Kenneth D. Hatch
Laparoscopy has been widely accepted for numerous operative procedures. Laparoscopy performed with the assistance of video monitors has become the preferred technique because the surgeon can view the operation in real time.
Laparoscopic procedures such as cholecystectomy were adapted by surgeons and accepted by the public because of an associated short hospital stay, quick recovery time, and rapid return to full activity. However, this procedure was incorporated into surgical practice before prospective trials could be established to evaluate its feasibility, morbidity, and costeffectiveness compared with the standard laparotomy. The same criticisms may be made in gynecology for such procedures as laparoscopically assisted vaginal hysterectomy, removal of adnexal masses, and management of endometriosis, which are widely performed by virtually thousands of gynecologists.
In gynecologic oncology, there has been a unique opportunity to study the use of operative laparoscopy in a prospective fashion because of the limited number of specialists performing the procedures, and the need to perform pelvic and/or paraaortic lymphadenectomy to stage several gynecologic malignancies. As surgeons adopt minimally invasive technologies, innovations in instrumentation from bipolar coagulation devices to large-scale robotic-assist devices continue to facilitate the use of laparoscopic approaches for increasingly complex surgical procedures.
Laparoscopic Pelvic and Paraaortic Lymphadenectomy
The performance of a pelvic and paraaortic lymphadenectomy, either a partial lymphadenectomy (lymph node sampling) or complete lymphadenectomy, is the key procedure for the staging of gynecologic malignancies.
In 1989, Dargent and Salvat (1) in France used the laparoscope to perform limited pelvic lymphadenectomy in women with cervical cancer. This was not widely accepted because of its limited access to the pelvic lymph nodes and the inability to evaluate the lymph nodes in the common iliac and paraaortic chains. In 1991, Childers and Surwit (2) described pelvic and paraaortic lymphadenectomy performed in conjunction with a laparoscopically assisted vaginal hysterectomy and bilateral salpingo-oophorectomy in two women with endometrial cancer. In 1992, Nezhat et al. (3) published a case of laparoscopic radical hysterectomy and pelvic and paraaortic lymphadenectomy, although the dissection went only 2 cm above the aortic bifurcation, an inadequate evaluation. These early publications were limited case reports that gave no information on morbidity, mortality, or complications.
Querleu et al. (4) performed transperitoneal laparoscopic pelvic lymphadenectomy on 39 patients with cervical cancer. Five patients had metastatic lymph nodes and were treated with radiation therapy. Thirty-two patients underwent abdominal radical hysterectomy and evaluation of the completeness of the laparoscopic lymphadenectomy. The sensitivity for node positivity by laparoscopy was 100%. However, the number of additional lymph nodes found at laparotomy was not stated.
Childers et al. (5) reported 59 patients with endometrial cancer who were staged laparoscopically, followed by vaginal hysterectomy and bilateral salpingo-oophorectomy. Of the 31 patients deemed candidates for staging based on criteria including high-grade or deep myometrial invasion, lymph node dissection was completed in 29 patients (obesity precluded it in two patients), for a feasibility rate of 93%. Three major and three minor complications were reported. The surgical complications were experienced early in the series and led to alternative techniques as the series progressed. The average hospital stay was 2.9 days, but the operative time, lymph node counts, and cost analysis were lacking.
These early series emphasized pelvic lymphadenectomy, but it remained necessary to do paraaortic lymphadenectomy for laparoscopy to be fully accepted as a technique to stage all gynecologic malignancies. In 1992, Childers et al. (6) reported their initial experience with pelvic and paraaortic lymphadenectomy extending from the duodenum to the bifurcation in 18 patients with cervical cancer. They subsequently summarized the Arizona experience in paraaortic lymphadenectomy through 1993 with a report of 61 women with cervical, endometrial, or ovarian cancer (7). In three patients (5%), obesity prevented the completion of the surgery, and in one patient (0.8%), adhesions were responsible for failure. Lymph node counts were available in 23 patients: For the right-sided dissection, there was an average lymph node count of three. The operating time for the six patients who underwent a bilateral paraaortic lymphadenectomy ranged from 25 to 70 minutes, depending on whether or not a unilateral or bilateral procedure was performed (these times included only the time to complete the lymphadenectomy). The hospital stay for the 33 patients undergoing laparoscopic lymphadenectomy was 1.3 days. There was one vena caval injury that required transfusion and laparotomy, a complication rate comparable with that of open surgery.
In 1995, Spirtos et al. (8) reported 40 patients who underwent bilateral partial paraaortic lymphadenectomy (sampling). Five laparotomies were performed: two to remove unsuspected metastases, two for control of hemorrhage, and one because of equipment failure. In two patients, the left-sided dissection was judged to be inadequate, which was an overall failure rate of 12.5%. An average of eight paraaortic lymph nodes were removed: four from the right side and four from the left side. Most of the patients also underwent a pelvic lymphadenectomy and hysterectomy. The mean operative time was 3 hours, 13 minutes, and the average hospital stay was 2.9 days.
In the early series (4,5,6,9), laparotomy was used to confirm the accuracy of the lymphadenectomy, and in each report, all positive lymph nodes were identified.
Possover et al. (10) reported 84 patients who underwent laparoscopic pelvic and paraaortic lymphadenectomy for cervical cancer. The surgeon classified the lymph nodes as positive or negative by visualization. The sensitivity and specificity of visualization was 92.3%. When frozen-section analysis was combined with laparoscopic assessment, 100% of the positive lymph nodes were identified. In 13 of the 84 patients (15.5%), the treatment plan was altered during surgery based on these findings.
Possover et al. (11) analyzed videotapes of 112 paraaortic lymphadenectomies and detailed the ventral tributaries of the infrarenal vena cava (Fig. 21.1). They divided the vena cava into three levels based on the distribution of venous tributaries. This is a significant contribution to anatomic knowledge and is an important guide for beginning laparoscopic surgeons.
A perforator of the inferior vena cava at the level of the bifurcation of the aorta is shown in Fig. 21.1A. A diagram of the most common sites where perforators are encountered during a paraaortic lymphadenectomy is shown in Fig. 21.1B.
Figure 21.1 Perforators of the vena cava. A: A vena cava perforator at the level of the bifurcation of the aorta. B. Diagram of the most common sites where perforators are encountered during the performance of a paraaortic lymphadenectomy. The figure shows the anatomic distribution of 237 venous tributaries in 112 patients undergoing laparoscopic lymphadenectomy according to different levels of the inferior vena cava. (From Possover M, Plaul K, Krause N, Schneider A. Left-sided laparoscopic para-aortic lymphadenectomy: anatomy of the ventral tributaries of the infrarenal vena cava. Am J Obstet Gynecol 1998;179:1295-1297, with permission.)
Multiple recent studies continue to report the adequacy and safety of laparoscopic pelvic and paraaortic node dissections in gynecologic cancer (12,13,14,15,16,17,18,19). In one of the largest series to date, Koehler et al. reported on 650 patients undergoing laparoscopic transperitoneal pelvic (n = 499) or paraaortic (n = 468) (combined pelvic and paraarotic n= 362) lymphadenectomies (19). The mean number of pelvic lymph nodes removed from 1994 to 2003 remained fairly constant (16.9-21.9). However, the mean number of paraaortic lymph nodes increased from 5.5 in 1994 to 18.5 in 2003, reflecting improvements in technique and extensive training. Intraoperative complications (bowel or vessel injury) occurred in 2.9% of patients while 5.8% had postoperative complications for an overall complication rate of 8.7%. The authors reported that no major intraoperative complications were encountered during the last five years of the study.
Querleu et al. subsequently reported on their experience with transperitoneal and extraperitoneal lymph node dissections in 1,000 gynecologic cancer patients (18). This study included 777 pelvic (757 transperitoneal, 20 extraperitoneal) and 415 aortic lymphadenectomies (155 transperitoneal, 260 extraperitoneal) in patients with early cervical carcinoma (n = 456); advanced cervical carcinoma (n = 219); vaginal carcinoma (n = 4); endometrial carcinoma (n = 182); and ovarian carcinoma (n = 139). The mean number of pelvic lymph nodes removed was 18 via a transperitoneal approach and the mean number of paraaortic lymph nodes removed was 17 via a transperitoneal approach versus 21 via an extraperitoneal approach. The authors reported an increase in the number of lymph nodes removed with increasing experience, yielding an average of 24 pelvic and 22 aortic lymph nodes in 2003. Intraoperative complications occurred in 2% of patients including injury to vascular structures (1.1%); bowel (0.3%); ureter (0.3%); and nerves (0.3%). Five patients underwent conversion to laparotomy for completion of the lymph node dissection, secondary to fixed nodes or extensive adhesions. Conversion to laparotomy occurred in an additional two patients secondary to bowel or ureteric injury. Five patients required a second surgical intervention due to postoperative complications, most commonly bowel obstruction (n = 4).
These studies have demonstrated the ability of laparoscopic surgeons to perform pelvic and paraaortic lymphadenectomy. The American Medical Association Physicians Current Procedure Terminology (CPT 2007) lists a total of four laparoscopic lymph node dissection procedures, including total pelvic lymphadenectomy and paraaortic lymph node sampling. Laparoscopic surgery has been used by many oncologic surgeons, and has been applied to nearly every disease site in gynecologic oncology.
Indications for Laparoscopic Surgery
Most women with endometrial cancer present with disease confined to the uterus. The treatment consists of total hysterectomy, bilateral salpingo-oophorectomy, and surgical staging, which includes peritoneal washings, inspection of the abdomen, and retroperitoneal lymph node sampling. Surgical staging with operative laparoscopy followed by vaginal hysterectomy or laparoscopic total hysterectomy has been proposed as an alternative to laparotomy (2,12,13,14,15,20,21).
Childers et al. (5,6) reported two patients in 1992 who underwent laparoscopic staging of the retroperitoneal nodes followed by vaginal hysterectomy and bilateral salpingo-oophorectomy, and they presented the first large series in 1993 (7). Laparoscopic staging was performed successfully in 93% of the patients, with obesity noted as a limiting factor. Two patients had complications related to the hysterectomy: One had a transected ureter caused by the endoscopic stapler, and one had a cystotomy.
Spirtos et al. (20) reported 13 patients who underwent laparoscopic staging and hysterectomy and compared them with 17 patients who underwent laparotomy. The laparotomy group required significantly longer hospitalization, (6.3 vs. 2.4 days, p <0.001), incurred higher overall hospital costs ($19,158 vs. $13,988, p <0.05), and took longer to return to normal activity (5.3 weeks vs. 2.4 weeks, p <0.0001). The patients having laparotomy were significantly more obese and had a higher body mass index (BMI) (30.2 vs. 24.2).
The effect of surgical experience has been demonstrated by Melendez et al. (21). In the first 100 patients with endometrial cancer, the operative time for staging decreased from a mean of 196 minutes for the first 25 patients to 128 minutes for the last 25 patients. Hospital stay decreased from 3.2 days to 1.8 days. The decrease in operative time and hospital stay, coupled with the diminished use of expensive, disposable instruments, has led to a significant cost savings for laparoscopy. More important are the social benefits to the individual patient.
Table 21.1 Recurrence Rates for Laparoscopic Surgery versus Laparotomy For Endometrial Cancer
Subsequent publications have continued to show a decrease in operative time, hospital stay, and total cost for laparoscopic treatment of endometrial cancer (15,22,23).
Women with endometrial cancer are often obese with BMI greater than 35 (24). This has been thought to be a limiting factor in using laparoscopy to stage and treat endometrial cancer. As surgical skills have grown, laparoscopy has been used successfully in these women. Holub et al. (25) have completed staging and hysterectomies successfully in 94.4% of 33 patients with BMIs of 30 to 40. Eltabbakh et al. (26) have completed staging in 88% of 42 women with BMIs of 28 to 60. In both studies, the benefits of shorter hospital stay with faster recovery were verified. However, many retrospective studies comparing laparoscopy and laparotomy have reported a significantly lower BMI in patients undergoing laparoscopic management, reflecting a selection bias favoring open procedures in obese patients. Results of further prospective studies are awaited to assess the impact of BMI and obesity in laparoscopic versus open surgical approaches.
Long-term survival has been reported in several papers (Table 21.1; 15,23,27,28,29,30,31,32,33,34). In these series, more than 900 patients have been studied for a median of 16 to 53 months. Inclusion criteria were heterogeneous, with some authors reporting only on outcomes of patients with stage I and II disease. The recurrence rate following laparoscopic management ranged from 0% to 9%. When compared with historical controls undergoing laparotomy in these papers and adjusting for factors such as stage, grade, age, and weight, there was no difference in survival. While these studies yield promising data, comparison of operative morbidity and shortand long-term outcomes between laparotomy and laparoscopy requires an adequately powered, randomized clinical study.
To answer these questions, The Gynecologic Oncology Group (GOG) (35) conducted the LAP 2 trial, a large, prospective, randomized trial designed to determine equivalency in early-stage endometrial cancer outcomes in laparoscopically assisted vaginal hysterectomy and bilateral salpingo-oophorectomy with surgical staging, when compared with traditional open surgery. This study enrolled 2,616 patients, completing accrual in 2005. There were 1,696 patients randomized to laparoscopy and 920 to laparotomy in a two-to-one randomization ratio. Preliminary data were presented at the Annual Meeting of the Society of Gynecologic Oncologists in 2006. Length of stay was shorter in the laparoscopic arm (median 3 days, range 0-95) versus laparotomy arm (median 4 days, range 1-49). Operative time was increased with laparoscopic procedures (3.3 hours versus 2.2 hours) and approximately 23% of patients randomized to laparoscopy required laparotomy to complete staging.
Results of this important study will help answer many questions regarding feasibility, appropriate patient selection, and short- and long-term oncologic outcomes of laparoscopy in the management of endometrial cancer.
The concept of sentinel node removal has been studied in endometrial cancer in several small studies. Techniques utilized for the detection of sentinel lymph nodes include injection of blue dye, injection of a radiocolloid, or both. To date, studies in endometrial cancer have focused on injection of the tracer into the uterine corpus (subserosal or myometrial); the cervix; the endometrium using hysteroscopy; or combined sites (36). The sentinel lymph node detection rates using injection of the subserosal myometrium alone ranged from 0-92% (37,38,39,40,41,42). Altgassen reported the highest detection rate (92%) utilizing eight injection sites in contrast to the 1 to 3 sites reported by several other authors. Similarly, Li et al. (39) reported a 75% detection rate using three subserosal myometrial sites and two subserosal isthmic sites. Reported detection rates using cervical injection alone ranged from 80% to100% (43,44,45,46). Holub et al. reported a detection rate using cervical and subserosal myometrial injections of 84% (47). Hysteroscopic injection of the endometrium has yielded detection rates of 50% to 100 % (48,49,50,51). While the possibility of laparoscopic assessment of sentinel lymph nodes and targeted sampling is interesting, sentinel lymph node detection in endometrial cancer remains investigational.
The use of laparoscopy in the treatment of cervical cancer was initially limited by the fact that there was no apparent advantage to laparoscopic lymphadenectomy because the standard operation for the primary cervical tumor was radical abdominal hysterectomy.
Dargent (52) first suggested that laparoscopic pelvic lymphadenectomy could be followed by a Schauta radical vaginal hysterectomy and has published long-term results, reporting a 95.5% 3-year survival rate in 51 patients with negative pelvic lymph nodes. Querleu (53) reported eight patients and demonstrated an average blood loss of less than 300 mL, an average hospital stay of 4.2 days, and decreased pain from the elimination of an abdominal incision.
Hatch et al. (54) reported 37 patients treated by laparoscopic pelvic and paraaortic lymphadenectomy followed by radical vaginal hysterectomy. The mean operative time was 225 minutes, the mean blood loss was 525 mL, and the average hospital stay was 3 days. Blood transfusion was required in 11% of the patients, compared with the range of 35% to 95% reported in the literature for radical abdominal hysterectomy. Complications occurred early in the series and included two cystostomies repaired at surgery without an increase in hospital stay. In two patients (5.4%), ureterovaginal fistulae developed that were treated by ureteral stents. These were removed 6 weeks later without further operative intervention.
Schneider and colleagues (55) reported 33 patients in whom bipolar techniques were used for lymphadenectomy and to transect the cardinal ligaments and uterine vessels. Hysterectomy was completed by the Schauta-Stoeckel technique. There were five (15%) intraoperative injuries managed successfully without subsequent sequelae. Four patients required transfusion. Numerous retrospective studies comparing laparoscopically assisted radical hysterectomy versus radical abdominal hysterectomy have reported increased operative time but decreased blood loss, decreased transfusion rates, and hospital length of stay in patients undergoing the minimally invasive procedures (56,57,58,59,60).
Studies have shown that the complication rates go down as the operator's experience increases (61,62). Long-term survival has been reported by Hertel et al. (14) for 200 patients, with a mean follow-up of 40 months. The projected 5-year survival was 83%. For the 100 patients who were stage I, lymphovascular space-negative and lymph node-negative, the survival was 98%.
Laparoscopic Radical Hysterectomy
Although most initial reports in the literature detailed some form of laparoscopically assisted radical vaginal hysterectomy, there are increasing reports of laparoscopic radical hysterectomy. Spirtos et al. (13) reported laparoscopic radical hysterectomy (type III) with aortic and pelvic lymphadenectomy in 78 patients. The average operative time was 205 minutes, length of hospitalization was 3.2 days, and blood loss was 225 mL; one transfusion was necessary. There were acceptable intraoperative and postoperative complications. With a minimum of 3 years followup, the disease-free survival was 95%. Since then, numerous authors have reported similar findings (41,63,64,65,66,67). In one of the largest series, Puntambekar (65) reported on 248 patients with early stage cervical cancer, noting a median operative time of 292 minutes, median number of resected pelvic lymph nodes of 18, median blood loss of 165 mL, and a median length of stay of 3 days. Other studies generally have reported longer operating times ranging from 196 to 344 minutes (41,60,67,68).
The issue of blood loss and transfusion has become very important to patients and surgeons since the identification of the human immunodeficiency virus and other blood-borne pathogens. Every report on laparoscopic lymphadenectomy and radical hysterectomy has noted a significant decrease in blood loss and transfusion rates. Other societal advantages are the decreased hospital stay and rapid return to normal function, even with radical surgery.
Laparoscopic Nerve-Sparing Radical Hysterectomy
Preservation of the superior hypogastric nerve plexus, the hypogastric nerve and the inferior hypogastric plexus are important for function of the bladder and rectum. The superior hypogastric plexus is composed of sympathetic nerves that allow the bladder to store urine. If this is damaged, the bladder will have a small volume and high pressure under parasympathetic control. The inferior hypogastric plexus is composed of parasympathetic nerves that initiate urination. If the inferior hypogastric plexus is damaged, the patient will have lack of sensation and be unable to initiate urination. The hypogastric nerve connects the two plexes. If it is severed, the patient will have a mixed pattern, consisting of an initial small volume, high-pressure phase followed by a hypotonic high-residual state. This may lead to the need for chronic self-catheterization.
Damage to the superior hypogastric plexus can occur at the time of paraaortic node dissection or presacral node dissection. Injury to the hypogastric nerve may occur when clamping the cardinal or uterosacral ligaments. Injury to the inferior hypogastric plexus may occur with lateral dissection of the cardinal ligaments, dissection of the posterior vesicouterine ligament, or with removal of paravaginal tissue.
Recent description of the anatomy of these nerves and the surrounding blood vessels has established the role of nerve-sparing techniques in performing radical hysterectomy (69). The laparoscope is an excellent technique to identify and preserve the nerves, because it magnifies small vessels and nerves so that the surgeon can more easily identify them.
The technique performed at the University of Arizona is described below (Figs. 21.2, 21.3, 21.4, 21.5, 21.6, 21.7 and 21.8).
Figure 21.2 The superior hypogastric nerve plexus is preserved during the paraaortic node dissection.
Figure 21.3 Traction on the peritoneum overlying the bifurcation of the aorta helps identify the course of the hypogastric nerve 2 to 3 centimeters medial to the ureter at the pelvic brim.
Figure 21.4 The hypogastric nerve continues into the pelvis along the ureter with branches to the rectum, uterus and inferior hypogastric plexus.
Figure 21.5 The hypogastric nerve is 1 to 2 centimeters dorsal to the uterine artery and can be preserved if mass ligature of the cardinal ligament is avoided.
Figure 21.6 The vessels and connective tissue medial to the ureter is dissected so that the ureter can be retracted laterally out of the cardinal ligament tunnel.
Figure 21.7 The ureter is now retracted laterally with the hypogastric nerve. The cardinal ligament can now be divided.
Figure 21.8 The anterior vesicouterine vessels and connective tissue are transected.
The operation can be completed by the vaginal route or with further dissection through the laparoscope. The vaginal route allows more precise removal of the vaginal margin.
The urethral catheter is left in place for 48 hours. The patient is then allowed to void and a postvoid residual is obtained. If she had good sensation of bladder fullness and a residual urine amount less than 60 cc, the catheter is left out. If she is unable to void, has no sensation of filling, or if the residual urine is over 60 cc, the catheter is left in place for 7 days. The nerve sparing operation has been performed in a total of 33 patients. Twenty-one of these patients had a laparoscopic node dissection and radical vaginal trachelectomy. Twelve had a laparoscopic node dissection and radical hysterectomy completed by laparoscopy or by vaginal assistance. Eight of the patients were unable to void with residual of less than 60 cc at 7 days. All of the patients were able to void at 21 days.
Laparoscopic staging of cervical cancer before treatment planning has been proposed (15,70,71). Vidaurreta et al. (69) staged 91 patients, stages IIB, IIIA, IIIB, and IVA. Computed tomography (CT) was performed in 49 patients, with 38 read as normal and 11 as positive. Histologic evaluation revealed metastases in 18 of the 38 (47.4%) patients with negative scans, and no metastases were found in 5 of the 11 (45.5%) with positive scans. Hertel et al. (71) compared laparoscopic surgical staging with magnetic resonance imaging (MRI) and CT scan in 101 patients, 91 of whom had a CT scan, 67 of whom had an MRI scan, and 49 of whom had both. False-positive or false-negative results were found in 22% of patients. Ten patients had false-positive paraaortic nodal metastases.
The initial studies of sentinel node detection showed sensitivity, negative predictive value, and accuracy of 100% (72,73,74). Subsequent studies have used both blue dye and Technetium 99m detection methods with varying success (74,75,76,77,78,79,80,81,82,83,84,85,86) (Table 21.2).
Radical Vaginal Trachelectomy with Laparoscopic Lymphadenectomy
Radical hysterectomy, with or without adjuvant therapy for patients with early cervical cancer, is associated with high cure rates. However, in young patients desiring fertility, alternative surgical options have been increasingly explored.
In 1994, Dargent et al. (87) first presented a series of 28 patients who underwent laparoscopic pelvic lymphadenectomy followed by radical vaginal trachelectomy. After a median follow-up of 36 months, there was only one recurrence in the paraaortic nodes of a 27-year-old patient with stage IB adenocarcinoma. The pelvic lymph nodes had been negative, and the margins were free. Among the eight patients who attempted pregnancy, three had cesarean section at 36 weeks gestation, and three had a spontaneous abortion.
The second report on radical vaginal trachelectomy was published in 1998 by Roy and Plante (88). Thirty patients underwent laparoscopic pelvic lymphadenectomy and radical vaginal trachelectomy; only six women had attempted pregnancy at the time of reporting, and four had healthy infants delivered by cesarean section.
Since these initial reports, several authors have now reported their experience with radical trachelectomy (Table 21.3; 89,90,91,92,93,94,95,96,97,98). The indications commonly used are:
Table 21.2 Sentinel Node Detection for Patients with Cervical Cancer (Series with ≥30 Patients)
Risk factors for recurrence include lesion size greater than 2 cm, depth of invasion greater than 1 cm, and lymph-vascular space invasion (76). However, some authors suggest that conservative fertility-sparing surgery may be appropriate for select patients with larger lesions that are clearly exophytic (99).
A recent systematic review of 504 women undergoing radical trachelectomy described the gestational outcomes of 200 pregnancies resulting in 133 third-trimester deliveries. The first trimester abortion rate was 19%, which is similar to the general population. The second trimester spontaneous abortion rate was 9.5% (19 of 200) when defined as <24 weeks. Overall, 84 of 200 pregnancies (42%) resulted in term deliveries of viable infants, but 25% (49 of 200) of total pregnancies or 37% of third-term deliveries were preterm. Moreover, 13 of 133 (9.8%) patients delivered between 24 and 28 weeks and 14 (10.5%) delivered between 24 and 34 weeks (100). These pregnancies should be managed by maternal fetal medicine specialists.
Laparoscopy has been used for several decades to manage adnexal masses and as a second-look procedure to avoid laparotomy in patients with persistent disease after primary chemotherapy. More recently, it has been reported to be useful for staging apparently early cancer of the ovary.
Table 21.3 Radical Trachelectomy for Fertility Preservation in Patients with Early-Stage Cervical Cancer: Oncologic Outcomes
The ability to perform retroperitoneal evaluation has seen it advocated again for second-look procedures.
Evaluation of the Suspicious Adnexal Mass
Laparotomy is accepted as the standard of care for management of the suspicious adnexal mass. However, it is possible to mismanage adnexal masses regardless of whether laparotomy or laparoscopy is used.
The incidence with which an unexpected malignancy is encountered when managing an adnexal mass is reported to be between 0.4% and 2.9% (101,102,103). Childers et al. (104) and Canis et al. (105) used laparoscopy for management of suspicious adnexal masses and reported malignancy rates of 14% and 15%, respectively. More than 80% of the masses were managed by laparoscopy. All of the malignancies were properly diagnosed and treated, including 13 staged by laparoscopy. A frozen section should be obtained so that surgical staging and appropriate treatment are not delayed.
Staging requires an infracolic omentectomy, peritoneal washings, multiple biopsies from the peritoneal surfaces and hemidiaphragms, and pelvic and paraaortic lymph node biopsies. Laparoscopic omentectomy has been described using a stapling technique (106) but has been simplified further with the advent of laparoscopic bipolar vessel sealing devices which incorporate both bipolar cautery and cutting functions.
Several investigators have reported their experiences with staging of early ovarian, fallopian tube, or primary peritoneal cancers. In 1994, Querleu and LeBlanc (107) described the first adequate laparoscopic surgical staging for ovarian carcinoma in eight patients undergoing pelvic and paraaortic lymph node sampling up to the level of the renal veins. An average of nine nodes (range 6-17) were removed, with an average operative time of 111 minutes, postoperative stay of 2.8 days, and blood loss of less than 300 mL. None of the lymph nodes were positive.
In 1995, Childers et al. (108) reported 14 patients undergoing staging for presumed early ovarian cancer. Metastatic disease was discovered in eight patients (57%) and the appropriate treatment instituted. Subsequent series by Pomel (n = 10) (109); Tozzi (n = 24) (110); Chi (n = 20) (111); Spirtos (n = 73) (112); and Leblanc (n = 44; 36 epithelial, 8 germ cell or granulosa cell) (113) have confirmed similar feasibility.
Results of these small series are encouraging and support the development of studies with larger sample sizes and long-term follow-up. However, the low incidence of early stage disease underscores the difficulty of clinical trial development in this patient group. At this time, laparoscopy for the staging of ovarian cancer remains investigational. It may be considered for patients with apparent early stage disease at presentation, for staging of unstaged patients, or for patients who are candidates for fertility-sparing oophorectomy and staging alone.
The two major concerns over the use of laparoscopy for adnexal masses are (i) delay in diagnosis and thus treatment; and (ii) rupture of the adnexal mass that is subsequently found to be malignant, which converts the stage from a possible IA to IC. Studies on laparotomy show that if the tumor is removed and proper treatment instituted, rupture does not affect the outcome (114,115,116), but it is prudent to avoid rupture to minimize any theoretical increase in the risk. If the tumor is ruptured, and the treatment is delayed, the prognosis is worsened (117). Thus, the use of laparoscopy should be limited to suspicious masses that are small enough to be removed intact or utilizing endoscopic bags to allow for cyst aspiration without the leakage of cyst contents into the peritoneal cavity.
Laparoscopy was initially used before planned second-look laparotomy to identify residual disease and thus avoid the laparotomy. This strategy resulted in a reduction in the need for laparotomy in 50% of patients (118).
Improvements in laparoscopic equipment encouraged some investigators to perform the entire second-look procedure by laparoscopy. Childers et al. (108) reported 44 reassessment laparoscopies in 40 women. Twenty-four of the procedures were positive, including five that were only microscopically positive. Five patients (11%) had inadequate laparoscopies because of adhesions, and recurrent disease developed in all of them. Eight of the 20 patients (40%) who were negative later developed recurrent disease. All of these data were similar to those obtained with second-look laparotomy.
Abu-Rustum et al. (119) reported 31 women having second-look laparoscopy, and compared them with 70 patients who had laparotomy and 8 who had both. The rates of positivity were 54.8%, 61.4%, and 62.5%, respectively. The recurrence rates after a negative second look were 14.8% for laparoscopy versus 14.3% for laparotomy. Clough et al. (120) reported 20 patients who had laparoscopy followed by laparotomy at the same surgery, with a positive predictive value for laparoscopy of 86% (12 of 14 patients).
The effects of the CO2 pneumoperitoneum and laparoscopy on the long-term survival of women undergoing second-look operations have been reported by Abu-Rustum and associates (121). Over an 11-year period, 289 patients had positive second-look operations. There were 131 laparoscopies using CO2, 139 laparotomies, and 19 laparoscopies converted to laparotomy. The groups were controlled for age, stage, histology, grade, and size of disease found at second look. The median survival for patients who had laparoscopy was 41.1 months and for laparotomy 38.9 months (p = 0.742). Thus, the overall survival was independent of the surgical approach.
Second-look assessment has generally declined in practice, because approximately 50% of patients with negative second-look surgeries eventually recur (122). However, minimally invasive techniques continue to play a role in the management of patients with advanced ovarian cancer, including the placement of ports for intraperitoneal chemotherapy, if not performed at the initial surgery.
Complication rates of laparoscopy for malignant disease are higher than for benign disease (123). The rate depends on the type of case and the experience of the surgeon. Laparoscopic second looks have the highest rate of injury to bowel because of the adhesions from previous surgery. Vascular injuries from trocars or the dissection of lymph nodes can occur in any procedure.
Postoperative wound infection, ileus, and fever occur, but at lower rates than after laparotomy. Herniation of omentum or bowel into the trocar sites is a complication unique to laparoscopy. Boike et al. (124) reported 19 cases from 11 institutions. No patient had a hernia through a port smaller than 10 cm, and therefore it is recommended that all port sites greater than 10 mm be closed. Kadar et al. (125) reported a 0.17% rate of herniation among 3,560 laparoscopic operations.
Abdominal wall port-site implantations have been reported with nearly every tumor type, particularly ovarian cancer. However, in a review of 1,335 transperitoneal laparoscopies in 1,288 women with malignant disease, Abu-Rustum et al. reported that laparoscopy-related subcutaneous tumor implantation is rare, occurring in only 13 (0.97%) cases (126).
Patient preparation begins with a clear liquid diet the day before the surgical procedure. Evacuation of the bowel may be accomplished with a laxative or an oral gastrointestinal lavage solution. It is important for the bowel to be collapsed during the laparoscopic lymphadenectomy so that proper exposure can be obtained. This is particularly important if the patient is somewhat obese and paraaortic lymphadenectomy is planned.
The recommended technique of laparoscopy is as follows:
Figure 21.9 Patient position for laparoscopically assisted radical hysterectomy.
Figure 21.10 The position of the surgeons and placement of the trocars in the abdomen.
Figure 21.11 Completed right common iliac and paraaortic node dissection.
Figure 21.12 Partially completed right pelvic lymphadenectomy.
Detailed step-by-step images and details of the dissection can be found in the accompanying DVD. Video of the surgical procedure is available in LWW Laparoscopy for Gynecology and Oncology by Dr. Kenneth Hatch.
Patients are given liquids the day of surgery, and the diet is advanced rapidly. Early ambulation is encouraged. The patient's progress is usually rapid. Adynamic ileus is unusual after laparoscopic surgery, but any abdominal distention, worsening of pain, or vomiting must be taken seriously. Unsuspected bowel injuries manifest themselves by abdominal distention, pain, and free air in the peritoneal cavity. The CO2 should be absorbed within hours, so any free air in the abdomen is highly suspicious.
Traditional laparoscopy can be technically difficult in comparison with open procedures due to factors such as lack of depth perception, ergonomic difficulties with often counter-intuitive directional movements, tremor amplification, and relatively rigid instrumentation which lacks the flexibility of movement possible in open surgery. Even simple procedures such as suturing are associated with a much steeper learning curve, and can take longer to perform, hindering widespread adoption of minimally invasive techniques. More recently, robotic-assisted surgical devices have been developed to address these issues. It is hoped they will make minimally invasive techniques easier to learn, and will expand the scope of procedures that may be performed laparoscopically.
The DaVinci surgical system (Intuitive Surgical, Sunnyvale CA) is currently the only commercially available robotic surgical system that is FDA approved for gynecologic and other surgical procedures. It incorporates several components including a surgeon console, patient side cart, 3-D vision system, and proprietary EndoWrist instruments. The surgeon console is a remote unit that allows the seated surgeon to manipulate the robotic instruments while viewing a 3-D operative field. The next component is the patient-side cart,which provides three or four robotic arms, which in turn control the endoscope and two or three instrument arms. A major feature of the DaVinci system is the EndoWrist instrumentation, which provides seven degrees of motion, mimicking the full range of motion of a surgeon's hand and wrist. The final component is the in-vision system, which allows for three-dimensional imaging. The Vision System incorporates a high-resolution three-dimensional endoscope, which provides 3-D vision when the surgeon utilizes the stereoscopic viewer. An additional vision tower provides operative field visualization for the rest of the surgical team, although images are two dimensional.
Experience in gynecologic procedures is still limited, but increasing. Reynolds and Advincula reported their initial experience of 16 patients undergoing robotic-assisted laparoscopic-assisted vaginal hysterectomy (127). Oncologic indications included endometrial and ovarian carcinoma and cervical dysplasia. This series reported an operative time ranging from 270 to 600 minutes with a mean estimated blood loss (EBL) of 300 cc.
Several series have focused on the experience with hysterectomy for benign disease. Beste et al. reported on 11 patients undergoing robotic assisted total laparoscopic hysterectomy (128). Estimated blood loss ranged from 25 to 350 mL, with operative times ranging from 49 to 227 minutes. One patient required conversion to laparotomy because of bleeding at the level of the uterine arteries. Fiorentino et al. reported on 20 patients with a median EBL of 81 mL, median uterine weight of 98 gm, and median operative time of 200 minutes. Two patients were converted to laparotomy because of poor visualization, but 18 procedures were successful (129). Reporting their experience of 16 consecutive patients undergoing either complete laparoscopic hysterectomy or supracervical hysterectomy, Reynolds and Advincula similarly noted feasibility, with no conversions to laparotomy, a mean blood loss of 72.5 mL, and a mean operative time of 242 minutes (170 to 432 minutes). Thirteen of 16 patients required lysis of pelvic adhesions from prior pelvic surgeries (127).
More recently, the use of robotic radical hysterectomy in cervical cancer has been described. Kim reported on 10 patients with stage IA1-IB1 cervical cancer. The mean operative time was 207 minutes, mean docking time 26 minutes, mean EBL 355 mL, and mean pelvic lymph node count of 27.6. No conversions were required (130). Sert and Abeler compared operative results of seven consecutive patients undergoing robotic laparoscopic radical hysterectomy with eight previous laparoscopic procedures (131). Estimated blood loss (71 vs. 160 mL) and hospital stay (4 vs. 8 days) were decreased in the robotic vs. laparoscopic groups, while operative time (241 vs. 300 minutes) and lymph node count (13 vs. 15) were not significantly different.
In 2008, Magrina et al. described the prospective Mayo Clinic experience of 27 patients undergoing robotic radical hysterectomy (132). Patients were matched for age, BMI, type of malignancy, stage of disease, and type of radical hysterectomy. Patients who underwent robotic or laparoscopic surgery had decreased blood loss (133.1, 208.4 mL) and hospital stay (1.7, 2.4 days) when compared with laparotomy (443.6 mL, 3.6 days). However, operative time was shorter for open laparotomy (166.8 minutes) and robotic procedures (189.6 minutes) when compared with laparoscopy (220.4 minutes).
Challenges in robotic surgery include the lack of tactile sensation as well as the high cost of the system. However, while initial series are limited, these experiences suggest that robotics offer the same patient benefits afforded by traditional laparoscopy, while providing improved surgical dexterity and visualization. As with traditional laparoscopy, further prospective studies are required to determine appropriate indications, feasibility, and oncologic outcomes.
The skills to manage gynecologic malignancies by laparoscopic techniques are acquired through a commitment on the surgeon's part to learn the technique. It requires up-to-date equipment and a team familiar with the procedures. Hands-on experience in an animal laboratory and proctored learning in the operating suite are highly recommended. In the hands of experienced laparoscopic surgeons and with properly selected patients, laparoscopic surgery appears to result in shorter hospital stays, earlier return of function, and outcomes comparable with laparotomy. The results of prospective, randomized trials are awaited. Innovative technologies such as robotic-assisted surgical units may further expand the scope and use of minimally invasive techniques in the field of gynecologic oncology.