Perioperative and oncological outcome of laparoscopic resection of gastrointestinal stromal tumour (GIST) of the stomach.

Abstract

Background. Surgery remains the only curative treatment for gastrointestinal stromal tumour (GIST). Resection needs to ensure tumour-free margins while lymphadenectomy is not required. Thus, partial gastric resection is the treatment of choice for small gastric GISTs. Evidence on whether performing resection laparoscopically compromises outcome is limited. Methods. We compiled patients undergoing laparoscopic resection of suspected gastric GIST between 2003 and 2007. Follow-up was performed to obtain information on tumour recurrence. Results. Laparoscopic resection with free margins was performed in 21/22 patients. Histology confirmed GIST in 17 cases, 4 tumours were benign neoplasms. Median operation time and postoperative stay for GIST patients were 130 (range 80-201) mins and 7 (range 5-95) days. Two patients experienced stapler line leakage necessitating surgical revision. After median follow-up of 18 (range 1-53) months, no recurrence occurred. Conclusions. Laparoscopic resection of gastric GISTs yields good perioperative outcomes. Oncologic outcome needs to be assessed with longer follow-up. For posterior lesions, special precaution is needed. Laparoscopic resection could become standard for circumscribed gastric GISTs if necessary precautions for oncological procedures are observed.

Full text

Page 1

Hindawi Publishing Corporation
Diagnostic and Therapeutic Endoscopy
Volume 2009, Article ID 286138, 7 pages
doi:10.1155/2009/286138
Clinical Study
Perioperativeand OncologicalOutcomeof
LaparoscopicResectionof Gastrointestinal Stromal
Tumour(GIST) of the Stomach
UlrichRonellenfitsch,1,2Wilko Staiger,2Georg K¨ ahler,2Philipp Str¨ obel,3
Matthias Schwarzbach,2and PeterHohenberger1
1Division of Surgical Oncology and Thoracic Surgery, University Medicine Mannheim, Ruprecht-Karls-University Heidelberg,
Theodor-Kutzer-Ufer 1-3, 68135 Mannheim, Germany
2Department of Surgery, University Medicine Mannheim, Ruprecht-Karls-University Heidelberg, Theodor-Kutzer-Ufer 1-3,
68135 Mannheim, Germany
3Institute of Pathology, University Medicine Mannheim, Ruprecht-Karls-University Heidelberg, Theodor-Kutzer-Ufer 1-3,
68135 Mannheim, Germany
Correspondence should be addressed to Peter Hohenberger, peter.hohenberger@chir.ma.uni-heidelberg.de
Received 29 October 2008; Accepted 23 January 2009
Recommended by Yuri W. Novitsky
Background. Surgery remains the only curative treatment for gastrointestinal stromal tumour (GIST). Resection needs to ensure
tumour-free margins while lymphadenectomy is not required. Thus, partial gastric resection is the treatment of choice for small
gastricGISTs.Evidenceonwhetherperformingresectionlaparoscopicallycompromisesoutcomeislimited. Methods.Wecompiled
patients undergoing laparoscopic resection of suspected gastric GIST between 2003 and 2007. Follow-up was performed to obtain
information on tumour recurrence. Results. Laparoscopic resection with free margins was performed in 21/22 patients. Histology
confirmed GIST in 17 cases, 4 tumours were benign neoplasms. Median operation time and postoperative stay for GIST patients
were130(range80–201)minsand7(range5–95)days.Twopatientsexperiencedstaplerlineleakagenecessitatingsurgicalrevision.
After median follow-up of 18 (range 1–53) months, no recurrence occurred. Conclusions. Laparoscopic resection of gastric GISTs
yields good perioperative outcomes. Oncologic outcome needs to be assessed with longer follow-up. For posterior lesions, special
precaution is needed. Laparoscopic resection could become standard for circumscribed gastric GISTs if necessary precautions for
oncological procedures are observed.
Copyright © 2009 Ulrich Ronellenfitsch et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
1.Introduction
Gastrointestinal stromal tumours (GISTs) are the most
common mesenchymal neoplasms of the digestive tract
with an estimated annual incidence of 10–20 cases per
one million inhabitants [1, 2]. GISTs probably arise from
precursor cells of the interstitial cells of Cajal. Their defining
characteristic is a gain-of-function mutation in genes coding
for the KIT tyrosine kinase receptor, which is considered
the driving force of cell proliferation in this tumour [3].
Clinical presentation of GISTs ranges from indolent, hardly
proliferating to fast-growing, recurring and metastasising
tumours [1]. Fletcher et al. proposed a classification of
aggressive behaviour for GISTs based on their maximum
diameter and mitotic rate [4] (Table 1), factors which were
both shown to predict recurrence and survival [5, 6].
Treatment of choice for primary GISTs remains complete
resection. Whereas current National Comprehensive Cancer
Network (NCCN) guidelines [7] recommend surgery for
GISTs or supposed GISTs of any size, the latest European
Society for Medical Oncology (ESMO) Clinical Recommen-
dations stipulate frequent surveillance without surgery for
lesions with a diameter below 2cm [8]. In contrast to
resection of intestinal carcinomas, surgery of GISTs does
not require lymphadenectomy since lymphatic metastatic
spread is extremely rare in primary tumours [1]. Thus, local

Page 2

2Diagnostic and Therapeutic Endoscopy
Table 1: Classification of aggressive behaviour of GISTs proposed
by Fletcher et al. [4].
Tumour size
(largest
diameter)
<2cm
2–5cm
<5cm
5–10cm
>10cm
any size
>5cm
Mitotic count per
50 high power
fields
<5
<5
6–10
<5
any number
>10
>5
Very low risk
Low risk
Intermediate risk
High risk
resection of the tumour with clear margins is recommended.
Moreover, strict avoidance of intraoperative tumour rupture
is crucial for preventing tumour relapse.
About 50% of GISTs are located in the stomach which
makes it the most frequent site of manifestation [2, 9]. Due
to the often fragile consistence, particularly of pedunculated
GISTs, there is an ongoing debate whether surgical resection
of gastric GISTs can be performed laparoscopically without
increasingperioperativemorbidityandcompromisingonco-
logical outcome. The latest ESMO Clinical Recommenda-
tions consider a laparoscopic approach “if cancer surgery
principles are respected.” [8] Current NCCN guidelines
do not contain a clear statement on whether surgery for
GIST should be performed laparoscopically or through
open surgery but recommend that surgery should produce
minimal surgical morbidity [7].
The present study tries to evaluate whether laparoscopic
resection of gastric GISTs can become a standard treatment
for such tumours by analysing perioperative characteristics
and long-term oncological outcome.
2.MaterialandMethods
2.1. Study Population and Data Analysis. The study includes
all patients who underwent laparoscopic resection of a pri-
marytumourofthestomachdeemedtobeaGISTonclinical
assessment between January 1, 2003, when a laparoscopic
approach became our standard for the described lesions, and
December 31, 2007. Patients were eligible for laparoscopic
surgery if preoperative staging (endoscopy, endosonography
and CT scan) showed a localised, non-metastatic extra-
mucosal gastric lesion. Tumours were required to be of a
diameter and in a position which suggested resectability
through segmental or wedge resection. Preoperative histo-
logical confirmation of the diagnosis was not a prerequisite
for inclusion. Histological diagnosis of GIST was obtained
from the resection specimen by means of hematoxylin and
eosin staining and immunohistological assays for CD117
and CD34 and platelet-derived growth factor receptor alpha.
Mutational analysis was performed where required.
From a prospectively kept database, we extracted the
following characteristics: age, sex, length of postoperative
hospital stay, operation time, tumour location, tumour
size, classification of aggressive behaviour as defined in
Table 1, histopathological assessment of resection margins,
intraoperative blood loss, and incidence of perioperative
complications (cardiac, pulmonary, septic, anastomotic fail-
ure, reoperation needed). We present single values as well as
the median plus range or, where applicable, percentages, for
the respective variables.
All patients with confirmed GIST were followed up
regularly including upper GI endoscopy and abdominal CT
scans every three to six months depending on the risk for
malignant behaviour. From these visits we assessed vital
status and tumour recurrence. A few patients with very low-
risk GIST decided against adhering to this program and were
followed up through phone calls to their general practitioner
or themselves. In the former case, the family physician was
asked when the patient had last presented and if to his
knowledge any recurrence of the tumour was known. The
same question was asked directly to patients in case their
family physician was not available. Date of follow-up was
ascertained either as the date of the patient’s last visit to our
outpatient clinic or the date of the phone call to the family
physician or patient, respectively.
2.2. Surgical Methodology. All laparoscopic resections were
performed in a standardised manner. The patient was
placed in supine position with legs spread and a four-port
technique was used. After visual and tactile control of the
liver and the abdominal cavity for metastases, the lesion was
identified through visualisation and palpation. In case of
tumour location at the posterior wall of the stomach, the
gastrocolic ligament was dissected and the stomach inverted
in order to display the lesion. If necessary, intraoperative
endoscopy was performed to elevate the tumour and stain
its margins with ink (Figure 1). The tumour-bearing gastric
segment was resected with one or several 45 mm endoscopic
linear staplers. The resection specimen was placed into
a plastic retrieval bag and removed through one of the
port incisions. Stapler lines were visually controlled and, if
deemednecessarybythesurgeon,theirintegritywasassessed
through the application of methylene blue via a nasogastric
tube. If the lines showed leakage or were deemed to be at risk
for it, additional manual sewing of the line was performed. A
postoperative control of stapler lines by means of endoscopy
or CT scan after oral intake of contrast medium was only
done in case of clinical suspicion of leakage.
3.Results
22 patients had been planned for laparoscopic segment
resection of a suspected gastric GIST. Out of these, 21 were
resected with the technique. In one patient, conversion to
laparotomy was necessary due to extensive intraabdominal
adhesions with the larger omentum completely fixed to the
ventral abdominal wall. This patient was not included in the
analysis. One of the 21 patients had received preoperative
imatinib treatment for a GIST with a diameter of 8cm. He
had adverse characteristics for open surgery (pronounced
obesity) and was reluctant to undergo any resection at first.

Page 3

Diagnostic and Therapeutic Endoscopy3
Figure 1: Endoscopic ink staining of tumour margins.
Neoadjuvant imatinib caused a reduction of tumour size
which enabled a laparoscopic approach to which the patient
finally agreed. Histology confirmed a GIST in 17 of 21
cases. In one case histological workup showed pancreatic
heterotopy, in one case leimoyoma, in one case gastric
schwannoma, and in one case gastric wall lipoma.
Table 2 provides perioperative characteristics of the 17
GIST patients in which the intervention could be performed
laparoscopically. Both sexes were equally represented and
most patients were in their sixth or seventh decade of life
(age range 43–79 years). Tumours were located in all parts of
the stomach with a predominance of the corpus and antrum.
All tumours were resected with negative margins and there
was no intraoperative tumour rupture. Intraoperative blood
loss was below 200mL in all patients and no patient required
blood transfusion. The median duration of surgery was
130 (range 80–201) minutes and the median postoperative
hospital stay 7 (range 5–95) days. In two patients, a
postoperative complication occurred: one early stapler line
leakage requiring laparotomy for re-suturing and one late
stapler line leakage which led to a prolonged hospital stay
and finally resulted in Billroth II gastrectomy. Both patients
were obese (BMI > 30) and in both cases the tumour
was located on the posterior stomach wall. In the former
case, the leakage occurred on postoperative day 1 when the
patient presented with severe abdominal pain and preseptic
conditions. Diagnosis was made by application of methylene
blue through the nasogastric tube and its detection through
the indicative drainage which was still in place. The patient
was immediately re-operated through a small laparotomy
and the further clinical course was uneventful except for
subcutaneous wound infection. The latter patient was first
treated with re-laparoscopy on postoperative day 6 for
removal and drainage of an intraabdominal abscess. On
postoperative day 10, leakage was detected endoscopically
and treated with endoluminal stenting. After gradual clinical
improvement the patient was discharged on day 61 but was
re-admitted few days later with peritonitic signs. Releakage
was diagnosed and open subtotal gastrectomy had to be
performed. Finally, the patient was discharged in good
clinical condition 95 days after initial surgery.
Table 3 shows the results of the follow-up. After a median
period of 18 (range 1–53) months, in none of the patients
recurrence or distant metastases had been detected and all
patients were alive. In patients with an intermediate risk
GIST, median follow-up was 18 (range 6–46) months.
4.Discussion
The main principle of curative surgery for GISTs is en bloc
resection with negative tumour margins and strict avoidance
of intraoperative tumour rupture. Due to the extremely
low frequency of lymphatic metastasis, lymphadenectomy
is not required. Thus, segmental or wedge resection is the
treatment of choice for tumours whose size and location
technically allow for it [10, 11]. The aims of the present
analysis were to assess if the mentioned surgical principles
couldbesufficientlymetwithalaparoscopicapproachandto
evaluate perioperative and oncological outcomes of patients
resected with this method.
In line with previously published case series [12–28],
our results support the application of laparoscopy for wedge
resection of gastric GISTs. In all but one case deemed eligible
for the procedure based on staging exams, the intervention
could be performed without conversion to laparotomy.
Resection with tumour-free resection margins was possible
in all cases and there were no instances of intraoperative
tumour rupture. Although there are no respective empirical
data, we consider the usage of a retrieval bag for the
removal of the surgical specimen essential in order to
avoid spillage of tumour cells into the abdominal cavity
or port sites, thus preventing metastasis. Even small GIST
of 2-3cm in size harbour the risk of malignant behaviour
and consequently tumour resection should be performed
according to standards of laparoscopic resection for GI
malignancies.
In our series, overall perioperative morbidity was low
with virtually no blood loss and satisfying operation times.
As in other laparoscopic procedures, the existence of a
“learningcurve”mustbeassumedanditcanbeexpectedthat
operation times further decrease with growing experience
[29–31]. The median length of hospital stay of our patients
was relatively short, too. It can be speculated that patients
could have been discharged even earlier and that the
relatively lengthy stay was attributed to the rather recent
introduction of the methodology at our centre.
In two patients, severe postoperative morbidity occurred
duetoleakagefromthestaplerline.Inbothcases,thetumour
was partially located at the posterior wall of the stomach,
and both patients were obese (BMI > 30). In such patients,
laparoscopic resection is particularly challenging because it
requires extended mobilisation of the stomach and prepara-
tion through the omental bursa, hampered by impaired visi-
bility of the operation field. Technical difficulties in stapling
and/or suturing might be the consequence. Both leakages
led us to the routine use of fleece patches to strengthen the
suture line and we strongly recommend their application
in patients with the characteristics described. An alternative
for the resection of lesions of the posterior gastric wall,

Page 4

4Diagnostic and Therapeutic Endoscopy
Table 2: Perioperative and tumour characteristics of the patients.
Patient Sex Age
Tumour
localisation
Histology
Max. tumour
diameter
(cm)
5
4
Mitotic
figures/50
HPF
<5
2
Risk
classification§
Duration of
surgery
(minutes)
165
94
Postop.
hospital stay
Postoperative
morbidity
1
2
m
m
45
79
fundus
fundus
anterior
corpus
greater cur-
vature/ant.
corpus
lesser
curvature
GIST
GIST
low
low
5
7
none
none
3∗
f 43
GIST
0.8
<5 very low895none
4f 56
GIST
1.75 very low1756 none
5f 55
GIST
5.42intermediate 1136none
6m 59
posterior
antrum
GIST
3.53 low16195
late stapler
line leakage
resulting in
B-II-
gastrectomy
7f 74
lesser
curvature
anterior
antrum
GIST
2.1
<5 low1307none
8m 45
GIST
2.51low1736none
9f72
greater cur-
vature/post.
corpus
GIST
5.1
<5intermediate33312
early stapler
line leakage
resulting in
resuturing
through
laparotomy
10f52
anterior
corpus
greater cur-
vature/ant.
corpus
antrum
lesser
curvature
posterior
corpus
anterior
antrum
lesser
curvature
anterior
corpus
n/a
GIST
2
<2low1256none
11f46
GIST
2.13low807none
12#
f66GIST
2.92low1129 none
13m82
GIST
10
<5intermediate185 14none
14m 62
GIST
2
<2very low2018none
15m64
GIST
6
<5intermediate1056 none
16∗
f 67
GIST
1.8
<5very low18410none
17$
m66
GIST
4n/a n/a1438none
median n/a 56n/a
2.9n/an/a1307n/a
n/a: not applicable.
classification of aggressive behaviour not possible.§see Table 1.
∗Additional cholecystectomy for cholecystolithiasis.#Additional liver cyst deroofing.$Due to preoperative imatinib treatment
especially for lesions located close to the gastro-esophageal
junction, could be a transgastric approach through anterior
gastrotomy [15, 20, 32–34]. This technique, however, is
technically even more demanding [20], and there are reports
of postoperative complications [32] and incomplete tumour
resections [35]. As a complementary method, a combined
endoscopic-laparoscopic approach can be used. We have
applied this technique in several cases to facilitate tumour
identification and lifting into the stapler predominantly in
relatively small and non-protruding lesions. Its utility for
resection of gastric GIST, especially for tumours in the
proximal posterior part of the stomach, has been reported
in several series [28, 36, 37]. Recently, a new technique of
endoscopic full-thickness resection using a flexible stapler
was described. This approach seems particularly useful in
tumours of the posterior distal part of the stomach [38].
In summary, we do not consider any tumour location as
a strict contraindication towards laparoscopy if the necessary

Page 5

Diagnostic and Therapeutic Endoscopy5
Table 3: Results of the follow-up of operated patients.
Patient SexAge
Classification of
aggressive behaviour§
Follow-up
(months)
Tumour recurrence,
metastases or death at
end of follow-up?
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
n/a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17$
m
m
f
f
f
m
f
m
f
f
f
f
m
m
m
f
m
n/a
45
79
43
56
55
59
74
45
72
52
46
66
82
62
64
67
66
59
low
low
44
40
18
27
23
14
9
6
6
53
10
12
46
47
18
1
5
18
very low
very low
intermediate
low
low
low
intermediate
low
low
low
intermediate
very low
intermediate
very low
intermediate/high
n/a
median
n/a: not applicable;§see Table 1;$received preoperative imatinib treatment. Aggressive behaviour classified based on pre-treatment staging.
experience is given and if the required precautions are
met. Several additional technical approaches can be used
to facilitate safe resection. Nevertheless, it is important to
emphasizethatthethresholdforlaparotomyshouldberather
low if intraoperative difficulties are encountered.
One concern about wedge resection of the stomach is
the occurrence of postoperative gastric stenosis, especially
when parts of the gastro-oesophageal junction or the pylorus
are resected [17, 20, 23]. In our patients, we did not find
any early postoperative stenosis. There were, however, only
six cases where the tumour was located in the antrum
and none with a tumour directly at the gastro-oesophageal
junction. With regard to long-term functional results,
we were not confronted with symptoms or endoscopical
signs of reflux or stenosis during postoperative follow-
up.
For tumours with a larger diameter and/or unfavourable
location, primary wedge resection is often not possible and
total or subtotal gastrectomy would be required for resection
with tumour-free margins. For these cases, the NCCN
guidelines [7] and ESMO recommendations [8] suggest
neoadjuvant imatinib therapy to decrease tumour size, thus
allowing for organ-preserving surgery. The feasibility and
outcomes of this approach are currently evaluated in several
clinical trials [39]. Our case series includes one patient who
receivedsixmonthsofneoadjuvantimatinibtreatmentinthe
framework of the C STI571 BDE 43 trial, (“Apollon study”)
[11]. Laparoscopic resection of an originally large GIST of
the anterior corpus was made possible in this case thanks to
a considerable shrinkage of the tumour.
In our series, 4 out of 21 tumours preoperatively
suspected to be a GIST were histologically diagnosed as
pancreatic heterotopia, gastric lipoma, schwannoma, and
leiomyoma. Retrospectively, in these cases surgical resection
would have not been required for oncological reasons.
Preoperative histological diagnosis of submucosal gastric
lesions is however not always feasible through endoscopic
biopsy. Moreover, biopsy seems to be associated with a
certain risk of tumour haemorrhage and dissemination [1].
Modern imaging techniques such as endosonography or
CT gastrography with 3D reconstruction [40] can aid in
making the preoperative diagnosis without being invasive
towards the tumour. Nevertheless, a clear preoperative
diagnosis will still not be possible in all cases of submucosal
gastric tumours. We think that it is warranted to perform
surgical excision of submucosal gastric lesions without prior
histological ascertainment even if in a small percentage of
cases the tumour is not a GIST. In fact, the latest ESMO
recommendations [8] explicitly consider surgical excision of
tumours without prior histological confirmation of GIST
if they are larger than 2cm or show an increase in size.
Current NCCN guidelines even regard preoperative biopsy
as not appropriate in easily resectable lesions and state that it
is mandatory only if neoadjuvant treatment is planned [7].
Three of the four benign lesions were at a size of greater
than 2cm and by this met the criteria specified in both
guidelines.
Follow-up did not show any local recurrence or distant
metastases, and all patients were alive at the end of the
follow-up period. This finding is similar to results from