- Email: [email protected]
Radical oophorectomy with primary stapled colorectal anastomosis for resection of locally advanced epithelial ovarian cancer
Radical Oophorectomy with Primary Stapled Colorectal Anastomosis for Resection of Locally Advanced Epithelial Ovarian Cancer Robert E Bristow, MD, Marcela G del Carmen, MD, Howard S Kaufman, MD, FACS, Fredrick J Montz, MD, FACS The aim of this study was to describe the feasibility, associated morbidity, and efficacy of radical oophorectomy with primary stapled colorectal anastomosis among patients with locally advanced ovarian cancer with contiguous extension to or encasement of the reproductive organs, pelvic peritoneum, cul-de-sac, and sigmoid colon. STUDY DESIGN: Thirty-one consecutive patients undergoing radical oophorectomy as part of an initial maximal surgical effort for International Federation of Obstetrics and Gynecology (FIGO) stage IIIB–IV ovarian cancer were prospectively collected from October 1, 1997 through November 30, 2001. The surgical technique, associated morbidity, and clinical outcomes are described. RESULTS: The median age was 63 years. All patients had advanced-stage epithelial ovarian cancer: FIGO stage IIIB (6.5%), stage IIIC (64.5%), stage IV (29.0%). Median operating time was 240 minutes (range 165 to 330 minutes), and the median estimated blood loss was 700 mL (range 300 to 2,900 mL). All patients underwent en bloc rectosigmoid colectomy with primary stapled anastomosis without protective intestinal diversion. There was one (3.2%) anastomotic breakdown requiring reoperation and colostomy. Complete clearance of macroscopic pelvic disease was achieved in all cases. Overall, 87.1% of patients were left with optimal (ⱕ1 cm) residual disease and 61.3% were visibly disease free. There were no postoperative deaths, but major and minor postoperative morbidity occurred in 12.9% and 35.5% of patients, respectively. Blood product transfusion was required in 29.0% of cases. Thirty patients received multiagent platinum-based chemotherapy, with a median overall survival time of 39.5 months. CONCLUSIONS: Radical oophorectomy with primary stapled anastomosis is an effective technique for resection of locally advanced ovarian cancer and contributes significantly to a maximal cytoreductive surgical effort. The associated morbidity is acceptable, and protective intestinal diversion appears unnecessary. ( J Am Coll Surg 2003;197:565-574. © 2003 by the American College of Surgeons) BACKGROUND:
Ovarian cancer is the most lethal of all gynecologic malignancies. The American Cancer Society has estimated that 23,300 new cases will be diagnosed in the United
States in 2002, with 13,900 deaths directly attributable to this disease.1 For patients with International Federation of Gynecology and Obstetrics (FIGO) stage III/IV disease, survival determinants are multifactorial, but the strongest clinician-driven predictors of clinical outcomes are the amount of residual tumor after primary cytoreductive surgery and the administration of platinum-based chemotherapy.2-4 Locally advanced disease, with contiguous extension to or encasement of the reproductive organs, pelvic peritoneum, cul-de-sac, and sigmoid colon, can present a significant challenge to surgeons operating on women with ovarian cancer. Designed for intact removal of a fixed ovarian tumor en bloc with attached peritoneum and surrounding structures, the retroperitoneal technique of radical oophorec-
No competing interests declared.
This work supported by the Elizabeth Frost Ovarian Cancer Research Fund. Received December 26, 2002; Revised April 28, 2003; Accepted April 28, 2003. From The Kelly Gynecologic Oncology Service, Departments of Gynecology and Obstetrics and Oncology (Bristow, del Carmen, Montz), and the Department of Surgery (Kaufman, Montz), The Johns Hopkins Medical Institutions, Baltimore, MD. Dr Kaufman’s current affiliation is Division of Colorectal Surgery, Department of Surgery, University of Southern California Keck School of Medicine, Los Angeles, CA. Correspondence address: Robert E Bristow, MD, The Kelly Gynecologic Oncology Service, Departments of Gynecology and Obstetrics and Oncology, The Johns Hopkins Medical Institutions, 600 North Wolfe St, Phipps #289, Baltimore, MD 21205.
© 2003 by the American College of Surgeons Published by Elsevier Inc.
565
ISSN 1072-7515/03/$21.00 doi:10.1016/S1072-7515(03)00478-2
566
Bristow et al
Radical Oophorectomy for Ovarian Cancer
Abbreviations and Acronyms
CEEA ⫽ circular end-to-end anastomosis FIGO ⫽ International Federation of Gynecology and Obstetrics GOG ⫽ Gynecologic Oncology Group JHMI ⫽ Johns Hopkins Medical Institutions
tomy was first described more than three decades ago.5,6 Although the basic principles of the procedure remain the same, operative techniques, surgical instrumentation, and postoperative care have evolved during this time period. The objectives of this study were to describe the feasibility, associated morbidity, and efficacy of radical oophorectomy with primary stapled colorectal anastomosis for the clearance of pan-pelvic disease among patients with locally advanced ovarian cancer. METHODS Approval to conduct this study was obtained from the Johns Hopkins Medical Institutions (JHMI) Clinical Research Committee and Joint Committee on Clinical Investigation. All patients undergoing primary surgical intervention for a diagnosis of epithelial ovarian cancer during a 50-month time period October 1, 1997 through November 30, 2001 at the JHMI were prospectively entered into the Kelly Gynecologic Oncologic Service clinical database. Patients undergoing a radical pelvic resection for FIGO stage III or IV disease were specifically identified for additional study. Classification and description of surgical technique
To define the scope of surgical resection with uniform terminology, we have applied a descriptive classification system to the radical oophorectomy procedure. Briefly, a type I radical oophorectomy consists of a retrograde (described below) modified radical hysterectomy (resection of medial parametria and proximal vagina) with en bloc resection of the adnexae, pelvic cul-de-sac tumor, and involved pelvic peritoneum. The procedure is modified to include removal of the residual cervix if a supracervical hysterectomy was previously performed or resection of the vaginal apex if both uterus and cervix were previously removed. The procedure might include stripping of the peritoneum, serosa, or both, of the anterior sigmoid colon or a limited full-thickness segmental wedge-shaped resection of anterior sigmoid wall with primary closure. A type II radical oophorectomy consists of the previous procedure com-
J Am Coll Surg
bined with an en bloc resection of the rectosigmoid colon with complete parietal pelvic peritonectomy to encompass all pan-pelvic disease. Finally, a type III radical oophorectomy is an extension of the type II procedure, incorporating a portion of urinary bladder, pelvic ureter, or both with the en bloc resection. This report is devoted only to patients undergoing a type II or III radical oophorectomy, as these are most commonly performed in clinical practice. The primary indication for radical oophorectomy was FIGO stage III or IV epithelial ovarian cancer with confluent extension to and encasement of the reproductive organs, pelvic peritoneum, cul-de-sac of Douglas, and rectosigmoid colon (Fig. 1). Relative contraindications to the procedure included a Gynecologic Oncology Group (GOG) performance status score of 3 or greater (Karnofsky score ⱕ40) or tumor distribution precluding an attempt at optimal resection: extensive tumor infiltration of small bowel mesenteric root, celiac axis nodal involvement, unresectable involvement of the porta hepatis, largevolume (1 cm or more) unresectable extraabdominal metastasis (eg, pulmonary), or multiple unresectable parenchymal liver metastases. Radical oophorectomy was generally limited to patients in whom the procedure facilitated resection of all or nearly all (residual disease less than 1 cm in maximal diameter) of their disease and was not performed simply as palliation for a current or impending large bowel obstruction in the setting of bulky unresectable upper abdominal or extraabdominal disease. Preoperatively, all patients underwent a complete physical examination and routine biochemical and hematologic laboratory assessment. Routine radiographic imaging included a chest radiograph and abdominopelvic CT. Standard mammographic and colonoscopy screening recommendations were followed whenever possible. All patients received a full mechanical bowel preparation with either oral polyethylene glycol lavage or Fleets phosphosoda the day before surgery. Standard antibiotic prophylaxis consisted of cefotetan 2 g intravenously or clindamycin 600 mg intravenously 30 minutes before surgery. In patients with gross intraoperative fecal contamination, broad-spectrum antibiotic coverage was administered for 72 hours postoperatively. Antiembolism stockings and pneumatic compression devices were routinely placed before induction of anesthesia and continued until the patient was fully ambulatory. All operations were performed with the patient in the modified dorsal lithotomy position. Abdominal entry and exposure are achieved through a midline xyphopublic incision with placement of a self-
Vol. 197, No. 4, October 2003
Bristow et al
Radical Oophorectomy for Ovarian Cancer
567
Figure 1. Locally advanced ovarian cancer with confluent extension to and encasement of the reproductive organs, pelvic peritoneum (including vesicouterine peritoneal reflection), cul-de-sac of Douglas, and rectosigmoid colon. OvT, ovarian tumor; Ut, uterus.
retaining Bookwalter retractor (Codman Division, Johnson & Johnson). A comprehensive assessment of the extent of disease is conducted, with particular attention to the feasibility of upper abdominal cytoreduction. Initial efforts are directed toward bulky upper abdominal disease to facilitate exposure to the pelvis and ensure a reasonable likelihood of achieving an optimal (less than 1 cm) or complete (no macroscopic residual) overall resection before undertaking radical oophorectomy. Omentectomy, with mobilization of the ascending and descending colon, is followed by exploration of the abdominal retroperitoneum and debulking of peritoneal tumor implants. The radical oophorectomy procedure is initiated by carrying the paracolic gutter incisions caudally into the pelvis, along the psoas muscles bilaterally, moving ventromedially along the posterior margin of the symphisis pubis. All pan-pelvic disease is circumscribed and included within this peritoneal incision (Fig. 2). The pelvic dissection proceeds in a centripetal fashion. The round ligaments are exposed retroperitoneally, ligated, and divided as laterally as possible. The central tumor mass(es) should be devascular-
ized early in the course of the operation by securing the infundibulopelvic ligaments (containing the ovarian vessels) with suture ligatures and dividing them at or above the pelvic brim. The ureters are mobilized from their attachments to the medial leaf of the broad ligament, moving from the pelvic brim to the tunnel of Wertheim, and held for traction with vasa-loops. Medial mobilization of the sigmoid colon will provide additional exposure to the left pelvis and delineate the extent of bowel resection needed. The proximal sigmoid colon is divided 2 cm to 3 cm above the most proximal extent of gross tumor using an automated linear gastrointestinal (GIA) stapling device (US Surgical or Ethicon Inc). The mesentery of the sigmoid colon is then divided, ligating individual vessels, including the superior hemorrhoidal artery. The posterior pelvis is further mobilized by developing the retrorectal (presacral) space caudally to the level of pelvic floor musculature, clamping, dividing and securing the rectal pillars (including the middle hemorrhoidal artery) in the process. The pararectal and paravesicle spaces are further developed using a combination of sharp and blunt dissec-
568
Bristow et al
Radical Oophorectomy for Ovarian Cancer
J Am Coll Surg
Figure 2. Radical oophorectomy. A circumscribing peritoneal incision encompasses all panpelvic disease, the round ligaments and ovarian vessels are divided, the ureters are mobilized, and the anterior pelvic peritoneal tumor is dissected from the bladder muscularis.
tion, exposing the cardinal ligament. Resection of any bulky pelvic adenopathy at this point may facilitate exposure to the central pelvis. The uterine vascular pedicles are skeletonized, doubly ligated, and divided at the level of the ureters (in the fashion of a modified radical hysterectomy), allowing additional lateral displacement of the ureters from the central specimen7 (Fig. 2). The retropubic space of Retzuis is entered and a plane of dissection established between the muscular bladder wall and tumor-laden anterior pelvic peritoneum. The anterior pelvic peritoneum is deperitonealized (or “stripped”) in a ventral-to-dorsal direction, inclusive of all tumor implants occupying the anterior pelvis. Bovie electrocautery (Valleylab Inc) (30 to 35 W, coagulation current) or the argon beam coagulator (Conmed Corp) (80 W power density) facilitates dissection of the anterior pelvic peritoneum from the underlying bladder muscularis until the pubovesicocervical fascia is reached. The anterior pelvic peritoneal tumor is left attached to the uterine specimen. The hysterectomy is completed in a retrograde fashion. The bladder is sharply mobilized ventrocaudally to expose the proximal 2 cm to 3 cm of vagina. If the
anterior pelvic peritoneal tumor is densely adherent to the cervix, the paravesicle spaces can be developed in a lateral-to-medial direction below the level of the cervix until the vesicovaginal space is reached. The vesicovaginal space is virtually always free of disease and can be used to define the proper plane of dissection between the adherent tumor and the bladder wall. Occasionally, an intentional cystotomy in the bladder dome, with partial resection of the bladder wall, is required. At this juncture, the need for partial ureterectomy is also determined. Ureteral reimplantation by ureteroneocystotomy (with or without Boari flap and psoas hitch) is normally delayed until the central specimen has been removed. The proximal vagina is exposed and a transverse anterior colpotomy is created using electrocautery, exposing the inner vagina (Fig. 3A). An intraoperative bimanual examination facilitates selecting the proper site of vaginal incision. In cases in which a hysterectomy has previously been performed, transvaginal placement of a rectal dilator to elevate the vaginal cuff might facilitate anterior colpotomy. Heaney clamps are used to sequentially circumscribe the anterior and lateral vagina, dividing and securing each pedicle in sequence. Place-
Vol. 197, No. 4, October 2003
Bristow et al
Radical Oophorectomy for Ovarian Cancer
569
Figure 4. Radical oophorectomy. The rectovaginal space has been developed to a level 2 to 3 cm below the caudal-most extent of the cul-de-sac tumor mass, and the distal rectosigmoid colon is divided using an automated stapling device.
Figure 3. Radical oophorectomy. (A) The anterior pelvic peritoneal tumor has been dissected from the bladder dome, the proximal vagina is exposed, and a transverse anterior colpotomy is created using electrocautery to enter the vagina. (B) The remaining cardinal ligament attachments are divided between Heaney clamps working in a ventral-to-dorsal direction toward the cul-de-sac tumor mass. ligs., ligaments.
ment of a narrow malleable retractor in the vaginal tube will help to displace the bladder and distal ureters anterolaterally during this portion of the dissection. The remaining cardinal ligament attachments are divided between Heaney clamps working in an anterior (ventral) to posterior (dorsal) direction (Fig. 3B). The posterior vaginal wall is incised and the rectovaginal space developed sharply. Dissection is carried inferiorly to a level 2 to 3 cm below the lowermost extent of the cul-de-sac tumor mass, which is then mobilized cephalad to preserve max-
imal length of the proximal rectum after transection. At this point, the only remaining attachment is that of the distal rectosigmoid colon, which is divided using a GIA or thoracoabdominal (TA) automated stapling device (US Surgical or Ethicon Inc) (Fig. 4). The central pelvic tumor mass is removed en bloc with the rectosigmoid colon (Fig. 5), leaving a macroscopically tumor-free operative site (Fig. 6). Additional mobilization of the proximal sigmoid colon might be required to ensure a tension-free colorectal anastomosis.8 Intestinal continuity is reestablished, using a circular end-to-end anastomosis (CEEA) automated stapling device (Fig. 7) (US Surgical or Ethicon Inc). The security of the anastomosis is confirmed by inspecting the resection rings to ensure two complete “donuts” of colon and filling the pelvis with saline, while manually obstructing the proximal colon, and instilling 150 to 200 mL of air into the rectum to demonstrate an airtight anastomosis. In most cases, a closed suction drainage system is placed within the pelvis and removed when the output is less than 200 mL per 24-hour period. Nasogastric suction is not routinely used, and parenteral nutrition is administered at the discretion of the attending surgeon.
570
Bristow et al
Radical Oophorectomy for Ovarian Cancer
Figure 5. Radical oophorectomy. En bloc specimen, including uterus, adnexae, anterior pelvic peritoneal tumor, cul-de-sac tumor mass, and rectosigmoid colon. CT, cul-de-sac tumor; Ut, uterus.
Data abstraction
Individual subjects were identified prospectively at the time of surgery and data collected from inpatient and ambulatory medical records. Demographic data, operative findings, surgical procedures performed, amount of residual disease, and final pathology diagnosis were recorded for all patients. Followup information included significant postoperative morbidity within 30 days of
Figure 7. Radical oophorectomy. Intestinal continuity is reestablished using a circular end-to-end anastomosis (CEEA) automated stapling device.
J Am Coll Surg
Figure 6. Radical oophorectomy. The pelvis is macroscopically tumor-free after en bloc resection.
surgery, chemotherapeutic agents administered and duration of treatment, the date of clinical or radiographic progression of disease or recurrence, the date of last followup or death, and disease status at last followup. For overall survival analysis, event time distributions were estimated using the method of Kaplan and Meier.9 RESULTS Patient characteristics
Thirty-one consecutive patients underwent a primary maximal surgical effort at the JHMI that included Type II (n ⫽ 29) or Type III (n ⫽ 2) radical oophorectomy for epithelial ovarian cancer with confluent involvement of pelvic viscera during the study period. The FIGO stage distribution was as follows: stage IIIB (2 patients), stage IIIC (20 patients), stage IV (9 patients). Three patients had received three cycles of neoadjuvant chemotherapy with carboplatin and paclitaxel before their initial surgical procedure for FIGO stage IV serous ovarian carcinoma. One patient (stage IIIC disease) was treated with three cycles of carboplatin and paclitaxel chemotherapy after a suboptimal attempt at primary cytoreductive surgery at an outside institution before undergoing radical oophorectomy at JHMI. The remaining 27 patients all underwent surgical exploration before receiving chemotherapy. At the time of radical oopho-
Bristow et al
Vol. 197, No. 4, October 2003
Table 1. Clinicopathologic Patient Characteristics Characteristic
Median age, y (range) Serum CA125, U/mL Median Mean Range Performance status* 0 1 2 3 Ascites volume, mL Median Mean Range Histology, n (%) Serous Mixed serous and endometrioid Mucinous Clear cell Tumor grade, n (%) 1 2 3
Data
63 (40 to 85) 1,122 2,524 74 to 17,935 12 (38.7%) 14 (45.2%) 5 (16.1%)
2,700 2,494 100 to 7,000 26 (83.9) 2 (6.5) 2 (6.5) 1 (3.2) 2 (6.5) 8 (25.8) 21 (67.7)
*Gynecologic Oncology Group scoring.
rectomy, 22 of 31 patients (71%) had ascites volume greater than or equal to 1,000 mL. Additional clinicopathologic patient characteristics are shown in Table 1. Surgical results and postoperative treatment
Radical oophorectomy included hysterectomy in 21 patients, trachelectomy (after previous supracervical hysterectomy) in 2 patients, and uni- or bilateral salpingooophorectomy only (after earlier total hysterectomy) in 8 patients. In all cases, rectosigmoid colon resections were completed below the peritoneal reflection and intestinal continuity reestablished using a CEEA stapler (25 mm to 33 mm). Additional surgical procedures are shown in Table 2. Twenty-seven patients (87.1%) were left with residual disease measuring 1 cm or more in maximal diameter. Of these, 19 patients underwent complete cytoreduction of all visible disease, representing 61.3% of the total study group and 70.4% of those with optimal cytoreduction. Operative time ranged from 165 to 330 minutes, with a median time of 240 minutes (mean 235.3 minutes). The median estimated blood loss was 700 mL (mean 821 mL, range 300 to 2,900 mL). Estimated blood loss
Radical Oophorectomy for Ovarian Cancer
571
Table 2. Surgical Procedures Performed in 31 Patients Undergoing Radical Oophorectomy Procedure
n
%
Total omentectomy Infracolic omentectomy Retroperitoneal lymph node sampling/excision Peritoneal tumor implant excision or ablation Ileocecal or small bowel resection Diaphragm stripping or resection Right or left hemicolectomy Subtotal transverse colectomy† Splenectomy Liver resection Cholecystectomy
28 3
90.3 9.7
28
90.3*
23
74.2
9
29.0
7 3 2 2 1 1
22.6 9.7 6.5 6.5 3.2 3.2
*Retroperitoneal lymph nodes contained metastatic disease in 17 of 28 patients (60.7%). † En bloc with omentectomy.
in excess of 1,000 mL occurred in seven patients (22.6%). Nine patients (29.0%) required intraoperative or postoperative transfusion of packed red blood cells (range 1 to 5 U, median 2 U, mean 2.6 U). Twenty patients (64.5%) were transferred directly to the surgical intensive care unit immediately postoperatively, for a median stay of 2 days (range 1 to 12 days). The median time to tolerating a regular diet was 6 days (range 3 to 14 days). Thirteen patients received parenteral nutrition postoperatively (median 8 days, range 6 to 14 days). There were no perioperative deaths, but four patients (12.9%) experienced significant life-threatening postoperative complications (pulmonary embolism, sepsis, upper gastrointestinal hemorrhage). Minor postoperative morbidity (incisional cellulitis, urinary tract infection, ileus more than 7 days) occurred in 11 patients (35.5%). There were no instances of prolonged (beyond 7 postoperative days) bladder dysfunction. One patient (3.2%) experienced a postoperative breakdown of the colorectal anastomosis on postoperative day 5 and required reexploration, with diverting colostomy. This was the only patient in this series requiring reoperation. The median length of hospital stay was 11 days (range 4 to 23 days), and 22 patients received chemotherapy before discharge. Followup and survival analysis
Postoperatively, one patient declined further treatment and was lost to followup 43 days after surgery. Of the 30
572
Bristow et al
Radical Oophorectomy for Ovarian Cancer
J Am Coll Surg
their first site of progression, representing 16.7% of the 30 evaluable patients.
Figure 8. Overall survival of 30 patients undergoing radical oophorectomy for International Federation of Gynecology and Obstetrics stage III (n ⫽ 21) or stage IV (n ⫽ 9) epithelial ovarian cancer.
patients receiving chemotherapy, carboplatin plus paclitaxel was administered to 24; 5 patients received cisplatin plus paclitaxel plus topotecan as part of a cooperative group investigational study protocol (GOG protocol #9602). One patient received single agent carboplatin. All patients, including those treated initially with neoadjuvant chemotherapy, completed a minimum of six cycles after radical oophorectomy. Fifteen patients underwent second-look surgery (14 laparoscopy, 1 laparotomy). Seven of these patients (46.6%) had persistent disease detected at second-look surgery (six microscopically positive, one macroscopically positive). No patient was found to have persistent pelvic disease at the time of second-look surgery. Thirty patients were available for survival analysis. The median followup time for surviving patients was 21.7 months (range 7.2 months to 57.4 months). At last followup, 6 patients (20%) were dead of disease, 10 patients (33.3%) were alive with disease, and 14 patients (46.7%) were alive with no evidence of disease. The median overall survival time for all patients was 39.5 months (Fig. 8). At last followup, 15 patients (48.4%) had experienced clinical recurrence or progression of disease. Five patients had a pelvic component of disease as
DISCUSSION For patients with FIGO stage III/IV ovarian cancer, the amount of residual disease before initiating cytotoxic chemotherapy has been shown to be a powerful and consistent determinant of overall survival.3,10-17 Although the contribution of tumor biology to this observation cannot be overlooked, primary cytoreductive surgery is the accepted standard initial therapy for women with advanced epithelial ovarian cancer. The literature has consistently demonstrated that patients with no visible or only small volume residual tumor experience a two- to three-fold extension in median survival time compared with those left with bulky residual tumor.17-20 Contemporary 5-year survival rates now approach 50% for patients with optimally cytoreduced stage III/IV disease.17-19,21 Not uncommonly, advanced ovarian cancer will present with confluent disease extending to or encasing the reproductive organs and other pelvic viscera, which might lead to an abbreviated debulking procedure or abandonment of primary surgery altogether.15,22 In some series, as many as 47% of patients with advanced ovarian cancer might be left with suboptimal large-volume residual pelvic disease.23-25 Given the poor prognosis in such cases, every effort should be made to resect even the most locally advanced ovarian cancer when upper abdominal disease is not prohibitive. Radical extirpative procedures for locally advanced ovarian cancer have evolved since 1965, when Barber and Brunschwig26 first reported 22 patients undergoing pelvic exenteration. In this series, the postoperative mortality rate was 23% and there were only two longterm survivors. In 1968 and 1973, Hudson and Chir5,6 published two reports describing a technique they termed “radical oophorectomy,” specifically designed for the intact removal of a fixed ovarian tumor en bloc with attached peritoneum and surrounding structures. These authors observed that ovarian cancer tumor growth usually respects peritoneal demarcations, and they advocated a retroperitoneal approach, using the “false capsule” of the ovarian tumor within the pouch of Douglas to effect en bloc excision. During the past 30 years, varying terminology has been used to describe modifications of this procedure, including: en bloc rectosigmoid colectomy,20,27-31 reverse hysterocolposigmoidectomy,32 complete parietal and visceral peritonectomy,8
Vol. 197, No. 4, October 2003
Bristow et al
en bloc pelvic peritoneal resection of the intrapelvic viscera,33 and modified posterior exenteration.34 Terminology aside, the cardinal feature of the radical oophorectomy procedure is the retroperitoneal approach to ovarian cancer encasing the pelvic viscera, using the tendency of epithelial ovarian cancer to respect peritoneal planes of demarcation to surgical advantage. In this fashion, the retroperitoneal spaces, uninvolved by extensive intraperitoneal tumor, can be used to develop the dissection in a centripetal fashion with maximum safety to surrounding vital structures. The type II/III procedures performed in the current series achieved complete clearance of macroscopic pan-pelvic disease in all cases and contributed significantly to an overall maximal cytoreductive effort. It should be noted that the observed rates of optimal and complete cytoreduction in the current series might not reflect the surgical outcomes for all patients with advanced ovarian cancer. Specifically, only patients in whom a reasonable attempt at resection of upper abdominal disease could be undertaken were submitted to radical oophorectomy. Although this necessarily introduces a component of selection bias, other authors have reported equivalent rates of successful cytoreduction using a similar technique.8,16,20,21,25,30,31-35 In these reports, overall optimal residual disease (ⱕ1 cm or 2 cm) was achievable in 74% to 100% of patients, with almost universal clearance of advanced pelvic tumor. The literature also suggests a reduced risk of pelvic failure (5.3% to 9.4%) after radical oophorectomy, which is consistent with our observation that just 16.7% of patients manifested a pelvic component at first recurrence.35,36 Given the high proportion of patients with stage IV disease, the observed median survival time of 39.5 months compares favorably with other contemporary reports. The current data also demonstrate that the extensive surgery often required to achieve such clinical outcomes is associated with a predictably high, but acceptable, risk of perioperative morbidity. Among patients undergoing a radical oophorectomy procedure, the reported estimated blood loss ranges from 800 mL to 2,900 mL, with 12% to 49% of patients experiencing significant postoperative complications.16,20,21,25,28,29,31-34 Nevertheless, operative mortality of radical oophorectomy can be limited to 1.5% to 3.1% using an intensive postoperative care program.20,29,30,34 Historically, there have been concerns about reestablishing intestinal continuity in patients undergoing a radical debulking operation for ovarian cancer requiring
Radical Oophorectomy for Ovarian Cancer
573
resection of the sigmoid colon. Such patients often have large-volume ascites, can be nutritionally compromised from the metabolic effects of an extensive tumor burden, or have evidence of early bowel obstruction. Temporary or permanent intestinal diversion was typically performed in 12% to 59% of such patients.28-30,32,34,35,37 In this series, primary colorectal anastomosis was accomplished in all cases using a CEEA automated stapling device with only one instance (3.2%) of anastomotic dehiscence requiring reoperation and colostomy despite the fact that 71% of patients had ascites volume of 1,000 mL or more. Using both stapled and hand-sewn anastomoses, other authors have reported anastomotic breakdown in 0% to 8% of patients undergoing sigmoid colectomy for ovarian cancer.20,27,29-32,37 Among patients with large-volume ascites (500 mL or more), the incidence of anastomotic dehiscence is just 2.1% to 3.1%.30,34 Taken together, these data suggest that protective intestinal diversion is unnecessary and that terminal excretory function can be preserved in the majority of patients undergoing radical oophorectomy. In summary, surgeons operating on women with ovarian cancer are frequently confronted with locally advanced central pelvic disease with contiguous extension to or encasement of other pelvic viscera. Smallvolume residual disease has been consistently associated with improved survival, and for the majority of patients a maximal attempt at tumor cytoreduction will be warranted. Radical oophorectomy is highly effective for achieving complete clearance of macroscopic pelvic disease and, with advances in surgical technique and perioperative care, can be performed with acceptable morbidity. On the other hand, patients with unresectable upper abdominal disease are unlikely to derive a significant survival benefit from a radical pelvic resection. In such cases, intestinal diversion alone is the more appropriate surgical option. Author contributions
Study conception and design: Bristow, del Carmen, Kaufman, Montz Acquisition of data: Bristow, del Carmen Analysis and interpretation of data: Bristow, del Carmen, Kaufman, Montz Drafting of manuscript: Bristow, del Carmen, Kaufman, Montz Critical revision: del Carmen, Kaufman, Montz Statistical expertise: Bristow
574
Bristow et al
Radical Oophorectomy for Ovarian Cancer
Obtaining funding: Bristow Supervision: Kaufman, Montz 21. REFERENCES 1. Jemal A, Thomas A, Murray T, Thun M. Cancer statistics 2002. CA Cancer J Clin 2002;52:23–47. 2. Announcements. FIGO stages—1988 revision. Gynecol Oncol 1989;35:125–127. 3. Hoskins WJ, Bundy BN, Thigpen JT, Omura GA. The influence of cytoreductive surgery on recurrence-free interval and survival in small-volume Stage III epithelial ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1992;47:159–166. 4. Marsden DE, Friedlander M, Hacker NF. Current management of epithelial ovarian carcinoma: a review. Semin Surg Oncol 2000;19:11–19. 5. Hudson CN. A radical operation for fixed ovarian tumors. J Obstet Gynaecol Br Cwlth 1968;75:1155–1160. 6. Hudson CN, Chir M. Surgical treatment of ovarian cancer. Gynecol Oncol 1973;1:370–378. 7. Morrow CP, Curtin JP. Gynecologic cancer surgery, 1st ed. New York: Churchill Livingstone, Inc; 1996, 451–568. 8. Sugarbaker PH. Complete parietal and visceral peritonectomy of the pelvis for advanced primary and recurrent ovarian cancer. Cancer Treat Res 1996;81:75–87. 9. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–480. 10. Hacker NF, Berek JS, Lagasse LD, et al. Primary cytoreductive surgery for epithelial ovarian cancer. Obstet Gynecol 1983;61: 413–420. 11. Heintz APM, Hacker NF, Berek JS, et al. Cytoreductive surgery in ovarian carcinoma: feasibility and morbidity. Obstet Gynecol 1986;67:783–788. 12. Piver MS, Baker T. The potential for optimal (ⱕ2 cm) cytoreductive surgery in advanced ovarian carcinoma at a tertiary medical center: a prospective study. Gynecol Oncol 1986;24:1–8. 13. Piver MS, Lele SB, Marchetti DL, et al. The impact of aggressive debulking surgery and cisplatin-based chemotherapy on progression-free survival in stage III and IV ovarian carcinoma. J Clin Oncol 1988;6:983–989. 14. Bertelson K. Tumor reduction surgery and long-term survival in advanced ovarian cancer: a DACOVA study. Gynecol Oncol 1990;38:203–209. 15. Eisenkop SM, Spirtos NM, Montag TW, et al. The impact of subspecialty training on the management of advanced ovarian cancer. Gynecol Oncol 1992;47:203–209. 16. Guidozzi F, Ball JH. Extensive primary cytoreductive surgery for advanced epithelial ovarian cancer. Gynecol Oncol 1994;53: 326–330. 17. Eisenkop SM, Friedman RL, Wang HJ. Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced epithelial ovarian cancer: a prospective study. Gynecol Oncol 1998;69:103–108. 18. Chi DS, Liao JB, Leon LF, et al. Identification of prognostic factors in advanced epithelial ovarian carcinoma. Gynecol Oncol 2001;82:532–537. 19. Dauplat J, LeBoue¨dec G, Pomel C, Scherer C. Cytoreductive surgery for advanced stages of ovarian cancer. Semin Surg Oncol 2000;19:42–48. 20. Clayton RD, Obemair A, Hammond IG, et al. The western
22.
23. 24.
25. 26. 27. 28. 29. 30.
31.
32. 33.
34. 35.
36.
37.
J Am Coll Surg
Australian experience of the use of en bloc resection of ovarian cancer with concomitant rectosigmoid colectomy. Gynecol Oncol 2002;84:53–57. Scarabelli C, Gallo A, Franceschi S, et al. Primary cytoreductive surgery with rectosigmoid colon resection for patients with advanced epithelial ovarian carcinoma. Cancer 2000;88:389–397. Devi KU, Banfa UD, Ahuja V, Ramesh C. Induction chemotherapy and interval debulking surgery in advanced epithelial ovarian cancer. Int J Gynecol Cancer 2002;12:521 (Abstract #OV8). Webb MJ. Cytoreduction in ovarian cancer: achievability and results. Baillieres Clin Obstet Gynaecol 1989;3:83–94. Makar AP, Baekelandt M, Trope´ CG, Kristensen GB. The prognostic significance of residual disease, FIGO substage, tumor histology and grade in patients with FIGO Stage III ovarian cancer. Gynecol Oncol 1995;56:175–180. Benedetti-Panici P, Maneschi F, Scambia G, et al. The pelvic retroperitoneal approach in the treatment of advanced ovarian carcinoma. Obstet Gynecol 1996;87:532–538. Barber HRK, Brunschwig A. Pelvic exenteration for locally advanced and recurrent ovarian cancer. Surgery 1965;58:935– 937. Sonnendecker EWW, Beale PG. Rectosigmoid resection without colostomy during primary cytoreductive surgery for ovarian carcinoma. Int Surg 1989;74:10–12. Berek JS, Hacker NF, Lagasse LD. Rectosigmoid colectomy and reanastomosis to facilitate resection of primary and recurrent gynecologic cancer. Obstet Gynecol 1984;64:715–720. Soper JT, Couchman G, Berchuk A, Clarke-Pearson D. The role of partial sigmoid colectomy for debulking epithelial ovarian carcinoma. Gynecol Oncol 1991;41:239–244. Obermair A, Hagenauer S, Tamandl D, et al. Safety and efficacy of low anterior en bloc resection as part of cytoreductive surgery for patients with ovarian cancer. Gynecol Oncol 2001;83:115–120. Bridges JE, Leung Y, Hammond IG, McCartney AJ. En bloc resection of epithelial ovarian tumors with concomitant rectosigmoid colectomy: the KEMH experience. Int J Gynecol Cancer 1993;3:199–202. Barnes W, Johnson J, Waggoner S, et al. Reverse hysterocolposigmoidectomy (RHCS) for resection of panpelvic tumors. Gynecol Oncol 1991;42:151–155. Sainz de la Cuesta R, Goodman A, Halverson SS, Fuller AF. En bloc pelvic peritoneal resection of the intraperitoneal pelvic viscera in patients with advanced epithelial ovarian cancer. Cancer J Sci Am 1996;2:152–157. Eisenkop SM, Nalick RH, Teng NNH. Modified posterior exenteration for ovarian cancer. Obstet Gynecol 1991;78:879– 885. Hertel H, Diebolder H, Herrmann J, et al. Is the decision for colorectal resection justified by histopathologic findings: a prospective study of 100 patients with advanced ovarian cancer. Gynecol Oncol 2001;83:481–484. Spirtos NM, Eisenkop SM, Schlaerth JB, Ballon SC. Secondlook laparotomy after modified posterior exenteration: patterns of persistence and recurrence in patients with stage III and stage IV ovarian cancer. Am J Obstet Gynecol 2000;182:1321–1327. Tamussino KF, Lim PC, Webb MJ, et al. Gastrointestinal surgery in patients with ovarian cancer. Gynecol Oncol 2001;80: 79–84.
Ovarian cancer is the most lethal of all gynecologic malignancies. The American Cancer Society has estimated that 23,300 new cases will be diagnosed in the United
States in 2002, with 13,900 deaths directly attributable to this disease.1 For patients with International Federation of Gynecology and Obstetrics (FIGO) stage III/IV disease, survival determinants are multifactorial, but the strongest clinician-driven predictors of clinical outcomes are the amount of residual tumor after primary cytoreductive surgery and the administration of platinum-based chemotherapy.2-4 Locally advanced disease, with contiguous extension to or encasement of the reproductive organs, pelvic peritoneum, cul-de-sac, and sigmoid colon, can present a significant challenge to surgeons operating on women with ovarian cancer. Designed for intact removal of a fixed ovarian tumor en bloc with attached peritoneum and surrounding structures, the retroperitoneal technique of radical oophorec-
No competing interests declared.
This work supported by the Elizabeth Frost Ovarian Cancer Research Fund. Received December 26, 2002; Revised April 28, 2003; Accepted April 28, 2003. From The Kelly Gynecologic Oncology Service, Departments of Gynecology and Obstetrics and Oncology (Bristow, del Carmen, Montz), and the Department of Surgery (Kaufman, Montz), The Johns Hopkins Medical Institutions, Baltimore, MD. Dr Kaufman’s current affiliation is Division of Colorectal Surgery, Department of Surgery, University of Southern California Keck School of Medicine, Los Angeles, CA. Correspondence address: Robert E Bristow, MD, The Kelly Gynecologic Oncology Service, Departments of Gynecology and Obstetrics and Oncology, The Johns Hopkins Medical Institutions, 600 North Wolfe St, Phipps #289, Baltimore, MD 21205.
© 2003 by the American College of Surgeons Published by Elsevier Inc.
565
ISSN 1072-7515/03/$21.00 doi:10.1016/S1072-7515(03)00478-2
566
Bristow et al
Radical Oophorectomy for Ovarian Cancer
Abbreviations and Acronyms
CEEA ⫽ circular end-to-end anastomosis FIGO ⫽ International Federation of Gynecology and Obstetrics GOG ⫽ Gynecologic Oncology Group JHMI ⫽ Johns Hopkins Medical Institutions
tomy was first described more than three decades ago.5,6 Although the basic principles of the procedure remain the same, operative techniques, surgical instrumentation, and postoperative care have evolved during this time period. The objectives of this study were to describe the feasibility, associated morbidity, and efficacy of radical oophorectomy with primary stapled colorectal anastomosis for the clearance of pan-pelvic disease among patients with locally advanced ovarian cancer. METHODS Approval to conduct this study was obtained from the Johns Hopkins Medical Institutions (JHMI) Clinical Research Committee and Joint Committee on Clinical Investigation. All patients undergoing primary surgical intervention for a diagnosis of epithelial ovarian cancer during a 50-month time period October 1, 1997 through November 30, 2001 at the JHMI were prospectively entered into the Kelly Gynecologic Oncologic Service clinical database. Patients undergoing a radical pelvic resection for FIGO stage III or IV disease were specifically identified for additional study. Classification and description of surgical technique
To define the scope of surgical resection with uniform terminology, we have applied a descriptive classification system to the radical oophorectomy procedure. Briefly, a type I radical oophorectomy consists of a retrograde (described below) modified radical hysterectomy (resection of medial parametria and proximal vagina) with en bloc resection of the adnexae, pelvic cul-de-sac tumor, and involved pelvic peritoneum. The procedure is modified to include removal of the residual cervix if a supracervical hysterectomy was previously performed or resection of the vaginal apex if both uterus and cervix were previously removed. The procedure might include stripping of the peritoneum, serosa, or both, of the anterior sigmoid colon or a limited full-thickness segmental wedge-shaped resection of anterior sigmoid wall with primary closure. A type II radical oophorectomy consists of the previous procedure com-
J Am Coll Surg
bined with an en bloc resection of the rectosigmoid colon with complete parietal pelvic peritonectomy to encompass all pan-pelvic disease. Finally, a type III radical oophorectomy is an extension of the type II procedure, incorporating a portion of urinary bladder, pelvic ureter, or both with the en bloc resection. This report is devoted only to patients undergoing a type II or III radical oophorectomy, as these are most commonly performed in clinical practice. The primary indication for radical oophorectomy was FIGO stage III or IV epithelial ovarian cancer with confluent extension to and encasement of the reproductive organs, pelvic peritoneum, cul-de-sac of Douglas, and rectosigmoid colon (Fig. 1). Relative contraindications to the procedure included a Gynecologic Oncology Group (GOG) performance status score of 3 or greater (Karnofsky score ⱕ40) or tumor distribution precluding an attempt at optimal resection: extensive tumor infiltration of small bowel mesenteric root, celiac axis nodal involvement, unresectable involvement of the porta hepatis, largevolume (1 cm or more) unresectable extraabdominal metastasis (eg, pulmonary), or multiple unresectable parenchymal liver metastases. Radical oophorectomy was generally limited to patients in whom the procedure facilitated resection of all or nearly all (residual disease less than 1 cm in maximal diameter) of their disease and was not performed simply as palliation for a current or impending large bowel obstruction in the setting of bulky unresectable upper abdominal or extraabdominal disease. Preoperatively, all patients underwent a complete physical examination and routine biochemical and hematologic laboratory assessment. Routine radiographic imaging included a chest radiograph and abdominopelvic CT. Standard mammographic and colonoscopy screening recommendations were followed whenever possible. All patients received a full mechanical bowel preparation with either oral polyethylene glycol lavage or Fleets phosphosoda the day before surgery. Standard antibiotic prophylaxis consisted of cefotetan 2 g intravenously or clindamycin 600 mg intravenously 30 minutes before surgery. In patients with gross intraoperative fecal contamination, broad-spectrum antibiotic coverage was administered for 72 hours postoperatively. Antiembolism stockings and pneumatic compression devices were routinely placed before induction of anesthesia and continued until the patient was fully ambulatory. All operations were performed with the patient in the modified dorsal lithotomy position. Abdominal entry and exposure are achieved through a midline xyphopublic incision with placement of a self-
Vol. 197, No. 4, October 2003
Bristow et al
Radical Oophorectomy for Ovarian Cancer
567
Figure 1. Locally advanced ovarian cancer with confluent extension to and encasement of the reproductive organs, pelvic peritoneum (including vesicouterine peritoneal reflection), cul-de-sac of Douglas, and rectosigmoid colon. OvT, ovarian tumor; Ut, uterus.
retaining Bookwalter retractor (Codman Division, Johnson & Johnson). A comprehensive assessment of the extent of disease is conducted, with particular attention to the feasibility of upper abdominal cytoreduction. Initial efforts are directed toward bulky upper abdominal disease to facilitate exposure to the pelvis and ensure a reasonable likelihood of achieving an optimal (less than 1 cm) or complete (no macroscopic residual) overall resection before undertaking radical oophorectomy. Omentectomy, with mobilization of the ascending and descending colon, is followed by exploration of the abdominal retroperitoneum and debulking of peritoneal tumor implants. The radical oophorectomy procedure is initiated by carrying the paracolic gutter incisions caudally into the pelvis, along the psoas muscles bilaterally, moving ventromedially along the posterior margin of the symphisis pubis. All pan-pelvic disease is circumscribed and included within this peritoneal incision (Fig. 2). The pelvic dissection proceeds in a centripetal fashion. The round ligaments are exposed retroperitoneally, ligated, and divided as laterally as possible. The central tumor mass(es) should be devascular-
ized early in the course of the operation by securing the infundibulopelvic ligaments (containing the ovarian vessels) with suture ligatures and dividing them at or above the pelvic brim. The ureters are mobilized from their attachments to the medial leaf of the broad ligament, moving from the pelvic brim to the tunnel of Wertheim, and held for traction with vasa-loops. Medial mobilization of the sigmoid colon will provide additional exposure to the left pelvis and delineate the extent of bowel resection needed. The proximal sigmoid colon is divided 2 cm to 3 cm above the most proximal extent of gross tumor using an automated linear gastrointestinal (GIA) stapling device (US Surgical or Ethicon Inc). The mesentery of the sigmoid colon is then divided, ligating individual vessels, including the superior hemorrhoidal artery. The posterior pelvis is further mobilized by developing the retrorectal (presacral) space caudally to the level of pelvic floor musculature, clamping, dividing and securing the rectal pillars (including the middle hemorrhoidal artery) in the process. The pararectal and paravesicle spaces are further developed using a combination of sharp and blunt dissec-
568
Bristow et al
Radical Oophorectomy for Ovarian Cancer
J Am Coll Surg
Figure 2. Radical oophorectomy. A circumscribing peritoneal incision encompasses all panpelvic disease, the round ligaments and ovarian vessels are divided, the ureters are mobilized, and the anterior pelvic peritoneal tumor is dissected from the bladder muscularis.
tion, exposing the cardinal ligament. Resection of any bulky pelvic adenopathy at this point may facilitate exposure to the central pelvis. The uterine vascular pedicles are skeletonized, doubly ligated, and divided at the level of the ureters (in the fashion of a modified radical hysterectomy), allowing additional lateral displacement of the ureters from the central specimen7 (Fig. 2). The retropubic space of Retzuis is entered and a plane of dissection established between the muscular bladder wall and tumor-laden anterior pelvic peritoneum. The anterior pelvic peritoneum is deperitonealized (or “stripped”) in a ventral-to-dorsal direction, inclusive of all tumor implants occupying the anterior pelvis. Bovie electrocautery (Valleylab Inc) (30 to 35 W, coagulation current) or the argon beam coagulator (Conmed Corp) (80 W power density) facilitates dissection of the anterior pelvic peritoneum from the underlying bladder muscularis until the pubovesicocervical fascia is reached. The anterior pelvic peritoneal tumor is left attached to the uterine specimen. The hysterectomy is completed in a retrograde fashion. The bladder is sharply mobilized ventrocaudally to expose the proximal 2 cm to 3 cm of vagina. If the
anterior pelvic peritoneal tumor is densely adherent to the cervix, the paravesicle spaces can be developed in a lateral-to-medial direction below the level of the cervix until the vesicovaginal space is reached. The vesicovaginal space is virtually always free of disease and can be used to define the proper plane of dissection between the adherent tumor and the bladder wall. Occasionally, an intentional cystotomy in the bladder dome, with partial resection of the bladder wall, is required. At this juncture, the need for partial ureterectomy is also determined. Ureteral reimplantation by ureteroneocystotomy (with or without Boari flap and psoas hitch) is normally delayed until the central specimen has been removed. The proximal vagina is exposed and a transverse anterior colpotomy is created using electrocautery, exposing the inner vagina (Fig. 3A). An intraoperative bimanual examination facilitates selecting the proper site of vaginal incision. In cases in which a hysterectomy has previously been performed, transvaginal placement of a rectal dilator to elevate the vaginal cuff might facilitate anterior colpotomy. Heaney clamps are used to sequentially circumscribe the anterior and lateral vagina, dividing and securing each pedicle in sequence. Place-
Vol. 197, No. 4, October 2003
Bristow et al
Radical Oophorectomy for Ovarian Cancer
569
Figure 4. Radical oophorectomy. The rectovaginal space has been developed to a level 2 to 3 cm below the caudal-most extent of the cul-de-sac tumor mass, and the distal rectosigmoid colon is divided using an automated stapling device.
Figure 3. Radical oophorectomy. (A) The anterior pelvic peritoneal tumor has been dissected from the bladder dome, the proximal vagina is exposed, and a transverse anterior colpotomy is created using electrocautery to enter the vagina. (B) The remaining cardinal ligament attachments are divided between Heaney clamps working in a ventral-to-dorsal direction toward the cul-de-sac tumor mass. ligs., ligaments.
ment of a narrow malleable retractor in the vaginal tube will help to displace the bladder and distal ureters anterolaterally during this portion of the dissection. The remaining cardinal ligament attachments are divided between Heaney clamps working in an anterior (ventral) to posterior (dorsal) direction (Fig. 3B). The posterior vaginal wall is incised and the rectovaginal space developed sharply. Dissection is carried inferiorly to a level 2 to 3 cm below the lowermost extent of the cul-de-sac tumor mass, which is then mobilized cephalad to preserve max-
imal length of the proximal rectum after transection. At this point, the only remaining attachment is that of the distal rectosigmoid colon, which is divided using a GIA or thoracoabdominal (TA) automated stapling device (US Surgical or Ethicon Inc) (Fig. 4). The central pelvic tumor mass is removed en bloc with the rectosigmoid colon (Fig. 5), leaving a macroscopically tumor-free operative site (Fig. 6). Additional mobilization of the proximal sigmoid colon might be required to ensure a tension-free colorectal anastomosis.8 Intestinal continuity is reestablished, using a circular end-to-end anastomosis (CEEA) automated stapling device (Fig. 7) (US Surgical or Ethicon Inc). The security of the anastomosis is confirmed by inspecting the resection rings to ensure two complete “donuts” of colon and filling the pelvis with saline, while manually obstructing the proximal colon, and instilling 150 to 200 mL of air into the rectum to demonstrate an airtight anastomosis. In most cases, a closed suction drainage system is placed within the pelvis and removed when the output is less than 200 mL per 24-hour period. Nasogastric suction is not routinely used, and parenteral nutrition is administered at the discretion of the attending surgeon.
570
Bristow et al
Radical Oophorectomy for Ovarian Cancer
Figure 5. Radical oophorectomy. En bloc specimen, including uterus, adnexae, anterior pelvic peritoneal tumor, cul-de-sac tumor mass, and rectosigmoid colon. CT, cul-de-sac tumor; Ut, uterus.
Data abstraction
Individual subjects were identified prospectively at the time of surgery and data collected from inpatient and ambulatory medical records. Demographic data, operative findings, surgical procedures performed, amount of residual disease, and final pathology diagnosis were recorded for all patients. Followup information included significant postoperative morbidity within 30 days of
Figure 7. Radical oophorectomy. Intestinal continuity is reestablished using a circular end-to-end anastomosis (CEEA) automated stapling device.
J Am Coll Surg
Figure 6. Radical oophorectomy. The pelvis is macroscopically tumor-free after en bloc resection.
surgery, chemotherapeutic agents administered and duration of treatment, the date of clinical or radiographic progression of disease or recurrence, the date of last followup or death, and disease status at last followup. For overall survival analysis, event time distributions were estimated using the method of Kaplan and Meier.9 RESULTS Patient characteristics
Thirty-one consecutive patients underwent a primary maximal surgical effort at the JHMI that included Type II (n ⫽ 29) or Type III (n ⫽ 2) radical oophorectomy for epithelial ovarian cancer with confluent involvement of pelvic viscera during the study period. The FIGO stage distribution was as follows: stage IIIB (2 patients), stage IIIC (20 patients), stage IV (9 patients). Three patients had received three cycles of neoadjuvant chemotherapy with carboplatin and paclitaxel before their initial surgical procedure for FIGO stage IV serous ovarian carcinoma. One patient (stage IIIC disease) was treated with three cycles of carboplatin and paclitaxel chemotherapy after a suboptimal attempt at primary cytoreductive surgery at an outside institution before undergoing radical oophorectomy at JHMI. The remaining 27 patients all underwent surgical exploration before receiving chemotherapy. At the time of radical oopho-
Bristow et al
Vol. 197, No. 4, October 2003
Table 1. Clinicopathologic Patient Characteristics Characteristic
Median age, y (range) Serum CA125, U/mL Median Mean Range Performance status* 0 1 2 3 Ascites volume, mL Median Mean Range Histology, n (%) Serous Mixed serous and endometrioid Mucinous Clear cell Tumor grade, n (%) 1 2 3
Data
63 (40 to 85) 1,122 2,524 74 to 17,935 12 (38.7%) 14 (45.2%) 5 (16.1%)
2,700 2,494 100 to 7,000 26 (83.9) 2 (6.5) 2 (6.5) 1 (3.2) 2 (6.5) 8 (25.8) 21 (67.7)
*Gynecologic Oncology Group scoring.
rectomy, 22 of 31 patients (71%) had ascites volume greater than or equal to 1,000 mL. Additional clinicopathologic patient characteristics are shown in Table 1. Surgical results and postoperative treatment
Radical oophorectomy included hysterectomy in 21 patients, trachelectomy (after previous supracervical hysterectomy) in 2 patients, and uni- or bilateral salpingooophorectomy only (after earlier total hysterectomy) in 8 patients. In all cases, rectosigmoid colon resections were completed below the peritoneal reflection and intestinal continuity reestablished using a CEEA stapler (25 mm to 33 mm). Additional surgical procedures are shown in Table 2. Twenty-seven patients (87.1%) were left with residual disease measuring 1 cm or more in maximal diameter. Of these, 19 patients underwent complete cytoreduction of all visible disease, representing 61.3% of the total study group and 70.4% of those with optimal cytoreduction. Operative time ranged from 165 to 330 minutes, with a median time of 240 minutes (mean 235.3 minutes). The median estimated blood loss was 700 mL (mean 821 mL, range 300 to 2,900 mL). Estimated blood loss
Radical Oophorectomy for Ovarian Cancer
571
Table 2. Surgical Procedures Performed in 31 Patients Undergoing Radical Oophorectomy Procedure
n
%
Total omentectomy Infracolic omentectomy Retroperitoneal lymph node sampling/excision Peritoneal tumor implant excision or ablation Ileocecal or small bowel resection Diaphragm stripping or resection Right or left hemicolectomy Subtotal transverse colectomy† Splenectomy Liver resection Cholecystectomy
28 3
90.3 9.7
28
90.3*
23
74.2
9
29.0
7 3 2 2 1 1
22.6 9.7 6.5 6.5 3.2 3.2
*Retroperitoneal lymph nodes contained metastatic disease in 17 of 28 patients (60.7%). † En bloc with omentectomy.
in excess of 1,000 mL occurred in seven patients (22.6%). Nine patients (29.0%) required intraoperative or postoperative transfusion of packed red blood cells (range 1 to 5 U, median 2 U, mean 2.6 U). Twenty patients (64.5%) were transferred directly to the surgical intensive care unit immediately postoperatively, for a median stay of 2 days (range 1 to 12 days). The median time to tolerating a regular diet was 6 days (range 3 to 14 days). Thirteen patients received parenteral nutrition postoperatively (median 8 days, range 6 to 14 days). There were no perioperative deaths, but four patients (12.9%) experienced significant life-threatening postoperative complications (pulmonary embolism, sepsis, upper gastrointestinal hemorrhage). Minor postoperative morbidity (incisional cellulitis, urinary tract infection, ileus more than 7 days) occurred in 11 patients (35.5%). There were no instances of prolonged (beyond 7 postoperative days) bladder dysfunction. One patient (3.2%) experienced a postoperative breakdown of the colorectal anastomosis on postoperative day 5 and required reexploration, with diverting colostomy. This was the only patient in this series requiring reoperation. The median length of hospital stay was 11 days (range 4 to 23 days), and 22 patients received chemotherapy before discharge. Followup and survival analysis
Postoperatively, one patient declined further treatment and was lost to followup 43 days after surgery. Of the 30
572
Bristow et al
Radical Oophorectomy for Ovarian Cancer
J Am Coll Surg
their first site of progression, representing 16.7% of the 30 evaluable patients.
Figure 8. Overall survival of 30 patients undergoing radical oophorectomy for International Federation of Gynecology and Obstetrics stage III (n ⫽ 21) or stage IV (n ⫽ 9) epithelial ovarian cancer.
patients receiving chemotherapy, carboplatin plus paclitaxel was administered to 24; 5 patients received cisplatin plus paclitaxel plus topotecan as part of a cooperative group investigational study protocol (GOG protocol #9602). One patient received single agent carboplatin. All patients, including those treated initially with neoadjuvant chemotherapy, completed a minimum of six cycles after radical oophorectomy. Fifteen patients underwent second-look surgery (14 laparoscopy, 1 laparotomy). Seven of these patients (46.6%) had persistent disease detected at second-look surgery (six microscopically positive, one macroscopically positive). No patient was found to have persistent pelvic disease at the time of second-look surgery. Thirty patients were available for survival analysis. The median followup time for surviving patients was 21.7 months (range 7.2 months to 57.4 months). At last followup, 6 patients (20%) were dead of disease, 10 patients (33.3%) were alive with disease, and 14 patients (46.7%) were alive with no evidence of disease. The median overall survival time for all patients was 39.5 months (Fig. 8). At last followup, 15 patients (48.4%) had experienced clinical recurrence or progression of disease. Five patients had a pelvic component of disease as
DISCUSSION For patients with FIGO stage III/IV ovarian cancer, the amount of residual disease before initiating cytotoxic chemotherapy has been shown to be a powerful and consistent determinant of overall survival.3,10-17 Although the contribution of tumor biology to this observation cannot be overlooked, primary cytoreductive surgery is the accepted standard initial therapy for women with advanced epithelial ovarian cancer. The literature has consistently demonstrated that patients with no visible or only small volume residual tumor experience a two- to three-fold extension in median survival time compared with those left with bulky residual tumor.17-20 Contemporary 5-year survival rates now approach 50% for patients with optimally cytoreduced stage III/IV disease.17-19,21 Not uncommonly, advanced ovarian cancer will present with confluent disease extending to or encasing the reproductive organs and other pelvic viscera, which might lead to an abbreviated debulking procedure or abandonment of primary surgery altogether.15,22 In some series, as many as 47% of patients with advanced ovarian cancer might be left with suboptimal large-volume residual pelvic disease.23-25 Given the poor prognosis in such cases, every effort should be made to resect even the most locally advanced ovarian cancer when upper abdominal disease is not prohibitive. Radical extirpative procedures for locally advanced ovarian cancer have evolved since 1965, when Barber and Brunschwig26 first reported 22 patients undergoing pelvic exenteration. In this series, the postoperative mortality rate was 23% and there were only two longterm survivors. In 1968 and 1973, Hudson and Chir5,6 published two reports describing a technique they termed “radical oophorectomy,” specifically designed for the intact removal of a fixed ovarian tumor en bloc with attached peritoneum and surrounding structures. These authors observed that ovarian cancer tumor growth usually respects peritoneal demarcations, and they advocated a retroperitoneal approach, using the “false capsule” of the ovarian tumor within the pouch of Douglas to effect en bloc excision. During the past 30 years, varying terminology has been used to describe modifications of this procedure, including: en bloc rectosigmoid colectomy,20,27-31 reverse hysterocolposigmoidectomy,32 complete parietal and visceral peritonectomy,8
Vol. 197, No. 4, October 2003
Bristow et al
en bloc pelvic peritoneal resection of the intrapelvic viscera,33 and modified posterior exenteration.34 Terminology aside, the cardinal feature of the radical oophorectomy procedure is the retroperitoneal approach to ovarian cancer encasing the pelvic viscera, using the tendency of epithelial ovarian cancer to respect peritoneal planes of demarcation to surgical advantage. In this fashion, the retroperitoneal spaces, uninvolved by extensive intraperitoneal tumor, can be used to develop the dissection in a centripetal fashion with maximum safety to surrounding vital structures. The type II/III procedures performed in the current series achieved complete clearance of macroscopic pan-pelvic disease in all cases and contributed significantly to an overall maximal cytoreductive effort. It should be noted that the observed rates of optimal and complete cytoreduction in the current series might not reflect the surgical outcomes for all patients with advanced ovarian cancer. Specifically, only patients in whom a reasonable attempt at resection of upper abdominal disease could be undertaken were submitted to radical oophorectomy. Although this necessarily introduces a component of selection bias, other authors have reported equivalent rates of successful cytoreduction using a similar technique.8,16,20,21,25,30,31-35 In these reports, overall optimal residual disease (ⱕ1 cm or 2 cm) was achievable in 74% to 100% of patients, with almost universal clearance of advanced pelvic tumor. The literature also suggests a reduced risk of pelvic failure (5.3% to 9.4%) after radical oophorectomy, which is consistent with our observation that just 16.7% of patients manifested a pelvic component at first recurrence.35,36 Given the high proportion of patients with stage IV disease, the observed median survival time of 39.5 months compares favorably with other contemporary reports. The current data also demonstrate that the extensive surgery often required to achieve such clinical outcomes is associated with a predictably high, but acceptable, risk of perioperative morbidity. Among patients undergoing a radical oophorectomy procedure, the reported estimated blood loss ranges from 800 mL to 2,900 mL, with 12% to 49% of patients experiencing significant postoperative complications.16,20,21,25,28,29,31-34 Nevertheless, operative mortality of radical oophorectomy can be limited to 1.5% to 3.1% using an intensive postoperative care program.20,29,30,34 Historically, there have been concerns about reestablishing intestinal continuity in patients undergoing a radical debulking operation for ovarian cancer requiring
Radical Oophorectomy for Ovarian Cancer
573
resection of the sigmoid colon. Such patients often have large-volume ascites, can be nutritionally compromised from the metabolic effects of an extensive tumor burden, or have evidence of early bowel obstruction. Temporary or permanent intestinal diversion was typically performed in 12% to 59% of such patients.28-30,32,34,35,37 In this series, primary colorectal anastomosis was accomplished in all cases using a CEEA automated stapling device with only one instance (3.2%) of anastomotic dehiscence requiring reoperation and colostomy despite the fact that 71% of patients had ascites volume of 1,000 mL or more. Using both stapled and hand-sewn anastomoses, other authors have reported anastomotic breakdown in 0% to 8% of patients undergoing sigmoid colectomy for ovarian cancer.20,27,29-32,37 Among patients with large-volume ascites (500 mL or more), the incidence of anastomotic dehiscence is just 2.1% to 3.1%.30,34 Taken together, these data suggest that protective intestinal diversion is unnecessary and that terminal excretory function can be preserved in the majority of patients undergoing radical oophorectomy. In summary, surgeons operating on women with ovarian cancer are frequently confronted with locally advanced central pelvic disease with contiguous extension to or encasement of other pelvic viscera. Smallvolume residual disease has been consistently associated with improved survival, and for the majority of patients a maximal attempt at tumor cytoreduction will be warranted. Radical oophorectomy is highly effective for achieving complete clearance of macroscopic pelvic disease and, with advances in surgical technique and perioperative care, can be performed with acceptable morbidity. On the other hand, patients with unresectable upper abdominal disease are unlikely to derive a significant survival benefit from a radical pelvic resection. In such cases, intestinal diversion alone is the more appropriate surgical option. Author contributions
Study conception and design: Bristow, del Carmen, Kaufman, Montz Acquisition of data: Bristow, del Carmen Analysis and interpretation of data: Bristow, del Carmen, Kaufman, Montz Drafting of manuscript: Bristow, del Carmen, Kaufman, Montz Critical revision: del Carmen, Kaufman, Montz Statistical expertise: Bristow
574
Bristow et al
Radical Oophorectomy for Ovarian Cancer
Obtaining funding: Bristow Supervision: Kaufman, Montz 21. REFERENCES 1. Jemal A, Thomas A, Murray T, Thun M. Cancer statistics 2002. CA Cancer J Clin 2002;52:23–47. 2. Announcements. FIGO stages—1988 revision. Gynecol Oncol 1989;35:125–127. 3. Hoskins WJ, Bundy BN, Thigpen JT, Omura GA. The influence of cytoreductive surgery on recurrence-free interval and survival in small-volume Stage III epithelial ovarian cancer: a Gynecologic Oncology Group study. Gynecol Oncol 1992;47:159–166. 4. Marsden DE, Friedlander M, Hacker NF. Current management of epithelial ovarian carcinoma: a review. Semin Surg Oncol 2000;19:11–19. 5. Hudson CN. A radical operation for fixed ovarian tumors. J Obstet Gynaecol Br Cwlth 1968;75:1155–1160. 6. Hudson CN, Chir M. Surgical treatment of ovarian cancer. Gynecol Oncol 1973;1:370–378. 7. Morrow CP, Curtin JP. Gynecologic cancer surgery, 1st ed. New York: Churchill Livingstone, Inc; 1996, 451–568. 8. Sugarbaker PH. Complete parietal and visceral peritonectomy of the pelvis for advanced primary and recurrent ovarian cancer. Cancer Treat Res 1996;81:75–87. 9. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;53:457–480. 10. Hacker NF, Berek JS, Lagasse LD, et al. Primary cytoreductive surgery for epithelial ovarian cancer. Obstet Gynecol 1983;61: 413–420. 11. Heintz APM, Hacker NF, Berek JS, et al. Cytoreductive surgery in ovarian carcinoma: feasibility and morbidity. Obstet Gynecol 1986;67:783–788. 12. Piver MS, Baker T. The potential for optimal (ⱕ2 cm) cytoreductive surgery in advanced ovarian carcinoma at a tertiary medical center: a prospective study. Gynecol Oncol 1986;24:1–8. 13. Piver MS, Lele SB, Marchetti DL, et al. The impact of aggressive debulking surgery and cisplatin-based chemotherapy on progression-free survival in stage III and IV ovarian carcinoma. J Clin Oncol 1988;6:983–989. 14. Bertelson K. Tumor reduction surgery and long-term survival in advanced ovarian cancer: a DACOVA study. Gynecol Oncol 1990;38:203–209. 15. Eisenkop SM, Spirtos NM, Montag TW, et al. The impact of subspecialty training on the management of advanced ovarian cancer. Gynecol Oncol 1992;47:203–209. 16. Guidozzi F, Ball JH. Extensive primary cytoreductive surgery for advanced epithelial ovarian cancer. Gynecol Oncol 1994;53: 326–330. 17. Eisenkop SM, Friedman RL, Wang HJ. Complete cytoreductive surgery is feasible and maximizes survival in patients with advanced epithelial ovarian cancer: a prospective study. Gynecol Oncol 1998;69:103–108. 18. Chi DS, Liao JB, Leon LF, et al. Identification of prognostic factors in advanced epithelial ovarian carcinoma. Gynecol Oncol 2001;82:532–537. 19. Dauplat J, LeBoue¨dec G, Pomel C, Scherer C. Cytoreductive surgery for advanced stages of ovarian cancer. Semin Surg Oncol 2000;19:42–48. 20. Clayton RD, Obemair A, Hammond IG, et al. The western
22.
23. 24.
25. 26. 27. 28. 29. 30.
31.
32. 33.
34. 35.
36.
37.
J Am Coll Surg
Australian experience of the use of en bloc resection of ovarian cancer with concomitant rectosigmoid colectomy. Gynecol Oncol 2002;84:53–57. Scarabelli C, Gallo A, Franceschi S, et al. Primary cytoreductive surgery with rectosigmoid colon resection for patients with advanced epithelial ovarian carcinoma. Cancer 2000;88:389–397. Devi KU, Banfa UD, Ahuja V, Ramesh C. Induction chemotherapy and interval debulking surgery in advanced epithelial ovarian cancer. Int J Gynecol Cancer 2002;12:521 (Abstract #OV8). Webb MJ. Cytoreduction in ovarian cancer: achievability and results. Baillieres Clin Obstet Gynaecol 1989;3:83–94. Makar AP, Baekelandt M, Trope´ CG, Kristensen GB. The prognostic significance of residual disease, FIGO substage, tumor histology and grade in patients with FIGO Stage III ovarian cancer. Gynecol Oncol 1995;56:175–180. Benedetti-Panici P, Maneschi F, Scambia G, et al. The pelvic retroperitoneal approach in the treatment of advanced ovarian carcinoma. Obstet Gynecol 1996;87:532–538. Barber HRK, Brunschwig A. Pelvic exenteration for locally advanced and recurrent ovarian cancer. Surgery 1965;58:935– 937. Sonnendecker EWW, Beale PG. Rectosigmoid resection without colostomy during primary cytoreductive surgery for ovarian carcinoma. Int Surg 1989;74:10–12. Berek JS, Hacker NF, Lagasse LD. Rectosigmoid colectomy and reanastomosis to facilitate resection of primary and recurrent gynecologic cancer. Obstet Gynecol 1984;64:715–720. Soper JT, Couchman G, Berchuk A, Clarke-Pearson D. The role of partial sigmoid colectomy for debulking epithelial ovarian carcinoma. Gynecol Oncol 1991;41:239–244. Obermair A, Hagenauer S, Tamandl D, et al. Safety and efficacy of low anterior en bloc resection as part of cytoreductive surgery for patients with ovarian cancer. Gynecol Oncol 2001;83:115–120. Bridges JE, Leung Y, Hammond IG, McCartney AJ. En bloc resection of epithelial ovarian tumors with concomitant rectosigmoid colectomy: the KEMH experience. Int J Gynecol Cancer 1993;3:199–202. Barnes W, Johnson J, Waggoner S, et al. Reverse hysterocolposigmoidectomy (RHCS) for resection of panpelvic tumors. Gynecol Oncol 1991;42:151–155. Sainz de la Cuesta R, Goodman A, Halverson SS, Fuller AF. En bloc pelvic peritoneal resection of the intraperitoneal pelvic viscera in patients with advanced epithelial ovarian cancer. Cancer J Sci Am 1996;2:152–157. Eisenkop SM, Nalick RH, Teng NNH. Modified posterior exenteration for ovarian cancer. Obstet Gynecol 1991;78:879– 885. Hertel H, Diebolder H, Herrmann J, et al. Is the decision for colorectal resection justified by histopathologic findings: a prospective study of 100 patients with advanced ovarian cancer. Gynecol Oncol 2001;83:481–484. Spirtos NM, Eisenkop SM, Schlaerth JB, Ballon SC. Secondlook laparotomy after modified posterior exenteration: patterns of persistence and recurrence in patients with stage III and stage IV ovarian cancer. Am J Obstet Gynecol 2000;182:1321–1327. Tamussino KF, Lim PC, Webb MJ, et al. Gastrointestinal surgery in patients with ovarian cancer. Gynecol Oncol 2001;80: 79–84.