Laparoscopic Bilateral Pelvic and Paraaortic Lymph Node Sampling:
Table 1. Gynecologic Oncology Group
To date, reports describing laparoscopic lymph node sampling in the management of patients with gynecologic malignancies have yet to describe a method to sample left-sided aortic lymph nodes between the aortic bifurcation and the inferior mesenteric artery in a large series of patients. We herein submit our experience using evolving techniques that allow for excellent visualization and resection of both left and right aortic and pelvic lymph nodes in forty consecutive patients which conforms to the GOG guidelines.
From April 1, 1992 to August 31, 1993, forty patients with gynecologic malignancies underwent laparoscopy for the purpose of surgical staging. Twenty-three of twenty-seven patients with previously untreated endometrial cancer clinically limited to the endometrium underwent laparoscopically-assisted vaginal hysterectomy, bilateral salpingo-oophorectomy, in addition to an aortic and pelvic lymph node dissection. Seven of eight patients with endometrial cancer having previously undergone hysterectomy and bilateral salpingo-oophorectomy underwent a laparoscopic aortic and pelvic lymph node dissection. Four patients with ovarian cancer and one with fallopian tube cancer had a complete lymph node dissection as part of a laparoscopic procedure performed to complete surgical staging. In five patients, the intended procedure was not completed laparoscopically.
In addition to a routine preoperative evaluation, a Quetelet Index (weight in kilograms/height in meters squared) was calculated as a means of assessing the patients associated with laparoscopy. (2) Patients with an index of greater than 30 were excluded. All patients were brought to the operating room, placed on a Skytron Electric or Ansco Model 1080 table and given a general anesthetic without the use of nitrous oxide. Shoulder braces were locked into place with the arms of the patient tucked to the sides. The patient's legs were supported with Allen Universal stirrups only if a hysterectomy was to be performed. Otherwise, patients remained in a supine position. TED hose and sequential compression devices were used on all patients for thrombosis prophylaxis.
A Verres needle was used to access the peritoneal cavity at the level of the umbilicus and carbon dioxide is infused to obtain a sufficient pneumoperitoneum. The current configuration of disposable trocars is as follows : A ten millimeter through the umbilicus, a twelve millimeter in the midline just above the symphysis, two five millimeter trocars at the level of the superior iliac crest just lateral to the inferior epigastric vessels and, finally, a ten millimeter in the midline approximately eight centimeters above the umbilicus.
A 3-Chip camera is attached to a ten millimeter laparoscope and passed through the umbilical port. Either Endoshears (United States Surgical Corp., USA) in three patients or the Argon beam coagulator (ABC) (Birtcher, USA) endoscopic probe in 18 patients was placed through the lowest midline port while grasping forceps and irrigating suction instruments were placed laterally. The ABC was set at 85 watts of power and a flow rate between four and eight liters per minute was used.
With the patient in approximately 30 degrees of Trendelenburg position, the small bowel was displaced into the right upper quadrant. The ABC or Endoshears were used to coagulate and separate the peritoneum overlying the psoas muscles bilaterally. The round ligaments were coagulated and separated. The broad ligament was opened and the ureters at the pelvic brim were visualized. If present, the ovarian vessels were transected with Endo GIA's (United States Surgical Corp., USA) or with bipolar cautery and single Endoloop sutures (United States Surgical Corp., USA). Then the pararectal and paravesical spaces were developed. Lymph node bearing tissue overlying the external, internal, and common iliac vessels was dissected using the Endoshears or ABC to coagulate the lymphatic channels and small perforating blood vessels. After identifying the obturator nerve, the lymph nodes anterior to it were then sampled, thus completing the pelvic node dissection. Rarely, Endoclips (United States Surgical Corp., USA) were used to secure larger perforators (greater than 3 mm) during this dissection. The right ureter was then dissected free of the peritoneum and reflected laterally superior to the iliac vessels. The peritoneal incision was extended superiorly overlying the right common iliac artery and distal aorta to the level of the origin of the inferior mesenteric artery exposing the most distal portion of the aorta and vena cava. Continuing to incise peritoneum superior to the origin of the inferior mesenteric artery towards the patient's left side allowed for exposure of the duodenum which was mobilized superiorly. In most cases, optimal exposure was achieved by inserting an Endobabcock (United States Surgical Corp., USA) through the supraumbilical trocar and grasping the superior edge of the incised peritoneum over the duodenum and reflecting it superiorly. The first step in removing the left para-aortic nodes was to establish a superior boundary by skeletonizing the inferior mesenteric artery for about 6 cm from its origin by blunt dissection. As the adventitia and para-aortic lymph nodes were separated and traction with the Endobabcock placed upward on the artery, the left ureter was easily visualized as the lateral boundary of the dissection. Returning the aorta, the adventitia and lymph nodes were then separated from medial to lateral from the origin of the inferior mesenteric artery to the midpoint of the left common iliac artery (the inferior boundary). This, in addition to increasing the Trendelenburg position to approximately 40 degrees, allowed for resection of the lymph nodes overlying the anterior and left lateral aspect of the aorta.
Figure 1. Laparoscopic and Diagrammatic View of Left Aortic Lymph Node Dissection. (A)-Aorta. (LCA)-Left Common Iliac Artery. (RCA)-Right Common Iliac Artery. (IMA)-Inferior Mesenteric Artery. (U)-Ureter. (D)-Duodenum.
Figure 2. Laparoscopic and Diagrammatic View of Right Aortic Lymph Node Dissection. (A)-Aorta. (RCA)-Right Common Iliac Artery. (VC)-Vena Cava. (U)-Ureter.
On the right side, the dissection was easier. After establishing the third portion of the duodenum as the superior boundary, the right ureter as the lateral boundary, and the midportion of the right common iliac artery as the inferior boundary, the lymph node-bearing tissue was resected beginning in the interspace between the aorta and vena cava. A commonly found small vein draining the precaval fat pad into the vena cava was routinely searched for and, when found, controlled by coagulation with the ABC. The lymph node dissection was completed by removing the nodal tissue overlying the left common iliac vein in the triangle between the aortic bifurcation and the midpoints of the common iliac arteries. (Figure 2) All nodal tissue was removed via the 12 mm trocar. In patients with endometrial cancer undergoing hysterectomy, the procedure was completed vaginally. Pneumoperitoneum was then re-established and the pedicles were inspected and hemostasis assured via the laparoscope. In patients with ovarian cancer, appropriate peritoneal biopsies and infracolic omentectomy were performed.
Starting with the twenty-first patient, the order of the dissection was reversed and the aortic lymph node dissection was undertaken prior to the pelvic lymph node dissection. Additionally, the camera position was changed to the suprapubic port during the aortic lymph node dissection and then returned to the umbilical site for the pelvic procedures.
A complete bilateral aortic and pelvic lymph node dissection was performed laparoscopically in all but seven patients. Five patients required laparotomy at the time of their initial surgery. Two patients underwent exploration to control bleeding; in one patient, we were unable to control bleeding from a perforator arising from the vena cava and in another, bleeding from a perforator arising lateral to the right iliac vessel necessitated laparotomy. Two other patients required laparotomy to remove unsuspected intra-abdominal metastases. Equipment failure (non-functioning television monitor) necessitated laparotomy in the fifth patient. In two other patients, the sampling of the left aortic nodes was inadequate, as the entire area for sampling could not be exposed.
In the thirty-five patients in which the laparoscopic procedure was done, the nodal count ranged from twelve to forty-two lymph nodes with an average of 27.7 nodes per patient. An average of 20.8 pelvic lymph nodes were removed (11.0 right and 9.8 left). An average of 7.9 aortic lymph nodes were removed (3.8 right and 4.1 left). Two patients with endometrial cancer had microscopically positive lymph nodes; one had a single positive pelvic node and the other had positive aortic and pelvic nodes.
Four patients had postoperative complications. Two patients developed deep vein thromboses approximately one week postoperatively. Two other patients developed small bowel obstructions secondary to the herniation of the small bowel through the laterally placed 12 mm trocar sites diagnosed on postoperative day six and seven respectively. Both successfully underwent laparoscopic herniorrhaphy.
Operative times were reviewed and include all procedures performed. No attempt to determine a separate lymphadenectomy time was made. The mean operative time was three hours and thirteen minutes with a range from one hour and ten minutes to seven hours and twenty minutes. The mean operative time, skin to skin, for the first fifteen patients having their procedure completed laparoscopically was 3 hours and 55 minutes (range 1 hour and 55 minutes to 7 hours and 20 minutes). After we began to perform the aortic lymph node dissection first with the camera being placed through the suprapubic port, the mean operative time decreased to 2 hours and 26 minutes (range 57 minutes to 5 hours and 40 minutes).
For the thirty-five patients who had their procedures performed laparoscopically, the estimated blood losses ranged from "minimal" to 100 milliliters. The two patients undergoing laparotomy for unsuspected metastases and the one patient undergoing laparotomy secondary to equipment failure did not require transfusion. Only the patient with the vena caval injury was transfused. This patient received four units of packed cells (including two units autologous) and was discharged with a hemoglobin level of 7gm/dl.
The average hospital stay was 2.69 days with a range of one to six days. The hospital stay for the first fifteen patients averaged 2.80 days with a range from one to seven days. The average hospital stay for the last 20 patients was 2.40 days with a range of one-half to four and one-half days.
As interesting as early reports of laparoscopic lymph node dissection were, their applicability to gynecologic oncology was limited since the technique of aortic lymph node removal had not been developed. Aortic lymph node metastases can be the sole extrapelvic site of spread in cancers of the female genital tract and they are not usually included in the radiation treatment fields because the morbidity is too great to warrant doing this as a routine. Usually, histologic proof of metastases is required before aortic lymph node radiation is given. Radiographically-directed fine needle aspiration can document aortic lymph node metastases, but these are necessarily larger lymph nodes and less likely to be controlled or to represent the sole site of extrapelvic metastases. It is the smaller metastases, not identifiable radiographically, that have the best chance to be controlled by radiation therapy. Recent reports suggest that a survival benefit is possible with treatment to the aortic lymph nodes in both cervical and endometrial cancer. (3,4)
Early attempts of laparoscopic lymph node dissection did not obtain aortic lymph nodes bilaterally. Querleu et al. described only pelvic lymph node dissection in their initial report. (5) Nezhat et al. did not describe or document that the node dissection undertaken followed any particular guidelines. (6,7) Childers et al. described two patients with endometrial cancer who underwent laparoscopic staging, including hysterectomy, however, only right-sided aortic lymph nodes were sampled. (8) Subsequently, Childers et al. described the use of laparoscopy to stage patients with cervical cancer and, at the time, did not dissect the left aortic nodes. (9) In their most recent report, Childers et al. describe their experience with laparoscopic aortic lymphadenectomy in 61 patients, only 12 of which were bilateral. (10)
In February, 1992, we sought to expand our capabilities beyond laparoscopic pelvic lymph node sampling which was gaining popularity then. Our interest was to perform the same aortic lymph node sampling done at laparotomy via laparoscopy. Three different surgical approaches were considered. 1. An extraperitoneal approach beginning at the pelvic brim and continuing superiorly. 2. Transperitoneally mobilizing the sigmoid colon medially and approaching the aorta from the colonic gutter. 3. Dissecting through the mesentery of the small bowel directly over the aorta. Herd et al. could not dissect the left aortic lymph nodes when using an extraperitoneal approach in animals. (11) Attempts to dissect directly through the bowel mesentery have not been undertaken as this is not a typical approach used in gynecology oncology to dissect the left aortic nodes. Initial attempts to visualize the left lateral aspect of the aorta by mobilizing the rectosigmoid medially by incising the peritoneum along the left colonic gutter failed for mechanical reasons. In the first two patients in this series, we attempted to mobilize the rectosigmoid medially by incising the peritoneum along the left colonic gutter to gain access to the left lateral aspect of the aorta. We failed as the mass and proportions of the colon and patient position precluded adequate retraction and exposure. In the same two patients, a peritoneal incision over the lower aorta below the root of the small bowel mesentery was made in an attempt to resect the left aortic nodes. In our estimation, the lymph node sampling was suboptimal because the risks of hemorrhage would be substantially increased by any further efforts to increase exposure. Nevertheless, of all the approaches to the left aortic nodes, this inframesenteric approach held the most promise. In early May, 1992, on our third patient, adequate exposure was achieved without hemorrhage. Using the blunt tip of the ABC, all of the tissues, including peritoneum lying anterior to the aorta, were bluntly dissected free. Inserting the Endobabcock through a supraumbilical port and retracting the inferior mesenteric artery anteriorly and superiorly proved to be a maneuver which provided adequate exposure of the lower left aortic lymph nodes without adverse sequelae. It then became apparent that mobilization and cephalad retraction of the third portion of the duodenum and retraction of it superiorly improved exposure even more. This technique allowed us to successfully complete a bilateral pelvic and aortic lymph node dissection that is identical to that performed at laparotomy in the next thirteen patients via laparoscopy.
Our technique continued to evolve. The next significant change occurred when we selected to perform the aortic lymph node dissection at the beginning of the procedure and the camera was moved to the suprapubic port with video monitors placed at the head of the table. This change in operative perspective greatly improved the orientation for both the surgeon and assistant.
This change in camera position has also lessened/decreased the need of the fifth (most superior) port. Now we only use the fifth port in patients in whom intraoperatively we are unable to retract the bowel adequately so as to allow dissection of both the left and right aortic nodes at least to the level of the inferior mesenteric artery. The use of the fifth trocar is still quite useful to place instrumentation that improves exposure in the upper abdomen and/or retroperitoneal spaces. As noted, we have used the ABC almost exclusively for our dissection of both aortic and pelvic lymph nodes. We use the 10 mm probe because it is stiffer and we can actually use it as a dissectional probe, as well.
Care should be taken to limit the flow of argon gas to four liters/minute as higher flow rates lead to markedly increased intra-abdominal pressures. Even at four liters/minute, the abdomen needs to be vented to keep pressure under 15 mm/Hg. In our experience, we have found that by using two CO2 insufflators and venting one 10 mm port, we can generally keep the abdominal pressure equilibrated without using additional suction or turning the insufflators on and off intermittently.
We removed an average of 27.7 lymph nodes per patient (range 12-42). Adequacy of a lymph node sampling is commonly felt to be reflected in the number of lymph nodes removed. Even when performing therapeutic lymph node dissections, as is done for early stage cervical cancer, the number of lymph nodes removed may vary widely from patient to patient although the same operation is performed. This variation is noted not only within an institution, but also between institutions and can probably be attributed to the lack of agreement amongst pathologists as to how to best evaluate and count lymph nodes. As a result, comparing results based on lymph node counts alone is probably meaningless and speaks to the importance of providing pictorial support, ideally videotape or still photographs of the operative site following sampling as a means of quality assurance. This is particularly important when minimal access surgical procedures are proffered as equivalent to standard procedures.
Our two intraoperative complications resulting in immediate laparotomy and transfusion in a single patient, resulted from our inability to laparoscopically control hemorrhage. In the ninth patient in our series three attempts to endoscopically clip a vena cava perforator were unsuccessful. Unfortunately, an endoscopic right-angle clamp was not available in our facility at that time. It is our opinion that had this clamp been available, the bleeding site could have been clamped and subsequently successfully clipped. In the fifteenth patient, we were unsuccessful at clipping a vessel lateral to the right iliac vessels.
The ninth patient who underwent laparotomy to control hemorrhage from the vena cave then developed deep vein thrombosis, as did another patient in our series. Both patients had TED hose and sequential compression devices used intra- and postoperatively. With shortened hospital stays, perhaps continuation of anti-thrombosis prophylaxis beyond discharge should be considered.
Two other patients in our series were re-admitted to the hospital on postoperative day six and seven respectively with small bowel obstruction secondary to herniation of the small bowel through laterally placed 12 mm trocars. In both patients, screws to stabilize the sleeves were used adding to the diameter of the puncture site. It should be noted that in both cases, attempts had been made to close the fascial defects in anticipation of this potential complication. (12) We have made two changes in our procedure to prevent this complication. We no longer place 12 mm ports laterally and we close any 12 mm port site under direct laparoscopic visualization.
There was an unexpected degree of subcutaneous emphysema noted after the completion of surgery. These changes were noted in the truncal areas spreading in a cephalad direction so as to cause distention of the subcutaneous tissues around the neck. The initial two patients that demonstrated this emphysema were not extubated immediately, as neither the anesthesiologists nor surgeons were clear as to the clinical significance of this finding. Fortunately, within one to two hours, the majority of the emphysematous changes resolved and both patients were extubated without incident. It is probable that the subcutaneous emphysema is caused by trocar slippage and/or manipulation associated with high intraperitoneal pressures. As our operative times have decreased, the degree of emphysema has been decreased. Similarly, in an attempt to limit this side effect, we have changed our method of placing the primary (umbilical) trocar. By using a direct puncture technique, we believe that the chance of carbon dioxide dissecting into the subcutaneous tissue is decreased. To date we have noted a decrease in subcutaneous emphysema with this technique without an increase in morbidity.
Concerns about extreme Trendelenburg position deserve mention even though postoperative problems could not be directly linked to this in this series. Avoidance of brachial plexus and peroneal nerve injuries requires meticulous care in positioning patients. Similarly, problems relating to ventilation during and after surgery did not occur. The overall good health and lower body mass of these patients were probably the main reasons for the good outcomes in this regard. Undoubtedly, there will be complications associated with any surgical procedure whether it is accomplished endoscopically or in a traditional fashion. Our series supports early claims that complications rate associated with this procedure are acceptable. (10) As expected, with increasing experience, our complication rate continues to decrease. In our last twenty patients we have had no complications. Similarly, we have documented that the hospital stay associated with these procedures is short and decreasing as our experience increases. Part of this reduction reflects our initial reluctance to send patients home in the early postoperative period despite uncomplicated courses. However, as the initial operative times were much longer, bowel function returned more slowly and probably helped to contribute to the initial average hospital stay of almost three days.
Important in achieving the benefits of endoscopic staging is proper patient selection, availability of state-of-the-art equipment, and surgical experience. Obesity, in and of itself seems to be a limiting factor in our ability to perform aortic lymph node dissections. Childers et al. have been reluctant to perform aortic lymph node sampling in any patient weighing over 184 pounds. (10) We have used a Quetelet index of 30 or less as a means of deciding if a patient is a candidate for laparoscopic aortic lymph node sampling or not. Patient selection is most important especially as one is just beginning to perform this procedure. Patients having undergone previous surgery, especially if this included a retroperitoneal exploration, will be more difficult surgically as will those patients having undergone whole-pelvis radiation therapy. Although adhesions alone might not prevent the completion of surgical staging, their existence may add significantly to operative time, morbidity, and the frustration level of the operating surgeon.
Our operative times ranged from one hour and fifty-five minutes to seven hours and twenty minutes (average three hours and thirteen minutes). It should be noted that the operative time has consistently decreased as our experience has increased and the operative times are now much more reproducible.
Laparoscopic node dissection is very technology dependent. Therefore, the need for the operating team to be able to trouble shoot the equipment is magnified. It was necessary for one of our patients to undergo celiotomy simply because of equipment failure. This case was started late in the afternoon after the dedicated endoscopy team had gone and the less experienced crew was unable to maintain adequate video output which necessitated our abandoning the endoscopic approach.Clearly, the role of endoscopic surgery in gynecologic oncology is evolving and our experience performing bilateral aortic and pelvic lymph node sampling provides a foundation from which to expand. Potential benefits from pelvic and aortic lymph node biopsies by a laparoscopic technique are less postoperative pain, shortened hospital stay, and earlier return to full activity. Possible drawbacks are longer operating times and higher surgical costs. Care, therefore, should be taken at this time not to overstate the benefits of this approach compared to a standard laparotomy as no prospective study has compared the two techniques.
American College of Obstetricians and Gynecologists. Gynecologic Oncology Group Surgical Procedures Manual. Washington, D.C. American College of Obstetricians and Gynecologists, 1989:48.
hosla I, Lowe CR. Indices of obesity derived from body weight and height. BR. J. PREV MED SOC 1967;21:122.
Downey GO, Potish RA, Adcock LL, Prem KA, Twiggs LB. Pretreatment surgical staging in cervical carcinoma. Therapeutic efficacy of pelvic lymph node resection. AM J OBSTET GYNECOL 1989;160:1055.
Potish RA, Twiggs LB, Adcock LL, Savage JE, Levitt SH, Prem KA. Para-aortic Lymph Node Radiotherapy in Cancer of the Uterine Corpus. OBSTET GYNECOL 1985;65:251.
Querleu D, LeBanc E, Castelain B. Laparoscopic Pelvic Lymphadenectomy in the Staging of Early Carcinoma of the Cervix. AM J OBSTET GYNECOL 1991;164:579-581.
Nezhat CR, Burrell MO, Nezhat FR, et al. Laparoscopic Radical Hysterectomy with Para-Aortic and Pelvic Lymph Node Dissection. AM J OBSTET GYNECOL 1992;166:864-865.
Spirtos, NM. Laparoscopic radical hysterectomy with para-aortic lymph node dissection? Letter to the Editor. AM J OBSTET GYNECOL 1993;168:1643.
Childers J, Surwit E. Combined laparoscopic and vaginal surgery for the management of Stage I endometrial cancer. Two case reports. GYNECOL ONCOL 1992;45:46-51.
Childers JM, Hatch K, Surwit EA. The role of laparoscopic lymphadenectomy in the management of cervical carcinoma. GYNECOL ONCOL 1992;47:38-43.
Childers JM, Hatch KD, Tran Ai-Nhi, Surwit EA. Laparoscoic Para-Aortic Lymphadenectomy in Gynecologic Malignancies. OBSTET GYNECOL 1993;82:741-7.
Herd J, Fowler JM, Shenson D, Lacy S, Montz FJ. Laparoscopic Para-Aortic Lymph Node Sampling. Development of a Technique. GYNECOL ONCOL 1992;44:271-276.
Kadar N, Liu CY, Manko J, Gimpelson R. Incisional hernia after major laparoscopic gynecologic procedure. AM J OBSTET GYNECOL 1993;168:4193-5.