- Email: [email protected]
Perioperative clinical thromboembolic events after radical or partial nephrectomy
ADULT UROLOGY
PERIOPERATIVE CLINICAL THROMBOEMBOLIC EVENTS AFTER RADICAL OR PARTIAL NEPHRECTOMY JOSEPH A. PETTUS, SCOTT E. EGGENER, AHMAD SHABSIGH, BRENT YANKE, MARK E. SNYDER, ANGEL SERIO, ANDREW VICKERS, PAUL RUSSO, AND S. MACHELE DONAT
ABSTRACT Objectives. To evaluate the incidence of, and identify the risk factors for, clinical thromboembolic events after radical/partial nephrectomy. Cancer is an established risk factor for deep vein thrombosis (DVT) and pulmonary embolism (PE); however, their incidence after nephrectomy for renal tumors has been poorly studied. Methods. We reviewed our prospective institutional renal database and identified 2208 patients who underwent renal tumor surgery from January 1989 to July 2005. The clinical parameters evaluated were age, sex, race, body mass index, smoking history, medical comorbidities, American Society of Anesthesia grade, procedure type, estimated blood loss, and length of hospitalization. Hospital records, discharge “International Classification of Diseases, Ninth Revision” codes, and 30-day postoperative morbidity and mortality data were reviewed to identify patients diagnosed with perioperative DVT or PE. Results. A total of 34 (1.5%, 95% confidence interval 1.1% to 2.1%) thromboembolic events (20 PEs and 14 DVTs) were identified in 33 patients. Patients with a preoperative history of arrhythmia (P ⫽ 0.02) or prior DVT (P ⫽ 0.053) were more likely to experience PE. The estimated blood loss was directly associated with an increased risk of PE (P ⫽ 0.001). Patients with coronary artery disease (P ⫽ 0.050) or of advanced age (P ⫽ 0.02) were more likely to experience DVT (P ⫽ 0.02). Conclusions. To our knowledge, this is the first study on the incidence of thromboembolic events after nephrectomy. Thromboembolic events are rare but are more likely to occur in patients with coronary artery disease, cardiac arrhythmia, prior DVT, Stage 3 or 4 tumors, or a large estimated blood loss. UROLOGY 68: 988–992, 2006. © 2006 Elsevier Inc.
P
atients with cancer undergoing surgical procedures are at increased risk of deep venous thrombosis (DVT) or pulmonary embolus (PE) during the perioperative period.1–3 Elderly patients and those undergoing abdominal or pelvic surgery are at particularly high risk; therefore, prophylactic measures are generally recommended.4,5 The incidence of venothromboembolism (VTE) has been well studied in general surgical and gynecologic studies but not urologic reports.5,6 Therefore, recommendations for urologic patients have been largely extrapolated from the general surgery data.5 From the Departments of Urology and Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, New York Reprint requests: Machele Donat, M.D., Department of Urology, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. E-mail: [email protected] Submitted: February 28, 2006, accepted (with revisions): June 15, 2006 © 2006 ELSEVIER INC. 988
ALL RIGHTS RESERVED
Compared with patients without cancer, screened surgical patients with cancer have a greater postoperative incidence of proximal lower-extremity DVTs (10% to 20% versus 2% to 4%) and calf DVTs (40% to 80% versus 10% to 20%).7 Although the overall complication rate after renal tumor surgery has been reported to be 17%,8 the VTE rate in patients after partial or radical nephrectomy has not been well studied. Despite the recommendations supporting the use of heparin,2–5,9 we have been impressed by the rarity of VTE in patients treated only with pneumatic compression boots. The purpose of this study was to evaluate a large cohort of patients from a single institution to determine the incidence of perioperative clinical VTE in patients undergoing radical or partial nephrectomy for renal masses and to define risk factors that might influence the decision to administer prophylactic perioperative anticoagulation. 0090-4295/06/$32.00 doi:10.1016/j.urology.2006.06.026
TABLE I.
Patient characteristics at nephrectomy
Age (yr) BMI (kg/m2) EBL (mL) OR time (min) Hospital stay (days) Sex Male Female Race White Black Hispanic Asian Black Hispanic Not recorded ASA grade 1 2 3 4 Missing Tumor laterality Right Left Missing Type of surgery Open radical Open partial Laparoscopic radical Laparoscopic partial Pathologic stage T1 T2 T3 T4 Other* Missing N stage N0/Nx N⫹ M stage M0/Mx M⫹
62.6 (53.2, 70.5) 28 (25, 31) 300 (200, 600) 185 (145, 235) 5 (4, 7) 1402 (64) 806 (37) 1763 (80) 103 (5) 59 (3) 42 (2) 9 (0.4) 232 (10) 126 (6) 1078 (49) 679 (31) 13 (1) 312 (14) 1120 (51) 1087 (49) 1 (0.05) 1325 (60) 770 (35) 77 (3) 33 (2) 1293 (59) 212 (10) 539 (24) 22 (1) 121 (5) 21 (1) 2147 (97) 61 (3) 2071 (94) 137 (6)
KEY: BMI ⫽ body mass index; EBL ⫽ estimated blood loss; OR ⫽ operating room; ASA ⫽ American Society of Anesthesiologists. Data presented as median, with interquartile range in parentheses, or as number, with percentages in parentheses. * T stage did not apply.
MATERIAL AND METHODS After institutional review board approval, we reviewed our prospectively collected departmental renal tumor database for all partial and radical nephrectomies performed for primary renal cortical tumors from January 1989 to July 2005. Demographic information, preoperative morbidities, American Society of Anesthesia grade, tumor laterality, smoking history, body mass index, procedure type (ie, laparoscopic versus open, radical versus partial nephrectomy), operative time, estimated blood loss (EBL), pathologic stage, and postoperative hospital stay were recorded for each patient (Table I). The UROLOGY 68 (5), 2006
postoperative hospital stay was calculated as the time in days from the date of surgery to discharge. Admission before surgery was not considered. We collapsed the comorbidities into general categories to make the statistical analysis more robust. For instance, all cardiac arrhythmias were grouped together. The pathologic stage was assigned only to patients with renal cell carcinoma in accordance with the 2002 TMN staging system. We conducted a chart review to identify patients who experienced VTE, defined as either clinically evident DVT and/or PE during the 30-day perioperative period. We also reviewed the discharge “International Classification of Diseases, Ninth Revision” codes and the departmental morbidity and mortality reports during the study period to identify patients who had experienced a clinically apparent VTE. During the study period, the DVT prophylaxis protocol consisted only of intraoperative and postoperative pneumatic compression devices (Venodyne boots) but no heparin. Patients taking an anticoagulation agent preoperatively started taking the drug again after resumption of oral intake and the hemoglobin was stable. Patients who had nonprimary renal cortical tumors were excluded. We used Fisher’s exact test to assess the relationship between outcome and each comorbidity. The associated confidence intervals were determined using exact methods. For continuous predictors, logistic regression analysis was used to assess their relationship with outcome. We used Stata, version 8.2 (Stata, College Station, Tex) and StatXact (Cytel, Cambridge, Mass) for all statistical analysis. Missing data were excluded from analysis on a case-by-case basis. We considered P ⬍0.05 to be significant.
RESULTS A total of 2208 patients met the study criteria. The demographic and clinical/pathologic characteristics of the study population are listed in Table I. Each VTE, except for one, was documented with imaging, either computed tomography angiography or ventilation-perfusion scan for PE and duplex ultrasonography for DVT. The patient without imaging studies was diagnosed with PE during surgery by arterial blood gas changes and died before the operation concluded. A total of 33 patients (1.5%) experienced a clinical VTE at a median of 7 days (interquartile range 3, 13) postoperatively (Table II), with 1 patient experiencing both DVT and PE. The only significant operative variable associated with having VTE was the EBL (P ⱕ0.001), with the operative time approaching significance (P ⫽ 0.06). Arrhythmia was the only medical comorbidity significantly associated with VTE (P ⫽ 0.02). Lymph node disease was marginally associated with having an event (P ⫽ 0.06) but neither T stage nor M stage was significant. The operative type and approach were not significantly associated with either event. No postoperative deaths occurred as a result of VTE. DVT occurred in 14 patients (0.6%) at a median of 10 days (interquartile range 8, 14) postoperatively and was confirmed by imaging done on the basis of evaluation of clinically suspicious physical examination findings or patient symptoms. The likelihood of clinical DVT increased with age (P ⫽ 989
TABLE II.
Clinicopathologic and comorbidity associations with VTE VTE/Total* (n)
Age (per yr) BMI (per kg/m2 unit) EBL (per 100 mL) OR time (per 10 min) Length of stay T stage ⬎T2 N stage ⬎N0 M stage ⬎M0 Malignant Year of surgery ASA† Comorbidity Arrhythmia Cancer history Clotting disorder CAD CHF COPD CVA Diabetes Previous DVT Hypercholesterolemia Hypertension Smoking TIA
26/1855
4/66 7/628 0/13 3/135 0/12 1/40 0/24 2/228 1/6 4/195 17/1059 19/1068 1/13
OR 1.02 1.0 1.02 1.03 1.032 1.6 3.7 1.2 0.77 1.0
4.7 0.67 0 2.2 0 1.7 0 1.2 14 1.4 1.3 1.3 5.6
95% CI
P Value
1.0–1.1 0.91–1.1 1.01–1.02 1.0–1.1 1.0–1.1 0.71–3.4 0.7–13 0.23–3.9 0.29–2.6 0.96–1.2
0.08 0.8 ⬍0.001 0.06 0.005 0.2 0.06 0.7 0.6 0.35 0.2
1.2–14 0.25–1.6 0–23 0.54–6.3 0–25 0–11 0–11 0.3–3.5 0.28–126 0.36–4.1 0.63–2.9 0.63–2.9 0.13–40
0.02 0.4 1 0.14 1 0.5 1 0.8 0.09 0.5 0.7 0.5 0.18
KEY: VTE ⫽ venothromboembolism; OR ⫽ odds ratio; CI ⫽ confidence interval; CAD ⫽ coronary artery disease; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebral vascular accident; TIA ⫽ transient ischemic attack; other abbreviations as in Table I. * Total number of comorbidities. † No OR presented because ASA has 4 levels.
0.02) and a history of coronary artery disease (P ⫽ 0.050). The mean age of the patients with DVT was 70 versus 61 years for those without DVT (P ⫽ 0.02). Patients with DVT were more likely to have nonorgan-confined disease (Stage T2 or worse, P ⫽ 0.02). A marginally significant association was seen between DVT and cardiac arrhythmia (P ⫽ 0.06). We observed a trend between increasing operative blood loss and DVT, such that the odds of DVT increased by a factor of 1.017 for each 100 mL of EBL (P ⫽ 0.08). Tumor laterality, nodal involvement, metastatic disease, body mass index, and operative time were not significantly associated with DVT. Procedure type (laparoscopic versus open, and partial versus radical nephrectomy) had no significant impact on DVT incidence. The annual incidence of DVT did not change during the course of the study. PE occurred in a total of 20 patients (0.9%) at a median of 4 days (interquartile range 2, 10) postoperatively. As the EBL increased, so did the risk of perioperative PE (P ⫽ 0.001), with the odds of experiencing a clinical PE increasing by a factor of 1.025 for each 100 mL of EBL. Patients with a prior history of DVT (P ⫽ 0.053) or arrhythmia (P ⫽ 0.02) were also more likely to experience PE. Again, procedure type and approach, TMN stage, 990
body mass index, and operative time were not significantly associated with PE, and the annual incidence did not change significantly. COMMENT Malignancy is a well-established risk factor for developing VTE. The American College of Chest Physicians currently recommends that all patients undergoing major urologic surgery for malignancy be treated with elastic stockings and/or pneumatic compression devices and either subcutaneous unfractionated heparin or low-molecular-weight heparin in the perioperative period.5 These recommendations have been extrapolated from thromboembolic events in large radical prostatectomy and general surgical series. In this study, we examined the incidence of thromboembolic events in all patients undergoing radical or partial nephrectomy for a renal cortical tumor. The incidence of perioperative DVT in asymptomatic screened patients has been reported to be 25% in a pooled analysis of general surgical patients without any prophylaxis and 4% in patients with pneumatic compression devices.5 A population-based study of the California Discharge Database found the 90-day perioperative VTE rate to be UROLOGY 68 (5), 2006
0.4% and 3% for patients undergoing nephrectomy for benign and malignant disease, respectively, but the prophylaxis status was unknown in that study.10 In this study, we analyzed clinically evident DVTs in patients treated only with pneumatic compression devices and found the rate to be quite low at 0.6%. This low rate is likely explained by the underreporting of events. Our patients were unscreened, and we only included the 30-day perioperative period, when most events occur. Had we extended the study period to include the 90-day perioperative period, the rate may have been greater. Patients may have had more events in the postdischarge period and either did not seek care or sought care elsewhere and did not report it to us. We most likely would have identified more patients with DVT had we used routine Doppler screening of asymptomatic patients postoperatively, but the significance of asymptomatic postoperative DVT is unknown. DVT has been reported to be very prevalent in patients with advanced cancer. Johnson and colleagues11 reported a 52% incidence among 258 patients with advanced cancer. Our study showed a relationship between perioperative DVT development and nonorgan-confined disease, but nodal and metastatic disease in renal cell carcinoma was not associated with DVT. The relationship between locally advanced disease and DVT may be explained by the venous stasis caused by the presence of tumor in the renal vein or vena cava, as well as mechanical compression of the vena cava by the tumor. It is certainly possible that some DVTs were present preoperatively but not detected until after surgery. The benefit of lymphadenectomy in radical or partial nephrectomy has not been established,12,13 and we have not consistently included lymphadenectomy at our institution. Therefore, the small portion of patients with lymph node involvement may have been a result of undersampling, and the true relationship between nodal status and VTE may not have been revealed in our study population. Finally, performing nephrectomy in the setting of metastatic disease is usually done as a cytoreductive procedure before initiating immunotherapy, and these patients are highly selected in terms of metastatic disease burden and performance status. We did not compare patients with benign tumors with those with malignant tumors, but analyzed them on an intent-to-treat basis. This was done because all renal cortical tumors in this study were assumed to be renal cell carcinoma at surgery on the basis of clinical and radiographic findings preoperatively. Preoperative renal biopsies and intraoperative frozen section analysis findings are often inaccurate and therefore not routinely performed in the absence of another primary malignancy.14,15 UROLOGY 68 (5), 2006
The pathologic findings are not typically available for several days after surgery. Therefore, pathologic confirmation of malignancy is not available to make decisions regarding thromboembolic prophylaxis. Patients with a cardiac arrhythmia history had greater rates of VTE, particularly PE. This may simply reflect that patients with significant comorbidities are at a greater risk of VTE in the perioperative period.16 However, suboptimal documentation about the nature of specific arrhythmias combined with the small number of events made additional analysis difficult, and we could not rule out the influence of unmeasured variables. The risk of VTE has previously been shown to increase with age.17 We also found a correlation between age and DVT and a marginal correlation with any VTE. The lack of correlation with PE could be explained by the underdetection of VTEs. DVT is detectable by compression ultrasonography in only 50% of patients with PE,18 and Moser and colleagues19 reported that up to 40% of patients with DVT may have asymptomatic PE. Thus, some of the patients with PE may have had DVT that was not diagnosed or had completely embolized, and some patients with DVT may have had undetected PE. EBL correlated strongly with VTE. Large-volume blood loss can lead to coagulopathy and hypothermia, both of which are well-known risk factors for developing clots. Therefore, it seems logical that increasing blood loss would correlate with VTE. Schultz and colleagues20 noted a proclivity for PEs in trauma patients with large blood losses. Transfusion with blood products, especially fresh frozen plasma, has been linked to VTE by Abu-Rustum and coworkers.21 However, we were unable to reliably capture the transfusion status of each patient to study this variable adequately. In addition, many factors are involved in operative blood loss, including tumor size (or stage), operative approach, and operative time. Thus, the likelihood of interaction with the significant variables is high. However, a larger cohort would be necessary for meaningful multivariate analysis. One concern in using perioperative heparin is the increased risk of bleeding. The risk of bleeding is of particular concern in patients undergoing partial nephrectomy. As a result of stage migration in renal cortical tumors, we have observed a marked trend toward more nephron-sparing surgery (unpublished data). However, with this technique, hemostasis is often tenuous, and the hemoglobin must be monitored carefully postoperatively. Decreasing the natural clotting ability may lead to a greater incidence of hemorrhage from the surgical field in this population. The published urologic reports are devoid of studies supporting or refuting 991
perioperative heparin use, and adopting policies from other operative experiences may not be appropriate without solid evidence of benefit in this patient population. Many variables contribute to the length of hospital stay. The present study found that patients with DVT or PE tended to have longer hospital stays (mean 11.4 days versus 6.3 days; P ⫽ 0.01). Whether the VTE is the cause or the result is more difficult to determine. Many factors contribute to longer hospital stays after nephrectomy, including poorly controlled pain, failure to ambulate, ileus, or any of a myriad of postoperative complications. Clearly, early mobilization is crucial in preventing DVT. In recent years, there has been a trend toward less invasive interventions, shorter hospitalization, and better acute pain management. All these factors should have a salutary impact on the incidence of VTE; however, no difference was noted between the laparoscopic and open approaches, and the incidence of VTE did not change during the course of this study. Our study raises the question of whether patients undergoing nephrectomy or partial nephrectomy would benefit from heparin, as advocated by the American College of Chest Physicians, in addition to pneumatic compression. Although it is not appropriate to make treatment recommendations on the basis of a single retrospective series without a control group, our results raise the possibility that renal cell carcinoma and nephrectomy do not necessarily pose the same VTE risk as do other malignancies and that perhaps only those with increased risk factors for VTE such as advanced age, high T stage, large perioperative blood loss, or a history of cardiac arrhythmia, coronary artery disease, or DVT should be considered for more extensive prophylaxis. CONCLUSIONS Clinically evident VTE after partial or radical nephrectomy for renal cortical tumors are rare. Pneumatic compression devices appear to provide excellent prophylaxis against DVT and PE. Heparin may not be necessary in all patients undergoing these procedures. REFERENCES 1. Kakkar AK, de Lorenzo F, Pineo GF, et al: Venous thromboembolism and cancer. Baillieres Clin Haematol 11: 675– 687, 1998.
992
2. Rasmussen MS: Preventing thromboembolic complications in cancer patients after surgery: a role for prolonged thromboprophylaxis. Cancer Treat Rev 28: 141–144, 2002. 3. Bergqvist D, Agnelli G, Cohen AT, et al: Duration of prophylaxis against venous thromboembolism with enoxaparin after surgery for cancer. N Engl J Med 346: 975–980, 2002. 4. Thromboembolic Risk Factors (THRIFT) Consensus Group: Risk of and prophylaxis for venous thromboembolism in hospital patients. BMJ 305: 567–574, 1992. 5. Geerts WH, Heit JA, Clagett GP, et al: Prevention of venous thromboembolism. Chest 119: 132S–175S, 2001. 6. Proctor MC, and Greenfield LJ: Thromboprophylaxis in an academic medical center. Cardiovasc Surg 9: 426 – 430, 2001. 7. Clagett GP: Prevention of postoperative venous thromboembolism: an update. Am J Surg 168: 515–522, 1994. 8. Stephenson AJ, Hakimi AA, Snyder ME, et al: Complications of radical and partial nephrectomy in a large contemporary cohort. J Urol 171: 130 –134, 2004. 9. Clagett GP, Anderson FA Jr, Geerts W, et al: Prevention of venous thromboembolism. Chest 114: 531S–560S, 1998. 10. White RH, and Romano PS: Incidence of symptomatic venous thromboembolism after different elective or urgent surgical procedures. Thromb Haemost 90: 446 – 455, 2003. 11. Johnson MJ, Sproule MW, and Paul J: The prevalence and associated variables of deep venous thrombosis in patients with advanced cancer. Clin Oncol (R Coll Radiol) 11: 105– 110, 1999. 12. Joslyn SA, Sirintrapun SJ, and Konety BR: Impact of lymphadenectomy and nodal burden in renal cell carcinoma: retrospective analysis of the National Surveillance, Epidemiology, and End Results database. Urology 65: 675– 680, 2005. 13. Giannakopoulos X, Charalabopoulos K, Charalabopoulos A, et al: The role of lymphadenectomy in renal cancer surgery: an update. Exp Oncol 26: 261–264, 2004. 14. Dechet CB, Zincke H, Sebo TJ, et al: Prospective analysis of computerized tomography and needle biopsy with permanent sectioning to determine the nature of solid renal masses in adults. J Urol 169: 71–74, 2003. 15. Dechet CB, Sebo T, Farrow G, et al: Prospective analysis of intraoperative frozen needle biopsy of solid renal masses in adults. J Urol 162: 1282–1285, 1999. 16. Kikura M, Takada T, and Sato S: Preexisting morbidity as an independent risk factor for perioperative acute thromboembolism syndrome. Arch Surg 140: 1210 –1218, 2005. 17. Anderson FA Jr, Wheeler HB, Goldberg RJ, et al, for the Worcester DVT Study: A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. Arch Intern Med 151: 933– 938, 1991. 18. Kearon C: Diagnosis of pulmonary embolism. Can Med Assoc J 168: 183–194, 2003. 19. Moser KM, Fedullo PF, LitteJohn JK, et al: Frequent asymptomatic pulmonary embolism in patients with deep venous thrombosis. JAMA 271: 223–235, 1994. 20. Schultz DJ, Brasel KJ, Washington L, et al: Incidence of asymptomatic pulmonary embolism in moderately to severely injured trauma patients. J Trauma 56: 727–733, 2004. 21. Abu-Rustum NR, Richard S, Wilton A, et al: Transfusion utilization during adnexal or peritoneal cancer surgery: effects on symptomatic venous thromboembolism and survival. Gynecol Oncol 99: 320 –326, 2005.
UROLOGY 68 (5), 2006
PERIOPERATIVE CLINICAL THROMBOEMBOLIC EVENTS AFTER RADICAL OR PARTIAL NEPHRECTOMY JOSEPH A. PETTUS, SCOTT E. EGGENER, AHMAD SHABSIGH, BRENT YANKE, MARK E. SNYDER, ANGEL SERIO, ANDREW VICKERS, PAUL RUSSO, AND S. MACHELE DONAT
ABSTRACT Objectives. To evaluate the incidence of, and identify the risk factors for, clinical thromboembolic events after radical/partial nephrectomy. Cancer is an established risk factor for deep vein thrombosis (DVT) and pulmonary embolism (PE); however, their incidence after nephrectomy for renal tumors has been poorly studied. Methods. We reviewed our prospective institutional renal database and identified 2208 patients who underwent renal tumor surgery from January 1989 to July 2005. The clinical parameters evaluated were age, sex, race, body mass index, smoking history, medical comorbidities, American Society of Anesthesia grade, procedure type, estimated blood loss, and length of hospitalization. Hospital records, discharge “International Classification of Diseases, Ninth Revision” codes, and 30-day postoperative morbidity and mortality data were reviewed to identify patients diagnosed with perioperative DVT or PE. Results. A total of 34 (1.5%, 95% confidence interval 1.1% to 2.1%) thromboembolic events (20 PEs and 14 DVTs) were identified in 33 patients. Patients with a preoperative history of arrhythmia (P ⫽ 0.02) or prior DVT (P ⫽ 0.053) were more likely to experience PE. The estimated blood loss was directly associated with an increased risk of PE (P ⫽ 0.001). Patients with coronary artery disease (P ⫽ 0.050) or of advanced age (P ⫽ 0.02) were more likely to experience DVT (P ⫽ 0.02). Conclusions. To our knowledge, this is the first study on the incidence of thromboembolic events after nephrectomy. Thromboembolic events are rare but are more likely to occur in patients with coronary artery disease, cardiac arrhythmia, prior DVT, Stage 3 or 4 tumors, or a large estimated blood loss. UROLOGY 68: 988–992, 2006. © 2006 Elsevier Inc.
P
atients with cancer undergoing surgical procedures are at increased risk of deep venous thrombosis (DVT) or pulmonary embolus (PE) during the perioperative period.1–3 Elderly patients and those undergoing abdominal or pelvic surgery are at particularly high risk; therefore, prophylactic measures are generally recommended.4,5 The incidence of venothromboembolism (VTE) has been well studied in general surgical and gynecologic studies but not urologic reports.5,6 Therefore, recommendations for urologic patients have been largely extrapolated from the general surgery data.5 From the Departments of Urology and Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, New York Reprint requests: Machele Donat, M.D., Department of Urology, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. E-mail: [email protected] Submitted: February 28, 2006, accepted (with revisions): June 15, 2006 © 2006 ELSEVIER INC. 988
ALL RIGHTS RESERVED
Compared with patients without cancer, screened surgical patients with cancer have a greater postoperative incidence of proximal lower-extremity DVTs (10% to 20% versus 2% to 4%) and calf DVTs (40% to 80% versus 10% to 20%).7 Although the overall complication rate after renal tumor surgery has been reported to be 17%,8 the VTE rate in patients after partial or radical nephrectomy has not been well studied. Despite the recommendations supporting the use of heparin,2–5,9 we have been impressed by the rarity of VTE in patients treated only with pneumatic compression boots. The purpose of this study was to evaluate a large cohort of patients from a single institution to determine the incidence of perioperative clinical VTE in patients undergoing radical or partial nephrectomy for renal masses and to define risk factors that might influence the decision to administer prophylactic perioperative anticoagulation. 0090-4295/06/$32.00 doi:10.1016/j.urology.2006.06.026
TABLE I.
Patient characteristics at nephrectomy
Age (yr) BMI (kg/m2) EBL (mL) OR time (min) Hospital stay (days) Sex Male Female Race White Black Hispanic Asian Black Hispanic Not recorded ASA grade 1 2 3 4 Missing Tumor laterality Right Left Missing Type of surgery Open radical Open partial Laparoscopic radical Laparoscopic partial Pathologic stage T1 T2 T3 T4 Other* Missing N stage N0/Nx N⫹ M stage M0/Mx M⫹
62.6 (53.2, 70.5) 28 (25, 31) 300 (200, 600) 185 (145, 235) 5 (4, 7) 1402 (64) 806 (37) 1763 (80) 103 (5) 59 (3) 42 (2) 9 (0.4) 232 (10) 126 (6) 1078 (49) 679 (31) 13 (1) 312 (14) 1120 (51) 1087 (49) 1 (0.05) 1325 (60) 770 (35) 77 (3) 33 (2) 1293 (59) 212 (10) 539 (24) 22 (1) 121 (5) 21 (1) 2147 (97) 61 (3) 2071 (94) 137 (6)
KEY: BMI ⫽ body mass index; EBL ⫽ estimated blood loss; OR ⫽ operating room; ASA ⫽ American Society of Anesthesiologists. Data presented as median, with interquartile range in parentheses, or as number, with percentages in parentheses. * T stage did not apply.
MATERIAL AND METHODS After institutional review board approval, we reviewed our prospectively collected departmental renal tumor database for all partial and radical nephrectomies performed for primary renal cortical tumors from January 1989 to July 2005. Demographic information, preoperative morbidities, American Society of Anesthesia grade, tumor laterality, smoking history, body mass index, procedure type (ie, laparoscopic versus open, radical versus partial nephrectomy), operative time, estimated blood loss (EBL), pathologic stage, and postoperative hospital stay were recorded for each patient (Table I). The UROLOGY 68 (5), 2006
postoperative hospital stay was calculated as the time in days from the date of surgery to discharge. Admission before surgery was not considered. We collapsed the comorbidities into general categories to make the statistical analysis more robust. For instance, all cardiac arrhythmias were grouped together. The pathologic stage was assigned only to patients with renal cell carcinoma in accordance with the 2002 TMN staging system. We conducted a chart review to identify patients who experienced VTE, defined as either clinically evident DVT and/or PE during the 30-day perioperative period. We also reviewed the discharge “International Classification of Diseases, Ninth Revision” codes and the departmental morbidity and mortality reports during the study period to identify patients who had experienced a clinically apparent VTE. During the study period, the DVT prophylaxis protocol consisted only of intraoperative and postoperative pneumatic compression devices (Venodyne boots) but no heparin. Patients taking an anticoagulation agent preoperatively started taking the drug again after resumption of oral intake and the hemoglobin was stable. Patients who had nonprimary renal cortical tumors were excluded. We used Fisher’s exact test to assess the relationship between outcome and each comorbidity. The associated confidence intervals were determined using exact methods. For continuous predictors, logistic regression analysis was used to assess their relationship with outcome. We used Stata, version 8.2 (Stata, College Station, Tex) and StatXact (Cytel, Cambridge, Mass) for all statistical analysis. Missing data were excluded from analysis on a case-by-case basis. We considered P ⬍0.05 to be significant.
RESULTS A total of 2208 patients met the study criteria. The demographic and clinical/pathologic characteristics of the study population are listed in Table I. Each VTE, except for one, was documented with imaging, either computed tomography angiography or ventilation-perfusion scan for PE and duplex ultrasonography for DVT. The patient without imaging studies was diagnosed with PE during surgery by arterial blood gas changes and died before the operation concluded. A total of 33 patients (1.5%) experienced a clinical VTE at a median of 7 days (interquartile range 3, 13) postoperatively (Table II), with 1 patient experiencing both DVT and PE. The only significant operative variable associated with having VTE was the EBL (P ⱕ0.001), with the operative time approaching significance (P ⫽ 0.06). Arrhythmia was the only medical comorbidity significantly associated with VTE (P ⫽ 0.02). Lymph node disease was marginally associated with having an event (P ⫽ 0.06) but neither T stage nor M stage was significant. The operative type and approach were not significantly associated with either event. No postoperative deaths occurred as a result of VTE. DVT occurred in 14 patients (0.6%) at a median of 10 days (interquartile range 8, 14) postoperatively and was confirmed by imaging done on the basis of evaluation of clinically suspicious physical examination findings or patient symptoms. The likelihood of clinical DVT increased with age (P ⫽ 989
TABLE II.
Clinicopathologic and comorbidity associations with VTE VTE/Total* (n)
Age (per yr) BMI (per kg/m2 unit) EBL (per 100 mL) OR time (per 10 min) Length of stay T stage ⬎T2 N stage ⬎N0 M stage ⬎M0 Malignant Year of surgery ASA† Comorbidity Arrhythmia Cancer history Clotting disorder CAD CHF COPD CVA Diabetes Previous DVT Hypercholesterolemia Hypertension Smoking TIA
26/1855
4/66 7/628 0/13 3/135 0/12 1/40 0/24 2/228 1/6 4/195 17/1059 19/1068 1/13
OR 1.02 1.0 1.02 1.03 1.032 1.6 3.7 1.2 0.77 1.0
4.7 0.67 0 2.2 0 1.7 0 1.2 14 1.4 1.3 1.3 5.6
95% CI
P Value
1.0–1.1 0.91–1.1 1.01–1.02 1.0–1.1 1.0–1.1 0.71–3.4 0.7–13 0.23–3.9 0.29–2.6 0.96–1.2
0.08 0.8 ⬍0.001 0.06 0.005 0.2 0.06 0.7 0.6 0.35 0.2
1.2–14 0.25–1.6 0–23 0.54–6.3 0–25 0–11 0–11 0.3–3.5 0.28–126 0.36–4.1 0.63–2.9 0.63–2.9 0.13–40
0.02 0.4 1 0.14 1 0.5 1 0.8 0.09 0.5 0.7 0.5 0.18
KEY: VTE ⫽ venothromboembolism; OR ⫽ odds ratio; CI ⫽ confidence interval; CAD ⫽ coronary artery disease; CHF ⫽ congestive heart failure; COPD ⫽ chronic obstructive pulmonary disease; CVA ⫽ cerebral vascular accident; TIA ⫽ transient ischemic attack; other abbreviations as in Table I. * Total number of comorbidities. † No OR presented because ASA has 4 levels.
0.02) and a history of coronary artery disease (P ⫽ 0.050). The mean age of the patients with DVT was 70 versus 61 years for those without DVT (P ⫽ 0.02). Patients with DVT were more likely to have nonorgan-confined disease (Stage T2 or worse, P ⫽ 0.02). A marginally significant association was seen between DVT and cardiac arrhythmia (P ⫽ 0.06). We observed a trend between increasing operative blood loss and DVT, such that the odds of DVT increased by a factor of 1.017 for each 100 mL of EBL (P ⫽ 0.08). Tumor laterality, nodal involvement, metastatic disease, body mass index, and operative time were not significantly associated with DVT. Procedure type (laparoscopic versus open, and partial versus radical nephrectomy) had no significant impact on DVT incidence. The annual incidence of DVT did not change during the course of the study. PE occurred in a total of 20 patients (0.9%) at a median of 4 days (interquartile range 2, 10) postoperatively. As the EBL increased, so did the risk of perioperative PE (P ⫽ 0.001), with the odds of experiencing a clinical PE increasing by a factor of 1.025 for each 100 mL of EBL. Patients with a prior history of DVT (P ⫽ 0.053) or arrhythmia (P ⫽ 0.02) were also more likely to experience PE. Again, procedure type and approach, TMN stage, 990
body mass index, and operative time were not significantly associated with PE, and the annual incidence did not change significantly. COMMENT Malignancy is a well-established risk factor for developing VTE. The American College of Chest Physicians currently recommends that all patients undergoing major urologic surgery for malignancy be treated with elastic stockings and/or pneumatic compression devices and either subcutaneous unfractionated heparin or low-molecular-weight heparin in the perioperative period.5 These recommendations have been extrapolated from thromboembolic events in large radical prostatectomy and general surgical series. In this study, we examined the incidence of thromboembolic events in all patients undergoing radical or partial nephrectomy for a renal cortical tumor. The incidence of perioperative DVT in asymptomatic screened patients has been reported to be 25% in a pooled analysis of general surgical patients without any prophylaxis and 4% in patients with pneumatic compression devices.5 A population-based study of the California Discharge Database found the 90-day perioperative VTE rate to be UROLOGY 68 (5), 2006
0.4% and 3% for patients undergoing nephrectomy for benign and malignant disease, respectively, but the prophylaxis status was unknown in that study.10 In this study, we analyzed clinically evident DVTs in patients treated only with pneumatic compression devices and found the rate to be quite low at 0.6%. This low rate is likely explained by the underreporting of events. Our patients were unscreened, and we only included the 30-day perioperative period, when most events occur. Had we extended the study period to include the 90-day perioperative period, the rate may have been greater. Patients may have had more events in the postdischarge period and either did not seek care or sought care elsewhere and did not report it to us. We most likely would have identified more patients with DVT had we used routine Doppler screening of asymptomatic patients postoperatively, but the significance of asymptomatic postoperative DVT is unknown. DVT has been reported to be very prevalent in patients with advanced cancer. Johnson and colleagues11 reported a 52% incidence among 258 patients with advanced cancer. Our study showed a relationship between perioperative DVT development and nonorgan-confined disease, but nodal and metastatic disease in renal cell carcinoma was not associated with DVT. The relationship between locally advanced disease and DVT may be explained by the venous stasis caused by the presence of tumor in the renal vein or vena cava, as well as mechanical compression of the vena cava by the tumor. It is certainly possible that some DVTs were present preoperatively but not detected until after surgery. The benefit of lymphadenectomy in radical or partial nephrectomy has not been established,12,13 and we have not consistently included lymphadenectomy at our institution. Therefore, the small portion of patients with lymph node involvement may have been a result of undersampling, and the true relationship between nodal status and VTE may not have been revealed in our study population. Finally, performing nephrectomy in the setting of metastatic disease is usually done as a cytoreductive procedure before initiating immunotherapy, and these patients are highly selected in terms of metastatic disease burden and performance status. We did not compare patients with benign tumors with those with malignant tumors, but analyzed them on an intent-to-treat basis. This was done because all renal cortical tumors in this study were assumed to be renal cell carcinoma at surgery on the basis of clinical and radiographic findings preoperatively. Preoperative renal biopsies and intraoperative frozen section analysis findings are often inaccurate and therefore not routinely performed in the absence of another primary malignancy.14,15 UROLOGY 68 (5), 2006
The pathologic findings are not typically available for several days after surgery. Therefore, pathologic confirmation of malignancy is not available to make decisions regarding thromboembolic prophylaxis. Patients with a cardiac arrhythmia history had greater rates of VTE, particularly PE. This may simply reflect that patients with significant comorbidities are at a greater risk of VTE in the perioperative period.16 However, suboptimal documentation about the nature of specific arrhythmias combined with the small number of events made additional analysis difficult, and we could not rule out the influence of unmeasured variables. The risk of VTE has previously been shown to increase with age.17 We also found a correlation between age and DVT and a marginal correlation with any VTE. The lack of correlation with PE could be explained by the underdetection of VTEs. DVT is detectable by compression ultrasonography in only 50% of patients with PE,18 and Moser and colleagues19 reported that up to 40% of patients with DVT may have asymptomatic PE. Thus, some of the patients with PE may have had DVT that was not diagnosed or had completely embolized, and some patients with DVT may have had undetected PE. EBL correlated strongly with VTE. Large-volume blood loss can lead to coagulopathy and hypothermia, both of which are well-known risk factors for developing clots. Therefore, it seems logical that increasing blood loss would correlate with VTE. Schultz and colleagues20 noted a proclivity for PEs in trauma patients with large blood losses. Transfusion with blood products, especially fresh frozen plasma, has been linked to VTE by Abu-Rustum and coworkers.21 However, we were unable to reliably capture the transfusion status of each patient to study this variable adequately. In addition, many factors are involved in operative blood loss, including tumor size (or stage), operative approach, and operative time. Thus, the likelihood of interaction with the significant variables is high. However, a larger cohort would be necessary for meaningful multivariate analysis. One concern in using perioperative heparin is the increased risk of bleeding. The risk of bleeding is of particular concern in patients undergoing partial nephrectomy. As a result of stage migration in renal cortical tumors, we have observed a marked trend toward more nephron-sparing surgery (unpublished data). However, with this technique, hemostasis is often tenuous, and the hemoglobin must be monitored carefully postoperatively. Decreasing the natural clotting ability may lead to a greater incidence of hemorrhage from the surgical field in this population. The published urologic reports are devoid of studies supporting or refuting 991
perioperative heparin use, and adopting policies from other operative experiences may not be appropriate without solid evidence of benefit in this patient population. Many variables contribute to the length of hospital stay. The present study found that patients with DVT or PE tended to have longer hospital stays (mean 11.4 days versus 6.3 days; P ⫽ 0.01). Whether the VTE is the cause or the result is more difficult to determine. Many factors contribute to longer hospital stays after nephrectomy, including poorly controlled pain, failure to ambulate, ileus, or any of a myriad of postoperative complications. Clearly, early mobilization is crucial in preventing DVT. In recent years, there has been a trend toward less invasive interventions, shorter hospitalization, and better acute pain management. All these factors should have a salutary impact on the incidence of VTE; however, no difference was noted between the laparoscopic and open approaches, and the incidence of VTE did not change during the course of this study. Our study raises the question of whether patients undergoing nephrectomy or partial nephrectomy would benefit from heparin, as advocated by the American College of Chest Physicians, in addition to pneumatic compression. Although it is not appropriate to make treatment recommendations on the basis of a single retrospective series without a control group, our results raise the possibility that renal cell carcinoma and nephrectomy do not necessarily pose the same VTE risk as do other malignancies and that perhaps only those with increased risk factors for VTE such as advanced age, high T stage, large perioperative blood loss, or a history of cardiac arrhythmia, coronary artery disease, or DVT should be considered for more extensive prophylaxis. CONCLUSIONS Clinically evident VTE after partial or radical nephrectomy for renal cortical tumors are rare. Pneumatic compression devices appear to provide excellent prophylaxis against DVT and PE. Heparin may not be necessary in all patients undergoing these procedures. REFERENCES 1. Kakkar AK, de Lorenzo F, Pineo GF, et al: Venous thromboembolism and cancer. Baillieres Clin Haematol 11: 675– 687, 1998.
992
2. Rasmussen MS: Preventing thromboembolic complications in cancer patients after surgery: a role for prolonged thromboprophylaxis. Cancer Treat Rev 28: 141–144, 2002. 3. Bergqvist D, Agnelli G, Cohen AT, et al: Duration of prophylaxis against venous thromboembolism with enoxaparin after surgery for cancer. N Engl J Med 346: 975–980, 2002. 4. Thromboembolic Risk Factors (THRIFT) Consensus Group: Risk of and prophylaxis for venous thromboembolism in hospital patients. BMJ 305: 567–574, 1992. 5. Geerts WH, Heit JA, Clagett GP, et al: Prevention of venous thromboembolism. Chest 119: 132S–175S, 2001. 6. Proctor MC, and Greenfield LJ: Thromboprophylaxis in an academic medical center. Cardiovasc Surg 9: 426 – 430, 2001. 7. Clagett GP: Prevention of postoperative venous thromboembolism: an update. Am J Surg 168: 515–522, 1994. 8. Stephenson AJ, Hakimi AA, Snyder ME, et al: Complications of radical and partial nephrectomy in a large contemporary cohort. J Urol 171: 130 –134, 2004. 9. Clagett GP, Anderson FA Jr, Geerts W, et al: Prevention of venous thromboembolism. Chest 114: 531S–560S, 1998. 10. White RH, and Romano PS: Incidence of symptomatic venous thromboembolism after different elective or urgent surgical procedures. Thromb Haemost 90: 446 – 455, 2003. 11. Johnson MJ, Sproule MW, and Paul J: The prevalence and associated variables of deep venous thrombosis in patients with advanced cancer. Clin Oncol (R Coll Radiol) 11: 105– 110, 1999. 12. Joslyn SA, Sirintrapun SJ, and Konety BR: Impact of lymphadenectomy and nodal burden in renal cell carcinoma: retrospective analysis of the National Surveillance, Epidemiology, and End Results database. Urology 65: 675– 680, 2005. 13. Giannakopoulos X, Charalabopoulos K, Charalabopoulos A, et al: The role of lymphadenectomy in renal cancer surgery: an update. Exp Oncol 26: 261–264, 2004. 14. Dechet CB, Zincke H, Sebo TJ, et al: Prospective analysis of computerized tomography and needle biopsy with permanent sectioning to determine the nature of solid renal masses in adults. J Urol 169: 71–74, 2003. 15. Dechet CB, Sebo T, Farrow G, et al: Prospective analysis of intraoperative frozen needle biopsy of solid renal masses in adults. J Urol 162: 1282–1285, 1999. 16. Kikura M, Takada T, and Sato S: Preexisting morbidity as an independent risk factor for perioperative acute thromboembolism syndrome. Arch Surg 140: 1210 –1218, 2005. 17. Anderson FA Jr, Wheeler HB, Goldberg RJ, et al, for the Worcester DVT Study: A population-based perspective of the hospital incidence and case-fatality rates of deep vein thrombosis and pulmonary embolism. Arch Intern Med 151: 933– 938, 1991. 18. Kearon C: Diagnosis of pulmonary embolism. Can Med Assoc J 168: 183–194, 2003. 19. Moser KM, Fedullo PF, LitteJohn JK, et al: Frequent asymptomatic pulmonary embolism in patients with deep venous thrombosis. JAMA 271: 223–235, 1994. 20. Schultz DJ, Brasel KJ, Washington L, et al: Incidence of asymptomatic pulmonary embolism in moderately to severely injured trauma patients. J Trauma 56: 727–733, 2004. 21. Abu-Rustum NR, Richard S, Wilton A, et al: Transfusion utilization during adnexal or peritoneal cancer surgery: effects on symptomatic venous thromboembolism and survival. Gynecol Oncol 99: 320 –326, 2005.
UROLOGY 68 (5), 2006