Abstract
Background/Aim: Hepatectomy with extrahepatic bile duct resection (Hx+EBDR), pancreaticoduodenectomy (PD), and occasionally hepatopancreaticoduodenectomy (HPD) are required for the treatment of advanced gallbladder cancer (GBC). This study aimed to clarify the clinical value of these extended surgeries for GBC.
Patients and Methods: We retrospectively reviewed the medical records of patients who underwent curative resection (Surg-G, n=59), and their survival rates were compared with those of patients with unresectable GBC who underwent chemotherapy (CTx-G, n=63).
Results: We performed PD (n=15), Hx+EBDR (n=37), and HPD (n=7). The postoperative complication and death rates were as follows: PD, 40% and 7%, respectively; Hx+EBDR, 89% and 14%, respectively; and HPD, 57% and 0%, respectively. Concomitant vascular resection (VR) was required in 61% of the patients. The 5-year overall survival rate and median survival time (MST) for Surg-G were 25.1% and 26 months, respectively, whereas those for CTx-G were 4.6% and 14.4 months, respectively. There were no significant differences between the surgical procedures. Patients who underwent VR had similar prognoses (5-year overall survival rate and MST: 14.5% and 22.3 months, respectively) as the patients in CTx-G.
Conclusion: Although extended surgery may be considered for patients with GBC, careful patient selection and new therapeutic strategies are required, especially for those requiring VR.
- Hepatopancreaticoduodenectomy
- hepatectomy with extrahepatic bile duct resection
- pancreaticoduodenectomy
- gallbladder carcinoma
- prognosis
- surgical indication
Introduction
Kondo et al. (1) previously identified six types of gallbladder cancer (GBC), and various surgical techniques have been indicated for GBC depending on the site of origin. In particular, when a tumor extends from the neck of the gallbladder to the extrahepatic bile duct, the anatomical features of the gallbladder allow it to easily extend into the right hepatic artery and/or portal vein, requiring major hepatectomy with bile duct resection (1-7). Procedures for these types of GBC include major hepatectomy with extrahepatic bile duct resection (Hx+EBDR), pancreaticoduodenectomy (PD), and hepatopancreaticoduodenectomy (HPD). Although these techniques are widely used for perihilar or distal cholangiocarcinoma, a high morbidity rate but with an acceptable mortality rate has been reported by Japanese institutes (8-11). Previous studies on advanced GBC have reported debatable indications for surgery (8, 12) because these procedures not only have high morbidity and mortality but also provide poor long-term prognosis. A study showed that patients with positive peripancreatic lymph node had a similar prognosis to that of a cohort with unresectable GBC (13). However, another study showed that patients with peripancreatic lymph node metastasis had similar prognosis to those who had regional lymph node metastasis. These patients had obviously better survival than unresectable patients (median survival: 28 months and 5 year survival rate 36%, respectively) (6). Furthermore, recent studies showed alternative treatment strategies (such as chemotherapy including molecular targeting therapy or radiation therapy) do not guarantee long-term survival (14). To solve this dilemma, it is important to identify the group with poor prognosis in upfront surgery for GBC and establish a borderline resection group, as has been done for pancreatic cancer (15).
This single-institution retrospective study aimed to clarify the clinical value and patient selection for these extended surgeries for GBC.
Patients and Methods
Institutional Review Board Statement. This retrospective study was approved by the Institutional Review Board of Hokkaido University (021-0161) and was conducted in accordance with the Declaration of Helsinki. The study was registered at UMIN-CTR (UMIN000046548).
Patients. We retrospectively reviewed the medical records of patients who underwent surgery with curative intent for GBC between January 2001 and December 2022. In the surgical group (Surg-G), we evaluated consecutive patients who underwent major hepatectomy with or without EBDR, HPD, and PD procedures. We also retrospectively reviewed the data of patients with unresectable GBC who underwent chemotherapy (CTX-G) between January 2010 and December 2022.
Preoperative management. Cholangiography, multidetector row computed tomography, ultrasonography, and magnetic resonance cholangiopancreatography were performed to evaluate clinical stage and precise longitudinal cancer spread. At Hokkaido University Hospital, a weekly multidisciplinary hepatobiliary disease management conference is held, during which all biliary malignancies are discussed. Preoperative management for hepatectomy with or without EBDR or HPD was similar to that of perihilar cholangiocarcinoma, which has been described in our previous report (12). Briefly, preoperative biliary decompression was carried out in patients with obstructive jaundice with the aim of reducing the serum bilirubin level to <2 mg/dl and controlling segmental cholangitis using endoscopic nasobiliary drainage (ENBD) as the standard procedure. Since 2004, percutaneous transhepatic biliary drainage (PTBD) has only been performed when sufficient decompression could not be achieved with ENBD. The standard practice is to replace the bile drained using ENBD or PTBD orally until the day of surgery (8). Portal vein embolization was performed in patients who underwent right hepatectomy or trisectionectomy (8).
Surgery. Patients without obvious periaortic lymph nodes, peritoneal dissemination, or liver metastases were considered surgical candidates. Figure 1 shows the algorithm used to select surgical procedures. Prior to radical resection, we performed para-aortic lymph node excisional biopsy and intraoperative pathological diagnosis of these nodes. We performed cholecystectomy with or without EBDR for patients without perihilar invasion, intrapancreatic invasion, or bulky peripancreatic lymph node metastasis. Cholecystectomy with EBDR was recommended for a limited number of patients with cystic ductal lesions or a limited lesion in the extrahepatic bile duct (Figure 1). Before curative resection, we performed excisional biopsy of the para-aortic lymph node. We selected patients who underwent Hx+EBDR for GBC with perihilar invasion (hepatic, bed, and hilar types) (1). We selected patients undergoing PD with intrapancreatic bile duct invasion and/or bulky peripancreatic lymph node metastasis. For the cystic duct type, various procedures were selected based on tumor spread (Hx+EBDR, PD, HPD, and cholecystectomy with EBDR) (1).
Surgical procedure algorithm. GBC: Gallbladder carcinoma; LNM: lymph node metastasis; HPD: hepatopancreaticoduodenectomy; Hx: major hepatectomy; EBDR: extrahepatic bile duct resection; PD: pancreaticoduodenectomy; Chole: cholecystectomy.
We performed surgery for patients with GBC based on the preoperative diagnosis. Surgery was contraindicated in patients with metastases to other organs or peritoneal metastases. Patients who had locally advanced disease were excluded from surgical indication. We planned curative resection for patients with peripancreatic lymph node metastasis outside the hepatoduodenal ligament metastasis (no. 8 or 13 in the Japanese Biliary Classification) (16). All types of procedures, and the most minimally invasive procedures to obtain an R0, were considered. We did not perform prophylactic extended resection such as minor liver resections (S4/5) or PD for prophylactic dissection of peripancreatic lymph nodes.
Surgical procedures and perioperative management of Hx+EBDR and HPD were reported in our previous study on cholangiocarcinoma (8). Operative procedures for PD were initiated with subtotal stomach-preserving PD with lymph node dissection around the superior mesenteric artery, followed by lymph node dissection in the hepatoduodenal ligament. All patients underwent resection of the confluence of the left and right hepatic ducts, but some patients underwent hilar plate resection or high-cut PD (17, 18). Some patients with PD who underwent minor hepatectomy (such as S4 segmentectomy for tumor involvement) were included in the PD group. For digestive tract reconstruction, pancreatojejunostomy, hepaticojejunostomy, and gastrojejunostomy were performed using the jejunal limb, according to the modified Child method. Pancreatic reconstructions were based on surgeons’ preferences.
Follow-up. Patients underwent regular follow-up examinations, including physical examination, laboratory tests, tumor marker evaluation, and computed tomography, at 3- to 6-month intervals.
Statistical analyses. All statistical analyses were performed using JMP version 15.2 (SAS Institute, Cary, NC, USA). Chi-squared, Fisher’s exact, and Mann-Whitney U-tests were performed as appropriate. Statistical significance was set at p<0.05. The clinical characteristics used to determine morbidity or mortality were assessed using receiver operating characteristic curve analysis, and the areas under the curve were compared. Overall survival curves were generated using the Kaplan-Meier method, and differences were compared using the log-rank test.
Results
We performed curative intent surgery on 189 patients with GBC (Figure 1 and Figure 2) based on our institutional surgical indications. Distant metastasis was observed in 39 patients. Cholecystectomy and lymphadenectomy, with or without EBDR, were performed in 150 patients. The remaining 59 patients were included in the study (Surg-G). The characteristics of the patients in Surg-G are presented in Table I. Two patients underwent neoadjuvant chemotherapy before surgery. We performed PD (n=15), Hx+EBDR (n=37), and HPD (n=7). Surgical procedures in patients who underwent minor hepatectomies were as follows: (H4′5′6′ resection: three cases, H5′6′ resection: one case).
Operative procedures in our department from January 2001 to December 2022. Of 189 patients, 39 underwent exploratory laparotomy and 59 underwent Hx+EBDR, PD or HPD. GBC: Gallbladder carcinoma; Ex-lap: exploratory laparotomy; Hx+EBDR: hepatectomy with extrahepatic bile duct resection; PD: pancreaticoduodenectomy; HPD: hepatopancreaticoduodenectomy; Chole: cholecystectomy; LND: lymph nodal resection; EBDR: extrahepatic bile duct resection.
Patient characteristics.
In these patients, we did not perform definitive adjuvant therapy after surgery. Three patients underwent adjuvant therapy according to their preferences. Six patients died after surgery because of postoperative liver failure (Hx+EBDR group, n=5) and postoperative hemorrhage (PD group, n=1). More than half of the patients in Surg-G had cystic duct carcinoma based on postoperative pathology. One patient presented with a benign lesion. Five patients were diagnosed with a distal or perihilar cholangiocarcinoma. Three of the seven patients with HPD (with preoperative diagnosis of cystic carcinoma) had perihilar cholangiocarcinoma. Distant metastases (liver or peritoneal) were found in 6 of 59 patients. The R0 resection rate was 83%. Table II shows the patient characteristics in CTx-G. Locally advanced disease (29%) and distant metastases (71%) were the reasons for unresectability in CTx-G (n=63). None of the patients received molecular targeted therapy.
Patient characteristics in the non-surgical treatment group.
Median survival time (MST) and 5-year overall survival rate (5Y OS) were 24% and 26 months for Surg-G, respectively, and 4.6% and 14.4 months for CTx-G, respectively (Figure 3A). Disease-specific survival in this study is shown in Figure 3B. The survival rate in Surg-G was significantly higher than that in CTx-G. Survival was not significantly different between patients with cystic duct carcinoma (CDC) (n=28) and those with non-CDC (n=25) (Figure 4). There were no significant differences in OS between the surgical procedures (Figure 5). Patients who underwent vascular resection (VR) had a poor prognosis, but they had a better prognosis than those who underwent CTx-G (Figure 6). Patients with or without pathological lymph node metastasis had no significant difference in survival but had better prognosis than CTx-G (Figure 7). Patients with incidental distant metastases [M(+)] were not significantly different from patients without incidental distant metastases [M(−)], but they had a similar prognosis as that of CTx-G (Figure 8). All these distant lesions were found by pathological examination after surgery.
Five-year overall survival rates of the surgery and chemotherapy groups. 5Y OS: 5-year overall survival rate; MST: median survival time; CTx: chemotherapy.
Overall survival after surgery based on tumor location. There were no differences between patients with cystic duct carcinoma (CDC) and those with non-CDC tumor. 5Y OS: 5-Year overall survival rate; MST: median survival time.
Overall survival based on surgical procedures and chemotherapy. Hx+EBDR: Hepatectomy with extrahepatic bile duct resection; PD: pancreaticoduodenectomy; HPD: hepatopancreaticoduodenectomy; CTx: chemotherapy; 5Y OS: 5-year overall survival rate; MST: median survival time.
Overall survival with and without vascular resection (VR). There were significant differences between the three groups. 5Y OS: 5-Year overall survival rate; MST: median survival time; CTx: chemotherapy.
Overall survival with and without pathological lymph node metastasis. There were no significant differences between patients with and without pathological metastasis, but there were significant differences between pN(+) and CTxG. pN: Pathological lymph node status; 5Y OS: 5-year overall survival rate; MST: median survival time; CTx: chemotherapy.
Overall survival with and without distant metastasis. There were no significant differences among the three groups, but no patients with distant metastasis survived over 48 months. M: Pathological distant metastasis; 5Y OS: 5-year overall survival rate; MST: median survival time; CTx: chemotherapy.
Discussion
There are several problems with the indication for surgery for advanced GBC. The first is whether surgery is indicated at all because of the high complication rate. The second is whether surgical resection is indicated for HPD with maximal invasion as a benefit. If surgical resection is beneficial, is there an indication for HPD, which is the most invasive procedure? Third, is cholangiocarcinoma the only specific type of GBC? Fourth, is multidisciplinary treatment in combination with chemotherapy and other therapies useful? Can we define a BR for aggressive combination with chemotherapy in the future?
The present study partly answered these questions. Our data showed the surgical results of GBC treated with extended resection. These procedures provide high R0 resection rates and better survival than chemotherapy alone. Patients who required VR had poor prognosis, but they had better prognosis than the patients in CTx-G. These findings suggest that patients requiring concomitant VR would be thought of as borderline resectable for the GBC.
Recently, Creasy et al. (19) reported a change in the trend of avoiding extended resection for GBC at their institution. Advances in chemotherapy for biliary tract malignancies may allow” downsizing” of the tumor extension. There may have been a shift from extended to limited resections in some patients. Another possible reason may be that surgeons avoid performing “demanding” procedures, as these procedures are difficult to perform and are associated with high postoperative complications and low survival rates. In other words, they may avoid surgery because they do not perceive the benefits of highly invasive procedures. In the viewpoint of morbidity and mortality, our series indicated high morbidity and mortality rates, similar to those of a previous report (5). In particular, the mortality rates after Hx+EBDR in our series and in a previous report are very high (5). A previous report showed significant differences in morbidity and mortality between simple hepatectomy and Hx+EBDR; the morbidity and mortality rates for biliary malignancies were high (20). Our previous report showed that the morbidity and mortality rates after Hx+EBDR for perihilar cholangiocarcinoma were 51.9% (109 of 210 patients) and 4.8% (10 of 210 patients), respectively (8). Although a recent international collaborative study showed that the benchmark ranges of morbidity and mortality after Hx+EBDR for perihilar cholangiocarcinoma were less than 70% and 13%, respectively, the morbidity and mortality rates in this series were significantly high (21). The reasons for the high rate of postoperative complications in this study are unknown. A previous report suggested simultaneous resection of multiple organs (e.g., the colon or duodenum) as a possible reason (5). A similar trend was observed in the results from the University of Nagoya. Their surgical outcomes for perihilar cholangiocarcinoma are probably the best in the world, with a low operative mortality rate (1.9% in a recent report) (22). However, a recent report on major hepatectomy for advanced GBC reported a relatively high mortality rate (5%) (5). The reasons for these high morbidity and mortality rates in GBC remain unknown.
Obviously, the disadvantages of surgical treatment are high complication and mortality rates. Assuming that high complication (mortality) rates are an impediment to choosing surgical treatment, are the outcomes of other treatment modalities for GBC superior to those of surgical treatment? We will discuss the superiority of chemotherapy, assuming that it has a small number of side effects, but that these are not as serious as those of surgery. The problem with chemotherapy is its poor survival rate. No large trials of chemotherapy focused exclusively on GBC have been reported, but a recent systematic review showed that chemotherapy would not lead to long-term survival (14). The efficacy of molecular-targeted agents in HER2-positive gallbladder cancer has also been suggested, but in any case, they do not significantly outperform surgery (23). There was no evidence of radiation therapy for advanced GBC.
Therefore, regardless of the demanding of the procedure, and high morbidity and mortality rates, overall survival rates after surgery were significantly better than those of CTx patients. Surgeries for selected patients could provide advanced GBC patients chance to achieve long-term survival. The next question is whether HPD with maximum invasiveness would be indicated for advanced GBC? A previous report showed that HPD for GBC had higher morbidity and mortality rates than Hx+EBDR (5). However, these results were obtained from high-volume centers for biliary surgeries. Therefore, some authors have suggested that surgery is contraindicated in patients with advanced GBC who require HPD (5, 24). Because there are few reports, it is unclear whether HPD for GBC has much higher morbidity and mortality rates. Sakamoto et al. (24) suggested that GBC requiring HPD was a contraindication to surgery. All patients died within two years. Mizuno et al. (5) emphasized that the surgical indication of HPD for patients with GBC should be strictly limited because patients who require HPD have unfavorable tumor biology, such as widespread lymph node metastasis and tumor involvement of the liver, hepatoduodenal ligament, and pancreas. They also suggested that only patients with cystic ductal carcinoma, with morphological features similar to those of perihilar cholangiocarcinoma, had long-term survival after HPD. However, Yamamoto et al. (3) suggested that HPD itself did not influence long-term survival. They reported that patients with GBC who underwent major hepatectomy or HPD might have a better prognosis than those who underwent nonsurgical treatments. However, our data suggested patients who required PD had high R1 resection rate or poor long-term survival. These data suggested that patients who required PD will be borderline resectable. Our results and a previous report for VR suggested that patients with arterial involvement were contraindicated for upfront surgery (5).
Compared with previous reports, our data showed several differences. Our data showed no significant differences in the tumor location or surgical procedures (5). Although the reasons for the differences between previous reports and ours are unclear (as the authors of the previous reports did not present their strategies), one possible reason could be the performance of para-aortic lymph node excisional biopsy before radical resection. We excluded patients with distant metastasis (para-aortic lymph node metastasis), correctly. Another reason could be our exclusion of patients diagnosed with perihilar or distal cholangiocarcinoma based on postoperative pathological examination. Therefore, whether CDC is unique or not cannot be determined from a single-center study alone; a multicenter analysis is needed. However, the question remains: What exactly is CDC? It is unclear which diagnostic criteria are used at other institutions; thus, it will be necessary to unify the pathological diagnostic criteria to advance this discussion.
Our results showed that patients who underwent VR had a very poor prognosis, but their outcomes were better than those in the CTx-G group. This finding supports that of a previous study (12, 25). This suggests that patients who require concomitant VR, almost all T4 cases, would be considered borderline resectable. The surgical results of concomitant hepatic artery resection for perihilar cholangiocarcinoma, as shown by us and the other authors, were better than those for GBC (10, 25). The results of concomitant portal vein resection were acceptable (5), and the results of other authors differ from ours in this regard. We have previously reported that invasion of the main portal vein or contralateral portal vein involving the tunica intima is associated with a very poor prognosis (26). The prognosis of our patients who required HPD with concomitant portal vein involvement was also very poor (12). T4 GBC has a poor prognosis, but to a different extent, and is considered borderline resectable.
Multidisciplinary therapy may be useful; however, evidence is lacking. The question is to what extent can downsizing be achieved after chemotherapy and can surgery be performed safely? There are no reports of conversion rate in unresectable GBC, but there are reports of conversion rate in advanced biliary tract cancer. Kato et al. (27) showed that only 9 of 50 cases became operable after chemotherapy. Future reports on the safety and benefits of multidisciplinary therapy are awaited.
Study limitations. Some of the data have been difficult to interpret because of the small number of cases. Our data suggest that patients with incidental distant metastases had a prognosis similar to that of CTx. It also suggested that pM(−) did not show significant differences with pM(+) (Figure 8). The reason for this discrepancy could be limited by small sample sizes. Previously, Mizuno et al. (5) indicated that patients who underwent major hepatectomy with distant metastases had a poor prognosis. This indicates that survival would not be improved by prophylactic liver resection, which could resect incidental liver metastases at S4/5. Recently, several meta-analyses have shown that prophylactic liver resection results in increased postoperative complications with no improvement in prognosis (28). Indication of HPD would be the same. Our surgical results of HPD for cholangiocarcinoma showed high morbidity and mortality (12). Therefore, indication of HPD for GBC should be discussed by a multicenter collaborative study. Furthermore, this was a retrospective cohort analysis involving a small number of patients from a single institution. Although we included CTx G, we did not evaluate all patients with GBC at our institution, including those who were not referred to a surgeon or an oncologist. Furthermore, patients who were surgical candidates and those who were not (and who received chemotherapy) might not be the best groups to compare; thus, the result of this comparison could be open to interpretation. Furthermore, these analyses were limited to patients who underwent exploration.
Recently BILCAP, ABC02 trials and ASCOT study recommended the adjuvant therapy for biliary malignancies (29, 30). However, there is no evidence supporting this recommendation for preoperative therapy. As emphasized in a previous study, a prospective study with clear inclusion criteria for locally advanced and high-risk patients would better determine the role of neoadjuvant chemotherapy in GBC (19).
In conclusion, although extended surgery may be considered for patients with GBC, careful patient selection and new therapeutic strategies are required, especially for patients who require VR.
Footnotes
Authors’ Contributions
Conceptualization: T.No. and S.H.; methodology: T.No. and S.H.; data curation: T.No, M.W., K.T., A.M., Y.N., T.A., T.Na, K.H, Y.H. S.T.; writing – original draft preparation: T.No.; writing – review and editing: M.W., K.T., A.M., Y.N., T.A., T.Na.; visualization: T.No.; supervision: S.H.; project administration: T.No. All Authors have read and agreed to the published version of the manuscript.
Funding
This research received no external funding.
Conflicts of Interest
All Authors declare no conflicts of interest in relation to this study.
- Received December 17, 2024.
- Revision received December 30, 2024.
- Accepted January 2, 2025.
- Copyright © 2025 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).














