ORIGINAL RESEARCH

Evaluation of a Prospective Surgical Strategy of ExtendedResection to Achieve R0 Status in Gall Bladder Cancer

Biju Pottakkat & Abhimanyu Kapoor & Anand Prakash &

Rajneesh Kumar Singh & Anu Behari & Ashok Kumar &

Vinay K. Kapoor & Rajan Saxena

Published online: 18 September 2012# Springer Science+Business Media, LLC 2012

AbstractIntroduction Radical resection to achieve R0 status remainsthe only potential curative option in patients with gall bladdercancer (GBC). This study was aimed to evaluate the efficacyof an extended criterion of radical resection to achieve R0status in GBC.Methods A triple-phase CTwith 3D reconstruction was donein all patients. A standard resectability criterion was followedin all patients. A minimum of liver segment 4B + 5 resectionand radical lymphadenectomy including the para-aortic areaswere undertaken in all patients. Adjacent organectomy wasadded as required.Results Between November 2008 and April 2011, 59 patientswith GBC underwent operation and 40 (resectability, 68 %)underwent resection. The resectional procedures performedwere segmentectomy 4B + 5 in 31 (78 %), median sectorec-tomy in 2 (5 %), extended right hepatectomy in 3 (8 %), andhepatopancreaticoduodenectomy in 4 (10 %) patients. Post-operative complications occurred in 24 (60 %) patients. Twopatients died postoperatively. A total of 829 lymph nodes wereharvested and the median lymph node count was 18 (4–77).Twenty-three (58 %) patients had lymph node metastases.Twenty-eight of 40 (70 %) had disease limited till N1 nodes.Metastases up to N2 lymph nodes were seen in 12 (30 %).American Joint Committee on Cancer seventh edition stageswere I—2 (5 %) patients, II—5 (13 %), III—19 (48 %), andIV—14 (35 %). R0 resection was achieved in 33 (83 %)patients. Four patients had recurrence and one died of recur-rence. All other patients are alive till the last follow-up.

Conclusions Assessment with triple-phase CT with 3D re-construction can produce high resectability rate in GBC.Extended criterion of radical resection results in R0 statusin more than 80 % of patients with GBC.

Keywords Cancer . Carcinoma . Gall bladder . Survival

Introduction

Gall bladder cancer (GBC) carries poor prognosis. Surgicalresection remains the only available potentially curativeoption, but only 10 % of patients are considered as surgicalcandidates [1]. The chance of resectability is precluded bydistant metastasis. Lymph nodes, liver, bile duct, duodenum,and colon are frequent sites of loco-regional spread. Theproportion of patients undergoing resection varies fromcenter to center based on their respective resectability crite-ria. Discrepancy exists between the eastern and westernliterature in terms of what constitutes an acceptable extentof resection [2].

GBC is common in northern India. It is the commonestgastrointestinal malignancy in women in northern India. Wehave earlier reported that R0 resection is the most importantfactor influencing survival in patients with GBC in whom anextended cholecystectomy—EC (cholecystectomy + 2 cmliver wedge resection + lymphadenectomy along the hepa-toduodenal ligament (HDL)—behind the duodenum andbehind the pancreatic head) was performed [3]. Among thepatients with non-metastatic disease, R0 resection can beachieved with EC in stages I and II, but these patientsrepresent only minority of the cases, while majority of casesbelong to stages III and IV and are outside the scope of EC.Most of the published reports of extended resections areretrospective and patients underwent nonuniform resections.Although resectional strategies tailored to the T stage have

B. Pottakkat :A. Kapoor :A. Prakash : R. K. Singh :A. Behari :A. Kumar :V. K. Kapoor (*) : R. SaxenaDepartment of Surgical Gastroenterology,Sanjay Gandhi Post Graduate Institute of Medical Sciences,Rae Bareli Road,Lucknow 226014, UP, Indiae-mail: vkkapoor.india@gmail.com

J Gastrointest Canc (2013) 44:33–40DOI 10.1007/s12029-012-9432-z

been suggested by some authors, an accurate T staging isdifficult on preoperative imaging and is possible only athistopathology after resection, and hence, prospective ap-plication of these strategies is difficult. A strategy forextended resection has to be employed to include bothearly as well as advanced non-metastatic GBC. Here, weevaluate the feasibility and efficacy of a prospective strategyof extended resection to achieve R0 status in all patients withnon-metastatic GBC.

Methods

Assessment of Resectability

The diagnosis of GBC was entertained upon the detection ofa mass in the gallbladder (GB) on imaging. Patients withsuspected GBC were initially assessed for metastatic dis-ease. An ultrasonogram of the abdomen was performed inall patients. Those with ascites or secondaries in the liver/peritoneum without jaundice were considered for palliativechemotherapy. Those with distant metastasis (liver and peri-toneum) with biliary obstruction and jaundice were consid-ered for palliation of jaundice by metallic stenting. Beforeany palliation, diagnosis of the malignancy was establishedby fine needle aspiration cytology. Those without distantmetastasis or ascites were subjected to a triple-phase multi-slice (64/128), contrast-enhanced computed tomogram(CECT) to assess the resectability. Presence of obstructivejaundice and/or gastric outlet obstruction was not consideredas a contraindication for resection. Patients with distantmetastasis detected on CECT were offered non-resectionalpalliative therapies. CECT data in the remaining patientswere analyzed in detail for the extent of loco-regionalspread. Patients with more than 1-cm-sized lymph nodesalong the aorta, tumor involvement of the main portal vein(MPV), common hepatic artery (CHA), or bilobar secondorder bile ducts were considered unresectable.

3D reconstruction was derived from the acquired CECTdata in patients considered for resection. The segmental extentof local infiltration to the liver was assessed. The course oflobar and sectoral hepatic arteries was reconstructed by a CTarteriography protocol and looked for tumor infiltration. Theportovenogram was reconstructed to look for involvement ofthe portal vein (PV) branches. The branches of segment 4Bfrom the pars umbilicus portion of the left PV were specificallydelineated. A hepatic venography was also done for mappingof Couinaud's segments. In patients with dilated bile ducts, anon-contrast CT cholangiography was also performed. A per-cutaneous transhepatic biliary drainage (PTBD) tube cholan-giogram was done in patients with PTBD tube in situ placedearlier for obstructive jaundice. Magnetic resonance cholangio-pancreaticogram (MRCP) was performed in some patients in

whom a clear delineation of the proximal biliary system wasneeded. Based upon all the information available, a surgicalplan was derived on a case-to-case basis. Tissue diagnosiswas not attempted in patients considered for resection as thedisease is very common in this part of the country; more-over, a negative tissue diagnosis would not have altered theplan of resection.

In patients with jaundice, the need for a preoperative biliarydrainage was decided based on the total serum bilirubin. PTBDwas the preferred mode of biliary drainage and the endoscopicstentingwas done in patients with biliary obstruction below thehilum. Every effort was made to bring down the total serumbilirubin to less than 5 mg/dL before the planned resection.Preoperative portal vein embolisation (PVE) was consideredonly in patients planned for extended right hepatectomy (ERH)without the adequate calculated post-resection residual livervolume. Patients with poor performance status even after ag-gressive nutritional support were excluded from the resection.

Surgical Strategy

A staging laparoscopy was performed on table before theplanned operation. In the event of detection of distant metas-tasis or ascites, tissue/fluid was sent for frozen section histol-ogy/cytology and further procedure was abandoned after apositive report of malignancy. In patients with no dissemina-tion, a laparotomy was performed. Although local extent ofthe tumor was assessed from the outset at laparotomy, in viewof the difficulties in technical assessment of vascular struc-tures and upper bile duct on table, especially in locally ad-vanced GBC, the planned resection was proceeded based onthe preoperative imaging.

Resection of the Liver The planned resection included thearea supplied by the PV branch to the corresponding segmentsharboring the GB. The minimum extent of liver resectionplanned was central inferior hepatic sub-segmentectomy (seg-mentectomy 4B + 5). An extended right hepatectomy (seg-mentectomy 4B + 5–8) was planned whenever the tumorextended beyond the central inferior segment, in cases whereright portal vein (RPV) was involved, or if the sole arterialsupply to the entire right hemiliver was involved by thegrowth. In cases where only one of the sectoral arteries wasinvolved, it was ligated. The liver resection was done withCavitron Ultrasonic Surgical Aspirator. Bile duct excision wasdone only in patients with direct bile duct involvement by thetumor. The proximal and distal bile duct margins were sent forfrozen section and a negative margin was always desired.Adjacent organectomy was done if the preoperative CT dem-onstrated infiltration. A hepatopancreaticoduodenectomy(HPD) was performed in patients with large lymph nodes inthe retropancreatic region or in patients with extensive duode-nal or any pancreatic infiltration.

34 J Gastrointest Canc (2013) 44:33–40

Lymphadenectomy A philosophy of extended lymphade-nectomy was adopted irrespective of the local extent of thetumor. The resected lymph nodes included the possiblelymphatic pathways reported by Ito et al. [4]. The lym-phadenectomy regions are delineated in Fig. 1. This cor-responds with the N1 and N2 lymph nodes in theAmerican Joint Committee on Cancer Staging seventhedition including the celiac axis and superior mesentericartery lymph nodes [5]. In the para-aortic region, pre-caval,inter-aorto-caval, and pre-aortic areas were cleared of alltissues till the aortic bifurcation. A bipolar diathermy wasused for lymphadenectomy.

Patients were categorized as R0 if the all the marginswere free of tumor and the para-aortic lymph nodes were notinvolved. Those with margin positivity or para-aortic lymphnode positivity were considered as R1; involvement of otherN2 lymph nodes was not considered as R1 resection. R2resections were never considered as an intended choice.Patients with T2 and above lesions and any patient withlymph node metastasis were given the option of postoperativechemoradiation.

Results

This study included patients operated based upon this prospec-tive strategy between November 2008 and April 2011. Sixtypatients were considered for resection. In one patient, a centralinferior sub-segmentectomy with extended lymphadenectomywas done as the suspicion of malignancy was very highon preoperative and intra-operative assessment, but the finalhistopathology of the GB revealed xantho-granulomatous

cholecystitis with no evidence of malignancy; this patientwas excluded from further analysis. There were 16 (27 %)males and 43 (73 %) females. The median age was 51 (31–73)years. Thirteen (22 %) patients were referred with a patholog-ical diagnosis of GBC after cholecystectomy done elsewhere.Three patients underwent cholecystectomy at our center andwere detected to have GBC (incidental GBC) on pathologicalexamination. The main clinical features were pain (n051,86 %), weight loss (n032, 54 %), palpable abdominal mass(n024, 41 %), jaundice (n021, 36 %), and gastric outletobstruction (n04, 7 %). Forty-one (70 %) patients hadassociated gall stones. The median hemoglobin was 11.8(7.8–15.5) g/dL and the median serum albumin was 3.9(1.9–5.0) g/dL. Preoperative biliary drainage was done in 18(31%) patients (endoscopic stenting in nine and PTBD in nine)and two underwent preoperative portal vein embolization.PTBD gram was done in six (10 %) patients and seven(12 %) patients underwent MRCP. All patients were put onaggressive oral nutritional support for optimization beforeoperation.

Fourteen (24 %) patients were found to have disseminationon staging laparoscopy and dissemination was revealed in onemore patient on laparotomy; two patients had main portal veininvolvement by the tumor detected during dissection. In these17 (29 %) patients, the original plan for resection was aban-doned. Resection was performed in remaining 42 (71 % resect-ability) patients. In two patients with multiple comorbidities,the planned procedure could not be performed because of theintra-operative cardiovascular instability—one patient under-went extended lymphadenectomy but only with a 2-cm wedgeresection of the liver and the other underwent segmentectomy4B + 5 but only with standard lymphadenectomy (in the

Fig. 1 Extent of operation inextended lymphadenectomy: aN1 lymph nodes removed (areato the left of the dotted line)include those in thehepatoduodenal ligament—cystic nodes, pericholedochalnodes, nodes along the properhepatic artery (HAP), and itsbranches, periportal nodesaround the main portal vein(MPV). N2 nodes removed (areato the right of the dotted line)include lymph node along thecommon hepatic artery (CHA),celiac axis (CA), retropancreatic(RP) nodes. b N2 lymph nodesalso include nodes along thesuperior mesenteric artery (SMA)and in the aortocaval (AC) region

J Gastrointest Canc (2013) 44:33–40 35

hepatoduodenal ligament). These two patients were excludedfrom further analysis.

Overall, 40 (68 %) patients underwent the planned resec-tion. The tumor location in the GB is given in Table 1.Fourteen (35 %) patients had liver infiltration by the tumor;13 (33 %) had bile duct involvement and bile duct wasresected in these 13 patients. Multiple organs were involvedin 11 (28 %) patients. Seven (18 %) patients had righthepatic artery (RHA) involvement, one had right PV in-volvement, and three had involvement of more than onevascular structure. ERH was done in three patients. Righthepatectomy was not added in the remaining patients withRHA involvement as they had the presence of an accessoryartery to the right lobe or sector,

Variations in the vascular anatomy were seen in 16(40 %) patients. Fourteen (35 %) patients had variations inthe HA anatomy. One patient underwent segmental colec-tomy because of colonic involvement. Five (13 %) patientsunderwent segmental duodenal resection for tumor infiltra-tion to the first part of duodenum. Four (10 %) patientsunderwent HPD—indications were extensive duodenal infil-tration (n02), for lymph node clearance (n01), and associatedbile duct carcinoma (n01). The procedures performed aregiven in Table 2. Central inferior hepatic sub-segmentectomywas performed in 35 patients (four with pancreaticoduode-nectomy (PD)); median sectorectomy had to be performedin two patients as they did not have standard Couinaud'sanatomy as shown by 3D CT reconstruction (in one pa-tient, the MPV was branching into three and the mediansector supplied by the middle branch of MPV harboringthe GB was resected; in the other patient with right-sidedround ligament, there were three PV branches and hence amedian sectorectomy was done); ERH had to be performedin three patients. Five (13 %) patients had more than onebile duct in the remnant liver (two ducts in four and threeducts in one patient). After performing ductoplasties, asingle anastomosis (Roux-en-Y hepaticojejunostomy) was

achieved in 16 (40 %) patients; one underwent two bilio-enteric anastomoses.

Four (10 %) patients had intra-operative complications—segmental bile duct injury (n02, 5 %), inadvertent lefthepatic artery ligation (n01, 3 %), and injuries to rightgonadal vein, right posterior sectoral artery, and cisternachyli (n01, 3 % each). All these were tackled during theoperation itself. The median duration of operation was 474(240–960) min. The median blood loss was 200 (50–1,000)ml and blood transfusion was required in 12 (30 %) patients.Postoperative complications occurred in 24 (60 %) patients(Table 3); none required reexploration. All intra-abdominalcollections were managed by percutaneous drainage andother complications were managed expectantly. Twopatients died postoperatively. The first patient died of pneu-monitis and paroxysmal supraventricular tachycardia onpostoperative day13, and the second patient died on day8 of cardiogenic shock which resulted from ventricularfibrillation. All other patients recovered well. The medianduration of postoperative hospital stay in the remaining patientswas 9 (4–36) days.

Table 1 Location of the tumor in the gall bladder (n040)

Location N (%)

Fundus 9 (23)

Fundus and body 8 (20)

Body 3 (8)

Neck 9 (23)

Entire gall bladder 7 (18)

Multicenter 2 (5)

Not knowna 2 (5)

a Referred aftercholecystectomy from elsewhere

Table 2 Resectional procedures performed in patients with gall blad-der cancer (n040)

Operative procedure N (%)

Central inferior hepatic sub-segmentectomy(segmentectomy 4B + 5)

31 (78)

Median sectorectomy 2 (5)

Extended right hepatectomy (segmentectomy 4B + 5–8) 3 (8)

Hepatopancreaticoduodenectomy (segmentectomy 4B + 5and pancreaticoduodenectomy)

4 (10)

Bile duct resection (13), segmental duodenal resection (5),and colectomy (1)

Table 3 Postoperative complications after extended resections for gallbladder cancer (n040)

Postoperative complication N (%)

Wound infection 12 (30)

Intra-abdominal collection 9 (22.5)

Pulmonary complications 9 (22.5)

Delayed gastric emptying 8 (20)

Wound dehiscence 5 (12.5)

Bile leak 5 (12.5)

Ascites 4 (10)

Liver failure 3 (7.5)

Chyle leak 2 (5)

Cardiac complications 2 (5)

Intra-abdominal bleeding 1 (2.5)

Enterocutaneous fistula 1 (2.5)

36 J Gastrointest Canc (2013) 44:33–40

The histopathology revealed adenocarcinoma in 39 (98 %)and adenosquamous carcinoma in one; 6 (15 %) out of 40were poorly differentiated carcinomas. Twenty-three (58 %)patients had metastases to lymph nodes. Metastases up to N2lymph nodes were seen in 12 (30 %) patients with fourpatients who had skip metastasis to the N2 lymph nodes. Atotal of 829 lymph nodes were harvested. The median lymphnode count was 18 (4–77). The final stage of the GBC is givenin Table 4. One patient who underwent excision of infiltratedbile duct had positive distal common bile duct margin on finalhistopathology; the rest of the patients had negative tumormargins. Twenty-eight of 40 (70%) had disease limited till N1nodes. R0 resection was achieved in 33 (83 %) patients(including six patients who underwent median sectorectomy,ERH, or HPD). Postoperative chemoradiotherapy was givento all patients except with those with a T1 lesion withoutlymph node metastasis. Radiation (45–50.4 Gy) was givenin 4.5 to 5.5 weeks along with concurrent chemotherapy with5-fluorouracil at a dose of 370 mg/m2 from day1 to 3 and day29 to 31 of radiotherapy. Post-radiotherapy, adjuvant 5-fluorouracil was given at a dose of 370 mg/m2 on day1–5four weekly for six cycles.

Follow-up information is available for all patients till thepreparation of this manuscript. Five (13 %) patients hadrecurrence and one died of recurrence in the liver 8 monthsafter a hepatopancreatoduodenectomy. Other patients are onsecond-line chemoradiation; one among them underwent aresection for an abdominal wall recurrence. Two had lymphnode recurrence and another developed liver metastasis.With a median follow-up of 15 (4–33) months after theoperation, the mean survival is 31 months.

Discussion

Management of patients with GBC is a challenge because offrequent adjacent organ involvement, low resectability rate,and lack of effective chemotherapy resulting in poor surviv-al. The reported median survival of patients who could notundergo resection for GBC is less than 4 months [6]. Sur-gery for GBC is associated with a reported overall 5-yearsurvival rate of 26 % and 5-year survival rate of 60 % aftercurative resection [6], one of the best reported in the litera-ture. The median survival was 34 months for the curativetreatment group versus 3 months for the palliative treatmentgroup in another series [7]. A population-based analysisincluding 2,955 patients with GBC showed an overall re-sectability rate of only 13 % [8]. A multicenter study fromthree institutions from different countries showed an R0resectability rate of 61 % in patients submitted for operation[9]. The major reasons for unresectability are distant metas-tasis and loco-regionally advanced nature of the disease.Distant metastasis is a contraindication for any resectionaltherapy, but the extent of loco-regional disease which con-traindicates resection varies according to the protocol fol-lowed in various centers. In centers which follow anaggressive resectional strategy, local infiltration to adjacentorgans such as bile duct, colon, and duodenum is not con-sidered as a contraindication for resection. Various reportshave shown that resection status, especially resection withnegative margins, was positively associated with increasedsurvival [6, 8, 10, 11].

In T1a (lamina propria) GBC, simple cholecystectomyalone is an adequate treatment. The optimal operation forT1b (muscle) GBC is controversial. We have advocated ECfor T1b disease as some patients with T1b have local recur-rence after simple cholecystectomy alone [12]. Moreover, itis difficult to differentiate between T1 and T2 lesions onimaging or at operation, and hence, it is prudent to considerany GBC as beyond T1 till the histopathology report isavailable. A reoperation can be avoided only in thosepatients who report after a simple cholecystectomy for aT1aN0 incidental GBC and this finding is to be confirmedby an expert pathologist. Reoperation is required for tumorsinvolving the muscle (T1b) and beyond (T2 and above)because of the associated lymph node involvement andlymphatic and peri-neural invasion. EC (ref para 2 of“Introduction” section) is considered as an option ofoncological resection in GBC. A Japanese multicenter studyinvolving 485 patients with pT2 and pT3 who underwent R0resection concluded that resection of the GB bed can beemployed only in patients with disease not invading the liveror hepatoduodenal ligament [13]. But majority of the patientsare above this stage at diagnosis. Only seven (18 %) patientsin our series belong to T2N0M0 or lesser stage. Variouscenters follow different strategies in deciding the extent of

Table 4 Final pathological staging of gall bladder cancer after extend-ed resections (n040): as per AJCC seventh [5] edition

Stage group pTNM N (%)

0 Tis N0 M0 0 (0)

I T1 N0 M0 2 (5)

II T2 N0 M0 5 (12.5)

IIIA T3 N0 M0 10 (25)

IIIB T1–3 N1 M0 9 (22.5)

IVA T4 N0–1 M0 1 (2.5)

IVB Any TN2 M0 13 (32.5)Any T any N M1

Tis carcinoma in situ, T1 tumor invades lamina propria or muscularlayer, T2 tumor invades perimuscular connective tissue; no extensionbeyond serosa or into the liver, T3 tumor perforates serosa and/ordirectly invades the liver and/or one other adjacent organ, T4 tumorinvades main portal vein or hepatic artery or invades two or moreextrahepatic organs, N0 no regional lymph node metastasis, N1 metas-tases to nodes along cystic duct, common bile duct, hepatic artery, and/or portal vein, N2 metastases to periaortic, pericaval, superior mesen-teric artery, and/or celiac artery lymph nodes

J Gastrointest Canc (2013) 44:33–40 37

liver resection. Japanese and western literature generally differin the extent of resection in stage-matched GBC. Data fromGerman registry as well as other studies showed a survivalbenefit for segmentectomy 4B + 5 compared to a 3-cm liverwedge resection in patients with T2 tumors [11, 14–16]. Inpatients with nodal disease, the para-cystic duct lymphaticpathways get obliterated at an early stage and the lymph flowsthrough alternate pathways through central inferior subseg-ment (segments 4B + 5) [17]. In this series, we adopted astrategy of central inferior hepatic sub-segmentectomy as theminimum liver resection even in early stages; with this ap-proach, the liver margins turned out to be negative in allpatients. Moreover, the low mortality and acceptable morbid-ity related to central inferior hepatic sub-segmentectomy inour patients justify this approach.

Survival benefit has been reported by many authors inpatients with T2/T3 and stage III/IV GBC who underwentextended lymphadenectomy [16, 18–20], although the ben-efit is not evident in patients with positive para-aortic lymphnodes on final histopathology [21, 22]. Although para-aorticlymph node involvement has been considered as equivalentto a distant metastasis in terms of survival, thus recommend-ing only sampling by some [22], others reported bettersurvival after extended lymphadenectomy (including thepara-aortic region) in patients with positive para-aorticlymph nodes compared to those with distant metastasis[20, 21, 23, 24]. The reported incidence of lymph nodepositivity is 42–73 % and that of para-aortic nodes is 22–38 % [22, 25]. This corresponds to our experience. As theinter-aorto-caval lymph nodes have been shown to be thelast regional basin in GBC by lymphatic studies [26], weopted to remove all inter-aorto-caval lymph nodes both forthe purpose of staging and prognosis as part of our strategyof the management of GBC. The morbidity related to ex-tended lymphadenectomy (N1 + N2 nodes) was acceptablein our series. Moreover, with this approach, we could cor-rectly stage all the patients. A 10 % incidence of skip lymphnodal metastasis in N2 lymph nodes is an additional argu-ment in adopting extended lymphadenectomy.

We did not resect extrahepatic bile duct for better lym-phatic clearance and advocate it only when there is directinfiltration of the common bile duct by the tumor. A similarphilosophy is being followed even in centers where an ag-gressive resectional strategy is being followed [19, 27, 28]. Afrozen section of cystic duct margin in all non-fundal tumorsensured R0 resection in our series. PD was added by us inselected patients. There are conflicting reports about addingPD for better lymphatic clearance citing very low survivalrates after HPD, although some authors recommend this inpatients without positive para-aortic lymph nodes [19, 20,25, 29, 30]. Reports of vascular resection and reconstructionin advanced GBC do not recommend the procedure citingreasons of poor survival benefit unlike cholangiocarcinoma

[20, 31]. We also consider MPV and CHA involvement inGBC as a contraindication for resection.

Triple-phase CT with 3D reconstruction is the gold stan-dard imaging to stage the disease and plan surgical strategy.This specifically helped us in three areas—delineating thevascular structures in the proposed line of liver resection, indetecting accessory arteries thus avoiding reconstructionafter resection of lobar/sectoral HA, and detection of arterialaberrations preoperatively, thereby providing a clear road-map of the arteries in the HDL during lymphadenectomy.The presence of arterial aberrations in 35 % of our patientsclearly underlines the importance of a preoperative arterio-gram. All the branches and subdivisions of the celiac axisand superior mesenteric artery need to be delineated beforeplanning extended resections in GBC. CT correctly delin-eated the extent of loco-regional spread in all patients,except two. The imaging–operation interval was more than1 month in both these patients which might be the reasonwhy we encountered further progression of the local diseaseat laparotomy, not to forget the speedy spread of GBC. ButCT failed to detect peritoneal disease in 14 (24 %) and livermetastasis in five (8 %) patients which were later revealedby laparoscopy. These could have been detected by PETscan, which we could not use due to its non-availability atour center. We have not used intra-operative ultrasound toassess the extent of liver involvement with the tumor as thiswas correctly assessed in CT. We have also been advocatinguse of upper GI endoscopy to look for duodenal involve-ment in patients with symptoms of gastric outlet obstructionand/or suspicion of duodenal involvement on CT, but it wasnot done in this group of patients as duodenal involvementwas not considered as a contraindication for resection. Wehave earlier reported that an on-table staging laparoscopy ismandatory in all patients with GBC being taken up forresection. We strongly recommend it as this obviated theneed of a laparotomy in 24 % of patients planned forresection earlier [32]. We do not use laparoscopy to assessthe lymph nodal disease as we found CT more useful in thisregard. CT combined with laparoscopy could correctly de-termine the stage in 97 % of patients with GBC in our series.In both the patients who underwent PVE, we could notproceed with resection because of the metastases seen atlaparoscopy which were not evident on the pre-embolizationstaging laparoscopy. Post-PVE preoperative repeat staginglaparoscopy may be recommended. The three patients whounderwent ERH did not undergo preoperative PVE.

Although we had a 60 % complication rate, none of ourpatients required reoperation; most of the patients recoveredwell and were discharged. Our mortality rate was 5 %. Withthis aggressive surgical strategy, we could achieve a resect-ability rate of 71% in operated patients and an R0 rate of 83%in resected patients. With this strategy of performing extendedresections in all patients with GBC, 26 patients with no liver

38 J Gastrointest Canc (2013) 44:33–40

infiltration (which could have been managed with GB bed2 cm liver wedge resection as in EC) underwent major hepaticresection; 17 patients with no lymph node involvement (whichcould have been managed with standard lymphadenectomy asin EC) underwent extended lymphadenectomy. If preopera-tive imaging and/or intra-operative evaluation can accuratelydetect liver infiltration, patients can be stratified for either GBbed 2 cm liver wedge resection (as in EC) or for no liverinfiltration (T1 and T2) or for major hepatic resection (as incentral inferior sub-segmentectomy) for liver infiltration (T3).During operation, sampling and frozen section of N1 and N2lymph nodes can further stratify them for standard lymphade-nectomy (as in EC) for node-negative (N0) or for extendedlymphadenectomy for node-positive (N1) patients.

Stages I and II represent only minority of GBC patients andmajority of patients belong to stages III and IV.0 surgicalstrategy of EC may benefit patients with early (I and II) stageGBC only and will deny a larger proportion of patients (stagesIII and IV) the chance to achieve R0 resection. An aggressivesurgical strategy can achieve R0 resection in patients withadvanced GBC and improve survival. The morbidity andmortality associated with extended resections in GBC areacceptable. Our experience shows the feasibility of performingextended resections in patients with GBC; whether it is justi-fied and useful will be proven later when medium- and long-term survival results available. If patients with GBC can bestratified, either preoperatively or intra-operatively, intothose with early (I and II) stage and those with advanced(III and IV) stage, then they can be treated with either EConly or extended resections, respectively. If this is notpossible, a surgical strategy of extended resections for allpatients is the only way to achieve R0 resection, long-termsurvival, and possible cure in GBC.

Conflict of Interest The authors state that there is no conflict ofinterest.

References

1. Zhu AX, Hong TS, Hezel AF, Kooby DA. Current management ofgallbladder carcinoma. Oncologist. 2010;15(2):168–81. Epub2010 Feb 10. Review.

2. Pilgrim C, Usatoff V, Evans PM. A review of the surgical strate-gies for the management of gallbladder carcinoma based on T stageand growth type of the tumour. Eur J Surg Oncol. 2009;35(9):903–7. Epub 2009 Mar 4.

3. Balachandran P, Agarwal S, Krishnani N, Pandey CM, Kumar A,Sikora SS, Saxena R, Kapoor VK. Predictors of long-term survivalin patients with gallbladder cancer. J Gastrointest Surg. 2006;10(6):848–54.

4. Ito M, Mishima Y, Sato T. An anatomical study of the lymphaticdrainage of the gallbladder. Surg Radiol Anat. 1991;13(2):89–104.

5. American Joint Committee on Cancer. Gallbladder. In: AJCC cancerstaging manual. 7th ed. New York: Springer; 2010 211–214.

6. Liang JW, Dong SX, Zhou ZX, Tian YT, Zhao DB, Wang CF,Zhao P. Surgical management for carcinoma of the gallbladder: asingle-institution experience in 25 years. Chin Med J (Engl).2008;121(19):1900–5.

7. Chan SY, Poon RT, Lo CM, Ng KK, Fan ST. Management ofcarcinoma of the gallbladder: a single-institution experience in16 years. J Surg Oncol. 2008;97(2):156–64.

8. Mayo SC, Shore AD, Nathan H, Edil B, Wolfgang CL, Hirose K,Herman J, Schulick RD, Choti MA, Pawlik TM. National trends inthe management and survival of surgically managed gallbladderadenocarcinoma over 15 years: a population-based analysis. JGastrointest Surg. 2010;14(10):1578–91.

9. Butte JM, Matsuo K, Gönen M, D'Angelica MI, Waugh E, AllenPJ, Fong Y, DeMatteo RP, Blumgart L, Endo I, De La Fuente H,Jarnagin WR. Gallbladder cancer: differences in presentation, sur-gical treatment, and survival in patients treated at centers in threecountries. J Am Coll Surg. 2011;212(1):50–61.

10. Chakravarty KD, Yeh CN, Jan YY, Chen MF. Factors influencinglong-term survival in patients with T3 gallbladder adenocarcino-ma. Digestion. 2009;79(3):151–7. Epub 2009 Mar 30.

11. Kai M, Chijiiwa K, Ohuchida J, Nagano M, Hiyoshi M, Kondo K.A curative resection improves the postoperative survival rate evenin patients with advanced gallbladder carcinoma. J GastrointestSurg. 2007;11(8):1025–32.

12. Kapoor VK. Incidental gall bladder cancer. Am J Gastroenterol.2001;96(3):627–9.

13. Araida T, Higuchi R, Hamano M, Kodera Y, Takeshita N, Ota T,Yoshikawa T, Yamamoto M, Takasaki K. Hepatic resection in 485R0 pT2 and pT3 cases of advanced carcinoma of the gallbladder:results of a Japanese Society of Biliary Surgery survey—a multi-center study. J Hepatobiliary Pancreat Surg. 2009;16(2):204–15.Epub 2009 Feb 14.

14. Goetze TO, Paolucci V. Adequate extent in radical re-resection ofincidental gallbladder carcinoma: analysis of the German Registry.Surg Endosc. 2010;24(9):2156–64. Epub 2010 Feb 23.

15. Morine Y, Shimada M, Imura S, Fujii M, Ikemoto T, Soejima Y,Utsunomiya T, Kurita N, Miyake H, Tashiro S. Surgical strategyfor advanced gallbladder carcinoma according to invasive depth ofthe tumor. Hepatogastroenterology. 2008;55(88):1965–70.

16. Kohya N, Miyazaki K. Hepatectomy of segment 4a and 5 com-bined with extra-hepatic bile duct resection for T2 and T3 gall-bladder carcinoma. J Surg Oncol. 2008;97(6):498–502.

17. Terazawa T, Miyake H, Kurahashi M, Tashiro S. Direct lymphaticspreading route into the liver from the gallbladder: an animalexperiment using pig. J Med Invest. 2004;51(3–4):210–7.

18. Wang JD, Liu YB, Quan ZW, Li SG, Wang XF, Shen J. Role ofregional lymphadenectomy in different stage of gallbladder carci-noma. Hepatogastroenterology. 2009;56(91–92):593–6.

19. Kokudo N, Makuuchi M, Natori T, Sakamoto Y, Yamamoto J, SekiM, Noie T, Sugawara Y, Imamura H, Asahara S, Ikari T. Strategiesfor surgical treatment of gallbladder carcinoma based on informa-tion available before resection. Arch Surg. 2003;138(7):741–50.discussion 750.

20. Kondo S, Nimura Y, Hayakawa N, Kamiya J, Nagino M, UesakaK. Extensive surgery for carcinoma of the gallbladder. Br J Surg.2002;89(2):179–84.

21. Shimada H, Endo I, Fujii Y, Kamiya N, Masunari H, Kunihiro O,Tanaka K, Misuta K, Togo S. Appraisal of surgical resection ofgallbladder cancer with special reference to lymph node dissection.Langenbecks Arch Surg. 2000;385(8):509–14.

22. Kondo S, Nimura Y, Hayakawa N, Kamiya J, Nagino M, UesakaK. Regional and para-aortic lymphadenectomy in radical surgeryfor advanced gallbladder carcinoma. Br J Surg. 2000;87(4):418–22.

23. Nishio H, Nagino M, Ebata T, Yokoyama Y, Igami T, Nimura Y.Aggressive surgery for stage IV gallbladder carcinoma; what are

J Gastrointest Canc (2013) 44:33–40 39

the contraindications? J Hepatobiliary Pancreat Surg. 2007;14(4):351–7. Epub 2007 Jul 30.

24. Chijiiwa K, Kai M, Nagano M, Hiyoshi M, Ohuchida J, Kondo K.Outcome of radical surgery for stage IV gallbladder carcinoma. JHepatobiliary Pancreat Surg. 2007;14(4):345–50. Epub 2007 Jul30.

25. Sasaki R, Itabashi H, Fujita T, Takeda Y, Hoshikawa K, TakahashiM, Funato O, Nitta H, Kanno S, Saito K. Significance of extensivesurgery including resection of the pancreas head for the treatmentof gallbladder cancer—from the perspective of mode of lymphnode involvement and surgical outcome. World J Surg. 2006;30(1):36–42.

26. Shirai Y, Yoshida K, Tsukada K, Ohtani T, Muto T. Identificationof the regional lymphatic system of the gallbladder by vital stain-ing. Br J Surg. 1992;79(7):659–62.

27. Araida T, Higuchi R, Hamano M, Kodera Y, Takeshita N, Ota T,Yoshikawa T, Yamamoto M, Takasaki K. Should the extrahepaticbile duct be resected or preserved in R0 radical surgery for ad-vanced gallbladder carcinoma? Results of a Japanese Society ofBiliary Surgery Survey: a multicenter study. Surg Today. 2009;39(9):770–9. Epub 2009 Sep 24.

28. Yagi H, ShimazuM, Kawachi S, TanabeM, Aiura K,Wakabayashi G,UedaM, Nakamura Y, KitajimaM. Retrospective analysis of outcomein 63 gallbladder carcinoma patients after radical resection. J Hepato-biliary Pancreat Surg. 2006;13(6):530–6. Epub 2006 Nov 30.

29. Araida T, Yoshikawa T, Azuma T, Ota T, Takasaki K, Hanyu F.Indications for pancreatoduodenectomy in patients undergoinglymphadenectomy for advanced gallbladder carcinoma. J Hepato-biliary Pancreat Surg. 2004;11(1):45–9.

30. Sasaki R, Takahashi M, Funato O, Nitta H, Murakami M, KawamuraH, Suto T, Kanno S, Saito K. Hepatopancreatoduodenectomy withwide lymph node dissection for locally advanced carcinoma of thegallbladder—long-term results. Hepatogastroenterology. 2002;49(46):912–5.

31. Shimada H, Endo I, Sugita M, Masunari H, Fujii Y, Tanaka K,Misuta K, Sekido H, Togo S. Hepatic resection combined withportal vein or hepatic artery reconstruction for advanced carcinomaof the hilar bile duct and gallbladder. World J Surg. 2003;27(10):1137–42. Epub 2003 Aug 21.

32. Agrawal S, Sonawane RN, Behari A, Kumar A, Sikora SS, SaxenaR, Kapoor VK. Laparoscopic staging in gallbladder cancer. DigSurg. 2005;22(6):440–5. Epub 2006 Feb 10.

40 J Gastrointest Canc (2013) 44:33–40