Original Paper
Urologia Internationalis
Received: January 27, 2014 Accepted after revision: April 28, 2014 Published online: August 7, 2014
Urol Int 2015;94:83–87 DOI: 10.1159/000363606
Laparoscopic versus Open Partial Nephrectomy for Multilocular Cystic Renal Cell Carcinoma: A Direct Comparison Based on Single-Center Experience Ben Xu Jun-jie Wang Yue Mi Li-qun Zhou Jie Jin Qian Zhang Department of Urology, Peking University First Hospital and Institute of Urology, Peking University, National Urological Cancer Center, Beijing, China
Key Words Kidney neoplasm · Kidney tumor · Laparoscopy · Microsurgery · Multilocular cystic renal cell carcinoma · Open partial nephrectomy · Nephron-sparing surgery · Retroperitoneoscopy
Abstract Objectives: To compare the treatment of laparoscopic (LPN) versus open partial nephrectomy (OPN) in patients with multilocular cystic renal cell carcinoma (MCRCC). Methods: Thirty-seven patients diagnosed with MCRCC were reviewed retrospectively between January 2007 and January 2013 at our institution. They were divided into two groups: group 1 (LPN) consisted of 19 patients (51.4%) and group 2 (OPN) of 18 patients (48.6%). RENAL and the Preoperative Aspects and Dimensions Used for an Anatomical classification were applied to predict perioperative complications, which were graded based on the Clavien-Dindo classification. Results: The two groups were comparable with regard to all of the patients’ baseline characteristics. In group 1, the mean operative time was 142.1 min, including the mean warm ischemia time (WIT) of 32.6 min; the mean estimated blood loss (EBL) was 96.1 ml, the mean retroperitoneal drainage lasted 3.6 days, and the mean postoperative hospital stay was 5.3 days. In group 2, the figures were 126.6 and 23.5 min, 223.3
© 2014 S. Karger AG, Basel 0042–1138/14/0941–0083$39.50/0 E-Mail [email protected] www.karger.com/uin
ml, and 4.6 and 8.7 days, respectively. The differences in WIT, EBL, drainage days and hospitalization were statistically significant between both groups (p < 0.05). No recurrence or new lesions occurred in these patients during a mean followup of 37.8 months. Conclusions: Our single-center experience suggests that although it remains technically complex, demanding and challenging for MCRCC, LPN can still induce favorable perioperative results and survival rates in MCRCC are comparable with OPN. © 2014 S. Karger AG, Basel
Introduction
First recognized in 1982, multilocular cystic renal cell carcinoma (MCRCC) is an uncommon tumor of the kidney, accounting for less than 5% of RCCs [1]. Most of the patients are asymptomatic and the tumors are usually incidental findings, while a few patients can present with a palpable mass or abdominal discomfort [2]. Pathologically, MCRCC is a tumor of the kidney composed of multiple cysts with clear cells in the septa [2]. Low Fuhrman grade and stage are characteristic features of MCRCC [3,
Ben Xu and Jun-jie Wang contributed equally to this paper.
Qian Zhang Department of Urology, Peking University First Hospital and Institute of Urology Peking University, National Urological Cancer Center 8 Xishiku Street, Xicheng District, Beijing 100034 (China) E-Mail zhangqian @ bjmu.edu.cn
4]. Usually, there is no nodal or metastatic spread at the time of diagnosis and vascular invasion is not found [3– 5]. Due to its low epithelial tumor volume and low Fuhrman grade and stage in the majority of cases, MCRCC is considered a specific entity with a good prognosis according to the 2004 World Health Organization (WHO) classification of kidney tumors [6]. Many studies have shown no evidence of disease during follow-up and excellent 5-year disease-related survival rates of 100% [4, 7, 8]. As a matter of fact, quite rare cases of MCRCC have been reported when reviewing the literature (mostly case reports or small series). In addition, optimal management of MCRCC is still subject to controversy and thus needs clarification to avoid unnecessary overtreatment, such as radical nephrectomy (RN) or open partial nephrectomy (OPN) in some simple cases. Recently, growing experience with laparoscopic partial nephrectomy (LPN) for RCC has demonstrated its potential to duplicate the techniques and outcome of OPN [9]. In our opinion, LPN can similarly be generalized to MCRCC although handling the cystic lesions is a more challenging procedure than RCC because of the greater potential for inadvertent cyst puncture and tumor cell spread [10]. Unfortunately, there is still no series report of a direct comparison between LPN and OPN for MCRCC, and the efficiency and superiority of LPN in the treatment of MCRCC has not been analyzed. To the best of our knowledge, this is the first study comparing treatment outcome of LPN versus OPN in MCRCC patients, and, more importantly, our single-center experience in the laparoscopic technique is presented.
Materials and Methods Between January 2007 and January 2013, we retrospectively reviewed the database of 37 patients diagnosed with MCRCC based on surgical-pathological findings at our institution. Relevant clinical data were obtained from our prospectively maintained database after obtaining approval from our institutional review board. In all patients, written informed consent was obtained. Patients with bilateral lesions or previous renal surgery were excluded from the study. They were divided into two groups based on the detailed surgical approach applied. Group 1 (LPN) consisted of 19 patients (14 males and 5 females, 51.4%) and group 2 (OPN) of 18 patients (13 males and 5 females, 48.6%). Preoperative evaluations, including urine analysis, estimated glomerular filtration rate (eGFR), renal B-scan ultrasonography and computed tomography (CT), were routinely applied in all patients. At least two urologists evaluated the CT images independently and consultation with at least one radiologist was necessary to improve the accuracy of preop-
84
Urol Int 2015;94:83–87 DOI: 10.1159/000363606
erative diagnosis. With the advent of the RENAL nephrometry scoring system and the Preoperative Aspects and Dimensions Used for an Anatomical classification (PADUA), both reproducible standardized classification systems quantitated the anatomy of MCRCC and could help clinicians stratify patients suitable for LPN, OPN, even RN into subgroups with different risks and complications. Patients’ baseline demographics (age, body mass index, tumor size, RENAL/PADUA scores and preoperative eGFR), intraoperative variables, including operative time (OT), warm ischemia time (WIT) and estimated blood loss (EBL), and postoperative outcome (postoperative eGFR, drainage days, hospitalization days, tumor stage and grade, complications and follow-up information) were reported and analyzed. The complications were recorded according to the Clavien-Dindo grading system. Pathologically, all of the specimens were examined by at least two experienced pathologists. The microscopic features included important factors of tumor margin, the longest diameter, 2009 tumor node metastasis stage and Fuhrman grade. Clinical follow-up, including physical examination, eGFR, scintigraphy renal scan, abdominal CT and chest X-ray, was obtained from patient charts and referring physicians. All of the 37 patients received a retroperitoneal approach with minimal handling of the abdominal cavity. After induction of general anesthesia, the patients were placed in the left or right lateral decubitus position at 90°. In group 1, the laparoscopic procedure was performed with a 3-trocar technique only. Our surgical technique for MCRCC followed basically the standard steps applied to manage other renal tumors, as previously reported [11]. A 1-cm skin incision was initially made at the cross point of the line 2 cm above the iliac crest and the longitudinal line close to the anterior auxiliary line. A handmade balloon-dilating device was then placed to create a retroperitoneal space, and the first 13mm port was penetrated into the retroperitoneal space at a 30° angle to the perpendicular line. After maintaining retroperitoneal cavity pressure at 13–15 mm Hg, the laparoscope was inserted through the first port, and the two other ports were subsequently placed at cross points of the subcostal line, and anterior and posterior auxiliary lines. Afterwards, the fat was excised from Gerota’s fascia, which was subsequently opened and incised away from the tumor to facilitate excision and suturing. By cautious dissection, the renal vessels were dissected free, and the renal artery was completely clamped with laparoscopic bulldog clamps, which was of utmost importance in the treatment of MCRCC. By complete occlusion, the tension of the affected kidney decreased immediately and the surgeon could obtain a different feeling to avoid invading the cystic components incautiously when contacting the cystic tumor. Subsequently, the cystic lesion was sharply excised with cold scissors in an almost bloodless field. Afterwards, hemostasis was achieved by suturing any pelvicaliceal entry and using an absorbable oxidized regenerated cellulose bolster. Following renorrhaphy, the vascular clamp was removed, and hemostasis was evaluated. A suction drain was finally placed in the retroperitoneal cavity, and all port site skin edges were closed with skin adhesive. As for OPN in group 2, we had no special modification or innovation to the previous open retroperitoneal experience in renal surgery and the conventional standard surgical techniques were followed for mobilization, exposure, occlusion and resection. By statistical analysis, descriptive data were presented as numbers and percentages. Continuous data were shown as
Xu/Wang/Mi/Zhou/Jin/Zhang
means ± SD. Cancer-specific survival was determined from the date of surgery to the date of cancer-related death. The groups were compared for continuous variables using the independent t test, Pearson’s χ2 test or Fisher’s exact test. All of the data were analyzed using SPSS 17.0 (Statistical Package for Social Sciences, USA), and a value of p < 0.05 was considered statistically significant.
Results
Table 1. A direct comparison of patients’ baseline characteristics
in groups 1 and 2 Variable
Group 1 (n = 19)
Group 2 (n = 18)
Gender Male 14 (73.7%) 13 (72.2%) Female 5 (26.3%) 5 (27.8%) Age, years 49.3±9.0 46.2±10.3 Body mass index, kg/m2 25.0±2.7 26.0±3.0 Tumor size, cm 3.0±1.1 3.5±1.5 RNS scores 5.4±1.2 5.1±1.4 PADUA scores 7.4±1.3 7.1±1.3 Preoperative eGFR, 110.6±13.8 111.1±25.8 ml/min/1.73 m2
p value 0.92 0.55 0.59 0.32 0.64 0.73
The detailed patients’ baseline demographics are listed in table 1, with no significant difference between both groups. Intraoperative variables and postoperative outcome are listed in table 2, which shows that all operations were performed successfully without severe complications. Only 2 patients in group 2 experienced mild postoperative complications (1 urinary infection and 1 excessive drainage), yet they were treated successfully by infection control and postponing drainage tube withdrawal. In group 1, the mean OT was 142.1 min, including the mean WIT of 32.6 min; the mean EBL was 96.1 ml, the mean retroperitoneal drainage was 3.6 days, and the mean postoperative hospital stay was 5.3 days. eGFR levels were all within normal limits, with mean pre- and postoperative eGFR levels of 110.6 and 95.9 ml/ min/1.73 m2, respectively. There was no significant difference between pre- and postoperative eGFR levels. In group 2, the figures were 126.6 and 23.5 min, 223.3 ml, and 4.6 and 8.7 days. eGFR levels were also within normal limits, with mean pre- and postoperative eGFR levels of 111.1 and 103.1 ml/min/1.73 m2, respectively. There was also no significant difference between pre- and postoperative eGFR levels. In a direct comparison, the differences in WIT, EBL, drainage days and hospitalization were statistically significant between both groups (p < 0.05). No significance (p > 0.05) was observed in the other parameters, including baseline patient characteristics, OT, the decline in eGFR and perioperative complications. Pathologically, all of these 37 tumors were confirmed MCRCC with a negative margin. Among them, in group 1, 19 (100%) of the tumors presented with T1N0M0 stage, while a total of 17 (94.4%) presented with T1N0M0 stage and the remaining 1 (5.6%) presented with T2aN0M0 stage in group 2. In group 1, 18 patients (94.7%) were classified as Fuhrman grade 1, with only 1 case (5.3%) being Fuhrman grade 2. In group 2, 15 patients (83.3%) were classified as Fuhrman grade 1, with the other 3 cases (16.7%) being Fuhrman grade 2. Regarding tumor stage and grade, there was no difference between both groups. Additionally, there was no vascular
involvement, capsular penetration or lymph node metastasis. The patients with available follow-up information had no evidence of disease by abdominal CT, and they attained normal renal function by eGFR and renal scintigraphy at a mean follow-up of 37.8 months. The cancerspecific survival rates after LPN and OPN both reached 100% in the observation periods.
Laparoscopic versus Open Partial Nephrectomy
Urol Int 2015;94:83–87 DOI: 10.1159/000363606
0.94
RNS = R.E.N.A.L. nephometry scores.
Table 2. Direct comparison of intraoperative variables and post-
operative outcomes in groups 1 and 2 Variable
Group 1 (n = 19)
Group 2 (n = 18)
p value
OT, min 142.1±74.6 126.6±28.1 0.49 WIT, min 32.6±11.1 23.5±6.3 0.006 EBL, ml 96.1±108.2 223.3±152.7 0.01 Drainage, days 3.6±1.3 4.6±0.9 0.02 Hospitalization, days 5.3±2.1 8.7±2.8
Urologia Internationalis
Received: January 27, 2014 Accepted after revision: April 28, 2014 Published online: August 7, 2014
Urol Int 2015;94:83–87 DOI: 10.1159/000363606
Laparoscopic versus Open Partial Nephrectomy for Multilocular Cystic Renal Cell Carcinoma: A Direct Comparison Based on Single-Center Experience Ben Xu Jun-jie Wang Yue Mi Li-qun Zhou Jie Jin Qian Zhang Department of Urology, Peking University First Hospital and Institute of Urology, Peking University, National Urological Cancer Center, Beijing, China
Key Words Kidney neoplasm · Kidney tumor · Laparoscopy · Microsurgery · Multilocular cystic renal cell carcinoma · Open partial nephrectomy · Nephron-sparing surgery · Retroperitoneoscopy
Abstract Objectives: To compare the treatment of laparoscopic (LPN) versus open partial nephrectomy (OPN) in patients with multilocular cystic renal cell carcinoma (MCRCC). Methods: Thirty-seven patients diagnosed with MCRCC were reviewed retrospectively between January 2007 and January 2013 at our institution. They were divided into two groups: group 1 (LPN) consisted of 19 patients (51.4%) and group 2 (OPN) of 18 patients (48.6%). RENAL and the Preoperative Aspects and Dimensions Used for an Anatomical classification were applied to predict perioperative complications, which were graded based on the Clavien-Dindo classification. Results: The two groups were comparable with regard to all of the patients’ baseline characteristics. In group 1, the mean operative time was 142.1 min, including the mean warm ischemia time (WIT) of 32.6 min; the mean estimated blood loss (EBL) was 96.1 ml, the mean retroperitoneal drainage lasted 3.6 days, and the mean postoperative hospital stay was 5.3 days. In group 2, the figures were 126.6 and 23.5 min, 223.3
© 2014 S. Karger AG, Basel 0042–1138/14/0941–0083$39.50/0 E-Mail [email protected] www.karger.com/uin
ml, and 4.6 and 8.7 days, respectively. The differences in WIT, EBL, drainage days and hospitalization were statistically significant between both groups (p < 0.05). No recurrence or new lesions occurred in these patients during a mean followup of 37.8 months. Conclusions: Our single-center experience suggests that although it remains technically complex, demanding and challenging for MCRCC, LPN can still induce favorable perioperative results and survival rates in MCRCC are comparable with OPN. © 2014 S. Karger AG, Basel
Introduction
First recognized in 1982, multilocular cystic renal cell carcinoma (MCRCC) is an uncommon tumor of the kidney, accounting for less than 5% of RCCs [1]. Most of the patients are asymptomatic and the tumors are usually incidental findings, while a few patients can present with a palpable mass or abdominal discomfort [2]. Pathologically, MCRCC is a tumor of the kidney composed of multiple cysts with clear cells in the septa [2]. Low Fuhrman grade and stage are characteristic features of MCRCC [3,
Ben Xu and Jun-jie Wang contributed equally to this paper.
Qian Zhang Department of Urology, Peking University First Hospital and Institute of Urology Peking University, National Urological Cancer Center 8 Xishiku Street, Xicheng District, Beijing 100034 (China) E-Mail zhangqian @ bjmu.edu.cn
4]. Usually, there is no nodal or metastatic spread at the time of diagnosis and vascular invasion is not found [3– 5]. Due to its low epithelial tumor volume and low Fuhrman grade and stage in the majority of cases, MCRCC is considered a specific entity with a good prognosis according to the 2004 World Health Organization (WHO) classification of kidney tumors [6]. Many studies have shown no evidence of disease during follow-up and excellent 5-year disease-related survival rates of 100% [4, 7, 8]. As a matter of fact, quite rare cases of MCRCC have been reported when reviewing the literature (mostly case reports or small series). In addition, optimal management of MCRCC is still subject to controversy and thus needs clarification to avoid unnecessary overtreatment, such as radical nephrectomy (RN) or open partial nephrectomy (OPN) in some simple cases. Recently, growing experience with laparoscopic partial nephrectomy (LPN) for RCC has demonstrated its potential to duplicate the techniques and outcome of OPN [9]. In our opinion, LPN can similarly be generalized to MCRCC although handling the cystic lesions is a more challenging procedure than RCC because of the greater potential for inadvertent cyst puncture and tumor cell spread [10]. Unfortunately, there is still no series report of a direct comparison between LPN and OPN for MCRCC, and the efficiency and superiority of LPN in the treatment of MCRCC has not been analyzed. To the best of our knowledge, this is the first study comparing treatment outcome of LPN versus OPN in MCRCC patients, and, more importantly, our single-center experience in the laparoscopic technique is presented.
Materials and Methods Between January 2007 and January 2013, we retrospectively reviewed the database of 37 patients diagnosed with MCRCC based on surgical-pathological findings at our institution. Relevant clinical data were obtained from our prospectively maintained database after obtaining approval from our institutional review board. In all patients, written informed consent was obtained. Patients with bilateral lesions or previous renal surgery were excluded from the study. They were divided into two groups based on the detailed surgical approach applied. Group 1 (LPN) consisted of 19 patients (14 males and 5 females, 51.4%) and group 2 (OPN) of 18 patients (13 males and 5 females, 48.6%). Preoperative evaluations, including urine analysis, estimated glomerular filtration rate (eGFR), renal B-scan ultrasonography and computed tomography (CT), were routinely applied in all patients. At least two urologists evaluated the CT images independently and consultation with at least one radiologist was necessary to improve the accuracy of preop-
84
Urol Int 2015;94:83–87 DOI: 10.1159/000363606
erative diagnosis. With the advent of the RENAL nephrometry scoring system and the Preoperative Aspects and Dimensions Used for an Anatomical classification (PADUA), both reproducible standardized classification systems quantitated the anatomy of MCRCC and could help clinicians stratify patients suitable for LPN, OPN, even RN into subgroups with different risks and complications. Patients’ baseline demographics (age, body mass index, tumor size, RENAL/PADUA scores and preoperative eGFR), intraoperative variables, including operative time (OT), warm ischemia time (WIT) and estimated blood loss (EBL), and postoperative outcome (postoperative eGFR, drainage days, hospitalization days, tumor stage and grade, complications and follow-up information) were reported and analyzed. The complications were recorded according to the Clavien-Dindo grading system. Pathologically, all of the specimens were examined by at least two experienced pathologists. The microscopic features included important factors of tumor margin, the longest diameter, 2009 tumor node metastasis stage and Fuhrman grade. Clinical follow-up, including physical examination, eGFR, scintigraphy renal scan, abdominal CT and chest X-ray, was obtained from patient charts and referring physicians. All of the 37 patients received a retroperitoneal approach with minimal handling of the abdominal cavity. After induction of general anesthesia, the patients were placed in the left or right lateral decubitus position at 90°. In group 1, the laparoscopic procedure was performed with a 3-trocar technique only. Our surgical technique for MCRCC followed basically the standard steps applied to manage other renal tumors, as previously reported [11]. A 1-cm skin incision was initially made at the cross point of the line 2 cm above the iliac crest and the longitudinal line close to the anterior auxiliary line. A handmade balloon-dilating device was then placed to create a retroperitoneal space, and the first 13mm port was penetrated into the retroperitoneal space at a 30° angle to the perpendicular line. After maintaining retroperitoneal cavity pressure at 13–15 mm Hg, the laparoscope was inserted through the first port, and the two other ports were subsequently placed at cross points of the subcostal line, and anterior and posterior auxiliary lines. Afterwards, the fat was excised from Gerota’s fascia, which was subsequently opened and incised away from the tumor to facilitate excision and suturing. By cautious dissection, the renal vessels were dissected free, and the renal artery was completely clamped with laparoscopic bulldog clamps, which was of utmost importance in the treatment of MCRCC. By complete occlusion, the tension of the affected kidney decreased immediately and the surgeon could obtain a different feeling to avoid invading the cystic components incautiously when contacting the cystic tumor. Subsequently, the cystic lesion was sharply excised with cold scissors in an almost bloodless field. Afterwards, hemostasis was achieved by suturing any pelvicaliceal entry and using an absorbable oxidized regenerated cellulose bolster. Following renorrhaphy, the vascular clamp was removed, and hemostasis was evaluated. A suction drain was finally placed in the retroperitoneal cavity, and all port site skin edges were closed with skin adhesive. As for OPN in group 2, we had no special modification or innovation to the previous open retroperitoneal experience in renal surgery and the conventional standard surgical techniques were followed for mobilization, exposure, occlusion and resection. By statistical analysis, descriptive data were presented as numbers and percentages. Continuous data were shown as
Xu/Wang/Mi/Zhou/Jin/Zhang
means ± SD. Cancer-specific survival was determined from the date of surgery to the date of cancer-related death. The groups were compared for continuous variables using the independent t test, Pearson’s χ2 test or Fisher’s exact test. All of the data were analyzed using SPSS 17.0 (Statistical Package for Social Sciences, USA), and a value of p < 0.05 was considered statistically significant.
Results
Table 1. A direct comparison of patients’ baseline characteristics
in groups 1 and 2 Variable
Group 1 (n = 19)
Group 2 (n = 18)
Gender Male 14 (73.7%) 13 (72.2%) Female 5 (26.3%) 5 (27.8%) Age, years 49.3±9.0 46.2±10.3 Body mass index, kg/m2 25.0±2.7 26.0±3.0 Tumor size, cm 3.0±1.1 3.5±1.5 RNS scores 5.4±1.2 5.1±1.4 PADUA scores 7.4±1.3 7.1±1.3 Preoperative eGFR, 110.6±13.8 111.1±25.8 ml/min/1.73 m2
p value 0.92 0.55 0.59 0.32 0.64 0.73
The detailed patients’ baseline demographics are listed in table 1, with no significant difference between both groups. Intraoperative variables and postoperative outcome are listed in table 2, which shows that all operations were performed successfully without severe complications. Only 2 patients in group 2 experienced mild postoperative complications (1 urinary infection and 1 excessive drainage), yet they were treated successfully by infection control and postponing drainage tube withdrawal. In group 1, the mean OT was 142.1 min, including the mean WIT of 32.6 min; the mean EBL was 96.1 ml, the mean retroperitoneal drainage was 3.6 days, and the mean postoperative hospital stay was 5.3 days. eGFR levels were all within normal limits, with mean pre- and postoperative eGFR levels of 110.6 and 95.9 ml/ min/1.73 m2, respectively. There was no significant difference between pre- and postoperative eGFR levels. In group 2, the figures were 126.6 and 23.5 min, 223.3 ml, and 4.6 and 8.7 days. eGFR levels were also within normal limits, with mean pre- and postoperative eGFR levels of 111.1 and 103.1 ml/min/1.73 m2, respectively. There was also no significant difference between pre- and postoperative eGFR levels. In a direct comparison, the differences in WIT, EBL, drainage days and hospitalization were statistically significant between both groups (p < 0.05). No significance (p > 0.05) was observed in the other parameters, including baseline patient characteristics, OT, the decline in eGFR and perioperative complications. Pathologically, all of these 37 tumors were confirmed MCRCC with a negative margin. Among them, in group 1, 19 (100%) of the tumors presented with T1N0M0 stage, while a total of 17 (94.4%) presented with T1N0M0 stage and the remaining 1 (5.6%) presented with T2aN0M0 stage in group 2. In group 1, 18 patients (94.7%) were classified as Fuhrman grade 1, with only 1 case (5.3%) being Fuhrman grade 2. In group 2, 15 patients (83.3%) were classified as Fuhrman grade 1, with the other 3 cases (16.7%) being Fuhrman grade 2. Regarding tumor stage and grade, there was no difference between both groups. Additionally, there was no vascular
involvement, capsular penetration or lymph node metastasis. The patients with available follow-up information had no evidence of disease by abdominal CT, and they attained normal renal function by eGFR and renal scintigraphy at a mean follow-up of 37.8 months. The cancerspecific survival rates after LPN and OPN both reached 100% in the observation periods.
Laparoscopic versus Open Partial Nephrectomy
Urol Int 2015;94:83–87 DOI: 10.1159/000363606
0.94
RNS = R.E.N.A.L. nephometry scores.
Table 2. Direct comparison of intraoperative variables and post-
operative outcomes in groups 1 and 2 Variable
Group 1 (n = 19)
Group 2 (n = 18)
p value
OT, min 142.1±74.6 126.6±28.1 0.49 WIT, min 32.6±11.1 23.5±6.3 0.006 EBL, ml 96.1±108.2 223.3±152.7 0.01 Drainage, days 3.6±1.3 4.6±0.9 0.02 Hospitalization, days 5.3±2.1 8.7±2.8
