Ann Surg Oncol (2015) 22:811–818 DOI 10.1245/s10434-014-4044-4
ORIGINAL ARTICLE – HEPATOBILIARY TUMORS
Lymph Node Metastases in Patients Undergoing Surgery for a Gallbladder Cancer. Extension of the Lymph Node Dissection and Prognostic Value of the Lymph Node Ratio David Je´re´mie Birnbaum, MD1, Luca Vigano`, MD1,2, Nadia Russolillo, MD1, Serena Langella, MD1, Alessandro Ferrero, MD1, and Lorenzo Capussotti, MD1 Department of HPB and Digestive Surgery, Ospedale Mauriziano ‘‘Umberto I’’, Turin, Italy; 2Department of Hepato-biliary Surgery, Humanitas Research Hospital, Milan, Rozzano, Italy 1
ABSTRACT Background. Lymph node (LN) status is one of the strongest prognostic factors after gallbladder cancer (GBC) resection. The adequate extension of LN dissection and the stratification of the prognosis in N? patients have been debated. The present study aims to clarify these issues. Methods. A total of 112 consecutive patients who underwent operations for GBC with LN dissection were analyzed. Twenty-five patients (22.3 %) had D1 dissection (hepatic pedicle), and 87 (77.7 %) had D2 dissection (hepatic pedicle, celiac and retro-pancreatic area). The LN ratio (LNR) was computed as follows: number of metastatic LNs/number of retrieved LNs. Results. The median number of retrieved LNs was 7 (1– 35). Fifty-nine patients (52.7 %) had LN metastases (22 N2). D2 dissection allowed the retrieval of more LNs (8 vs. 3, p = 0.0007), with similar short-term outcomes. Common bile duct (CBD) resection (n = 41) did not increase the number of retrieved LNs. In five patients, D2 dissection identified skip LN metastases that otherwise would have been missed. LN metastases negatively impacted survival (5-years survival 57.2 % if N0 vs. 12.4 % if N?, p \ 0.0001), but N1 and N2 patients had similar survival rates. The number of LN? (1–3 vs. C4) did not impact prognosis. An LNR = 0.15 stratified the prognosis of N?
David Je´re´mie Birnbaum and Luca Vigano` contributed equally. Ó Society of Surgical Oncology 2014 First Received: 20 May 2014; Published Online: 9 September 2014 L. Vigano`, MD e-mail: [email protected]
patients: 5-years survival 32.7 % if LNR B 0.15 vs. 10.3 % if LNR [ 0.15 (multivariate analysis p = 0.007). Conclusions. A D2 LN dissection is recommended in all patients, because it allows for better staging. CBD resection does not improve LN dissection. An LNR = 0.15, not the site of metastatic LNs, stratified the prognoses of N? patients.
Gallbladder cancer (GBC) is the most common biliary tract malignancy worldwide.1 It presents with lymph node (LN) metastases in a high proportion of patients, up to 60– 80 % of T3–4 tumors.2–7 LN status is consistently one of the strongest predictors of a poor outcome after surgery, but the adequate extent of LN dissection is debated.2,5,7–9 Even among high-volume HPB centers, a large discrepancy in the number of retrieved LNs after GBC resection occurs.10–13 This is mainly related to the inclusion or omission of celiac and retropancreatic LNs in the dissection. In addition, the need for a systematic common bile duct (CBD) resection to optimize LN dissection is unclear.10,14 A further debated issue is the stratification of the prognoses of N? patients. The most common scoring system is based on the LN metastatic site.15 However, the prognostic value of the LN metastatic site has not been confirmed by all authors. The LN ratio (the ratio of the number of positive LNs to the number of retrieved LNs) has been recently proposed, as well as the number of harvested LNs and the number of positive LNs.11,16–18 The present study aims to elucidate these aspects. First, the authors compared two different LNs dissections (D1 vs. D2) in terms of the number of retrieved LNs, LN status assessments, and outcomes. The contribution of CBD resection to LN dissection also was evaluated.
812
Second, the different scoring systems for N? patients were compared to identify an optimal manner of stratifying prognoses.
D. J. Birnbaum et al.
Definitions
From January 1990 to December 2011, 126 consecutive patients underwent resections for GBC at the author’s institution. A total of 115 (91.3 %) patients received an LN dissection. Patients with the simultaneous diagnosis of other organ tumors (n = 2), and those undergoing R2 resections (n = 1) were excluded from the present analysis. Finally, 112 patients were included. The study was approved by the local ethical committee.
All of the patients were staged according to the 7th edition of TNM staging system.15 LN metastases confined to hepatic pedicle LNs were staged N1, whereas celiac or retropancreatic LN metastases were staged N2. Major hepatectomy was defined as the resection of C3 Couinaud segments. Operative mortality was defined as death within 90 days after surgery or before discharge from the hospital. Morbidity included all postoperative complications and was assessed according to the Dindo–Clavien classification.20 The LN ratio was computed as the ratio between the number of positive LNs and the number of retrieved LNs.
Preoperative Staging and Surgical Procedures
Statistical Analysis
Preoperative staging systematically included carcinoembryonic antigen (CEA) and Ca 19-9 values and thoracoabdominal computed tomography (CT). Since 2000, positron emission tomography (PET)-CT was performed in selected cases, e.g., patients with aggressive GBC to exclude distant metastases or patients with uncertain extrahepatic lesions or enlarged aortocaval LNs. Preoperative N staging did not include endoscopic ultrasonography. Surgery was indicated only if complete resection was achievable. Distant metastases, even if they were hepatic, contraindicated surgery. Since 1998, a diagnostic laparoscopy with laparoscopic ultrasonography was systematically performed before laparotomy to exclude peritoneal carcinomatosis and liver metastases. After laparotomy, aortocaval LNs were systematically sampled and sent for frozen section; if aortocaval LNs were metastatic the resection was not performed. The standard surgical procedure comprised the resection of segments 4b and 5, en bloc cholecystectomy, and LN dissection.19 LN dissection of the hepatic pedicle (D1) was the standard up to 1998; since 1999 LN dissection was extended to the celiac and retropancreatic area (D2).
All patient data were prospectively collected and retrospectively analyzed. Continuous variables were compared between groups by the unpaired t test or Mann–Whitney U test, as appropriate; categorical variables were compared by the Chi-square test or Fisher exact test, as appropriate. A p value \ 0.05 was considered significant for all tests. Patients with 90-days mortalities were excluded from the survival analysis. The multivariate analysis of prognostic factors of disease-specific overall survival (OS) was performed by including all variables significant or borderline significant (p \ 0.1) at univariate analysis into a stepwise regression model. To determine the most appropriate cutoff values for continuous data variables, such as the LN ratio and the number of positive LNs, the cut-point survival analysis was performed with an outcome-oriented approach.
METHODS
Adjuvant Therapy and Follow-up Since 1998, adjuvant chemotherapy was systematically considered in N? patients and in those with advanced GBC. Adjuvant radiotherapy was not delivered to GBC patients. All patients were reviewed every 3 months, including a physical examination, CEA and CA 19-9 levels, and abdominal ultrasonography or thoracoabdominal CT. No patient was lost during the follow-up period. The follow-up was updated to December 31, 2012.
RESULTS Between January 1990 and December 2011, 112 patients affected by GBC underwent surgical resection and LN dissection with curative intent. Table 1 summarizes the patient characteristics. Thirty-seven (33.0 %) patients were male; the median age was 64 years (range 32–83). The GBC diagnosis was made postoperatively after a simple cholecystectomy in 47 (42.0 %) patients. All of them had a reoperation including liver resection and LN dissection. Twenty-four (21.4 %) patients underwent major hepatectomy, eight after a preoperative portal vein embolization. Fifty-five patients (49.1 %) underwent an associated resection of the contiguous organs, including CBD resection in 41. The tumor staging was T3 in 61 (54.5%) and T4 in 16 (14.3%). Ten (8.9%) patients had an R1 resection (positive biliary margins in five patients with CBD resection).
Lymph Node Metastases in Gallbladder Cancers
813
TABLE 1 Patient characteristics and surgical details Patients (n = 112) n (%)
The median number of retrieved LNs was 7 (1–35), 3 (1–8) after D1 vs. 8 (3–35) after D2 (p = 0.0007). Among the D2 patients, the number of retrieved LNs was similar in patients with and without CBD resection (9 vs. 7, p = 0.095), while it was higher in those with associated pancreatoduodenectomy (14.5, p = 0.001).
Age (years)
64 (32–83)
Male sex
37 (33.0)
CEA (ng/mL)
2 (0.6–657)
CA 19-9 (U/mL)
27.5 (0.8–4778)
Short-term Outcomes
Cholecystectomy
2 (1.8)
Sg4b-5 resection Right hepatectomy [±Sg1]
86 (76.7) 8 (7.1)
Right hepatectomy ? Sg4 [±Sg1]
13 (11.6)
The 90-days mortality rate was 5.4 % (n = 6). The causes of death were as follows: hepatic failure in two patients, hemoperitoneum in one, sepsis in one, acute myocardial infarction in one, and acute respiratory distress syndrome in one. The overall morbidity rate was 47.3 % (n = 53), including 20 severe complications (grade III– IV). In details, severe complications included liver dysfunction in four patients, bile leak requiring interventional procedures in five, bile leak associated with hemoperitoneum requiring reoperation in four, abdominal collection requiring percutaneous drainage in two, duodenal leak associated with hemoperitoneum requiring reoperation in one, renal dysfunction in one, myocardial infarction in one, acute respiratory distress in one, and pleural effusion requiring percutaneous drainage in one. D1 and D2 patients had similar overall morbidity rates (40.0 vs. 49.4 %), severe morbidity rates (grade III–V complications, 16.0 vs. 25.3 %), and hospital stays (12.5 vs. 11 days). Patients undergoing standard surgical procedures (bisegmentectomy Sg4b-5 and LN dissection, n = 53) had no mortality and three grade III/IV complications (5.7 %). Patients requiring resection of contiguous organs (n = 55) had higher mortality (11.0 vs. 0 %, p = 0.027) and grade III/ IV morbidity rates (32.7 vs. 5.7 %, p = 0.0009). Among the patients requiring pancreatoduodenectomy (n = 10), there was one operative mortality (10 %) and no grade III–IV morbidities.
Surgical procedures
Left hepatectomy ? Sg5 [±Sg1]
3 (2.7)
Associated Sg1 resection
9 (8.0)
Associated resections Common bile duct
41 (36.6)
Pancreatoduodenectomy
10 (8.9)a
Right colectomy
4 (3.6)a
Distal gastrectomy
2 (1.8)
T stageb T1b
4 (3.6)
T2
31 (27.7)
T3
61 (54.5)
T4
16 (14.3)
Completeness of surgery R0 R1 Adjuvant chemotherapy
102 (91.1) 10 (8.9) 39 (34.8)
Continuous variables are reported as median value (range) CEA carcinoembryonic antigen, GBC gallbladder cancer, Sg segment a
Two patients had combined pancreatoduodenectomy and right colectomy
b
According to the 7th edition of the AJCC manual15
LN Dissection LN Status Table 2 summarizes the data on the LN dissections. A D2 was performed in 87 (77.7 %) patients. D1 and D2 patients had similar characteristics. To achieve complete LN excision, two patients required a CBD resection and three a pancreatoduodenectomy. The remaining indications for a CBD resection were as follows: suspect macroscopic neoplastic infiltration in 33 patients, noncontiguous neoplastic tissue along the hepatic pedicle in three, and cystic duct stump infiltration in three. Seven additional patients required a pancreatoduodenectomy because of macroscopic neoplastic infiltration of the duodenopancreatic area by GBC. All ten patients undergoing pancreatoduodenectomy had concomitant liver resection.
Table 2 summarizes the data on the LN status. Fiftynine patients (52.7 %) had LN metastases, including 22 (19.6 %) N2. In five patients, a noncontinuous LN metastases distribution was observed: LN metastases in the retropancreatic or celiac area without LN metastases in the hepatic pedicle. Among N? patients, the median number of positive LNs was 2 (1–28), and the median LN ratio was 0.33 (0.08–1). The LN ratio was similar in D1 and D2 patients (0.31 vs. 0.33). LN status correlated with the T stage: N? patients were 0 % in T1b group, 29.0 % in T2, 60.6 % in T3, and 81.3 % in T4.
814
D. J. Birnbaum et al.
TABLE 2 Details about LN dissection and N status Patients (n = 112) n (%)
Prognostic Impact of the LN Status and Stratification of Prognosis in N? Patients
Lymph node dissection D1
25 (22.3)
D2
87 (77.7)
Number of retrieved lymph nodes After D1* Lymph node staging N0
7 (1–35) 3 (1–8)*
After D2*
8 (3–35)* a
N?
53 (47.4) 59 (52.7)
N1
37 (33.0)
N2
22 (19.6)
Number of metastatic lymph nodes
2 (1–28)
Lymph node ratio Overall
0.08 (0–1)
In N? patients
0.33 (0.08–1)
Number of metastatic LNs 1
32 (28.6)
2–3
12 (10.7)
[3
15 (13.4)
Continuous variables are reported as median (range) * Number of retrieved lymph nodes after D1 vs. D2, p = 0.0007 a
Overall, the ten patients who required pancreatoduodenectomy had 0 % OS at 2 years after surgery.
According to the 7th edition of the AJCC manual15
Long-term Outcomes and Prognostic Impact of the LN Dissection After a median follow-up of 71.7 months, 3- and 5-years OS rates were 40.6 and 33.2 %, respectively (median OS 23.8 months). Twenty-three (20.5 %) patients survived more than 5 years, including three N1 and one N2 patients. Seventy-one patients (63.4 %) developed recurrences: hepatic only in 32, extrahepatic only in 34, and both hepatic and extrahepatic in 5. The recurrence site was unrelated to the LN status, LN metastatic site, and LN ratio value. D1 and D2 patients had similar prognoses (5-years OS, 26.2 vs. 38.2 %). The number of harvested LNs (1-6 vs. [6) did not impact OS (28.0 vs. 31.7 %). Two patients required CBD resections because of hepatic pedicle LN metastases. One is still alive and disease-free 81 months after resection, while the other died of disease 14 months after surgery. The CBD resection was associated with worse OS, both in the whole series (5-years OS 9.6 vs. 42.7 % in patients without CBD resection, p \ 0.0001) and in N0 patients (17.4 vs. 59.0 %, p = 0.017). The three patients who had pancreatoduodenectomies due to retropancreatic LN metastases died of disease at 8, 15, and 15 months after surgery, respectively.
LN metastases negatively impacted survival: 5-years OS 57.2 % in N0 patients vs. 12.4 % in N? ones (p \ 0.0001; Fig. 1a). Among N? patients, N1 and N2 patients had similar OS (15.3 % in N1 patients vs. 6.6 % in N2; Fig. 1a). The LN ratio stratified the prognoses in N? patients. The cut-point survival analysis identified 0.15 as the best cutoff value. The 5-years OS in patients with LN ratio 0.01–0.15 was 32.7 vs. 10.3 % in those with LN ratio [0.15 (p = 0.0001; Fig. 1b). The prognostic value of the LN ratio was independent from the extent of LN dissection (significant predictor of OS after both D1 and D2 dissection). An LN ratio 0.15 stratified survival outcomes, even in T3-4 tumors. The number of metastatic LNs (1 vs. 2–3 vs. [3) was not associated with the prognosis (5-years OS 13.7 % if one metastatic LN vs. 13.9 % if 2–3 vs. 16.7 % if [3, p = 0.641; Fig. 2). Prognostic Factors of OS: Multivariate Analysis Table 3 provides the results of univariate and multivariate analysis of the prognostic factors of OS after surgery. LN ratio, T stage, and R1 resection were independent prognostic factors of OS (p = 0.007, 0.008, and 0.016, respectively). The study period (1990–1998 vs. 1999–2011) did not impact OS. DISCUSSION LN metastases represent a crucial issue in the management of patients affected by GBC. First, LN metastases are common, and their rate increases along with the T stage, up to 60–80 % in T3–4.2–7 In the present series, 53 % of patients were N?, 65 % in T3–4 tumors. Second, LN metastases are one of the strongest prognostic determinants after surgery.2,5–9 In the present series, survival expectancy was drastically reduced in N? patients (5-years OS 12 vs. 57 % in N0 patients). Many issues about LN metastases in GBC still await clarification. The first of these is the extent of LN dissection. Two recent population-based U.S. studies demonstrated that, in the majority of surgical centers, GBC patients receive very limited or even no LN dissection.21,22 The AJCC recommends the analysis of at least three LNs to adequately assess the N stage.15 A higher cutoff also has been recently
Lymph Node Metastases in Gallbladder Cancers
a
1.0 0.9 0.8 0.7
% Overall Survival
FIG. 1 a Overall survival curves of patients undergoing operations for gallbladder cancer according to the N status (N0 vs. N?, p \ 0.0001; N1 vs. N2, p = 0.612). b Overall survival curves of patients undergoing operations for gallbladder cancer according to the lymph node (LN) ratio (p = 0.007)
815
0.6 0.5 0.4 0.3 0.2 0.1 0.0
0
12
24
36
48
60
48
60
Months N0
b
N2
N1
1.0 0.9 0.8
% Overall Survival
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
0
12
24
36
Months LN Ratio = 0 (N0 patients)
proposed: in N0 patients, survival was improved if at least six LNs were retrieved.12,16 Despite this theoretical agreement, the adequate extent of LN dissection remains controversial (D1 vs. D2). Even among high-volume HPB centers, a large discrepancy in the number of retrieved LNs can be observed. Pawlik et al. and Ito et al. reported a median of three LNs retrieved per patient, whereas Shirai et al. achieved 14 LNs and Sasaki et al. achieved 22.10–13 In the present series, both D1 and D2 dissections were performed, and their results were compared. Several data suggest that D2 should be the standard. First, the median number of LNs retrieved after D1 dissection was three, the
LN Ratio 0.01-0.15
LN Ratio > 0.15
minimum recommended by AJCC, but far from the number of six recommended by recent papers.12,16 Second, in five patients, skip metastases were detected, e.g., LN metastases in the retropancreatic or celiac area without LN metastases in the hepatic pedicle. These patients would have been classified as N0 after a D1 dissection. Finally, the D2 dissection did not increase operative risk. D2 LN dissection is part of an adequate radical oncological treatment of GBC. Some Japanese authors have recommended a systematic CBD resection in patients operated on for a GBC to improve LN dissection.14 The Makuuchi group even reported a survival benefit after CBD resection in patients
816
D. J. Birnbaum et al.
FIG. 2 Overall survival curves of patients undergoing operations for gallbladder cancer according to the number of metastatic lymph nodes (p = 0.641)
1.0 0.9 0.8
% Overall Survival
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
12
24
36
48
60
Months 1 metastatic LN
with perineural infiltration.14 These data have not been confirmed by the most recent papers.10,16 In the present series, the CBD resection did not modify the quality of LN dissection; exactly the same number of LNs was retrieved in patients with and without biliary resection. Patients with CBD resection had worse survival compared with those without it. However, in the present series, the CBD was resected if necessary to achieve a complete surgery. This population cannot be compared with the one reported by Makuuchi et al. (de principe resection). A randomized trial is advocated to clarify the prognostic value of the systematic CBD resection. In the present series, three patients required pancreatoduodenectomy to remove metastatic retropancreatic LNs. Seven additional patients had concomitant liver resection and pancreatoduodenectomy because of neoplastic infiltration of the duodenopancreatic area. The short-term outcomes were good, but survival results were extremely poor (0 % OS at 2 years). In the literature, a debate is ongoing about the benefit of pancreatoduodenectomy for GBC.23–26 According to present data, in the author center patients requiring a pancreatoduodenectomy for GBC are now excluded from surgery. A second major issue is the stratification of prognosis in N? patients. Different proposals have been put forth, as follows: the LN metastases site, the number of positive LNs, the LN ratio.11,15–18,27 The anatomical location of positive LNs is the most common solution and has been adopted by the AJCC (TNM 7th edition) and the Japanese Society of Biliary Surgery.15,27 The stratification of patient prognoses on the basis of LN? site may allow pre- and intraoperative predictions of outcomes but
2-3 metastatic LNs
>3 metastatic LNs
has several limitations: it depends on the extent of LN dissection; the farthest metastatic LN is assessed, but the burden of the lymphatic dissemination is not considered; the LN site can be difficult to determine on the specimen. Furthermore, recent studies did not confirm the prognostic value of LN metastatic sites, as the present series did (N1 and N2 patients had exactly the same OS).11,16–18 As an alternative, some authors have proposed the number of positive LNs.11,17,18 However, in the literature, a number of long-term survivors after resection for GBC with multiple positive LNs have been reported.23,26,28,29 In the present series, the number of positive LNs failed to stratify prognoses, and four 5-years survivors with multiple positive LNs were observed. Finally, the LN ratio can be considered, as it has been recently adopted in other gastrointestinal tumors.30–32 Negi et al. first reported its prognostic role in GBC patients over a small series of 57 patients.16 In a larger setting (112 patients), the present series confirmed the LN ratio as the best way to stratify prognosis of N? patients. The best cutoff value was 0.15. The LN ratio can be assessed only at the pathological examination but is independent from the extent of the LN dissection, considers the burden of the lymphatic dissemination, and does not require the identification of the LN site. Shirai et al. suggested that the LN ratio value decreases together with the extension of LN dissection.11 This was not confirmed in the present series: the LN ratio had the same prognostic value after both D1 and D2 dissections. Some limitations of the present study could be argued. First, it is a retrospective series, collecting 112 patients
Lymph Node Metastases in Gallbladder Cancers TABLE 3 Uni- and multivariate analysis of prognostic factors of overall survival Parameter
Overall survival Univ.
Multivariate analysis
p
p
RR (CI 95 %)
Age (years) B70
0.624
–
0.864
–
0.021
n.s.
[70 Sex Male Female Major hepatectomy N 0.002
n.s.
0.0001
0.008
1 2.561 (1.275–5.143)
\0.0001
0.007
1
N T stage
Lymph-node ratio 0 0.01–0.15
1.313 (0.551–3.127)
[0.15
2.627 (1.426–4.842)
Complete resection R0
0.001
0.016
R1
1 2.383 (1.175–4.834)
Grading 1–2
0.011
n.s.
0.450
–
0.098
n.s.
3–4 Adjuvant CTx Y N Study period 1990–1998
None to declare.
REFERENCES
Associated resection
1–2 3–4
benefit of surgery, especially in N2 patients. However, N2 patients had survival outcomes similar to N1 ones and superior to survival rates reported in the literature for unresected patients (median survival 4–7 months).1,33,34 In conclusion, N status is one of the strongest prognostic determinants in patients undergoing operations for GBC. A D2 LN dissection is recommended in all patients because it allows for better patient staging without worsening shortterm outcomes. A systematic CBD resection is not required to improve LN dissection. An LN ratio of 0.15, not the location or the number of positive LNs, is the best method of stratifying prognoses in N? patients. CONFLICTS OF INTEREST
Y
Y
817
1999–2011 Bold values are statistically significant Univ. univariate, RR relative risk, CTx chemotherapy, n.s. not significant
during a 20-years period. This limitation is common to all published series on GBC due to the rarity of the disease and its low resectability rate. Second, the extent of the LN dissection was modified during the study period. However, no additional major modifications in GBC management occurred in the two study periods. The two groups (D1 and D2) had similar characteristics, and the study period had no impact on OS. Finally, the outcomes of resected patients were not compared to those of unresected patients. This comparison is mandatory to definitively ascertain the
1. Misra S, Chaturvedi A, Misra NC, et al. Carcinoma of the gallbladder. Lancet Oncol. 2003;4:167–76. 2. Miyakawa S, Ishihara S, Horiguchi A, Takada T, Miyazaki M, Nagakawa T. Biliary tract cancer treatment: 5,584 results from the Biliary Tract Cancer Statistics Registry from 1998 to 2004 in Japan. J Hepatobiliary Pancreat Surg. 2009;16(1):1–7. 3. Tsukada K, Kurosaki I, Uchida K, Shirai Y, Oohashi Y, Yokoyama N, et al. Lymph node spread from carcinoma of the gallbladder. Cancer. 1997;80(4):661–7. 4. Shimada H, Endo I, Togo S, Nakano A, Izumi T, Nakagawara G. The role of lymph node dissection in the treatment of gallbladder carcinoma. Cancer. 1997;79(5):892–9. 5. Fong Y, Wagman L, Gonen M, Crawford J, Reed W, Swanson R, et al. Evidence-based gallbladder cancer staging: changing cancer staging by analysis of data from the National Cancer Database. Ann Surg. 2006;243(6):767–71. 6. Shirai Y, Wakai T, Hatakeyama K. Radical lymph node dissection for gallbladder cancer: indications and limitations. Surg Oncol Clin N Am. 2007;16(1):221–32. 7. Kondo S, Takada T, Miyazaki M, Miyakawa S, Tsukada K, Nagino M, et al. Guidelines for the management of biliary tract and ampullary carcinomas: surgical treatment. J Hepatobiliary Pancreat Surg. 2008;15(1):41–54. 8. Benoist S, Panis Y, Fagniez PL. Long-term results after curative resection for carcinoma of the gallbladder. French University Association for Surgical Research. Am J Surg. 1998;175(2):118–22. 9. Jensen EH, Abraham A, Jarosek S, Habermann EB, Al-Refaie WB, Vickers SA, et al. Lymph node evaluation is associated with improved survival after surgery for early stage gallbladder cancer. Surgery. 2009;146(4):706–11. 10. 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 re-resection. J Gastrointest Surg. 2007;11(11):1478–86. 11. Shirai Y, Sakata J, Wakai T, Ohashi T, Ajioka Y, Hatakeyama K. Assessment of lymph node status in gallbladder cancer: location, number, or ratio of positive nodes. World J Surg Oncol. 2012;10:87. 12. Ito H, Ito K, D’Angelica M, Gonen M, Klimstra D, Allen P, et al. Accurate staging for gallbladder cancer: implications for surgical therapy and pathological assessment. Ann Surg. 2011;254(2):320–5. 13. Sasaki E, Nagino M, Ebata T, Oda K, Arai T, Nishio H, et al. Immunohistochemically demonstrated lymph node micrometastasis and prognosis in patients with gallbladder carcinoma. Ann Surg. 2006;244(1):99–105.
818 14. Sakamoto Y, Kosuge T, Shimada K, Sano T, Hibi T, Yamamoto J, et al. Clinical significance of extrahepatic bile duct resection for advanced gallbladder cancer. J Surg Oncol. 2006;94(4):298– 306. 15. Edge SBB, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC cancer staging manual. 7th ed. New York, NY: Springer; 2010. p. 211–7. 16. Negi SS, Singh A, Chaudhary A. Lymph nodal involvement as prognostic factor in gallbladder cancer: location, count or ratio? J Gastrointest Surg. 2011;15(6):1017–25. 17. Endo I, Shimada H, Tanabe M, Fujii Y, Takeda K, Morioka D, et al. Prognostic significance of the number of positive lymph nodes in gallbladder cancer. J Gastrointest Surg. 2006;10(7):999– 1007. 18. Sakata J, Shirai Y, Wakai T, Ajioka Y, Hatakeyama K. Number of positive lymph nodes independently determines the prognosis after resection in patients with gallbladder carcinoma. Ann Surg Oncol. 2010;17(7):1831–40. 19. Muratore A, Amisano M, Vigano` L, et al. Gallbladder cancer invading the perimuscular connective tissue: results of reresection after prior non-curative operation. J Surg Oncol. 2003;83(4):212–5. 20. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240(2):205–13. 21. Coburn NG, Cleary SP, Tan JC, Law CH. Surgery for gallbladder cancer: a population-based analysis. J Am Coll Surg. 2008;207(3): 371–82. 22. Mayo SC, Shore AD, Nathan H, Edil B, Wolfgang CL, Hirose K, et al. National trends in the management and survival of surgically managed gallbladder adenocarcinoma over 15 years: a population-based analysis. J Gastrointest Surg. 2010;14(10): 1578–91. 23. Shirai Y, Ohtani T, Tsukada K, Hatakeyama K. Combined pancreaticoduodenectomy and hepatectomy for patients with locally advanced gallbladder carcinoma: long term results. Cancer. 1997;80(10):1904–9.
D. J. Birnbaum et al. 24. Kondo S, Nimura Y, Hayakawa N, Kamiya J, Nagino M, Uesaka K. Extensive surgery for carcinoma of the gallbladder. Br J Surg. 2002;89(2):179–84. 25. Araida T, Yoshikawa T, Azuma T, Ota T, Takasaki K, Hanyu F. Indications for pancreatoduodenectomy in patients undergoing lymphadenectomy for advanced gallbladder carcinoma. J Hepatobiliary Pancreat Surg. 2004;11(1):45–9. 26. Sasaki R, Itabashi H, Fujita T, Takeda Y, Hoshikawa K, Takahashi M, et al. Significance of extensive surgery including resection of the pancreas head for the treatment of gallbladder cancer -from the perspective of mode of lymph node involvement and surgical outcome. World J Surg. 2006;30(1):36–42. 27. Japanese Society of Biliary Surgery (JSBS). Classification of Biliary Tract Carcinoma. English ed. Tokyo: Kanehara & Co., Ltd; 2004. p. 33–50. 28. Shirai Y, Yoshida K, Tsukada K, Muto T, Watanabe H. Radical surgery for gallbladder carcinoma. Long-term results. Ann Surg. 1992;216(5):565–8. 29. Kondo S, Nimura Y, Kamiya J, Nagino M, Kanai M, Uesaka K, et al. Five-year survivors after aggressive surgery for stage IV gallbladder cancer. J Hepatobiliary Pancreat Surg. 2001;8(6):511–7. 30. Pawlik TM, Gleisner AL, Cameron JL, Winter JM, Assumpcao L, Lillemoe KD, et al. Prognostic relevance of lymph node ratio following pancreaticoduodenectomy for pancreatic cancer. Surgery. 2007;141(5):610–8. 31. Bando E, Yonemura Y, Taniguchi K, Fushida S, Fujimura T, Miwa K. Outcome of ratio of lymph node metastasis in gastric carcinoma. Ann Surg Oncol. 2002;9(8):775–84. 32. Lee HY, Choi HJ, Park KJ, Shin JS, Kwon HC, Roh MS, et al. Prognostic significance of metastatic lymph node ratio in nodepositive colon carcinoma. Ann Surg Oncol. 2007;14(5):1712–7. 33. Nishio H, Ebata T, Yokoyama Y, Igami T, Sugawara G, Nagino M. Gallbladder cancer involving the extrahepatic bile duct is worthy of resection. Ann Surg. 2011;253(5):953–60. 34. Hueman MT, Vollmer CM Jr, Pawlik TM. Evolving treatment strategies for gallbladder cancer. Ann Surg Oncol. 2009;16(8):2101–15.
ORIGINAL ARTICLE – HEPATOBILIARY TUMORS
Lymph Node Metastases in Patients Undergoing Surgery for a Gallbladder Cancer. Extension of the Lymph Node Dissection and Prognostic Value of the Lymph Node Ratio David Je´re´mie Birnbaum, MD1, Luca Vigano`, MD1,2, Nadia Russolillo, MD1, Serena Langella, MD1, Alessandro Ferrero, MD1, and Lorenzo Capussotti, MD1 Department of HPB and Digestive Surgery, Ospedale Mauriziano ‘‘Umberto I’’, Turin, Italy; 2Department of Hepato-biliary Surgery, Humanitas Research Hospital, Milan, Rozzano, Italy 1
ABSTRACT Background. Lymph node (LN) status is one of the strongest prognostic factors after gallbladder cancer (GBC) resection. The adequate extension of LN dissection and the stratification of the prognosis in N? patients have been debated. The present study aims to clarify these issues. Methods. A total of 112 consecutive patients who underwent operations for GBC with LN dissection were analyzed. Twenty-five patients (22.3 %) had D1 dissection (hepatic pedicle), and 87 (77.7 %) had D2 dissection (hepatic pedicle, celiac and retro-pancreatic area). The LN ratio (LNR) was computed as follows: number of metastatic LNs/number of retrieved LNs. Results. The median number of retrieved LNs was 7 (1– 35). Fifty-nine patients (52.7 %) had LN metastases (22 N2). D2 dissection allowed the retrieval of more LNs (8 vs. 3, p = 0.0007), with similar short-term outcomes. Common bile duct (CBD) resection (n = 41) did not increase the number of retrieved LNs. In five patients, D2 dissection identified skip LN metastases that otherwise would have been missed. LN metastases negatively impacted survival (5-years survival 57.2 % if N0 vs. 12.4 % if N?, p \ 0.0001), but N1 and N2 patients had similar survival rates. The number of LN? (1–3 vs. C4) did not impact prognosis. An LNR = 0.15 stratified the prognosis of N?
David Je´re´mie Birnbaum and Luca Vigano` contributed equally. Ó Society of Surgical Oncology 2014 First Received: 20 May 2014; Published Online: 9 September 2014 L. Vigano`, MD e-mail: [email protected]
patients: 5-years survival 32.7 % if LNR B 0.15 vs. 10.3 % if LNR [ 0.15 (multivariate analysis p = 0.007). Conclusions. A D2 LN dissection is recommended in all patients, because it allows for better staging. CBD resection does not improve LN dissection. An LNR = 0.15, not the site of metastatic LNs, stratified the prognoses of N? patients.
Gallbladder cancer (GBC) is the most common biliary tract malignancy worldwide.1 It presents with lymph node (LN) metastases in a high proportion of patients, up to 60– 80 % of T3–4 tumors.2–7 LN status is consistently one of the strongest predictors of a poor outcome after surgery, but the adequate extent of LN dissection is debated.2,5,7–9 Even among high-volume HPB centers, a large discrepancy in the number of retrieved LNs after GBC resection occurs.10–13 This is mainly related to the inclusion or omission of celiac and retropancreatic LNs in the dissection. In addition, the need for a systematic common bile duct (CBD) resection to optimize LN dissection is unclear.10,14 A further debated issue is the stratification of the prognoses of N? patients. The most common scoring system is based on the LN metastatic site.15 However, the prognostic value of the LN metastatic site has not been confirmed by all authors. The LN ratio (the ratio of the number of positive LNs to the number of retrieved LNs) has been recently proposed, as well as the number of harvested LNs and the number of positive LNs.11,16–18 The present study aims to elucidate these aspects. First, the authors compared two different LNs dissections (D1 vs. D2) in terms of the number of retrieved LNs, LN status assessments, and outcomes. The contribution of CBD resection to LN dissection also was evaluated.
812
Second, the different scoring systems for N? patients were compared to identify an optimal manner of stratifying prognoses.
D. J. Birnbaum et al.
Definitions
From January 1990 to December 2011, 126 consecutive patients underwent resections for GBC at the author’s institution. A total of 115 (91.3 %) patients received an LN dissection. Patients with the simultaneous diagnosis of other organ tumors (n = 2), and those undergoing R2 resections (n = 1) were excluded from the present analysis. Finally, 112 patients were included. The study was approved by the local ethical committee.
All of the patients were staged according to the 7th edition of TNM staging system.15 LN metastases confined to hepatic pedicle LNs were staged N1, whereas celiac or retropancreatic LN metastases were staged N2. Major hepatectomy was defined as the resection of C3 Couinaud segments. Operative mortality was defined as death within 90 days after surgery or before discharge from the hospital. Morbidity included all postoperative complications and was assessed according to the Dindo–Clavien classification.20 The LN ratio was computed as the ratio between the number of positive LNs and the number of retrieved LNs.
Preoperative Staging and Surgical Procedures
Statistical Analysis
Preoperative staging systematically included carcinoembryonic antigen (CEA) and Ca 19-9 values and thoracoabdominal computed tomography (CT). Since 2000, positron emission tomography (PET)-CT was performed in selected cases, e.g., patients with aggressive GBC to exclude distant metastases or patients with uncertain extrahepatic lesions or enlarged aortocaval LNs. Preoperative N staging did not include endoscopic ultrasonography. Surgery was indicated only if complete resection was achievable. Distant metastases, even if they were hepatic, contraindicated surgery. Since 1998, a diagnostic laparoscopy with laparoscopic ultrasonography was systematically performed before laparotomy to exclude peritoneal carcinomatosis and liver metastases. After laparotomy, aortocaval LNs were systematically sampled and sent for frozen section; if aortocaval LNs were metastatic the resection was not performed. The standard surgical procedure comprised the resection of segments 4b and 5, en bloc cholecystectomy, and LN dissection.19 LN dissection of the hepatic pedicle (D1) was the standard up to 1998; since 1999 LN dissection was extended to the celiac and retropancreatic area (D2).
All patient data were prospectively collected and retrospectively analyzed. Continuous variables were compared between groups by the unpaired t test or Mann–Whitney U test, as appropriate; categorical variables were compared by the Chi-square test or Fisher exact test, as appropriate. A p value \ 0.05 was considered significant for all tests. Patients with 90-days mortalities were excluded from the survival analysis. The multivariate analysis of prognostic factors of disease-specific overall survival (OS) was performed by including all variables significant or borderline significant (p \ 0.1) at univariate analysis into a stepwise regression model. To determine the most appropriate cutoff values for continuous data variables, such as the LN ratio and the number of positive LNs, the cut-point survival analysis was performed with an outcome-oriented approach.
METHODS
Adjuvant Therapy and Follow-up Since 1998, adjuvant chemotherapy was systematically considered in N? patients and in those with advanced GBC. Adjuvant radiotherapy was not delivered to GBC patients. All patients were reviewed every 3 months, including a physical examination, CEA and CA 19-9 levels, and abdominal ultrasonography or thoracoabdominal CT. No patient was lost during the follow-up period. The follow-up was updated to December 31, 2012.
RESULTS Between January 1990 and December 2011, 112 patients affected by GBC underwent surgical resection and LN dissection with curative intent. Table 1 summarizes the patient characteristics. Thirty-seven (33.0 %) patients were male; the median age was 64 years (range 32–83). The GBC diagnosis was made postoperatively after a simple cholecystectomy in 47 (42.0 %) patients. All of them had a reoperation including liver resection and LN dissection. Twenty-four (21.4 %) patients underwent major hepatectomy, eight after a preoperative portal vein embolization. Fifty-five patients (49.1 %) underwent an associated resection of the contiguous organs, including CBD resection in 41. The tumor staging was T3 in 61 (54.5%) and T4 in 16 (14.3%). Ten (8.9%) patients had an R1 resection (positive biliary margins in five patients with CBD resection).
Lymph Node Metastases in Gallbladder Cancers
813
TABLE 1 Patient characteristics and surgical details Patients (n = 112) n (%)
The median number of retrieved LNs was 7 (1–35), 3 (1–8) after D1 vs. 8 (3–35) after D2 (p = 0.0007). Among the D2 patients, the number of retrieved LNs was similar in patients with and without CBD resection (9 vs. 7, p = 0.095), while it was higher in those with associated pancreatoduodenectomy (14.5, p = 0.001).
Age (years)
64 (32–83)
Male sex
37 (33.0)
CEA (ng/mL)
2 (0.6–657)
CA 19-9 (U/mL)
27.5 (0.8–4778)
Short-term Outcomes
Cholecystectomy
2 (1.8)
Sg4b-5 resection Right hepatectomy [±Sg1]
86 (76.7) 8 (7.1)
Right hepatectomy ? Sg4 [±Sg1]
13 (11.6)
The 90-days mortality rate was 5.4 % (n = 6). The causes of death were as follows: hepatic failure in two patients, hemoperitoneum in one, sepsis in one, acute myocardial infarction in one, and acute respiratory distress syndrome in one. The overall morbidity rate was 47.3 % (n = 53), including 20 severe complications (grade III– IV). In details, severe complications included liver dysfunction in four patients, bile leak requiring interventional procedures in five, bile leak associated with hemoperitoneum requiring reoperation in four, abdominal collection requiring percutaneous drainage in two, duodenal leak associated with hemoperitoneum requiring reoperation in one, renal dysfunction in one, myocardial infarction in one, acute respiratory distress in one, and pleural effusion requiring percutaneous drainage in one. D1 and D2 patients had similar overall morbidity rates (40.0 vs. 49.4 %), severe morbidity rates (grade III–V complications, 16.0 vs. 25.3 %), and hospital stays (12.5 vs. 11 days). Patients undergoing standard surgical procedures (bisegmentectomy Sg4b-5 and LN dissection, n = 53) had no mortality and three grade III/IV complications (5.7 %). Patients requiring resection of contiguous organs (n = 55) had higher mortality (11.0 vs. 0 %, p = 0.027) and grade III/ IV morbidity rates (32.7 vs. 5.7 %, p = 0.0009). Among the patients requiring pancreatoduodenectomy (n = 10), there was one operative mortality (10 %) and no grade III–IV morbidities.
Surgical procedures
Left hepatectomy ? Sg5 [±Sg1]
3 (2.7)
Associated Sg1 resection
9 (8.0)
Associated resections Common bile duct
41 (36.6)
Pancreatoduodenectomy
10 (8.9)a
Right colectomy
4 (3.6)a
Distal gastrectomy
2 (1.8)
T stageb T1b
4 (3.6)
T2
31 (27.7)
T3
61 (54.5)
T4
16 (14.3)
Completeness of surgery R0 R1 Adjuvant chemotherapy
102 (91.1) 10 (8.9) 39 (34.8)
Continuous variables are reported as median value (range) CEA carcinoembryonic antigen, GBC gallbladder cancer, Sg segment a
Two patients had combined pancreatoduodenectomy and right colectomy
b
According to the 7th edition of the AJCC manual15
LN Dissection LN Status Table 2 summarizes the data on the LN dissections. A D2 was performed in 87 (77.7 %) patients. D1 and D2 patients had similar characteristics. To achieve complete LN excision, two patients required a CBD resection and three a pancreatoduodenectomy. The remaining indications for a CBD resection were as follows: suspect macroscopic neoplastic infiltration in 33 patients, noncontiguous neoplastic tissue along the hepatic pedicle in three, and cystic duct stump infiltration in three. Seven additional patients required a pancreatoduodenectomy because of macroscopic neoplastic infiltration of the duodenopancreatic area by GBC. All ten patients undergoing pancreatoduodenectomy had concomitant liver resection.
Table 2 summarizes the data on the LN status. Fiftynine patients (52.7 %) had LN metastases, including 22 (19.6 %) N2. In five patients, a noncontinuous LN metastases distribution was observed: LN metastases in the retropancreatic or celiac area without LN metastases in the hepatic pedicle. Among N? patients, the median number of positive LNs was 2 (1–28), and the median LN ratio was 0.33 (0.08–1). The LN ratio was similar in D1 and D2 patients (0.31 vs. 0.33). LN status correlated with the T stage: N? patients were 0 % in T1b group, 29.0 % in T2, 60.6 % in T3, and 81.3 % in T4.
814
D. J. Birnbaum et al.
TABLE 2 Details about LN dissection and N status Patients (n = 112) n (%)
Prognostic Impact of the LN Status and Stratification of Prognosis in N? Patients
Lymph node dissection D1
25 (22.3)
D2
87 (77.7)
Number of retrieved lymph nodes After D1* Lymph node staging N0
7 (1–35) 3 (1–8)*
After D2*
8 (3–35)* a
N?
53 (47.4) 59 (52.7)
N1
37 (33.0)
N2
22 (19.6)
Number of metastatic lymph nodes
2 (1–28)
Lymph node ratio Overall
0.08 (0–1)
In N? patients
0.33 (0.08–1)
Number of metastatic LNs 1
32 (28.6)
2–3
12 (10.7)
[3
15 (13.4)
Continuous variables are reported as median (range) * Number of retrieved lymph nodes after D1 vs. D2, p = 0.0007 a
Overall, the ten patients who required pancreatoduodenectomy had 0 % OS at 2 years after surgery.
According to the 7th edition of the AJCC manual15
Long-term Outcomes and Prognostic Impact of the LN Dissection After a median follow-up of 71.7 months, 3- and 5-years OS rates were 40.6 and 33.2 %, respectively (median OS 23.8 months). Twenty-three (20.5 %) patients survived more than 5 years, including three N1 and one N2 patients. Seventy-one patients (63.4 %) developed recurrences: hepatic only in 32, extrahepatic only in 34, and both hepatic and extrahepatic in 5. The recurrence site was unrelated to the LN status, LN metastatic site, and LN ratio value. D1 and D2 patients had similar prognoses (5-years OS, 26.2 vs. 38.2 %). The number of harvested LNs (1-6 vs. [6) did not impact OS (28.0 vs. 31.7 %). Two patients required CBD resections because of hepatic pedicle LN metastases. One is still alive and disease-free 81 months after resection, while the other died of disease 14 months after surgery. The CBD resection was associated with worse OS, both in the whole series (5-years OS 9.6 vs. 42.7 % in patients without CBD resection, p \ 0.0001) and in N0 patients (17.4 vs. 59.0 %, p = 0.017). The three patients who had pancreatoduodenectomies due to retropancreatic LN metastases died of disease at 8, 15, and 15 months after surgery, respectively.
LN metastases negatively impacted survival: 5-years OS 57.2 % in N0 patients vs. 12.4 % in N? ones (p \ 0.0001; Fig. 1a). Among N? patients, N1 and N2 patients had similar OS (15.3 % in N1 patients vs. 6.6 % in N2; Fig. 1a). The LN ratio stratified the prognoses in N? patients. The cut-point survival analysis identified 0.15 as the best cutoff value. The 5-years OS in patients with LN ratio 0.01–0.15 was 32.7 vs. 10.3 % in those with LN ratio [0.15 (p = 0.0001; Fig. 1b). The prognostic value of the LN ratio was independent from the extent of LN dissection (significant predictor of OS after both D1 and D2 dissection). An LN ratio 0.15 stratified survival outcomes, even in T3-4 tumors. The number of metastatic LNs (1 vs. 2–3 vs. [3) was not associated with the prognosis (5-years OS 13.7 % if one metastatic LN vs. 13.9 % if 2–3 vs. 16.7 % if [3, p = 0.641; Fig. 2). Prognostic Factors of OS: Multivariate Analysis Table 3 provides the results of univariate and multivariate analysis of the prognostic factors of OS after surgery. LN ratio, T stage, and R1 resection were independent prognostic factors of OS (p = 0.007, 0.008, and 0.016, respectively). The study period (1990–1998 vs. 1999–2011) did not impact OS. DISCUSSION LN metastases represent a crucial issue in the management of patients affected by GBC. First, LN metastases are common, and their rate increases along with the T stage, up to 60–80 % in T3–4.2–7 In the present series, 53 % of patients were N?, 65 % in T3–4 tumors. Second, LN metastases are one of the strongest prognostic determinants after surgery.2,5–9 In the present series, survival expectancy was drastically reduced in N? patients (5-years OS 12 vs. 57 % in N0 patients). Many issues about LN metastases in GBC still await clarification. The first of these is the extent of LN dissection. Two recent population-based U.S. studies demonstrated that, in the majority of surgical centers, GBC patients receive very limited or even no LN dissection.21,22 The AJCC recommends the analysis of at least three LNs to adequately assess the N stage.15 A higher cutoff also has been recently
Lymph Node Metastases in Gallbladder Cancers
a
1.0 0.9 0.8 0.7
% Overall Survival
FIG. 1 a Overall survival curves of patients undergoing operations for gallbladder cancer according to the N status (N0 vs. N?, p \ 0.0001; N1 vs. N2, p = 0.612). b Overall survival curves of patients undergoing operations for gallbladder cancer according to the lymph node (LN) ratio (p = 0.007)
815
0.6 0.5 0.4 0.3 0.2 0.1 0.0
0
12
24
36
48
60
48
60
Months N0
b
N2
N1
1.0 0.9 0.8
% Overall Survival
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0
0
12
24
36
Months LN Ratio = 0 (N0 patients)
proposed: in N0 patients, survival was improved if at least six LNs were retrieved.12,16 Despite this theoretical agreement, the adequate extent of LN dissection remains controversial (D1 vs. D2). Even among high-volume HPB centers, a large discrepancy in the number of retrieved LNs can be observed. Pawlik et al. and Ito et al. reported a median of three LNs retrieved per patient, whereas Shirai et al. achieved 14 LNs and Sasaki et al. achieved 22.10–13 In the present series, both D1 and D2 dissections were performed, and their results were compared. Several data suggest that D2 should be the standard. First, the median number of LNs retrieved after D1 dissection was three, the
LN Ratio 0.01-0.15
LN Ratio > 0.15
minimum recommended by AJCC, but far from the number of six recommended by recent papers.12,16 Second, in five patients, skip metastases were detected, e.g., LN metastases in the retropancreatic or celiac area without LN metastases in the hepatic pedicle. These patients would have been classified as N0 after a D1 dissection. Finally, the D2 dissection did not increase operative risk. D2 LN dissection is part of an adequate radical oncological treatment of GBC. Some Japanese authors have recommended a systematic CBD resection in patients operated on for a GBC to improve LN dissection.14 The Makuuchi group even reported a survival benefit after CBD resection in patients
816
D. J. Birnbaum et al.
FIG. 2 Overall survival curves of patients undergoing operations for gallbladder cancer according to the number of metastatic lymph nodes (p = 0.641)
1.0 0.9 0.8
% Overall Survival
0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0
12
24
36
48
60
Months 1 metastatic LN
with perineural infiltration.14 These data have not been confirmed by the most recent papers.10,16 In the present series, the CBD resection did not modify the quality of LN dissection; exactly the same number of LNs was retrieved in patients with and without biliary resection. Patients with CBD resection had worse survival compared with those without it. However, in the present series, the CBD was resected if necessary to achieve a complete surgery. This population cannot be compared with the one reported by Makuuchi et al. (de principe resection). A randomized trial is advocated to clarify the prognostic value of the systematic CBD resection. In the present series, three patients required pancreatoduodenectomy to remove metastatic retropancreatic LNs. Seven additional patients had concomitant liver resection and pancreatoduodenectomy because of neoplastic infiltration of the duodenopancreatic area. The short-term outcomes were good, but survival results were extremely poor (0 % OS at 2 years). In the literature, a debate is ongoing about the benefit of pancreatoduodenectomy for GBC.23–26 According to present data, in the author center patients requiring a pancreatoduodenectomy for GBC are now excluded from surgery. A second major issue is the stratification of prognosis in N? patients. Different proposals have been put forth, as follows: the LN metastases site, the number of positive LNs, the LN ratio.11,15–18,27 The anatomical location of positive LNs is the most common solution and has been adopted by the AJCC (TNM 7th edition) and the Japanese Society of Biliary Surgery.15,27 The stratification of patient prognoses on the basis of LN? site may allow pre- and intraoperative predictions of outcomes but
2-3 metastatic LNs
>3 metastatic LNs
has several limitations: it depends on the extent of LN dissection; the farthest metastatic LN is assessed, but the burden of the lymphatic dissemination is not considered; the LN site can be difficult to determine on the specimen. Furthermore, recent studies did not confirm the prognostic value of LN metastatic sites, as the present series did (N1 and N2 patients had exactly the same OS).11,16–18 As an alternative, some authors have proposed the number of positive LNs.11,17,18 However, in the literature, a number of long-term survivors after resection for GBC with multiple positive LNs have been reported.23,26,28,29 In the present series, the number of positive LNs failed to stratify prognoses, and four 5-years survivors with multiple positive LNs were observed. Finally, the LN ratio can be considered, as it has been recently adopted in other gastrointestinal tumors.30–32 Negi et al. first reported its prognostic role in GBC patients over a small series of 57 patients.16 In a larger setting (112 patients), the present series confirmed the LN ratio as the best way to stratify prognosis of N? patients. The best cutoff value was 0.15. The LN ratio can be assessed only at the pathological examination but is independent from the extent of the LN dissection, considers the burden of the lymphatic dissemination, and does not require the identification of the LN site. Shirai et al. suggested that the LN ratio value decreases together with the extension of LN dissection.11 This was not confirmed in the present series: the LN ratio had the same prognostic value after both D1 and D2 dissections. Some limitations of the present study could be argued. First, it is a retrospective series, collecting 112 patients
Lymph Node Metastases in Gallbladder Cancers TABLE 3 Uni- and multivariate analysis of prognostic factors of overall survival Parameter
Overall survival Univ.
Multivariate analysis
p
p
RR (CI 95 %)
Age (years) B70
0.624
–
0.864
–
0.021
n.s.
[70 Sex Male Female Major hepatectomy N 0.002
n.s.
0.0001
0.008
1 2.561 (1.275–5.143)
\0.0001
0.007
1
N T stage
Lymph-node ratio 0 0.01–0.15
1.313 (0.551–3.127)
[0.15
2.627 (1.426–4.842)
Complete resection R0
0.001
0.016
R1
1 2.383 (1.175–4.834)
Grading 1–2
0.011
n.s.
0.450
–
0.098
n.s.
3–4 Adjuvant CTx Y N Study period 1990–1998
None to declare.
REFERENCES
Associated resection
1–2 3–4
benefit of surgery, especially in N2 patients. However, N2 patients had survival outcomes similar to N1 ones and superior to survival rates reported in the literature for unresected patients (median survival 4–7 months).1,33,34 In conclusion, N status is one of the strongest prognostic determinants in patients undergoing operations for GBC. A D2 LN dissection is recommended in all patients because it allows for better patient staging without worsening shortterm outcomes. A systematic CBD resection is not required to improve LN dissection. An LN ratio of 0.15, not the location or the number of positive LNs, is the best method of stratifying prognoses in N? patients. CONFLICTS OF INTEREST
Y
Y
817
1999–2011 Bold values are statistically significant Univ. univariate, RR relative risk, CTx chemotherapy, n.s. not significant
during a 20-years period. This limitation is common to all published series on GBC due to the rarity of the disease and its low resectability rate. Second, the extent of the LN dissection was modified during the study period. However, no additional major modifications in GBC management occurred in the two study periods. The two groups (D1 and D2) had similar characteristics, and the study period had no impact on OS. Finally, the outcomes of resected patients were not compared to those of unresected patients. This comparison is mandatory to definitively ascertain the
1. Misra S, Chaturvedi A, Misra NC, et al. Carcinoma of the gallbladder. Lancet Oncol. 2003;4:167–76. 2. Miyakawa S, Ishihara S, Horiguchi A, Takada T, Miyazaki M, Nagakawa T. Biliary tract cancer treatment: 5,584 results from the Biliary Tract Cancer Statistics Registry from 1998 to 2004 in Japan. J Hepatobiliary Pancreat Surg. 2009;16(1):1–7. 3. Tsukada K, Kurosaki I, Uchida K, Shirai Y, Oohashi Y, Yokoyama N, et al. Lymph node spread from carcinoma of the gallbladder. Cancer. 1997;80(4):661–7. 4. Shimada H, Endo I, Togo S, Nakano A, Izumi T, Nakagawara G. The role of lymph node dissection in the treatment of gallbladder carcinoma. Cancer. 1997;79(5):892–9. 5. Fong Y, Wagman L, Gonen M, Crawford J, Reed W, Swanson R, et al. Evidence-based gallbladder cancer staging: changing cancer staging by analysis of data from the National Cancer Database. Ann Surg. 2006;243(6):767–71. 6. Shirai Y, Wakai T, Hatakeyama K. Radical lymph node dissection for gallbladder cancer: indications and limitations. Surg Oncol Clin N Am. 2007;16(1):221–32. 7. Kondo S, Takada T, Miyazaki M, Miyakawa S, Tsukada K, Nagino M, et al. Guidelines for the management of biliary tract and ampullary carcinomas: surgical treatment. J Hepatobiliary Pancreat Surg. 2008;15(1):41–54. 8. Benoist S, Panis Y, Fagniez PL. Long-term results after curative resection for carcinoma of the gallbladder. French University Association for Surgical Research. Am J Surg. 1998;175(2):118–22. 9. Jensen EH, Abraham A, Jarosek S, Habermann EB, Al-Refaie WB, Vickers SA, et al. Lymph node evaluation is associated with improved survival after surgery for early stage gallbladder cancer. Surgery. 2009;146(4):706–11. 10. 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 re-resection. J Gastrointest Surg. 2007;11(11):1478–86. 11. Shirai Y, Sakata J, Wakai T, Ohashi T, Ajioka Y, Hatakeyama K. Assessment of lymph node status in gallbladder cancer: location, number, or ratio of positive nodes. World J Surg Oncol. 2012;10:87. 12. Ito H, Ito K, D’Angelica M, Gonen M, Klimstra D, Allen P, et al. Accurate staging for gallbladder cancer: implications for surgical therapy and pathological assessment. Ann Surg. 2011;254(2):320–5. 13. Sasaki E, Nagino M, Ebata T, Oda K, Arai T, Nishio H, et al. Immunohistochemically demonstrated lymph node micrometastasis and prognosis in patients with gallbladder carcinoma. Ann Surg. 2006;244(1):99–105.
818 14. Sakamoto Y, Kosuge T, Shimada K, Sano T, Hibi T, Yamamoto J, et al. Clinical significance of extrahepatic bile duct resection for advanced gallbladder cancer. J Surg Oncol. 2006;94(4):298– 306. 15. Edge SBB, Compton CC, Fritz AG, Greene FL, Trotti A. AJCC cancer staging manual. 7th ed. New York, NY: Springer; 2010. p. 211–7. 16. Negi SS, Singh A, Chaudhary A. Lymph nodal involvement as prognostic factor in gallbladder cancer: location, count or ratio? J Gastrointest Surg. 2011;15(6):1017–25. 17. Endo I, Shimada H, Tanabe M, Fujii Y, Takeda K, Morioka D, et al. Prognostic significance of the number of positive lymph nodes in gallbladder cancer. J Gastrointest Surg. 2006;10(7):999– 1007. 18. Sakata J, Shirai Y, Wakai T, Ajioka Y, Hatakeyama K. Number of positive lymph nodes independently determines the prognosis after resection in patients with gallbladder carcinoma. Ann Surg Oncol. 2010;17(7):1831–40. 19. Muratore A, Amisano M, Vigano` L, et al. Gallbladder cancer invading the perimuscular connective tissue: results of reresection after prior non-curative operation. J Surg Oncol. 2003;83(4):212–5. 20. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240(2):205–13. 21. Coburn NG, Cleary SP, Tan JC, Law CH. Surgery for gallbladder cancer: a population-based analysis. J Am Coll Surg. 2008;207(3): 371–82. 22. Mayo SC, Shore AD, Nathan H, Edil B, Wolfgang CL, Hirose K, et al. National trends in the management and survival of surgically managed gallbladder adenocarcinoma over 15 years: a population-based analysis. J Gastrointest Surg. 2010;14(10): 1578–91. 23. Shirai Y, Ohtani T, Tsukada K, Hatakeyama K. Combined pancreaticoduodenectomy and hepatectomy for patients with locally advanced gallbladder carcinoma: long term results. Cancer. 1997;80(10):1904–9.
D. J. Birnbaum et al. 24. Kondo S, Nimura Y, Hayakawa N, Kamiya J, Nagino M, Uesaka K. Extensive surgery for carcinoma of the gallbladder. Br J Surg. 2002;89(2):179–84. 25. Araida T, Yoshikawa T, Azuma T, Ota T, Takasaki K, Hanyu F. Indications for pancreatoduodenectomy in patients undergoing lymphadenectomy for advanced gallbladder carcinoma. J Hepatobiliary Pancreat Surg. 2004;11(1):45–9. 26. Sasaki R, Itabashi H, Fujita T, Takeda Y, Hoshikawa K, Takahashi M, et al. Significance of extensive surgery including resection of the pancreas head for the treatment of gallbladder cancer -from the perspective of mode of lymph node involvement and surgical outcome. World J Surg. 2006;30(1):36–42. 27. Japanese Society of Biliary Surgery (JSBS). Classification of Biliary Tract Carcinoma. English ed. Tokyo: Kanehara & Co., Ltd; 2004. p. 33–50. 28. Shirai Y, Yoshida K, Tsukada K, Muto T, Watanabe H. Radical surgery for gallbladder carcinoma. Long-term results. Ann Surg. 1992;216(5):565–8. 29. Kondo S, Nimura Y, Kamiya J, Nagino M, Kanai M, Uesaka K, et al. Five-year survivors after aggressive surgery for stage IV gallbladder cancer. J Hepatobiliary Pancreat Surg. 2001;8(6):511–7. 30. Pawlik TM, Gleisner AL, Cameron JL, Winter JM, Assumpcao L, Lillemoe KD, et al. Prognostic relevance of lymph node ratio following pancreaticoduodenectomy for pancreatic cancer. Surgery. 2007;141(5):610–8. 31. Bando E, Yonemura Y, Taniguchi K, Fushida S, Fujimura T, Miwa K. Outcome of ratio of lymph node metastasis in gastric carcinoma. Ann Surg Oncol. 2002;9(8):775–84. 32. Lee HY, Choi HJ, Park KJ, Shin JS, Kwon HC, Roh MS, et al. Prognostic significance of metastatic lymph node ratio in nodepositive colon carcinoma. Ann Surg Oncol. 2007;14(5):1712–7. 33. Nishio H, Ebata T, Yokoyama Y, Igami T, Sugawara G, Nagino M. Gallbladder cancer involving the extrahepatic bile duct is worthy of resection. Ann Surg. 2011;253(5):953–60. 34. Hueman MT, Vollmer CM Jr, Pawlik TM. Evolving treatment strategies for gallbladder cancer. Ann Surg Oncol. 2009;16(8):2101–15.
