrent venous or pulmonary embolism, hereditary hypercoagulable states, mechanical heart valves, recently placed bioprosthetic valves (less than 3 months), left ventricular dysfunction and atrial fibrillation are at high risk for thromboembolism if not continuously anticoagulated and, thus, they require bridging anticoagulation perioperatively.11 Even in patients at high risk perioperative management must be individualized. For example, patients with mitral valve prostheses are at twice the risk for thromboem-

1022

LAPAROSCOPIC RENAL/ADRENAL SURGERY AND ORAL ANTICOAGULANT THERAPY

bolism as that in those with aortic valve prostheses.12 Similarly patients with AF with risk factors such as heart failure, hypertension, diabetes, a dilated left atrium, impaired cardiac function and a history of stroke or transient ischemic attack (TIA), are at increased risk compared with patients with AF who are young (less than 65years) and have no risk factors for stroke or thromboembolism.13 AF and a mitral prosthetic heart valve were the most frequent indications for chronic anticoagulant therapy in our study. Only 2 of 25 patients who were on chronic anticoagulation were considered to be at low risk for thromboembolism and, thus, they did not require any type of bridging anticoagulation therapy. Appropriate anticoagulation management with early mobilization may effectively prevent thromboembolic events. There is little consensus regarding optimal perioperative treatment in patients on long-term warfarin therapy.14 Although the transition from warfarin therapy to prophylaxis with LMW heparin is more convenient and cost-effective because it can be done on an outpatient basis, it is reserved for patients at lesser thromboembolic risk.15 Intravenous UFH has been used in the majority of patients and it accounts for the significantly longer hospital stay in the anticoagulated group. Overlapping warfarin and UFH is necessary postoperatively until oral anticoagulation takes full effect, usually 36 to 72 hours. Warfarin can be restarted when there is less risk of postoperative bleeding, usually 24 to 48 hours after the procedure. The UFH-warfarin transition usually requires hospitalization and it can be challenging. For example, a prolonged hospital stay of 18 days was necessary in 1 patient because of difficulty determining therapeutic anticoagulation levels on warfarin after surgery. Overall transfusion rates were statistically higher in patients with anticoagulation. Although a third of patients were transfused due to postoperative bleeding, the increased overall transfusion rate is probably attributable to a lower threshold for transfusion in patients with associated cardiac comorbidities. Healthy patients can tolerate a decrease in postoperative hemoglobin better than patients with cardiac comorbidities and, thus, transfusion is less frequently necessary. Increased surgical experience as well as new, improved hemostatic adjuvants, such as tissue sealants and glues, the harmonic scalpel and other new instrumentation, has allowed laparoscopic surgery to be performed with the potential for less blood loss. In addition, pneumoperitoneum can

tamponade venous bleeding, aiding in hemostasis. As such, it is important to lower pressures at the conclusion of surgery to assess bleeding. Small bleeding sites obscured by pneumoperitoneum may be missed, leading to late postoperative bleeding. Meticulous hemostasis is particularly important in this subset of cases, as it is routine in all laparoscopic cases. Careful monitoring of postoperative parameters such as hemoglobin, urine output and patient vital signs are crucial for the early detection of postoperative bleeding. However, one must not forget that patients on chronic anticoagulation frequently have cardiac comorbidities and, therefore, they are on anti-arrhythmic medication. This may prevent tachycardia even in the case of postoperative bleeding, thus, making the diagnosis challenging. The incidence of postoperative bleeding events was statistically higher in patients with anticoagulation. The 2 cases occurred early in the postoperative period prior to the commencement of warfarin. Therefore, it is possible that bleeding occurred during a small window of over anticoagulation while on heparin. Heparin has a short half-life and its effect can easily be reversed. Despite frequent monitoring through activated partial thromboplastin time over anticoagulation has been shown to occur frequently with UFH.16 One must always have in mind the synergistic effect that this anticoagulant may have with commonly used antibiotics, eg amoxicillin, clarithromycin, norfloxacin and trimethoprim-sulfamethoxazole. CONCLUSIONS

Laparoscopic renal/adrenal surgery is feasible in patients with anticoagulation, as indicated by equivalent operating times and conversion rates. The incidence of intraoperative bleeding is the same, although the risk of postoperative bleeding is higher along with the possibility of transfusion. Meticulous hemostasis and close monitoring of coagulation parameters may help prevent bleeding in such patients. Scrupulous postoperative monitoring allows the early detection of postoperative bleeding. The transfusion threshold in these patients is usually lower due to the frequent coexistence of cardiac comorbidities. Close cooperation with the medical service will allow the development of optimal perioperative anticoagulation bridging management. Careful adherence to principles will help minimize embolic events, while avoiding intraoperative and postoperative bleeding.

APPENDIX: PATIENT RISK STRATIFICATION FOR PERIOPERATIVE ANTICOAGULATION

High risk Anticoagulation strongly recommended Mechanical prosthetic heart valve

Chronic AF

Previous episode of VT

Recent (less than 1 month) stroke or TIA Any mitral valve Caged ball or single leaflet tilting disc aortic valve Recent (less than 1 month) stroke or TIA Rheumatic mitral valvular heart disease Recent (less than 3 months) episode of VT Active cancer Major comorbid disease Antiphospholipid antibody REFERENCES

1. Katholi, R. E., Nolan, S. P. and McGuire, L. B.: Living with prosthetic heart valves. Subsequent noncardiac operations and the risk of thromboembolism or hemorrhage. Am Heart J, 92: 162, 1976

Moderate risk Anticoagulation should be considered Bileaflet tilting disc Aortic valve and 2 or more stroke risk factors

Bileaflet tilting disc Aortic valve and less than 2 stroke risk factors

Chronic AF and 2 or more stroke risk factors

Chronic AF and less than 2 stroke risk factors

VT within the past 6 months VT occurring in association with previous interruption of warfarin therapy

None of the previously stated

Low risk Anticoagulation is optional

2. Poller, L. and Thomson J.: Evidence for “rebound” hypercoagulability after stopping anticoagulants. Lancet, 39: 62, 1964 3. Torn, M. and Rosendaal, F. R.: Oral anticoagulation in surgical procedures: risks and recommendations. Br J Haematol, 123: 676, 2003

LAPAROSCOPIC RENAL/ADRENAL SURGERY AND ORAL ANTICOAGULANT THERAPY 4. Katholi, R. E., Nolan, S. P. and McGuire, L. B.: The management of anticoagulation during noncardiac operations in patients with prosthetic heart valves. A prospective study. Am Heart J, 96: 163, 1978 5. Fitzgerald, S. D., Bailey, P. V., Liebscher, G. J. and Andrus, C. H.: Laparoscopic cholecystectomy in anticoagulated patients. Surg Endosc, 5: 166, 1991 6. Yoshida, T., Kitano, S., Matsumoto, T., Bandoh, T., Baatar, D., Ninomiya, K. et al: Laparoscopic cholecystectomy in patients undergoing anticoagulant therapy. Surg Today, 28: 308, 1998 7. Heit, J. A.: Perioperative management of the chronically anticoagulated patient. J Thromb Thrombolysis, 12: 81, 2001 8. Watterson, J. D., Girvan, A. R., Cook, A. J., Beiko, D. T., Nott, L., Auge, B. K. et al: Safety and efficacy of holmium: YAG laser lithotripsy in patients with bleeding diatheses. J Urol, 168: 442, 2002 9. Morris, A. and Elder, M. J.: Warfarin therapy and cataract surgery. Clin Experiment Ophthalmol, 28: 419, 2000 10. Evans, I. L., Sayers, M. S., Gibbons, A. J., Price, G., Snooks, H. and Sugar, A. W.: Can warfarin be continued during dental extraction? Results of a randomized controlled trial. Br J Oral Maxillofac Surg, 40: 248, 2002 11. Douketis, J. D.: Perioperative anticoagulation management in patients who are receiving oral anticoagulant therapy: a practical guide for clinicians. Thrombosis Res, 108: 3, 2002 12. Cannegieter, S. C., Rosendaal, F. R. and Briet, E.: Thromboembolic and bleeding complications in patients with mechanical heart valve prostheses. Circulation, 89: 635, 1994 13. Bleeding during antithrombotic therapy in patients with atrial fibrillation. The Stroke Prevention in Atrial Fibrillation Investigators. Arch Intern Med, 156: 409, 1996 14. Douketis, J. D., Crowther, M. A. and Cherian, S. S.: Perioperative anticoagulation in patients with chronic atrial fibrillation who are undergoing elective surgery: results of a physician survey. Can J Cardiol, 16: 326, 2000 15. Amorosi, S. L., Tsilimingras, K., Thompson, D., Fanikos, J., Weinstein, M. C. et al: Cost analysis of “bridging therapy” with low-molecular-weight heparin versus unfractionated heparin during temporary interruption of chronic anticoagulation. Am J Cardiol, 93: 509, 2004 16. Fanikos, J., Stapinski, C., Koo, S., Kucher, N., Tsilimingras, K. and Goldhaber, S. Z.: Medication errors associated with anticoagulant therapy in the hospital. Am J Cardiol, 94: 532, 2004 EDITORIAL COMMENT These authors have provided us with some sobering data about laparoscopic surgery in patients on anticoagulant therapy. They report that there were 2 patients of the 25 (8%) in the anticoagulant therapy group with postoperative hemorrhage. I think that the authors are underestimating the problem and the rate of postoperative

1023

hemorrhage is actually closer to the 24% rate that they report for postoperative blood transfusions. The authors admit giving postoperative blood transfusions in 6 of the 25 patients (24%) but claim that 4 of these 6 did not have postoperative hemorrhage and instead received transfusions for low postoperative hematocrit and cardiac comorbidities. While the anticoagulated patients indeed were older and sicker than the rest of the patients, if these 4 did not have postoperative hemorrhage, then the cause of low postoperative hematocrits must have been low preoperative hematocrit, over hydration and subsequent dilutional anemia or large intraoperative blood loss. The last is unlikely, given that EBL was lower in anticoagulated patients. In my opinion any transfusion postoperatively after abdominal surgery that is not accounted for by operative blood loss or another definable site of blood loss, such as gastrointestinal hemorrhage or hemolytic anemia, indicates postoperative (abdominal) hemorrhage. My conclusion is similar to that of the authors, namely that anticoagulated patients are at greater risk for postoperative hemorrhage and blood transfusions. I am simply arguing that the magnitude of the risk is greater than the authors make it out to be. I read in their data that 1 of 4 such patients require blood transfusion for postoperative (abdominal) hemorrhage. This is a significant risk. Personally I am willing to perform laparoscopic total nephrectomy or adrenalectomy in such patients with an understanding of the increased hemorrhagic risk. However, I avoid laparoscopic partial nephrectomy in patients in whom full anticoagulation with bridging heparin is required early postoperatively. In such patients alternatives such as percutaneous radio frequency ablation are preferred. J. Stuart Wolf, Jr. Department of Urology University of Michigan Ann Arbor, Michigan REPLY BY AUTHORS To clarify, in 4 of 6 patients transfused intraoperative blood loss was not increased over controls and postoperative computerized tomography demonstrated no bleeding. These patients started with hemoglobin less than normal due to chronic disease. We value the input of the medical service that closely follows these patients who believed that the comorbidities warranted transfusion. We cannot agree that total nephrectomy or even ablation can be substituted in all chronically anticoagulated patients whose disease would be amenable to partial nephrectomy. Indeed in 3 of the 4 patients Wolf cites radical nephrectomy was performed. Is a blood transfusion more of an evil than potentially impacting on renal function or inadequate tumor control?