Calvin Harris And Disciples | Introduction of EarReaders. | One Piece - E 845

Perioperative morbidity of laparoscopic radical prostatectomy compared with open radical retropubic prostatectomy

Perioperative morbidity of laparoscopic radical prostatectomy compared with open radical retropubic prostatectomy

Urologic Oncology: Seminars and Original Investigations 22 (2004) 102–106 Original article Perioperative morbidity of laparoscopic radical prostatec...

51KB Sizes 0 Downloads 18 Views

Urologic Oncology: Seminars and Original Investigations 22 (2004) 102–106

Original article

Perioperative morbidity of laparoscopic radical prostatectomy compared with open radical retropubic prostatectomy James A. Brown*,1, Christopher Garlitz, Leonard G. Gomella, David E. McGinnis, Stuart M. Diamond, Stephen E. Strup Department of Urology, Thomas Jefferson University, Philadelphia, PA 19107, USA Received 21 March 2003; accepted 21 May 2003

Abstract The objective of the study was to compare the perioperative complication rates of our initial 60 laparoscopic radical prostatectomy (LRP) patients and our most recent 60 sequential open radical retropubic prostatectomy (RRP) patients. Sixty sequential LRP and 60 sequential RRP patients treated between March 2000 and March 2002 were retrospectively evaluated. Patients who received neo-adjuvant hormonal therapy or had metastatic disease and 3 LRP patients converted to open RRP were excluded. Estimated blood loss (EBL), transfusion rates, hemoglobin level, serum and drain fluid creatinine levels, hospital stay and complication rates were analyzed. There were 15 (25%) and 11 (18.3%) complications in the LRP and RRP cohorts, respectively. There were 3 (ulnar neuropathy, ureteral stricture, anastomotic leak with ureteral obstruction requiring reoperation), and 4 [2 bladder neck contractures (BNC) and 2 deep venous thromboses (DVT)] major complications, respectively. Minor complications included rectus hematoma, superficial wound infections, ileus and anastomotic urine leaks. A higher incidence of the latter (10 patients) was noted in the LRP cohort. One (1.7%) LRP and 31 (52%) RRP cohort patients received intraoperative or postoperative transfusions. The mean (median) EBL was 317 (250) and 1355 (1100) for the LRP and RRP cohorts, respectively. A transient, insignificant increase in serum creatinine from a median of 1.0 to 1.2 mg/dL was observed only in the LRP cohort. We concluded that our initial 60 LRP patients had a similar, but not improved, rate of perioperative complications when compared with 60 sequential open RRP patients of nearly identical age, preoperative PSA and prostate size. The types of complications differed between the LRP and RRP cohorts. © 2004 Elsevier Inc. All rights reserved. Keywords: Radical prostatectomy; Laparoscopic prostatectomy; Morbidity; Complications; Perioperative

1. Introduction For laparoscopic radical prostatectomy (LRP) to be a reasonable alternative to open radical prostatectomy, oncologic equivalence and a decrease in morbidity must be demonstrated to justify its expense and steep learning curve [1]. Guillonneau et al. have reported positive margin rates of approximately 15% in large European series [2,3]. A recent study demonstrated 11% and 32% positive margin rates in 37 LRP and 37 radical retropubic prostatectomy (RRP) patients, respectively [4]. Thus, initial series have demon-

strated that LRP appears to be equivalent to RRP in terms of positive margin rates. European centers have also reported LRP complication rates of 12 to 25%, but these series were not compared with open RRP series [3,5,6]. The goal of this study, therefore, is to compare the perioperative complication rates of our initial 60 LRP patients with our most recent 60 RRP cohort.

2. Methods 2.1. Cohort selection

* Corresponding author. Tel.: ⫹1-706-721-9977; Fax: ⫹1-706-7212548. E-mail address: [email protected] (J.A. Brown). 1 Current address: Section of Urology, Medical College of Georgia, Augusta, GA. 1078-1439/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/S1078-1439(03)00101-7

We performed 68 LRPs at our institution between March 2000 and March 2002. Three patients required open conversion: one for ureteral injury, one for obesity and technical difficulty and one for hypercapnia. These patients as well as 5 patients receiving neo-adjuvant hormonal therapy

J.A. Brown et al. / Urologic Oncology: Seminars and Original Investigations 22 (2004) 102–106

were excluded, leaving 60 patients in our sequential LRP cohort. Among our most recent 72 RRP patients (performed between June 2001 and March 2002), 2 with lymph node metastases and 10 receiving neo-adjuvant hormonal therapy were excluded, leaving 60 for a comparative analysis. Clinical and pathologic data including patient age, preoperative PSA, stage and grade, number of anastomotic sutures, operative time (in the LRP cohort), estimated blood loss (EBL), serum hemoglobin level (preop, POD 1 and discharge), serum creatinine level (preoperative, evening postop, POD 1, discharge and when present at follow-up), drain fluid creatinine level (when obtained), length of hospital stay, prostate weight, margin status, and complications were recorded. Major complications were defined as those that were potentially life threatening, required reoperation, or resulted in permanent injury or deformity. Minor complications were defined as those, which were self-limited, nonlife threatening and required no additional procedures under anesthesia. 2.2. Operative technique The open RRP cases were performed by or under the direction of 2 staff surgeons (LGG and SMD). Open RRP was performed in the standard fashion with simultaneous performance of a modified bilateral pelvic lymph node dissection (BPLND). The deep dorsal vein complex was bunched and ligated both proximally and distally before division. Unilateral or bilateral nerve sparing was performed when indicated before division of the urethra at the prostate apex. The prostate was dissected off the rectum in a posterior peel fashion with anterior division of the bladder neck. The bladder neck was reconstructed with mucosal eversion before performing a 6-point anastomosis with interrupted 3-0 Monocryl sutures over a 20 French Foley catheter. The LRP cases were performed by or under the direction of 2 staff surgeons (SES and DEM). LRP was performed as described by Guillonneau and Vallancien [2]. A simultaneous BPLND was performed in 11 patients. In brief, the patient was positioned on a beanbag with the arms tucked within arm sleds and with padded shoulder support braces. Twelve-millimeter infraumbilical and right lateral rectus ports were placed, as were 5 mm ports at the left lateral rectus border and bilaterally in the lower quadrants. With the patients placed in the steep Trendelenburg position, initial seminal vesicle and vas deferens dissection was performed followed by development of the space of Retzius, dorsal vein complex ligation and prostatectomy with initial anterior dissection off of the bladder neck. Hemostasis was maintained with the harmonic scalpel and bipolar cautery, and the urethral anastomosis was typically performed with 8 interrupted 3-0 Vicryl sutures over a 20 French Foley catheter with bladder reconstruction or narrowing as indicated. After testing for a watertight anastomosis, urachal recon-

103

struction and BPLND were performed in selected patients before specimen removal and incision closure. 2.3. Statistical analysis The GraphPad InStat software program was used to determine mean, median, standard deviation, range and probability values when indicated.

3. Results The patient age, preoperative serum PSA, clinical stage, transrectal needle biopsy grade, prostate weight, pathologic stage and grade, positive margin rate, OR time, estimated blood loss (EBL), transfusion rate and change in serum hemoglobin, and creatinine level for each cohort are listed in Table 1. This data demonstrates that the patients in each cohort were similar with very similar prostate malignancies. The skin-skin operative time was recorded for the LRP cohort but not the open RRP cohort. Open RRP typically takes between 3 and 4 h with an average skin-skin operative time of approximately 210 min at our center. The mean (median) LRP operative time was 5.8 (5.5) h or 348 (330) min. The average operative time for the first 10, 20, and 30 LRP patients was 7.6, 6.7, and 6.4 h, respectively. The average operative time for the last 30 LRP patients was 5.1 h. Ten (16.9%) of 59 evaluable LRP patients had positive surgical margins while 12 (20%) of 60 RRP cohort patients had positive margins. The mean (median) EBL for the LRP and RRP cohorts was 317 (250) and 1355 (1100) mL, respectively. Only 1 LRP patient received one unit of blood (1.7%) intraoperatively. Conversely, 31 (51.7%) of the RRP patients received transfusions, approximately two-thirds being autologous and one-third donated packed red blood cells. Excluding patients who received transfusions, serum hemoglobin assessment was performed for each cohort preoperatively, on POD 1 and at dismissal. The mean (median) preoperative hemoglobin in nontransfused patients was 14.2 (14.2) and 13.5 (13.5) g/dL in the LRP and RRP cohorts, respectively. The hemoglobin dropped by a mean (median) of 3.1 (3.0) and 3.7 (3.6) g/dL to a level of 11.1 (11.2) and 9.8 (9.9) in each cohort, respectively. Serum creatinine levels were also recorded preoperatively and postoperatively. The mean (median) preoperative serum creatinine was 1.0 (1.0), and 0.99 (1.0) mg/dL in each cohort, respectively. The mean (median) serum creatinine rose to a value of 1.13 (1.2) mg/dL the evening of surgery in the LRP cohort while it remained unchanged at 1.0 (1.0) mg/dL in the open RRP cohorts. The mean (median) serum creatinine in the LRP cohort decreased to 1.09 (1.1) mg/dL the morning of POD1 and it returned to its baseline value of 1.0 (1.0) mg/dL by discharge. Fifteen (25%) LRP and 11 (18.3%) RRP cohort patients

104

J.A. Brown et al. / Urologic Oncology: Seminars and Original Investigations 22 (2004) 102–106

Table 1 LRP versus RRP patient cohort profiles LRP cohort No. of patients Mean Age (years) (median) Mean PSA (ng/ml) (median) Clinical Stage T1a/b T1c T2a T2b Biopsy gleason grade 3⫹2 3⫹3 3⫹4 4⫹3 4⫹4 Pathologic stage T2a T2b T3a T3b T4 Gleason grade-final 3⫹2 3⫹3 3⫹4 3⫹5 4⫹3 4⫹4 4⫹5 5⫹3 Margin (⫹) Mean OR time (min) (median) Mean specimen weight (gm) (median) Mean EBL (mL) (median) Transfusion rate Mean Hgb (g/dL) (median)* Preoperative Postoperative Decrease Mean Serum Creatinine (mg/dl) (median) Preoperative Postoperative POD#1 At Discharge

RRP cohort

60 58.8 (58.5) 6.4 (6.0)

60 59.0 (59.0) 5.6 (5.1)

0 47 13 0

1 45 11 3

1 46 11 2 0

2 39 16 2 1

14 34 8 2 1

13 39 4 3 1

1 37 15 0 4 1 1 0 16.90% 348 (330) 46.5 (41.5) 317 (250) 1.70%

1 34 15 0 6 2 1 1 20% not recorded 46.9 (41.5) 1355 (1100) 52%

14.2 (14.2) 11.1 (11.2) 3.1 (3.0)

13.5 (13.5) 9.8 (9.9) 3.7 (3.6)

1.0 (1.0) 1.13 (1.2) 1.09 (1.1) 1.0 (1.0)

0.99 (1.0) 1.0 (1.0) 1.0 (1.0) 1.0 (1.0)

plus an additional 2 anastomotic leaks requiring 3 to 4 weeks of JP drainage and Foley catheterization (JP creatinine remained elevated on POD12 and 16). Ten additional LRP patients were noted to have an elevated JP creatinine without increased drain output on POD 1, which returned to normal by POD 2 to POD 4, and these patients were not listed as complications. JP creatinine levels were routinely obtained postoperatively in the LRP cohort to assist in determining the timing of JP drain and Foley catheter removal. Three patients with inadvertent cystotomies, which were closed uneventfully intraoperatively, were also not listed as complications. No LRP patient has developed a bladder neck contracture (BNC) to date. The RRP cohort had 4 patients with complications classified as major. Two patients developed deep venous thromboses (DVT), one of which required readmission. Two other patients developed BNCs, which required additional surgical management. The minor complications in this cohort included 3 patients with ileus, 2 with a superficial wound infection, and 2 patients who developed clinically significant anastomotic leaks, one of which required readmission for increasing output. The mean (median) hospital stay for the LRP and RRP cohorts was 2.8 (2.0) and 3.0 (3.0) days, respectively. All RRP patients were hospitalized between 2 and 5 days. The LRP cohort had a wider range of hospitalization with 6 patients dismissed on POD 1, a 15-day hospitalization after reoperation for a patient with ureteral obstruction and marked urinary extravasation and an 8-day hospitalization for a patient with an anastomotic leak and ileus.

4. Comments Laparoscopic radical prostatectomy burst onto the urologic scene 3 yrs ago when the Montsouris group reported their early promising results [2,5,7]. Table 3 presents the perioperative and postoperative complication rates observed in several LRP series reported to date [2,3,5–9]. The mean operative time for various series has ranged from 210 to 300

* Includes only patients not transfused.

had complications (Table 2). Three and 4 in each cohort were classified as major. In the LRP cohort, the major complications included Patient 5 who suffered right ureteral obstruction and left ureteral orifice incorporation into the anastomosis with a high volume urine leak which required reoperation (cystoscopy with anastamotic stitch division and bilateral ureteral stent placement) and a 15-day hospitalization. Patient 2 developed an ulnar neuropathy. A third patient developed a right ureteral stricture postoperatively. Minor complications in the LRP cohort included a rectus hematoma, 2 cases of ileus (1 required readmission), 7 clinically significant urine leaks with elevated JP drain creatinine levels that did not return to normal by POD 4 but spontaneously resolved within 1 to 2 weeks postoperatively

Table 2 Perioperative morbidity comparison LRP versus RRP cohorts

Complications Major Ureteral injury Ulnar neuropathy Deep venous thrombosis Bladder neck contracture Rectal injury Minor Prolonged Ileus Rectus hematoma Anastomotic urine leak Superficial wound infection

LRP cohort

RRP cohort

15 (25%) 3 (5%) 2 1 0 0 0 12 (20%) 2 1 9 0

11 (18.3%) 4 (6.7%) 0 0 2 2 0 7 (12.6%) 3 0 2 2

J.A. Brown et al. / Urologic Oncology: Seminars and Original Investigations 22 (2004) 102–106

105

Table 3 Perioperative morbidity comparison of LRP series

No. pts. Complications Major complications Hemoperitoneum Epigastric artery injury Obturator nerve injury Compressive neurapraxia Anastomotic urine leak Ureteral injury Ureteral stricture Bladder neck contracture Obstructive anuria Rectal injury Rectal fistula Ileal injury Sigmoid injury Prolonged ileus Wound dehiscence Minor complications Prolonged ileus Urinary tract infection Anastomotic urine leak Acute retention Rectus hematoma Pelvic hematoma Trocar hernia Rectal injury Bladder injury Ureteral injury Compressive nuerapraxia Axonal degeneration Lymphorrhea Thrombosis Epigastric artery injury

Present series

Guillonneau et al.5

Guillonneau and Vallancien2

Rassweiler et al.3

Rassweiler et al.3

60 15 (25%) 3 (5%) 0 0 0 1 (1.6%) 0 1 (1.6%) 1 (1.6%) 0 0 0 0 0 0 0 0 12 (20%) 2 (3.3%) 0 9 (15%) 0 1 (1.6%) 0 0 0 0 0 0 0 0 0 0

567 105 (17.1%) 21 (3.7%) 5 (0.88%) 1 (0.17%) N/R 0 1 (0.17%) 2 (0.35%) N/R N/R 1 (0.17%) 2 (0.35%) N/R 2 (0.35%) 1 (0.17%) 1 (0.17%) 4 (0.7%) 83 (14.6%) 5 (0.9%) N/R 57 (10%) N/R N/R N/R N/R 6 (1%) 9 (1.6%) 1 (0.17%) 2 (0.35%) 1 (0.17%) 1 (0.17%) 2 (0.35%) 2 (0.35%)

120 30 (25%) 3 (2.5%) N/R 1 (0.83%) 1 (0.83%) N/R 1 (0.83%) N/R N/R N/R N/R 1 (0.83%) N/R N/R N/R N/R N/R 27 (22.5%) 2 (1.6%) 14 (10%) 8 (6.6%) 3 (2.5%) N/R N/R N/R N/R N/R N/R N/R N/R N/R N/R N/R

60 14 (23.3%) N/R N/R N/R N/R N/R N/R N/R N/R 4 (6.6%) N/R N/R 0 N/R N/R N/R N/R N/R 2 (3.3%) N/R 1 (1.6%) N/R N/R 7 (11.6%) 0 N/R N/R N/R N/R N/R N/R N/R N/R

180 34 (18.9%) N/R N/R N/R N/R N/R N/R N/R N/R 6 (3.3%) N/R N/R 2 (1.1%) N/R N/R N/R N/R N/R 5 (2.7%) N/R 4 (6.6%) N/R N/R 16 (8.8%) 1 (0.5%) N/R N/R N/R N/R N/R N/R N/R N/R

min. Excluding the Rassweiler et al.’s series with a 31% transfusion rate, LRP series have reported 2% to 6% transfusion rates with EBLs of a few hundred milliliters. The open conversion rates and PSA recurrence rates are ⬍2% and 5%, respectively. The mean Foley catheter times have been 5.5 to 7 days. Complication rates have been reported to occur in 11.7% to 25% of patients, with a major complication rate of approximately 4%. A rectal injury rate of 2% and BNC rates of 3.3% and 8% have been reported. An anastomotic leakage rate, determined by cystogram at Foley removal, of 17% was reported in one series [3]. This study is unique in that we compare our initial series of patients undergoing LRP to a cohort of patients undergoing open RRP at our institution during an overlapping time period. In an effort to minimize bias, we excluded patients with metastatic disease or status post neo-adjuvant hormonal therapy from the evaluation. Our mean 317 mL EBL and 1.7% transfusion rate in the LRP cohort is on par with previously reported LRP series. Our 50% intraoperative or postoperative transfusion rate is greater than the majority of reported RRP series in the literature. We believe

that this is largely because of our routine practice of having patients donate autologous blood and a relative willingness to transfuse these units back to the patients during or after surgery. Approximately two-thirds of the transfused patients received only autologous blood. Our 5.5-h median LRP operative time and 25% complication rate is similar to other initial series [3]. Our anastomotic leak rate of 33% (20 patients) is greater than reported previously. Rassweiler et al. reported a 17% leak rate 1 week postoperatively at the time of Foley removal. Abbou et al. reported that 4 of 43 (9.3%) initial patients required open surgical repair of anastomotic leakages [8], and they subsequently switched from an interrupted to a running hemicircumferencial anastomotic suture associated with no bladder neck contracture at 6 month follow-up [10]. Unlike European series where patients remain hospitalized until the catheter is removed, we evaluated drain fluid creatinine POD1 in an effort to assess anastomotic integrity to determine which LRP patients would have their JP drain removed before dismissal. Therefore, our greater anastomotic leak rate is likely, at least in part, the result of our routine

106

J.A. Brown et al. / Urologic Oncology: Seminars and Original Investigations 22 (2004) 102–106

analysis of JP drain fluid for creatinine level on POD 1 (other series have not reported doing this). It may also be related to other factors including: a more capacious bladder neck opening anastomosed using interrupted sutures (with increased distance between sutures), initial routine placement of a urachal reanastomosis suture (with possible tension on and separation of the anastomosis) and minimization of IV fluids to approximately 2500 mL during the case to minimize urine output and facilitate anastomosis reconstruction. We believe the latter may have contributed to clot retention and anastomotic rupture in a few cases. We, therefore, now routinely space our anastomotic sutures around the posterior and lateral bladder neck (when capacious) with subsequent figure-of-eight reapproximation and narrowing of the anterior bladder neck in a tennis racket configuration. While we initially believed urachal reconstruction might reduce the incidence of severe postoperative urge incontinence (noted in one case), we no longer routinely reconstruct the urachus. We are also now carefully monitoring patients for postoperative hematuria and clot retention and liberally providing IV fluids to maintain good urine output in the postoperative period. Finally, while a running anastomotic suture as described by Hoznek et al would likely reduce the incidence of anastomotic leakage, as we currently have a 0% BNC rate, we have not adopted this modification [10]. 5. Conclusions In conclusion, comparison of our initial 60 LRP cases with our most recent 60 open RRP cohort has demonstrated that LRP has lower blood loss and transfusion rates, is

associated with an insignificant transient serum creatinine increase, has a greater anastomotic leak rate (not associated with BNC formation) but an otherwise similar rate of varying complications. Additional studies to further evaluate and compare the perioperative and postoperative morbidity of laparoscopic and open radical prostatectomy are indicated.

References [1] Cadeddu JA, Kavoussi LR. Laparoscopic radical prostatectomy: is it feasible and reasonable? Urol Clin North Am 2001;28:655– 61. [2] Guillonneau B, Vallancien G. Laparoscopic radical prostatectomy: the Montsouris Experience. J Urol 2000;163:418 –22. [3] Rassweiler J, Sentker L, Seemann O, Hatzinger M, Rumpelt HJ. Laparoscopic radical prostatectomy with the heilbronn technique: an analysis of the first 180 cases. J Urol 2001;166:2101– 8. [4] Mitka M. Laparoscopic prostate surgery suggested. JAMA 2001;286: 2224 –5. [5] Guillonneau B, Rozet F, Cathelineau X, et al. Perioperative complications of laparoscopic radical prostatectomy: the Montsouris 3-year experience. J Urol 2002;167:51– 6. [6] Turk I, Deger IS, Winkelmann B, Roigas J, Schonberger B, Loening SA. Laparoscopic radical prostatectomy. Experiences with 145 interventions. Urologe A 2001;40:199 –206. [7] Guillonneau B, Cathelineau X, Doublet JD, Vallancien G. Laparoscopic radical prostatectomy: the lessons learned. J Endourol 2001; 15:441–5; discussion 447– 8. [8] Abbou CC, Salomon L, Hoznek A, et al. Laparoscopic radical prostatectomy: preliminary results. Urology 2000;55:630 – 4. [9] Olsson LE, Salomon L, Nadu A, et al. Prospective patient-reported continence after laparoscopic radical prostatectomy. Urology 2001; 58:570 –2. [10] Hoznek A, Salomon L, Rabii R, et al. Vesical urethral anastomosis during laparoscopic radical prostatectomy: the running suture method. J Endourol 2000;14:749 –53.