Abstract
Background:
Duodenal involvement occurs frequently in gallbladder cancer (GBC) as a result of the proximity of the duodenum to the gallbladder.
Methods:
The study group included 74 GBC patients assessed between August 2009 and March 2011 in whom computed tomography (CT) of the abdomen indicated suspicion for duodenal involvement.
Results:
Of 172 patients with resectable GBC, 74 (43.0%) had suspected duodenal involvement on imaging. Of these, 51 (68.9%) had suspected duodenal involvement on upper gastrointestinal endoscopy (UGIE). Symptoms of gastric outlet obstruction (GOO) were present in only 14 (18.9%) patients. Thirteen (17.6%) patients underwent staging laparoscopy alone. Of the 61 patients who underwent laparotomy, 31 (50.8%) were found to have actual duodenal involvement. The positive predictive value (PPV) of CT of the abdomen for duodenal involvement was 50.8% (31 of 61 patients). The addition of UGIE increased the PPV to 65.9% (27 of 41 patients). In the subgroup with evidence of duodenal mural thickening or mucosal irregularity on CT of the abdomen (n= 9) or duodenal mucosal infiltration on UGIE (n= 14), the PPV increased to 100%. A total of 33 (44.6%) patients underwent curative resection. The resectability rate was significantly lower in patients with symptoms of GOO [two of 14 (14.3%) vs. 31 of 60 (51.7%); P= 0.010], CT findings of duodenal mural thickening or mucosal irregularity compared with only loss of the fat plane [two of 12 (16.7%) vs. 31 of 62 (50.0%); P= 0.032], and UGIE evidence of duodenal infiltration compared with extrinsic compression or normal endoscopic findings [three of 16 (18.8%) vs. 18 of 35 (51.4%) and 12 of 23 (52.2%), respectively; P= 0.027 and P= 0.036, respectively].
Conclusions:
Overall, CT of the abdomen demonstrated a PPV of 50.8% in detecting duodenal involvement, which increased to 65.9% with the addition of UGIE. The combined presence of GOO symptoms, CT findings of duodenal mural thickening and mucosal irregularity, and UGIE findings of infiltration of the duodenal mucosa significantly decreases resectability but does not preclude resection.
Introduction
Gallbladder cancer (GBC) is the most common malignancy of the biliary tract and frequently presents at an advanced stage with adjacent organ involvement.1 Although the liver is the most commonly involved adjacent organ, duodenal involvement occurs frequently as a result of the proximity of the duodenum to the gallbladder.2 Whereas some authors have considered duodenal involvement in GBC to represent a sign of unresectable advanced disease, others have shown that radical resection improves survival in these patients.3–7 However, there is controversy in the literature regarding the extent of resection, which ranges from limited resection of the stomach and duodenum to hepatopancreatoduodenectomy (HPD).5–7 Although limited resection of the stomach and duodenum can be performed with minimal morbidity, HPD is associated with significant morbidity and mortality.7 Cross-sectional imaging [computed tomography (CT) of the abdomen, magnetic resonance imaging (MRI) of the abdomen] and upper gastrointestinal endoscopy (UGIE) are commonly used investigations to detect duodenal involvement.7–10 It is of paramount importance to ascertain whether and to what extent the duodenum is infiltrated by the tumour in order to establish a management strategy. However, the data available on the accuracy of these investigations to predict actual duodenal involvement are limited. Hence, this study was undertaken to assess the predictive value of imaging and UGIE in detecting duodenal involvement. The significance of duodenal involvement on preoperative evaluation was assessed by correlating such findings with resectability rates and extents of surgery.
Materials and methods
This was a prospective study conducted over 20 months from August 2009 to March 2011. All patients suspected to have GBC underwent an ultrasound examination of the abdomen. Patients without evidence of dissemination underwent a dual-phase CT scan of the abdomen using a 16-slice CT scanner. Patients in whom CT showed suspicions of duodenal involvement formed the study group. These patients were further evaluated with UGIE followed by staging laparoscopy (SL). The criterion for duodenal involvement on dual-phase CT of the abdomen was defined as loss of the fat plane between the gallbladder mass and the duodenum with or without evidence of duodenal mural thickening and mucosal irregularity (Figs 1 and 2). These patients were further evaluated with UGIE. Findings on UGIE were classified as normal, demonstrative of extrinsic compression without mucosal involvement, or indicative of obvious tumour infiltration into the duodenal lumen with or without complete gastric outlet obstruction (GOO) (Fig. 3). Endoscopic biopsy was taken when there was evidence of tumour infiltration into the duodenum on UGIE.
Figure 1.
(a,b) Computed tomography (CT) of the abdomen showing only loss of the fat plane between the gallbladder mass and the duodenum without mural thickening (arrow)
Figure 2.
Computed tomography (CT) of the abdomen showing loss of the fat plane between the gallbladder mass and the duodenum with mural thickening and mucosal irregularity (arrow)
Figure 3.
Upper gastrointestinal endoscopy showing mucosal infiltration with luminal narrowing
All patients deemed resectable on preoperative imaging underwent an SL prior to a laparotomy. Suspicious metastatic lesions on SL were biopsied and confirmed by frozen-section analysis. Palliative gastrojejunostomy (GJ) was performed in patients symptomatic for GOO (Fig. 3). In patients in whom no metastatic disease was found on SL, laparotomy was performed through a right subcostal incision. Routine sampling of inter-aortocaval lymph nodes was performed in all patients and a positive result on frozen-section histopathology was considered to indicate metastatic disease and the resection was abandoned. The type of resection was determined by the loco-regional extent of the tumour according to liver and adjacent organ infiltration (Fig. 4). In patients with limited duodenal involvement, a sleeve resection of the duodenum was performed. In patients with more extensive involvement (without pancreatic infiltration) and those in whom a luminal compromise was expected after sleeve resection of the duodenum, a distal gastrectomy with proximal duodenectomy (DGPD) was performed. This included resection of a variable length of the first and suprapapillary portion of the second part of the duodenum (depending upon the extent of duodenal involvement). Bowel continuity in these patients was restored by a gastrojejunostomy. Patients with limited pancreatic infiltration underwent a wedge resection of the pancreas. Patients with more extensive pancreatic involvement or involvement of the duodenal papilla underwent an HPD. Common bile duct (CBD) excision was selectively performed when there was direct infiltration or extensive nodal disease in the hepatoduodenal ligament and in patients with GBC with an underlying choledochal cyst. Patients with colonic involvement underwent either a limited colonic resection or a right hemicolectomy depending upon the extent of colonic infiltration. Standard lymphadenectomy was performed in all patients. This included the clearance of nodes in the hepatoduodenal ligament, nodes anterior and posterior to the head of the pancreas, and nodes along the hepatic artery until its origin from the coeliac axis. Symptoms of GOO, CT scan findings and UGIE findings were correlated with intraoperative findings and resectability rates in these patients. Statistical analysis was performed using GraphPad instat Version 4 (GraphPad Software, Inc., La Jolla, CA, USA). Categorical variables were summarized using proportions and were compared using Fisher's exact test. A P-value of <0.05 was considered to indicate statistical significance.
Figure 4.
Algorithm for the management of gallbladder cancer patients with suspected duodenal involvement on imaging. UGIE, upper gastrointestinal endoscopy
Results
During the study period, 318 patients with GBC were assessed for operability. Of these, 146 (45.9%) patients were deemed to be unresectable according to either metastatic disease or poor general condition. Of the 172 patients considered to be resectable on preoperative imaging, 74 (43.0%) demonstrated suspicions of duodenal involvement on CT of the abdomen; these patients represented the study group. The mean ± standard deviation (SD) age in this group was 53.4 ± 8.3 years (range: 29–86 years). A total of 59 (79.7%) patients were female (female : male ratio: 3.9:1). Abdominal pain was the most common clinical symptom and was reported by 100% of patients. Symptoms suggestive of GOO were present in 14 (18.9%) patients and jaundice was present in 16 (21.6%) patients (Table 1).
Table 1.
Demographics and clinical features of gallbladder cancer patients with duodenal involvement on imaging (n= 74)
| Clinical parameter | |
|---|---|
| Age, years, mean ± SD | 53.4 ± 8.3 |
| Sex, male : female | 1:3.9 |
| Abdominal pain | 100% (74/74) |
| Gastric outlet obstruction | 18.9% (14/74) |
| Jaundice | 21.6% (16/74) |
| Anorexia/weight loss | 94.6% (70/74) |
| Palpable gallbladder mass | 64.9% (48/74) |
SD, standard deviation.
On CT scan of the abdomen, all 74 patients were found to have loss of the fat plane between the GB mass and the duodenum. Twelve (16.2%) patients also had duodenal mural thickening with mucosal irregularity. Examination by UGIE revealed extrinsic compression without mucosal involvement in 35 (47.3%) patients and obvious infiltration of the tumour into the duodenal lumen with or without complete luminal obstruction in 16 (21.6%) patients. The remaining 23 (31.1%) patients demonstrated normal findings on UGIE (Table 2). Of the 74 patients who underwent SL, 24 (32.4%) were found to have disseminated disease in the form of surface liver metastasis (n= 10) or peritoneal deposits (n= 14). Of these, 11 patients with symptoms of GOO underwent palliative GJ and 13 patients in whom duodenal infiltration by the tumour could not be assessed did not undergo any further procedure. Hence, intraoperative assessment for duodenal involvement was possible in only 61 (82.4%) patients, 31 (50.8%) of whom were found to have actual duodenal involvement by the GBC. Using CT scan findings to correlate with intraoperative assessment, if only loss of the fat plane was taken as the criterion for duodenal involvement, the positive predictive value (PPV) was 50.8% (31 of 61 patients). The PPVs of duodenal mural thickening with mucosal irregularity on CT of the abdomen, and extrinsic compression and obvious infiltration of the tumour into the duodenal lumen on UGIE are summarized in Table 2.
Table 2.
Positive predictive value of preoperative investigations for duodenal involvement
| Preoperative investigation | Patients, n | Positive predictive value | ||
|---|---|---|---|---|
| Suspected duodenal involvement | Intraoperative assessment possible | Actual involvement (intraoperative assessment) | ||
| Computed tomography of the abdomen | ||||
| Loss of fat plane | 74 | 61 | 31 | 50.8% |
| Duodenal mural thickening with mucosal irregularity | 12 | 9 | 9 | 100% |
| Upper gastrointestinal endoscopy | ||||
| Overall | 51 | 41 | 27 | 65.9% |
| Extrinsic compression | 35 | 27 | 13 | 48.2% |
| Duodenal infiltration | 16 | 14 | 14 | 100% |
Types of surgery performed and overall resectability rates in these patients are summarized in Table 3. Of the 74 patients who underwent SL, 24 (32.4%) had disseminated disease. Of these 24 patients, 13 underwent SL alone because they did not require palliation and the remaining 11 patients underwent laparotomy to achieve a palliative GJ. Of the 50 patients who underwent laparotomy with curative intent, 17 (34.0%) were found to have unresectable disease as a result of liver metastasis (n= 1), inter-aortocaval lymph node metastasis (n= 7) or locally advanced unresectable disease (n= 9). Overall, curative resection was performed in 33 (44.6%) patients. Radical cholecystectomy with resection of segments IVb and V was performed in 31 patients, resection of segments IVb, V and VI was performed in one patient and extended right hepatectomy was performed in one patient. Standard lymphadenectomy (as previously described) was performed in all patients. Of the 33 patients who underwent curative resection, 18 were found to have duodenal involvement on intraoperative assessment. Eight patients in whom a small area of contact with the duodenum was identified underwent a duodenal sleeve resection; 10 patients in whom the area of contact with the duodenum was wide underwent a DGPD. A wedge resection of the pancreas was performed in two patients with limited pancreatic infiltration. Indications for CBD excision and types of colonic resection are summarized in Table 3. A palliative procedure was performed in 19 patients, of whom 17 underwent a GJ alone and two underwent an additional ileo-transverse bypass performed to palliate obstruction of the large bowel secondary to infiltration of the hepatic flexure.
Table 3.
Extent of resection in patients who underwent curative resection (n= 33)
| Extent of resection | Patients, n |
|---|---|
| Liver resection | |
| Segments IVb and V | 31 |
| Segments IVb, V and VI/extended right hepatectomy | 2 |
| Duodenal involvement | 18 |
| Duodenal sleeve resection | 8 |
| Distal gastrectomy with proximal duodenectomy | 10 |
| Wedge pancreatectomy | 2 |
| Common bile duct excision | 13 |
| Common bile duct involvement with jaundice | 5 |
| Choledochal cyst | 2 |
| Positive cystic duct margin | 2 |
| Lymph node clearance | 4 |
| Colon resection | |
| Segmental colectomy/right hemicolectomy | 7 |
The significance of duodenal involvement was analysed by correlating resectability rates with symptoms, and CT and UGIE findings (Table 4). The resectability rate was significantly lower in patients with symptoms of GOO compared with patients without these symptoms [two of 14 (14.3%) vs. 31 of 60 (51.7%); P= 0.010]. Correlations of findings in CT scans of the abdomen showed the resectability rate to be lower in patients with mural thickening and mucosal irregularity compared with those with only loss of the fat plane between the gallbladder mass and the duodenum [two of 12 (16.7%) vs. 31 of 62 (50.0%); P= 0.032]. Patients with evidence of duodenal involvement on CT combined with mucosal infiltration on UGIE had a significantly lower resectability rate than patients with only CT evidence of duodenal involvement [three of 16 (18.8%) vs. 12 of 23 (52.2%); P= 0.036] or CT evidence with extrinsic compression on UGIE [three of 16 (18.8%) vs. 18 of 35 (51.4%); P= 0.027]. An analysis of the causes of unresectability showed the incidence of metastatic disease to be significantly higher in patients with imaging evidence of duodenal involvement and endoscopic evidence of duodenal infiltration compared with patients with an extrinsic bulge or a normal endoscopy [11 of 16 (68.8%) vs. 21 of 58 (36.2%); P= 0.020]. After excluding patients with metastatic disease, the incidence of locally advanced unresectable disease was higher in patients with endoscopic evidence of duodenal infiltration than in patients with an extrinsic bulge or a normal endoscopy [two of five (40.0%) vs. seven of 37 (18.9%)], although this difference did not reach statistical significance (P= 0.288).
Table 4.
Correlations of resectability rate with symptoms and findings on computed tomography (CT) and upper gastrointestinal endoscopy (UGIE)
| Parameter | Resectability rate | P-value |
|---|---|---|
| Symptoms of gastric outlet obstruction | ||
| Present (n= 14) | 14.3% (2/14) | 0.010 |
| Absent (n= 60) | 51.7% (31/60) | |
| CT findings | ||
| Loss of fat plane/mural thickening (n= 74) | 44.6% (33/74) | 0.032a |
| Only loss of fat plane (n= 62) | 50.0% (31/62) | |
| Loss of fat plane with duodenal mural thickening with mucosal irregularity (n= 12) | 16.7% (2/12) | |
| UGIE findings | ||
| CT involvement + normal UGIE | 52.2% (12/23) | 0.584b |
| CT involvement + extrinsic compression on UGIE | 51.4% (18/35) | 0.027c |
| CT involvement + duodenal mucosal infiltration | 18.8% (3/16) | 0.036d |
Only loss of fat plane vs. loss of fat plane with duodenal mural thickening with mucosal irregularity.
CT involvement + normal UGIE vs. CT involvement + extrinsic compression on UGIE.
CT involvement + extrinsic compression on UGIE vs. CT involvement + duodenal mucosal infiltration.
CT involvement + normal UGIE vs. CT involvement + duodenal mucosal infiltration.
Discussion
Although the liver is the most commonly involved adjacent organ, duodenal involvement occurs frequently in patients with GBC as a result of the proximity of the duodenum to the gallbladder.2 Cross-sectional imaging in the form of CT or MRI of the abdomen and UGIE are commonly used to diagnose duodenal involvement preoperatively in patients with GBC.7–10 There is, however, only limited evidence in the available literature on the accuracy of preoperative investigations in predicting duodenal involvement. Some reports have considered duodenal involvement as a sign of unresectable advanced disease.3,4,11 Others have considered it as an indication for extensive surgery in the form of HPD, a procedure which has been reported to incur significant morbidity and mortality.5,6 It has been previously reported that duodenal involvement alone should not be considered as a sign of unresectable disease and that a negative margin (R0) resection can be achieved in many of these patients by performing a duodenal sleeve resection or DGPD.7 Hence, the present study was conducted to prospectively analyse the accuracy of preoperative assessment of duodenal involvement based on symptoms, imaging and UGIE and to ascertain its management implications.
Clinically, patients with duodenal involvement may present with symptoms of GOO manifested by vomiting. The information available in the literature on the predictive value of clinical symptoms for duodenal involvement is limited. In the present study, all patients with symptoms of GOO in whom an intraoperative assessment was possible showed evidence of duodenal involvement. By contrast, not all patients with imaging and endoscopic evidence of duodenal mucosal infiltration had symptoms of GOO (n= 2). To summarize, although the presence of complete GOO symptoms had high PPV for detecting duodenal involvement, the absence of GOO symptoms does not exclude it.
Computed tomography of the abdomen is commonly used in the staging of GBC. Yoshimitsu et al. reported an accuracy of 83–86% for stage T2 disease and above, but reported poor sensitivity for T1 lesions.12 Kim et al. reported an overall accuracy of 71% for assessing T-stage in 100 consecutive patients with GBC.8 However, few studies have focused on assessing the accuracy of CT of the abdomen in detecting duodenal involvement. The criterion most commonly used in these studies is loss of the fat plane between the gallbladder mass and the duodenum. Ohtani et al. reported a sensitivity of 57% for detecting duodenal involvement on a CT scan based on an analysis of seven patients with GBC.13 Kalra et al. reported an accuracy of 85% for detecting duodenal involvement, based on an analysis of 11 patients.9 In the present study, when loss of the fat plane between the gallbladder tumour and the duodenum on CT was used as the sole criterion for predicting duodenal involvement, the PPV was 50.8%. However, when duodenal mural thickening with mucosal irregularity was used as an additional criterion to loss of the fat plane, the PPV rose to 100% in the small subgroup of nine patients.
Upper gastrointestinal endoscopy is commonly used to confirm the finding of suspected duodenal involvement on imaging. However, current data on the predictive value of endoscopic findings are limited. In UGIE, duodenal involvement may be seen as either an extrinsic compression or infiltration of the duodenal mucosa. In the current study, the presence of extrinsic compression alone had a PPV of 48.1%, which increased to 100% when there was infiltration of the duodenal mucosa. This may reflect a large gallbladder tumour that abuts or indents the duodenum without actually infiltrating it. This finding (extrinsic compression) is significant because curative resection can be performed in many of these patients without adding a duodenal sleeve resection or DGPD.
In order to ascertain the significance of preoperative assessment of duodenal involvement, findings on preoperative evaluation were correlated with resectability rates. Some previous studies have suggested that the presence of GOO symptoms is indicative of advanced disease and that only palliative therapy is indicated.3,4,11 In the present study, patients with symptoms of GOO had a lower resectability rate than those without these symptoms (14.3% vs. 51.7%; P= 0.010). However, two of the 14 patients with complete GOO symptoms underwent curative resection and hence the presence of GOO symptoms should not be considered a sign of unresectability in isolation. The correlation of CT and UGIE findings with resectability rates showed that resectability rates in patients with evidence of duodenal involvement on CT combined with mucosal infiltration on UGIE were significantly lower than those in patients with only CT evidence of duodenal involvement (18.8% vs. 52.2%; P= 0.036) or CT evidence with extrinsic compression on UGIE (18.8% vs. 51.4%; P= 0.027). An analysis of the causes of unresectability showed the frequency of metastatic disease was significantly higher in patients with endoscopic evidence of duodenal mucosal infiltration compared with patients with an extrinsic bulge or normal endoscopy (68.8% vs. 36.2%; P= 0.020). To summarize, the presence of symptoms of GOO, and evidence of duodenal mural thickening with mucosal irregularity on CT of the abdomen and mucosal infiltration on UGIE have a high PPV for duodenal involvement and are associated with a higher incidence of metastatic disease and a significantly lower rate of resectability.
There is heterogeneity in the literature regarding the magnitude of surgery required in patients with GBC with duodenal involvement. Moreover, the majority of published studies have analysed simultaneous duodenopancreatic involvement and have advocated an HPD in all such instances.14,15 This procedure, however, has been reported to be associated with high rates of morbidity and mortality. Ogura et al. analysed 1686 patients with GBC who underwent resection in 172 major hospitals in Japan.16 Of these, 150 patients underwent HPD, in which the overall rate of major complications was 54%.16 A previous study reported that duodenal sleeve resection or DGPD can be safely performed in such patients with minimal morbidity and equivalent longterm survival.7 Kondo et al. also described sleeve resection of the duodenum in GBC with limited duodenal involvement.17 In the present series, of the 18 patients with actual duodenal involvement, R0 resection was achieved in eight patients by performing a duodenal sleeve resection and in 10 by DGPD. In a significant proportion of patients with combined duodenopancreatic involvement, a DGPD with a wedge of pancreas can achieve an R0 resection and avoid an HPD. Hirano et al. compared operative and longterm outcomes in patients who underwent wedge resection (n= 9) and HPD (n= 8) and found no differences in local recurrence or cumulative survival rates.18 In the present series, two patients with limited pancreatic involvement were managed with wedge resection of the pancreas without substantially increasing the morbidity associated with the procedure.
Conclusions
In the present study, CT of the abdomen was found to have a PPV of 50.8% in detecting duodenal involvement. The addition of UGIE in these patients increased the PPV to 65.9%. In the subgroup with duodenal mural thickening and mucosal irregularity on CT of the abdomen, and duodenal mucosal infiltration on UGIE, the PPV increased to 100%. The presence of symptoms suggestive of complete GOO, duodenal mural thickening or mucosal irregularity on CT of the abdomen and infiltration of the duodenal mucosa on UGIE significantly decreases resectability in these patients, but does not preclude resection. In the majority of patients with actual duodenal infiltration, R0 resection can be achieved with limited resection of the stomach and duodenum.
Conflicts of interest
None declared.
References
- 1.Wilkinson DS. Carcinoma of the gallbladder – an experience and review of the literature. Aust N Z J Surg. 1995;65:724–727. doi: 10.1111/j.1445-2197.1995.tb00545.x. [DOI] [PubMed] [Google Scholar]
- 2.Fahim RB, McDonald JR, Richards JC, Ferris DO. Carcinoma of the gallbladder: a study of its modes of spread. Ann Surg. 1962;156:114–124. doi: 10.1097/00000658-196207000-00021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Agrawal S, Sonawane RN, Behari A, Kumar A, Sikora SS, Saxena R. Laparoscopic staging in gallbladder cancer. Dig Surg. 2005;22:440–445. doi: 10.1159/000091447. et al. [DOI] [PubMed] [Google Scholar]
- 4.Singh B, Kapoor VK, Sikora SS, Kalawat TC, Das BK, Kaushik SP. Malignant gastroparesis and outlet obstruction in carcinoma gallbladder. Trop Gastroenterol. 1998;19:37–39. [PubMed] [Google Scholar]
- 5.Shimizu H, Kimura F, Yoshidome H, Ohtsuka M, Kato A, Yoshitomi H. Aggressive surgical approach for stage IV gallbladder carcinoma based on Japanese Society of Biliary Surgery classification. J Hepatobiliary Pancreat Surg. 2007;14:358–365. doi: 10.1007/s00534-006-1188-z. et al. [DOI] [PubMed] [Google Scholar]
- 6.D'Angelica M, Martin RC, 2nd, Jarnagin WR, Fong Y, DeMatteo RP, Blumgart LH. Major hepatectomy with simultaneous pancreatectomy for advanced hepatobiliary cancer. J Am Coll Surg. 2004;198:570–576. doi: 10.1016/j.jamcollsurg.2003.11.022. [DOI] [PubMed] [Google Scholar]
- 7.Agarwal AK, Mandal S, Singh S, Sakhuja P, Puri S. Gallbladder cancer with duodenal infiltration: is it still resectable? J Gastrointest Surg. 2007;11:1722–1727. doi: 10.1007/s11605-007-0320-y. [DOI] [PubMed] [Google Scholar]
- 8.Kim SJ, Lee JM, Lee JY, Choi JY, Kim SH, Han JK. Accuracy of preoperative T-staging of gallbladder carcinoma using MDCT. AJR Am J Roentgenol. 2008;190:74–80. doi: 10.2214/AJR.07.2348. et al. [DOI] [PubMed] [Google Scholar]
- 9.Kalra N, Suri S, Gupta R, Natarajan SK, Khandelwal N, Wig JD. MDCT in the staging of gallbladder carcinoma. AJR Am J Roentgenol. 2006;186:758–762. doi: 10.2214/AJR.04.1342. et al. [DOI] [PubMed] [Google Scholar]
- 10.Khilani MT, Wolf BS, Finkel M. Roentgen features of carcinoma of the gallbladder on barium-meal examination. Radiology. 1962;79:264–273. doi: 10.1148/79.2.264. [DOI] [PubMed] [Google Scholar]
- 11.Chaudhary A, Dhar P, Sachdev A, Agarwal A. Gastric outlet obstruction in carcinoma gall bladder. Indian J Gastroenterol. 1999;18:101–103. [PubMed] [Google Scholar]
- 12.Yoshimitsu K, Honda H, Shinozaki K, Aibe H, Kuroiwa T, Irie H. Helical CT of the local spread of carcinoma of the gallbladder: evaluation according to the TNM system in patients who underwent surgical resection. AJR Am J Roentgenol. 2002;179:423–428. doi: 10.2214/ajr.179.2.1790423. et al. [DOI] [PubMed] [Google Scholar]
- 13.Ohtani T, Shirai Y, Tsukada K, Muto T, Hatakeyama K. Spread of gallbladder carcinoma: CT evaluation with pathologic correlation. Abdom Imaging. 1996;21:195–201. doi: 10.1007/s002619900045. [DOI] [PubMed] [Google Scholar]
- 14.Kokudo N, Makuuchi M, Natori T, Sakamoto Y, Yamamoto J, Seki M. Strategies for surgical treatment of gallbladder carcinoma based on information available before resection. Arch Surg. 2003;138:741–750. doi: 10.1001/archsurg.138.7.741. et al. [DOI] [PubMed] [Google Scholar]
- 15.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;11:45–49. doi: 10.1007/s00534-003-0867-2. [DOI] [PubMed] [Google Scholar]
- 16.Ogura Y, Mizumoto R, Isaji S, Kusuda T, Matsuda S, Tabata M. Radical operations for carcinoma of the gallbladder. World J Surg. 1991;15:337–343. doi: 10.1007/BF01658725. [DOI] [PubMed] [Google Scholar]
- 17.Kondo S, Nimura Y, Hayakawa N, Kamiya J, Nagino M, Uesaka K. Extensive surgery for carcinoma of the gallbladder. Br J Surg. 2002;89:179–184. doi: 10.1046/j.0007-1323.2001.02001.x. [DOI] [PubMed] [Google Scholar]
- 18.Hirano S, Tanaka E, Shichinohe T, Saitoh K, Takeuchi M, Senmaru N. Feasibility of en-bloc wedge resection of the pancreas and/or the duodenum as an alternative to pancreatoduodenectomy for advanced gallbladder cancer. J Hepatobiliary Pancreat Surg. 2007;14:149–154. doi: 10.1007/s00534-006-1109-1. et al. [DOI] [PubMed] [Google Scholar]