Conference PaperPDF Available

Long-Term Outcomes of Periampullary Adenocarcinomas After Pancreaticoduodenectomy - A Peruvian Multi-Institutional Study

Authors:

Abstract and Figures

Background: Reports of periampullary adenocarcinomas are scarce in Latin-America. Pancreatic adenocarcinoma (PA) has the highest incidence in western countries, but also worse survival compared to distal bile duct (DBD), ampullary (AA) and duodenal (DA) adenocarcinomas, even after being resected by pancreaticoduodenectomy (PD). In this study, we analyze the survival of patients with resected periampullary adenocarcinomas and determined factors that influence this outcome. Methods: We retrospectively analyzed patients undergoing PD for periampullary adenocarcinoma from 2002 to 2011 in the three major pancreatic surgery centers in Peru. Postoperative mortalities (90-day) were excluded. Clinicopathologic features that may influence overall survival (OS) were collected and analyzed by Kaplan Meier method. Log-Rank test and Cox regression modelling were used for univariate and multivariate analysis, respectively. Results: 473 Periampullary adenocarcinomas (153 PA, 258 AA, 33 DBD and 29 DA) treated by PD were identified. 5-year OS was 34% with a median survival of 28 months for the entire cohort. PA was associated with the worst 5-year OS (11%) compared to DBD (24%), DA (41%) and AA (47%), respectively (p < 0.01). Predictors of poor survival were site of primary, age � 70 years, tumor size � 2.5 cm, high-grade tumors, nodal metastases, lymphovascular invasion, perineural invasion and positive resection margin. On multivariate analysis, site of primary and nodal metastases were the strongest predictors of survival.Conclusion: We report the oncologic outcomes of the largest multi-institutional series of resected periampullary adenocarcinomas in Peru and Latin-America. AA is the most frequently resected tumor and portent the best prognosis, while PA has the worst OS. Patient selection could account for differences with reports of developed countries.
Content may be subject to copyright.
long-term outcomes were compared for patients who
received gemcitabine versus either 5-FU or capecitabine.
Results: A total of 57 patients were included for analysis.
Thirty patients received gemcitabine, 23 received capecita-
bine, and 4 received 5-FU infusion. Full results are detailed
in Table 1. The 5- FU/capecitabine and gemcitabine groups
were comparable with regard to age, gender, and carbohy-
drate antigen 19-9 levels. There were no signicant differ-
ences in peri-operative complication or readmission rates.
There were more node positive resections in the 5FU/
capecitabine group (59 vs 20%, p = 0.006), while other
pathologic outcomes were comparable. Median disease-free
survival was 15.5 months in the 5-FU/capecitabine group
versus 14.3 months in the gemcitabine group (p = 0.61);
overall survival was 29.2 versus 26.5 months (p = 0.61).
Conclusion: Neoadjuvant chemoradiation with gemcita-
bine demonstrated an advantage over 5-FU or capecitabine
in BRPC, with more node negative resections. This
improvement in pathology did not translate to a survival
benet in this study.
P 36.
LONG-TERM OUTCOMES OF
PERIAMPULLARY
ADENOCARCINOMAS AFTER
PANCREATICODUODENECTOMY eA
PERUVIAN MULTI-INSTITUTIONAL
STUDY
D. Callacondo, J. L. Arenas, G. Reaño-Paredes,
J. De Vinatea de Cñ¡rdenas, H. Robledo,
J. Quesada-Olarte, J. Targarona, W. Robles, L. Barreda,
L. Taxa-Rojas, J. Celis, E. Ruiz and I. Chavez
* Corresponding author. David Callacondo, Hospital Nacional
Guillermo Almenara IrigoyenPeru
Background: Reports of periampullary adenocarcinomas
are scarce in Latin-America. Pancreatic adenocarcinoma
(PA) has the highest incidence in western countries, but
also worse survival compared to distal bile duct (DBD),
ampullary (AA) and duodenal (DA) adenocarcinomas,
even after being resected by pancreaticoduodenectomy
(PD). In this study, we analyze the survival of patients with
resected periampullary adenocarcinomas and determined
factors that inuence this outcome.
Methods: We retrospectively analyzed patients undergoing
PD for periampullary adenocarcinoma from 2002 to 2011
in the three major pancreatic surgery centers in Peru.
Postoperative mortalities (90-day) were excluded. Clini-
copathologic features that may inuence overall survival
(OS) were collected and analyzed by KaplaneMeier
method. Log-Rank test and Cox regression modelling were
used for univariate and multivariate analysis, respectively.
Results: 473 Periampullary adenocarcinomas (153 PA,
258 AA, 33 DBD and 29 DA) treated by PD were
identied. 5-year OS was 34% with a median survival of
28 months for the entire cohort. PA was associated with
the worst 5-year OS (11%) compared to DBD (24%), DA
(41%) and AA (47%), respectively (p <0.01). Predictors
of poor survival were site of primary, age 70 years,
tumor size 2.5 cm, high-grade tumors, nodal metasta-
ses, lymphovascular invasion, perineural invasion and
positive resection margin. On multivariate analysis, site
of primary and nodal metastases were the strongest pre-
dictors of survival.
Table 1
5-FU/
Capecitabine
Gemcitabine p-value
n[27 n [30
Demographics
Age (year) 66 68 0.52
Gender: male 15 (56) 19(63) 0.60
Body mass index
(kg/rn
2
)
25.4 26.2 0.67
CA 199 (U/mL)
Presenting 202 111.5 0.58
Post-NAT 86.5 61 0.99
Operative
Operative time (min) 437 366 0.012*
Estimated blood
loss (ml)
350 385 0.50
Vein resection 12 (44) 6 (20) 0.086
Arterial resection 2 (7) 2 (7) 0.99
Peri-Operative
POPF 0 1 (3) 0.99
Chyle leak 2 (7) 3 (10) 0.99
SSI or seroma 9 (33) 12 (40) 0.78
Clavien-Dindo
grade 3-5
6 (22) 6 (20) 0.99
Length of stay (day) 6 7 0.58
90-day readmission 4 (17) 9 (32) 0.34
Pathology
Tumor size (cm) 3.0 3.2 0.53
Treatment response 0.24
Complete or
marked
5 (19) 10 (33)
Moderate or poor 22 (81) 20 (67)
Margin negative 21 (78) 22 (76) 0.99
T stage 0.32
No residual tumor 0 4 (13)
T1 1 (4) 2 (7)
Table 1
(continued )
5-FU/
Capecitabine
Gemcitabine p-value
n[27 n [30
T2 5 (19) 3 (10)
T3 19 (70) 19 (63)
T4 2 (7) 2 (7)
N stage 0.006*
N0 11 (41) 24 (80)
N1 16 (59) 6 (20)
Adequate nodal
yield (12)
23 (85) 23 (77) 0.51
Disease free survival
(mo.)
15.5 14.3 0.61
Overall survival (mo.) 29.2 26.5 0.61
Reported as median or number (percent). POPF, post-operative
pancreatic stula; DGE, delayed gastric emptying; SSI, surgical
site infection.
HPB 2018, -(-), 1e54
ePoster Abstracts 9
Conclusion: We report the oncologic outcomes of the
largest multi-institutional series of resected periampullary
adenocarcinomas in Peru and Latin-America. AA is the
most frequently resected tumor and portent the best
prognosis, while PA has the worst OS. Patient selection
could account for differences with reports of developed
countries.
P 37.
DOES SURGICAL VOLUME TRUMP
DISPARITIES IN PREDICTING
OUTCOMES FOLLOWING
HEPATOPANCREATICOBILIARY (HPB)
SURGERY?
B. R. Harris, J. S. Cardinal, J. T. Stemple, L. Swain,
A. Thomay and S. Cho
* Corresponding author. Britney Harris, West Virginia
University, USA
Background: Research has shown that high volume hos-
pitals (HVH) are associated with improved outcomes
following HPB surgery. It has also been shown that health
care disparities can also contribute to outcomes following
HPB surgery. The purpose of this study is to examine
outcomes following HPB surgery in a HVH which services
a population of patients at signicant risk for health care
disparities.
Methods: A retrospective analysis of patients undergoing
HPB surgical procedure at a single high volume HPB
center from 2012 to 17 was performed. National Surgical
Quality Improvement Program (NSQIP) data on the cohort
of patients included in the univariate analysis were
analyzed to determine if gender, race, insurance status,
distance to nearest HVH, primary care physician (PCP)
access, rural versus urban population, and history of mental
illness were associated with outcome(s).
Results: One hundred eighty-two patients were included in
this study. There was no statistical difference in outcome(s)
in regards to the presence of leak rates, post-operative
Table 1
Univariate and multivariate analysis of predictors of OS in patients undergoing PD for periampullary adenocarcinomas
Variable Number of patients 5 years OS Univariate analysis Multivariate analysis
Hazzard ratio (95% CI) P Hazzard ratio (95% CI) P
Sex
Man 217 80(37%)
Woman 256 79 (31%) 1.1 (0.9 1.4) 0.304 ––
Age
<70 years 332 118 (36%)
70 years 141 41 (29%) 1.3 (1.11.7) 0.017 1.4 (1.11.8) 0.008
Tumor size
<2.5 cm 146 76 (41%)
2.5 cm 287 83 (29%) 1.5 (1.2 1.9) 0.001 1.1 (0.81.4) 0.549
Site of primary
Pancreas 153 17 (11%) Reference Reference
Distal Bile Duct 33 3 (24%) 0.6(0.40.9) <0.001 0.7 (0.4 1.0) 0.008
Duodenum 29 12 (41%) 0.3 (0.2 0.6) 0.008 0.6 (0.30.9) 0.036
Ampulla of Vater 258 122 (47%) 0.3 (0.20.4) <0.001 0.5 (0.40.7) <0.001
Nodal metastases
Negative 237 116 (49%)
Positive 236 43 (18%) 2.5 (2.03.1) <0.001 1.7 (1.3 2.2). <0.001
Degree of differentiation
G1 157 72 (46%)
G2 + G3 + G4 316 87 (23%) 1.8 (1.4 2.3) <0.001 1.4 (1.11.8) 0.016
Lymphovascular invasion
Negative 172 94 (55%)
Positive 301 65 (22%) 2.7 (2.13.4) <0.001 15.(1.12.1) 0.011
Invasionperineural
Negative 240 115 (48%)
Positive 233 44 (19%) 2.4 (1.93.0) <0.001 1.2 (0.9 1.6) 0.205
Resection margin
Negative 419 153 (37%)
Positive 54 6 (11%) 2.8 (2.4 3.9) <0.001 1.6 (1.11.3) 0.007
OS, overall survival; PD, pancreaticoduodenectomy; CI, condence interval; G1, well differentiated; G2 moderately differentiated; G3
poorly differentiated; G4, undifferentiated.
HPB 2018, -(-), 1e54
10 ePoster Abstracts
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.