Trends in port-site metastasis after laparoscopic resection of incidental gallbladder cancer: A systematic review

ARTICLE IN PRESS

Trends in port-site metastasis after laparoscopic resection of incidental gallbladder cancer: A systematic review David Berger-Richardson, MD, CM,a,b,c Tyler R. Chesney, MD, MSc,a Marina Englesakis, MLIS,d Anand Govindarajan, MD, MSc, FRCSC,a,c,e Sean P. Cleary, MD, MSc, MPH, FRCSC,a,b,c and Carol J. Swallow, MD, PhD, FRCSC, FACS,a,b,c Toronto, Ontario, Canada

Background. The risk of port-site metastasis after laparoscopic removal of incidental gallbladder cancer was previously estimated to be 14–30%. The present study was designed to determine the incidence of port-site metastasis in incidental gallbladder cancer in the modern era (2000–2014) versus the historic era (1991–1999). We also investigated the site of port-site metastasis. Methods. Using PRISMA, a systematic review was conducted to identify papers that addressed the development of port-site metastasis after laparoscopic resection of incidental gallbladder cancer. Studies that described cancer-specific outcomes in $5 patients were included. A validated quality appraisal tool was used, and a weighted estimate of the incidence of port-site metastasis was calculated. Results. Based on data extracted from 27 papers that met inclusion criteria, the incidence of port-site metastasis in incidental gallbladder cancer has decreased from 18.6% prior to 2000 (95% confidence interval 15.3–21.9%, n = 7) to 10.3% since then (95% confidence interval 7.9–12.7%, n = 20) (P < .001). The extraction site is at significantly higher risk than nonextraction sites. Conclusion. The incidence of port-site metastasis in incidental gallbladder cancer has decreased but remains high relative to other primary tumors. Any preoperative finding that raises the suspicion of gallbladder cancer should prompt further investigation and referral to a hepato-pancreato-biliary specialist. (Surgery 2016;j:j-j.) From the Division of General Surgery, Department of Surgery,a and the Institute of Medical Science,b University of Toronto; the Lunenfeld-Tanenbaum Research Institute,c Mount Sinai Hospital; the Library and Information Services,d University Health Network; and the Institute of Health Policy, Management, and Evaluation,e University of Toronto, Toronto, Ontario, Canada

THE LANDSCAPE OF OPERATIVE MANAGEMENT of gallbladder disease was significantly changed after the advent of laparoscopy. In the United States, the proportion of cholecystectomies performed laparoscopically increased from 2.5% to 76.6% for elective cases and from 0.7% to 67.5% for urgent cases between the years 1988 and 1997.1 Though some surgeons expressed concerns about insufficiently mature and objective data regarding complications and unexpected sequelae, this rapid Presented at the Academic Surgical Congress 2016, Jacksonville, FL, February 4, 2016. Accepted for publication August 5, 2016. Reprint requests: Carol J. Swallow, MD, PhD, FRCSC, FACS, Division of General Surgery, University of Toronto, and Mount Sinai Hospital, 600 University Avenue #1225, Toronto, Ontario, Canada. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2016 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2016.08.007

adoption was fuelled largely by improved patientreported outcomes.2 Shortly after the initial application of the laparoscopic approach to cancer resection in the 1990s, there were an alarming number of early reports of recurrence within an access port wound, coined port-site metastasis (PSM). This pattern of recurrence has now been observed in gastrointestinal, gynecologic, urologic, and thoracic malignancies.3 Reports of PSM after laparoscopic resection of gallbladder cancer seemed to be especially frequent, prompting a literature review by Paolucci4 in which he attempted to estimate the incidence of PSM in gallbladder cancer case series and reports published up to January 2000. His review of 4 international surveys plus 75 case reports concluded that PSM occurred in 14–30% of patients after laparoscopic cholecystectomy for presumed benign conditions, in whom gallbladder cancer was discovered postoperatively (so-called incidental gallbladder cancer [IGBCA]). The SURGERY 1

ARTICLE IN PRESS 2 Berger-Richardson et al

incidence of IGBCA in gallbladders removed laparoscopically is in turn estimated to be 0.2–3.3%.5 In the United States, approximately 3,700 new cases of gallbladder cancer are documented annually, 50% of which are diagnosed as IGBCA after laparoscopic cholecystectomy.6,7 This represents a large cohort of patients who are potentially at risk for PSM. This is important because the development of PSM is an indicator of poor prognosis and may contribute to inferior outcomes. In a study of 113 patients with IGBCA, median survival in patients with T2/T3 disease who developed PSM was 17 months, compared to 42 months in similar patients without PSM.8 The presence of PSM may also contribute to the burden of intractable symptoms in such patients. No cases of PSM after resection of IGBCA had been reported prior to 1991. Paolucci’s review4 of cases that had been described in papers published between 1991 and 1999 has been frequently cited and was important in raising awareness of the high incidence of PSM in gallbladder cancer, possibly precipitating changes in operative technique (eg, routine use of a specimen retrieval bag, avoidance of bile spillage). It was not clear whether the incidence of PSM subsequently declined. In the present systematic review, we sought to determine whether the incidence of PSM has changed from the historic era (1991–1999) to the modern era (2000–2014). We also examined the relative incidence of extraction and nonextraction site PSMs, a comparison that might provide a mechanistic clue to inform the development of future strategies to reduce the risk of PSM. METHODS With the collaboration of a Health Sciences library information specialist (M.E.), search strategies were designed to retrieve all published articles and abstracts related to wound recurrence after cancer resection. Five electronic databases were searched from inception until August 14, 2014, limited to human studies published in English: Cochrane Central Register of Controlled Trials, Cochrane Database of Systematic Reviews, Embase, Medline, and Medline In-Process. The detailed search strategy can be found online in Supplementary Figure 1. The search strategy was validated with 10 test articles that were independently identified while manually surveying the literature. The titles and abstracts yielded by the search were screened by 2 independent reviewers (D.B.R., T.R.C.), and only those related to wound recurrence in gallbladder cancer were retained. The full-text manuscripts of retained studies were obtained for comprehensive

Surgery j 2016

review. If an article cited another publication that described wound recurrence in gallbladder cancer that had not been captured by our search strategy, we included the latter for analysis if it met inclusion criteria. If only the abstract of a study was available, it was retained if the relevant data were provided in the abstract itself. Although the search was limited to English language publications, some non-English articles had English abstracts and were captured by our search strategy; if the relevant data were provided in English, the study was retained. We sought to compare the incidence of PSM after laparoscopic resection of unsuspected gallbladder cancer reported in papers published in the historic era (1991–1999) and the modern era (2000–2014). For this purpose, the inclusion criteria were English language papers that reported the presence or absence of PSM following a minimum of 5 laparoscopic cholecystectomies in which the gallbladder was found to harbor an unsuspected adenocarcinoma. Papers that explicitly reported no PSM after laparoscopic cholecystectomy in the setting of IGBCA were included in the analysis of PSM incidence. Multiple publications by the same author(s) were closely reviewed to eliminate duplicate cases. Individual cases that were converted from laparoscopic to open resections were excluded from subsequent analysis. Within each manuscript, only cases of unsuspected gallbladder cancer were retained for estimating the incidence of PSM. In addition to reporting a raw, pooled incidence of PSM in IGBCA, to compensate for the different levels of quality among manuscripts, we used a validated quality appraisal tool for case series.9 We included 18 of the 20 elements of the checklist for our analysis and excluded 2 that were not relevant to case series reporting the incidence of PSM in IGBCA. This quality appraisal tool does not suggest a cut-off to distinguish high- from low-quality papers; interpretation of results is left to the tool user. Therefore, we separated the 18 elements into 3 tiers according to their importance in assessing quality for manuscripts reporting on PSM in IGBCA (Table I). Consensus agreement was reached among all authors as to which elements would be included in each tier. Two independent reviewers (D.B.R., T.R.C.) reviewed the manuscripts that met inclusion criteria and determined which checklist elements were present in the text. Discrepancies were resolved by discussion. Points were then awarded to each manuscript to reflect the number of checklist elements that had been included, with

ARTICLE IN PRESS Berger-Richardson et al 3

Surgery Volume j, Number j

Table I. Modified quality assessment tool for case series Criteria Tier 1: each element worth 3 points Was the study conducted prospectively? Were patients recruited consecutively? Were the relevant outcomes measured using appropriate objective/subjective methods? Was follow-up long enough for important events and outcomes to occur? Tier 2: each element worth 2 points Were the cases collected in more than one center? Were the characteristics of the patients included in the study described? Were additional interventions (cointerventions) clearly described? Were relevant outcome measures established a priori? Were losses to follow-up reported? Tier 3: each element worth 1 point Were the eligibility criteria (ie, inclusion and exclusion criteria) for entry into the study clearly stated? Did patients enter the study at a similar point in the disease? Was the hypothesis/aim/objective of the study clearly stated? Was the intervention of interest clearly described? Were outcome assessors blinded to the intervention that patients received? Were the relevant outcome measures made before and after the intervention? Did the study provide estimates of random variability in the data analysis of relevant outcomes? Were the adverse events reported? Were both competing interests and sources of support for the study reported? Excluded criteria Were the conclusions of the study supported by the results? Were the statistical tests used to assess the relevant outcomes appropriate?

each element worth 1–3 points depending on its assigned tier (Table I). For the purpose of determining the location of PSM (extraction port versus nonextraction port), inclusion criteria were English language papers that specified the location of PSM after laparoscopic cholecystectomy, with no limitation on publication date or number of cases reported, and including any method of PSM detection. If it could not be determined whether the PSM was at an extraction or nonextraction port, the case was excluded from analysis. Additionally, when explicitly described, survival data on patients who developed PSM after resection of IGBCA were extracted to investigate their prognostic significance. RESULTS The initial search for all published articles and abstracts related to wound recurrence after cancer resection yielded 5,945 abstracts. Of these, 4,211 were identified by both reviewers as having no relevance to wound recurrence after a cancer operation, and a further 528 were identified as duplicates. Of the remaining 1,206 abstracts, 108 were identified by both reviewers as relevant to gallbladder carcinoma, and the corresponding articles were reviewed in full (Fig). A review of

the literature cited within these 108 articles yielded an additional 31 publications suitable for full review, totaling 139 for our library. Review articles that captured no additional cases beyond the primary sources were excluded from further analysis, though they were useful for cross-referencing purposes. A total of 89 papers met inclusion criteria for analysis (Supplementary References). Incidence of port-site metastasis in incidental gallbladder cancer. Of the retained papers, 7 published in the historic era (up to January 2000) and 20 published in the modern era (January 2000 to August 2014) met the additional inclusion criteria for estimating the incidence of PSM in IGBCA. Details of the parameters extracted from these 27 papers are shown in Table II. The incidence of PSM reported in the 7 individual historic era papers ranged from 0–43%. Combining the data, there were 93 cases of PSM in 524 cases of IGBCA diagnosed after laparoscopic cholecystectomy, yielding a pooled, estimated incidence of 17.7% (95% confidence interval [CI] 14.4–21.0). The quality scores assigned to each individual study are shown in Supplementary Table I. The incidence of PSM in IGBCA documented in each paper was multiplied by the quality score of the paper; these results were summed and

ARTICLE IN PRESS 4 Berger-Richardson et al

Surgery j 2016

Databases searched for wound recurrence in cancer surgery: 5945 papers identified CENTRAL: 22 CDSR: 58 Embase: 2816 Medline: 2756 Medline In-Process: 293

Papers for review of titles and abstracts: 5945

Included Cross-referenced articles: 31

Excluded No relevance to gallbladder cancer by title and abstract review or full article when uncertain: 1098

Papers for review of full text: 139

Papers with original case(s) PSM and/or explicit mention of absence of PSM in GBCA: 89

Papers published before Paolucci’s 1999 report: 45 Papers meeting inclusion criteria for estimate of historic incidence: 7 1- Article published before January, 2000 2- English language 3- Human patients 4- Pathology: Gallbladder adenocarcinoma 5- Articles reporting the incidence of PSM following a minimum of 5 laparoscopic cholecystectomies.

Excluded Duplicates: 528 No relevance to wound recurrence: 4211

Papers with possible relevance to wound recurrence: 1206

Excluded No original case(s) of PSM or no explicit mention of absence of PSM in GBCA: 50

Papers published after Paolucci’s 1999 report: 44

Location of PSM Extraction vs Non-extraction Port

Incidence of PSM (historic)

Papers meeting inclusion criteria for estimate of modern incidence: 20 1- Article published after 1999 2- English language 3- Human patients 4- Pathology: Gallbladder adenocarcinoma 5- Articles reporting the incidence of PSM following a minimum of 5 laparoscopic cholecystectomies.

Incidence of PSM (modern)

Fig. PRISMA flowchart of systematic review. Flow chart illustrating the selection of studies reporting on PSM after laparoscopic resection of incidental gallbladder cancer.

then divided by the sum of quality scores. This yielded a quality weighted incidence of PSM in IGBCA of 18.6% (95% CI 15.3–21.9%) in the historic era. The incidence of PSM reported in the 20 individual modern era papers ranged from 0–38%. Combining the data, there were 62 cases of PSM in 622 cases of IGBCA diagnosed after laparoscopic cholecystectomy, yielding a pooled estimated incidence of PSM in IGBCA of 10.0% (95% CI 7.6–12.4%). Using the methods described above, the quality weighted incidence of PSM in IGBCA was found to be 10.3% (95% CI 7.9–12.7%) in the modern era. Location of PSM after laparoscopic resection of incidental gallbladder cancer. The location of PSM can be classified as (1) the extraction port, which could imply direct seeding of the wound by the specimen, or (2) nonextraction port(s), which would imply other mechanisms. To determine the distribution of PSM, all 89 papers in which PSM in gallbladder cancer was reported as an outcome were included. The majority of these papers were case reports, and the year of publication ranged from 1991 to 2014. More than 305 individual port sites have been reported to harbor recurrence in 236 patients with

gallbladder cancer (number of PSM/person = 1–4). Of the 190 individual port sites with a location decipherable as extraction or nonextraction, 101 (53%) occurred at extraction ports compared to 89 (47%) at nonextraction ports. Assuming that a typical laparoscopic cholecystectomy would have 1 extraction and 3 nonextraction ports, extraction ports appear to be at increased risk. In only 57 of the patients with PSM after resection of IGBCA was it explicitly stated whether the extraction port was involved and how many of the nonextraction ports were involved. Of these, 36 patients (63%) had isolated extraction site recurrences, 15 patients (26%) had both extraction and nonextraction port recurrences (1–3 nonextraction ports/ patient), and only 6 (11%) had nonextraction site recurrences in 1–3 instrument ports but no extraction port recurrence. This distribution highlights the increased risk of PSM at extraction compared to nonextraction sites. In these 57 patients, assuming a standard 4-port set-up, 51 of 57 extraction ports were sites of tumor seeding (90%) compared to 33 of 171 (19%) nonextraction ports. Survival data. In the present systematic review, the outcome of 81 patients with PSM after

First author

Early GBCA* (n)

Advanced GBCA* (n)

Median follow-up (mo)

Patients with PSMy (n)

Incidence of PSM (%)

1995 1996 1997 1997 1998 1999 1999

7 14 13 21 23A 409 37 524

6 12 12 25 55 (77.5%)

1 2 1 12 16 (22.5%)

11 Mean 27 24 11 Mean 27 -

3 0 1 6 4 70 9 93

43 0 8 29 17 17 24 17.7

2000 2000 2000 2000 2002 2002 2003 2003 2004 2005 2005 2006 2006 2010 2011 2012 2013 2013 2013 2014

7 8 19 40B 27 10 142C 7 7 8 14 13 10 22 96 9 54 5 10 114 622

5 7 17 40 27 10 114 6 7 8 20 62 9 28 360 (79.3%)

2 1 2 1 0 0 25 1 0 0 2 34 0 26 94 (20.7%)

18 59 15 31 68 20 27 25 21 60 54 25 32 24 60 26 Mean 23 20 -

1 1 2 4 2 2 10 0 0 3 0 0 1 1 14 1 1 0 3 16 62

14 13 11 10 7 20 7 0 0 38 0 0 10 5 15 11 2 0 30 14 10.0

*Early GBCA, Tis–T2; Advanced GBCA, T3–4. yNumber of patients who developed PSM during follow-up. A Excludes 14 cases converted from laparoscopic to open resection. B Excludes one case in which the cancer was recognized intra-operatively. C Includes 3 cases with unknown T stage. LC, Laparoscopic cholecystectomy; -, not reported.

Berger-Richardson et al 5

LC for IGBCA (n)

ARTICLE IN PRESS

Historical era Wibbenmeyer10 Yamaguchi11 Mori Ricardo12 Z’Graggen Paolucci13 Lundberg14 Total Modern era Romano Yoshida15 Frauenschuh16 Suzuki30 Wakai17 Weiland18 Paolucci19 Yano Yeh Cucinotta20 Sun21 Chan22 Giuliante23 Kim24 Maker8 Cavallaro Fuks Gumbs Hu25 Kalayarasan26 Total

Year

Surgery Volume j, Number j

Table II. Studies that meet inclusion criteria for estimating the incidence of PSM in IGBCA: comparison of historical and modern eras

ARTICLE IN PRESS 6 Berger-Richardson et al

resection of IGBCA was ascertained (Supplemental references1-4,7,10-14,16,19,22,24-25,27,29,32,34,36-39,41,43,4849,51-53,55-56,62-63,65,68,71-72,78-79,81-82,84,88 ). A total of 57 (70%) died of cancer during follow-up, at a median of 13 months (range 1–36). In the group for whom date of death was known (n = 53), 49 (92%) died within 2 years of laparoscopic cholecystectomy. In the group of patients who were alive at a minimum follow-up of 2 years (n = 16), 14 had no evidence of disease recurrence after resection of the involved port site(s) at 31–82 months post cholecystectomy.10,14,26-34 Interestingly, one of these long-term survivors had residual tumor in the liver bed at reresection and another had a metastatic lymph node in the axilla that was resected.27,32 The possibility of achieving a long, disease-free interval after resection of PSM suggests it is reasonable to consider aggressive operative control in selected cases. However, in the vast majority of patients, PSM is associated with concurrent or future diffuse peritoneal disease and early death.30 DISCUSSION Laparoscopy has been widely adopted as an approach to resection of many types of cancer. Early concern about PSM led to comparative studies that examined the incidence of wound recurrence in open surgery versus laparoscopic surgery. In the Clinical Outcomes of Surgical Therapy study, 872 patients with colon cancer were randomized to either a laparoscopic or open resection, and the incidence of wound recurrence in both arms was <1% after a median follow-up of 4.4 years.35 Although oncologic outcomes in open and laparoscopic resection may be equivalent for some cancer types/sites, the current recommendations for resection of known gallbladder cancer stipulate an open approach.36 This is based largely on expert opinion, rather than evidence of superior outcomes. Interestingly, the incidence of wound recurrence following open cholecystectomy for preoperatively known or incidental GBCA is approximately 7% (Supplementary Table II),12,19,20,22,25,37 which approaches the incidence of PSM after laparoscopic cholecystectomy for IGBCA in the modern era we describe here (10.3%). This reinforces the adverse biology of GBCA and the necessity to better understand the molecular determinants of peritoneal and wound recurrence. When the diagnosis of gallbladder cancer is made preoperatively, curative resection should consist of cholecystectomy with resection of the adjacent liver segments to obtain an R0 resection and lymphadenectomy based on evidence that completion radical

Surgery j 2016

cholecystectomy has a survival advantage over simple cholecystectomy without reresection in $T1b gallbladder cancer.17,36,38-40 Simple laparoscopic cholecystectomy uses the subserosal layer as the dissection plane on the liver side; for lesions >T1a, there is a risk of residual malignant cells without completion radical cholecystectomy.41 In a study of 115 patients who underwent completion radical cholecystectomy and lymphadenectomy, 46% of patients had residual disease, most of whom had >T1 gallbladder cancer.42 Moreover, the risk of lymph node metastasis in T1a cancers is <3% but increases significantly with lesions $T1b.43 There is, however, a significant challenge in identifying gallbladder cancer preoperatively. The symptoms typically mimic benign disease, leading to incidental discovery of malignancy upon pathologic examination of specimens removed laparoscopically.30 Indeed IGBCA accounts for approximately 50% of newly diagnosed cases of GBCA in the United States each year.7,39 Similarly, in a series of 196 consecutive patients treated for GBCA at the University of Munich, 50% had a preoperative diagnosis of GBCA, while the remaining cases were identified either intraoperatively (29%) or incidentally postoperatively (21%).44 Intraoperative suspicion of GBCA should prompt an immediate reconsideration of the planned procedure; if this occurs early enough, the procedure should be aborted, with dissection minimized. Appropriate investigation and referral can then be undertaken. If suspicion of GBCA arises after the dissection is complete, the pathologist should be alerted in order to process the specimen accordingly. The development of PSM in the setting of IGBCA is associated with poor prognosis. We attempted to determine whether there has been a change in the incidence of PSM in IGBCA since the time of a previous review reporting on literature published through January 2000. In papers published in the historic era (1991–1999) that met inclusion criteria, the estimated incidence of PSM in IGBCA was 17.7%. When corrected for study quality, the estimate was 18.6% (95% CI 15.3–21.9%). In papers published from January 2000 through August 2014 that met inclusion criteria, the uncorrected estimated incidence was 10.0%, with a quality-adjusted estimate of 10.3% (95% CI 7.9–12.7%). On this basis, it appears that the incidence of PSM in IGBCA has decreased by approximately 45% but remains high when compared to other primary cancer types. Whether the decline is attributable to better preoperative recognition, improved precautionary operative techniques, or other variables is challenging to ascertain. Laparoscopic equipment has evolved, and modern trocars

ARTICLE IN PRESS Surgery Volume j, Number j

are less prone to physical displacement and gas leakage, which could reduce local wound exposure. One factor that may contribute to the high incidence of PSM in IGBCA is that wound protection measures may not be routinely employed when the surgeon does not expect malignancy. Of note, the recommended methods to minimize wound tumor seeding in this setting (avoiding bile spillage,15,18,45 use of retrieval bags, port site excision) are not strongly supported in the literature. According to a retrospective survey of 409 cases of IGBCA, 59 of 70 patients (84%) who developed PSM had an intact, unopened gallbladder during the procedure and during extraction.4 While retrospective identification of intraoperative bile spillage is undoubtedly inaccurate, this suggests that bile spillage is probably not the leading cause of PSM. The routine use of retrieval bags has been proposed as a method to prevent dissemination of malignant cells upon extraction.11,16,27,33,45-49 However, the present review demonstrates that ;50% of PSM occur at nonextraction ports, and many cases of PSM have been reported despite the use of a retrieval bag.13,50 The recommended management of port sites after laparoscopic cholecystectomy for what proves to be an IGBCA has evolved from an initial recommendation for routine port site excision (PSE) to prevent PSM. Of 113 patients with IGBCA managed at Memorial Sloan-Kettering Cancer Center, 69 underwent “prophylactic” PSE and 44 did not. Although the discovery of malignant cells within the PSE specimen was of prognostic significance, PSE was not associated with improved survival or reduced recurrence.8 Perhaps related to that experience, routine PSE for IGBCA has fallen out of favor and is not recommended in the 2014 consensus statement issued by the American Hepato-Pancreato-Biliary Association.36 However, if a patient does develop a PSM in the absence of diffuse, unresectable disease, port site metastasectomy has been associated with the potential for long-term, recurrence-free survival.10,14,26-34 For cases of PSM in gallbladder cancer described since 1991, an equivalent number were reported at extraction and nonextraction ports (53% and 47% respectively, n = 190 PSM). However, for each patient, there would typically be 1 extraction port plus 3 nonextraction ports at risk, suggesting that extraction ports are in fact at relatively higher risk. Nevertheless, the large number of PSM at port sites that had not been in contact with the specimen suggests that mechanisms other than direct contact between the specimen

Berger-Richardson et al 7

and wound are involved. Given that no definite mechanism(s) contributing to PSM has been identified, it is not surprising that the incidence of PSM in gallbladder cancer remains high despite increased awareness of this pattern of recurrence. It has been suggested that PSM is not related to operative technique but rather a manifestation of disseminated disease.8 GBCA is associated with aggressive tumor biology, which may explain its propensity to recur in wounds. The generally poor prognosis is illustrated by 10 years of experience with GBCA at Memorial Sloan-Kettering Cancer Center (n = 435), in which 37% of patients presented with American Joint Committee on Cancer stage IV disease, while 16% had unresectable disease without distant metastasis. Median survival in the entire cohort was 10.3 months; for stage Ia–III, it was 12.9 months, and for stage IV, 5.8 months.7 In a review of 83 case reports of PSM in gallbladder cancer, however, only 25% were associated with peritoneal carcinomatosis and 6% had concurrent liver metastasis.4 The frequency of PSM as the lone site of recurrence suggests that it may well be an iatrogenic event. Future studies investigating possible mechanisms of PSM in gallbladder cancer are warranted. These include local tissue factors within wounds that may predispose to tumor seeding, effects of pneumoperitoneum on cancer cell dissemination, and surgical gloves and instruments that may harbor malignant cells capable of seeding wounds. The main limitation of this review is the generally poor quality of papers reviewed. The development of PSM after resection of IGBCA was usually not the primary outcome of interest but was included in these manuscripts as a secondary outcome. Furthermore, accurately estimating the incidence of PSM in IGBCA relies heavily on physician reporting and publication. In the era we defined as historic, the literature was dominated by case reports, whereas in the modern era, these are much less common and case series predominate. This heterogeneity reflects submission and publication biases that favor the unexpected; as novelty declines, higher quality is required to achieve publication status. Considering that well-designed studies on this topic are limited, we wanted to be inclusive and so included manuscripts following as few as 5 patients with IGBCA. Sixteen of 27 manuscripts followed <20 patients. The duration of follow-up was stated in 23 of 27 papers (median 11–68 months); however, the method/interval of follow-up was described in only 8 of 27 papers. Four of these described the frequency and modality of surveillance imaging, 3 used patient interviews for

ARTICLE IN PRESS 8 Berger-Richardson et al

follow-up data collection,18,21,25 and 1 contacted the treating physician to obtain follow-up data.30 It is likely that if patients were systematically followed with physical exam and routine imaging at prescribed intervals, more cases of PSM would be discovered. Furthermore, many patients die of disease within a few months after resection of gallbladder cancer and may have developed PSM that was not recognized. Thus the pooled and quality-weighted estimates of the incidence of PSM may well be underestimates. In patients with recurrent colorectal cancer as a cause of death, autopsy studies have shown wound recurrence rates as high as 16.6%, notably higher than that detected antemortem.9 As a further illustration of this underestimation, in 1,711 patients who were followed prospectively after colorectal cancer resection, 11 cases of wound recurrence were identified, but only 4 of the cases were detected clinically, whereas 7 were found at relaparotomy.51 Similarly, in a study comparing patients undergoing routine PSE following diagnosis of IGBCA and patients undergoing observation without routine PSE, PSM was identified in 19% of patients in the routine PSE group (n = 69) compared to 5% in the observation group (n = 44), showing that the incidence of PSM is higher when it is intentionally looked for early on than when observed until clinically apparent.8 Another source of heterogeneity was the T stage of gallbladder cancer included in the individual studies; some included patients with Tis (carcinoma in situ),16,20,22,24,30 and 4 studies included only cases of early stage GBCA (T1–T2) and categorically excluded cases of T3 or greater.18,21-23 However, as shown in Table II, when decipherable, the proportion of early GBCA and advanced GBCA was similar in the historic and modern periods (early GBCA 77.5% and 79.3%, respectively; advanced GBCA 22.5% and 20.7%, respectively). Thus, the pathologic stage of IGBCA is unlikely to be responsible for the apparent decrease in risk of developing PSM in recent years. Furthermore, when data are extracted from studies in which stage is clearly reported for each case, we find that 10% of patients with early IGBCA developed PSM (n = 266) compared to 13% of patients with advanced IGBCA (n = 77) (P = .46, v2); thus, there is no clear correlation between T stage and risk of PSM in IGBCA. We concede that our search strategy may have failed to identify some relevant articles. However, we were thorough with cross-referencing, and 17 of 89 included articles (19%) were identified in this way. Despite these limitations, to our knowledge, this is

Surgery j 2016

the most comprehensive assessment of PSM in IGBCA reported to date. We acknowledge the limitations of the existing literature, and through the use of a quality appraisal tool, we provide the best possible estimate of the incidence of PSM in IGBCA. In conclusion, occasionally, a gallbladder removed laparoscopically for a presumed benign condition may unexpectedly harbor gallbladder cancer. The estimated incidence of PSM in this setting was approximately 18.6% in the early years of laparoscopy and is now approximately 10.3%. The decreased incidence of PSM in IGBCA in the modern compared to the historic era, taken together with the increased risk of PSM at extraction compared to nonextraction ports, support the hypothesis that PSM is an iatrogenic event with a modifiable risk. While PSM can be avoided if cases of gallbladder cancer are diagnosed preoperatively and patients undergo an open resection, GBCA is still associated with a relatively high incidence of wound recurrence. Early gallbladder cancer is difficult to diagnose radiologically, and it is important to have an experienced radiologist diligently review the images in suspicious cases. Surgeons should have heightened awareness and diligently investigate patients with atypical features (eccentric gallbladder wall thickening, diffuse gallbladder wall thickening in the absence of pain, etc) prior to operative intervention to minimize unplanned laparoscopic resection of GBCA. SUPPLEMENTARY DATA Supplementary data related to this article can be found online at http://dx.doi.org/10.1016/j.surg.2016.08.007. REFERENCES 1. Kemp JA, Zuckerman RS, Finlayson SR. Trends in adoption of laparoscopic cholecystectomy in rural versus urban hospitals. J Am Coll Surg 2008;206:28-32. 2. Schirmer BD, Edge SB, Dix J, Hyser MJ, Hanks JB, Jones RS. Laparoscopic cholecystectomy. Treatment of choice for symptomatic cholelithiasis. Ann Surg 1991;213:665-76 discussion 77. 3. Reymond MA, Bonjer HJ, K€ ockerling F. Port-site and wound recurrences in cancer surgery: incidence, pathogenesis, prevention. Berlin (Germany): Springer; 2000. 4. Paolucci V. Port site recurrences after laparoscopic cholecystectomy. J Hepatobiliary Pancreat Surg 2001;8:535-43. 5. Sujata J, S R, Sabina K, Mj H, Jairajpuri ZS. Incidental gall bladder carcinoma in laparoscopic cholecystectomy: a report of 6 cases and a review of the literature. J Clin Diagn Res 2013;7:85-8. 6. Henley SJ, Weir HK, Jim MA, Watson M, Richardson LC. Gallbladder cancer incidence and mortality, United States 1999-2011. Cancer Epidemiol Biomarkers Prev 2015;24: 1319-26. 7. Duffy A, Capanu M, Abou-Alfa GK, Huitzil D, Jarnagin W, Fong Y, et al. Gallbladder cancer (GBC): 10-year experience

ARTICLE IN PRESS Surgery Volume j, Number j

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

at Memorial Sloan-Kettering Cancer Centre (MSKCC). J Surg Oncol 2008;98:485-9. Maker AV, Oxenberg J, Butte J, Kuk D, Gonen M, Fong Y, et al. Is port site resection necessary in the surgical management of gallbladder cancer? Ann Surg Oncol 2011;18:S12. Moga CG, Schopflocher D, Harstall C. Development of a quality appraisal tool for case series studies using a modified Delphi technique. Edmonton (Alberta): Institute of Health Economics; 2012. Wibbenmeyer LA, Wade TP, Chen RC, Meyer RC, Turgeon RP, Andrus CH. Laparoscopic cholecystectomy can disseminate in situ carcinoma of the gallbladder. J Am Coll Surg 1995;181:504-10. Yamaguchi K, Chijiiwa K, Ichimiya H, Sada M, Kawakami K, Nishikata F, et al. Gallbladder carcinoma in the era of laparoscopic cholecystectomy. Arch Surg 1996;131:981-5. Ricardo AE, Feig BW, Ellis LM, Hunt KK, Curley SA, MacFadyen BV Jr, et al. Gallbladder cancer and trocar site recurrences. Am J Surg 1997;174:619-23. Paolucci V, Schaeff B, Schneider M, Gutt C. Tumor seeding following laparoscopy: International survey. World J Surg 1999;23:989-97. Lundberg O, Kristoffersson A. Port site metastases from gallbladder cancer after laparoscopic cholecystectomy. Results of a Swedish survey and review of published reports. Eur J Surg 1999;165:215-22. Yoshida T, Matsumoto T, Sasaki A, Morii Y, Ishio T, Bandoh T, et al. Laparoscopic cholecystectomy in the treatment of patients with gallbladder cancer. J Am Coll Surg 2000;191:158-63. Frauenschuh D, Greim R, Kraas E. How to proceed in patients with carcinoma detected after laparoscopic cholecystectomy. Langenbecks Arch Surg 2000;385:495-500. Wakai T, Shirai Y, Hatakeyama K. Radical second resection provides survival benefit for patients with T2 gallbladder carcinoma first discovered after laparoscopic cholecystectomy. World J Surg 2002;26:867-71. Weiland ST, Mahvi DM, Niederhuber JE, Heisey DM, Chicks DS, Rikkers LF. Should suspected early gallbladder cancer be treated laparoscopically? J Gastrointest Surg 2002;6:50-6; discussion 6-7. Paolucci V, Neckell M, Gotze T. [Unsuspected gallbladder carcinoma–the CAE-S/CAMIC registry]. Zentralbl Chir 2003;128:309-12; German. Cucinotta E, Lorenzini C, Melita G, Iapichino G, Curro G. Incidental gall bladder carcinoma: does the surgical approach influence the outcome? ANZ J Surg 2005;75: 795-8. Sun CD, Zhang BY, Wu LQ, Lee WJ. Laparoscopic cholecystectomy for treatment of unexpected early-stage gallbladder cancer. J Surg Oncol 2005;91:253-7. Chan KM, Yeh TS, Jan YY, Chen MF. Laparoscopic cholecystectomy for early gallbladder carcinoma: long-term outcome in comparison with conventional open cholecystectomy. Surg Endosc 2006;20:1867-71. Giuliante F, Ardito F, Vellone M, Clemente G, Nuzzo G. Port-sites excision for gallbladder cancer incidentally found after laparoscopic cholecystectomy. Am J Surg 2006;191: 114-6. Kim JH, Kim WH, Kim JH, Yoo BM, Kim MW. Unsuspected gallbladder cancer diagnosed after laparoscopic cholecystectomy: focus on acute cholecystitis. World J Surg 2010; 34:114-20. Hu L, Wang B, Liu X, Lv Y. Unsuspected gallbladder cancer: a clinical retrospective study. Arch Iran Med 2013;16:631-5.

Berger-Richardson et al 9

26. Kalayarasan R, Javed A, Sakhuja P, Agarwal A. Abdominal wall scar recurrence in incidental gallbladder cancer is not always disseminated disease. HPB (Oxford) 2014;16: 337-8. 27. Ohmura Y, Yokoyama N, Tanada M, Takiyama W, Takashima S. Port site recurrence of unexpected gallbladder carcinoma after a laparoscopic cholecystectomy: Report of a case. Surg Today 1999;29:71-5. 28. Nakagawa S, Tada T, Furukawa H, Abe M, Hatakeyama K. Late-type recurrence at the port site of unexpected gallbladder carcinoma after a laparoscopic cholecystectomy: Report of a case. Surg Today 2000;30:853-5. 29. Suzuki K, Kimura T, Hashimoto H, Nishihira T, Ogawa H. Port site recurrence of gallbladder cancer after laparoscopic surgery: two case reports of long-term survival. Surg Laparosc Endosc 2000;10:86-8. 30. Suzuki K, Kimura T, Ogawa H. Long-term prognosis of gallbladder cancer diagnosed after laparoscopic cholecystectomy. Surg Endosc 2000;14:712-6. 31. Tan YM, Lim DT. Gallstone embedded within a port site metastasis–report of a case. Singapore Med J 2002;43:637-9. 32. Povoski SP, Ouellette JR, Chang WW, Jarnagin WR. Axillary lymph node metastasis following resection of abdominal wall laparoscopic port site recurrence of gallbladder cancer. J Hepatobiliary Pancreat Surg 2004;11:197-202. 33. Ciulla A, Romeo G, Genova G, Tomasello G, Agnello G, Cstronovo G. Gallbladder carcinoma late metastases and incisional hernia at umbilical port site after laparoscopic cholecystectomy. G Chir 2006;27:214-6. 34. Piekarski JH, Kusinska R, Nejc D, Pluta P, Sek P, Bilski A, et al. Recurrence of cholangiogenous carcinoma in portsites two years after laparoscopic removal of noncancerous gallbladder. Eur J Gastroenterol Hepatol 2008;20:474-7. 35. Clinical Outcomes of Surgical Therapy Study Group. A comparison of laparoscopically assisted and open colectomy for colon cancer. N Engl J Med 2004;350:2050-9. 36. Aloia TA, Jarufe N, Javle M, Maithel SK, Roa JC, Adsay V, et al. Gallbladder cancer: expert consensus statement. HPB (Oxford) 2015;17:681-90. 37. Lundberg O, Kristoffersson A. Wound recurrence from gallbladder cancer after open cholecystectomy. Surgery 2000; 127:296-300. 38. Hardiman KM, Sheppard BC. What to do when the pathology from last week’s laparoscopic cholecystectomy is malignant and T1 or T2. J Gastrointest Surg 2009;13:2037-9. 39. Shih SP, Schulick RD, Cameron JL, Lillemoe KD, Pitt HA, Choti MA, et al. Gallbladder cancer: the role of laparoscopy and radical resection. Ann Surg 2007;245:893-901. 40. Ouchi K, Mikuni J, Kakugawa Y. Laparoscopic cholecystectomy for gallbladder carcinoma: results of a Japanese survey of 498 patients. J Hepatobiliary Pancreat Surg 2002;9:256-60. 41. Gallbladder cancer - Who needs reresection and to which extend after incidental finding at pathology. Eur J Surg Oncol 2010;36:847. 42. Pawlik TM, Gleisner AL, Vigano L, Kooby DA, Bauer TW, Frilling A, et al. Incidence of finding residual disease for incidental gallbladder carcinoma: implications for reresection. J Gastrointest Surg 2007;11:1478-86. 43. Townsend B, Evers M, Mattox KL, editors. Sabiston textbook of surgery the biological basis of modern surgical practice. Philadelphia (PA): Elsevier Saunders; 2012. p. 1508. 44. Lohe F, Meimarakis G, Schauer C, Angele M, Jauch KW, Schauer RJ. The time of diagnosis impacts surgical management but not the outcome of patients with gallbladder carcinoma. Eur J Med Res 2009;14:345-51.

ARTICLE IN PRESS 10 Berger-Richardson et al

45. Steinert R, Nestler G, Sagynaliev E, Muller J, Lippert H, Reymond MA. Laparoscopic cholecystectomy and gallbladder cancer. J Surg Oncol 2006;93:682-9. 46. Horvath LZ, Flautner LE, Tihanyi TF, Miklos IJ. Trocar site metastasis of gall bladder cancer after laparoscopic cholecystectomy. Minim Invasiv Ther 1996;5:193-6. 47. Martinez J, Taragona EM, Balague C, Pera M, Trias M. Port site metastasis. An unresolved problem in laparoscopic surgery. Int Surg 1995;80:315-21. 48. Nally C, Preshaw RM. Tumour implantation at umbilicus after laparoscopic cholecystectomy for unsuspected gallbladder carcinoma. Can J Surg 1994;37:243-4.

Surgery j 2016

49. Raznatovic ZJ, Zaric ND, Galun DA, Lekic NS, Micev M, Djordjevic VR, et al. Multiple port-site metastasis of incidental gallbladder carcinoma after laparoscopic cholecystectomy. Acta Chir Iugosl 2012;59:105-9. 50. Razzetta F, Borgonovo G, Cagnazzo A, Bianchi C, Mattioli FP. Laparoscopic cholecystectomy and gallbladder cancer: a diagnostic and therapeutic dilemma. Eur J Surg Oncol 1997;23:84-5. 51. Reilly WT, Nelson H, Schroeder G, Wieand HS, Bolton J, O’Connell MJ. Wound recurrence following conventional treatment of colorectal cancer. A rare but perhaps underestimated problem. Dis Colon Rectum 1996;39:200-7.