Available online at www.sciencedirect.com
ScienceDirect EJSO 40 (2014) 85e89
www.ejso.com
Predictive factors of hemorrhagic complications after partial nephrectomy T. Fardoun a,*, D. Chaste a, E. Oger b, R. Mathieu a, B. Peyronnet a, N. Rioux-Leclercq c, G. Verhoest a, J.J. Patard d, K. Bensalah a a
Department of Urology, Rennes University Hospital, 2, rue Henri Le Guilloux, 35033 Rennes Cedex, France Department of Biostatistics, Rennes University Hospital, 2, rue Henri Le Guilloux, 35033 Rennes Cedex, France c Department of Pathology, Rennes University Hospital, 2, rue Henri Le Guilloux, 35033 Rennes Cedex, France d Department of Urology, Bicetre University Hospital, 78, rue du General Leclerc, 94270 Le Kremlin-Bic^etre, Paris, France b
Accepted 3 November 2013 Available online 13 November 2013
Abstract Objectives: To identify the predictive factors of hemorrhagic complications (HC) in a contemporary cohort of patients who underwent partial nephrectomy (PN). Materials and methods: Records of 199 consecutive patients who underwent PN between 2008 and 2012 at our institution were retrospectively analyzed. HC was defined as a hematoma requiring transfusion, an arterio-veinous fistula, a false aneurysm or a post-operative decrease of hemoglobin >3 g/dl. Patients with or without HC were compared using Wilcoxon and Fisher exact tests for continuous and categorical variables, respectively. We performed a univariate and multivariate analysis with a logistic regression model using the occurrence of an HC as the dependent variable. Results: 54% of the patients were male with a median age of 61 (22e86) years. Median BMI was 26 (18e47) kg/m2. Surgery was done open, laparoscopically or with robotic assistance in 106, 54 and 39 cases, respectively. Global complication rate was 40% including 21.6% HC. There were more complex tumors (75.6% vs. 66.5%, p ¼ 0.04) and median length of stay was increased (11 days compared to 7 days, p < 0.0001) in case of a HC. In univariate analysis, imperative indication ( p ¼ 0.08), RENAL score ( p ¼ 0.07), operating time ( p ¼ 0.07) and operative blood loss > 250 ml ( p ¼ 0.002) were statistically relevant. In multivariate analysis, only operative blood loss >250 ml was identified as a predictive factor of HC ( p ¼ 0.0007). Conclusion: Patients who underwent a procedure with estimated blood loss >250 ml should be carefully monitored in the postoperative course. Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. Keywords: Partial nephrectomy; Hemorrhage; Complications; RENAL score
Introduction Nephron sparing surgery (NSS) is the optimal management of small localized renal cell carcinomas (RCC).1e3 Partial nephrectomy (PN) provides oncological outcomes similar to that of radical nephrectomy (RN)4e6 with better preservation of renal function.7,8 However, PN
Abbreviations: NSS, nephron sparing surgery; RCC, renal cell carcinoma; PN, partial nephrectomy; RN, radical nephrectomy; HC, hemorrhagic complication; RS, RENAL Score; MDRD, modification of diet in renal disease. * Corresponding author. Tel.: þ33 2 99 28 42 69; fax: þ33 2 99 28 41 13. E-mail address: [email protected] (T. Fardoun).
remains underused and mostly performed in tertiary care centers.9,10 A possible explanation for this lack of diffusion is the potential morbidity of PN.5,11e13 In many studies, complication rates of PN lie between 10 and 50%, according to tumor complexity.14e16 The most frequent complications of PN are hemorrhagic17 (ie: blood loss, perirenal hematoma, arterio-venous fistula and false aneurysm). These hemorrhagic complications (HC) can lead to a need for transfusion, embolization and/or reoperation that can raise morbidity, increase length of hospital stay and delay return to normal activities. Our objective was to evaluate predictive factors of HC after PN in a contemporary cohort of patients operated both open and laparoscopically.
0748-7983/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejso.2013.11.006
86
T. Fardoun et al. / EJSO 40 (2014) 85e89
Materials and methods
Results
This was a retrospective study of 199 patients who underwent PN at our institution for a renal tumor between January 2008 and March 2012. After informed consent, all their records were prospectively entered into a specific RCC database. All cases were done consecutively. Patients were mostly operated by two surgeons. Operative technique (open, laparoscopic with or without robotic assistance) was decided by the surgeon based on his experience and tumor complexity. Data regarding patient demographics, tumor characteristics, and surgical outcomes were reviewed. Comorbidities were graded using Charlson index. The presence of anticoagulant therapy (either vitamin K antagonists or antiplatelet agent) was noted. Tumor complexity was evaluated with the RENAL nephrometry score (RS)18 and tumors were classified as: low complexity (RENAL score between 4 and 6), moderate complexity (RENAL score between 7 and 9) and high complexity (RENAL score > 9 or hilar tumors). Patients were operated either open or laparoscopically with or without robotic assistance. The open access was performed extraperitoneally through a flank incision. The tumor was dissected free from the perirenal fat. Most of the time, tumor resection was performed under pedicle clamping. The resection margin was assessed macroscopically as we do not routinely ask for frozen examination. A 2.0 running suture on the resection bed was made to ensure hemostasis and closure of the collecting system. Renal reconstruction was achieved with interrupted 0 sutures. All laparoscopic interventions were performed transperitoneally. The principles of tumor resection were the same as with the open access. For all operation, hemostatic agents were used at the surgeon’s discretion if needed. We studied the use of the following hemostatic agents: gelatin matrix with human thrombin (FlosealÒ), cellulose surgical bolster (SurgicelÒ) and biological surgical glue (BioglueÒ). All drugs could have been used separately or combined. Surgeon’s experience was appraised considering that an experienced surgeon had performed more than 20 PN with a specific surgical approach (for example, even if a surgeon had performed more than 20 laparoscopic PN, we considered that he was still unexperienced in his first 20 robotic cases). Complications that occurred within one month after surgery were classified using the ClavieneDindo system. We distinguished minor (grade I and II) from major complications (grade IIIeV). HC were defined as the occurrence of a perirenal hematoma requiring transfusion, an arterio-venous fistula, a false aneurysm or a decrease of hemoglobin >3 g/dl.
Patients data
Statistical analysis We compared patients with or without HC using the Wilcoxon test for continuous variables and the Fisher exact test for categorical variables. We performed a univariate and multivariate analysis with a logistic regression model using the occurrence of a HC as the dependent variable.
Demographics are shown in Table 1. 108 patients were male. Median age was 61 (22e86) years old. Median BMI was 26 (18e47) kg/m2. Median Charlson comorbidity index was 4 (0e11). Median baseline GFR was 86 (31e268) ml/min/1.73 m2 and 26 patients had chronic renal disease (defined as MDRD < 15 ml/min). 34 patients were receiving anticoagulants before surgery. Pathological data Tumor characteristics are summarized in Table 1. Median tumor size was 35 mm (10e110). 24 patients underwent imaging guided renal biopsy before surgery. On final pathologic exam, 165 tumors were malignant. According to RENAL score, 60, 83 and 46 tumors were classified as low, moderately and highly complex, respectively. Operative and perioperative data Details are summarized in Table 2. The indication was elective in 166 cases. 106 PN were done open, 54 by laparoscopy and 39 with robotic assistance. The surgeon was considered experimented in 142 procedures. Median Table 1 Patient and tumor characteristics. Variables Gender: male, n (%) Median age, years (range) Median BMI, kg/m2 (range) Median Charlson Index (range) Charlson index, n (%) 0e1 2 3 4 Chronic renal disease at surgery, n (%) Median baseline MDRD, ml/min (range) Anticoagulant treatment before surgery, n (%) Preoperative biopsy, n (%) Malignant tumors, n (%) Histology, n (%) Clear cell Papillary Chromophobe Other Median tumor size, mm (range) Fuhrman grade, n (%) 1 2 3 4 RENAL score, n (%) Low complexity Intermediate complexity High complexity
108 61 26 4
(54) (22e86) (18e47) (0e11)
50 39 39 71 26 86 34 24 165
(25) (20) (20) (35) (13) (31e268) (17) (12) (82.9)
122 23 15 5 35
(74) (14) (9) (3) (10e110)
14 95 51 3
(9) (58) (32) (1)
60 (32) 83 (44) 46 (24)
T. Fardoun et al. / EJSO 40 (2014) 85e89 Table 2 Peri-operative outcomes.
87
of stay was significantly increased (11 days compared to 7 days, p < 0.0001) in case of a HC.
Variables Imperative indication, n (%) Type of surgery, n (%) Open Laparoscopic Robotic Surgeon’s experience, n (%) Experimented Less than 20 procedures Hemostatic agent, n (%) None FlosealÒ BioglueÒ SurgicelÒ Type of clamping, n (%) None Parenchyma Pedicle Median clamping time, min (range) Median estimated blood loss, ml (range) Median operative time, min (range) Complications, n (%) Major Minor Hemorrhagic Complications, n (%) Hematoma False aneurysm Arterio-venous fistulae Loss of hemoglobin None Hemorrhagic Complications, n (%) Surgical complications Urinary fistula Wound infection - UTI Wound dehiscence Acute urinary retention Ileus Medical complications Pulmonary embolism Deep vein thrombosis Peptic ulcer Acute tubular necrosis Median length of stay, days (range)
33 (17) 106 (53) 54 (27) 39 (20) 142 (71) 57 (29) 4 113 13 160
(2) (57) (6) (80)
22 27 146 20 300 180
(11) (14) (75) (0e55) (20e2000) (70e360)
Regression analysis In univariate analysis (Table 4), imperative indication ( p ¼ 0.08), RENAL score ( p ¼ 0.07), operating time ( p ¼ 0.07) and operative blood loss ( p ¼ 0.002) were identified as statistically relevant to be included in the multivariable model. Neither obesity ( p ¼ 0.59), anticoagulant treatment ( p ¼ 0.45), Charlson comorbidity index ( p > 0.05), GFR status ( p ¼ 0.74), use of hemostatic agent ( p ¼ 0.353), surgical approach ( p > 0.05) or surgeon’s experience ( p ¼ 0.38) could foretell the occurrence of a HC. In multivariate analysis (Table 5), only blood loss > 250 ml was identified as a predictive factor of HC ( p ¼ 0.0007). Discussion
26 (13) 55 (27)
NSS is the recommended treatment of small renal masses.19,20 However, it remains underused: a recent study
15 (8) 11 (5) 3 (1) 14(7)
Table 3 Comparison of patients with and without hemorrhagic complications (HC).
6(3) 5 (2) e 13 (6) 2(1) 3(1) 5(2) 1 1 1 1 8
(0.5) (0.5) (0.5) (0.5) (3e45)
operative time was 180 (70e360) min with a median estimated blood loss of 300 (20e2000) ml. Clamping of the renal pedicle was done in 146 procedures and median ischemia time was 20 (0e55) min. Global complication rate was 40%. 81 complications (26 major and 55 minor) occurred. 43 were HC. There were 15 perirenal hematomas requiring transfusion, 11 false aneurysms, 3 arterio-venous fistulas and 14 patients who lost more than 3 points of hemoglobin after surgery. Comparison of patients with or without HC (Table 3) Patients were comparable for age, BMI, renal function and comorbidities. Median blood loss was higher in the HC group but the difference was not significant (400 vs. 300 ml, respectively, p ¼ 0.2). There were more complex tumors (75.6% vs. 66.5%, p ¼ 0.04) and median length
Group
No HC
Number of patients (%) Median age, years (range) Median BMI, kg/m2 (range) Median Charlson index (range) Chronic Renal Disease (%) Patients receiving anticoagulants (%) Median tumor size, mm (range) Experienced surgeon (%) Surgical approach Open Laparoscopy Robot Renal score category Low Intermediate High Hemostatic agents BioglueÒ FlosealÒ SurgicelÒ Median blood loss, ml (range) Median operating time, min (range) Median length of stay, days (range)
156 (78) 61 (22e82) 25.9 (18e47)
HC
P value
43 (22) 59 (32e86)
0.79
25(18e38.1)
0.76
3 (0e11)
0.41
21 (13.5)
5 (11.6)
1.00
25 (16)
9 (20.9)
0.49
35 (10e90)
0.51
109 (70)
33 (76)
0.44
81 (52) 45 (28) 30 (19)
25 (58) 9 (21) 9 (21)
0.62
50 (33.8) 58 (39.2) 40 (27)
10 (24.4) 25 (61) 6 (14.6)
0.04
4 (0e11)
33 (10e110)
10 87 124 300
(6.4) (56) (79) (20e1500)
3 26 36 400
(7) (60) (84) (40e2000)
1.00 0.6 0.66 0.20
180 (70e360)
180 (90e360)
0.20
7 (3e35)
11 (3e45)
250 ml)
0.077
2.1 (0.92e4.75)
0.314 0.949 0.379 0.353
1.0 0.65 0.97 1.42 1.35
0.691 0.721 0.800 0.07 0.002
1.26 (0.40e3.99) 1.29 (0.31e5.29) 0.99 (0.96e1.03) 1.00 (1.00e1.01) 1.51(1.16e2.00)
(0.37e2.55) (0.22e1.75) (0.27e1.53) (0.59e3.24) (0.30e2.37) (0.98e1.01)
(0.28e1.51) (0.41e2.32) (0.65e3.12) (0.71e2.57)
of almost 25 000 patients showed that even when patients were at risk of chronic renal disease, NSS was performed in less than 20% of the cases.21 Furthermore, NSS is mostly done in tertiary care centers.10 It has been suggested that this underutilization was related to the competition between NSS and laparoscopic radical nephrectomy, which have been both developed and promoted over the same time
Table 5 Results of multivariate analysis. Variables
P value
OR (95%CI)
Imperative indication
0.288
1.73 (0.63e4.81)
Renal score Low Intermediate High
0.247 0.090
1.0 1.76 (0.67e4.61) 0.29 (0.07e1.21)
Operative time
0. 505
1.00 (0.99e1.01)
Operative blood loss >250 ml
0.0007
1.64 (1.23e2.18)
period.9,22 In fact, the most probable explanation is that: 1) PN is technically demanding (particularly when performed laparoscopically) because it requires to remove the tumor an repair the kidney in a limited time and 2) its morbidity is superior to that of radical nephrectomy with HC that can occur in up to 30% of complicated cases.11,14 About 20% of our patients had a HC. In the literature, reported rates of bleeding events are highly variable ranging from 0 to 12%.12,15,23,24 However, all these studies are hardly comparable since the definition of bleeding varies greatly. Some authors don’t distinguish per-operative from post-operative bleeding.25 Some define it as a blood loss requiring at least one unit of blood transfusion.26 Others rely on a biological definition (acute decrease in hemoglobin).27 We chose to define a HC as a hematoma requiring transfusion, the occurrence of a pseudo-aneurysm or an arterio-veinous fistula and a decrease of hemoglobin >3 g/dl since these events are those that directly impact the post-operative course and patient recovery. Our rate of HC is higher than other studies. It can be explained by our definition of HC that was both clinical and biological. Most of the patients who have a significant loss of hemoglobin will not require transfusion. Our study showed that the main predictor of HC after PN was per-operative blood loss, which can be seen as an indirect sign of surgical difficulty. Every 250 ml, there was a 1.6-augmented risk of post-operative HC and when OBL exceed 1000 ml, there was a 12.5 increased risk of post-operative HC. Surprisingly, complexity of tumor was not identified as an independent factor of HC. This finding is in accordance with a recent study that found that there were no raise of incidence of complications when PN was performed for complex tumors.28 Interestingly, we did not find any impact of the surgical technique. Compared to open surgery, laparoscopic and robotic PN did not yield more post-operative HC. This is in contradiction with large multicentric retrospective studies that have suggested that even in expert centers, laparoscopic PN was associated with more complications, particularly HC.29 Other variables that could be felt to have an influence on bleeding were not significant in our statistical model. We did not find any impact of age, comorbidities, anticoagulant treatment, type of clamping, or ischemia time. Similarly, the importance of per-operative blood loss was not associated with post-operative hemorrhage. The surgeon’s experience (defined as more than 20 cases) was not associated to an increased frequency of HC. It might be explained by the fact that at our institution, the vast majority of cases are done or supervised by experienced surgeons (open or laparoscopically). There is some debate concerning the utility of hemostatic agents to ensure hemostasis during PN. It has been suggested that hemostatic gelatin matrix could decrease the
T. Fardoun et al. / EJSO 40 (2014) 85e89
incidence of vascular complications.30 Our analysis did not find any influence of these agents on the prevention of HC. However, we used such agents in a large majority of patients (98%) and it could just be that we don’t have enough patients operated without those agents to show a significant difference. Our study has some limitations. It is a retrospective monocentric study, which reflects the experience of mostly two surgeons. However, we believe the findings are of interest since it is a contemporary study with a good balance between open, laparoscopic and robotic cases that reflect the contemporary management of many tertiary centers. In conclusion, we found that the main predictor of HC after PN was operative blood loss. Several factors that are usually associated with bleeding such as anticoagulant treatment use of hemostatic agents or surgeon experience did not increase the risk of vascular complications. For this reason, patients who underwent a procedure with estimated blood loss >250 ml should be carefully monitored in the postoperative course. Conflicts of interest statement None. References 1. MacLennan S, Imamura M, Lapitan MC, et al. Systematic review of oncological outcomes following surgical management of localised renal cancer. Eur Urol 2012;61:972–93. 2. Ljungberg B, Cowan NC, Hanbury DC, et al. EAU guidelines on renal cell carcinoma: the 2010 update. Eur Urol 2010;58:398–406. 3. Gill IS, Aron M, Gervais DA, Jewett MA. Clinical practice. small renal mass. N Engl J Med 2010;362:624–34. 4. Tan H-J, Norton EC, Ye Z, Hafez KS, Gore JL, Miller DC. Long-term survival following partial vs radical nephrectomy among older patients with early-stage kidney cancer. JAMA 2012;307:1629–35. 5. Patard JJ, Pantuck AJ, Crepel M, et al. Morbidity and clinical outcome of nephron-sparing surgery in relation to tumor size and indication. Eur Urol 2007;52:148–54. 6. Becker F, Siemer S, Hack M, Humke U, Ziegler M, Stockle M. Excellent long-term cancer control with elective nephron-sparing surgery for selected renal cell carcinomas measuring more than 4 cm. Eur Urol 2006;49:1058–63. [discussion 1063e64]. 7. Huang WC. Impact of nephron sparing on kidney function and nononcologic mortality. Urol Oncol 2010;28:568–74. 8. Huang WC, Levey AS, Serio AM, et al. Chronic kidney disease after nephrectomy in patients with renal cortical tumors: a retrospective cohort study. Lancet Oncol 2006;7:735–40. 9. Abouassaly R, Alibhai SM, Tomlinson G, Timilshina N, Finelli A. Unintended consequences of laparoscopic surgery on partial nephrectomy for kidney cancer. J Urol 2010;183:467–72. 10. Hollenbeck BK, Taub DA, Miller DC, Dunn RL, Wei JT. National utilization trends of partial nephrectomy for renal cell carcinoma: a case of underutilization? Urology 2006;67:254–9.
89
11. Van Poppel H, Da Pozzo L, Albrecht W, et al. A prospective randomized EORTC intergroup phase 3 study comparing the complications of elective nephron-sparing surgery and radical nephrectomy for lowstage renal cell carcinoma. Eur Urol 2007;51:1606–15. 12. Pasticier G, Timsit MO, Badet L, et al. Nephron-sparing surgery for renal cell carcinoma: detailed analysis of complications over a 15year period. Eur Urol 2006;49:485–90. 13. Touijer K, Jacqmin D, Kavoussi LR, et al. The expanding role of partial nephrectomy: a critical analysis of indications, results, and complications. Eur Urol 2010;57:214–22. 14. Simhan J, Smaldone MC, Tsai KJ, et al. Objective measures of renal mass anatomic complexity predict rates of major complications following partial nephrectomy. Eur Urol 2011;60:724–30. 15. Ficarra V, Bhayani S, Porter J, et al. Predictors of warm ischemia time and perioperative complications in a multicenter, international series of robot-assisted partial nephrectomy. Eur Urol 2012;61:395–402. 16. Mottrie A, Schatteman P, De Wil P, De Troyer B, Novara G, Ficarra V. Validation of the preoperative aspects and dimensions used for an anatomical (PADUA) score in a robot-assisted partial nephrectomy series. World J Urol 2013;31:799–804. 17. Huber J, Pahernik S, Hallscheidt P, et al. Risk factors and clinical management of haemorrhage after open nephron-sparing surgery. BJU Int 2010;106:1488–93. 18. Kutikov A, Uzzo RG. The R.E.N.A.L. nephrometry score: a comprehensive standardized system for quantitating renal tumor size, location and depth. J Urol 2009;182:844–53. 19. Zini L, Patard JJ, Capitanio U, et al. The use of partial nephrectomy in European tertiary care centers. Eur J Surg Oncol 2009;35:636–62. 20. Miller DC, Hollingsworth JM, Hafez KS, Daignault S, Hollenbeck BK. Partial nephrectomy for small renal masses: an emerging quality of care concern? J Urol 2006;175:853–7. [discussion 858]. 21. Abouassaly R, Finelli A, Tomlinson GA, Urbach DR, Alibhai SM. How often are patients with diabetes or hypertension being treated with partial nephrectomy for renal cell carcinoma? A populationbased analysis. BJU Int 2011;108:1806–12. 22. Bhayani SB, Clayman RV, Sundaram CP, et al. Surgical treatment of renal neoplasia: evolving toward a laparoscopic standard of care. Urology 2003;62:821–6. 23. Heye S, Maleux G, Van Poppel H, Oyen R, Wilms G. Hemorrhagic complications after nephron-sparing surgery: angiographic diagnosis and management by transcatheter embolization. Am J Roentgenol 2005;184:1661–4. 24. Gratzke C, Seitz M, Bayrle F, et al. Quality of life and perioperative outcomes after retroperitoneoscopic radical nephrectomy (RN), open RN and nephron-sparing surgery in patients with renal cell carcinoma. BJU Int 2009;104:470–5. 25. Zimmermann R, Janetschek G. Complications of laparoscopic partial nephrectomy. World J Urol 2008;26:531–7. 26. Richstone L, Montag S, Ost MC, et al. Predictors of hemorrhage after laparoscopic partial nephrectomy. Urology 2011;77:88–91. 27. Wheat JC, Roberts WW, Hollenbeck BK, Wolf Jr JS, Weizer AZ. Complications of laparoscopic partial nephrectomy. Urol Oncol 2013;31:57–62. 28. Dulabon LM, Kaouk JH, Haber GP, et al. Multi-institutional analysis of robotic partial nephrectomy for hilar versus nonhilar lesions in 446 consecutive cases. Eur Urol 2011;59:325–30. 29. Gill IS, Kavoussi LR, Lane BR, et al. Comparison of 1,800 laparoscopic and open partial nephrectomies for single renal tumors. J Urol 2007;178:41–6. 30. Gill IS, Ramani AP, Spaliviero M, et al. Improved hemostasis during laparoscopic partial nephrectomy using gelatin matrix thrombin sealant. Urology 2005;65:463–6.
ScienceDirect EJSO 40 (2014) 85e89
www.ejso.com
Predictive factors of hemorrhagic complications after partial nephrectomy T. Fardoun a,*, D. Chaste a, E. Oger b, R. Mathieu a, B. Peyronnet a, N. Rioux-Leclercq c, G. Verhoest a, J.J. Patard d, K. Bensalah a a
Department of Urology, Rennes University Hospital, 2, rue Henri Le Guilloux, 35033 Rennes Cedex, France Department of Biostatistics, Rennes University Hospital, 2, rue Henri Le Guilloux, 35033 Rennes Cedex, France c Department of Pathology, Rennes University Hospital, 2, rue Henri Le Guilloux, 35033 Rennes Cedex, France d Department of Urology, Bicetre University Hospital, 78, rue du General Leclerc, 94270 Le Kremlin-Bic^etre, Paris, France b
Accepted 3 November 2013 Available online 13 November 2013
Abstract Objectives: To identify the predictive factors of hemorrhagic complications (HC) in a contemporary cohort of patients who underwent partial nephrectomy (PN). Materials and methods: Records of 199 consecutive patients who underwent PN between 2008 and 2012 at our institution were retrospectively analyzed. HC was defined as a hematoma requiring transfusion, an arterio-veinous fistula, a false aneurysm or a post-operative decrease of hemoglobin >3 g/dl. Patients with or without HC were compared using Wilcoxon and Fisher exact tests for continuous and categorical variables, respectively. We performed a univariate and multivariate analysis with a logistic regression model using the occurrence of an HC as the dependent variable. Results: 54% of the patients were male with a median age of 61 (22e86) years. Median BMI was 26 (18e47) kg/m2. Surgery was done open, laparoscopically or with robotic assistance in 106, 54 and 39 cases, respectively. Global complication rate was 40% including 21.6% HC. There were more complex tumors (75.6% vs. 66.5%, p ¼ 0.04) and median length of stay was increased (11 days compared to 7 days, p < 0.0001) in case of a HC. In univariate analysis, imperative indication ( p ¼ 0.08), RENAL score ( p ¼ 0.07), operating time ( p ¼ 0.07) and operative blood loss > 250 ml ( p ¼ 0.002) were statistically relevant. In multivariate analysis, only operative blood loss >250 ml was identified as a predictive factor of HC ( p ¼ 0.0007). Conclusion: Patients who underwent a procedure with estimated blood loss >250 ml should be carefully monitored in the postoperative course. Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. Keywords: Partial nephrectomy; Hemorrhage; Complications; RENAL score
Introduction Nephron sparing surgery (NSS) is the optimal management of small localized renal cell carcinomas (RCC).1e3 Partial nephrectomy (PN) provides oncological outcomes similar to that of radical nephrectomy (RN)4e6 with better preservation of renal function.7,8 However, PN
Abbreviations: NSS, nephron sparing surgery; RCC, renal cell carcinoma; PN, partial nephrectomy; RN, radical nephrectomy; HC, hemorrhagic complication; RS, RENAL Score; MDRD, modification of diet in renal disease. * Corresponding author. Tel.: þ33 2 99 28 42 69; fax: þ33 2 99 28 41 13. E-mail address: [email protected] (T. Fardoun).
remains underused and mostly performed in tertiary care centers.9,10 A possible explanation for this lack of diffusion is the potential morbidity of PN.5,11e13 In many studies, complication rates of PN lie between 10 and 50%, according to tumor complexity.14e16 The most frequent complications of PN are hemorrhagic17 (ie: blood loss, perirenal hematoma, arterio-venous fistula and false aneurysm). These hemorrhagic complications (HC) can lead to a need for transfusion, embolization and/or reoperation that can raise morbidity, increase length of hospital stay and delay return to normal activities. Our objective was to evaluate predictive factors of HC after PN in a contemporary cohort of patients operated both open and laparoscopically.
0748-7983/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejso.2013.11.006
86
T. Fardoun et al. / EJSO 40 (2014) 85e89
Materials and methods
Results
This was a retrospective study of 199 patients who underwent PN at our institution for a renal tumor between January 2008 and March 2012. After informed consent, all their records were prospectively entered into a specific RCC database. All cases were done consecutively. Patients were mostly operated by two surgeons. Operative technique (open, laparoscopic with or without robotic assistance) was decided by the surgeon based on his experience and tumor complexity. Data regarding patient demographics, tumor characteristics, and surgical outcomes were reviewed. Comorbidities were graded using Charlson index. The presence of anticoagulant therapy (either vitamin K antagonists or antiplatelet agent) was noted. Tumor complexity was evaluated with the RENAL nephrometry score (RS)18 and tumors were classified as: low complexity (RENAL score between 4 and 6), moderate complexity (RENAL score between 7 and 9) and high complexity (RENAL score > 9 or hilar tumors). Patients were operated either open or laparoscopically with or without robotic assistance. The open access was performed extraperitoneally through a flank incision. The tumor was dissected free from the perirenal fat. Most of the time, tumor resection was performed under pedicle clamping. The resection margin was assessed macroscopically as we do not routinely ask for frozen examination. A 2.0 running suture on the resection bed was made to ensure hemostasis and closure of the collecting system. Renal reconstruction was achieved with interrupted 0 sutures. All laparoscopic interventions were performed transperitoneally. The principles of tumor resection were the same as with the open access. For all operation, hemostatic agents were used at the surgeon’s discretion if needed. We studied the use of the following hemostatic agents: gelatin matrix with human thrombin (FlosealÒ), cellulose surgical bolster (SurgicelÒ) and biological surgical glue (BioglueÒ). All drugs could have been used separately or combined. Surgeon’s experience was appraised considering that an experienced surgeon had performed more than 20 PN with a specific surgical approach (for example, even if a surgeon had performed more than 20 laparoscopic PN, we considered that he was still unexperienced in his first 20 robotic cases). Complications that occurred within one month after surgery were classified using the ClavieneDindo system. We distinguished minor (grade I and II) from major complications (grade IIIeV). HC were defined as the occurrence of a perirenal hematoma requiring transfusion, an arterio-venous fistula, a false aneurysm or a decrease of hemoglobin >3 g/dl.
Patients data
Statistical analysis We compared patients with or without HC using the Wilcoxon test for continuous variables and the Fisher exact test for categorical variables. We performed a univariate and multivariate analysis with a logistic regression model using the occurrence of a HC as the dependent variable.
Demographics are shown in Table 1. 108 patients were male. Median age was 61 (22e86) years old. Median BMI was 26 (18e47) kg/m2. Median Charlson comorbidity index was 4 (0e11). Median baseline GFR was 86 (31e268) ml/min/1.73 m2 and 26 patients had chronic renal disease (defined as MDRD < 15 ml/min). 34 patients were receiving anticoagulants before surgery. Pathological data Tumor characteristics are summarized in Table 1. Median tumor size was 35 mm (10e110). 24 patients underwent imaging guided renal biopsy before surgery. On final pathologic exam, 165 tumors were malignant. According to RENAL score, 60, 83 and 46 tumors were classified as low, moderately and highly complex, respectively. Operative and perioperative data Details are summarized in Table 2. The indication was elective in 166 cases. 106 PN were done open, 54 by laparoscopy and 39 with robotic assistance. The surgeon was considered experimented in 142 procedures. Median Table 1 Patient and tumor characteristics. Variables Gender: male, n (%) Median age, years (range) Median BMI, kg/m2 (range) Median Charlson Index (range) Charlson index, n (%) 0e1 2 3 4 Chronic renal disease at surgery, n (%) Median baseline MDRD, ml/min (range) Anticoagulant treatment before surgery, n (%) Preoperative biopsy, n (%) Malignant tumors, n (%) Histology, n (%) Clear cell Papillary Chromophobe Other Median tumor size, mm (range) Fuhrman grade, n (%) 1 2 3 4 RENAL score, n (%) Low complexity Intermediate complexity High complexity
108 61 26 4
(54) (22e86) (18e47) (0e11)
50 39 39 71 26 86 34 24 165
(25) (20) (20) (35) (13) (31e268) (17) (12) (82.9)
122 23 15 5 35
(74) (14) (9) (3) (10e110)
14 95 51 3
(9) (58) (32) (1)
60 (32) 83 (44) 46 (24)
T. Fardoun et al. / EJSO 40 (2014) 85e89 Table 2 Peri-operative outcomes.
87
of stay was significantly increased (11 days compared to 7 days, p < 0.0001) in case of a HC.
Variables Imperative indication, n (%) Type of surgery, n (%) Open Laparoscopic Robotic Surgeon’s experience, n (%) Experimented Less than 20 procedures Hemostatic agent, n (%) None FlosealÒ BioglueÒ SurgicelÒ Type of clamping, n (%) None Parenchyma Pedicle Median clamping time, min (range) Median estimated blood loss, ml (range) Median operative time, min (range) Complications, n (%) Major Minor Hemorrhagic Complications, n (%) Hematoma False aneurysm Arterio-venous fistulae Loss of hemoglobin None Hemorrhagic Complications, n (%) Surgical complications Urinary fistula Wound infection - UTI Wound dehiscence Acute urinary retention Ileus Medical complications Pulmonary embolism Deep vein thrombosis Peptic ulcer Acute tubular necrosis Median length of stay, days (range)
33 (17) 106 (53) 54 (27) 39 (20) 142 (71) 57 (29) 4 113 13 160
(2) (57) (6) (80)
22 27 146 20 300 180
(11) (14) (75) (0e55) (20e2000) (70e360)
Regression analysis In univariate analysis (Table 4), imperative indication ( p ¼ 0.08), RENAL score ( p ¼ 0.07), operating time ( p ¼ 0.07) and operative blood loss ( p ¼ 0.002) were identified as statistically relevant to be included in the multivariable model. Neither obesity ( p ¼ 0.59), anticoagulant treatment ( p ¼ 0.45), Charlson comorbidity index ( p > 0.05), GFR status ( p ¼ 0.74), use of hemostatic agent ( p ¼ 0.353), surgical approach ( p > 0.05) or surgeon’s experience ( p ¼ 0.38) could foretell the occurrence of a HC. In multivariate analysis (Table 5), only blood loss > 250 ml was identified as a predictive factor of HC ( p ¼ 0.0007). Discussion
26 (13) 55 (27)
NSS is the recommended treatment of small renal masses.19,20 However, it remains underused: a recent study
15 (8) 11 (5) 3 (1) 14(7)
Table 3 Comparison of patients with and without hemorrhagic complications (HC).
6(3) 5 (2) e 13 (6) 2(1) 3(1) 5(2) 1 1 1 1 8
(0.5) (0.5) (0.5) (0.5) (3e45)
operative time was 180 (70e360) min with a median estimated blood loss of 300 (20e2000) ml. Clamping of the renal pedicle was done in 146 procedures and median ischemia time was 20 (0e55) min. Global complication rate was 40%. 81 complications (26 major and 55 minor) occurred. 43 were HC. There were 15 perirenal hematomas requiring transfusion, 11 false aneurysms, 3 arterio-venous fistulas and 14 patients who lost more than 3 points of hemoglobin after surgery. Comparison of patients with or without HC (Table 3) Patients were comparable for age, BMI, renal function and comorbidities. Median blood loss was higher in the HC group but the difference was not significant (400 vs. 300 ml, respectively, p ¼ 0.2). There were more complex tumors (75.6% vs. 66.5%, p ¼ 0.04) and median length
Group
No HC
Number of patients (%) Median age, years (range) Median BMI, kg/m2 (range) Median Charlson index (range) Chronic Renal Disease (%) Patients receiving anticoagulants (%) Median tumor size, mm (range) Experienced surgeon (%) Surgical approach Open Laparoscopy Robot Renal score category Low Intermediate High Hemostatic agents BioglueÒ FlosealÒ SurgicelÒ Median blood loss, ml (range) Median operating time, min (range) Median length of stay, days (range)
156 (78) 61 (22e82) 25.9 (18e47)
HC
P value
43 (22) 59 (32e86)
0.79
25(18e38.1)
0.76
3 (0e11)
0.41
21 (13.5)
5 (11.6)
1.00
25 (16)
9 (20.9)
0.49
35 (10e90)
0.51
109 (70)
33 (76)
0.44
81 (52) 45 (28) 30 (19)
25 (58) 9 (21) 9 (21)
0.62
50 (33.8) 58 (39.2) 40 (27)
10 (24.4) 25 (61) 6 (14.6)
0.04
4 (0e11)
33 (10e110)
10 87 124 300
(6.4) (56) (79) (20e1500)
3 26 36 400
(7) (60) (84) (40e2000)
1.00 0.6 0.66 0.20
180 (70e360)
180 (90e360)
0.20
7 (3e35)
11 (3e45)
250 ml)
0.077
2.1 (0.92e4.75)
0.314 0.949 0.379 0.353
1.0 0.65 0.97 1.42 1.35
0.691 0.721 0.800 0.07 0.002
1.26 (0.40e3.99) 1.29 (0.31e5.29) 0.99 (0.96e1.03) 1.00 (1.00e1.01) 1.51(1.16e2.00)
(0.37e2.55) (0.22e1.75) (0.27e1.53) (0.59e3.24) (0.30e2.37) (0.98e1.01)
(0.28e1.51) (0.41e2.32) (0.65e3.12) (0.71e2.57)
of almost 25 000 patients showed that even when patients were at risk of chronic renal disease, NSS was performed in less than 20% of the cases.21 Furthermore, NSS is mostly done in tertiary care centers.10 It has been suggested that this underutilization was related to the competition between NSS and laparoscopic radical nephrectomy, which have been both developed and promoted over the same time
Table 5 Results of multivariate analysis. Variables
P value
OR (95%CI)
Imperative indication
0.288
1.73 (0.63e4.81)
Renal score Low Intermediate High
0.247 0.090
1.0 1.76 (0.67e4.61) 0.29 (0.07e1.21)
Operative time
0. 505
1.00 (0.99e1.01)
Operative blood loss >250 ml
0.0007
1.64 (1.23e2.18)
period.9,22 In fact, the most probable explanation is that: 1) PN is technically demanding (particularly when performed laparoscopically) because it requires to remove the tumor an repair the kidney in a limited time and 2) its morbidity is superior to that of radical nephrectomy with HC that can occur in up to 30% of complicated cases.11,14 About 20% of our patients had a HC. In the literature, reported rates of bleeding events are highly variable ranging from 0 to 12%.12,15,23,24 However, all these studies are hardly comparable since the definition of bleeding varies greatly. Some authors don’t distinguish per-operative from post-operative bleeding.25 Some define it as a blood loss requiring at least one unit of blood transfusion.26 Others rely on a biological definition (acute decrease in hemoglobin).27 We chose to define a HC as a hematoma requiring transfusion, the occurrence of a pseudo-aneurysm or an arterio-veinous fistula and a decrease of hemoglobin >3 g/dl since these events are those that directly impact the post-operative course and patient recovery. Our rate of HC is higher than other studies. It can be explained by our definition of HC that was both clinical and biological. Most of the patients who have a significant loss of hemoglobin will not require transfusion. Our study showed that the main predictor of HC after PN was per-operative blood loss, which can be seen as an indirect sign of surgical difficulty. Every 250 ml, there was a 1.6-augmented risk of post-operative HC and when OBL exceed 1000 ml, there was a 12.5 increased risk of post-operative HC. Surprisingly, complexity of tumor was not identified as an independent factor of HC. This finding is in accordance with a recent study that found that there were no raise of incidence of complications when PN was performed for complex tumors.28 Interestingly, we did not find any impact of the surgical technique. Compared to open surgery, laparoscopic and robotic PN did not yield more post-operative HC. This is in contradiction with large multicentric retrospective studies that have suggested that even in expert centers, laparoscopic PN was associated with more complications, particularly HC.29 Other variables that could be felt to have an influence on bleeding were not significant in our statistical model. We did not find any impact of age, comorbidities, anticoagulant treatment, type of clamping, or ischemia time. Similarly, the importance of per-operative blood loss was not associated with post-operative hemorrhage. The surgeon’s experience (defined as more than 20 cases) was not associated to an increased frequency of HC. It might be explained by the fact that at our institution, the vast majority of cases are done or supervised by experienced surgeons (open or laparoscopically). There is some debate concerning the utility of hemostatic agents to ensure hemostasis during PN. It has been suggested that hemostatic gelatin matrix could decrease the
T. Fardoun et al. / EJSO 40 (2014) 85e89
incidence of vascular complications.30 Our analysis did not find any influence of these agents on the prevention of HC. However, we used such agents in a large majority of patients (98%) and it could just be that we don’t have enough patients operated without those agents to show a significant difference. Our study has some limitations. It is a retrospective monocentric study, which reflects the experience of mostly two surgeons. However, we believe the findings are of interest since it is a contemporary study with a good balance between open, laparoscopic and robotic cases that reflect the contemporary management of many tertiary centers. In conclusion, we found that the main predictor of HC after PN was operative blood loss. Several factors that are usually associated with bleeding such as anticoagulant treatment use of hemostatic agents or surgeon experience did not increase the risk of vascular complications. For this reason, patients who underwent a procedure with estimated blood loss >250 ml should be carefully monitored in the postoperative course. Conflicts of interest statement None. References 1. MacLennan S, Imamura M, Lapitan MC, et al. Systematic review of oncological outcomes following surgical management of localised renal cancer. Eur Urol 2012;61:972–93. 2. Ljungberg B, Cowan NC, Hanbury DC, et al. EAU guidelines on renal cell carcinoma: the 2010 update. Eur Urol 2010;58:398–406. 3. Gill IS, Aron M, Gervais DA, Jewett MA. Clinical practice. small renal mass. N Engl J Med 2010;362:624–34. 4. Tan H-J, Norton EC, Ye Z, Hafez KS, Gore JL, Miller DC. Long-term survival following partial vs radical nephrectomy among older patients with early-stage kidney cancer. JAMA 2012;307:1629–35. 5. Patard JJ, Pantuck AJ, Crepel M, et al. Morbidity and clinical outcome of nephron-sparing surgery in relation to tumor size and indication. Eur Urol 2007;52:148–54. 6. Becker F, Siemer S, Hack M, Humke U, Ziegler M, Stockle M. Excellent long-term cancer control with elective nephron-sparing surgery for selected renal cell carcinomas measuring more than 4 cm. Eur Urol 2006;49:1058–63. [discussion 1063e64]. 7. Huang WC. Impact of nephron sparing on kidney function and nononcologic mortality. Urol Oncol 2010;28:568–74. 8. Huang WC, Levey AS, Serio AM, et al. Chronic kidney disease after nephrectomy in patients with renal cortical tumors: a retrospective cohort study. Lancet Oncol 2006;7:735–40. 9. Abouassaly R, Alibhai SM, Tomlinson G, Timilshina N, Finelli A. Unintended consequences of laparoscopic surgery on partial nephrectomy for kidney cancer. J Urol 2010;183:467–72. 10. Hollenbeck BK, Taub DA, Miller DC, Dunn RL, Wei JT. National utilization trends of partial nephrectomy for renal cell carcinoma: a case of underutilization? Urology 2006;67:254–9.
89
11. Van Poppel H, Da Pozzo L, Albrecht W, et al. A prospective randomized EORTC intergroup phase 3 study comparing the complications of elective nephron-sparing surgery and radical nephrectomy for lowstage renal cell carcinoma. Eur Urol 2007;51:1606–15. 12. Pasticier G, Timsit MO, Badet L, et al. Nephron-sparing surgery for renal cell carcinoma: detailed analysis of complications over a 15year period. Eur Urol 2006;49:485–90. 13. Touijer K, Jacqmin D, Kavoussi LR, et al. The expanding role of partial nephrectomy: a critical analysis of indications, results, and complications. Eur Urol 2010;57:214–22. 14. Simhan J, Smaldone MC, Tsai KJ, et al. Objective measures of renal mass anatomic complexity predict rates of major complications following partial nephrectomy. Eur Urol 2011;60:724–30. 15. Ficarra V, Bhayani S, Porter J, et al. Predictors of warm ischemia time and perioperative complications in a multicenter, international series of robot-assisted partial nephrectomy. Eur Urol 2012;61:395–402. 16. Mottrie A, Schatteman P, De Wil P, De Troyer B, Novara G, Ficarra V. Validation of the preoperative aspects and dimensions used for an anatomical (PADUA) score in a robot-assisted partial nephrectomy series. World J Urol 2013;31:799–804. 17. Huber J, Pahernik S, Hallscheidt P, et al. Risk factors and clinical management of haemorrhage after open nephron-sparing surgery. BJU Int 2010;106:1488–93. 18. Kutikov A, Uzzo RG. The R.E.N.A.L. nephrometry score: a comprehensive standardized system for quantitating renal tumor size, location and depth. J Urol 2009;182:844–53. 19. Zini L, Patard JJ, Capitanio U, et al. The use of partial nephrectomy in European tertiary care centers. Eur J Surg Oncol 2009;35:636–62. 20. Miller DC, Hollingsworth JM, Hafez KS, Daignault S, Hollenbeck BK. Partial nephrectomy for small renal masses: an emerging quality of care concern? J Urol 2006;175:853–7. [discussion 858]. 21. Abouassaly R, Finelli A, Tomlinson GA, Urbach DR, Alibhai SM. How often are patients with diabetes or hypertension being treated with partial nephrectomy for renal cell carcinoma? A populationbased analysis. BJU Int 2011;108:1806–12. 22. Bhayani SB, Clayman RV, Sundaram CP, et al. Surgical treatment of renal neoplasia: evolving toward a laparoscopic standard of care. Urology 2003;62:821–6. 23. Heye S, Maleux G, Van Poppel H, Oyen R, Wilms G. Hemorrhagic complications after nephron-sparing surgery: angiographic diagnosis and management by transcatheter embolization. Am J Roentgenol 2005;184:1661–4. 24. Gratzke C, Seitz M, Bayrle F, et al. Quality of life and perioperative outcomes after retroperitoneoscopic radical nephrectomy (RN), open RN and nephron-sparing surgery in patients with renal cell carcinoma. BJU Int 2009;104:470–5. 25. Zimmermann R, Janetschek G. Complications of laparoscopic partial nephrectomy. World J Urol 2008;26:531–7. 26. Richstone L, Montag S, Ost MC, et al. Predictors of hemorrhage after laparoscopic partial nephrectomy. Urology 2011;77:88–91. 27. Wheat JC, Roberts WW, Hollenbeck BK, Wolf Jr JS, Weizer AZ. Complications of laparoscopic partial nephrectomy. Urol Oncol 2013;31:57–62. 28. Dulabon LM, Kaouk JH, Haber GP, et al. Multi-institutional analysis of robotic partial nephrectomy for hilar versus nonhilar lesions in 446 consecutive cases. Eur Urol 2011;59:325–30. 29. Gill IS, Kavoussi LR, Lane BR, et al. Comparison of 1,800 laparoscopic and open partial nephrectomies for single renal tumors. J Urol 2007;178:41–6. 30. Gill IS, Ramani AP, Spaliviero M, et al. Improved hemostasis during laparoscopic partial nephrectomy using gelatin matrix thrombin sealant. Urology 2005;65:463–6.
