Received: 16 April 2022 Revised: 11 December 2022 Accepted: 14 December 2022

DOI: 10.1002/cac2.12399

ORIGINAL ARTICLE

Standard pancreatoduodenectomy versus extended

pancreatoduodenectomy with modified retroperitoneal
nerve resection in patients with pancreatic head cancer: a
multicenter randomized controlled trial

Qing Lin1,2,3,# Shangyou Zheng1,2,3,# Xianjun Yu4,# Meifu Chen5,#

Yu Zhou1,# Quanbo Zhou2,3 Chonghui Hu1,2,3 Jing Gu6 Zhongdong Xu3,7
Lin Wang3,8 Yimin Liu3,9 Qingyu Liu3,10 Min Wang11 Guolin Li12
He Cheng4 Dongkai Zhou13 Guodong Liu14 Zhiqiang Fu2,3 Yu Long2,3
Yixiong Li14 Weilin Wang13,15,16 Renyi Qin11 Zhihua Li3,9 Rufu Chen1,2,3
1 Department
of Pancreas Center, Department of General Surgery, Guangdong Provincial People’s Hospital, Guangdong Academy of Medical Sciences,
Guangzhou, Guangdong, People’s Republic of China
2 Department of Pancreatobiliary Surgery, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, Guangdong, People’s Republic of

China
3 Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation Medical Research Center, Sun Yat-Sen Memorial
Hospital, Sun Yat-Sen University, Guangzhou, Guangdong, People’s Republic of China
4 Department of Pancreatic Surgery, Fudan University Shanghai Cancer Center, Shanghai, People’s Republic of China
5 HunanResearch Center of Biliary Disease/Department of Hepatobiliary Surgery, Hunan Provincial People’s Hospital/The First Affiliated Hospital of
Hunan Normal University, Changsha, Hunan, People’s Republic of China
6 Department of Medical Statistics, School of Public Health, Sun Yat-sen University, Guangzhou, Guangdong, People’s Republic of China
7 Department of Anesthesiology, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, Guangdong, People’s Republic of China
8 Department of Pathology, Sun Yat-sen Memorial Hospital, Guangzhou, Guangdong, People’s Republic of China
9 Department of Oncology, Sun Yat-sen Memorial Hospital, Sun Yat-sen University Guangzhou, Guangdong, People’s Republic of China
10 Department of Radiology, Sun Yat-sen Memorial Hospital, Sun Yat-sen University, Guangzhou, Guangdong, People’s Republic of China
11 Department
of Biliary-Pancreatic Surgery, Affiliated Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology,
Wuhan, Hubei, People’s Republic of China
12 Department of Hepatobiliary, Pancreatic and Splenic surgery, the Sixth Affiliated Hospital of Sun Yat-sen University, Guangzhou, Guangdong,

People’s Republic of China

Abbreviations: AG, albumin bound paclitaxel plus gemcitabine; AJCC, American Joint Committee on Cancer; ALT, alanine aminotransferase; AST,
aspartate aminotransferase; CA19–9, carbohydrate antigen 19–9; CEA, carcinoembryonic antigen; CSPAC, Chinese Study Group for Pancreatic Cancer;
CT, computed tomography; CI, confidence interval; Cr, creatinine; CTA, computed tomography angiography; DFS, disease-free survival; ECOG,
Eastern Cooperative Oncology Group; EPD, extended pancreatoduodenectomy; FOLFIRINOX, fluorouracil, leucovorin, irinotecan, and oxaliplatin;
HR, hazard ratio; ITT, intention-to-treat; LNs, lymph nodes; MIPD, minimally invasive pancreatoduodenectomy; MD, moderate differentiation; NRS,
numerical rating scale; NSAIDs, non-steroidal anti-inflammatory drugs; OPD, open pancreatoduodenectomy; PSM, propensity score matching; PNI,
Perineural invasion; PD, poor differentiation; PDAC, pancreatic adenocarcinoma; OS, overall survival; PV, portal vein; RCTs, randomized controlled
trials; RBC, red blood cell; RCS, restricted cubic splines; ROC, receiver operating characteristic; SF-36, short form 36; SD, standard deviation; SPD,
standard pancreatoduodenectomy; SPSS, Statistical Package For The Social Sciences; SMA, superior mesenteric artery; SMV, superior mesenteric vein;
TNM, tumor-node-metastasis; WD, well differentiation.
# These authors contributed equally to this work.

This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium,
provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
© 2022 The Authors. Cancer Communications published by John Wiley & Sons Australia, Ltd. on behalf of Sun Yat-sen University Cancer Center.

Cancer Communications. 2023;43:257–275. wileyonlinelibrary.com/journal/cac2 257

258 LIN et al.

13 Hepatobiliary
and Pancreatic Interventional Treatment Center, Division of Hepatobiliary and Pancreatic Surgery, The First Affiliated Hospital,
College of Medicine, Zhejiang University, Hangzhou, Zhejiang, People’s Republic of China
14 Department of General Surgery, Xiangya Hospital, Central South University, Changsha, Hunan, People’s Republic of China
15 Departmentof Hepatobiliary and Pancreatic Surgery, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang,
People’s Republic of China
16 Key
Laboratory of Precision Diagnosis and Treatment for Hepatobiliary and Pancreatic Tumor of Zhejiang Province, Hangzhou, Zhejiang, People’s
Republic of China

Correspondence
Rufu Chen, Department of General Abstract
Surgery, Guangdong Provincial People’s Background: The extent of pancreatoduodenectomy for pancreatic head can-
Hospital, Guangdong Academy of
Medical Sciences, Guangzhou, 510080, cer remains controversial, and more high-level clinical evidence is needed. This
Guangdong, People’s Republic of China. study aimed to evaluate the outcome of extended pancreatoduodenectomy (EPD)
E-mail: chenrufu@gdph.org.cn.
with retroperitoneal nerve resection in pancreatic head cancer.
Zhihua, Li, Department of Oncology, Sun Methods: This multicenter randomized trial was performed at 6 Chinese high-
Yat-sen Memorial Hospital, Sun Yat-sen volume hospitals that enrolled patients between October 3, 2012, and September
University Guangzhou, 510120,
21, 2017. Four hundred patients with stage I or II pancreatic head cancer
Guangdong, People’s Republic of China.
E-mail: lzhdoct@163.com; and without specific pancreatic cancer treatments (preoperative chemotherapy
or chemoradiation) within three months were randomly assigned to undergo
Renyi Qin, Department of
Biliary-Pancreatic Surgery, Affiliated
standard pancreatoduodenectomy (SPD) or EPD, with the latter followed by
Tongji Hospital, Tongji Medical College, dissection of additional lymph nodes (LNs), nerves and soft tissues 270◦ on
Huazhong University of Science and the right side surrounding the superior mesenteric artery and celiac axis. The
Technology, Wuhan, 430030, Hubei,
People’s Republic of China. primary endpoint was overall survival (OS) by intention-to-treat (ITT). The sec-
E-mail: ryqin@tjh.tjmu.edu.cn; ondary endpoints were disease-free survival (DFS), mortality, morbidity, and
postoperative pain intensity.
Weilin Wang, Department of
Hepatobiliary and Pancreatic Surgery, The Results: The R1 rate was slightly lower with EPD (8.46%) than with SPD (12.56%).
Second Affiliated Hospital, Zhejiang The morbidity and mortality rates were similar between the two groups. The
University School of Medicine,
median OS was similar in the EPD and SPD groups by ITT in the whole study
Hangzhou, 310009, Zhejiang, People’s
Republic of China. cohort (23.0 vs. 20.2 months, P = 0.100), while the median DFS was superior
E-mail: wam@zju.edu.cn; in the EPD group (16.1 vs. 13.2 months, P = 0.031). Patients with preoperative
Yixiong Li, Department of General
CA19–9 < 200.0 U/mL had significantly improved OS and DFS with EPD (EPD
Surgery, Xiangya Hospital, Central South vs. SPD, 30.8 vs. 20.9 months, P = 0.009; 23.4 vs. 13.5 months, P < 0.001). The EPD
University, Changsha, 410008, Hunan, group exhibited significantly lower locoregional (16.48% vs. 35.20%, P < 0.001)
People’s Republic of China.
E-mail: liyixiong2011@hotmail.com. and mesenteric LN recurrence rates (3.98% vs. 10.06%, P = 0.022). The EPD group
exhibited less back pain 6 months postoperation than the SPD group.
Registration number:
Conclusions: EPD for pancreatic head cancer did not significantly improve OS,
ChiCTR-TRC-12002548
(ClinicalTrials.gov, but patients with EPD treatment had significantly improved DFS. In the sub-
http://www.chictr.org.cn). group analysis, improvements in both OS and DFS in the EPD arm were observed
in patients with preoperative CA19–9 < 200.0 U/mL. EPD could be used as an
Funding information
Sun Yat-sen University Clinical Research effective surgical procedure for patients with pancreatic head cancer, especially
5010 Program, Grant/Award Number: those with preoperative CA19–9 < 200.0 U/mL.
2012007; National Natural Science
Foundation of China, Grant/Award
KEYWORDS
Number: 81871945; National Key Clinical
disease-free survival, extended, lymph nodes, nerve resection, overall survival, pancreatic head
Specialty Construction Project,
cancer, pancreatoduodenectomy, standard
Grant/Award Number: 2022YW030009
LIN et al. 259

1 INTRODUCTION

Pancreatic cancer is a lethal disease, with a 5-year overall 2.2 Patients and study design
survival (OS) rate of approximately 10% [1]. Concerning the
localized disease status, radical resection remains a poten- This multicenter, randomized, controlled, nonblinded,
tially curable treatment option for patients with clinical parallel-group trial compared SPD versus EPD with
stage I to II pancreatic cancer. However, the high risk of retroperitoneal nerve resection for pancreatic head cancer.
local and/or distant recurrence for most resectable diseases The protocol complied with the Declaration of Helsinki
makes surgeons dismayed by the value of surgery. In recent and was approved and overseen by the clinical ethics
decades, several randomized controlled trials (RCTs) on committee of each participating hospital. This study was
the surgical management of pancreatic cancer have been registered at the Chinese Clinical Trial Registry (http://
performed by comparing the outcomes of standard pancre- www.chictr.org.cn, no: ChiCTR-TRC-12002548). Patients
aticoduodenectomy (SPD) to those of extended pancreati- with resectable pancreatic head tumors eligible to undergo
coduodenectomy (EPD). The results showed that, despite pancreatoduodenectomy were enrolled between October
its theoretical advantages, EPD does not have survival 3, 2012, and September 21, 2017. All listed authors had
advantages over standard Whipple surgery [2–6]. Never- access to the study data and approved the final manuscript.
theless, some limitations have made these RCTs uncon- The study sponsors had no role in the design and conduct
vincing, such as the small number of enrolled patients, lack of the study.
of intention-to-treat (ITT) analyses, differential definitions
of standard and extended lymphadenectomy, different sur-
gical procedures and techniques among surgeons and 2.3 Inclusion and exclusion criteria
institutions, discrepancies in adjuvant treatment, differ-
ential definitions of outcome parameters and complica- Patients were included if they (1) were 18 to 80 years of age
tions, and inclusion of patients with nonpancreatic ductal (the upper limit age was changed from 70 to 80 years due to
adenocarcinoma. an increase in the number of adults 70 years and older, and
Perineural invasion (PNI) is a pathohistological hall- pancreatic cancer tends to occur at an older age), regard-
mark of pancreatic cancer and serves as an alternative less of gender; (2) had potentially curable cancer of the
route for dissemination in addition to the lymphatic and pancreatic head (stage I or II according to the American
vascular systems [7]. However, a consensus on the extent Joint Committee on Cancer [AJCC] 7th edition), as shown
of nerve plexus dissection in pancreatoduodenectomy has on preoperative imaging examinations (enhanced com-
not been reached. puted tomography, magnetic resonance imaging/magnetic
Due to the ominous features of PNI in pancreatic resonance cholangiopancreatography, endoscopic ultra-
ductal adenocarcinoma (PDAC), the lack of a surgical sound, endoscopic retrograde cholangiopancreatography,
consensus and limited results from previous RCTs, we positron emission tomography-computed tomography or
performed this randomized controlled study to compare fine-needle aspiration biopsy); (3) had a Karnofsky per-
the outcomes between SPD and EPD with retroperi- formance status score > 70; (4) had Loyer grade A to
toneal nerve resection in patients with pancreatic head D type; (5) had no obvious surgical contraindications;
cancer. and (6) provided written informed consent. Patients were
excluded if they (1) had an unresectable condition or
metastasis (stage III or IV) found during surgery; (2) had
2 METHODS a pathologic diagnosis of a benign tumor of the head
of the pancreas or a tumor in the tail of the pancreas;
2.1 Participating hospitals (3) had surgical contraindications; (4) had a history of
other malignancies; (5) were pregnant (urine human
Initially, 3 hospitals agreed to participate in this study. chorionic gonadotropin (HCG) > 2500 IU/L, diagnosed
Then, 3 other centers showed interest in participating and as early pregnancy), planned to become pregnant or were
joined the study one year later. Ultimately, patients were lactating; (6) had received other specific pancreatic can-
enrolled from 6 tertiary hospitals (all are high-volume cen- cer treatments (preoperative chemotherapy or chemora-
ters in pancreatic cancer diagnosis and treatment in China, diation) within three months; (7) had mental disease;
Supplementary Table S1). (8) participated in other clinical trials 3 months before; or
260 LIN et al.

(9) had impaired visceral function (cardiac function 3-4, by independent expert evaluation was performed; (3) 6
alanine aminotransferase[ALT] and/or aspartate amino- meetings (April 15, May 20, June 24, July 15, August 12
transferase [AST] exceeding 3 times the upper limit, and September 15, 2012) were held to discuss the surgi-
creatinine [Cr] beyond the upper limit). cal videos and unify the surgical procedures before patient
enrollment; and (4) all randomized cases were evalu-
ated for heterogeneity through objective data, such as
2.4 Study treatment surgical images (Supplementary Figure S1) and videos,
by third-party experts to ensure homogeneity of patient
In the SPD group, lymph nodes (LNs) around the gastric enrollment.
pylorus (stations 5, 6), LNs around the pancreatic head
(stations 13a, 13b, 17a, and 17b), LNs anterior to the com-
mon hepatic artery (station 8a), LNs on the right side of 2.5 Randomization
the hepatoduodenal ligament (stations 12b, 12c), and LNs
on the right side of the root of the superior mesenteric Patients were randomly assigned at a 1:1 ratio to undergo
artery (SMA) (stations 14a, 14b) were dissected without SPD or EPD with retroperitoneal nerve resection using
retroperitoneal nerve resection. During EPD, the follow- a stratified permuted block method. Patients were strati-
ing nerve tissues at the retroperitoneum and LNs around fied according to the coordinating hospitals and predefined
the pancreatic head were dissected: (I) nerves and soft preoperative cancer antigen 19-9 (CA19-9) (< 200.0 U/mL
tissues between the inferior vena cava (including the aor- or ≥ 200.0 U/mL). Computer-generated random assign-
tic plexus) and abdominal aorta (including LN stations ment lists were created at the School of Public Health,
16a2 and 16b1); (II) all the nerves and soft tissues around Sun Yat-sen University (Guangzhou, Guangdong, China).
the hepatoduodenal ligament, which were completely dis- The assignments were placed in sealed envelopes, labeled
sected and skeletonized (including whole LN station 12); by stratum, and unsealed only after exploratory laparo-
(III) the common hepatic artery, which was isolated, and tomy. Investigators at each center enrolled participants and
its surrounding nerves and soft tissues (including LN sta- assigned them to their corresponding interventions. Each
tions 8a and 8p); (IV) the celiac trunk, which was isolated, patient received a unique study number that remained
and its nerves and soft tissues 270◦ around the right longi- unchanged throughout the trial.
tudinal axis (including LN station 9); (V) the root of the
SMA, which was dissected to open the vascular sheath,
and the proximal uncinate process mesentery, which was 2.6 Study outcomes
removed along its right side, the nerves and soft tissues
270◦ around the right longitudinal axis (including LN sta- The primary endpoint of this study was OS by ITT, cal-
tions 14a and 14b); and (VI) the nerves and soft tissues culated from the date of randomization until the date
in the dense postpancreatic connective tissues that fixed of death from any cause. The secondary endpoints were
the pancreas at the celiac trunk-aorta-SMA artery axis. disease-free survival (DFS), mortality, morbidity, and post-
Both the portal vein (PV) and the superior mesenteric operative pain intensity. DFS was calculated from the date
vein (SMV) were resected and reconstructed in patients of randomization until the date of the first event of either
with PV/SMV involvement to achieve radical resection. recurrence or death from any cause. The censored data
Differences in the extent of resection are summarized in were defined as patients without events at study termi-
Supplementary Table S2 and Figure 1. nation or those lost to follow-up at any time during the
First, all participating hospitals were high-volume pan- study. Prespecified subgroups were defined for preopera-
creatic surgery centers (completing more than 100 pan- tive CA19–9 (< 200.0 U/mL or ≥ 200.0 U/mL based on
creatoduodenectomy operations each year), and all the the predictive value of CA19–9 from previous reports [9–11]
surgeons participating in this trial had rich surgical and discussions of the researchers), and we performed
experience (the cumulative number of pancreatoduo- subgroup analyses with this cutoff for OS and DFS. More-
denectomies exceeded 300 surgeries). Second, for the over, we performed post hoc subgroup analyses of other
homogenization of the surgical technique, the Pancreatic factors, including age (< 60/≥ 60), gender (male/female),
Cancer Committee of the Chinese Anti-Cancer Associ- preoperative carcinoembryonic antigen (CEA) (< 5.00/≥
ation conducted multiple unified training sessions and 5.00 ng/mL), PV/SMV resection (yes/no), type of resection
assessments on the operators before the start of the (R0/R1), N stage (N0/N1), tumor-node-metastasis (TNM)
study [8]. Specifically, (1) unified surgical specifications stage (I/IIA/IIB), histology (well differentiation [WD] and
and procedures were formulated; (2) strict assessment of moderate differentiation [MD]/poor differentiation [PD]),
EPD and SPD surgical proficiency from each operator perineural invasion (+/-) and adjuvant treatment (yes/no),
LIN et al. 261

FIGURE 1 Schematic diagram of the dissection of surgery. Abbreviations: LN, lymph node.

for OS and DFS in the ITT population. Pain intensity 2.7 Pathological analysis
was assessed with a numerical rating scale (NRS) [12].
An 11-point scale from 0 to 10 was used to describe pain The LNs and nerve tissues removed during dissection were
(minimum to maximum). The NRS was explained to the marked with the exact location and sent for a pathological
patients, and they were asked to circle the number best examination after resection. Pathology reports contained
representing the degree of their pain on the last day of pre- the primary pathologic diagnosis, the extent of disease,
operation (baseline, day 0), 1 month postoperation, and 6 margin status, LN status and overall pathological stage as
months postoperation. The patients were followed up once previously described. The stages of the resected specimens
a month from 1 to 6 months after surgery, every 3 months were classified according to the AJCC TNM classification
from 6 months to 3 years after surgery, and every 6 months (7th edition). All participating surgeons observed intraop-
for 3 years after surgery. The follow-up examinations erative photographs or videos of both groups to ensure
included routine blood tests, biochemical tests, digestive consistency of the extent of nerve, tissue and LN dissection.
tract tumor index detection, B-ultrasound, and enhanced
computed tomography (CT) scan examinations. The short
form 36 (SF-36) was used to assess the patients’ general 2.8 Postoperative chemotherapy
health status. The SF-36 is composed of eight multi-item
scales (physical functioning, physical role, pain, general Chemotherapy was recommended for all patients except
health, vitality, social function, emotional role, and mental for those with a poor performance status (grade 3-5
health), with scores for each of these scales (or dimen- by ECOG [Eastern Cooperative Oncology Group] per-
sions) ranging from 0 to 100, whereby a higher score formance status scale) [14] or organ dysfunction and
indicates a higher health-related quality of life [13]. The those who refused adjuvant treatment. The first postop-
patients completed the SF-36 questionnaire with the assis- erative chemotherapy cycle started within 2 months of
tance of the follower. The SF-36 score and pain score of the surgery. The chemotherapy regimen was as follows: gem-
patients were assessed preoperatively and every 3 months citabine 1000 mg/m2 , dissolved in 100 mL normal saline,
after surgery. Each follow-up included gathering infor- intravenous infusion, completed within 30 minutes; once
mation on the chemotherapy regimen, adjuvant therapy weekly, after 2 consecutive times, rest for 1 week, every 3
regimen and adverse reactions. The last date of follow-up weeks for a course of treatment, for 8 consecutive courses.
was defined as three calendar years since the last enrolled
patient underwent surgery.
An interim analysis was conducted on the primary 2.9 Postoperative pain control strategy
endpoint when 50% of patients were randomized and com-
pleted the 6-month follow-up. The interim analysis was In this study, for postoperative pain control, all our
performed by an independent statistician blinded to the patients followed the principle of pain ladder treatment.
treatment allocation. The trial was calculated to end using For mild pain, we recommended using nonopioid adju-
symmetric stopping boundaries at P < 0.001. vant analgesics (ibuprofen, aspirin). For moderate pain, we
262 LIN et al.

recommended using weak opioids plus or minus nons- progression survival, a multivariable Cox model with
teroidal anti-inflammatory drugs (NSAIDs) plus or minus restricted cubic splines (RCS) was built. RCS has been
adjuvant pain relievers (tramadol, celebrex, codeine, etc.). widely described as a valid strategy for analyzing the
For severe pain, we recommended using opioids plus relationship between survival and independent variables
or minus nonsteroidal adjuvant painkillers (morphine, [17, 18]. The analysis was performed with R software
duloxetine, etc.) version 4.2.1. To minimize the effect of confounding
factors and potential bias between the SPD and EPD
groups, the propensity score was calculated using logistic
2.10 Sample size and statistical analysis regression, and we performed 1:1 patient matching by the
nearest-neighbor matching method without replacement.
Based on previous RCTs [3, 6, 15] and our pre-experiment We used a caliper radius equal to 0.2 of the standard devi-
analysis, our trial was powered for the superiority of OS ation to prevent poor matching. The variables included
data at 3 years according to the operation, assuming that in the matching model were age, gender, preoperative
the 3-year OS rate (35%) of patients from the EPD group CA19–9 value, PV/SMV resection, N stage and T stage.
was 15% higher than that of patients from the SPD group All statistical analyses were performed using Statistical
(20%). Our estimation showed that an enrollment of 180 Package For The Social Sciences (SPSS) version 23.0 (SPSS
patients would provide 80% power to detect the superiority Inc., Chicago, IL, USA), and a 2-sided P value less than
of a procedure with a 1-sided α = 0.05 and β = 0.2. Tak- 0.05 was considered statistically significant.
ing into account an estimated 10% dropout rate, the sample
size was increased to 200 participants per group (n = 400).
The ITT population was defined as all randomly 2.11 Data management
assigned patients. The qualified population was defined
as patients enrolled in randomization who underwent After random assignment, the following data were col-
appropriate surgery without rule violation for the extent lected from all patients: clinical and pathological infor-
of surgical dissection, with ductal adenocarcinoma as the mation, details of the operative procedure (including
final pathology, with a complete case report form and with- photographs of the operation field and a surgeon question-
out loss to follow-up. Analyses were conducted using the naire detailing the operative findings), and other relevant
ITT principle, irrespective of any protocol deviations or information. Adverse events were graded according to
violations. the Common Terminology Criteria for Adverse Events
Patients who underwent the operation to which they (CTCAE, version 4.0). Follow-up was obtained from hos-
were originally allocated and satisfied the criteria for pitalization or outpatient records, telephone calls, and
optimal surgery based on photographs uploaded to our assistance from the Chinese public security administra-
data center were evaluated. Categorical variables are tion. Assessors were blinded to the treatment groups (i.e.,
expressed as proportions, whereas continuous vari- the SPD group vs. the EPD group).
ables are expressed as the medians (range, minimum
to maximum) or means (standard deviation, SD) where
appropriate. Missing data were handled using a Markov 3 RESULTS
chain Monte Carlo multiple imputation approach with
the assumption of missing data at random [16]. Contin- 3.1 Patients
uous variables were reported as the median and range,
where appropriate, and compared using Student’s t-test Between October 3, 2012, and September 21, 2017, 468
(when the data conformed to a normal distribution) or patients from six Chinese centers were screened (Figure
Mann‒Whitney U test (when the data did not conform to a 2). After enrollment, 68 patients were excluded for
normal distribution). Nominal data were compared using the following reasons: 28 patients refused to participate
χ2 tests or Fisher’s exact test. Survival outcomes were despite initially agreeing, 32 patients had unresectable or
calculated using the Kaplan‒Meier method and compared metastatic tumors during the intraoperative exploration,
using the log-rank test (stratified for predefined preopera- and 8 patients actually received distal pancreatectomy.
tive CA19–9 [< 200.0 U/mL or ≥ 200.0 U/mL]). Variables Thus, 400 eligible patients were enrolled and randomly
revealed as statistically significant by the univariate anal- allocated into two groups (Figure 2). The last follow-up
ysis were included in the multivariate analysis, which was date was expected to be on September 21, 2020, but to fur-
performed using a Cox proportional hazards regression ther improve the integrity of the data, we extended the
model. Both univariate and multivariate analyses were follow-up time to November 30, 2020.
performed for the treatment method. To evaluate the Of the remaining 400 patients, 199 were randomized to
association between levels of CA19–9 and disease- receive SPD, 201 were randomized to receive EPD with
LIN et al. 263

no differences between the two groups in tumor size, T

stage or AJCC TNM stage. As expected, the total num-
ber of retrieved LNs was significantly higher in the EPD
group (median [range], 20 [7–63]) than in the SPD group
(median [range], 15 [7–46]; P < 0.001). Moreover, a sig-
nificantly higher number of positive LNs was observed in
the EPD group than in the SPD group (P = 0.005), but the
LN (+) ratio was not significantly different between the
two groups (P = 0.101).

3.3 Morbidity and mortality

Although there was a trend for higher morbidity in the

EPD group, there was no significant difference between
the two groups (EPD vs. SPD, 37.81% vs. 34.17%; P = 0.467;
Table 2). Moreover, no significant difference was found
between the two groups in postoperative hospital stay
(EPD vs. SPD, median [range], 18 [6-105] vs. 18 [9-49], P =
F I G U R E 2 Study flowchart illustrating patients’
randomization and group allocation for treatments with SPD and
0.232; Supplementary Table S3). The incidence of diarrhea
EPD. in the EPD group within 3 months of surgery was 7.46%
Abbreviations: SPD, standard pancreatoduodenectomy; EPD, compared with 5.03% in the SPD group; no significant dif-
extended pancreatoduodenectomy; CRF, case report form. ference was found (P = 0.409), suggesting that dissection
of the retroperitoneal plexus 270◦ on the right side sur-
retroperitoneal nerve resection and constituted the ITT rounding the celiac trunk and SMA might not significantly
population. Of them, 21 were excluded for the following affect bowel movement. The inpatient mortality was simi-
reasons: (1) having undergone inadequate or inappropriate lar between the two groups (EPD vs. SPD, 0.50% vs. 0.50%,
surgery, as revealed by reviewing the images of the surgi- Table 2). Regarding the specific cause of death, 1 patient
cal field (n = 6); (2) having a nonductal adenocarcinoma in the EPD group died of severe sepsis with gastroduode-
(n = 2), such as ampullary adenocarcinoma and duode- nal artery rupture, and 1 patient in the SPD group died of
nal adenocarcinoma; or (3) having inadequate case report cirrhotic liver failure.
forms (n = 9) or withdrawing consent (n = 4, two by the
patients and two by the surgeons). Consequently, 379 qual-
ified participants were included, with 187 in the SPD group 3.4 Survival and recurrence data
and 192 in the EPD group. The two groups were well bal-
anced in the following characteristics at baseline: median For the ITT analysis, all randomly enrolled patients were
age, gender, preoperative CEA and CA19–9 levels, preop- included. Kaplan‒Meier analysis of OS showed a better
erative total bilirubin, and resection rate of the PV/SMV prognostic trend for the EPD group than for the SPD group,
(Table 1). The operation time was 45 minutes longer in but the difference was not significant (Figure 3A). The
the EPD group than in the SPD group (P < 0.001), with median OS (ITT population) was 23.0 months in the EPD
no difference in either estimated blood loss (EPD vs. SPD, group versus 20.2 months in the SPD group (hazard ratio
median [range, mL], 300 [20–2,500] vs. 300 [20–6,000], (HR), 0.84; 95% CI, 0.68 to 1.04; P = 0.100; Table 2). The OS
P = 0.113; Table 1) or red blood cell (RBC) transfusion (P = rates in the EPD and SPD groups were 76.62% and 70.35%
0.870; Table 1). Overall, 343 patients underwent adjuvant at 1 year, 45.77% and 40.20% at 2 years, 29.35% and 22.61% at
treatment after surgical resection (Table 1). 3 years, and 9.95% and 9.05% at 5 years, respectively (Table
2). DFS was significantly better in the EPD group by ITT
(EPD vs. SPD, 16.1 months vs. 13.2 months; HR, 0.80; 95%
3.2 Postoperative pathological CI, 0.66 to 0.98; P = 0.031; Figure 3B and Table 2).
differences between groups

Table 1 also shows the pathological characteristics of the 3.5 Subgroup analysis
two groups. The percentage of patients who underwent
R1 resection was similar in the SPD group (12.56%) and The predefined subgroup with a preoperative CA19–9 <
the EPD group (8.46%; P = 0.195). In addition, there were 200.0 U/mL showed a significantly improved median OS
264 LIN et al.

TA B L E 1 Demographic and pathologic findings between two groups (ITT population)

Clinical variables SPD group EPD group P value
Age, median (range), year 60 (28-79) 59 (19-79) 0.916
Gender, male/female, male% 111/88, 55.78% 117/84, 58.21% 0.686
Preoperative CEA, median 3.37 (0.33-1256.00) 3.30 (0.20-422.40) 0.659
(range), ng/mL
Preoperative CA19–9, median 187.2 (0.6-12000.0) 180.5 (0.6-15968.0) 0.769
(range), U/mL
Preoperative total bilirubin, 91.8 (4.4-599.2) 94.7 (0.9-637.4) 0.515
median (range), μmol/L
OP time, median (range), min 330 (180-1031) 375 (240-625) < 0.001
EBL, median (range), mL 300 (20-6000) 300 (20-2500) 0.113
Transfusion (RBC pack), 0 (0-25.5) 0 (0-8.0) 0.870
median (range), unit
PV/SMV resection, cases (%) 44 (22.11) 41 (20.40) 0.715
Pathologic variables
R1 resection, cases (%) 25 (12.56) 17 (8.46) 0.195
Histology differentiation, 0.109
cases (%)
Well 8 (4.02) 14 (6.97)
Moderate 94 (47.24) 108 (53.73)
Poor 97 (48.74) 79 (39.30)
Tumor size, median (range), 3.2 (0.8-10.0) 3.0 (0.3-8.0) 0.400
cm
T stage, cases (%) 0.762
T1 7 (3.52) 10 (4.98)
T2 21 (10.55) 20 (9.95)
T3 171 (85.93) 171 (85.07)
LN (+), cases (%) 105 (52.76) 123 (61.19) 0.106
Total retrieved LNs, median 15 (7-46) 20 (7-63) < 0.001
(range)
No. positive LNs, median 1 (0-11) 1 (0-29) 0.005
(range)
LN (+) ratio, median (range) 0.06 (0-0.75) 0.07 (0-0.91) 0.101
TNM stage*, cases (%) 0.106
IA 4 (2.01) 7 (3.48)
IB 9 (4.52) 12 (5.97)
IIA 81 (40.70) 59 (29.35)
IIB 105 (52.76) 123 (61.19)
Adjuvant treatment, cases (%) 173 (86.93) 170 (84.58) 0.568
Complete adjuvant treatment, 130 (65.33) 125 (62.19) 0.534
cases (%)
*The stages of the resected specimens were classified according to the AJCC TNM classification (7th edition), I/IIB and IIA/IIB in the TNM stage are equivalent
to N0/N1, so multivariate analysis is not included.
Abbreviations: ITT, intention-to-treat; SPD, standard pancreatoduodenectomy; EPD, extended pancreatoduodenectomy; CEA, carcinoembryonic antigen; CA19–9,
carbohydrate antigen 19–9; OP, operation; EBL, estimated blood loss; RBC indicates red blood cells; PV, portal vein; SMV, superior mesenteric vein; LN, lymph node;
pancre1atoduodenectomy; AJCC, American Joint Committee on Cancer.
LIN et al. 265

F I G U R E 3 The outcome of the SPD and EPD groups in the ITT population. (A) OS and (B) DFS in the ITT population. (C) OS and (D)
DFS in the ITT population for the prespecified subgroup of preoperative CA19–9 < 200.0 U/mL. (E) OS and (F) DFS in the ITT population for
the prespecified subgroup of preoperative CA19–9 ≥ 200.0 U/mL.
Abbreviations: SPD, standard pancreatoduodenectomy; EPD, extended pancreatoduodenectomy; OS, overall survival; DFS, disease-free
survival; ITT, intention-to-treat; CA19–9, carbohydrate antigen 19–9; HR, hazard ratio; CI, confidence interval.
266 LIN et al.

TA B L E 2 ITT analysis of primary and secondary end points for both treatment groups
Outcome SPD(n = 199) EPD(n = 201) HR (95% CI) OR (95% CI) P value
Primary
Median OS, months 20.2 23.0 0.84 (0.68 to 1.04) N/A 0.100
Secondary
Median DFS, months 13.2 16.1 0.80 (0.66 to 0.98) N/A 0.031
OS rate, %
At 1 year 70.35 76.62 N/A 0.72 (0.46 to 1.12) 0.174
At 2 year 40.20 45.77 N/A 0.80 (0.54 to 1.17) 0.268
At 3 year 22.61 29.35 N/A 0.70 (0.44 to 1.10) 0.139
At 5 year 9.05 9.95 N/A 0.90 (0.46 to 1.73) 0.865
Safety, n (%)
Morbidity 68 (34.17) 76 (37.81) N/A 1.17 (0.78 to 1.78) 0.467
Mortality 1 (0.50)* 1 (0.50)** N/A 0.99 (0.05 to 18.90) NS
Mean (SD)***
Abdominal pain upper -2.05 (1.90) -1.77 (2.09) N/A 0.28 (-0.11 to 0.67) 0.163
changing in 1 month from
baseline****
Back pain changing in 1 -0.77 (1.71) -0.77 (1.72) N/A 0.02 (-0.33 to 0.37) 0.987
month from baseline
Abdominal pain upper -0.90 (3.12) -0.70 (2.97) N/A 0.20 (-0.43 to 0.82) 0.533
changing in 6 months from
baseline*****
Back pain changing in 6 0.48 (2.83) -0.35 (2.43) N/A -0.82 (-1.37 to -0.29) 0.003
months from baseline
*One died of cirrhotic liver failure.
**One died of severe sepsis with gastroduodenal artery rupture.
***Data of pain intensity were shown as mean (standard deviation).
****The data was calculated as pain intensity in 1 month postoperation minus pain intensity baseline. The baseline representsthe last day of preoperation.
*****The data was calculated as pain intensity in 6 months postoperation minus pain intensity baseline.
Abbreviations: ITT, intention-to-treat; SPD, standard pancreatoduodenectomy; EPD, extended pancreatoduodenectomy; HR, hazard ratio; CI, confidence interval;
OS, overall survival; DFS, disease-free survival; OR, odds ratio; NS, not significant; NRS, numerical rating scale; SD, standard deviation.

and DFS for EPD (EPD vs. SPD; OS: 30.8 months vs. 20.9 terms of median OS and DFS (OS: HR, 0.72; 95% CI, 0.54
months; HR, 0.68; 95% CI, 0.50 to 0.92; P = 0.009; DFS: to 0.96, P = 0.020, Supplementary Table S4; DFS: HR,
23.4 months vs. 13.5 months; HR, 0.62; 95% CI, 0.47 to 0.69; 95% CI, 0.52 to 0.91, P = 0.006, Supplementary Table
0.83; P < 0.001, Figure 3C-D and Table 3). In the subgroup S5). In the subgroup analysis (Supplementary Table S5),
with a preoperative CA19–9 < 200.0 U/mL, the respec- with the results of some prognostic factors presented in
tive 1-, 2-, 3-, and 5-year OS rates were 82.24%, 59.81%, Supplementary Figure S2 and Supplementary Figure S3,
40.19% and 16.82% in the EPD arm and 72.55%, 42.16%, survival benefits for DFS were found from EPD treatment
26.47%, and 10.78% in the SPD arm (Table 3). The prede- in patients with preoperative CEA < 5.00 ng/mL (HR, 0.75;
fined subgroup of patients with a preoperative CA19–9 ≥ 95% CI, 0.59 to 0.96, P = 0.018, Supplementary Figure S2B),
200.0 U/mL showed no significant difference in OS or DFS without PV/SMV resection (HR, 0.79; 95% CI, 0.63 to 0.99,
(Figure 3E-F and Table 4). Post hoc analysis of subgroups P = 0.038, Supplementary Figure S2F), with a resectable
by age (< 60/≥ 60), gender (male/female), PV/SMV resec- stage (HR, 0.74; 95% CI, 0.58 to 0.94, P = 0.013, Supplemen-
tion (yes/no), preoperative CEA (< 5.00/≥ 5.00 ng/mL), tary Figure S3H), with R0 resection (HR, 0.77; 95% CI, 0.62
resectability (resectable/borderline resectable), type of to 0.95, P = 0.014, Supplementary Figure S2H), and with
resection (R0/R1), N stage (N0/N1), TNM stage (I/IIA/IIB), postoperative adjuvant chemotherapy (HR, 0.77; 95% CI,
perineural invasion (positive/negative) and postoperative 0.62 to 0.96, P = 0.019, Supplementary Figure S2J).
adjuvant treatment (with/without) showed no differences As mentioned in the methods section, the prespecified
in OS (Supplementary Table S4). In addition, in the sub- threshold of the subgroup for CA19–9 was 200.0 U/mL.
group with well and moderate pathological differentiation, The rationale for this threshold was examined through
the EPD group showed superior oncologic outcomes in receiver operating characteristic (ROC) and RCS analyses.
LIN et al. 267

TA B L E 3 ITT analysis of primary and secondary endpoints for predefined subgroups preoperative CA19–9 level < 200.0 U/mL
Outcome SPD(n = 102) EPD(n = 107) HR (95% CI) OR (95% CI) P value
Primary
Median OS, months 20.9 30.8 0.68 (0.50 to 0.92) N/A 0.009
Secondary
Median DFS, months 13.5 23.4 0.62 (0.47 to 0.83) N/A <0.001
OS rate, %
At 1 year 72.55 82.24 N/A 0.57 (0.30 to 1.11) 0.100
At 2 year 42.16 59.81 N/A 0.49 (0.28 to 0.84) 0.013
At 3 year 26.47 40.19 N/A 0.54 (0.29 to 0.94) 0.041
At 5 year 10.78 16.82 N/A 0.60 (0.28 to 1.35) 0.234
Safety, n (%)
Morbidity 38 (37.25) 34 (31.78) N/A 0.78 (0.45 to 1.36) 0.467
Mean (SD)*
Abdominal pain upper -2.11(1.80) -1.78(2.01) N/A 0.33 (-0.19 to 0.85) 0.210
changing in 1 month from
baseline**
Back pain changing in 1 -0.63(1.56) -0.71(1.61) N/A -0.07 (-0.51 to 0.36) 0.740
month from baseline
Abdominal pain upper -0.89(3.09) -1.01(2.74) N/A -0.11 (-0.93 to 0.71) 0.782
changing in 6 months from
baseline***
Back pain changing in 6 0.59(2.93) -0.29(2.32) N/A -0.89 (-1.63 to -0.14) 0.020
months from baseline
*Data of pain intensity were shown as mean (standard deviation).
**The data was calculated as pain intensity in 1 month postoperation minus pain intensity baseline.
***The data was calculated as pain intensity in 6 months postoperation minus pain intensity baseline.
Abbreviations: ITT, intention-to-treat; SPD, standard pancreatoduodenectomy; EPD, extended pancreatoduodenectomy; CA19–9, carbohydrate antigen 19–9; HR,
hazard ratio; CI, confidence interval; OS, overall survival; DFS, disease-free survival; OR, odds ratio; NS, not significant; NRS, numerical rating scale; SD, standard
deviation.

The ROC analysis identified that the best cutoff value of levels could affect the benefits of adjuvant chemother-
preoperative CA19–9 was 198.7 U/mL for predicting tumor apy. In the subgroup with preoperative CA19–9 < 200.0
DFS, a value very close to 200.0 U/mL (Figure 4A-B); in U/mL, a significant improvement in median OS and DFS
addition, as shown in the RCS model (Figure 4C), the HR was found in the EPD group in patients with adjuvant
values for DFS in cases with CA19–9 ≥ 200.0 U/mL were chemotherapy (Supplementary Table S7). However, in
consistently greater than those for cases with CA19–9 < the subgroup with preoperative CA19–9 ≥ 200.0 U/mL,
200.0 U/mL, both supporting the prespecified threshold of EPD showed no advantage regardless of whether patients
200.0 U/mL. received adjuvant chemotherapy (Supplementary Table
To evaluate the benefit of adjuvant chemotherapy S8). In addition, we conducted a 1:1 propensity match-
according to different CA19–9 levels, we compared the ing analysis and further performed survival and subgroup
prognosis of patients stratified by adjuvant chemotherapy analyses for the treatments after matching. Consistent with
application in the patients with CA19–9 < 200.0 U/mL our aforementioned findings, no difference was found in
and those with CA19–9 ≥ 200.0 U/mL. In the subgroup the propensity score matching (PSM) population between
with preoperative CA19–9 < 200.0 U/mL, there was no the treatment groups in both OS and DFS (Supplemen-
difference in OS and DFS between those receiving or not tary Table S9), while patients with predefined preoperative
receiving adjuvant chemotherapy. In contrast, in the sub- CA19–9 < 200.0 U/mL showed a better outcome in the
group with preoperative CA19–9 ≥ 200.0 U/mL, adjuvant EPD group (Supplementary Table S10). In terms of the
chemotherapy improved both OS and DFS (Supplemen- subgroup analysis in the PSM population, EPD treatment
tary Table S6). Since DFS benefits were found from EPD significantly improved the OS in patients with WD & MD
treatment for patients who received adjuvant chemother- (Supplementary Table S11) and DFS in patients with preop-
apy (Supplementary Table S5), subgroup analyses were erative CEA < 5.00 ng/mL, with resectable stage, with R0
further performed to evaluate whether different CA19–9 resection or with WD & MD (Supplementary Table S12).
268 LIN et al.

TA B L E 4 ITT analysis of primary and secondary endpoints for predefined subgroups preoperative CA19–9 level ≥ 200.0 U/mL
Outcome SPD(n = 97) EPD(n = 94) HR (95% CI) OR (95% CI) P value
Primary
Median OS, months 19.9 18.8 1.13 (0.84 to 1.51) N/A 0.395
Secondary
Median DFS, months 12.7 12.6 1.17 (0.88 to 1.56) N/A 0.282
OS rate, %
At 1 year 68.04 70.21 N/A 0.90 (0.50 to 1.63) 0.757
At 2 year 38.14 29.79 N/A 1.45 (0.79 to 2.60) 0.285
At 3 year 18.56 17.02 N/A 1.11 (0.54 to 2.35) 0.851
At 5 year 7.22 2.13 N/A 3.58 (0.76 to 17.32) 0.170
Safety, n (%)
Morbidity 30 (30.93) 42 (44.68) N/A 1.80 (1.01 to 3.32) 0.054
Mean (SD)*
Abdominal pain upper -1.99(2.01) -1.77(2.18) N/A 0.22 (-0.38 to 0.82) 0.462
changing in 1 month
from baseline**
Back pain changing in 1 -0.91(1.85) -0.84(1.85) N/A 0.08 (-0.45 to 0.60) 0.774
month from baseline
Abdominal pain upper -0.91(3.18) -0.35(3.19) N/A 0.56 (-0.41 to 1.53) 0.256
changing in 6 months
from baseline***
Back pain changing in 6 0.35(2.72) -0.41(2.55) N/A -0.76 (-1.56 to 0.04) 0.064
months from baseline
*Data of pain intensity were shown as mean (standard deviation).
**The data was calculated as pain intensity in 1 month postoperation minus pain intensity baseline.
***The data was calculated as pain intensity in 6 months postoperation minus pain intensity baseline.
Abbreviations: ITT, intention-to-treat; SPD, standard pancreatoduodenectomy; EPD, extended pancreatoduodenectomy; CA19–9, carbohydrate antigen 19–9; HR,
hazard ratio; CI, confidence interval; OS, overall survival; DFS, disease-free survival; OR, odds ratio; NS, not significant; NRS, numerical rating scale; SD, standard
deviation.

3.6 Prognostic factors 3.7 Recurrence pattern and pain

intensity
Preoperative CEA ≥ 5.00 ng/mL, preoperative CA19–9 ≥
200.0 U/mL, PV/SMV resection, borderline resectable, R1 There were no differences in the total recurrence rate
resection, N1 stage, poor differentiation, and absence of between the EPD and SPD groups (Supplementary Table
adjuvant treatment were identified as adverse prognostic S15). As expected, the locoregional recurrence rate was sig-
factors for OS in the univariate analysis (ITT population, nificantly lower in the EPD group than in the SPD group
Supplementary Table S13). When these factors combined (16.48% vs. 35.20%; P < 0.001; Supplementary Table S15).
with treatment methods were included in the multivariate No difference was found in the total systemic recurrence
Cox proportional hazards model, preoperative CA19–9 ≥ rate between the two groups. However, the recurrence
200.0 U/mL, N1 stage and poor differentiation remained rate of mesenteric LNs was significantly higher in the
indicators for poor OS (Supplementary Table S13). SPD group (10.06% vs. 3.98%; P = 0.022; Supplementary
In the EPD group, the retroperitoneal nerve plexus was Table S15), while peritoneal seeding was more frequently
removed during the surgery and marked with the exact detected in the EPD group (17.61% vs. 8.94%; P = 0.029;
location for a separate pathological examination. The uni- Supplementary Table S15).
variate analysis of OS (qualified population in the EPD Concerning the pain intensity analysis between the EPD
group) showed no significant difference in patients with or and SPD groups, baseline pain intensity was compara-
without nerve invasion in celiac axis plexus, SMA plexus, ble, and no significant difference was found in changes
hepatoduodenal ligament plexus, common hepatic artery in either upper abdominal pain or back pain intensity
plexus or post-pancreatic plexus. In particular, nerve inva- at 1 month postoperation from baseline between the two
sion of the aortic plexus suggested poor OS (HR, 0.58; 95% groups (ITT population, Supplementary Table S16). Simi-
CI, 0.36 to 0.94; P = 0.006; Supplementary Table S14). larly, no significant differences were found in changes in
LIN et al. 269

F I G U R E 4 Analysis for the best cutoff value of preoperative CA19–9. (A) ROC curve for preoperative CA19–9. The red diagonal
represents sensitivity plus specificity = 1. The blue polyline represents the ROC curve of preoperative CA 19–9 in predicting DFS. (B) The
cutoff value of preoperative CA19–9 from ROC analysis. (C) Relationship between preoperative CA19–9 levels and DFS by RCS model. The red
curve represents the result of RCS, which indicates the HR value of CA19–9 on DFS. The black vertical dotted line indicates the zero value of
CA19–9; the black horizontal dotted line represents the reference HR value of 1.0; the red vertical dotted line indicates the position of CA19–9
of 200.0 U/mL.
Abbreviations: CA19–9, carbohydrate antigen 19–9; ROC, receiver operating characteristic; AUC, area under the curve; DFS, disease-free
survival; RCS, restricted cubic splines; HR, hazard ratio; CI, confidence interval.

upper abdominal pain intensity at 6 months postoperation to improve OS and may have even led to increased mor-
from baseline between the two groups (ITT population, bidity. Concerning the overall data analysis, our study
Supplementary Table S17). In particular, at 6 months post- was similar to previous RCTs, and we concluded that
operation, in the EPD group, back pain intensity decreased extended dissection did not benefit all patients with poten-
from 1.02 (SD, 1.83) at baseline to 0.71 (SD, 1.51), while in tially curable pancreatic head cancer. Nevertheless, the
the SPD group, back pain intensity increased from 1.07 (SD, secondary endpoint DFS was superior to EPD. As safe
1.73) at baseline to 1.43 (SD, 2.42), with a significant change and reliable as SPD, EPD led to a significant improvement
of -0.82 (95% CI, -1.37 to -0.29; P = 0.003; Supplementary in DFS for patients with pancreatic head cancer. Specif-
Table S17). ically, EPD significantly increased both OS and DFS in
patients with a low chance (preoperative CA19–9 < 200.0
U/mL) of systemic metastasis in pancreatic head cancer.
4 DISCUSSION Our study provides high-level evidence for a significant
benefit from EPD in patients with preoperative CA19–9
For many years, the delicate surgical procedures for pan- < 200.0 U/mL. Together with the predefined subgroup
creatic head cancer have not been unified, and the range of analysis of patients with a preoperative CA19–9 < 200.0
surgical resection (including the LNs and nerves), margin U/mL, this suggests a clinically relevant benefit of EPD in
of resection and usefulness of combined vascular resec- patients with pancreatic head cancer with a low probability
tion remain controversial. An updated meta-analysis from of micrometastasis.
Wang et al. [19], which included 8 studies involving 687 Regarding the morbidity of extended pancreatoduo-
(342 vs. 345) patients, showed that radical dissection failed denectomy, Wang et al. showed that the incidence of
270 LIN et al.

diarrhea (three months postoperatively) was significantly malnutrition and affecting the quality of life. Similar
higher with EPD than with SPD [19]. This finding was results were obtained from our trial, whereby we found
attributed mainly to the circumferential dissection of the that extended retroperitoneal LN dissection and right-
nerve plexus around the celiac axis and SMA in RCTs sided 270◦ dissection of the nerve plexus around the celiac
from the Mayo Clinic and Japan. In our study, we dis- axis and SMA did not increase the morbidity or mortality
sected the nerves and soft tissues at 270◦ on the right rates. The rate of postoperative diarrhea at 3 months was
side surrounding the right longitudinal axis of the celiac- similar between the SPD (5.03%) and EPD (7.46%) groups,
aorta-SMA artery axis. Our results showed that the rate with none of these patients experiencing severe intractable
of diarrhea was comparable between the SPD (5.03%) and diarrhea.
EPD (7.46%) groups (Supplementary Table S3), similar to Although the invasion of the extrapancreatic nerve in
an RCT from Korea, in which the right half of the nerve pancreatic cancer has been indicated as a poor prognostic
plexus was dissected [2]. Retroperitoneal nerve dissection factor in previous studies [7, 20, 21], the prognostic val-
involves opening the arterial sheath. Although EPD tended ues of different levels of nerve plexus involvement remain
to increase the rate of postoperative pseudoaneurysm (EPD unclear. In our study, we obtained the invasion status of
vs. SPD, 1.99% vs.1.01%, Supplementary Table S3), there different nerve plexuses in the EPD group. The univariate
was no significant difference between the two groups. analysis of OS according to retroperitoneal plexus inva-
However, by routinely performing abdominal enhance- sion showed that patients with positive invasion of the
ment computed tomography angiography (CTA) within aortic plexus showed worse OS in EPD treatment (Sup-
one week of surgery, we could effectively detect pseu- plementary Table S14), which suggested that this group
doaneurysms and prevent life-threatening complications of patients may need more individualized follow-up and
caused by the rupture of pseudoaneurysms through angio- adjuvant therapy after surgery. The perioperative CA19–9
graphic embolization or membrane stent implantation level is one of the most reliable tumor markers for assess-
(only one patient died of severe sepsis with gastroduode- ing pancreatic cancer. An elevated preoperative CA19–9
nal artery rupture in the EPD group, and the proportions of level is believed to be an independent predictor of early
aneurysm-related reoperations of EPD vs. SPD were 1.00% postoperative recurrence and metastasis, even if R0 surgi-
vs. 1.01% [Supplementary Table S3], with no significant cal resection is achieved [9, 22]. Forsmark et al. reported
difference). that CA 19–9 levels greater than 300 U/mL indicated an
Surgical techniques that have been developed to refine advanced stage of pancreatic cancer and increased the
oncological resections and surgeons may be able to impact risk of unresectability, but their small sample limited their
local control by the radicality and quality of surgical resec- research results [10]. Furthermore, multivariate regression
tion. Techniques used to achieve local radicality include analysis demonstrated that the independent contributing
artery-first approaches, the triangle operation, extended factor to resectability (R0 resection) was a preoperative
resections and level-3 dissection with the removal of the CA 19–9 level < 92.77 U/mL [11]. Our results showed that
nerves and soft tissues surrounding the artery. To avoid the sufficient dissection of nerve tissues at the retroperi-
complications from extended resections, we modified our toneum and LNs around the pancreatic head significantly
procedure to refine the range to dissect the nerve and soft improved the prognosis of patients with pancreatic head
tissues at 270◦ around the right longitudinal axis of the cancer whose preoperative CA19–9 level was < 200.0
celiac axis and SMA. U/mL. The actual OS rate at 3 years was highly similar
At present, evidence on the extent of nerve dissection to the predicted rate and was approximately 15% higher in
for pancreatic head cancer remains insufficient. There the EPD group than in the SPD group (40.19% vs. 26.47%,
are currently only 3 RCTs on the extent of nerve dis- P = 0.041; Table 3). In line with our findings, an analy-
section for pancreatic head cancer. As early as the last sis from Japan showed that for pancreatic cancer patients
decade, Japanese scholars emphasized that 360◦ circum- with LN16 positivity, surgical resection and extended lym-
ferential dissection of the nerve plexus around the celiac phadenectomy significantly improved the OS of those with
axis and SMA improved R0 resection, but both RCTs a preoperative CA19–9 ≤ 360 U/mL compared with bypass
from Japan reported that extended pancreatectomy led surgery [23]. Kim et al. reported that markedly elevated
to intractable diarrhea, malnutrition and low quality preoperative CA19–9 levels might reflect unresectability in
of life, which in turn affected patient prognosis. Based pancreatic adenocarcinoma patients who were thought to
on the modified extent of nerve plexus dissection in have resectable disease on preoperative imaging [11]. The
Japan, nerve plexus dissection with a right-sided range PRODIGE 24 trial showed that the modified “fluorouracil,
of 180◦ was utilized in the late-stage RCT performed by leucovorin, irinotecan, and oxaliplatin (FOLFIRINOX)”
Korean scholars [2]. The results suggested no increase in regimen led to significantly longer survival than gemc-
the risk of intractable diarrhea leading to postoperative itabine among patients with resected pancreatic cancer.
LIN et al. 271

Meanwhile, to minimize the risk of the incorrect inclusion resection in patients with pancreatic head cancer signifi-
of patients with metastatic disease, only patients with post- cantly improved oncological outcomes in those in an early
operative serum CA19–9 levels < 180 U/mL were included stage with lower potential metastasis than SPD.
[24]. In this study, according to the RCS model and ROC Approximately 20% of the patients in this study required
analysis, we recommend the cutoff point of CA19–9 as vein reconstruction (Table 1), which means that this group
200.0 U/mL for stratified treatment options in pancreatic was in the borderline stage and had a higher chance of
head cancer. micrometastasis. Our subgroup analysis confirmed that
In our trial, we found no prognostic difference between EPD did not show an advantage over SPD for patients with
the two groups of patients with preoperative CA19–9 ≥ PV/SMV reconstruction (Supplementary Table S5), but for
200.0 U/mL, which indicates that micrometastasis might the subgroup who did not require vein reconstruction, we
be the main contributor to prognosis. In our trial, when found a significant improvement in DFS in the EPD group
the rate of the recurrence pattern manifested only as sys- (Supplementary Figure S2F). For patients requiring vascu-
temic metastasis, the proportion was as high as 80.43% (148 lar resection, neoadjuvant chemotherapy should be given
out of 184) in the subgroup with preoperative CA19–9 ≥ priority according to current guidelines. An interesting
200.0 U/mL, which probably resulted in an underpowered new topic is whether EPD combined with vascular resec-
study. The key point that might need to be discussed in tion is a better option after neoadjuvant therapy, which we
this subgroup is upfront surgery or neoadjuvant chemora- will explore in the future. In the absence of RCTs to sup-
diotherapy. A previous retrospective analysis showed that port the extent of dissection and the treatment sequencing
normalizing postoperative CA19–9 levels was a strong strategy in subgroups of patients with resectable pancre-
prognostic marker for long-term survival [25]. In regard to atic head cancer, hopefully, our data will encourage more
patients who had high preoperative CA19–9 ≥ 200.0 U/mL, open trials to provide a more precise strategy for differ-
we found that patients whose CA19–9 levels decreased to ent subgroups of patients with resectable pancreatic head
normal 2 weeks after surgery had a significantly better cancer.
outcome than those with elevated CA19–9 levels (Supple- In this study, most of the patients (343/400, 85.75%)
mentary Figure S3K-L), further emphasizing on the impor- accepted postoperative adjuvant chemotherapy, with a
tant predictive value of CA19–9 for prognosis. Nonetheless, comparable rate in the SPD group and the EPD group
patients with pancreatic head cancer are often associated (86.93% vs.84.58%, P = 0.568, Table 1) and the proportion
with cholangitis, pancreatitis and obstructive jaundice, of patients who received complete chemotherapy between
resulting in the elevation of CA19–9 [26, 27]. In our trial, the two groups was similar (EPD vs. SPD, 62.19% vs. 65.33%,
patients with jaundice should be tested for baseline CA19– P = 0.534, Table 1), which indicates that EPD did not
9 after adequate jaundice-reducing treatment and/or bile significantly affect the postoperative chemotherapy accep-
drainage. tance rate of patients. The median OS in the adjuvant
Moreover, in the subgroup with well and moderate chemotherapy group (mainly gemcitabine single-agent
pathological differentiation (WD and MD, n = 224), the chemotherapy) was similar to that reported in previous
median OS (subgroup from the ITT population) in the EPD phase 3 trials of adjuvant therapy (24.4 months vs. 20.1
group was 6.2 months longer than that in the SPD group (P to 26.5 months), although the median DFS was slightly
= 0.020, Supplementary Figure S2C). Correspondingly, the longer in our trial (17.7 months vs. 11.3 to 15.3 months)
median DFS in the EPD group was 5.8 months longer than [28–32]. Of the 343 patients who underwent postopera-
that in the SPD group (P = 0.006, Supplementary Figure tive adjuvant chemotherapy, DFS was significantly longer
S2D). Furthermore, in the EPD group, the subgroup with a in the EPD group than in the SPD group, although the
preoperative CEA level less than 5.00 ng/mL (Supplemen- difference was only 3.2 months (17.7 months vs. 14.5
tary Figure S2B), with a resectable stage (Supplementary months, P = 0.019, Supplementary Figure S2J). These
Figure S3H), without PV/SMV resection (Supplementary results were inconsistent with those from a Korean study
Figure S2F) or who underwent R0 resection (Supplemen- [2], in which adjuvant treatment had no effect on sur-
tary Figure S2H) and who received postoperative adjuvant vival in the extended resection group. Vascular injury by
chemotherapy (Supplementary Figure S2J) achieved bet- extensive dissection was considered by a Korean study
ter outcomes based on the DFS analysis (Supplementary to reduce the postoperative effects of chemoradiation,
Table S5). Interestingly, factors including postoperative whereas both our study and the Korean study showed
CA19–9 < 200.0 U/mL, N0 stage, and well and moderate that the major type of pancreatic cancer recurrence after
pathological differentiation significantly predicted a good surgery was systemic recurrence (locoregional vs. systemic
prognosis in both the univariate and multivariate analyses for extended resection, 96.6% vs. 25.9% in the Korean study,
(Supplementary Table S13). These results also strength- 89.77% vs. 16.48% in our study, Supplementary Table S15),
ened our conclusion that EPD with retroperitoneal nerve and the main effect of chemotherapeutics should be to
272 LIN et al.

patients to undergo EPD, which has more advantages in

achieving adequate dissection of retroperitoneal nerve tis-
sues and lymph nodes and R0 resection. Therefore, we
believe it may be the best choice for patients to receive sys-
temic treatment based on adequate radical surgery (EPD
is more likely to achieve local radical resection than SPD).
In addition, pancreatoduodenectomy surgery has entered
the era of minimally invasive surgery, including laparo-
scopic and robotic pancreatoduodenectomy. At present,
minimally invasive PD (MIPD) and open pancreatoduo-
denectomy (OPD) are basically the same in terms of the
selection of patients, except for the unique contraindica-
tions of minimally invasive procedures such as the inability
to tolerate pneumoperitoneum or the inability to safely
establish pneumoperitoneum [36]. Several studies have
shown that minimally invasive EPD with vascular replace-
ment is safe and feasible with highly experienced surgeons
F I G U R E 5 A flow chart of treatment options for patients with and in high-volume pancreatic centers [37–39], although
TNM stage I and II pancreatic cancer. Abbreviations: CA19–9, one study reported that a higher proportion of complex seg-
carbohydrate antigen 19–9. mental resections was performed in the OPD group than
in the LPD group [39]. Although the data are limited, we
believe that the inherent advantages of minimally invasive
preferentially reduce tumor spread or metastasis rather approaches and delicate manipulation will lead to accu-
than locoregional recurrence. rate results for the EPD of pancreatic head cancer. Our
Furthermore, in the subgroup with CA19–9 < 200.0 team is conducting RCTs to explore the short-term and
U/mL, the benefit of EPD compared with SPD was the long-term oncological effects of MIPD in treating pancre-
most significant in the subgroup that received chemother- atic head cancer [40]. In the future, we will further explore
apy (Supplementary Table S7), while for the subgroup with the safety of minimally invasive EPD and which subgroups
CA19–9 ≥ 200.0 U/mL, EPD showed no significant ben- of patients can truly benefit from minimally invasive EPD.
efit in the population with and without adjuvant therapy In addition, our PSM analysis suggested that there were
(Supplementary Table S8). Based on this study, for pancre- no significant confounding factors in this study, which
atic head cancer in stages I and II, we recommend EPD further affirmed the validity of the randomization and
combined with adjuvant chemotherapy for patients with the reliability of the results of this study. Furthermore,
preoperative CA19–9 < 200.0 U/mL and SPD combined we supplemented the table with data on loss to follow-
with adjuvant chemotherapy for patients with CA19–9 ≥ up, which was only 2.25% (Supplementary Table S18). The
200.0 U/mL (Figure 5) table shows that the missing data were mainly within 1
Considering that a large proportion of PDACs deemed year of follow-up, and the time period of loss to follow-up
resectable by conventional imaging are likely metastatic in the two groups was very similar. The Gehan-Breslow-
[9], neoadjuvant treatment in high-risk PDAC patients Wilcoxon test analysis was used to extra weight for early
may provide a chance for tumor downsizing or even time points, and the results showed that the missing data
downstaging, thereby improving surgical outcomes. Thus, did not significantly impact the results in the ITT popula-
the critical question raised by our trial is whether we tion (Supplementary Table S19) or in subpopulation based
can achieve a better outcome when extended surgery on a predefined preoperative CA19–9 level [< 200.0 U/mL
is performed for patients with high-risk micrometasta- or ≥ 200.0 U/mL] (Supplementary Table S20), further
sis after preoperative chemotherapy, which could indeed indicating the reliability of the study.
treat micrometastases and/or limit tumor spread during The goal of a cure for patients with resectable pancre-
surgery [33]. Recent studies have suggested the possibil- atic cancer is careful management of the surgical margin
ity of using preoperative CA19–9 levels to discern patient to ensure adequate tumor clearance. Although there was
subgroups that would benefit from upfront surgery or no difference in the total recurrence rate between the two
neoadjuvant therapy [34, 35]. Currently, neoadjuvant ther- groups, the locoregional recurrence rate was significantly
apy is the standard strategy for patients with borderline lower in the EPD group than in the SPD group (EPD vs.
pancreatic cancer. After neoadjuvant therapy to control SPD, 16.48% vs. 35.20%, P < 0.001, Supplementary Table
or exclude micrometastases, it is more meaningful for S15). Moreover, the recurrence rate of mesenteric LNs was
LIN et al. 273

significantly higher in the SPD group (SPD vs. EPD, 10.06% Qing Lin, Shangyou Zheng, Yu Zhou, Zhihua Li: acquisi-
vs. 3.98%, P = 0.022, Supplementary Table S15), while peri- tion of data; analysis and interpretation of data; statistical
toneal seeding was more frequently detected in the EPD analysis; drafting of the manuscript, approval of the final
group (EPD vs. SPD, 17.61% vs. 8.94%, P = 0.029, Sup- version of the manuscript; informed consent and data veri-
plementary Table S15), which may have been caused by fication. Xianjun Yu, Meifu Chen Yixiong Li, Weilin Wang,
prolonged and extended manipulation around the tumor. Renyi Qin: study concept and design; surgeon performing
Nevertheless, this study has limitations. First, the procedures; acquisition of data; analysis and interpreta-
patients enrolled in this study included patients with bor- tion of data; drafting of the manuscript; critical revision
derline resectable disease who had not been screened for of the manuscript for important intellectual content;
neoadjuvant therapy and were at high risk of micrometas- administrative, technical, or material support; study super-
tasis. Second, the postoperative chemotherapy regimen vision; approval of the final version of the manuscript;
used in this study was gemcitabine as a single agent, which informed consent and data verification. Quanbo Zhou,
may be less effective in controlling systemic recurrence Chonghui Hu, Zhongdong Xu, Lin Wang, Yimin Liu, Min
than albumin-bound paclitaxel plus gemcitabine (AG)and Wang, Guolin Li, He Cheng, Dongkai Zhou, Guodong
FOLFININOX regimens. Intolerance after chemotherapy Liu, Zhiqiang Fu, Yu Long: acquisition of data; analysis
in some patients and changes in the regimen after recur- and interpretation of data; approval of the final version
rence may have also caused a bias in survival prognosis of the manuscript. Jing Gu: Analysis and interpretation
between the two groups. Third, the subgroup analysis was of data; statistical analysis; approval of the final version
stratified based on a predefined preoperative CA19–9 level of the manuscript. Qingyu Liu: Radiologist interpret-
[< 200.0 U/mL or ≥ 200.0 U/mL], and there was no prior ing cross-sectional imaging in the trial; critical revision
detection of Lewis negative status. Approximately 5 to 10% of the manuscript; approval of the final version of the
of the population who are negative for the Lewis anti- manuscript. The corresponding authors are responsible for
gen have no or scarce secretion of CA19–9, causing false all aspects of this study, including the entire contents and
negative results. A previous study reported that among all data.
Lewis antigen-negative individuals with pancreatic cancer,
high levels of CEA and CA125 were associated with a high AC K N OW L E D G M E N T S
risk of micrometastasis [41] and that the effect of radical The authors would like to thank all patients who consented
surgery was attenuated in this subgroup of patients. Due to be screened and who participated in the study. We are
to the limited number of patients, we did not perform fur- grateful to our colleagues and research staff who partici-
ther stratified analyses of these confounding factors in our pated in the enrollment and evaluation of the patients at
study. each center. We also thank Prof. Jie Wang and Prof. Quanx-
In this multicenter RCT comparing EPD and SPD for ing Ni for advice on the study design; Prof. Chunyou Wang
pancreatic head cancer, the difference in OS was not sta- and Prof. Zhengxi Jin from the Pancreatic Cancer Commit-
tistically significant, but we observed that the patients in tee of the Chinese Anti-Cancer Association as third-party
the EPD arm had significantly improved DFS. In the sub- experts for the independent evaluation to avoid intentional
group analysis, improvements in both OS and DFS in the or unintentional patient enrollment; Prof. Yamei Tang for
EPD arm were observed in patients with CA19–9 < 200.0 critical revision of the manuscript.
U/mL. Considering that there was no significant differ-
ence in mortality and morbidity between the two groups, CONFLICT OF INTEREST
EPD could be used as an effective surgical procedure for DISCLOSURES
patients with pancreatic head cancer, especially those with We declare no competing interests.
preoperative CA19–9 < 200.0 U/mL.
DATA S H A R I N G S TAT E M E N T
D E C L A R AT I O N S Individual de-identified participant data that underlie the
AU T H O R CO N T R I B U T I O N S results reported in this article and study protocol will be
Rufu Chen: study concept and design; surgeon performing shared with investigators whose proposed use of the data
procedures; acquisition of data; analysis and interpreta- has been approved by an independent review committee.
tion of data; drafting of the manuscript; critical revision Data can only be used to achieve aims in the approved pro-
of the manuscript for important intellectual content; posal. Data will be available 6-36 months after the article is
obtained funding; administrative, technical, or material published. To gain access, data requesters will need to sign
support; study supervision; approval of the final version a data access agreement. Proposals should be directed to
of the manuscript; informed consent and data verification. CRF (chenrufu@mail.sysu.edu.cn).
274 LIN et al.

E T H I C S A P P R O VA L A N D C O N S E N T T O 9. Haeno H, Gonen M, Davis MB, Herman JM, Iacobuzio-Donahue

PA R T I C I PA T E CA, Michor F. Computational modeling of pancreatic cancer
This trial was approved by all independent institutional reveals kinetics of metastasis suggesting optimum treatment
strategies. Cell. 2012;148(1-2):362-75.
review boards and ethics committees at each participat-
10. Forsmark CE, Lambiase L, Vogel SB. Diagnosis of pancreatic
ing site. All study-related procedures were performed with
cancer and prediction of unresectability using the tumor-
the written informed consent of all patients. This study associated antigen CA19-9. Pancreas. 1994;9(6):731-4.
complies with the Declaration of Helsinki and the relevant 11. Kim YC, Kim HJ, Park JH, Park DI, Cho YK, Sohn CI, et al.
provisions of the guidelines in both the collection of data Can preoperative CA19-9 and CEA levels predict the resectabil-
and the use of materials. ity of patients with pancreatic adenocarcinoma? J Gastroenterol
Hepatol. 2009;24(12):1869-75.
12. Ferreira-Valente MA, Pais-Ribeiro JL, Jensen MP. Validity of
C O N S E N T F O R P U B L I C AT I O N four pain intensity rating scales. Pain. 2011;152(10):2399-404.
Not applicable. 13. Ware JE, Jr, Gandek B. Overview of the SF-36 Health Survey and
the International Quality of Life Assessment (IQOLA) Project. J
ORCID Clin Epidemiol. 1998;51(11):903-12.
Xianjun Yu https://orcid.org/0000-0002-6697-7143 14. Oken MM, Creech RH, Tormey DC, Horton J, Davis TE,
McFadden ET, et al. Toxicity and response criteria of the Eastern
Renyi Qin https://orcid.org/0000-0002-4552-9720
Cooperative Oncology Group. Am J Clin Oncol. 1982;5(6):649-55.
Rufu Chen https://orcid.org/0000-0003-4568-1324
15. Yeo CJ, Cameron JL, Sohn TA, Coleman J, Sauter PK, Hruban
RH, et al. Pancreaticoduodenectomy with or without extended
REFERENCES retroperitoneal lymphadenectomy for periampullary adenocar-
1. Siegel RL, Miller KD, Fuchs HE, Jemal A. Cancer Statistics, 2021. cinoma: comparison of morbidity and mortality and short-term
CA Cancer J Clin. 2021;71(1):7-33. outcome. Ann Surg. 1999;229(5):613-22; discussion 22-4.
2. Jang JY, Kang MJ, Heo JS, Choi SH, Choi DW, Park SJ, et al. A 16. Kenward MG, Carpenter J. Multiple imputation: current per-
prospective randomized controlled study comparing outcomes spectives. Stat Methods Med Res. 2007;16(3):199-218.
of standard resection and extended resection, including dissec- 17. Durrleman S, Simon R. Flexible regression models with cubic
tion of the nerve plexus and various lymph nodes, in patients splines. Stat Med. 1989;8(5):551-61.
with pancreatic head cancer. Ann Surg. 2014;259(4):656-64. 18. Molinari N, Daures JP, Durand JF. Regression splines for thresh-
3. Pedrazzoli S, DiCarlo V, Dionigi R, Mosca F, Pederzoli P, old selection in survival data analysis. Stat Med. 2001;20(2):237-
Pasquali C, et al. Standard versus extended lymphadenec- 47.
tomy associated with pancreatoduodenectomy in the surgical 19. Wang W, He Y, Wu L, Ye L, Yao L, Tang Z. Efficacy of extended
treatment of adenocarcinoma of the head of the pancreas: a versus standard lymphadenectomy in pancreatoduodenectomy
multicenter, prospective, randomized study. Lymphadenectomy for pancreatic head adenocarcinoma. An update meta-analysis.
Study Group. Ann Surg. 1998;228(4):508-17. Pancreatology. 2019;19(8):1074-80.
4. Riall TS, Cameron JL, Lillemoe KD, Campbell KA, Sauter PK, 20. Chatterjee D, Katz MH, Rashid A, Wang H, Iuga AC,
Coleman J, et al. Pancreaticoduodenectomy with or without dis- Varadhachary GR, et al. Perineural and intraneural invasion in
tal gastrectomy and extended retroperitoneal lymphadenectomy posttherapy pancreaticoduodenectomy specimens predicts poor
for periampullary adenocarcinoma–part 3: update on 5-year sur- prognosis in patients with pancreatic ductal adenocarcinoma.
vival. J Gastrointest Surg. 2005;9(9):1191-204; discussion 204-6. Am J Surg Pathol. 2012;36(3):409-17.
5. Nimura Y, Nagino M, Takao S, Takada T, Miyazaki K, Kawarada 21. Chen JW, Bhandari M, Astill DS, Wilson TG, Kow L, Brooke-
Y, et al. Standard versus extended lymphadenectomy in radi- Smith M, et al. Predicting patient survival after pancreatico-
cal pancreatoduodenectomy for ductal adenocarcinoma of the duodenectomy for malignancy: histopathological criteria based
head of the pancreas: long-term results of a Japanese multicen- on perineural infiltration and lymphovascular invasion. HPB
ter randomized controlled trial. J Hepatobiliary Pancreat Sci. (Oxford). 2010;12(2):101-8.
2012;19(3):230-41. 22. Aoki S, Motoi F, Murakami Y, Sho M, Satoi S, Honda G,
6. Farnell MB, Pearson RK, Sarr MG, DiMagno EP, Burgart et al. Decreased serum carbohydrate antigen 19-9 levels after
LJ, Dahl TR, et al. A prospective randomized trial compar- neoadjuvant therapy predict a better prognosis for patients with
ing standard pancreatoduodenectomy with pancreatoduodenec- pancreatic adenocarcinoma: a multicenter case-control study of
tomy with extended lymphadenectomy in resectable pancreatic 240 patients. BMC Cancer. 2019;19(1):252.
head adenocarcinoma. Surgery. 2005;138(4):618-28; discussion 23. Asaoka T, Miyamoto A, Maeda S, Hama N, Tsujie M, Ikeda
28-30. M, et al. CA19-9 level determines therapeutic modality in pan-
7. Schorn S, Demir IE, Haller B, Scheufele F, Reyes CM, Tieftrunk creatic cancer patients with para-aortic lymph node metastasis.
E, et al. The influence of neural invasion on survival and tumor Hepatobiliary Pancreat Dis Int. 2018;17(1):75-80.
recurrence in pancreatic ductal adenocarcinoma - A systematic 24. Conroy T, Hammel P, Hebbar M, Ben Abdelghani M, Wei AC,
review and meta-analysis. Surg Oncol. 2017;26(1):105-15. Raoul JL, et al. FOLFIRINOX or Gemcitabine as Adjuvant Ther-
8. Qin RY, Cao XY, Zhu F, Wang X. A new method for integrated apy for Pancreatic Cancer. N Engl J Med. 2018;379(25):2395-406.
radical resection of uncinate process of the pancreas: a report of 25. Tsai S, George B, Wittmann D, Ritch PS, Krepline AN, Aldakkak
306 cases. Zhonghua Wai Ke Za Zhi. 2010;48(18):1379-82. M, et al. Importance of Normalization of CA19-9 Levels
LIN et al. 275

Following Neoadjuvant Therapy in Patients With Localized cancer treated with neoadjuvant chemoradiation. Ann Surg
Pancreatic Cancer. Ann Surg. 2020;271(4):740-7. Oncol. 2010;17(7):1794-801.
26. Lamerz R. Role of tumour markers, cytogenetics. Ann Oncol. 36. Wang M, Peng B, Liu J, Yin X, Tan Z, Liu R, et al. Practice
1999;10Suppl (4):145-9. Patterns and Perioperative Outcomes of Laparoscopic Pan-
27. Azizian A, Ruhlmann F, Krause T, Bernhardt M, Jo P, Konig A, creaticoduodenectomy in China: A Retrospective Multicenter
et al. CA19-9 for detecting recurrence of pancreatic cancer. Sci Analysis of 1029 Patients. Ann Surg. 2021;273(1):145-53.
Rep. 2020;10(1):1332. 37. Kendrick ML, Sclabas GM. Major venous resection during
28. Oettle H, Post S, Neuhaus P, Gellert K, Langrehr J, Ridwelski K, total laparoscopic pancreaticoduodenectomy. HPB (Oxford).
et al. Adjuvant chemotherapy with gemcitabine vs observation 2011;13(7):454-8.
in patients undergoing curative-intent resection of pancreatic 38. Yang E, Chong Y, Wang Z, Koh YX, Lim KI, BK PG. Minimally-
cancer: a randomized controlled trial. JAMA. 2007;297(3):267-77. invasive versus open pancreatoduodenectomies with vascular
29. Uesaka K, Boku N, Fukutomi A, Okamura Y, Konishi M, resection: A 1:1 propensity-matched comparison study. J Minim
Matsumoto I, et al. Adjuvant chemotherapy of S-1 versus Access Surg. 2022;18(3):420-5.
gemcitabine for resected pancreatic cancer: a phase 3, open- 39. Croome KP, Farnell MB, Que FG, Reid-Lombardo KM, Truty
label, randomised, non-inferiority trial (JASPAC 01). Lancet. MJ, Nagorney DM, et al. Pancreaticoduodenectomy with major
2016;388(10041):248-57. vascular resection: a comparison of laparoscopic versus open
30. Neoptolemos JP, Palmer DH, Ghaneh P, Psarelli EE, Valle JW, approaches. J Gastrointest Surg. 2015;19(1):189-94; discussion 94.
Halloran CM, et al. Comparison of adjuvant gemcitabine and 40. Pan S, Qin T, Yin T, Yu X, Li J, Liu J, et al. Laparoscopic versus
capecitabine with gemcitabine monotherapy in patients with open pancreaticoduodenectomy for pancreatic ductal adenocar-
resected pancreatic cancer (ESPAC-4): a multicentre, open-label, cinoma: study protocol for a multicentre randomised controlled
randomised, phase 3 trial. Lancet. 2017;389(10073):1011-24. trial. BMJ Open. 2022;12(4):e057128.
31. Sinn M, Bahra M, Liersch T, Gellert K, Messmann H, Bechstein 41. Luo G, Liu C, Guo M, Cheng H, Lu Y, Jin K, et al. Potential
W, et al. CONKO-005: Adjuvant Chemotherapy With Gemc- Biomarkers in Lewis Negative Patients With Pancreatic Cancer.
itabine Plus Erlotinib Versus Gemcitabine Alone in Patients Ann Surg. 2017;265(4):800-5.
After R0 Resection of Pancreatic Cancer: A Multicenter Ran-
domized Phase III Trial. J Clin Oncol. 2017;35(29):3330-7.
32. Neoptolemos JP, Stocken DD, Friess H, Bassi C, Dunn JA,
Hickey H, et al. A randomized trial of chemoradiotherapy and
S U P P O RT I N G I N F O R M AT I O N
chemotherapy after resection of pancreatic cancer. N Engl J Med. Additional supporting information can be found online
2004;350(12):1200-10. in the Supporting Information section at the end of this
33. Mokdad AA, Minter RM, Zhu H, Augustine MM, Porembka MR, article.
Wang SC, et al. Neoadjuvant Therapy Followed by Resection
Versus Upfront Resection for Resectable Pancreatic Cancer: A
Propensity Score Matched Analysis. J Clin Oncol. 2017;35(5):515-
22. How to cite this article: Lin Q, Zheng S, Yu X,
34. Bergquist JR, Puig CA, Shubert CR, Groeschl RT, Habermann Chen M, Zhou Y, Zhou Q, et al. Standard
EB, Kendrick ML, et al. Carbohydrate Antigen 19-9 Elevation pancreatoduodenectomy versus extended
in Anatomically Resectable, Early Stage Pancreatic Cancer Is pancreatoduodenectomy with modified
Independently Associated with Decreased Overall Survival and
retroperitoneal nerve resection in patients with
an Indication for Neoadjuvant Therapy: A National Cancer
pancreatic head cancer: a multicenter randomized
Database Study. J Am Coll Surg. 2016;223(1):52-65.
35. Katz MH, Varadhachary GR, Fleming JB, Wolff RA, Lee JE, controlled trial. Cancer Communications.
Pisters PW, et al. Serum CA 19-9 as a marker of resectability 2023;43:257–275. https://doi.org/10.1002/cac2.12399
and survival in patients with potentially resectable pancreatic