Customize your JAMA Network experience by selecting one or more topics from the list below.
Adam J, Jacquin A, Laurent C, et al. Laparoscopic Spleen-Preserving Distal Pancreatectomy: Splenic Vessel Preservation Compared With the Warshaw Technique. JAMA Surg. 2013;148(3):246–252. doi:10.1001/jamasurg.2013.768
Objective To compare preservation with the division of the splenic vessels
in the surgical management of laparoscopic spleen-preserving distal
Design Bicentric retrospective study.
Setting Prospectively maintained databases.
Patients Between January 1997 and January 2011, 140 patients who underwent
laparoscopic spleen-preserving distal pancreatectomy for benign or
low-grade malignant tumors in the body/tail of the pancreas were included.
Patients treated with the attempted splenic vessel preservation were
compared with patients treated with the attempted division of the
splenic vessels (Warshaw technique).
Main Outcome Measures Operative outcomes and postoperative morbidity were evaluated.
Results The outcomes of 55 patients in the splenic vessel preservation
group were compared with those of 85 patients in the Warshaw technique
group. The clinical characteristics were similar in both groups, except
for tumor size, which was significantly greater in the Warshaw technique
group (33.6 vs 42.5 mm; P < .001).
The mean operative time, mean blood loss, and rate of conversion to
the open procedure did not differ between the 2 groups. The rate of
successful spleen preservation was significantly improved following
the splenic vessel preservation technique (96.4% vs 84.7%; P = .03). Complications related to the
spleen only occurred in the Warshaw technique group (0% vs 10.5%; P = .03), requiring a splenectomy in
4 patients (4.7%). The mean length of stay was shorter in the splenic
vessel preservation group (8.2 vs 10.5 days; P = .01).
Conclusions The short-term benefits associated with the preservation of
the splenic vessels should lead to an increased preference for this
technique in selected patients undergoing laparoscopic spleen-preserving
distal pancreatectomy for benign or low-grade malignant tumors in
the body/tail of the pancreas.
Over the last 2 decades, laparoscopic pancreatic surgery has
become increasingly popular owing to improvements in surgical skills
and research regarding minimally invasive procedures. Distal pancreatectomy
is the most commonly reported laparoscopic pancreatic procedure in
the literature and is generally considered to be safe and practical.1-7
Spleen conservation during laparoscopic distal pancreatectomy
remains controversial based on the indications of pancreatic resections.8 However, in patients with benign or low-grade
malignant tumors in the body/tail of the pancreas, conservation of
the spleen eliminates the risk for overwhelming postsplenectomy sepsis
and other complications related to a splenectomy.9-11
Two surgical techniques are available to preserve the spleen.
First, Warshaw12 described a technique
in which splenic vessels are ligated with the preservation of the
short gastric and left gastroepiploic vessels. Second, the operation
can also be performed by sparing the splenic vessels, which assures
increased blood supply to the spleen.13
Neither one of these techniques is favored over the other, and
the reasons for selection remain unclear. Ligation of the splenic
vessels is known to have potential risks including postoperative infarction
of the spleen and the need for a reoperation.14,15 To our knowledge, only
a few studies with a small number of patients have compared both laparoscopic
The objective of this study was to compare splenic vessel preservation
(SVP) with the Warshaw technique (WT) in the surgical management of
patients undergoing laparoscopic spleen-preserving distal pancreatectomy
A retrospective analysis of prospectively maintained databases
from 2 academic medical centers was performed. Data from patients
undergoing LSPDP for benign or low-grade malignant tumors in the body/tail
of the pancreas between January 1997 and January 2011 were included.
Patients with suspected malignant tumors or planned preoperative splenopancreatectomies
were excluded. Patients were divided into 2 groups. In the SVP group,
SVP was performed on an intention-to-treat basis. In the WT group,
division of the splenic vessels was performed on an intention-to-treat
Both medical centers (University of Bordeaux Segalen and University
of Barcelona) provided specified preoperative, operative, and postoperative
data parameters using a common menu-driven database file. Informed
consent was obtained from each patient, and the study was approved
by the local ethics committee.
Patient selection, choice of procedure, technical operative
course, and use of surgical equipment were independently determined
by each surgeon based on the patients' preoperative findings and surgical
experience. Both approaches were performed in the 2 centers throughout
the study. The protocols for all of the surgical procedures were notably
similar between the institutions, as described in previous publications.16,17
In general, the patient is positioned supine with the legs apart
and arms tucked beside the body. The surgeon stands between the patient's
legs, with an assistant on the patient's left side and a scrub nurse
on the right side. A video monitor is placed to the left of the patient's
head. The procedure requires 4 trocars. A 10-mm port is placed at
the umbilicus for the 30° telescope, a second 10-mm port is located
along the left costal margin with a 5-mm right subcostal port and
a 5-mm subxiphoid epigastric port. The table is placed in a reverse
Trendelenburg position to facilitate displacement of the transverse
colon and small bowel from the operative field. The lesser sac is
opened by the lysis of the avascular plane between the transverse
colon and the greater omentum to expose the isthmus, body, and tail
of the pancreas. The posterior aspect of the stomach is grasped and
elevated with a fenestrated grasper introduced through the subxiphoid
port to enable exploration of the entire left side of the pancreas.
Localization of the tumor and delineation of its extent is performed
using intraoperative ultrasonography. The inferior border of the pancreas
is dissected at the level of the isthmus. After identification of
the mesentericoportal venous axis, the avascular plane between the
posterior aspect of the isthmus and the anterior wall of the portal
vein is dissected. After the upper edge of the pancreas has been freed,
the anterior wall of the portal vein lies within the triangle formed
by the common hepatic artery above, the gastroduodenal artery to the
right, and the upper edge of the pancreas below. The retro-isthmic
tunnel is completely dissected, and the pancreatic isthmus is encircled
and retracted with tape. The isolated isthmus is divided using a cutting
linear stapler or ultrasonic scissors (harmonic scalpel) to identify
and ligate the Wirsung duct separately.
The splenic artery and vein are isolated at their origins and
encircled with tape.
The splenic vein and splenic artery are progressively dissected
and freed from the tail of the pancreas from right to left. Small
venous branches supplying the pancreas must be clipped and divided
with ultrasonic scissors. Arterial branches are also clipped and divided
After the splenic artery has been clipped and divided, the inferior
border of the pancreas is freed from the transverse mesocolon all
the way to the tip of the pancreatic tail. The tail of the pancreas
is dissected and freed from the splenic hilum. The splenic artery
and vein appear above the upper edge of the pancreas at the junction
of the body and tail. The short gastric vessels must be identified
such that the splenic vessels can be divided immediately before the
first short gastric branches (
The specimen is placed in a sack and removed from the abdomen,
which may require enlargement of the left subcostal trocar site with
a transverse incision. A suction drain remains in the pancreatic bed.
Clinicopathologic characteristics and operative and postoperative
outcomes were examined. Tumor size and pathologic diagnoses were obtained
from the final histologic report. The type of procedure was defined
on an intention-to-treat basis. Blood loss and operative times were
obtained from the patient record. The need for an open conversion
and the occurrence of intraoperative complications were also reported.
Morbidity was defined as a complication occurring within 30
days after surgery or during the hospital stay. Complications were
graded using Dindo classification.18 A pancreatic fistula (PF) was defined as an amylase concentration—as
measured in the fluid collected at days 3 and 5 after a drain was
placed intraoperatively—that was more than 3 times greater than
the serum concentration. Pancreatic fistulas were classified according
to the clinical impact on the patient's course (grade A, B, or C)
using the definition from the International Study Group of Pancreatic
Postoperative management differed between the 2 academic medical
centers. One institution performed color Doppler ultrasonography on
all patients during the postoperative period. The second institution
conducted diagnostic imaging (color Doppler ultrasonography and/or
computed tomography) when patients were suspected of developing medical
or surgical complications.
Continuous data were expressed as means and standard deviations.
Categorical variables were analyzed with the chi-squared test or Fisher
exact test, and continuous variables were analyzed with the t test. Two-tailed P < .05
was considered statistically significant.
All statistical analyses were performed using GraphPad Prism
version 4 for Macintosh (GraphPad Software).
From January 1997 through January 2011, 140 LSPDPs were performed
(University of Bordeaux Segalen, n = 64; University of Barcelona,
n = 76).
Fifty-five patients (University of Bordeaux Segalen, n = 25;
University of Barcelona, n = 30) underwent LSPDP on an intention-to-treat
basis with SVP. Eighty-five patients (University of Bordeaux Segalen,
n = 39; University of Barcelona, n = 46) underwent
LSPDP with WT.
The clinical and pathologic characteristics of all the patients
are shown in Table 1.
No statistical differences were found between the SVP and WT
groups regarding patient age, sex, body mass index (calculated as
weight in kilograms divided by height in meters squared), and American
Society of Anesthesiology score. However, the mean (SD) tumor size
was significantly higher in the WT group (33.6 [19.7] mm vs 42.5 [29.9]
mm; P < .001).
Of the 140 LSPDPs performed, the most common diagnoses were
cystic neoplasms in 53 patients (37.9%) and neuroendocrine neoplasms
in 44 patients (31.4%), followed by intraductal papillary mucinous
neoplasms in 22 patients (15.7%). One patient in the SVP group had
an in situ intraductal adenocarcinoma. Other histologic diagnoses
in the WT group were epithelial cysts (n = 2), an intrapancreatic
spleen (n = 1), nesidioblastosis (n = 1), and
intraductal dystrophy (n = 1).
Between the SVP and WT groups, the mean (SD) operative time
(214.7 [66.7] minutes vs 199.2 [46.7] minutes; P = .11) and mean (SD) blood loss (342.8 [223.5 mL vs 288.9
[172.6] mL; P = .11) did not differ
(Table 2). The rates of conversion
to laparotomy were also similar (9% vs 13%; P = .48).
Splenic preservation represented 89.3% of all the LSPDPs. The
success rate of spleen preservation was significantly higher with
the SVP procedure (96.4% vs 84.7%;P = .03).
The splenic vessels could be preserved in 41 patients (75%). Two patients
(3.6%) required splenectomy for splenic decapsulation (n = 1)
and adhesions (n = 1), and 12 patients (21.8%) required
conversion to WT owing to splenic vessel injuries (n = 8)
and adhesions of the tumor to the splenic vessels (n = 4).
In the WT group, the reasons for splenectomy included splenic infarction
(n = 4), bleeding (n = 3), adhesions (n = 3),
tumor located close to the splenic hilum (n = 2), and tumor
with an unknown location (n = 1). The occurrence of intraoperative
complications requiring splenectomy (bleeding and splenic infarction)
did not differ between both groups (2% vs 8%; P = .15).
A summary of the morbimortality data of the cohort is shown
in Table 3. No deaths were
recorded during the 30-day period after surgery. The overall morbidity
rate related to all of the LSPDP procedures was 34.3%, and the rate
of PF was 22.1%.
Postoperative complications occurred in 15 patients (27.3%)
in the SVP group and 33 patients (38.8%) in the WT group (P = .16). The rate of minor and major
complications did not differ between the groups (P = .17).
The rate of PF was similar between the SVP and WT groups (16.3%
vs 25.9%; P = .18).
Spleen-related complications occurred only in the WT group.
Nine patients (10.5%) presented with symptomatic postoperative splenic
infarction in the WT group compared with none of the patients in the
SVP group (P = .03). Of the 9 patients
who presented with symptoms such as fever, abdominal pain, and/or
abscess, 4 (4.7%) required a splenectomy. In the institution that
performed systematic postoperative imaging examinations, 7 asymptomatic
patients were diagnosed and treated conservatively.
Three patients from the SVP group required a reoperation owing
to a small bowel perforation on day 1, bleeding of an inferior pancreatic
artery on day 2, and perforated duodenal ulcer on day 4 after surgery.
Five patients from the WT group required reoperation. A splenectomy
was performed owing to splenic infarction in 3 patients on day 7 and
in 1 patient on day 45. One patient required a reoperation 10 days
after surgery because of hemorrhagic shock owing to a grade C PF.
The mean (SD) hospital stay was statistically shorter for patients
who underwent LSPDP with SVP (8.2 [3.1] days vs 10.5 [6.6] days; P = .01).
Based on a nonintention-to-treat analysis of actual procedures
performed, the postoperative complications were also compared between
the 41 patients treated with SVP and the 84 patients treated with
WT (Table 4). The overall morbidity
rate (24.4% vs 38.1%; P = .16),
the rate of PF (14.6% vs 25%; P = .25),
and the rate of reoperation (7.3% vs 5.9%; P = .72) were similar. This analysis confirmed that patients
treated with WT had more spleen-related complications (0% vs 10.7%; P = .03) and an increased length of hospital
stay (7.9 [2.8] days vs 10 [6.4] days; P < .05).
The purpose of this study was to compare the short-term outcomes
between the preservation and the division of the splenic vessels during
LSPDP in patients with benign or low-grade malignant tumors in the
body/tail of the pancreas. This study represents one of the largest
series of distal pancreatectomies with splenic conservation performed
using a laparoscopic approach and, to our knowledge, the first study
to compare the division procedure with the preservation of the splenic
vessels using an intention-to-treat analysis and with a significant
number of patients.
Data related to this laparoscopic procedure and conservation
of the spleen are scarce. Our results from 140 LSPDPs confirm the
feasibility, safety, and efficiency previously described in the literature.17,20-24 The overall rates of morbidity and PF (34.3% and 22.1%, respectively),
as well as the rate of conversion to laparotomy (11.4%) were similar
to the rates described for laparoscopic distal pancreatectomy.
In patients with nonmalignant lesions in the body/tail of the
pancreas, the preservation of the spleen is preferable for avoiding
long-term complications related to a splenectomy,9-11 and several
studies have reported the benefits of LSPDP compared with laparoscopic
distal pancreatectomy in terms of intraoperative outcomes, postoperative
collections and infections, and length of hospital stays.20,25
Between both techniques available to preserve the spleen, WT
was reported to be faster and easier,3,17,20 but few
studies are available in the literature to determine which technique
is the most preferable for LSPDP (Table
5). Most of these studies included a small number of patients,
described results from both open and laparoscopic approaches, and
only reported outcomes from procedures that actually succeeded in
preserving the spleen. To our knowledge, our study is the largest
to compare WT with laparoscopic SVP with respect to the intention-to-treat
Based on our study, both techniques were similar regarding the
rates of overall morbidity, PF, and reoperation. Blood loss and operative
times were also similar contrary to what has been previously reported.17
The rate of spleen conservation was significantly higher in
patients who underwent the intention-to-treat SVP. Preservation of
the splenic vessels was successful in 75% of the cases. In 22% of
cases, splenic vessel injuries or difficulties during the dissection
required a conversion to WT. This procedure is technically challenging
and may not be possible if the tumor is located close to the splenic
hilum, the vessels are embedded in the pancreatic gland, or local
inflammation is present. However, failing to spare the splenic vessels
does not necessarily compromise the conservation of the spleen. The
spleen was able to be salvaged in a large proportion of patients by
conversion to WT. Regarding the intraoperative complications that
might have been attributable to surgical differences, the occurrence
of bleeding or splenic infarction, requiring en-bloc splenectomy,
did not differ between the 2 techniques.
Similarly to previous reports, spleen-related complications
were recorded after undergoing WT (Table
5). The division of the splenic vessels represents a potential
risk for splenic infarction when the blood supply to the spleen is
not sufficiently recovered from the short gastric and left gastroepiploic
vessels. With this procedure, 10.7% of our patients developed symptomatic
splenic infarcts, which required reoperation for splenectomy in around
half of these patients (4.7%). In a recent retrospective analysis
with long-term follow-up on 158 patients who had undergone open spleen-preserving
distal pancreatectomies using WT, the rate of reoperation for splenectomy
owing to splenic infarction was 1.9%. Among the 65 patients with postoperative
imaging, 23% had splenic infarcts and 25% had perigastric varices.29 It is still unclear whether postoperative
control of splenic perfusion must be performed in asymptomatic patients
undergoing LSPDP with WT. The occurrence of perigastric varices owing
to the excision of the splenic artery and vein with the potential
risk for bleeding could represent an additional argument for delayed
systematic postoperative imaging controls. Sinistral portal hypertension30 has been reported in other publications
that evaluated hemodynamic changes in splenogastric circulation after
ligation of the splenic vessels.31,32 Miura et al31 reported a 70% rate of perigastric varices
and a 20% rate of submucosal varices after WT, as well as 1 patient
with gastrointestinal bleeding.
The length of hospital stay was significantly shorter in patients
who underwent WT. This was already mentioned in a previous study26 and could be partially explained by the
occurrence of spleen-related complications associated with WT. Other
complications, especially PFs, did not differ between the techniques.
Furthermore, patients with splenic infarcts sometimes required a readmission
and reoperation for splenectomy, which increased the overall length
of hospital stay and affected the cost-effectiveness of the procedure.
Regarding the adverse events related to division, preservation
of the splenic vessels appears to be the preferable technique. Both
techniques are comparable in terms of intraoperative and postoperative
morbidity, but the SVP procedure provided the best chance to conserve
the spleen. We think that this procedure should be attempted when
possible and switched to WT in cases of accidental bleeding or difficulties
Biases cannot be excluded based on the retrospective nature
of this study. The mean tumor size was significantly higher in patients
who underwent sacrifice of the splenic vessels, which suggests that
preservation of the vessels might have been more technically challenging
in these cases. Prospective studies are necessary to confirm and validate
these results through long-term follow-up.
In conclusion, this large, bicentric retrospective study confirms
that LSPDP is safe, feasible, and efficient. The comparison of both
techniques for spleen conservation showed that the preservation of
the splenic vessels was associated with an improved rate of spleen
conservation and a reduced hospital stay without increased morbidity.
The spleen-related complications related to WT should lead to a surgical
preference for the laparoscopic spleen and splenic vessel-preserving
distal pancreatectomy in selected patients with benign or low-grade
malignant tumors in the body/tail of the pancreas.
Correspondence: Jean-Philippe Adam,
Service de Chirurgie Digestive, Hôpital Saint-André,
1 rue Jean Burguet, 33075 Bordeaux, France (email@example.com).
Accepted for Publication: September
Published Online: November 19, 2012.
Author Contributions:Study concept and design: Adam, Jacquin, Fernández-Cruz,
and Sa-Cunha. Acquisition of data: Adam,
Jacquin, Laurent, Collet, Masson, and Fernández-Cruz. Analysis and interpretation of data: Adam, Jacquin,
and Sa-Cunha. Drafting of the manuscript:
Adam and Sa-Cunha. Critical revision of the manuscript
for important intellectual content: Adam, Jacquin, Laurent,
Collet, Masson, Fernández-Cruz, and Sa-Cunha. Statistical analysis: Adam and Jacquin. Administrative, technical, and material support: Laurent and
Collet. Study supervision: Masson, Fernández-Cruz,
Conflict of Interest Disclosures: None
Previous Presentations: This article
was presented in part at the 7th French Congress of Digestive and
Hepatobiliary Surgery; November 30, 2011; Paris, France; at the Annual
French Meeting of Hepatogastroenterology and Digestive Oncology; March
15, 2012; Paris, France; and at the 10th World Congress of the International
Hepato-Pancreato-Biliary Association; July 5, 2012; Paris, France.
Create a personal account or sign in to: