Abu-Rustum, Nadeem R. MD; Rhee, Eunice H.; Chi, Dennis S. MD; Sonoda, Yukio MD; Gemignani, Mary MD; Barakat, Richard R. MD
Tumor implantation in the surgical wound is a well known but infrequent complication of cancer surgery. However, the occurrence of this complication in patients undergoing a minimally invasive surgical procedure as part of their overall oncologic care may be associated with concerns by health care providers and patients as to the risks of these surgical techniques and selection of the surgical approach. This aspect might limit the use of laparoscopic surgery in the treatment of cancer in general. Moreover, since its introduction in gynecologic cancer surgery, laparoscopy has been scrutinized as a possible cause of increased incision-related metastasis or trocar-related tumor implantation.
Tumor implantation in or near a surgical scar is not new in gynecology and is not limited to laparoscopy. For example, in cervical cancer, the gynecologic literature describes subcutaneous tumor implantation in traditional abdominal incisions, pelvic drain sites, and episiotomy scars.1–3 In addition, several case reports and case series have been published describing laparoscopic trocar-related tumor implantation of a variety of gynecologic malignancies, including cervical,4 corpus,5 and ovarian cancers.6
In ovarian cancer, laparoscopic trocar site implantation is frequently described and was recently reported to occur in 6 (19.4%) of 31 patients with ovarian cancer or borderline tumors who underwent primary laparoscopic surgery.7 However, when examined carefully, the reported incidence of laparoscopy trocar-related subcutaneous tumor implantation is generally low (less than 1%) in large surgical oncology practices,8 particularly if the denominator of all laparoscopic procedures is described, which is generally deficient in many of the smaller series and case reports frequently published. We conducted this review to describe the incidence of clinically detected laparoscopy incision-related subcutaneous tumor implantation in women with malignant disease who underwent a transperitoneal laparoscopy by the gynecologic oncology service.
MATERIALS AND METHODS
Using the Virginia K. Pierce Gynecology Service Database, we identified all patients who underwent a transperitoneal laparoscopic procedure by the gynecologic oncology service. We excluded all patients with benign disease or no history of malignancy. In addition, we excluded all patients with premalignant disease, those who had less than a 1-month postoperative follow-up, those who underwent an extraperitoneal laparoscopy, and those who had the laparoscopic procedure performed elsewhere.
Identifying subcutaneous implantation was performed by a thorough review of all available medical records, including operative notes, radiology reports, pathology reports, and progress notes from all involved services, such as the gynecologic oncology, radiation oncology, and medical oncology services. We also reviewed all outside correspondence, such as imaging tests or reports relating to possible physical findings that the patient may have had in combination with follow-up at our institution and reviewed patients’ discharge summaries. All medical record reviews were performed by 2 authors (N.A.R. and E.H.R.). In addition, incision-related complications, including subcutaneous, intraperitoneal catheter, and trocar-related complications, were identified after reviewing a comprehensive complications database maintained by the gynecology service and reviewed at a weekly disease management team conference.
All patients included in the analysis had a postoperative physical examination performed by a gynecologic oncologist, medical oncologist, or radiation oncologist. A clinically detected subcutaneous implantation was noted when the physician described a subcutaneous nodule near or at the previous laparoscopy-related scar. Pathologic and radiologic confirmation was noted when available. We used the term “isolated” subcutaneous tumor recurrence, as previously described by Reymond et al,9 who defined a port-site recurrence as an early tumor recurrence that develops locally in the abdominal wall, within the scar tissue of 1 or more trocar sites or an incision, after laparoscopy; these implantations are not associated with peritoneal carcinomatosis. All pathology specimens were reviewed at Memorial Sloan-Kettering Cancer Center.
Open transperitoneal laparoscopy technique was used in all cases with a Hasson trocar, usually placed at the umbilicus, often by using a previous abdominal scar. Technique aspects have been described in previous publications.10,11 A carbon dioxide pneumoperitoneum was used in all cases, with the maximum intraabdominal pressure set at 15 mm Hg. A variety of trocar types and laparoscopic instrumentation was used during the study period, and the number of trocars inserted per case varied depending on the procedure and the surgeon's preference. It was common to insert the initial trocar, by using the open technique, in a previous scar if available; moreover, it was also common to attempt to use previous scars when possible to insert lateral accessory trocars. All trocar sites more than 5 mm were closed at the fascia and skin levels. The 5-mm trocar incisions were usually closed at the skin level only. Closure of the peritoneum was variable. Specimens were generally removed via the trocar incisions. Small specimens and biopsies were removed directly through the trocars with 5-mm biopsy forceps, graspers, or a 10-mm spoon. Larger specimens were removed with laparoscopy retrieval bags, especially if there was concern for malignancy. The uterus and remaining adnexa were removed vaginally during laparoscopically assisted vaginal hysterectomy. We did not routinely irrigate the incisions with antiseptics or cytotoxic agents. There were no other standard trocar-incision care protocols used during the study period.
Semipermanent intraperitoneal catheters with subcutaneous reservoirs were inserted in selected patients for administration of intraperitoneal chemotherapy or radiocolloid. The catheter tip was usually inserted via a 5-mm lateral laparoscopic trocar and the proximal tip tunneled subcutaneously to the reservoir, which was fixed to the anterior left lower ribcage with interrupted sutures.12,13 The skin over the 5-mm trocar where the catheter was inserted and the skin over the reservoir were closed with absorbable sutures. The intraperitoneal catheters were generally accessed and used after a 2-week postoperative recovery period. The intraperitoneal catheters were generally kept in the patient for several months and were usually removed under local anesthesia after completion of the planned treatment by using the previous reservoir skin incision. The skin was reapproximated with absorbable suture, and there were no other standard incision care protocols used during the study period.
Dichotomous variables were analyzed by using the χ2 and Fisher exact tests. Dichotomous variables with small numbers in some cells were analyzed with the Fisher exact test, and the χ2 test was used for other dichotomous variables. Continuous variables were analyzed by using the t test. SPSS software (SPSS, Inc, Chicago, IL) was used for statistical analysis. Statistical significance was set at P < .05. The institutional review board at Memorial Sloan-Kettering Cancer Center approved the study.
Between July 1, 1991, and July 1, 2003, 2,593 laparoscopic procedures were performed by the gynecologic oncology service, including 1,335 transperitoneal laparoscopic procedures in 1,288 women with malignant disease. During the study period, approximately 3,237 previously untreated patients with malignancy and 4,602 patients with previously treated malignancy were admitted and managed by the gynecology service. Median patient age at the time of laparoscopy was 56 years (mean 55.9 years, range 15 to 91 years). Age did not significantly differ for patients who developed subcutaneous implantation (median 54 years, mean 54.8 years, range 42 to 70 years) when compared with patients who did not develop subcutaneous implantation (median 56 years, mean 55.9 years, range 15 to 91 years; P = .76). The majority of patients with malignant disease undergoing laparoscopy had ovarian cancer. Other malignant disease sites are summarized in Table 1. Patients with lymphoma, renal, gastric, colorectal, lung, skin, or other less frequently occurring cancers operated on by the gynecologic oncology service were listed as “other malignancy.”
For ovarian and peritoneal cancer, the incidence of residual persistent disease at the time of a second-look laparoscopy is approximately 54%.10,14 Semipermanent intraperitoneal catheters were frequently inserted in patients undergoing laparoscopy with a history of ovarian cancer and used in the setting of consolidation therapy (66%) or salvage intraperitoneal treatment of persistent residual disease (69%).14 In this series, intraperitoneal catheters with a subcutaneous reservoir were inserted at the time of laparoscopy for planned intraperitoneal chemotherapy in 359 cases and for the administration of intraperitoneal P32 in 1 patient.
Primary disease stage category was documented when available and included stage I (n = 376), stage II (n = 75), stage III (n = 386), and stage IV (n = 96). In 355 patients, the original staging data were incomplete, and stage category could not be determined.
The majority of patients (1,242 [96.4%] of 1,288) underwent only 1 laparoscopic procedure during the study period; however, 45 patients underwent 2 laparoscopies, and 1 patient underwent 3 laparoscopic procedures during the study period, resulting in a total of 1,335 procedures. For patients who underwent more than 1 laparoscopic procedure, the disease sites included the following: ovary (n = 36), peritoneum (n = 3), breast (n = 2), bladder (n = 1), fallopian tube (n = 1), colon (n = 1), uterine (n = 1), and unknown primary cancer (n = 1).
A variety of diagnostic or operative laparoscopic procedures was performed. If more than 1 procedure was performed at the same laparoscopy, only the main procedure was listed. Table 2 summarizes the last laparoscopic procedure performed on all patients during the study period.
For the purpose of this report, subcutaneous tumor implantation was classified into 2 categories: tumor implantation related to the intraperitoneal catheter incision (Table 3) and subcutaneous tumor implantation related to or near previous laparoscopy trocar scars (Table 4). Although intraperitoneal catheters were placed at the time of laparoscopy, the reservoir incision was always separate from the trocar sites, and the presence of a semipermanent foreign device connecting the peritoneal cavity to the subcutaneous tissue for several months was evaluated separately from an isolated trocar-related subcutaneous implantation independent of the presence of a foreign subcutaneous device.
There were no “isolated” trocar-related subcutaneous tumor implantations during the study period. All subcutaneous tumor implantations (n = 13 patients) occurred in women with previously documented advanced-stage or metastatic disease. In fact, 11 of the 13 patients who developed subcutaneous tumor implantation had diffuse carcinomatosis, ascites, or malignant peritoneal cytology at the time of the laparoscopy that preceded tumor implantation, and subcutaneous disease developed after overall progression of carcinomatosis (Tables 3 and 4).
In all, 7 (1.94%) of 360 patients (95% confidence interval [CI] 0.85%, 4.14%) developed subcutaneous tumor implantation at or near the site of the intraperitoneal catheter scar (Table 3). One of the 7 patients (Table 3, patient 5) also had synchronous implantation detected in the lateral trocar scar. This patient is counted as having both trocar-related implantation and intraperitoneal catheter-related implantation. All 7 patients had known advanced-stage malignancy before laparoscopy, including 5 stage IIIC ovarian cancers, 1 stage IIIC fallopian tube cancer, and 1 stage IIIA uterine corpus cancer (positive cytology). In addition, all had documented peritoneal disease present at the time of laparoscopy when the intraperitoneal catheter was inserted. The median interval from laparoscopy with intraperitoneal catheter insertion to developing clinically detected tumor implantation related to the intraperitoneal catheter site was 13 months (mean 11.8 months, range 4 to 21 months). The patients with the shortest interval (4 months) developed extensive diffuse progression of disease while on intraperitoneal chemotherapy, and the reservoir site was noted to be involved with tumor at the time of catheter removal.
Tumor implantation occurred at the reservoir site over the lower anterior chest wall in all 7 cases (Figure 1) and was confirmed pathologically in 4 cases. It should be emphasized that these were not isolated subcutaneous tumor implantations. All patients had synchronous metastases to other sites, including clinical progression of advanced carcinomatosis in 6 cases and splenic metastasis in 1 case at the time the subcutaneous tumor was noted. All patients received subsequent chemotherapy, usually started within 2 to 6 weeks from diagnosis, and in 3 cases, the subcutaneous tumor nodules were resected.
In addition, 7 (0.54%) of 1,288 patients (95% CI 0.24%, 1.17%) developed subcutaneous tumor nodules near or at a previous laparoscopy scar (including patient 5 from Table 3, who developed implantation at the trocar and intraperitoneal catheter site). Unlike intraperitoneal catheter reservoir incisions, which were always placed in a previously unscarred location, it was common to perform the laparoscopy trocar incisions over previous abdominal or pelvic skin scars when possible, and the operative reports and progress notes were used to identify if the scars from different procedures overlapped. Similar to cases that developed intraperitoneal catheter site tumor implantation, all patients who developed trocar-related subcutaneous nodules had known advanced-stage disease, including stage IIIC ovarian cancer,3 stage IV breast cancer metastatic to abdomen and pelvis,2 and recurrent cervical cancer.1 The median time from laparoscopy to development of clinically detected trocar incision-related tumor implantation was 8.5 months (mean 9 months, range 9 days to 18 months). The patient with the shortest interval had laparoscopy with drainage of malignant ascites and biopsy of extensive carcinomatosis that confirmed metastatic breast cancer to the abdomen. She developed clinically suspicious subcutaneous nodules noted at her postoperative visit, and it was not clear whether these clinical findings were inflammatory or neoplastic. In another patient with advanced ovarian cancer, the subcutaneous tumor was detected as part of a large umbilical hernia, and in 1 patient with recurrent advanced cervical cancer invading the pelvic sidewall and lower abdominal wall, the subcutaneous tumor could not be accurately attributed to the laparoscopy site or a previous transverse abdominal scar used for a radical abdominal hysterectomy (Table 4). Pathologic confirmation of the subcutaneous tumor was obtained in 5 of the 7 cases. All patients received chemotherapy, and 4 patients had the tumor resected before chemotherapy.
The overall incidence of clinically detected subcutaneous tumor implantation, based on subjects rather than procedure, was 13 (1.01%) of 1,288 (95% CI 0.56%, 1.77%). Implantation near trocar incisions did not appear to significantly differ when compared among patients with a history of adnexal/peritoneal cancer (4 [0.68%] of 584), corpus cancer (0 [0%] of 355), cervical cancer (1 [1.0%] of 100), or other malignancy (2 [0.8%] of 249; P = .27, Fisher exact test). The incidence of subcutaneous tumor implantation based on procedure in each disease site is summarized in Table 5.
Metastatic carcinoma to the skin is an uncommon finding in gynecology with the most common solid tumors metastasizing to the skin being breast, gastrointestinal, and ovarian cancers. However, cancer recurrence in the surgical incision is not new in gynecology and is not limited to laparoscopy. Subcutaneous tumor implantation has been reported in cervical, uterine, and ovarian cancers, usually as case reports with unknown denominators or small retrospective case series.
Trocar-related tumor implantation has been evaluated in other malignancies. In a large series recently reported by Memorial Sloan-Kettering Cancer Center,8 port site implantation was noted in only 0.79% of 1,650 laparoscopic cases for upper gastrointestinal malignancy compared with skin implantation in 0.86% of 1,040 laparotomy incisions for patients with similar disease. This led the authors to conclude that laparoscopy-related tumor implantation is uncommon, occurs in the setting of advanced disease, and does not seem to be different from open incision site recurrence.8
In ovarian cancer, cutaneous involvement is unusual at presentation (0.45%) and during the course of ovarian carcinoma treatment (4%).15 Laparoscopic trocar metastases have been reported with low-malignant potential tumors16 and invasive cancers. In 1994, Childers et al6 reported that trocar tumor implantation in women with cancer was infrequent and occurs at approximately 0.2% of puncture sites or 1% of procedures. In 1996, Kruitwagen17 reported that the occurrence of abdominal wall metastasis after laparoscopy or paracentesis was noted only in patients with stage IIIC-IV disease, which included the presence of ascites, similar to our experience, and that the development of subcutaneous metastasis did not affect survival. Also, in 1999, Wang et al18 reported that the majority of trocar skin recurrences were in advanced adenocarcinoma cases, often with carcinomatosis, ascites, or a diagnosis of palliative procedure for malignancy. This appears to be the common finding in the majority of ovarian cancer patients who develop subcutaneous metastasis. Skin implantation or recurrence is only a fraction of the overall progression of disease, which is usually associated with carcinomatosis and bulky intraabdominal metastasis. The effects of laparoscopy with carbon dioxide pneumoperitoneum on the overall survival of women with advanced ovarian cancer have also been investigated, and there appears to be no difference in survival between patients with persistent ovarian cancer as noted at the time of second-look operation with laparoscopy or laparotomy.19
In 1998, Reymond et al9 proposed a definition for port site recurrence as early tumor recurrence that develops locally in the abdominal wall, within the scar tissue of one or more trocar sites or an incision, after laparoscopy or thoracoscopy for cancer, and these implantations are not associated with peritoneal carcinomatosis. The authors also outlined that one of the conditions that may lead to the development of port site recurrence is the presence of tumor cells within the abdominal cavity at the time of laparoscopy, a condition that was commonly seen in the majority of patients who developed subcutaneous implantation from this series. The use of the above definition is important because it limits the number of cases reported to those with true isolated recurrences, none of which were seen in this series.
In uterine cervical cancer, subcutaneous implantation has been reported in laparotomy skin incisions, retroperitoneal drain sites, episiotomy scars, and laparoscopy trocar scars.1–4 Numerous case reports are available, but the denominator is usually unknown. However, these data suggest that the incidence with transperitoneal laparoscopy is near 1% and is not higher than that seen with other gynecologic or nongynecologic malignancies. Similarly, for uterine corpus cancer, few case reports are present5; however, the overall incidence appears generally low, possibly because of the fact that the majority of endometrial cancer patients undergoing laparoscopy have early-stage disease, unlike the case we described in this series, in which malignant peritoneal fluid (stage IIIA) was already present at the time of laparoscopy. The completion of the ongoing phase III trial conducted by the Gynecologic Oncology Group, in which women with operable uterine corpus cancer are randomly assigned to either open versus laparoscopic staging and hysterectomy with salpingo-oophorectomy, will hopefully give us the definitive conclusion to subcutaneous implantation in this disease.
Several investigators have contributed to the understanding of subcutaneous implantation and have suggested means to reduce its occurrence.9,18,20,21 It appears that the incidence of port site recurrence differs among surgeons, suggesting a difference in technique. In 1999, Van Dam et al21 confirmed that for diagnostic laparoscopy in women with advanced ovarian cancer, trocar recurrence was significantly lower if careful closure of the peritoneum, fascia, and skin was performed followed by planned cytoreductive surgery or chemotherapy. This is consistent with other surgical oncology data which suggests that with proper surgical technique, knowledge of the disease process, and patterns of tumor spread, it is possible to keep the incidence of subcutaneous metastasis after laparoscopy near 1%, which appears comparable with the incidence with open surgery in similar groups of patients.8,20 Preclinical investigations continue to contribute to our understanding of this process, as recently suggested by studies on animal xenograft ovarian cancer models, which have shown that peritoneal closure of port sites after carbon dioxide pneumoperitoneum decreases the risk of port site metastasis.22
In conclusion, laparoscopy-related subcutaneous tumor implantation is rare (0.97%) in women undergoing a transperitoneal laparoscopy with malignant disease. Subcutaneous implantation appears to occur in patients with known metastatic disease and is detected in the setting of synchronous advanced intraabdominal or pelvic metastasis and carcinomatosis, usually as part of overall progression of disease. The risk of an “isolated” subcutaneous tumor implantation appears to be even lower. Therefore, the risk of subcutaneous tumor implantation should not be used as an argument against laparoscopy in the majority of women with gynecologic malignancies managed by gynecologic oncologists. Large prospective cooperative group trials in gynecologic and nongynecologic malignancies will hopefully continue to contribute more to our understanding of this infrequent occurrence.
1.Stenson R, Jacobs AJ, Janney CG, Schmidt DA. Incisional recurrence of squamous cell cervical carcinoma following operative staging. Gynecol Oncol
2.Copas PR, Spann CO, Thoms WW, Horowitz IR. Squamous cell carcinoma of the cervix metastatic to a drain site. Gynecol Oncol
3.Copeland LJ, Saul PB, Sneige N. Cervical adenocarcinoma: tumor implantation in the episiotomy sites of two patients. Gynecol Oncol
4.Kohlberger PD, Edwards L, Collins C, Milross C, Hacker NF. Laparoscopic port-site recurrence following surgery for a stage IB squamous cell carcinoma of the cervix with negative lymph nodes. Gynecol Oncol
5.Wang PH, Yen MS, Yuan CC, Chao KC, Ng HT, Lee WL, et al. Port site metastasis after laparoscopic-assisted vaginal hysterectomy for endometrial cancer: possible mechanisms and prevention [review]. Gynecol Oncol
6.Childers JM, Aqua KA, Surwit EA, Hallum AV, Hatch KD. Abdominal-wall tumor implantation after laparoscopy for malignant conditions. Obstet Gynecol
7.Huang KG, Wang CJ, Chang TC, Liou JD, Hsueh S, Lai CH, et al. Management of port-site metastasis after laparoscopic surgery for ovarian cancer. Am J Obstet Gynecol
8.Shoup M, Brennan MF, Karpeh MS, Gillern SM, McMahon RL, Conlon KC. Port site metastasis after diagnostic laparoscopy for upper gastrointestinal tract malignancies: an uncommon entity. Ann Surg Oncol
9.Reymond MA, Schneider C, Kastl S, Hohenberger W, Kockerling F. The pathogenesis of port-site recurrences [review]. J Gastrointest Surg
10.Abu-Rustum NR, Barakat RR, Siegel PL, Venkatraman E, Curtin JP, Hoskins WJ. Second-look operation for epithelial ovarian cancer: laparoscopy or laparotomy. Obstet Gynecol
11.Abu-Rustum NR, Chi DS, Sonoda Y, DiClemente MJ, Bekker G, Gemignani M, et al. Transperitoneal laparoscopic pelvic and paraaortic lymph node dissection using the argon beam coagulator and monopolar instruments: an 8-year study and description of technique. Gynecol Oncol
12.Abu-Rustum NR, Chi DS, Curtin JP. Epithelial ovarian cancer [review]. Curr Probl Surg
13.Awtrey CS, Abu-Rustum NR. Second-look laparoscopy with intraperitoneal catheter placement. In: Levine DA, Barakat RR, Hoskins WJ, editors. Atlas of procedures in gynecologic oncology. London, England: Martin Dunitz; 2003. p. 161–72.
14.Husain A, Chi DS, Prasad M, Abu-Rustum N, Barakat RR, Brown CL, et al. The role of laparoscopy in second-look evaluations for ovarian cancer. Gynecol Oncol
15.Cormio G, Capotorto M, Vagno GD, Cazzolla A, Carriero C, Selvaggi L. Skin metastases in ovarian carcinoma: a report of nine cases and a review of the literature [review]. Gynecol Oncol
16.Hsiu JG, Given FT Jr, Kemp GM. Tumor implantation after diagnostic laparoscopic biopsy of serous ovarian tumors of low malignant potential. Obstet Gynecol
17.Kruitwagen RF, Swinkels BM, Keyser KG, Doesburg WH, Schijf CP. Incidence and effect on survival of abdominal wall metastases at trocar or puncture sites following laparoscopy or paracentesis in women with ovarian cancer. Gynecol Oncol
18.Wang PH, Yuan CC, Lin G, Ng HT, Chao HT. Risk factors contributing to early occurrence of port site metastases of laparoscopic surgery for malignancy [review]. Gynecol Oncol
19.Abu-Rustum NR, Sonoda Y, Chi DS, Teoman H, Dizon DS, Venkatraman E, et al. The effects of CO2
pneumoperitoneum on the survival of women with persistent metastatic ovarian cancer. Gynecol Oncol
20.Schneider C, Jung A, Reymond MA, Tannapfel A, Balli J, Franklin ME, et al. Efficacy of surgical measures in preventing port-site recurrences in a porcine model. Surg Endosc
21.Van Dam PA, DeCloedt J, Tjalma WA, Buytaert P, Becquart D, Vergote IB. Trocar implantation metastasis after laparoscopy in patients with advanced ovarian cancer: can the risk be reduced? Am J Obstet Gynecol
22.Agostini A, Robin F, Jais JP, Aggerbeck M, Vilde F, Blanc B, et al. Peritoneal closure reduces port site metastases: results of an experiment in a rat ovarian cancer model. Surg Endosc