- Email: [email protected]
Predictors of curative resection and long term survival of gallbladder cancer – A retrospective analysis
Accepted Manuscript Predictors of curative resection and long term survival of gallbladder cancer – A retrospective analysis Pramod Kumar Mishra, Sundeep Singh Saluja, Nabi Prithiviraj, Vaibhav Varshney, Neeraj Goel, Nilesh Patil PII:
S0002-9610(16)30863-7
DOI:
10.1016/j.amjsurg.2017.02.006
Reference:
AJS 12292
To appear in:
The American Journal of Surgery
Received Date: 14 November 2016 Revised Date:
2 February 2017
Accepted Date: 5 February 2017
Please cite this article as: Mishra PK, Saluja SS, Prithiviraj N, Varshney V, Goel N, Patil N, Predictors of curative resection and long term survival of gallbladder cancer – A retrospective analysis, The American Journal of Surgery (2017), doi: 10.1016/j.amjsurg.2017.02.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Predictors of curative resection and long term survival of gallbladder cancer – a
retrospective analysis. Name and initials of authors:
Saluja SS Prithiviraj N
Vaibhav Varshney MCh
Varshney V
Neeraj Goel MCh
Goel N
Nilesh Patil MCh
M AN U
Prithiviraj Nabi MS
SC
Sundeep Singh Saluja MCh
Mishra PK
RI PT
Pramod Kumar Mishra MS, PhD
Patil N
Postal address and email address:
Department of Gastrointestinal Surgery, Room No 218, 2nd floor, Academic Block GB Pant
Corresponding author Sundeep Singh Saluja
TE D
Hospital, New Delhi, India
Professor and Corresponding author,
EP
Department of Gastrointestinal Surgery
G B Pant Institute of Postgraduate Medical Education & Research 1, Jawaharlal Nehru Marg,
AC C
New Delhi 110002.
Email: [email protected] phone no 919718599259 Source of funding: No funding received Category: Original article
Key words: Carcinoma gallbladder, Predictors of resectability, Radical cholecystectomy, resectability in gallbladder cancer, Surgical Obstructive jaundice
ACCEPTED MANUSCRIPT
Background: Gallbladder cancer (GBC) is an aggressive malignancy. We analysed factors predicting resectability and survival of patients with GBC and the impact of surgical
RI PT
obstructive jaundice (SOJ).
Methods: Four hundred and thirty-seven patients with suspected GBC were analysed (52
SC
excluded: benign pathology n=35, missed GBC n=17). The remaining 385 patients were divided into non-SOJ (n=234) and SOJ (n=151) groups. Predictors of resectability and long
M AN U
term survival were analysed and compared with their subgroups..
Results: Patients with gastric outlet obstruction, abdominal lump, weight loss and SOJ were more likely to be unresectable (p:0.04,0.024,0.003 and 0.003, respectively). TNM stage, node positivity and adjacent organ involvement were predictors of survival (p<0.001,0.008 and<0.001). Metastatic (36.7%vs47.7%),
inoperable (1.7%vs12.6%) and unresectable
TE D
disease (9.8%vs24.5%) were more in the SOJ group and had lower curative resection rates (51.7%vs15.2%;p<0.0001). The 1,2 and 5-year survival rates were higher in patients in the
EP
non-SOJ than SOJ group (79.6%,65% and 52.9% vs 48.6,32.4%and 0%;p<0.001).
Conclusion: GBC with SOJ is more likely to be unresectable. SOJ, nodal involvement,
AC C
adjacent organ infiltration and higher TNM stage predict poor survival.
ACCEPTED MANUSCRIPT
Introduction Gallbladder cancer (GBC) is an aggressive disease with a poor prognosis.1 It usually presents late and often at an unresectable stage. Locally advanced GBC presents with a varied
RI PT
symptomatology depending on the extent of the adjacent organ or vascular involvement. About 25-50% of the patients of GBC present with jaundice.1-3 While few authors have reported long term survival in patients of GBC with jaundice4,5,6others consider it an ominous sign1,3,7.
SC
Resectability in these patients may vary from 7% to 58%.1-8Complete surgical removal offers the best chance of cure at present. Surgeons are under constant pressure to increase meaningful
M AN U
survival and possibly, deliver a cure. This has led to pushing the boundaries of resection at the cost of increased morbidity and mortality. Given the aggressive nature of the disease and the mortality/morbidity associated with extensive resections, the subgroup of patients who do not benefit from these resections should also be regularlyanalysed.1,7,9Various factors like stage of
TE D
disease, presence of jaundice, lymph node or visceral involvement have been analysed to determine which patients may actually benefit from these resections.1-7Identification of the factors that may predict unresectability can be used to avoid surgery in inoperable patients and
EP
the associated morbidity. We analysed our data to understand the predictors of curative resection, morbidity and mortality in patients undergoing resection, factors predicting long term survival in
AC C
GBC and the impact of jaundice.
ACCEPTED MANUSCRIPT
Patients and Methods The records of all patients with suspected GBC admitted in our unit from September 2003 till
RI PT
December 2014 were reviewed from theprospectively maintained database. Inclusion criteria
SC
Patients with a diagnosis of GBC deemed resectable on initial work up Exclusion criteria
M AN U
1.GBC masquarades who underwent radical resection for suspected GBC but had benign disease on histopathology.
2a) Missed GBC: defined as patients in whom the diagnosis of GBC was evident on preoperative imaging but who underwent only cholecystectomy elsewhere
TE D
2b) patients with incidental GBC that presented later with advanced disease. The patients were divided into two groups
EP
Non-surgical obstructive jaundice (SOJ) group: GBC with no SOJ. Patients with jaundice due to
AC C
CBD stones were included in the non-SOJ group. SOJ group: These were defined as patients with GBC presenting withSOJsecondary to direct infiltration or lymph node compression of the common bile duct. Preoperative work up
All patients underwent detailed clinical evaluation and blood investigations followed by imaging. Imaging modalities included ultrasound abdomen with Doppler, dual phase contrast enhanced
ACCEPTED MANUSCRIPT
CT scan (CECT).Magnetic resonance cholangio-pancreaticography (MRCP) was done in patients with SOJ. Upper gastrointestinal endoscopy was done routinely in patients suspected to have locally advanced tumour to assess gastro-duodenal infiltration. Endoscopic ultrasound was
RI PT
used in the latter part of the study to assess extent of disease and sampling of nodes in the interaortocaval region. Blood investigations including liver function tests and chest X-ray were done
SC
in all patients. Peri-operative strategy
M AN U
Patients deemed resectable at the time of evaluation in the outpatient were admitted for further detailed evaluation. Patients with obvious para-aortic nodes, main or left portal vein invasion or metastatic disease were considered unresectable. Patients with SOJ underwent percutaneous transhepatic biliary drainage/endoscopic retrograde cholangiography (ERCP) stenting to reduce
Surgical procedure
TE D
serum bilirubin to <5mg/dL. AJCC seventh edition was used for staging.10
Staging laparoscopy was done in all patients to rule out metastatic disease. In the presence of
EP
metastasis, a biopsy confirmation was done and the procedure abandoned. Inter-aortocaval
AC C
lymph node was sampled and a frozen section examination done before proceeding to assess local resectability. Palliative biliary bypass/gastrojejunostomy was done when required in patients found to have unresectable disease during surgery. In patients undergoing curative resection, R0 resection was achieved with resection of a 2 cm wedge of liver/segment IVB, V resection along with the gallbladder mass. In case of adjacent organ involvement either sleeve or segmental resection was done to obtain an R0 resection. Standard lymphadenectomy included clearance of lymph nodes along the hepatoduodenal ligament, and the pericholedochal,
ACCEPTED MANUSCRIPT
periportal, peripancreatic and right celiac nodes. Extended right hepatectomy was done in those in whom the hilum was involved with infiltration of the pedicle on the right side. The cystic duct margin was routinely sent for frozen section and the common bile duct was resected in patients
if required. Definitions Curative resection: Patients who underwent R0 resection.
SC
RI PT
with obvious infiltration by tumour, a positive cystic duct margin and for lymph nodal clearance,
M AN U
Inoperable disease: Patients with inadequate liver remnant or poor performance status to undergo major surgery
Locally advanced unresectable patients: Those with infiltration of the main portal vein, hepatic
TE D
artery or encasement of a major vascular pedicle precluding R0 resection. Metastatic: Any patient with peritoneal, omental, ovarian or liver metastases. Presence of positive inter-aortocaval (IAC) and lymph node on the left side of the coeliac axis were also
EP
considered as metastatic disease.
The predictors of curative resection and long term survival were analysed in the entire cohort.
AC C
Subgroup analysis among patients with or without SOJ was also done. The survival among patients in the SOJ group was compared with those with locally advanced(T3-4) non-SOJ disease to assess the impact of jaundice Factors predicting curative resection: Preoperative parameters such as presence of pain abdomen at time of presentation, loss of weight, loss of appetite, features of gastric outlet
ACCEPTED MANUSCRIPT
obstruction, jaundice, palpable abdominal lump, hepatomegaly, level of bilirubin were analysed by multivariable logistic regression analysis for their ability to predict curative resection.
RI PT
Morbidity and mortality Operative mortality included all patients who died within 30 days of the procedure or during the same hospital admission. Morbidity included all postoperative complications occurring within 30
SC
days of procedure or during the same hospital admission. Post-hepatectomy liver failure was defined according to definitions by ISGLS.11 Morbidity was classified according to the Clavien-
M AN U
Dindo classification.12 Factors predicting Survival
a) TNM stage was evaluated as per AJCC 7th edition b)Nodal involvement was defined
Follow up
TE D
aspresence of histologically positive lymph node c)Adjacent organ involvement
Patients were followed up every 3 months for the first year then every 6 months. They were
EP
assessed clinically, using laboratory tests including liver functions tests and abdominal ultrasound. A CECT scan of the abdomen was done at one year in patients who remained well. It
AC C
was done earlier if necessitated by clinical or ultrasound findings. All patients were treated by multimodality approach with gemcitabine-based chemotherapy in both the adjuvant and palliative setting. According to our protocol adjuvant chemotherapy was given to all patients with stage T2 disease and above, and for all lymph node positive tumours irrespective of the T stage. Best possible supportive care was given in the palliative setting.
ACCEPTED MANUSCRIPT
Statistical analysis Statistical analysis was done using SPSS software (IBM SPSS Statistics for Windows, Version
RI PT
22.0. Armonk, NY: IBM Corp). Categorical variables were compared with the χ2 test or Fisher’s exact test. Wherever appropriate, multivariate analysis was done using the logistic regression model to determine independent predictors of outcome. Long-term survival was calculated using
SC
the Kaplan-Meier method, and differences in groups of survivor were compared by the log-rank test. Multivariate analysis was done using the Cox regression model to determine independent
AC C
EP
TE D
M AN U
predictors of survival. A P value less than 0.05 was considered to indicate statistical significance.
ACCEPTED MANUSCRIPT
Results A total of 437 patients with suspected GBC were admitted after preliminary evaluation in the
RI PT
outpatient department from October 2003 to December 2014. Fifty-two patients were excluded either because of benign pathology on postoperative histology (n=35) or a missed GBC (n=17). Among 385 patients included in the final study, 234 patients (61%) had GBC with no SOJ (Non-
SC
SOJgroup ) while151 patients (43%) had GBC with SOJ(SOJ group).The patient profile is shown in Figure 1.
M AN U
Abdominal pain was the most common presenting symptom(89%) followed by weight loss (63%) and loss of appetite (60%) (Table 1). Thirty-one (8%) patients presented with gastric outlet obstruction. The incidence of anorexia, weight loss and adjacent colon/duodenal infiltration were significantly more in SOJ group (p<0.001, 0.001, 0.001 and 0.005,
TE D
respectively). The incidence of metastasis, inoperable disease and unresectable locally advanced disease was significantly more in the SOJ group while curative resection was more in the nonSOJ group (Table 2).Although the incidence of metastasis in lymph nodes and liver was higher
EP
in the SOJ group, the difference was not statistically significant (P=0.16). The major reason for inoperability in the SOJ group was deterioration of general condition during the waiting period
AC C
for fall of serum bilirubin following biliary drainage, while co-existing major medical illness not permitting surgery in patients with non-SOJ group.
Predictors of curative resection Overall, 154 patients underwent curative resection––curative resections were done significantly more in non-SOJ as compared to SOJ groups (121/234 vs 23/151; p<0.001). Factors predicting curative resection(Table 3)
ACCEPTED MANUSCRIPT
There was no difference in resectability rate among males compared to females and patients with <50 years compared to those >50 years of age. On univariate analysis, factors with negative prediction for resectability were presence of pain abdomen, anorexia, significant weight loss
RI PT
(>10% of body weight), gastric outlet obstruction, hepatomegaly, abdominal lump, presence of jaundice and bilirubin >1mg/dl. Age and history of cholangitis had no influence on resectability. On multivariate analysis preoperative gastric outlet obstruction, abdominal lump, weight loss and
SC
jaundice on presentation were significant negative predictors of resectability (p value: 0.04, 0.024, 0.003 and 0.003, respectively). The preoperative bilirubin did not correlate with
M AN U
resectability (area under curve of 0.597 in ROC). (Fig2). Direct infiltration into the hepatoduodenal ligament was present in 116 patients in our series with a patent biliary confluence in 70 patients and blocked in 46 patients. Porta hepatis lymph nodes compressing the mid/lower CBD as a cause of jaundice was seen in 32 patients. Resectability rates did not differ
TE D
with cause of SOJ (P=0.29).
On subgroup analysis of the non-SOJ and SOJ groups with respect to resectability, increasing age group had less resectable disease in the SOJ group compared with the non-SOJ
EP
group (P=0.004). Preoperative gastric outlet obstruction, abdominal lump as presentation and
AC C
weight loss were significant predictors of inoperability in non-SOJ group but not in the SOJ group (Table 3).
Surgical Procedure
Operative resection included radical cholecystectomy with liver wedge/IVB,V resection and standard lymphadenectomy alone in 66 cases, adjacent organ/CBD resection in 34 patients, CBD exploration in 5 patients and open cholecystectomy in one. Completion radical cholecystectomy
ACCEPTED MANUSCRIPT
for incidental GBC was done in 25 patients. Trisegmentectomy was done in 8 cases and hepatopancreato-duodenectomy was done in 5 cases. (Table 2).
RI PT
Morbidity, mortality and hospital stay The morbidity grading based on Clavien-Dindo classification of surgical complications in both groups are as given in table 4b. There was significantly less morbidity in the non-SOJ group
SC
compared to the SOJ group [19% (23/121)vs56 % (12/23) P<0.001] (Table 4a). The SOJ group had significant increase in major complications (Grades III and IV) in comparison with the non-
M AN U
SOJ group but the incidence of minor complications was similar between the two groups. The overall mortality of the entire study population was 6.9%. The 30-day mortality was significantly less in the non-SOJ compared to the SOJ group(4% vs 21%; p=0.002).Of the five perioperative deaths in the SOJ group, 3 were directly related to surgical complications, while in the other two
TE D
they were due to massive stroke and cardiac failure in the postoperative period. The median length of hospital stay was significantly shorter in patients in the non-SOJ compared to the SOJ
Survival
AC C
Non-SOJ group
EP
group [7(3-25) days vs 9.5 (6-28) p<0.001].
Of the 234 patients, 121 underwent curative surgery. The 30-day mortality was 5/121 (4%).The median survival of patients without SOJ was 61months, while the 1-, 2- and 5-year survival rate was 79.6%, 65%, and 52.9%, respectively (Figure3A). When further subgroup analysis of patients with locally advanced GBC (pathological T3 and T4 disease) was done, the median survival decreased to 19 months. Among the 60 patients with pathological T3 and T4 disease,
ACCEPTED MANUSCRIPT
there were two perioperative deaths (4%).The 1-, 2- and 5-year survival rates were 59.4%. 36.7% and 23.6% respectively (Figure3B).
RI PT
SOJ group Among 154 patients, 23 patients of GBC with SOJ underwent curative resection. The 30-day mortality was 5/23 (21%). All patients in this group were followed till death. The median
SC
survival was 12 months, while the 1- and 2-year survival rates were 48.6% and 32.4%, respectively (Figure 3B). However, there were no five-year survivors in this group.
M AN U
Comparison of survival in locally advanced (T3,T4) non-SOJ versus SOJ patients Patients with SOJ have at least T3 disease considering the CBD infiltration and was compared with T3, T4 disease in non-SOJ group to analyse the impact of jaundice. Survival curves of these two groups by log-rank test did not reveal any statistical significance despite a difference in the
TE D
median survival of these two groups (P value=0.44) (Figure 3B). Predictors of long-term survival
EP
Long-term survival was dependent on the TNM stage of the disease, the N status and involvement of adjacent organ. Survival of the entire cohort had significant difference across all
AC C
stages by log-rank test (P value<0.001) (Figure 4A).Stage-wise survival in the non-SOJ and SOJ groups was similar with declining median survival with increasing stages (Figure 4B ). Stage wise comparison between both groups showed no statistical difference in median survival (Stage IIIap=0.59 Stage IIIb p=0.53 and stage IV p=0.27). Nodal positivity had poor survival compared with node negative disease. Median survival in node negative disease was 62 months as against 14 months in node positive disease. (p=0.008 by
ACCEPTED MANUSCRIPT
log rank test)(Figure 4C). Median survival in patients with adjacent organ involvement was 14 months compared to 77 months in patients without any adjacent organ involvement with P value
RI PT
being statistically significant at P<0.001. (Figure 4D) Discussion
GBC is a biologically aggressive disease with late presentation and poor prognosis. Despite
SC
screening from the outpatient department 41% patients of the entire cohort were found to have metastatic disease at surgery. It is important that in the SOJ group only one of seven patients
M AN U
underwent curative resection while in the non-SOJ group one in every two patients underwent curative resection. Therefore, it is important to assess the predictors of resectability particularly in patients with SOJ. Resectability
TE D
This aspect of GBC has been seldom addressed in the literature. Varma et al.13found presence of abdominal lump and jaundice as signs of advanced disease but not of unresectability. On the contrary we found preoperative gastric outlet obstruction, abdominal lump as a presenting
EP
symptom, weight loss and jaundice on presentation as negative predictors of resectability on multivariate analysis. However these predictors lose their significance in SOJ group implying
AC C
presence of jaundice itself as the most important determinant. Soek et al.14from UK reported raised bilirubin and alkaline phosphatase along with raised total leukocyte & platelet count as predictors for unresectable disease. This suggests tacitly that jaundice is an important determinant of unresectability similar to our series. The utility of these predictors is little as they per se do not contraindicate surgery. However, careful preoperative workup in these subgroups
ACCEPTED MANUSCRIPT
may reduce unwarranted laparotomies and may help in planning neoadjuvant treatment in these subgroups.
RI PT
Gastric outlet obstruction may be a surrogate marker of advanced disease but duodenal infiltration per se is not a contraindication for resection. Distal gastrectomy or duodenal sleeve resection was done in 14 patients in our series.
SC
Curative resection
One hundred and forty four (37%)patients underwent curative resection in the present series.
M AN U
Resection rate significantly decreased in presence of jaundice. The resection ratein presence of jaundice in our series was better than the MSKCC group [23/151 (15%) vs 6/84 (7%)]1 while it was not dissimilar in absence of jaundice [121/234 (51%) vs 65/158 (41%)]. A recent French study demonstrated 45% resectability rate achieving R0 resection in 58% of these subjects.4 But
TE D
in this series even patients with positive interaortocaval nodes (usually considered metastatic) underwent radical resection. It is therefore imperative to have better comprehension of GBC patients with jaundice particularly in relation to the spread of disease and the cause of jaundice.
EP
Hepatoduodenal spread of tumour causing thickening of ligament blocking the porta hepatis is
AC C
the most common mechanism of jaundice. It differs from lymph nodal spread as described by Endo et al.15Our analysis showed three distinct mechanisms of jaundice. Infiltration into the porta with or without confluence block was more common than lymph node compression. The resectability rate was low in both HDL infiltration and LN compression without any significant difference among two mechanisms. The third mechanism of jaundice was, presence of CBD stones, seen in 12 (7.3%) patients in the present series. These patients behaved more like GBC patients without jaundice with a higher resectability and survival rate. Only one other large series
ACCEPTED MANUSCRIPT
has reported CBD stones in 2% of their cases of GBC with jaundice.1The essence lies in realizing that some patients with GBC with jaundice may actually have a CBD stone. This should be looked for as these patients are potentially resectable and have a survival equivalent to
RI PT
patients with non-SOJ GBC. In our series the median survival was 61 months for patients with jaundice due to CBD stones underscoring the importance of offering resection in this subgroup.
SC
Morbidity and mortality
Despite improvement in mortality rates across various studies the morbidity rates have been
M AN U
high.16-18 Chan et al reported an overall mortality rate of 1% to 8.3 % in patients undergoing radical resection for carcinoma gallbladder18. The mortality in our series is 6.9% for the entire cohort Patients in SOJ Group had a significantly higher mortality compared to non-SOJ group. This could probably be because jaundiced patients undergo more major resections and therefore
TE D
are more prone to post-operative complications.
Yang et al. found more complications in jaundiced patients in their study on 192 GBC patients(34.0% vs12.4%, p = 0.001).19Inthe present study the morbidity rates, especially major
EP
complications were significantly higher in jaundiced patients(P value = 0.003).Patients with advanced GBC have anorexia and weight loss at presentation and are nutritionally compromised.
AC C
High mortality and morbidity is expected when massive resection is undertaken in these patients, but with careful selection long term survival can be achieved in a few. Survival
Increasing survival trends have been observed with extended radical resections in advanced GBC patients.16-20Patients withR0 resections do better than margin positive resections. The median survival of patients in non-SOJ group was significantly higher than those in SOJ group. Since
ACCEPTED MANUSCRIPT
jaundiced patients belong to locally advanced group it would be appropriate to compare their survival with locally advanced patients in non-SOJ group to derive a meaningful conclusion. Median survival of advanced gallbladder cancers (T3, T4) without jaundice was 19 months
RI PT
compared to 12 months in those with jaundice with no five year survivors in the latter group. The difference in median survival was not statistically significant probably because of the smaller number of resected patients in SOJ group. High postoperative deaths with no 5 year survivor
SC
further makes this problem confounding.
M AN U
Nishio et al5 from Nagoya University, Japan, found significantly lower 5-year (23% vs 54%) and median survival (1.5 vs 15.4 years) for pathologically involved compared to uninvolved CBD.Notably12 patients survived more than 5 years, 6 of whom had lymph node involvement as well. Initial reports from MSKCC indicated median survival of GBC patients with jaundice to be only 6 months. Regimbeau et al4 reporting for the French study group investigated prognostic
TE D
value of jaundice in their 110/429 cases of GBC. They reported that 1 and 3 year survival for resected and non-resected jaundiced patients with GBC was 48% and 19% and 31% and 0% respectively. They concluded that R0 resection did not increase the overall survival in all
EP
resected patients, but R0 resection increased median survival in the subgroup of N0 patients (20
AC C
months versus 6 months, p = 0.01). Yang showed lower five-year survival rates in jaundiced compared tonon-jaundiced patients (6.0% and 36.0%, p < 0.001). However, their multivariate analysis showed that preoperative jaundice was not a significant risk factor for poor outcome (p=0.295).19Agarwal et al6 in a series of 14/51 patients of GBC with jaundice undergoing curative resection reported a mean disease free survival of 23.4 months with 50% patients surviving more than 2 years. Prognostic factors
ACCEPTED MANUSCRIPT
Various prognostic factors have been analysed and reported to be significant in different studies like age, gender, jaundice, liver infiltration, bile duct invasion, lymph node metastasis, distant metastasis, type of surgery (R0/R1/R2), incidental detection, lymphadenectomy, perineural
RI PT
invasion, vascular invasion, lymphatic invasion, grade of differentiation with stage of disease, nodal involvement, presence of jaundice and bile duct involvement.1,3,5,16,17 ,19 , 21 -23We found increasing stage, positive lymph nodes and involvement of adjacent organ as negative predictor
SC
of survival.In a group of 107 patients of gallbladder cancer, Shih et al found stage, incidental finding, jaundice, and lymphadenectomy to remain significant predictors of resection on
M AN U
multivariate analysis.23 Nodal disease
Node positivity remains one of the most important prognostic factor in GBC. The median
irrespective of the T stage .
TE D
survival in our study group was only 14 months in patients with node positive disease
Similar results were shown by Behari et al.24 where survival of patients without nodal disease (median not reached) was significantly better than those with nodal disease (median 17 months)
EP
(p =0.01). In their study none of the patients with nodal disease survived for 5 years compared to
AC C
a 58% 5-year survival in node negative patients. Chijiwa et al. have reported 50% 5-year survival in patients with stage III disease and N1 (hepatoduodenal) lymph node involvement with no 3 year survivors in patients with N2 (retropancreatic) or more distant lymph nodes even after aggressive operations25. Para aortic lymphadenectomy has not shown a definite survival advantage in advanced stage of gallbladder cancer patients undergoing extended resections.26Extended lymphadenectomy incorporating para aortic nodes does not offer survival benefit and we do not advocate it.
ACCEPTED MANUSCRIPT
Adjacent organ Adjacent organ involvement was a poor predictor of survival in our study with a median survival of 14 months vs 77 months in those without adjacent organ resection (P<0.001). Nishio et al.5
RI PT
reported that combined resection of adjacent organ (pancreas, colon and stomach) had16 % five year survival but a median survival of only 0.8 years in patients with involved CBD. They
concluded that although challenging, R0 resection should be done in GBC patients with jaundice
SC
especially when it did not necessitate adjacent organ resection.Lim and coworkers in their series of 279 patients of gallbladder cancer have commented that patients with stage IVb (N2, M0)
M AN U
undergoing aggressive surgical treatment such as hepato-pancreato-duodenectomy, concurrent resection of the adjacent organ or aortocaval node dissection did not have significantly improved prognosis(5-yr overall survival: 0%). This group avoided aggressive surgical treatment in patients with stage IVb (N2, M0) during the second half of the study.27
TE D
Conclusion
GBC is an aggressive disease with poor prognosis but careful selection and achieving R0 resection leads to better survival. Preoperative gastric outlet obstruction, abdominal lump as a
EP
presenting symptom, weight loss and jaundice are significant negative predictors of resectability.
AC C
Patients with jaundice have higher rates of locally advanced disease which is often unresectable or will have distant metastasis. Patients with SOJ undergoing curative resection have higher morbidity and mortality. Long term survival of patients with SOJ who underwent curative resection is less compared to resected patients with locally advanced disease without SOJ although this did not reach significance. Higher tumour stage, adjacent organ involvement and lymphnodal disease are poor prognostic factors. In the presence of these factors, extensive resections in SOJ patients should be considered judiciously.
ACCEPTED MANUSCRIPT
References 1. Hawkins WG, DeMatteo RP, Jarnagin WR, Ben-Porat L, Blumgart LH, Fong Y. Jaundice
RI PT
predicts advanced disease and early mortality in patients with gallbladder cancer. Ann Surg Oncol. 2004 Mar;11(3):310-5.
2. Batra Y, Pal S, Dutta U, Desai P, Garg PK, MakhariaG,et al. Gallbladder cancer in India:
SC
a dismal picture. J Gastroenterol Hepatol.2005 Feb;20(2):309-14.
3. Dwivedi M, Misra SP, Misra V. Clinical and ultrasonographic findings of carcinoma of
M AN U
gallbladder in Indian patients. J Assoc Physicians India 2000 Feb;48(2):192-5. 4. Regimbeau JM, Fuks D, Bachellier P, Le Treut YP, Pruvot FR, Navarro Fet al. Prognostic value of jaundice in patients with gallbladder cancer by the AFC-GBC-2009 study group. Eur J SurgOncol 2011 Jun;37(6):505-12.
5. Nishio H, Ebata T, Yokoyama Y, Igami T, Sugawara G, Nagino M. Gallbladder cancer
TE D
involving the extrahepatic bile duct is worthy of resection. Ann Surg 2011 May;253(5):953-60.
6. Agarwal AK, Mandal S, Singh S, Bhojwani R, Sakhuja P, Uppal R. Biliary obstruction in
EP
gall bladder cancer is not sine qua non of inoperability. Ann Surg Oncol. 2007 Oct;
AC C
14(10): 2831-7)
7. D'Angelica M, Dalal KM, DeMatteo RP, Fong Y, Blumgart LH, Jarnagin WR. Analysis of the extent of resection for adenocarcinoma of the gallbladder. Ann Surg Oncol. 2009 Apr; 16(4):806-16.
8. Chan KM, Yeh TS, Yu MC, Jan YY, Hwang TL, Chen MF. Gallbladder carcinoma with biliary invasion: clinical analysis of the differences from nonbiliary invasion. World J Surg. 2005 Jan;29(1):72-5.
ACCEPTED MANUSCRIPT
9. Sikora SS, Singh RK. Surgical strategies in patients with gallbladder cancer: nihilism to optimism. J Surg Oncol.2006 Jun 15; 93(8):670-81.
Edition. Chichago, IL: Springer, 211-217.
RI PT
10. American Joint Committee on cancer. (2010) AJCC Cancer Staging Manual seventh
11. Rahbari NN, Garden OJ, Padbury R Brooke-Smith M, Crawford M, Adam R et al.
Posthepatectomy liver failure: a definition and grading by the International Study Group
SC
of Liver Surgery (ISGLS). Surgery.2011 May;149(5):713-24.
12. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new
M AN U
proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205–213.
13. Varma V, Gupta S, Soin AS, Nundy S. Does the presence of jaundice and/or a lump in a patient with gall bladder cancer mean that the lesion is not resectable? Dig Surg.
TE D
2009;26(4):306-11.
14. Ong SL, Garcea G, Thomasset SC, Neal CP, Lloyd DM, Berry DP, Dennison AR. Tenyear experience in the management of gallbladder cancer from a single hepatobiliary and
EP
pancreatic centre with review of the literature. HPB (Oxford). 2008;10(6):446-58. 15. Endo I, Shimada H, Fujii Y, Sugita M, Masunari H, Miura Y. Indications for curative
AC C
resection of advanced gallbladder cancer with hepatoduodenal ligament invasion. J HepatobiliaryPancreat Surg. 2001;8(6):505-10.
16. Miyazaki M, Itoh H, Ambiru S, Shimizu H, Togawa A, Gohchi E. Radical surgery for advanced gallbladder carcinoma. Br J Surg. 1996 Apr;83(4):478-81.
ACCEPTED MANUSCRIPT
17. Kai M, Chijiiwa K, Ohuchida J, Nagano M, Hiyoshi M, Kondo K. A curative resection improves the postoperative survival rate even in patients with advanced gallbladder carcinoma. J GastrointestSurg 2007; 11: 1025-32.
RI PT
18. Chan SY, Poon RTP, Lo CM, Ng KK, Fan ST. Management of Carcinoma of
Gallbladder: A Single-Institution Experience in 16 years. J SurgOncol 2008; 97: 156-64. 19. Yang XW, Yuan JM, Chen JY, Yang J, Gao QG, Yan XZ, Zhang BH, Feng S, Wu MC.
SC
The prognostic importance of jaundice in surgical resection with curative intent for gallbladder cancer. BMC Cancer 2014,Sep 3;14:652.
Oncol2003;4:167–176.
M AN U
20. Misra S, Chaturvedi A, Misra NC, et al. Carcinoma of the gallbladder. Lancet
21. Balachandran P, Agarwal S, Krishnani N, Pandey CM, Kumar A, Sikora SS. Predictors of long-term survival in patients with gallbladder cancer. J Gastrointest Surg.2006
TE D
Jun;10(6):848-54.
22. Pais-Costa SR, Farah JF, Artigiani-Neto R, Franco MI, Martins SJ, Goldenberg A.Gallbladder adenocarcinoma: evaluation of the prognostic factors in 100 resectable
EP
cases in Brazil. Arq Bras Cir Dig. 2012 Jan-Mar;25(1):13-9 23. Shih SP, Schulick RD, Cameron JL, Lillemoe KD, Pitt HA, Choti MA, et al. Gallbladder
AC C
cancer: the role of laparoscopy and radical resection. Ann Surg 2007 Jun;245(6):893-901.
24. Behari A, Sikora SS, Wagholikar GD, et al. Long-term survival after extended resections in patients with gallbladder cancer. J Am CollSurg2003;196:82–88.
25. Chijiwa K, Tanaka M. Carcinoma of the gallbladder: an appraisal of surgical resection. Surgery 1994;115:751–756
ACCEPTED MANUSCRIPT
26. Kondo S, Nimura Y, Hayakawa N, et al. Regional and paraaortic lymphadenectomy in radical surgery for advanced gallbladder carcinoma. Br J Surg. 2000;87:418–422. 27. Prognostic factors in patients with gallbladder cancer after surgical resection: analysis of
AC C
EP
TE D
M AN U
SC
ClinGastroenterol. 2013 May-Jun;47(5):443-8
RI PT
279operated patients.Lim H, Seo DW, Park do H, Lee SS, Lee SK, Kim MH, Hwang S. J
ACCEPTED MANUSCRIPT
Table 1 – Demographic Profile of patients with comparison between non-SOJ and SOJ Entire cohort n=385 (%) 50(17-80)
Non-SOJ n=234 (%) 51(17-80)
SOJ n=151 (%) 50(25-76)
p-value
94:291
47:187
47:104
0.01
Pain abdomen
344(89)
222 (95)
122(81)
0.078
Anorexia
232(60)
116 (50)
116(77)
<0.001
Weight loss
237(63)
116 (53)
31(8)
15 (6)
M: F Symptoms
Imaging
121(80)
<0.001
16(11)
0.18
47 (20)
49 (33)
0.005
37(15)
7 (5)
0.001
170 (73)
101 (67)
0.31
M AN U
GOO
0.035
RI PT
Mean Age
SC
Parameters
96(25)
Colon infiltration
44(11)
Liver infiltration
271(70)
Gallstones
290(75)
174 (74)
116 (77)
0.74
5.0
1.0
12.5
<0.001
11.2(5.5-16.2)
11.8(5.5-16)
11 (6.2-16)
0.802
9.0(5-32)
8.3(5-25)
10127(4-32)
0.035
OT (IU/cc)
75(12-166)
49(13-267)
117(12-457)
0.001
PT (IU/cc)
65(35-445)
41(35-265)
104.2(50-587)
<0.001
323(38-5206)
230(38-3068)
608(95-5200)
<0.001
-
7(3-25) days
9.5 (6-28) days
<0.001
Mean Bilirubin Hb (g/dl) 3
ALP
AC C
TLC (x10 )
EP
Laboratory Parameters
TE D
Duodenal infiltration
Median length of hospital stay
ACCEPTED MANUSCRIPT
Table 2 Comparison of resectable and unresectable patients in GBC with no SOJ (non-SOJ Group ) with GBC with SOJ (SOJ Group )
Non-SOJ Group
n=151
n= 234
Curative resection
23 (15.2%)
121 (51.7%)
< 0.001
Inoperable
19 (12.6%)
4 ( 1.7%)
<0.001
Locally advanced
37 (24.5%)
23 (9.8%)
<0.001
Metastatic
72 (47.7%)
86 (36.7%)
<0.001
Metastatic
Metastatic
Peritoneal/diffuse
37 (51.4%)
47 (54.6%)
LN + liver
35 (48.6%)
39 (45.3%)
Procedure
n=23
n=121
AC C
Radical chole with CBD exploration
TE D
Radical Chole with CBD excision
SC 66
0
25
140
12
EP
Completion Radical chole
M AN U
0
Radical chole alone
5
0
Radical chole with adj organ resection
4
8
HPD
5
1
0
3
Trisegmentectomy Open chole
P value
RI PT
SOJ Group
1
Chole=cholecystectomy, HPD=Hepato-pancreato-duodenectomy
0.16
ACCEPTED MANUSCRIPT
Table 3 Predictors of curative resection in entire cohort and subgroups non-SOJ and SOJ Entire Cohort Mutl
HR
Pain
Uni
Multi
.048
.054
0.97 (0.93-1.0)
.008
.103
(0.45- 1.87)
.000
.001
(1.4- 5.7) 4.62 0.001
.024
0.48 .199
(0.16- 1.45) 0.79 0.218
(0.14-4.66) 2.98 .326 (0.34- 26.3)
0.96
.966
.42 .770
.497
(0.21-4.37)
(0.03-5.16) 2.60 .828
.259 (0.49-13.7)
TE D
(0.28- 2.2)
Abdominal Lump
.003
.772
.628
2.35
<0.001
.001
.000
2.73
(1.42- 3.9)
Hepatomegaly
EP
.070
SOJ status
<0.001
6.53
0.003 (1.91-22.2)
.39 .330
.164 (0.10-1.46)
0.94 .108
(0.95- 1.1)
(0.44-3.68)
(0.57-2.54)
1.01
.581
AC C
<0.001
.664
1.20
(0.52- 1.6)
Bilirubin
1.26 .785
(1.4-5.32)
0.92
<0.001
(0.44-11.7) .79
(0.76-51.9)
.138
Cholangitis
.252
2.27
.800
6.28
.088
(1.21- 17.5)
<0.001
.292
(0.92- 1.23)
.328
(1.86-12.8)
.003
Jaundice
.080
M AN U
Gastric oulet obstruction
.415
4.89
.004
HR 1.06
.538
(0.17-1.17)
2.82 <0.001
Multi
0.44
.825
Weight loss
Uni
(0.91-1.0)
0.92 <0.001
HR 0.95
.028
Anorexia
SOJ
RI PT
Uni
NON SOJ
SC
Parameter
.781
1.045 .248
(0.63-1.42)
.240 (0.97-1.12)
ACCEPTED MANUSCRIPT
Table 4 Comparison of surgical procedure among two groups(Removed) Non-SOJ Group
SOJ Group
n=121
n=23
RI PT
Procedure
66
0
Completion Radical Cholecystectomy
25
0
Radical Cholecystectomy with CBD excision ± duodenal
14
5
0
Radical Cholecystectomy with adjacent organ resection
8
0
Trisegmentectomy ± duodenal sleeve/ distal gastrectomy
3
5
1
2
0
2
1
0
HPD ± colonic sleeve resection
TE D
Radical Cholecystectomy with CBD exploration
12
M AN U
sleeve
SC
Radical Cholecystectomy
AC C
Open Cholecystectomy
EP
HPD with portal vein sleeve resection
ACCEPTED MANUSCRIPT
Table 4a Comparison of morbidity
8
SSI
8
Collection/bilioma
2
Sepsis/fever
1
Ascites
1
Others
3
4
2
1
1
2
M AN U
Bile leak
SOJ Group
RI PT
non-SOJ Group
SC
Complications
3
19% vs 56.5% - Statistically significant (P value < 0.001)
Table 4b
Comparison of morbidity based on Clavien Dindo classification
Grade II Grade III
n=121 (%)
n=23 (%)
16(13)
5(22)
0.14
5(4)
2(8.6)
0.25
2(1.6)
4(17)
0.004
0(0)
2(8.6)
0.003
5(4)
5(21.7)
0.002
AC C
Grade IV
SOJ Group
TE D
Grade I
non-SOJ Group
EP
Grading
Grade V
P value
ACCEPTED MANUSCRIPT
Figure 1:Profile of patients with gall bladder cancer (GBC)
RI PT
Patients with suspected GBC(n=437)
Excluded (n=52 patients) ♦ Missed GBCn=35 ♦ Benign etiology n=17
SC
GBC patients (n=385) study group
EP
Unresectable (n=113) • Inoperablen= 4 • Locally advanced unresectable n=23 • metastatic n=86 T
TE D
GBC without jaundice (non-SOJ) (n=234)
M AN U
Assessed for predictors of resectability andlong term survival
AC C
Curative resection (n= 121)
GBC with jaundice (SOJ) (n=151)
Unresectable n=128 • Inoperablen=19 • Locally advanced unresectable n=37 • metastaticn=72
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2- ROC curve for level of bilirubin and resectability. Area under curve- 0.597
ACCEPTED MANUSCRIPT
Figure 3A Kaplan-Meier survival analysis of GBC patients in Non-SOJ group
100
RI PT
Median survival - 61 months 1 year survival - 79.6% 2 year survival - 65% 5 year survival- 52.9%
80
SC
70
60
50
40 0
20
40
M AN U
Survival probability (%)
90
60
80
100
120
140
5
3
1
0
Time
43
25
11
AC C
EP
TE D
Number at risk 95
ACCEPTED MANUSCRIPT
deleted 100 90
RI PT
Median survival - 12 months 1 year survival - 46.6% 2 year survival- 32.4%
80 Survival probability (%) 70 60
SC
50 40
20 10 0
10
M AN U
30
20
30
40
1
0
Time
Number at risk 17
2
AC C
EP
TE D
9
Figure3B:Kaplan-Meier Survival analysis comparing SOJ vs Non SOJ T3-4 group
ACCEPTED MANUSCRIPT
100
Non-SOJ Median survival 19mths 1-yr Survival 59.4% 2-yr Survival 36.7% 5-yr Survival 23.6%
90 80 Survival probability (%) 70
SOJ 12 mths 48.6% 32.4% 0%
RI PT
SOJ Non SOJ T3 T4
60 50 40
SC
30 20 10 10
20
30
40
50
60
70
0
0
0
0
8
5
2
0
M AN U
0
Time
2
1
14
11
AC C
EP
TE D
Number at risk Group: SOJ 17 9 Group: Non SOJ T3 T4 45 30
ACCEPTED MANUSCRIPT
Figure4A: Stage I- median survival not reached, Stage II- 62 months, Stage IIIA- 20 months , Stage IIIB- 14 months, Stage IVA- 8 months, Stage IVB- 9 months (P value <0.001).
80
stage I II IIIA IIIB IVA IVB
RI PT
Survival probability (%)
100
60 40 20 0 20
40
60
80
100
120
140
3
2
1
0
5
3
16
11
5
10
5
1
8
4
2
1
0
0
0
0
0
AC C
EP
TE D
10
M AN U
Time Number at risk Stage I 17 Stage II 28 Stage IIIA 27 Stage IIIB 33 Stage IVA 5 Stage IVB 2
SC
0
1
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
ACCEPTED MANUSCRIPT
AC C
EP
delete
TE D
M AN U
SC
RI PT
Figure 4B: shows declining median survival in both non-SOJ and SOJ patients in stage wise survival curves
ACCEPTED MANUSCRIPT
Figure4C Kaplan-Meier survival curve comparing difference in survival in patients with node positive disease vs node negative 100
RI PT
90 Survival 80 probability (%)
N status
70
N0 62 months N+ 14 months
60
SC
p= 0.001
50
30 0
20
40
M AN U
40
60
80
100
120
140
4
2
1
0
1
1
0
0
Time
33
20
9
12
5
2
AC C
EP
TE D
Number at risk Group: N0 68 Group: N+ 44
Figure 4D Kaplan-Meier survival curve showing difference in survival among a patients with adjacent organ involvement compared to those without adjacent organ
ACCEPTED MANUSCRIPT
involvement 100 90
RI PT
80 Survival probability (%) 70
adjacent organ involvement 0 77 months 1 14 months
60 50
SC
40 30
10 0
20
40
60
80 Time
22
8
5
3
3
0
100
120
140
3
1
0
0
0
0
AC C
EP
TE D
Number at risk Group: 0 81 38 Group: 1 27 7
M AN U
20
ACCEPTED MANUSCRIPT
Highlights •
Gallbladder cancer (GBC) is an aggressive malignancy and we analysed factors predicting resectability and survival of patients with GBC and the impact of surgical
•
RI PT
obstructive jaundice (SOJ). Patients with gastric outlet obstruction, abdominal lump, weight loss and SOJ were more likely to be unresectable. •
Metastatic (36.7% vs 47.7%), inoperable (1.7%vs12.6%) and unresectable disease (9.8%vs24.5%) were more in the SOJ group and had lower curative resection rates
•
SC
(51.7%vs15.2%;p<0.0001)
The 1,2 and 5-year survival rates were higher in patients in the non-SOJ than SOJ
M AN U
group (79.6%,65% and 52.9% vs 48.6,32.4%and 0%;p<0.001). •
GBC with SOJ is more likely to be unresectable.
•
SOJ, nodal involvement, adjacent organ infiltration and higher TNM stage predict
AC C
EP
TE D
poor survival.
S0002-9610(16)30863-7
DOI:
10.1016/j.amjsurg.2017.02.006
Reference:
AJS 12292
To appear in:
The American Journal of Surgery
Received Date: 14 November 2016 Revised Date:
2 February 2017
Accepted Date: 5 February 2017
Please cite this article as: Mishra PK, Saluja SS, Prithiviraj N, Varshney V, Goel N, Patil N, Predictors of curative resection and long term survival of gallbladder cancer – A retrospective analysis, The American Journal of Surgery (2017), doi: 10.1016/j.amjsurg.2017.02.006. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
Title: Predictors of curative resection and long term survival of gallbladder cancer – a
retrospective analysis. Name and initials of authors:
Saluja SS Prithiviraj N
Vaibhav Varshney MCh
Varshney V
Neeraj Goel MCh
Goel N
Nilesh Patil MCh
M AN U
Prithiviraj Nabi MS
SC
Sundeep Singh Saluja MCh
Mishra PK
RI PT
Pramod Kumar Mishra MS, PhD
Patil N
Postal address and email address:
Department of Gastrointestinal Surgery, Room No 218, 2nd floor, Academic Block GB Pant
Corresponding author Sundeep Singh Saluja
TE D
Hospital, New Delhi, India
Professor and Corresponding author,
EP
Department of Gastrointestinal Surgery
G B Pant Institute of Postgraduate Medical Education & Research 1, Jawaharlal Nehru Marg,
AC C
New Delhi 110002.
Email: [email protected] phone no 919718599259 Source of funding: No funding received Category: Original article
Key words: Carcinoma gallbladder, Predictors of resectability, Radical cholecystectomy, resectability in gallbladder cancer, Surgical Obstructive jaundice
ACCEPTED MANUSCRIPT
Background: Gallbladder cancer (GBC) is an aggressive malignancy. We analysed factors predicting resectability and survival of patients with GBC and the impact of surgical
RI PT
obstructive jaundice (SOJ).
Methods: Four hundred and thirty-seven patients with suspected GBC were analysed (52
SC
excluded: benign pathology n=35, missed GBC n=17). The remaining 385 patients were divided into non-SOJ (n=234) and SOJ (n=151) groups. Predictors of resectability and long
M AN U
term survival were analysed and compared with their subgroups..
Results: Patients with gastric outlet obstruction, abdominal lump, weight loss and SOJ were more likely to be unresectable (p:0.04,0.024,0.003 and 0.003, respectively). TNM stage, node positivity and adjacent organ involvement were predictors of survival (p<0.001,0.008 and<0.001). Metastatic (36.7%vs47.7%),
inoperable (1.7%vs12.6%) and unresectable
TE D
disease (9.8%vs24.5%) were more in the SOJ group and had lower curative resection rates (51.7%vs15.2%;p<0.0001). The 1,2 and 5-year survival rates were higher in patients in the
EP
non-SOJ than SOJ group (79.6%,65% and 52.9% vs 48.6,32.4%and 0%;p<0.001).
Conclusion: GBC with SOJ is more likely to be unresectable. SOJ, nodal involvement,
AC C
adjacent organ infiltration and higher TNM stage predict poor survival.
ACCEPTED MANUSCRIPT
Introduction Gallbladder cancer (GBC) is an aggressive disease with a poor prognosis.1 It usually presents late and often at an unresectable stage. Locally advanced GBC presents with a varied
RI PT
symptomatology depending on the extent of the adjacent organ or vascular involvement. About 25-50% of the patients of GBC present with jaundice.1-3 While few authors have reported long term survival in patients of GBC with jaundice4,5,6others consider it an ominous sign1,3,7.
SC
Resectability in these patients may vary from 7% to 58%.1-8Complete surgical removal offers the best chance of cure at present. Surgeons are under constant pressure to increase meaningful
M AN U
survival and possibly, deliver a cure. This has led to pushing the boundaries of resection at the cost of increased morbidity and mortality. Given the aggressive nature of the disease and the mortality/morbidity associated with extensive resections, the subgroup of patients who do not benefit from these resections should also be regularlyanalysed.1,7,9Various factors like stage of
TE D
disease, presence of jaundice, lymph node or visceral involvement have been analysed to determine which patients may actually benefit from these resections.1-7Identification of the factors that may predict unresectability can be used to avoid surgery in inoperable patients and
EP
the associated morbidity. We analysed our data to understand the predictors of curative resection, morbidity and mortality in patients undergoing resection, factors predicting long term survival in
AC C
GBC and the impact of jaundice.
ACCEPTED MANUSCRIPT
Patients and Methods The records of all patients with suspected GBC admitted in our unit from September 2003 till
RI PT
December 2014 were reviewed from theprospectively maintained database. Inclusion criteria
SC
Patients with a diagnosis of GBC deemed resectable on initial work up Exclusion criteria
M AN U
1.GBC masquarades who underwent radical resection for suspected GBC but had benign disease on histopathology.
2a) Missed GBC: defined as patients in whom the diagnosis of GBC was evident on preoperative imaging but who underwent only cholecystectomy elsewhere
TE D
2b) patients with incidental GBC that presented later with advanced disease. The patients were divided into two groups
EP
Non-surgical obstructive jaundice (SOJ) group: GBC with no SOJ. Patients with jaundice due to
AC C
CBD stones were included in the non-SOJ group. SOJ group: These were defined as patients with GBC presenting withSOJsecondary to direct infiltration or lymph node compression of the common bile duct. Preoperative work up
All patients underwent detailed clinical evaluation and blood investigations followed by imaging. Imaging modalities included ultrasound abdomen with Doppler, dual phase contrast enhanced
ACCEPTED MANUSCRIPT
CT scan (CECT).Magnetic resonance cholangio-pancreaticography (MRCP) was done in patients with SOJ. Upper gastrointestinal endoscopy was done routinely in patients suspected to have locally advanced tumour to assess gastro-duodenal infiltration. Endoscopic ultrasound was
RI PT
used in the latter part of the study to assess extent of disease and sampling of nodes in the interaortocaval region. Blood investigations including liver function tests and chest X-ray were done
SC
in all patients. Peri-operative strategy
M AN U
Patients deemed resectable at the time of evaluation in the outpatient were admitted for further detailed evaluation. Patients with obvious para-aortic nodes, main or left portal vein invasion or metastatic disease were considered unresectable. Patients with SOJ underwent percutaneous transhepatic biliary drainage/endoscopic retrograde cholangiography (ERCP) stenting to reduce
Surgical procedure
TE D
serum bilirubin to <5mg/dL. AJCC seventh edition was used for staging.10
Staging laparoscopy was done in all patients to rule out metastatic disease. In the presence of
EP
metastasis, a biopsy confirmation was done and the procedure abandoned. Inter-aortocaval
AC C
lymph node was sampled and a frozen section examination done before proceeding to assess local resectability. Palliative biliary bypass/gastrojejunostomy was done when required in patients found to have unresectable disease during surgery. In patients undergoing curative resection, R0 resection was achieved with resection of a 2 cm wedge of liver/segment IVB, V resection along with the gallbladder mass. In case of adjacent organ involvement either sleeve or segmental resection was done to obtain an R0 resection. Standard lymphadenectomy included clearance of lymph nodes along the hepatoduodenal ligament, and the pericholedochal,
ACCEPTED MANUSCRIPT
periportal, peripancreatic and right celiac nodes. Extended right hepatectomy was done in those in whom the hilum was involved with infiltration of the pedicle on the right side. The cystic duct margin was routinely sent for frozen section and the common bile duct was resected in patients
if required. Definitions Curative resection: Patients who underwent R0 resection.
SC
RI PT
with obvious infiltration by tumour, a positive cystic duct margin and for lymph nodal clearance,
M AN U
Inoperable disease: Patients with inadequate liver remnant or poor performance status to undergo major surgery
Locally advanced unresectable patients: Those with infiltration of the main portal vein, hepatic
TE D
artery or encasement of a major vascular pedicle precluding R0 resection. Metastatic: Any patient with peritoneal, omental, ovarian or liver metastases. Presence of positive inter-aortocaval (IAC) and lymph node on the left side of the coeliac axis were also
EP
considered as metastatic disease.
The predictors of curative resection and long term survival were analysed in the entire cohort.
AC C
Subgroup analysis among patients with or without SOJ was also done. The survival among patients in the SOJ group was compared with those with locally advanced(T3-4) non-SOJ disease to assess the impact of jaundice Factors predicting curative resection: Preoperative parameters such as presence of pain abdomen at time of presentation, loss of weight, loss of appetite, features of gastric outlet
ACCEPTED MANUSCRIPT
obstruction, jaundice, palpable abdominal lump, hepatomegaly, level of bilirubin were analysed by multivariable logistic regression analysis for their ability to predict curative resection.
RI PT
Morbidity and mortality Operative mortality included all patients who died within 30 days of the procedure or during the same hospital admission. Morbidity included all postoperative complications occurring within 30
SC
days of procedure or during the same hospital admission. Post-hepatectomy liver failure was defined according to definitions by ISGLS.11 Morbidity was classified according to the Clavien-
M AN U
Dindo classification.12 Factors predicting Survival
a) TNM stage was evaluated as per AJCC 7th edition b)Nodal involvement was defined
Follow up
TE D
aspresence of histologically positive lymph node c)Adjacent organ involvement
Patients were followed up every 3 months for the first year then every 6 months. They were
EP
assessed clinically, using laboratory tests including liver functions tests and abdominal ultrasound. A CECT scan of the abdomen was done at one year in patients who remained well. It
AC C
was done earlier if necessitated by clinical or ultrasound findings. All patients were treated by multimodality approach with gemcitabine-based chemotherapy in both the adjuvant and palliative setting. According to our protocol adjuvant chemotherapy was given to all patients with stage T2 disease and above, and for all lymph node positive tumours irrespective of the T stage. Best possible supportive care was given in the palliative setting.
ACCEPTED MANUSCRIPT
Statistical analysis Statistical analysis was done using SPSS software (IBM SPSS Statistics for Windows, Version
RI PT
22.0. Armonk, NY: IBM Corp). Categorical variables were compared with the χ2 test or Fisher’s exact test. Wherever appropriate, multivariate analysis was done using the logistic regression model to determine independent predictors of outcome. Long-term survival was calculated using
SC
the Kaplan-Meier method, and differences in groups of survivor were compared by the log-rank test. Multivariate analysis was done using the Cox regression model to determine independent
AC C
EP
TE D
M AN U
predictors of survival. A P value less than 0.05 was considered to indicate statistical significance.
ACCEPTED MANUSCRIPT
Results A total of 437 patients with suspected GBC were admitted after preliminary evaluation in the
RI PT
outpatient department from October 2003 to December 2014. Fifty-two patients were excluded either because of benign pathology on postoperative histology (n=35) or a missed GBC (n=17). Among 385 patients included in the final study, 234 patients (61%) had GBC with no SOJ (Non-
SC
SOJgroup ) while151 patients (43%) had GBC with SOJ(SOJ group).The patient profile is shown in Figure 1.
M AN U
Abdominal pain was the most common presenting symptom(89%) followed by weight loss (63%) and loss of appetite (60%) (Table 1). Thirty-one (8%) patients presented with gastric outlet obstruction. The incidence of anorexia, weight loss and adjacent colon/duodenal infiltration were significantly more in SOJ group (p<0.001, 0.001, 0.001 and 0.005,
TE D
respectively). The incidence of metastasis, inoperable disease and unresectable locally advanced disease was significantly more in the SOJ group while curative resection was more in the nonSOJ group (Table 2).Although the incidence of metastasis in lymph nodes and liver was higher
EP
in the SOJ group, the difference was not statistically significant (P=0.16). The major reason for inoperability in the SOJ group was deterioration of general condition during the waiting period
AC C
for fall of serum bilirubin following biliary drainage, while co-existing major medical illness not permitting surgery in patients with non-SOJ group.
Predictors of curative resection Overall, 154 patients underwent curative resection––curative resections were done significantly more in non-SOJ as compared to SOJ groups (121/234 vs 23/151; p<0.001). Factors predicting curative resection(Table 3)
ACCEPTED MANUSCRIPT
There was no difference in resectability rate among males compared to females and patients with <50 years compared to those >50 years of age. On univariate analysis, factors with negative prediction for resectability were presence of pain abdomen, anorexia, significant weight loss
RI PT
(>10% of body weight), gastric outlet obstruction, hepatomegaly, abdominal lump, presence of jaundice and bilirubin >1mg/dl. Age and history of cholangitis had no influence on resectability. On multivariate analysis preoperative gastric outlet obstruction, abdominal lump, weight loss and
SC
jaundice on presentation were significant negative predictors of resectability (p value: 0.04, 0.024, 0.003 and 0.003, respectively). The preoperative bilirubin did not correlate with
M AN U
resectability (area under curve of 0.597 in ROC). (Fig2). Direct infiltration into the hepatoduodenal ligament was present in 116 patients in our series with a patent biliary confluence in 70 patients and blocked in 46 patients. Porta hepatis lymph nodes compressing the mid/lower CBD as a cause of jaundice was seen in 32 patients. Resectability rates did not differ
TE D
with cause of SOJ (P=0.29).
On subgroup analysis of the non-SOJ and SOJ groups with respect to resectability, increasing age group had less resectable disease in the SOJ group compared with the non-SOJ
EP
group (P=0.004). Preoperative gastric outlet obstruction, abdominal lump as presentation and
AC C
weight loss were significant predictors of inoperability in non-SOJ group but not in the SOJ group (Table 3).
Surgical Procedure
Operative resection included radical cholecystectomy with liver wedge/IVB,V resection and standard lymphadenectomy alone in 66 cases, adjacent organ/CBD resection in 34 patients, CBD exploration in 5 patients and open cholecystectomy in one. Completion radical cholecystectomy
ACCEPTED MANUSCRIPT
for incidental GBC was done in 25 patients. Trisegmentectomy was done in 8 cases and hepatopancreato-duodenectomy was done in 5 cases. (Table 2).
RI PT
Morbidity, mortality and hospital stay The morbidity grading based on Clavien-Dindo classification of surgical complications in both groups are as given in table 4b. There was significantly less morbidity in the non-SOJ group
SC
compared to the SOJ group [19% (23/121)vs56 % (12/23) P<0.001] (Table 4a). The SOJ group had significant increase in major complications (Grades III and IV) in comparison with the non-
M AN U
SOJ group but the incidence of minor complications was similar between the two groups. The overall mortality of the entire study population was 6.9%. The 30-day mortality was significantly less in the non-SOJ compared to the SOJ group(4% vs 21%; p=0.002).Of the five perioperative deaths in the SOJ group, 3 were directly related to surgical complications, while in the other two
TE D
they were due to massive stroke and cardiac failure in the postoperative period. The median length of hospital stay was significantly shorter in patients in the non-SOJ compared to the SOJ
Survival
AC C
Non-SOJ group
EP
group [7(3-25) days vs 9.5 (6-28) p<0.001].
Of the 234 patients, 121 underwent curative surgery. The 30-day mortality was 5/121 (4%).The median survival of patients without SOJ was 61months, while the 1-, 2- and 5-year survival rate was 79.6%, 65%, and 52.9%, respectively (Figure3A). When further subgroup analysis of patients with locally advanced GBC (pathological T3 and T4 disease) was done, the median survival decreased to 19 months. Among the 60 patients with pathological T3 and T4 disease,
ACCEPTED MANUSCRIPT
there were two perioperative deaths (4%).The 1-, 2- and 5-year survival rates were 59.4%. 36.7% and 23.6% respectively (Figure3B).
RI PT
SOJ group Among 154 patients, 23 patients of GBC with SOJ underwent curative resection. The 30-day mortality was 5/23 (21%). All patients in this group were followed till death. The median
SC
survival was 12 months, while the 1- and 2-year survival rates were 48.6% and 32.4%, respectively (Figure 3B). However, there were no five-year survivors in this group.
M AN U
Comparison of survival in locally advanced (T3,T4) non-SOJ versus SOJ patients Patients with SOJ have at least T3 disease considering the CBD infiltration and was compared with T3, T4 disease in non-SOJ group to analyse the impact of jaundice. Survival curves of these two groups by log-rank test did not reveal any statistical significance despite a difference in the
TE D
median survival of these two groups (P value=0.44) (Figure 3B). Predictors of long-term survival
EP
Long-term survival was dependent on the TNM stage of the disease, the N status and involvement of adjacent organ. Survival of the entire cohort had significant difference across all
AC C
stages by log-rank test (P value<0.001) (Figure 4A).Stage-wise survival in the non-SOJ and SOJ groups was similar with declining median survival with increasing stages (Figure 4B ). Stage wise comparison between both groups showed no statistical difference in median survival (Stage IIIap=0.59 Stage IIIb p=0.53 and stage IV p=0.27). Nodal positivity had poor survival compared with node negative disease. Median survival in node negative disease was 62 months as against 14 months in node positive disease. (p=0.008 by
ACCEPTED MANUSCRIPT
log rank test)(Figure 4C). Median survival in patients with adjacent organ involvement was 14 months compared to 77 months in patients without any adjacent organ involvement with P value
RI PT
being statistically significant at P<0.001. (Figure 4D) Discussion
GBC is a biologically aggressive disease with late presentation and poor prognosis. Despite
SC
screening from the outpatient department 41% patients of the entire cohort were found to have metastatic disease at surgery. It is important that in the SOJ group only one of seven patients
M AN U
underwent curative resection while in the non-SOJ group one in every two patients underwent curative resection. Therefore, it is important to assess the predictors of resectability particularly in patients with SOJ. Resectability
TE D
This aspect of GBC has been seldom addressed in the literature. Varma et al.13found presence of abdominal lump and jaundice as signs of advanced disease but not of unresectability. On the contrary we found preoperative gastric outlet obstruction, abdominal lump as a presenting
EP
symptom, weight loss and jaundice on presentation as negative predictors of resectability on multivariate analysis. However these predictors lose their significance in SOJ group implying
AC C
presence of jaundice itself as the most important determinant. Soek et al.14from UK reported raised bilirubin and alkaline phosphatase along with raised total leukocyte & platelet count as predictors for unresectable disease. This suggests tacitly that jaundice is an important determinant of unresectability similar to our series. The utility of these predictors is little as they per se do not contraindicate surgery. However, careful preoperative workup in these subgroups
ACCEPTED MANUSCRIPT
may reduce unwarranted laparotomies and may help in planning neoadjuvant treatment in these subgroups.
RI PT
Gastric outlet obstruction may be a surrogate marker of advanced disease but duodenal infiltration per se is not a contraindication for resection. Distal gastrectomy or duodenal sleeve resection was done in 14 patients in our series.
SC
Curative resection
One hundred and forty four (37%)patients underwent curative resection in the present series.
M AN U
Resection rate significantly decreased in presence of jaundice. The resection ratein presence of jaundice in our series was better than the MSKCC group [23/151 (15%) vs 6/84 (7%)]1 while it was not dissimilar in absence of jaundice [121/234 (51%) vs 65/158 (41%)]. A recent French study demonstrated 45% resectability rate achieving R0 resection in 58% of these subjects.4 But
TE D
in this series even patients with positive interaortocaval nodes (usually considered metastatic) underwent radical resection. It is therefore imperative to have better comprehension of GBC patients with jaundice particularly in relation to the spread of disease and the cause of jaundice.
EP
Hepatoduodenal spread of tumour causing thickening of ligament blocking the porta hepatis is
AC C
the most common mechanism of jaundice. It differs from lymph nodal spread as described by Endo et al.15Our analysis showed three distinct mechanisms of jaundice. Infiltration into the porta with or without confluence block was more common than lymph node compression. The resectability rate was low in both HDL infiltration and LN compression without any significant difference among two mechanisms. The third mechanism of jaundice was, presence of CBD stones, seen in 12 (7.3%) patients in the present series. These patients behaved more like GBC patients without jaundice with a higher resectability and survival rate. Only one other large series
ACCEPTED MANUSCRIPT
has reported CBD stones in 2% of their cases of GBC with jaundice.1The essence lies in realizing that some patients with GBC with jaundice may actually have a CBD stone. This should be looked for as these patients are potentially resectable and have a survival equivalent to
RI PT
patients with non-SOJ GBC. In our series the median survival was 61 months for patients with jaundice due to CBD stones underscoring the importance of offering resection in this subgroup.
SC
Morbidity and mortality
Despite improvement in mortality rates across various studies the morbidity rates have been
M AN U
high.16-18 Chan et al reported an overall mortality rate of 1% to 8.3 % in patients undergoing radical resection for carcinoma gallbladder18. The mortality in our series is 6.9% for the entire cohort Patients in SOJ Group had a significantly higher mortality compared to non-SOJ group. This could probably be because jaundiced patients undergo more major resections and therefore
TE D
are more prone to post-operative complications.
Yang et al. found more complications in jaundiced patients in their study on 192 GBC patients(34.0% vs12.4%, p = 0.001).19Inthe present study the morbidity rates, especially major
EP
complications were significantly higher in jaundiced patients(P value = 0.003).Patients with advanced GBC have anorexia and weight loss at presentation and are nutritionally compromised.
AC C
High mortality and morbidity is expected when massive resection is undertaken in these patients, but with careful selection long term survival can be achieved in a few. Survival
Increasing survival trends have been observed with extended radical resections in advanced GBC patients.16-20Patients withR0 resections do better than margin positive resections. The median survival of patients in non-SOJ group was significantly higher than those in SOJ group. Since
ACCEPTED MANUSCRIPT
jaundiced patients belong to locally advanced group it would be appropriate to compare their survival with locally advanced patients in non-SOJ group to derive a meaningful conclusion. Median survival of advanced gallbladder cancers (T3, T4) without jaundice was 19 months
RI PT
compared to 12 months in those with jaundice with no five year survivors in the latter group. The difference in median survival was not statistically significant probably because of the smaller number of resected patients in SOJ group. High postoperative deaths with no 5 year survivor
SC
further makes this problem confounding.
M AN U
Nishio et al5 from Nagoya University, Japan, found significantly lower 5-year (23% vs 54%) and median survival (1.5 vs 15.4 years) for pathologically involved compared to uninvolved CBD.Notably12 patients survived more than 5 years, 6 of whom had lymph node involvement as well. Initial reports from MSKCC indicated median survival of GBC patients with jaundice to be only 6 months. Regimbeau et al4 reporting for the French study group investigated prognostic
TE D
value of jaundice in their 110/429 cases of GBC. They reported that 1 and 3 year survival for resected and non-resected jaundiced patients with GBC was 48% and 19% and 31% and 0% respectively. They concluded that R0 resection did not increase the overall survival in all
EP
resected patients, but R0 resection increased median survival in the subgroup of N0 patients (20
AC C
months versus 6 months, p = 0.01). Yang showed lower five-year survival rates in jaundiced compared tonon-jaundiced patients (6.0% and 36.0%, p < 0.001). However, their multivariate analysis showed that preoperative jaundice was not a significant risk factor for poor outcome (p=0.295).19Agarwal et al6 in a series of 14/51 patients of GBC with jaundice undergoing curative resection reported a mean disease free survival of 23.4 months with 50% patients surviving more than 2 years. Prognostic factors
ACCEPTED MANUSCRIPT
Various prognostic factors have been analysed and reported to be significant in different studies like age, gender, jaundice, liver infiltration, bile duct invasion, lymph node metastasis, distant metastasis, type of surgery (R0/R1/R2), incidental detection, lymphadenectomy, perineural
RI PT
invasion, vascular invasion, lymphatic invasion, grade of differentiation with stage of disease, nodal involvement, presence of jaundice and bile duct involvement.1,3,5,16,17 ,19 , 21 -23We found increasing stage, positive lymph nodes and involvement of adjacent organ as negative predictor
SC
of survival.In a group of 107 patients of gallbladder cancer, Shih et al found stage, incidental finding, jaundice, and lymphadenectomy to remain significant predictors of resection on
M AN U
multivariate analysis.23 Nodal disease
Node positivity remains one of the most important prognostic factor in GBC. The median
irrespective of the T stage .
TE D
survival in our study group was only 14 months in patients with node positive disease
Similar results were shown by Behari et al.24 where survival of patients without nodal disease (median not reached) was significantly better than those with nodal disease (median 17 months)
EP
(p =0.01). In their study none of the patients with nodal disease survived for 5 years compared to
AC C
a 58% 5-year survival in node negative patients. Chijiwa et al. have reported 50% 5-year survival in patients with stage III disease and N1 (hepatoduodenal) lymph node involvement with no 3 year survivors in patients with N2 (retropancreatic) or more distant lymph nodes even after aggressive operations25. Para aortic lymphadenectomy has not shown a definite survival advantage in advanced stage of gallbladder cancer patients undergoing extended resections.26Extended lymphadenectomy incorporating para aortic nodes does not offer survival benefit and we do not advocate it.
ACCEPTED MANUSCRIPT
Adjacent organ Adjacent organ involvement was a poor predictor of survival in our study with a median survival of 14 months vs 77 months in those without adjacent organ resection (P<0.001). Nishio et al.5
RI PT
reported that combined resection of adjacent organ (pancreas, colon and stomach) had16 % five year survival but a median survival of only 0.8 years in patients with involved CBD. They
concluded that although challenging, R0 resection should be done in GBC patients with jaundice
SC
especially when it did not necessitate adjacent organ resection.Lim and coworkers in their series of 279 patients of gallbladder cancer have commented that patients with stage IVb (N2, M0)
M AN U
undergoing aggressive surgical treatment such as hepato-pancreato-duodenectomy, concurrent resection of the adjacent organ or aortocaval node dissection did not have significantly improved prognosis(5-yr overall survival: 0%). This group avoided aggressive surgical treatment in patients with stage IVb (N2, M0) during the second half of the study.27
TE D
Conclusion
GBC is an aggressive disease with poor prognosis but careful selection and achieving R0 resection leads to better survival. Preoperative gastric outlet obstruction, abdominal lump as a
EP
presenting symptom, weight loss and jaundice are significant negative predictors of resectability.
AC C
Patients with jaundice have higher rates of locally advanced disease which is often unresectable or will have distant metastasis. Patients with SOJ undergoing curative resection have higher morbidity and mortality. Long term survival of patients with SOJ who underwent curative resection is less compared to resected patients with locally advanced disease without SOJ although this did not reach significance. Higher tumour stage, adjacent organ involvement and lymphnodal disease are poor prognostic factors. In the presence of these factors, extensive resections in SOJ patients should be considered judiciously.
ACCEPTED MANUSCRIPT
References 1. Hawkins WG, DeMatteo RP, Jarnagin WR, Ben-Porat L, Blumgart LH, Fong Y. Jaundice
RI PT
predicts advanced disease and early mortality in patients with gallbladder cancer. Ann Surg Oncol. 2004 Mar;11(3):310-5.
2. Batra Y, Pal S, Dutta U, Desai P, Garg PK, MakhariaG,et al. Gallbladder cancer in India:
SC
a dismal picture. J Gastroenterol Hepatol.2005 Feb;20(2):309-14.
3. Dwivedi M, Misra SP, Misra V. Clinical and ultrasonographic findings of carcinoma of
M AN U
gallbladder in Indian patients. J Assoc Physicians India 2000 Feb;48(2):192-5. 4. Regimbeau JM, Fuks D, Bachellier P, Le Treut YP, Pruvot FR, Navarro Fet al. Prognostic value of jaundice in patients with gallbladder cancer by the AFC-GBC-2009 study group. Eur J SurgOncol 2011 Jun;37(6):505-12.
5. Nishio H, Ebata T, Yokoyama Y, Igami T, Sugawara G, Nagino M. Gallbladder cancer
TE D
involving the extrahepatic bile duct is worthy of resection. Ann Surg 2011 May;253(5):953-60.
6. Agarwal AK, Mandal S, Singh S, Bhojwani R, Sakhuja P, Uppal R. Biliary obstruction in
EP
gall bladder cancer is not sine qua non of inoperability. Ann Surg Oncol. 2007 Oct;
AC C
14(10): 2831-7)
7. D'Angelica M, Dalal KM, DeMatteo RP, Fong Y, Blumgart LH, Jarnagin WR. Analysis of the extent of resection for adenocarcinoma of the gallbladder. Ann Surg Oncol. 2009 Apr; 16(4):806-16.
8. Chan KM, Yeh TS, Yu MC, Jan YY, Hwang TL, Chen MF. Gallbladder carcinoma with biliary invasion: clinical analysis of the differences from nonbiliary invasion. World J Surg. 2005 Jan;29(1):72-5.
ACCEPTED MANUSCRIPT
9. Sikora SS, Singh RK. Surgical strategies in patients with gallbladder cancer: nihilism to optimism. J Surg Oncol.2006 Jun 15; 93(8):670-81.
Edition. Chichago, IL: Springer, 211-217.
RI PT
10. American Joint Committee on cancer. (2010) AJCC Cancer Staging Manual seventh
11. Rahbari NN, Garden OJ, Padbury R Brooke-Smith M, Crawford M, Adam R et al.
Posthepatectomy liver failure: a definition and grading by the International Study Group
SC
of Liver Surgery (ISGLS). Surgery.2011 May;149(5):713-24.
12. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new
M AN U
proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg. 2004;240:205–213.
13. Varma V, Gupta S, Soin AS, Nundy S. Does the presence of jaundice and/or a lump in a patient with gall bladder cancer mean that the lesion is not resectable? Dig Surg.
TE D
2009;26(4):306-11.
14. Ong SL, Garcea G, Thomasset SC, Neal CP, Lloyd DM, Berry DP, Dennison AR. Tenyear experience in the management of gallbladder cancer from a single hepatobiliary and
EP
pancreatic centre with review of the literature. HPB (Oxford). 2008;10(6):446-58. 15. Endo I, Shimada H, Fujii Y, Sugita M, Masunari H, Miura Y. Indications for curative
AC C
resection of advanced gallbladder cancer with hepatoduodenal ligament invasion. J HepatobiliaryPancreat Surg. 2001;8(6):505-10.
16. Miyazaki M, Itoh H, Ambiru S, Shimizu H, Togawa A, Gohchi E. Radical surgery for advanced gallbladder carcinoma. Br J Surg. 1996 Apr;83(4):478-81.
ACCEPTED MANUSCRIPT
17. Kai M, Chijiiwa K, Ohuchida J, Nagano M, Hiyoshi M, Kondo K. A curative resection improves the postoperative survival rate even in patients with advanced gallbladder carcinoma. J GastrointestSurg 2007; 11: 1025-32.
RI PT
18. Chan SY, Poon RTP, Lo CM, Ng KK, Fan ST. Management of Carcinoma of
Gallbladder: A Single-Institution Experience in 16 years. J SurgOncol 2008; 97: 156-64. 19. Yang XW, Yuan JM, Chen JY, Yang J, Gao QG, Yan XZ, Zhang BH, Feng S, Wu MC.
SC
The prognostic importance of jaundice in surgical resection with curative intent for gallbladder cancer. BMC Cancer 2014,Sep 3;14:652.
Oncol2003;4:167–176.
M AN U
20. Misra S, Chaturvedi A, Misra NC, et al. Carcinoma of the gallbladder. Lancet
21. Balachandran P, Agarwal S, Krishnani N, Pandey CM, Kumar A, Sikora SS. Predictors of long-term survival in patients with gallbladder cancer. J Gastrointest Surg.2006
TE D
Jun;10(6):848-54.
22. Pais-Costa SR, Farah JF, Artigiani-Neto R, Franco MI, Martins SJ, Goldenberg A.Gallbladder adenocarcinoma: evaluation of the prognostic factors in 100 resectable
EP
cases in Brazil. Arq Bras Cir Dig. 2012 Jan-Mar;25(1):13-9 23. Shih SP, Schulick RD, Cameron JL, Lillemoe KD, Pitt HA, Choti MA, et al. Gallbladder
AC C
cancer: the role of laparoscopy and radical resection. Ann Surg 2007 Jun;245(6):893-901.
24. Behari A, Sikora SS, Wagholikar GD, et al. Long-term survival after extended resections in patients with gallbladder cancer. J Am CollSurg2003;196:82–88.
25. Chijiwa K, Tanaka M. Carcinoma of the gallbladder: an appraisal of surgical resection. Surgery 1994;115:751–756
ACCEPTED MANUSCRIPT
26. Kondo S, Nimura Y, Hayakawa N, et al. Regional and paraaortic lymphadenectomy in radical surgery for advanced gallbladder carcinoma. Br J Surg. 2000;87:418–422. 27. Prognostic factors in patients with gallbladder cancer after surgical resection: analysis of
AC C
EP
TE D
M AN U
SC
ClinGastroenterol. 2013 May-Jun;47(5):443-8
RI PT
279operated patients.Lim H, Seo DW, Park do H, Lee SS, Lee SK, Kim MH, Hwang S. J
ACCEPTED MANUSCRIPT
Table 1 – Demographic Profile of patients with comparison between non-SOJ and SOJ Entire cohort n=385 (%) 50(17-80)
Non-SOJ n=234 (%) 51(17-80)
SOJ n=151 (%) 50(25-76)
p-value
94:291
47:187
47:104
0.01
Pain abdomen
344(89)
222 (95)
122(81)
0.078
Anorexia
232(60)
116 (50)
116(77)
<0.001
Weight loss
237(63)
116 (53)
31(8)
15 (6)
M: F Symptoms
Imaging
121(80)
<0.001
16(11)
0.18
47 (20)
49 (33)
0.005
37(15)
7 (5)
0.001
170 (73)
101 (67)
0.31
M AN U
GOO
0.035
RI PT
Mean Age
SC
Parameters
96(25)
Colon infiltration
44(11)
Liver infiltration
271(70)
Gallstones
290(75)
174 (74)
116 (77)
0.74
5.0
1.0
12.5
<0.001
11.2(5.5-16.2)
11.8(5.5-16)
11 (6.2-16)
0.802
9.0(5-32)
8.3(5-25)
10127(4-32)
0.035
OT (IU/cc)
75(12-166)
49(13-267)
117(12-457)
0.001
PT (IU/cc)
65(35-445)
41(35-265)
104.2(50-587)
<0.001
323(38-5206)
230(38-3068)
608(95-5200)
<0.001
-
7(3-25) days
9.5 (6-28) days
<0.001
Mean Bilirubin Hb (g/dl) 3
ALP
AC C
TLC (x10 )
EP
Laboratory Parameters
TE D
Duodenal infiltration
Median length of hospital stay
ACCEPTED MANUSCRIPT
Table 2 Comparison of resectable and unresectable patients in GBC with no SOJ (non-SOJ Group ) with GBC with SOJ (SOJ Group )
Non-SOJ Group
n=151
n= 234
Curative resection
23 (15.2%)
121 (51.7%)
< 0.001
Inoperable
19 (12.6%)
4 ( 1.7%)
<0.001
Locally advanced
37 (24.5%)
23 (9.8%)
<0.001
Metastatic
72 (47.7%)
86 (36.7%)
<0.001
Metastatic
Metastatic
Peritoneal/diffuse
37 (51.4%)
47 (54.6%)
LN + liver
35 (48.6%)
39 (45.3%)
Procedure
n=23
n=121
AC C
Radical chole with CBD exploration
TE D
Radical Chole with CBD excision
SC 66
0
25
140
12
EP
Completion Radical chole
M AN U
0
Radical chole alone
5
0
Radical chole with adj organ resection
4
8
HPD
5
1
0
3
Trisegmentectomy Open chole
P value
RI PT
SOJ Group
1
Chole=cholecystectomy, HPD=Hepato-pancreato-duodenectomy
0.16
ACCEPTED MANUSCRIPT
Table 3 Predictors of curative resection in entire cohort and subgroups non-SOJ and SOJ Entire Cohort Mutl
HR
Pain
Uni
Multi
.048
.054
0.97 (0.93-1.0)
.008
.103
(0.45- 1.87)
.000
.001
(1.4- 5.7) 4.62 0.001
.024
0.48 .199
(0.16- 1.45) 0.79 0.218
(0.14-4.66) 2.98 .326 (0.34- 26.3)
0.96
.966
.42 .770
.497
(0.21-4.37)
(0.03-5.16) 2.60 .828
.259 (0.49-13.7)
TE D
(0.28- 2.2)
Abdominal Lump
.003
.772
.628
2.35
<0.001
.001
.000
2.73
(1.42- 3.9)
Hepatomegaly
EP
.070
SOJ status
<0.001
6.53
0.003 (1.91-22.2)
.39 .330
.164 (0.10-1.46)
0.94 .108
(0.95- 1.1)
(0.44-3.68)
(0.57-2.54)
1.01
.581
AC C
<0.001
.664
1.20
(0.52- 1.6)
Bilirubin
1.26 .785
(1.4-5.32)
0.92
<0.001
(0.44-11.7) .79
(0.76-51.9)
.138
Cholangitis
.252
2.27
.800
6.28
.088
(1.21- 17.5)
<0.001
.292
(0.92- 1.23)
.328
(1.86-12.8)
.003
Jaundice
.080
M AN U
Gastric oulet obstruction
.415
4.89
.004
HR 1.06
.538
(0.17-1.17)
2.82 <0.001
Multi
0.44
.825
Weight loss
Uni
(0.91-1.0)
0.92 <0.001
HR 0.95
.028
Anorexia
SOJ
RI PT
Uni
NON SOJ
SC
Parameter
.781
1.045 .248
(0.63-1.42)
.240 (0.97-1.12)
ACCEPTED MANUSCRIPT
Table 4 Comparison of surgical procedure among two groups(Removed) Non-SOJ Group
SOJ Group
n=121
n=23
RI PT
Procedure
66
0
Completion Radical Cholecystectomy
25
0
Radical Cholecystectomy with CBD excision ± duodenal
14
5
0
Radical Cholecystectomy with adjacent organ resection
8
0
Trisegmentectomy ± duodenal sleeve/ distal gastrectomy
3
5
1
2
0
2
1
0
HPD ± colonic sleeve resection
TE D
Radical Cholecystectomy with CBD exploration
12
M AN U
sleeve
SC
Radical Cholecystectomy
AC C
Open Cholecystectomy
EP
HPD with portal vein sleeve resection
ACCEPTED MANUSCRIPT
Table 4a Comparison of morbidity
8
SSI
8
Collection/bilioma
2
Sepsis/fever
1
Ascites
1
Others
3
4
2
1
1
2
M AN U
Bile leak
SOJ Group
RI PT
non-SOJ Group
SC
Complications
3
19% vs 56.5% - Statistically significant (P value < 0.001)
Table 4b
Comparison of morbidity based on Clavien Dindo classification
Grade II Grade III
n=121 (%)
n=23 (%)
16(13)
5(22)
0.14
5(4)
2(8.6)
0.25
2(1.6)
4(17)
0.004
0(0)
2(8.6)
0.003
5(4)
5(21.7)
0.002
AC C
Grade IV
SOJ Group
TE D
Grade I
non-SOJ Group
EP
Grading
Grade V
P value
ACCEPTED MANUSCRIPT
Figure 1:Profile of patients with gall bladder cancer (GBC)
RI PT
Patients with suspected GBC(n=437)
Excluded (n=52 patients) ♦ Missed GBCn=35 ♦ Benign etiology n=17
SC
GBC patients (n=385) study group
EP
Unresectable (n=113) • Inoperablen= 4 • Locally advanced unresectable n=23 • metastatic n=86 T
TE D
GBC without jaundice (non-SOJ) (n=234)
M AN U
Assessed for predictors of resectability andlong term survival
AC C
Curative resection (n= 121)
GBC with jaundice (SOJ) (n=151)
Unresectable n=128 • Inoperablen=19 • Locally advanced unresectable n=37 • metastaticn=72
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
Figure 2- ROC curve for level of bilirubin and resectability. Area under curve- 0.597
ACCEPTED MANUSCRIPT
Figure 3A Kaplan-Meier survival analysis of GBC patients in Non-SOJ group
100
RI PT
Median survival - 61 months 1 year survival - 79.6% 2 year survival - 65% 5 year survival- 52.9%
80
SC
70
60
50
40 0
20
40
M AN U
Survival probability (%)
90
60
80
100
120
140
5
3
1
0
Time
43
25
11
AC C
EP
TE D
Number at risk 95
ACCEPTED MANUSCRIPT
deleted 100 90
RI PT
Median survival - 12 months 1 year survival - 46.6% 2 year survival- 32.4%
80 Survival probability (%) 70 60
SC
50 40
20 10 0
10
M AN U
30
20
30
40
1
0
Time
Number at risk 17
2
AC C
EP
TE D
9
Figure3B:Kaplan-Meier Survival analysis comparing SOJ vs Non SOJ T3-4 group
ACCEPTED MANUSCRIPT
100
Non-SOJ Median survival 19mths 1-yr Survival 59.4% 2-yr Survival 36.7% 5-yr Survival 23.6%
90 80 Survival probability (%) 70
SOJ 12 mths 48.6% 32.4% 0%
RI PT
SOJ Non SOJ T3 T4
60 50 40
SC
30 20 10 10
20
30
40
50
60
70
0
0
0
0
8
5
2
0
M AN U
0
Time
2
1
14
11
AC C
EP
TE D
Number at risk Group: SOJ 17 9 Group: Non SOJ T3 T4 45 30
ACCEPTED MANUSCRIPT
Figure4A: Stage I- median survival not reached, Stage II- 62 months, Stage IIIA- 20 months , Stage IIIB- 14 months, Stage IVA- 8 months, Stage IVB- 9 months (P value <0.001).
80
stage I II IIIA IIIB IVA IVB
RI PT
Survival probability (%)
100
60 40 20 0 20
40
60
80
100
120
140
3
2
1
0
5
3
16
11
5
10
5
1
8
4
2
1
0
0
0
0
0
AC C
EP
TE D
10
M AN U
Time Number at risk Stage I 17 Stage II 28 Stage IIIA 27 Stage IIIB 33 Stage IVA 5 Stage IVB 2
SC
0
1
0
0
0
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
ACCEPTED MANUSCRIPT
AC C
EP
delete
TE D
M AN U
SC
RI PT
Figure 4B: shows declining median survival in both non-SOJ and SOJ patients in stage wise survival curves
ACCEPTED MANUSCRIPT
Figure4C Kaplan-Meier survival curve comparing difference in survival in patients with node positive disease vs node negative 100
RI PT
90 Survival 80 probability (%)
N status
70
N0 62 months N+ 14 months
60
SC
p= 0.001
50
30 0
20
40
M AN U
40
60
80
100
120
140
4
2
1
0
1
1
0
0
Time
33
20
9
12
5
2
AC C
EP
TE D
Number at risk Group: N0 68 Group: N+ 44
Figure 4D Kaplan-Meier survival curve showing difference in survival among a patients with adjacent organ involvement compared to those without adjacent organ
ACCEPTED MANUSCRIPT
involvement 100 90
RI PT
80 Survival probability (%) 70
adjacent organ involvement 0 77 months 1 14 months
60 50
SC
40 30
10 0
20
40
60
80 Time
22
8
5
3
3
0
100
120
140
3
1
0
0
0
0
AC C
EP
TE D
Number at risk Group: 0 81 38 Group: 1 27 7
M AN U
20
ACCEPTED MANUSCRIPT
Highlights •
Gallbladder cancer (GBC) is an aggressive malignancy and we analysed factors predicting resectability and survival of patients with GBC and the impact of surgical
•
RI PT
obstructive jaundice (SOJ). Patients with gastric outlet obstruction, abdominal lump, weight loss and SOJ were more likely to be unresectable. •
Metastatic (36.7% vs 47.7%), inoperable (1.7%vs12.6%) and unresectable disease (9.8%vs24.5%) were more in the SOJ group and had lower curative resection rates
•
SC
(51.7%vs15.2%;p<0.0001)
The 1,2 and 5-year survival rates were higher in patients in the non-SOJ than SOJ
M AN U
group (79.6%,65% and 52.9% vs 48.6,32.4%and 0%;p<0.001). •
GBC with SOJ is more likely to be unresectable.
•
SOJ, nodal involvement, adjacent organ infiltration and higher TNM stage predict
AC C
EP
TE D
poor survival.