Disease-free survival following resection in non-ductal periampullary cancers: A retrospective multicenter analysis

Accepted Manuscript Disease-free survival following resection in non-ductal periampullary cancers: A retrospective multicenter analysis Stéphane Bourgouin, MD, Jacques Ewald, MD, Julien Mancini, MD, PhD, Vincent Moutardier, MD, PhD, Jean-Robert Delpero, MD, Yves-Patrice Le Treut, MD PII:

S1743-9191(17)30363-1

DOI:

10.1016/j.ijsu.2017.04.051

Reference:

IJSU 3773

To appear in:

International Journal of Surgery

Received Date: 31 January 2017 Revised Date:

8 April 2017

Accepted Date: 23 April 2017

Please cite this article as: Bourgouin S, Ewald J, Mancini J, Moutardier V, Delpero J-R, Le Treut YP, Disease-free survival following resection in non-ductal periampullary cancers: A retrospective multicenter analysis, International Journal of Surgery (2017), doi: 10.1016/j.ijsu.2017.04.051. 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

DISEASE-FREE SURVIVAL FOLLOWING RESECTION IN NONDUCTAL PERIAMPULLARY CANCERS: A RETROSPECTIVE

RI PT

MULTICENTER ANALYSIS

Shot title: Recurrence in non-ductal cancers

Authors: Stéphane BOURGOUIN, MD a,b, Jacques EWALD, MD c, Julien MANCINI, MD, PhD d, Vincent

Aix-Marseille University, APHM La Conception, Department of Surgery and Liver Transplantation, Marseille,

M AN U

a

SC

MOUTARDIER, MD, PhD e, Jean-Robert DELPERO, MD c and Yves-Patrice LE TREUT, MD a

France

Sainte Anne Military Teaching Hospital, Department of Oncologic and Digestive Surgery, Toulon, France

c

Aix-Marseille University, Institut Paoli-Calmettes, Department of Oncologic Surgery, Marseille, France

d

Aix-Marseille University, UMR912 SESSTIM, Inserm, IRD and APHM La Timone, Department of Public

Health, Marseille, France e

TE D

b

Aix-Marseille University, APHM Hôpital Nord, Department of Visceral Surgery, Marseille, France

EP

Corresponding author permanent address:

Stéphane Bourgouin, Sainte Anne Military Teaching Hospital, Department of Oncologic and Digestive Surgery,

AC C

Boulevard Sainte Anne, BP 20545, 83041 Toulon Cedex 09, France Tel : + 33 6 64 24 76 32 ; Fax : + 33 4 83 16 27 11 ; Mail : [email protected] Funding source:

This research did not receive any specific grant from funding agencies in the public, commercial, or not-forprofit sectors.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT ABSTRACT Background: Predictors of recurrence following pancreaticoduodenectomy are well described for ductal periampullary cancers but lack reliability for non-ductal tumors. The purpose of this study is to analyze the disease-free survival (DFS) and to define the predictors of recurrence following resection for ampullary (AC),

RI PT

bile duct (BDC) and duodenal cancers (DC). Materials and Methods: Clinico-pathological data of patients operated on between 2001 and 2011 were

retrospectively reviewed. The effect of lymphatic invasion was specified using the lymph node ratio (LNR) and the number of positive nodes (NPN), with thresholds calculated with the likelihood ratio. Kaplan–Meier disease-

analyses were performed to identify predictors of recurrence.

SC

free survival (DFS) curves were compared for all covariates by a log-rank test. Multivariate logistic regression

M AN U

Results: 135 patients were identified. Mean follow-up was 49±35 months. Median DFS was not reached for AC and was 36 and 18 months for DC and BDC, respectively. Five-year DFS was 52%, 43% and 32% for AC, DC and BDC, respectively. Predictors of recurrence were T4 tumors, neural invasion and preoperative biliary drainage for DC, ≥3 positive nodes and ≥4% loss of BMI for AC, and T3–T4 tumors for BDC. Conclusion: loss of BMI ≥4% is a strong predictor of recurrence in AC, and the recurrence risk increases with

TE D

the total number of lymph nodes invaded (0; 1–3; ≥4). Only T stage influences recurrence for BDC. Considering DC, the adverse effect of preoperative biliary drainage should be validated in randomized series. Keywords: Pancreaticoduodenectomy; Duodenal Neoplasms; Common Bile Duct Neoplasms; Ampulla of Vater

AC C

EP

Neoplasms; Disease-free Survival

1

ACCEPTED MANUSCRIPT 1.

INTRODUCTION

Periampullary cancers regroup a heterogeneous set of lesions including ductal cancers from the pancreatic head (PC), and non-ductal tumors as ampullary cancers (AC), distal bile duct cholangiocarcinomas (BDC) and duodenal cancers (DC). Regardless of the tumor origin, surgical resection by pancreaticoduodenectomy (PD) is

RI PT

the only curative option. However, despite a common treatment, these tumors have very different long-term postoperative prognoses reflecting, for each tumor, the effects of specific factors of aggressiveness.

These factors have been already well described for PC [1-6]. However, predictors of recurrence are less studied for non-ductal cancers, because: (i) AC, BDC and DC have a lower incidence rate; (ii) despite a different

SC

prognosis, AC, BDC and DC are most frequently analyzed together with PC to enable powerful statistical analyses; (iii) smaller series focusing separately on each tumor type lack of reproducibility.

M AN U

Recently, our team analyzed the prognostic factors of overall survival following PD for AC, BDC and DC using a multicenter retrospective database [7] . The aim of the present study is to extend the results of our first analysis to the disease-free survival (DFS) to separately identify for AC, BDC and DC the factors which could influence

AC C

EP

TE D

recurrence.

2

ACCEPTED MANUSCRIPT 2.

MATERIAL AND METHODS

Medical charts of patients who benefited from a PD between 2001 and 2011 for a periampullary tumor in three expert centers of the Aix-Marseille University, France, were retrospectively reviewed. Only patients with proven AC, BDC and DC on the definitive histological report were included. All the patients benefited from a

RI PT

preoperative diagnosis workup including contrast-enhanced thoraco-abdominal computed tomography, and, if necessary, magnetic resonance imaging or endoscopic ultrasonography. Patients were excluded from the present study when they had metastatic or locally advanced tumor at the time of diagnosis or when resection margins were macroscopically invaded on the definitive histological report. Due to the period inclusion, some patients

SC

received neoadjuvant chemoradiation either because of borderline tumors on preoperative imaging or in the context of a local trial: all these patients had resectable disease either at initial staging or at restaging before

complies with the STROBE guidelines. 2.1 Surgery

M AN U

surgery [8]. The study was approved by Institutional Review Board of the Aix-Marseille University, and

After thorough abdominal examination, intraoperative ultrasonography was routinely performed to assess resectability against the mesenterico-portal trunk. A classical or pylorus-preserving PD (PPPD) was then

TE D

performed with en bloc lymph node resection along the common hepatic artery, coeliac trunk, hepatoduodenal ligament and pancreaticoduodenal arteries. Frozen sections of the resection margins were extemporaneously examined. Reconstruction was achieved by an end-to-side pancreaticojejunostomy (PJ) or

EP

pancreaticogastrostomy (PG), hepaticojejunostomy and pylorojejunostomy or gastrojejunostomy. External transanastomotic duct stent of the pancreatic remnant and prophylactic octreotid were routinely used in two

AC C

centers. Drainage of the abdominal cavity was systematically achieved. 2.2 Study parameters

Demographic, operative and postoperative parameters retrieved included gender, age, Body Mass Index (BMI), American Society of Anesthesiologists (ASA) score, jaundice, preoperative biliary drainage, type of resection, pancreatic hardness, pancreatic and bile duct dilation (more than 3 mm and 8 mm respectively), postoperative outcomes according to the Dindo–Clavien classification [9], and adjuvant or neo-adjuvant treatment. Postoperative pancreatic fistula (POPF) and postpancreatectomy hemorrhage (PPH) were detailed according to the International Study Group of Pancreatic Surgery [10,11]. Histological data included the tumor origin, maximal tumor size (in cm), resection margin (negative, R0, or positive, R1), node stage (number of lymph

3

ACCEPTED MANUSCRIPT nodes examined, number of positive nodes (NPN), and lymph node ratio (LNR) defined as the ratio of positive to total number of examined nodes), tumor differentiation (good, moderate or poor), microangio and neural invasion. Disease staging was according to the TNM classification (American Joint Committee on Cancer 7th Edition 2010). The status of patients was retrieved at January 2015, the census date. Disease-free survival was

RI PT

determined from the date of operation to the date of tumor recurrence. 2.3 Statistical analysis

Statistical analyses were performed and reviewed by a biomedical statistician (JM) using SPSS 20.0 (SPSS Inc., Chicago, IL, USA). Categorical variables are described in terms of frequency (percentages) and continuous

SC

variables as the mean (± standard deviation) or the median (range). Owing to the retrospective nature of this work, some variables were missing. Percentages are thus recorded as valid percentages based on the number of

M AN U

patients with the data available for each parameter. Univariate analyses were conducted using a Student’s t-test or a Kruskal-Wallis test for continuous variables, and a Chi-square test or Fischer’s exact test for categorical variables, as appropriate. A two-tailed P-value ≤0.05 was considered statistically significant. To specifically assess the impact of lymph node status by means of the NPN and the LNR as categorical variables, best cut-off points were determined using Chi-square scores calculated with the likelihood ratio method. Disease-free

TE D

survival curves were constructed using the Kaplan–Meier method, then compared by log-rank analysis. Multivariate analyses were performed using the Cox proportional hazards regression model, including variables

AC C

EP

from univariate analysis most likely to impact survival (P ≤0.2).

4

ACCEPTED MANUSCRIPT 3.

RESULTS

From January 2001 to January 2011, 135 patients were identified: 55 AC, 55 BDC, and 25 DC. Sixty-six patients came from the Paoli-Calmettes Institute, 41 from the Conception Hospital and 28 from the North Hospital. The median age of the series was 66 (28–81) years (Table 1). There were no differences in demographics between

RI PT

groups, except for ASA score. Seventy patients (52%) benefited from endoscopic preoperative biliary drainage either because of clinical

jaundice (n = 62, 46%) or due to neoadjuvant treatment (n = 8, 6%): 26 patients (47%) in the AC group, 36 (65%) in the BDC group and 8 (32%) in the DC group. The median delay to surgery was 2.5 weeks (1-4) and 10

SC

weeks (8-12) in case of drainage for clinical jaundice and neoadjuvant treatment, respectively. Considering resection, 109 classical PD and 26 PPPD were performed. The pancreas was soft or friable for 72 patients (67%).

M AN U

Clavien III-IV-V complications occurred in 48 patients (36%). The global in-hospital mortality rate was 9% (12 patients). Seven patients (5%) deceased because of a surgical complication (six PPH, one multi-organ failure). The POPF rate was 27%. The median length of hospital stay was 20 days. Histological data are presented in Table 2. The DC group included larger and more T4 tumors. Thirteen patients (10%) presented positive resection margins. R1 resections were more frequent in the BDC group, which also included more neural invasion.

TE D

Fifteen patients were excluded from the DFS analysis: 12 died postoperatively, and three presented hepatic metastases on the definitive histological report despite negative extemporaneous analysis. Mean follow-up for the 120 remaining patients was 49 ±35 months. Up to January 2015, 67 patients experienced recurrences. In the AC group, three patients experienced pancreatic recurrences, three developed thoracic metastases, and all the

EP

others presented with hepatic or peritoneal metastases. In the BDC group, most recurrences involved the peritoneum or the liver, including four recurrences on the hepaticojejunal anastomosis, and eight patients

AC C

experienced thoracic or bone metastases. All the recurrences were abdominal in the DC group, except for three patients who presented thoracic metastases. Median DFS was not achieved in the AC group (Fig. 1), and 1-year, 2-year and 5-year DFS rates were 75%, 61% and 52% respectively. These results were not statistically different from the DC group (39 months, 79%, 70% and 43%) or the BDC group (18 months, 72%, 45% and 32%). Eighteen patients in the AC group had nodal invasion. The NPN ranged from 1 to 28 nodes (median 2.5 nodes), and the LNR ranged from 7.1% to 90.0% (median 19.4%). Best cut-off points were identified as three nodes for the NPN, and 15% for the LNR. Thus, DFS analyses were carried out using three categories for the NPN (0; 1–3; ≥4 nodes) and the LNR (0; 1–15; >15%). In the BDC group, the NPN and the LNR among the 22 patients with nodal disease ranged from 1 to 12 nodes (median 3 nodes) and from 5.6% to 62.5% (median 19.4%). Best cut-off

5

ACCEPTED MANUSCRIPT points were identified as two nodes (0; 1–2; ≥3 nodes) and 15% (0; 1–15; >15%). In the DC group, 14 patients had nodal metastases. The NPN and the LNR ranged from 1 to 15 nodes (median 4 nodes) and from 7.7% to 65.2% (median 28.9%), with best cut-off points identified as two nodes (0; 1–2; ≥ 3nodes) and 15% (0; 1–15; >15%), respectively. In the DC group, ≥ 4% loss of BMI, preoperative biliary drainage and T4 stage significantly impacted DFS on

RI PT

univariate analyses. Multivariate analysis showed that only T4 stage, preoperative biliary drainage and neural invasion were independent predictors (Table 3, Fig. 2). Correlation analysis was used to investigate the effect of preoperative biliary drainage. Half of the patients with biliary drainage benefited from neo-adjuvant

SC

chemotherapy. However, neither neo-adjuvant therapy, nor the other clinico-pathological covariates controlled for, including T stage and tumor size, were statistically associated with biliary drainage.

Considering AC, ≥4% loss of BMI, T3–T4 stages, N stage, NPN, and microangio invasion affected DFS on

M AN U

univariate analyses. In multivariate analysis, only the loss of BMI and more than three positive nodes were independently associated with poor DFS (Table 3, Fig. 3).

In the BDC group, preoperative biliary drainage, T3 and T4 stages, NPN, LNR, positive resection margins and neural invasion were all significant by univariate analysis, whereas only T3 and T4 stages were considered as independent predictors in multivariate model (Table 3, Fig. 4).

TE D

Factors evaluated but not found to be significant in any groups were age, gender, ASA score, operative

AC C

EP

parameters, tumor differentiation, postoperative outcomes and use of neoadjuvant or adjuvant chemotherapy.

6

ACCEPTED MANUSCRIPT 4.

DISCUSSION

This study analyzed the long-term results of 135 AC, BDC and DC resected over a ten-year period with a minimum of 49 months of follow-up to allow the detection of even late recurrences in long-term survivors. Results showed that AC patients had the best prognosis, followed by DC, then by BDC. Factors influencing

RI PT

long-term failure were T4 tumors, neural invasion and preoperative biliary drainage for DC, loss of BMI and more than 3 nodes invaded for AC, and T3–T4 tumors for BDC. These results highlight that, if biological factors of malignancy largely determine the long-term prognosis following resection, other factors such as inadequate preoperative nutritional management or invasive preoperative procedures could also alter survival. However, it

SC

should be noted that this is a retrospective study exposed to selection bias. Second, the inclusion period (10 years) was long, leading to the inclusion of a heterogeneous cohort. Finally, as for other rare tumors, statistical

M AN U

analyses may be underpowered because of the inclusion of too few patients. As a consequence, multivariate analyses have to be interpreted only to determine which factors independently predict recurrence, rather than to evaluate the predictive strength of the factor identified by means of the Hazard Ratios. High postoperative mortality in our series ensues from the inclusion of pathologies reputed to be at high risk of postoperative complications because they develop on soft pancreases (67% in our series). Secondly, the

TE D

mortality rate was calculated including all of the patients who experienced a Clavien V event, regardless of the delay; thus, this is different from the 30-day or 90-day mortality rate which has been reported in other studies. AC have a more favorable prognosis than other periampullary tumors [12]. Because AC develops just distal to the biliopancreatic junction, small tumors can be revealed by an early onset of symptoms, leading to rapid

EP

diagnosis with, consequently, a high level of complete (R0) resection. In this study, 47% of the patients in the AC group presented with jaundice. As a result, 57% of the patients presented T1-T2 tumors, and R0 resection

AC C

rate was 98%. Recognized predictors of survival are T stage, NPN, neural invasion, tumor differentiation and positive resection margins [13-16]. Nodal status has been proposed as a predictor of both overall survival [16] and DFS [17]. Our results are consistent with these findings, and also show that the NPN is a major predictor of recurrence. Moreover, the cut-off points used in our study (0; 1–3; ≥4 positive nodes), and determined by the likelihood ratio, are the same as those identified by Sakata et al. [16], and Sierzega et al. [18], who used different statistical methods. This finding indicates that lymph node status, especially the NPN, is a strong predictor in AC, and that stratifying patients into three categories of positive nodes is a reproducible predictor for overall and disease-free long-term survival.

7

ACCEPTED MANUSCRIPT The effect of weight loss in the onset of tumor recurrences has already been well discussed in previous studies [7,19-21]. Actually, weight loss is the reflection of the systemic inflammatory response, and is the consequence of major tumor–host interactions as part of the cancer anorexia–cachexia syndrome [22]. Studies have demonstrated that weight loss is usually well correlated with the magnitude of the inflammatory response and is a reliable predictor of survival, in ductal and non-ductal periampullary cancers [21]. In our study, weight loss

RI PT

was studied by stratifying patients by loss of BMI. The optimal threshold was found when loss of BMI exceeded 4% of the pre-illness BMI with more than two third of patients experiencing recurrences at five years of followup. A weight loss exceeding 4% of the initial BMI should thus be considered as a strong marker of tumor

SC

aggressiveness, and as a reliable predictor of recurrence in AC. However, effect of the regulation of metabolic disturbances on DFS in patients with AC should be clarified on prospective series.

This is the first time that endoscopic preoperative biliary drainage has been identified as an independent factor of

M AN U

recurrence in DC. Few reports have examined the effect of preoperative biliary drainage in periampullary cancers [23-27]. Murakami et al. [23] analyzed the long-term outcomes of 160 resected PC, and found that transhepatic percutaneous drainage increased peritoneal recurrences and was detrimental to survival, conversely to endoscopic drainage. Similar findings were reported by Martigoni et al.[24] about 190 resected AC and PC, and by Smith et al. [25] about 155 resected PC. Our results differed from these findings, but must be interpreted

TE D

in light of the biases inherent to our work. Secondly, although no significant correlations were found when controlling for clinico-pathological covariates, it should be noted that preoperative biliary drainage in DC was performed on half of the patients because of neo-adjuvant therapy, resulting in a supplementary delay to surgery.

EP

The effect of digestive tract tumor stenting has also been suggested to be detrimental in digestive tract cancers because stent insertion increases the rate of tumor circulating cells thus increasing the risk of recurrences [28-

AC C

31]. Thus, the negative impact of preoperative biliary drainage in DC could result from: (1) an increased delay between diagnosis and surgery due to neo-adjuvant therapy, suggesting that three months of preoperative FOLFOX treatment is not advantageous for DC; or (2) the increase of tumor seeding in DC because of the need to drill through the mass to bridge the tumor, explaining why the adverse effect of preoperative biliary drainage has not been found in AC and BDC, which were smaller than DC. In any case, further investigations on larger series are needed for firmer conclusions about the effect of preoperative biliary drainage in DC and, more generally, in periampullary tumors. However, in the time since this study, we have modified the management of patients with obstructive tumors in our centers, avoiding excessive drainage in those patients who tolerate

8

ACCEPTED MANUSCRIPT obstructive jaundice, and reserving endoscopic drainage to those who develop cholangitis or liver failure, instituting a principle of preoperative no-touch isolation technique. Despite a common treatment, AC, DC and BDC harbor very different long-term DFS following PD. This study highlights the importance of malignancy factors such as T stage, nodal status, neural and lymph node invasion in the establishment of oncological prognosis. However, our results also suggest that, if not controlled, preoperative

RI PT

management factors, such as nutritional status, could also be detrimental. Further investigations should be carried out for more robust conclusions about the effect of preoperative biliary drainage in DC, and more

AC C

EP

TE D

M AN U

SC

generally in the set of periampullary tumors.

9

ACCEPTED MANUSCRIPT 5. 1.

REFERENCES C.R. Ferrone, D.M. Finkelstein, S.P. Thayer, A. Muzikansky, C. Fernandez-delCastillo, A.L. Warshaw, Perioperative CA19-9 levels can predict stage and survival in patients with resectable pancreatic adenocarcinoma, J. Clin. Oncol. 24 (2006) 2897-2902. C.L. Hallemeier, M. Botros, M.M. Corsini, M.G. Haddock, L.L. Gunderson, R.C. Miller, Preoperative

RI PT

2.

CA 19-9 level is an important prognostic factor in patients with pancreatic adenocarcinoma treated with surgical resection and adjuvant concurrent chemoradiotherapy, Am. J. Clin. Oncol. 34 (2011) 567-572. 3.

W. Hartwig, T. Hackert, U. Hinz, A. Gluth, F. Bergmann, O. Strobel, et al., Pancreatic cancer surgery in

4.

SC

the new millennium: better prediction of outcome, Ann. Surg. 254 (2011) 311-319.

T. Schnelldorfer, A.L. Ware, M.G. Sarr, T.C. Smyrk, L. Zhang, R. Qin, et al., Long-term survival after

M AN U

pancreatoduodenectomy for pancreatic adenocarcinoma: is cure possible?, Ann. Surg. 247 (2008) 456462. 5.

M. Waraya, K. Yamashita, H. Katagiri, K. Ishii, Y. Takahashi, K. Furuta, et al., Preoperative serum CA19-9 and dissected peripancreatic tissue margin as determiners of long-term survival in pancreatic cancer, Ann. Surg. Oncol. 16 (2009) 1231-1240.

T. Zacharias, D. Jaeck, E. Oussoultzoglou, A. Neuville, P. Bachellier, Impact of lymph node

TE D

6.

involvement on long-term survival after R0 pancreaticoduodenectomy for ductal adenocarcinoma of the pancreas, J. Gastrointest. Surg. 11 (2007) 350-356. 7.

S. Bourgouin, J. Ewald, J. Mancini, V. Moutardier, J.R. Delpero, Y.P. Le Treut, Predictors of Survival

EP

in Ampullary, Bile Duct and Duodenal Cancers Following Pancreaticoduodenectomy: a 10-Year Multicentre Analysis, J. Gastrointest. Surg. 19 (2015) 1247-1255. O. Turrini, F. Viret, L. Moureau-Zabotto, J. Guiramand, V. Moutardier, B. Lelong, et al., Neoadjuvant 5

AC C

8.

fluorouracil-cisplatin chemoradiation effect on survival in patients with resectable pancreatic head adenocarcinoma: a ten-year single institution experience, Oncology. 76 (2009) 413-419.

9.

D. Dindo, N. Demartines, P.A. Clavien, Classification of surgical complications: a new proposal with

evaluation in a cohort of 6336 patients and results of a survey, Ann. Surg. 240 (2004) 205-213. 10.

C. Bassi, C. Dervenis, G. Butturini, A. Fingerhut, C. Yeo, J. Izbicki, et al., Postoperative pancreatic fistula: an international study group (ISGPF) definition, Surgery. 138 (2005) 8-13.

10

ACCEPTED MANUSCRIPT 11.

M.N. Wente, J.A. Veit, C. Bassi, C. Dervenis, A. Fingerhut, D.J. Gouma, et al., Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition, Surgery. 142 (2007) 20-25.

12.

F. Klein, D. Jacob, M. Bahra, U. Pelzer, G. Puhl, A. Krannich, et al., Prognostic factors for long-term

pancreaticoduodenectomy, HPB Surg. 2014 (2014) 970234. 13.

RI PT

survival in patients with ampullary carcinoma: the results of a 15-year observation period after

J.P. Duffy, O.J. Hines, J.H. Liu, C.Y. Ko, G. Cortina, W.H. Isacoff, et al., Improved survival for

adenocarcinoma of the ampulla of Vater: fifty-five consecutive resections, Arch. Surg. 138 (2003) 941-

14.

SC

948; discussion 948-950.

J. Klempnauer, G.J. Ridder, H. Maschek, R. Pichlmayr, Carcinoma of the ampulla of vater: determinants of long-term survival in 94 resected patients, HPB Surg. 11 (1998) 1-11. Q.L. Qiao, Y.G. Zhao, M.L. Ye, Y.M. Yang, J.X. Zhao, Y.T. Huang, et al., Carcinoma of the ampulla

M AN U

15.

of Vater: factors influencing long-term survival of 127 patients with resection, World J. Surg. 31 (2007) 137-143; discussion 144-136. 16.

J. Sakata, Y. Shirai, T. Wakai, Y. Ajioka, K. Akazawa, K. Hatakeyama, Assessment of the nodal status in ampullary carcinoma: the number of positive lymph nodes versus the lymph node ratio, World J.

17.

TE D

Surg. 35 (2011) 2118-2124.

J.S. Park, D.S. Yoon, K.S. Kim, J.S. Choi, W.J. Lee, H.S. Chi, et al., Factors influencing recurrence after curative resection for ampulla of Vater carcinoma, J. Surg. Oncol. 95 (2007) 286-290. M. Sierzega, K. Nowak, J. Kulig, A. Matyja, W. Nowak, T. Popiela, Lymph node involvement in

EP

18.

ampullary cancer: the importance of the number, ratio, and location of metastatic nodes, J. Surg. Oncol.

19.

AC C

100 (2009) 19-24.

J. Bachmann, M.W. Buchler, H. Friess, M.E. Martignoni, Cachexia in patients with chronic pancreatitis

and pancreatic cancer: impact on survival and outcome, Nutr. Cancer. 65 (2013) 827-833.

20.

J.S. Falconer, K.C. Fearon, J.A. Ross, R. Elton, S.J. Wigmore, O.J. Garden, et al., Acute-phase protein

response and survival duration of patients with pancreatic cancer, Cancer. 75 (1995) 2077-2082. 21.

H.R. Scott, D.C. McMillan, L.M. Forrest, D.J. Brown, C.S. McArdle, R. Milroy, The systemic inflammatory response, weight loss, performance status and survival in patients with inoperable nonsmall cell lung cancer, Br. J. Cancer. 87 (2002) 264-267.

11

ACCEPTED MANUSCRIPT 22.

I. Ronga, F. Gallucci, F. Riccardi, G. Uomo, Anorexia-cachexia syndrome in pancreatic cancer: recent advances and new pharmacological approach, Adv. Med. Sci. 59 (2014) 1-6.

23.

Y. Murakami, K. Uemura, Y. Hashimoto, N. Kondo, N. Nakagawa, H. Sasaki, et al., Does preoperative biliary drainage compromise the long-term survival of patients with pancreatic head carcinoma?, J. Surg. Oncol. 111 (2015) 270-276. M.E. Martignoni, M. Wagner, L. Krahenbuhl, C.A. Redaelli, H. Friess, M.W. Buchler, Effect of

RI PT

24.

preoperative biliary drainage on surgical outcome after pancreatoduodenectomy, Am. J. Surg. 181 (2001) 52-59; discussion 87.

R.A. Smith, K. Dajani, S. Dodd, P. Whelan, M. Raraty, R. Sutton, et al., Preoperative resolution of

SC

25.

jaundice following biliary stenting predicts more favourable early survival in resected pancreatic ductal adenocarcinoma, Ann. Surg. Oncol. 15 (2008) 3138-3146.

T.J. Strom, J.B. Klapman, G.M. Springett, K.L. Meredith, S.E. Hoffe, J. Choi, et al., Comparative long-

M AN U

26.

term outcomes of upfront resected pancreatic cancer after preoperative biliary drainage, Surg. Endosc. 29 (2015) 3273-3281. 27.

K. Uemura, Y. Murakami, S. Satoi, M. Sho, F. Motoi, M. Kawai, et al., Impact of Preoperative Biliary Drainage on Long-Term Survival in Resected Pancreatic Ductal Adenocarcinoma: A Multicenter

28.

TE D

Observational Study, Ann. Surg. Oncol. 22 Suppl 3 (2015) S1238-1246. K. Maruthachalam, G.E. Lash, B.K. Shenton, A.F. Horgan, Tumour cell dissemination following endoscopic stent insertion, Br. J. Surg. 94 (2007) 1151-1154. C. Sabbagh, F. Browet, M. Diouf, C. Cosse, O. Brehant, E. Bartoli, et al., Is stenting as "a bridge to

EP

29.

surgery" an oncologically safe strategy for the management of acute, left-sided, malignant, colonic

30.

AC C

obstruction? A comparative study with a propensity score analysis, Ann. Surg. 258 (2013) 107-115. C. Sabbagh, D. Chatelain, N. Trouillet, F. Mauvais, S. Bendjaballah, F. Browet, et al., Does use of a metallic colon stent as a bridge to surgery modify the pathology data in patients with colonic obstruction? A case-matched study, Surg. Endosc. 27 (2013) 3622-3631.

31.

E.M. Steinert, R.H. Schwartz, N.J. Singh, At low precursor frequencies, the T-cell response to chronic self-antigen results in anergy without deletion, Eur. J. Immunol. 42 (2012) 2875-2880.

12

ACCEPTED MANUSCRIPT 6.

CONFLICT OF INTEREST STATEMENT

AC C

EP

TE D

M AN U

SC

RI PT

Conflicts of interest: none

13

ACCEPTED MANUSCRIPT 7.

FIGURE LEGENDS

Fig.1 Disease-free survival following pancreaticoduodenectomy, by tumor type Fig.2 Disease-free survival for duodenal cancers, by a) preoperative biliary drainage, b) T stage and c) neural

RI PT

invasion Fig.3 Disease-free survival for ampullary cancers, by a) number of positive nodes and b) loss of BMI

AC C

EP

TE D

M AN U

SC

Fig.4 Disease-free survival for bile duct cancers, by T stage

14

ACCEPTED MANUSCRIPT Table 1. Demographic, operative, and postoperative data, by tumor type

Variable

Total 135 (%)

Ampullary 55 (%)

Bile Duct 55 (%)

Duodenal 25 (%)

p

Demographic data Age (years) a

66 (28-81)

66 (28-81)

65 (37-79)

69 (41-80)

ns

Men

73 (54)

25 (45)

35 (64)

13 (52)

ns

24 (15-43)

24 (18-33)

24 (15-43)

Loss of BMI ≥ 4%

56 (41)

21 (38)

24 (44)

ASA 3

43 (32)

11 (20)

19 (35)

Jaundice

62 (46)

26 (47)

31 (56)

Biliary drainage

70 (52)

26 (47)

36 (65)

8 (6)

1 (2)

1 (2)

Neoadjuvant therapy Classical PD Soft pancreas

109 b

72

41 (74.5)

(67)

29 (66)

59

(44)

36 (66)

Biliary dilation

114

(86)

51 (94)

Pancreatic stent

78

(58)

Postoperative data In-hospital mortality

12 (9)

Clavien III-IV-V

48 (36)

POPF

38 (28)

PPH

28 (21)

13 (52)

0.015

5 (20)

0.012

8 (32)

0.014

6 (24)

<0.001

20 (80.0)

ns

31 (69)

12 (67)

ns

16 (29)

7 (29)

<0.001

50 (91)

13 (54)

<0.001

31 (56)

32 (58)

15 (60)

5 (9)

6 (11)

1 (4)

ns

18 (33)

24 (44)

6 (24)

ns

ns

17 (31)

14 (25)

7 (28)

ns

10 (18)

14 (25)

4 (16)

ns

20 (2-133)

19 (8-133)

20 (2-80)

22 (3-56)

ns

Adjuvant chemotherapy

58 (43)

21 (38)

25 (45)

12 (48)

ns

32 (58)

12 (48)

ns

TE D

LOHS (days) a Recurrence

67 (50)

23 (42)

median (range); b data were retrieved for 107 patients (44 AC, 45 BDC, and 18 DC); LOHS: Length Of

EP

a

ns

48 (87.3)

M AN U

Pancreatic duct dilation

(80.7)

ns

11 (46)

SC

Operative data

24 (17-36)

RI PT

Pre-illness BMI (kg/m²) a

AC C

Hospital Stay; ns: Not Statistically Significant

ACCEPTED MANUSCRIPT Table 2. Histological data, by tumor type

Variable

Total 135 (%)

Tumor size (cm) a

Ampullary 55 (%)

2 (1-7)

Differentiation (well)

76 (56)

2 (1-5) 34 (62)

Bile Duct 55 (%)

Duodenal 25 (%)

2 (1-5)

4 (1-7)

0.004

9 (36)

ns

33 (60)

p

13 (10)

1 (2)

10 (18)

2 (8)

0.014

43 (32)

20 (36)

14 (25)

9 (36)

ns

Neural invasion

55 (41) 15 (11)

15 (27) 7 (13)

35 (64) 4 (7)

T stage 1 T stage 2

41 (30)

24 (44)

15 (27)

T stage 3

59 (44)

21 (38)

28 (51)

T stage 4

20 (15)

3 (5)

8 (15)

Lymph node ratio

a

N stage (N1)

13 (2-43)

3 (1-28)

2 (1-28)

0.2 (0.1-1.0)

0.2 (0.1-0.9)

61 (45)

20 (36)

UICC stage 1

42 (31)

UICC stage 2

64 (47)

UICC stage 3

25 (19)

UICC stage 4

4 (3)

24 (44)

2 (8)

0.005

10 (40) 9 (36)

12 (3-23)

2.5 (1-12)

4 (1-15)

0.001

ns

0.002 ns

ns

0.2 (0.1-1.0)

0.3 (0.1-0.6)

ns

26 (48)

15 (53)

ns

13 (24)

5 (20)

0.032

32 (58)

5 (20)

0.006

3 (5)

7 (13)

15 (60)

<0.001

1 (2)

3 (5)

0 (0)

TE D EP

14 (2-39)

ns

27 (49)

median (range); ns: Not Statistically Significant

AC C

a

13 (2-43)

5 (20) 4 (16)

SC

Number of positive nodes

a

M AN U

Number of examined nodes a

RI PT

Positive resection margins Microangio invasion

ns

ACCEPTED MANUSCRIPT Table 3. Multivariate Cox regression analysis for predictors of recurrence, by tumor type

Variables

Duodenal cancers

Ampullary cancers

Bile duct cancers

5-y.

5-y.

5-y.

p-

p-

p-

p-

p-

p-

DFS univ multi HR (95% CI) DFS univ multi HR (95% CI) DFS univ multi HR (95% CI) -

(ref) 73 0.002 0.041 1.0 35 0.656 29 2.9 (1.0-8.4) 41

-

59 0.052 ns 22

-

-

(ref)

70 0.001 ns 19

51 0.011 ns 20

67 0.001 0.001 1.0 (ref) 76 16 7.2 (2.4-21.5) 0 24.0 (6.0-96.5) 41 0.051 ns 23

(ref) 70 0.001 0.001 1.0 41 0.049 27 2.6 (0.9-7.7) 33 0 35.6 (5.7-222) 10 27 0.001 40 0.013 ns 11 8

EP

AC C

a

29 0.467 42

M AN U

80 0.014 ns 61 38 21.4 (3.6-126) 1.0

RI PT

30 0.386 33

(ref) 1.0 52 0.502 15.5 (2.7-87.3) 53

-

SC

56 0.593 47

TE D

Age (years) ≤ 65 64 0.104 ns > 65 29 Loss of BMI < 4% 75 0.034 ns ≥ 4% 14 ASA 1-2 56 0.259 3 24 Biliary drainage No 53 0.046 0.002 Yes 25 T stage a T1 75 0.001 0.001 T2 50 T3 65 T4 0 N stage N0 69 0.110 ns N1 27 Positive nodes 0 69 0.166 1-3 37 ≥4 22 1-2 37 0.166 ns ≥3 22 LNR 0 69 0.056 ns 1-15 100 > 15 18 Resection margins R0 47 0.002 ns R1 0 Micro-angio invasion No 48 0.372 Yes 31 Perineural invasion No 49 0.084 0.049 Yes 20

70 0.001 ns 25 17

41 0.015 ns 44 8

53 0.027 ns 0

36 0.021 ns 12

64 0.012 ns 32

37 0.455 14

(ref) 1.0 57 0.191 ns 4.6 (1.0-21.1) 38

-

61 0.006 ns 21

analysis performed for T1–2–3 vs T4 in DC, T1–2 vs T3–4 in AC, and T1–2 vs T3 vs T4 in BDC; ns: Not

Statistically Significant.

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT HIGHLIGHTS •

Despite a common treatment, ampullary, bile duct, duodenal and pancreatic cancers have very different long-term prognoses; These differences reflect for each tumor the effects of specific factors of aggressiveness;

•

Predictors of recurrence have been already well described for pancreatic cancers, but lack reliability for

RI PT

•

non-ductal tumors;

The identification of reliable predictors of recurrence for non-ductal tumors is mandatory to select high-

EP

TE D

M AN U

SC

risk patients and adapt management.

AC C

•