Lymphoceles after laparoscopic pelvic node dissection

LYMPHOCELES AFTER LAPAROSCOPIC PELVIC NODE DISSECTION RUSSELL M. FREID, DAVID SIEGEL, ARTHUR D. SMITH,

AND

GARY H. WEISS

ABSTRACT Objective. Lymphocele formation has been infrequently reported as a complication of laparoscopic pelvic lymph node dissection (LPLND). We determined the incidence of clinical and subclinical lymphocele formation in patients undergoing transperitoneal LPLND. Methods. Charts and radiological records of 111 patients undergoing transperitoneal LPLND at this institution between January 1991 and December 1995 were reviewed to determine the incidence of lymphocele formation. Results. Of 111 patients undergoing LPLND, 12.6% had positive lymph nodes and received hormonal therapy. Radical retropubic (12) or perineal (28) prostatectomy was performed either simultaneously or within 2 weeks in 41% of the node-negative patients. Radiation therapy was the treatment modality in the remaining node negative patients (N 5 57). Twenty-three patients undergoing radiation therapy had preplanning pelvic computed tomography (CT) scans 2 to 16 weeks (mean 8.2 weeks) after LPLND. These were reviewed by a single radiologist to determine the presence of subclinical lymphoceles. Seven patients (30.4%) had lymphoceles of varying sizes (3 large and 4 small). Although most were identified on CT scans 4 weeks after the procedure, two were identified on scans 12 and 16 weeks after the procedure (mean 6.5 weeks). None of these patients developed symptoms referable to or had treatment for the lymphocele during a 2 to 37 month follow-up (mean 20 months). Only two patients (3.5%) undergoing LPLND as an isolated procedure had clinical evidence of lymphocele formation, both of which were subsequently confirmed with CT scans (1 large, 1 small). One was treated with CT-guided drainage and sclerosis and the other resolved spontaneously. Conclusion. The clinical incidence of lymphocele formation following LPLND remains relatively low. Only a portion of these patients requires intervention. Subclinical lymphoceles, as detected on follow-up CT scans, occur with a much greater frequency. These seldom become symptomatic requiring treatment. Rather, they appear to resolve spontaneously. Nevertheless, clinical suspicion should remain high in order to detect and properly treat symptomatic lymphoceles when they occur. UROLOGY 51 (Suppl 5A): 131–134, 1998. © 1998, Elsevier Science Inc. All rights reserved.

S

ince its initial introduction in 1989, laparoscopic pelvic lymph node dissection (LPLND) has become an established technique to safely and accurately determine the nodal status of prostate cancer patients.1 The overall complication rate (minor and major) of a multi-institutional review of 500 patients was 15%. The most common major complication was intraoperative vascular injury (2%), while postoperative urinary retention was the most common minor complication (1.6%). The From the Departments of Urology and Radiology, Long Island Jewish Medical Center, New Hyde Park, New York Reprint requests: Russell M. Freid, MD, Department of Urology, Long Island Jewish Center, 270-05 76th Ave., New Hyde Park, NY 11040 © 1998, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED

incidence of clinically detected lymphocele formation in this series was only 1%. The incidence of clinically detected lymphoceles after open pelvic lymph node dissection (OPLND) has been reported to occur much more frequently, ranging from 4.7% to 14.8%.2,3 This higher frequency may be due to several factors, including more extensive dissection, lack of peritoneal communication, and the use of subcutaneous minidose heparin in many patients. Furthermore, when consecutive (n 5 22) patients who underwent OPLND were followed with serial pelvic sonography, Spring et al. found an even higher incidence (27%) of subclinical lymphoceles.4 Although most of these lymphoceles, especially small echo-free ones, 0090-4295/98/$19.00 PII S0090-4295(98)00074-0 131

resolved spontaneously, 41% of the total did require intervention. We attempted to determine the incidence of lymphocele formation (clinical and subclinical) in patients undergoing LPLND for prostate cancer staging.

TABLE I. Symptomatic and asymptomatic lymphoceles after series of 57 laparoscopic pelvic lymph node dissections Symptomatic Lymphoceles Patient

Lymphocele Size 3

Weeks Post-Op.

MATERIALS AND METHODS

A.M.

2 cm R

4

Between January 1991 and December 1995, LPLND was performed at our institution by three surgeons in 111 patients with prostate cancer. All procedures were done with the patient under general anesthesia. All node dissections were performed through a transperitoneal approach using a linear incision in the peritoneal surface lateral to the medial umbilical ligament. Hemoclips and electrocautery were used liberally to occlude lymphatic vessels prior to transection in all cases. No drains were placed in any patient. All patients had intraoperative sequential compression stockings, and none received perioperative subcutaneous heparin. Fourteen (12.6%) patients were found to have positive lymph nodes and received hormonal therapy. None developed clinically evident lymphoceles. The remaining patients underwent either radical prostatectomy or radiation therapy according to patient and physician choice. Although none had clinically evident lymphoceles, patients undergoing radical retropubic (n 5 12) or perineal (n 5 28) prostatectomy simultaneously with or shortly (,10 days) after LPLND were excluded from analysis to avoid the confounding effect of the other procedure. Charts, radiologic records, and clinical follow-up of the remaining patients (n 5 57) were reviewed to determine the presence of any lymphoceles. Lymphocele size was calculated from computed tomography (CT) scan images using the prolate ellipsoid formula:

A.N.

126 cm3 R

9

L * W * H * 0.53 where L, W, and H are the measured length, width, and height, respectively.

RESULTS The overall complication rate of our series was comparable to other reports. Major complications occurred in 3 patients (2.7%) and included ureteral transection, small bowel perforation, and small bowel obstruction. Minor complications occurred in 9%, with postoperative urinary retention occurring most commonly (4.5%). Clinically evident lymphoceles were noted in only 2 of 57 (3.5%) patients who underwent LPLND as an isolated procedure (Table I). One patient presented with a lower extremity deep venous thrombosis and fever 7 weeks after LPLND. CT scan subsequently revealed a large pelvic lymphocele, which was aspirated percutaneously and left to gravity drainage. After 7 days, the drain output was minimal, and absolute ethanol was injected into the cavity as a sclerosant prior to removing the drain. The patient also had an inferior vena cava filter placed to reduce the likelihood of pulmonary embolus. The second clinical presentation was that of clear fluid draining from the infraumbilical trocar site 4 weeks postoperatively. The pa132

Misc. Infraumbilical drainage Deep vein thrombosis fever

274 cm3 L Asymptomatic Lymphoceles Follow-up M.Z. W.G. B.T. R.V. H.G. W.G. A.H.

19 cm3 R 119 cm3 L 6 cm3 R 40 cm3 L 2.8 cm3 L 6 4 41 11

cm3 cm3 cm3 cm3

L L R L

4

37 months*

4 4 16

26 months† 36 months‡ aspirated at 16 weeks 17 months 2 months 2 months

12 4 4

* Persistant L40 cm3 lymphocele at 35 months. † Lymphocele unchanged at 12 weeks. ‡ Radiologic resolution at 5 months.

tient had no other symptoms and CT scan revealed only a small residual lymphocele, that resolved with conservative management. Twenty-three patients who were scheduled to receive prostate radiation therapy had preplanning pelvic CT scans available for review. The clinical characteristics (prostate-specific antigen [PSA], Gleason score, and stage) of the cancers in this group were representative of those found in patients undergoing other treatments. The CT scans were performed 2 to 16 weeks (mean 8.2 weeks) after LPLND. A single radiologist (D.S.) reviewed these films, and lymphoceles were noted in 7 (30.4%) patients (Table I). These were present on CT scans obtained 4 to 16 weeks (mean 6.5 weeks) postsurgery. This time interval was not significantly different from the negative studies that were obtained 3 to 20 weeks (mean 9.1 weeks) after surgery. Four patients had lymphoceles that were small in size (,40 cm3) and three had lymphoceles that were large ($40 cm3; Fig. 1). The mean size of these asymptomatic lymphoceles (35.5 cm3) was not statistically different from the mean size of the symptomatic ones (134.0 cm3) due to the small number of patients and the spontaneous drainage of one symptomatic lymphocele prior to CT measurement. None of these patients had identifiable risk factors for lymphocele formation. All patients had LPLND performed in a similar manner and no UROLOGY 51 (Supplement 5A), May 1998

FIGURE 1. Pelvic CT image of large right and small left lymphoceles of patient A.H. (Table 1) that, despite their pressure on the bladder, remained asymptomatic.

difference in the number of hemoclips visualized on CT was noted between patients with or without lymphoceles. Mean clinical follow-up for these seven patients was 20 months (range 2 to 37 months) after radiologic evidence of the lymphocele was present. No symptoms referable to the lymphocele developed during this period. However, one patient with a large lymphocele had palpable lower abdominal fullness when examined. One patient with a small asymptomatic lymphocele underwent CT-guided needle aspiration to rule out a necrotic lymph node. This lymphocele, which was located at the caudal aspect of the external iliac vein, had an atypical appearance with low attenuation and debris in its center. Aspirated fluid was straw colored and had scant white blood cells when examined microscopically. Additional radiologic studies for the purpose of monitoring lymphocele resolution were not ordered on any patient during the follow-up period. However, additional CT scans performed for other reasons were available in 3 patients (Table I). These studies revealed the partial resolution of a large lymphocele in one patient 34 months after its radiologic detection. A second large lymphocele was noted to completely resolve 4 months after initial diagnosis while a small lymphocele was unchanged 8 weeks after identification. COMMENT Although the indications for staging LPLND have been refined, it is still a commonly performed procedure due to the high prevalence of UROLOGY 51 (Supplement 5A), May 1998

prostate cancer. While major complications of the procedure are rare and related to the learning curve of the surgeon, minor complications occur with regularity and are responsible for most of the morbidity of the procedure. Lymphoceles have been reported as seldomly occurring after this procedure, as compared to their frequent occurrence after OPLND. We set out to determine whether this decrease was a true phenomenon as a result of potential differences in the two procedures or the result of a reduced detection rate. When performing LPLND, approximately the same number of lymph nodes are recovered as with OPLND.5,6 However, the latter is a more extensive procedure. The transperitoneal approach of most LPLND further distinguishes this technique from the OPLND, which generally has no peritoneal communication. Subcutaneous minidose heparin, which has been identified as a predisposing factor for lymphocele formation,7,8 has been eliminated from most LPLND perioperative protocols. Finally, a rapid return to full ambulation and activity is clearly a factor that is present only with LPLND. In light of these differences, one would expect an extremely low lymphocele rate after LPLND. However, in our limited group of patients who underwent radiologic surveillance, we found a surprisingly high (30.4%) rate of subclinical lymphoceles. In a similar study by Spring et al., using serial pelvic monographic evaluation (1 to 12 weeks postoperatively), found that 27% of patients who underwent OPLND had asymptomatic lymphoce133

les.4 Why, if similar subclinical lymphocele rates of the two procedures are nearly equal, would there be such a disparity between their clinical presentation? Since, with the exception of infection, symptoms of lymphoceles are due to the mass effect they create in the pelvis, small lymphoceles would rarely cause symptoms. One would expect the peritoneal incision of LPLND to not only reduce the incidence of lymphoceles but also reduce the size of those that occur. However, in our series, 43% of the lymphoceles were larger than 40 cm3, which is similar to the size distribution (42% large) that Spring et al. found. Several other studies further illustrate this point. For instance, in a series of 187 OPLND, clinical lymphoceles formed in 8.4% of extraperitoneal dissections but in only 1.1% of transperitoneal dissections.2 In a review of complications of LPLND, a lymphocele was reported in a patient who underwent closure of the peritoneum after completion of the dissection.9 In the few series of extraperitoneal LPLND that have been published, the incidence of clinical lymphoceles (6.5%) is several times higher than reports of transperitoneal LPLND.10 –12 In conclusion, subclinical lymphoceles, as detected on follow-up CT scans, occur frequently after LPLND. Only a small fraction of these (primarily large ones) become clinically evident and require treatment. A large incision in the peritoneum, such as the inverted ‘‘V’’ described by See et al.,13 may promote postoperative intraperitoneal lymphatic fluid drainage, thus reducing the likelihood of forming of a large lymphocele. Nevertheless, clinical suspicion should remain high to detect and properly treat symptomatic lymphoceles when they occur. REFERENCES 1. Schuessler WW, Vancaille TG, Reich H, and Griffith DP: Transperitoneal endosurgical lymphadenectomy in pa-

134

tients with localized prostate cancer. J Urol 145: 988 –991, 1991. 2. Sogani PC, Watson RC, and Whitmore WF Jr: Lymphocele after pelvic lymphadenectomy for urologic cancer. Urology 17(1): 3943, 1981. 3. Ojeda L, Sharifi R, Lee M, Mouli K, and Guinan P: Lymphocele formation after extraperitoneal pelvic lymphadenectomy: possible predisposing factors. J Urol 136: 616 – 618, 1986. 4. Spring DB, Schroeder D, Babu S, Agee R, and Gooding GAW: Ultrasonic evaluation of lymphocele formation after staging lymphadenectomy for prostate carcinoma. Radiology 141: 479 – 483, 1981. 5. Winfield HN, Donovan JF, See WA, Loening SA, and Williams RD: Laparoscopic pelvic lymph node dissection for genitourinary malignancies: indications, techniques and results. J Endourol 6: 103–111, 1992. 6. Parra RO, Andrus C, and Boullier J: Staging laparoscopic pelvic lymph node dissection: comparison of results with open pelvic lymphadenectomy (abst). J Urol 147(part 2): 875, 1992. 7. Bigg SW, and Catalona WJ: Prophylactic mini-dose heparin in patients undergoing radical retropubic prostatectomy. A prospective trial. Urology 39(4): 309 –313, 1992. 8. Koonce J, Selikowitz S, and McDougal WS: Complications of low-dose heparin prophylaxis following pelvic lymphadenectomy. Urology 28(1): 21–25, 1986. 9. Kavoussi LR, Sosa E, Chandoke P, et al: Complications of laparoscopic pelvic lymph node dissection. J Urol 149: 322– 325, 1993. 10. Das S, and Tashima M: Extraperitoneal laparoscopic staging pelvic lymph node dissection. J Urol 151: 1321–1323, 1994. 11. Villers A, Vannier JL, Abecassis R, et al: Extraperitoneal endosurgical lymphadenectomy with insufflation in the staging of bladder and prostate cancer. J Endourol 7: 229 –233, 1993. 12. Etwaru D, Raboy A, Ferzli G, and Albert P: Extraperitoneal endoscopic gasless pelvic lymph node dissection. J Laparoendoscop Surg 4: 113–116, 1994. 13. See WA, Cohen MB, and Winfield HN: Inverted V peritoneotomy significantly improves nodal yield in laparoscopic pelvic lymphadenectomy. J Urol 149: 772–775, 1993.

UROLOGY 51 (Supplement 5A), May 1998