Keywords
Wilms tumor - treatment - surgery - complications
Introduction
Wilms tumor (WT) is the most common kidney tumor in children. The annual incidence
of WT is between 7 and 10 cases per million in children younger than 15 years and
represents the fourth malignancy in children. Outcomes of children with WT have improved
drastically over the past decades, thanks to the contributions of the two major cooperative
working groups—the National Wilms' Tumor Study–Children's Oncology Group (NWTS-COG)
and the Société Internationaled' Oncologie Pédiatrique (SIOP) group. The overall survival
(OS) in developed countries is around 90%.[1] The different approaches adopted by the two cooperative groups, upfront nephrectomy
by COG and a preoperative chemotherapy by SIOP,[1] have not been translated in outcome differences in most recent studies (NWTS-4 and
SIOP 9).[2]
[3] Recognizing the excellent OS expected for most patients with WT, the research focus
has turned into decreasing treatment-associated morbidity.[4]
Ten years ago, we adopted the COG strategy of upfront surgery following diagnosis.
The patients were treated consecutively by a group of pediatric surgical oncology
surgeons in a referral pediatric university hospital. The aim of this study is to
(1) evaluate the outcome of patients with WT treated with upfront surgery in our institution
and (2) analyze predictive markers of surgical complications that could help with
improving treatment-associated complications and morbidity.
Methods
A retrospective review of medical records of WT patients treated in our institution
between 2003 and 2013 was performed; the review was done after a waiver of authorization
was obtained from our institutional review board. We excluded patients, who received
any preoperative treatment.
Following the COG protocol, a computed tomography (CT) of the thorax, abdomen, and
pelvis was performed as preoperative study. Radical upfront nephrectomy was performed
through a transverse laparotomy. After examining the abdominal cavity, the renal vessels
were exposed and ligated. The adrenal gland was only removed if it was adherent to
the tumor and to the ureter ligated as low as possible. If the tumor had infiltrated
local structures, they were excised in continuity with the tumor. Sampling of the
lymph nodes of the renal hilum and paraaortic area was always performed. The COG staging
system was used to assess tumor stage and the appropriate treatment was established
according to the contemporary COG protocols. Demographic data, clinical symptoms,
tumor size, tumor biology, surgical outcomes, and surgical complications based on
the Clavien–Dindo classification were analyzed. The therapy used to correct a specific
complication is the basis of this classification to rank a complication in an objective
and reproducible manner. It consists of seven grades (I, II, IIIa, IIIb, IVa, IVb,
and V). The introduction of the subclasses a and b allows a contraction of the classification
into five grades (I, II, III, IV, and V) ([Table 1]).[5]
Table 1
Clavien–Dindo classification
|
Grades
|
Definition
|
|
Grade I
|
Any deviation from the normal postoperative course without the need for pharmacological
treatment or surgical, endoscopic, and radiological interventions
Allowed therapeutic regimens are: drugs as antiemetics, antipyretics, analgetics,
diuretics, and electrolytes, as well as physiotherapy. This grade also includes wound
infections opened at the bedside
|
|
Grade II
|
Requiring pharmacological treatment with drugs other than such allowed for grade I
complications
Blood transfusions and total parenteral nutrition are also included
|
|
Grade III
|
Requiring surgical, endoscopic, or radiological intervention
|
|
IIIa
|
Intervention not under general anesthesia
|
|
IIIb
|
Intervention under general anesthesia
|
|
Grade IV
|
Life-threatening complication (including CNS complications)* requiring IC/ICU management
|
|
IVa
|
Single organ dysfunction (including dialysis)
|
|
IVb
|
Multiorgan dysfunction
|
|
Grade V
|
Death of a patient
|
|
Suffix “d”
|
If the patients suffer from a complication at the time of discharge, the suffix “d”
(for “disability”) is added to the respective grade of complication. This label indicates
the need for a follow-up to fully evaluate the complication
|
Abbreviations: CNS, central nervous system; ICU, intensive care unit.
We used Statistical Package for the Social Sciences (SPSS) 22.0 version (SPSS, Inc.,
Chicago, Illinois, United States) for the statistical analysis. The normality of the
variables was tested using the Shapiro–Wilk test. Normal variables were expressed
as mean and standard deviation. Nonnormal variables were expressed as median and range.
Qualitative variables were expressed as frequencies and percentages. Chi-square and
Fisher's tests were used to compare results between different patient's groups. Statistical
significance was set at p < 0.05. The event-free survival (EFS) and OS were assessed using the Kaplan–Meier
method and the log-rank test was used to compare differences between variables.
Results
Patient characteristics are outlined in [Table 2]. Forty-five patients with WT were treated with upfront nephrectomy. The mean age
at diagnosis was 3.8 years (45.3 ± 31.3 months). Twenty-four (53%) patients were female
and 21 (47%) were male. The most common clinical symptom at presentation was the presence
of an abdominal palpable mass in 21 (47%) patients. Nineteen (42%) patients presented
with abdominal pain or discomfort, 11 (24%) with fever, and 6 (13%) had macroscopic
hematuria. Two (4%) patients had WAGR (Wilms tumor, Aniridia, Genitourinary problems and Range of developmental delays) syndrome and one (2%) Beckwith–Wiedemann syndrome.
Table 2
Patient's characteristics
|
Gender
|
|
Female
|
24
|
53%
|
|
Male
|
21
|
47%
|
|
Age at diagnosis
|
3.8 y (45.3 ± 31.3 mo)
|
|
Clinical presentation
|
|
Palpable mass
|
21
|
47%
|
|
Abdominal pain or discomfort
|
19
|
42%
|
|
Fever
|
11
|
24%
|
|
Hematuria
|
6
|
13%
|
|
Syndromes
|
|
WAGR
|
2
|
4%
|
|
Beckwith–Wiedeman
|
1
|
2%
|
Tumor characteristics and staging are presented in [Table 3]. Thirteen (29%) patients had stage I, 12 (27%) stage II, 15 (33%) stage III, and
5 (11%) had stage IV WT. All stage IV patients had isolated lung metastases. The mean
tumor size was 10.8 cm (3.5–16 cm). The most common histological subtype in 75% of
cases was the triphasic pattern. Six patients (13%) showed anaplasia. Lymph node involvement
was found in a total of five cases (11%).
Table 3
Tumor's characteristics
|
Laterality
|
|
Left
|
21
|
47%
|
|
Right
|
24
|
53%
|
|
Size
|
10.8 cm (3.5–16 cm)
|
|
Stage
|
|
I
|
13
|
29%
|
|
II
|
12
|
27%
|
|
III
|
15
|
33%
|
|
IV
|
5
|
11%
|
|
Histology
|
|
Triphasic
|
34
|
76%
|
|
Epithelial
|
1
|
2%
|
|
Blastemal
|
7
|
16%
|
|
Stromal
|
1
|
2%
|
|
Mixed
|
2
|
4%
|
|
Anaplasia
|
|
Focal anaplasia
|
4
|
9%
|
|
Diffuse anaplasia
|
2
|
4%
|
|
Lymph node involvement
|
5
|
11%
|
|
Positive margins
|
3
|
7%
|
|
Spontaneous tumor rupture
|
5
|
11%
|
Surgical outcome: gross total resection (GTR) was achieved in 42 (93%) cases. Spontaneous
tumor rupture was found in five (11%), three of them with an abdominal trauma prior
to diagnosis and hemoperitoneum at the time of surgery. In one case, partial colectomy
was necessary because compression of the tumor caused by poor vascularization to the
cecum. One case required a distal pancreatectomy. Two (4%) cases had an intraoperative
tumor rupture ([Table 4]).
Table 4
Surgical complications
|
n
|
%
|
Clavien–Dindo
|
|
Intraoperative rupture
|
2
|
4
|
Nonapplicable
|
|
Other organ resection
|
2
|
4
|
Nonapplicable
|
|
Blood transfusion
|
20
|
44
|
Grade II
|
|
Chylous ascites
|
1
|
2
|
Grade II
|
|
Fever
|
1
|
2
|
Grade I
|
|
Oliguria
|
2
|
4
|
Grade I
|
We used the Clavien–Dindo classification to assess surgical complications. Two (4%)
patients had grade I complications that required diuretic treatment to treat oliguria.
Twenty-one (47%) patients had grade II complications including 20 cases (44%) of blood
transfusion; one case with chylous ascites that required total parenteral nutrition.
One case with fever and negative cultures required antibiotic treatment ([Table 4]). There were no instances of surgical reintervention or death.
The requirement of blood transfusion was increased in children with higher stage of
disease ([Table 5]). A tumor larger than 12 cm positively correlated with the need for blood transfusions
(p < 0.05). This size correlation was not observed with other complications. The rest
of the factors analyzed did not significantly correlate with the incidence of complications
([Table 5]). On the other hand, the presence of any degree of anaplasia was the only factor
that increased the risk of relapse. Other factors like lymph node invasion, tumor
rupture, or histological subtype, did not significantly influence prognosis ([Table 6]).
Table 5
Complications versus tumor and patient's characteristics
|
Blood transfusion
|
Complications
|
|
Age at diagnosis (mo)
|
51 (vs. 41)
|
p > 0.05
|
43 (vs. 46)
|
p > 0.05
|
|
Size
|
< 12
|
7/24
|
p < 0.05
|
6/24
|
p > 0.05
|
|
> 12
|
13/21
|
1/21
|
|
Stage
|
I
|
3/13
|
p < 0.05
|
2/13
|
p > 0.05
|
|
II
|
2/12
|
0/12
|
|
III
|
12/15
|
4/15
|
|
IV
|
3/5
|
1/5
|
|
Laterality
|
Left
|
9/21
|
p > 0.05
|
3/21
|
p > 0.05
|
|
Right
|
11/24
|
4/24
|
Table 6
Relapse versus tumor and patient's characteristics
|
Stage
|
I
|
1/13 (8%)
|
p > 0.05
|
|
II
|
0/12
|
|
III
|
2/15 (13%)
|
|
IV
|
1/5 (20%)
|
|
Histology
|
Triphasic
|
3/34 (9%)
|
p > 0.05
|
|
Epithelial
|
0/1
|
|
Stromal
|
1/7 (14%)
|
|
Blastemal
|
0/1
|
|
Mixed
|
0/2
|
|
Anaplasia
|
Yes
|
2/39 (5%)
|
p = 0.01
|
|
No
|
2/4 (50%)
|
|
Lymph node involvement
|
Yes
|
1/3 (33%)
|
p > 0.05
|
|
No
|
3/42 (7%)
|
|
Free tumor resection margins
|
No
|
0/5
|
p > 0.05
|
|
Yes
|
4/40 (10%)
|
|
Tumor rupture
|
Yes
|
1/7 (14%)
|
p > 0.05
|
|
No
|
3/38 (8%)
|
Outcome: Five-year rates for OS and EFS were 100 and 91.1%, respectively. The median
follow-up was 4.5 years (range, 32.6–55.6 months). As regards to EFS, four cases had
disease progression out of which, three in the lungs. Two of these patients had stage
III disease at diagnosis. In one of the cases, rupture of the tumor occurred during
surgery. In the other case, the resection margins were positive and, most importantly,
the tumor was anaplastic. Both cases were treated with chemotherapy and lung radiation.
The third patient had stage IV disease at diagnosis and presented new lung lesions
after the end of treatment. The patient was treated successfully with salvage chemotherapy,
surgical resection of the lesions, and lung radiation. The fourth case had a locoregional
relapse. The patient had stage I at diagnosis and diffuse anaplasia. Treatment of
relapse consisted of surgical resection of the retroperitoneal mass along with chemotherapy
and abdominal radiation. All four patients eventually achieved complete remission
after salvage treatment and remained disease free.
Discussion
One of the arguments for administering preoperative chemotherapy is that chemotherapy
reduces the size of the tumor facilitating surgical resection and thus decreasing
the risk of surgical complications.[6] Data regarding surgical complications in WT is not uniformly reported and spans
diverse surgical techniques and strategies that make comparisons extremely difficult.
Classification systems like the Clavien–Dindo system may help the analysis of data
comparison between different groups.[5] According to this system, there is an excessive bleeding when the patient needs
blood transfusion. This aspect is reflected in our work and therefore the incidence
of excessive bleeding appears to be higher than other publications; interestingly,
this fact is not well defined and well described. In our institution, the decision
for blood transfusion during surgery is led by the anesthesiologist and might differ
among professionals making comparisons difficult. For transfusions outside the operating
room, the standard threshold is a hemoglobin level lower than 7 g/dL. Recent articles
from cooperative groups regarding the surgical complications in WT, the incidence
of excessive bleeding appears lower (UKW-3, 2% and NWTS-4, 1.9%) than in our experience,
but these publications do not report whether excessive bleeding is observed with transfusion
or not.
In the study of the United Kingdom Children's Cancer Study Group, UKW3,[7] the definition of excessive bleeding is not well defined and considered only when
the surgeon would report it in the surgical note. In the NWTS-4 study,[4] they considered excessive bleeding when it exceeded 50 mL/kg but do not mention
the blood transfusion rate. The incidence of excessive bleeding therefore is difficult
to compare between publications.
In a review of WT treatment published in 2009 by Ko and Ritchey,[1] there is a reference from preliminary data of a prospective study that compared
surgical complications between SIOP-93–01 and NWTS-5. The overall incidence of complications
was 6.4% in the SIOP group and 9.8% in the NWTS group; all in all, the differences
were not statistically significant. However, when intraoperative complications like
tumor rupture (2.2% SIOP vs. 15.3% NWTS, p < 0.001), or resection of other organs (6.9% SIOP vs. 15% NWTS, p < 0.001) were evaluated, the differences between both groups were significant. This
finding from previous cooperative studies was noted in a recent publication.[6] Excluding excessive bleeding, the postoperative complication rate in our group was
9%, a lower rate than what was previously reported by the UKW-3 for children with
upfront nephrectomy (20.4%),[7] and the rate reported by the NWTS-4 (12.7%), but similar to the incidence of SIOP-93–01
and NWTS-5.[1]
When analyzing intraoperative complications, we had an accidental tumor rupture and
resection of affected organs of 4.4%, respectively. These rates are lower than rates
reported by other groups with upfront nephrectomy.[1] Preoperative chemotherapy offers other theoretical advantages like decreasing the
risk of tumor rupture during surgery. Several studies of the NWTS suggest that tumor
rupture and tumor spillage were an adverse prognostic factor related to worsen OS
for patients included in the NWTS-1 and OS and EFS for NWTS-2 and 3 studies.[8] Tumor rupture was one of the most important arguments used in the past to justify
the administration of preoperative chemotherapy. It is worthwhile to mention that
in the SIOP-1 study the randomization between preoperative chemotherapy and upfront
nephrectomy was stopped earlier than the planned timeline because of the higher incidence
of intraoperative rupture in the upfront nephrectomy group. However, they did not
find statistically significant differences in OS and EFS between both groups.[9] In the UKW3 study, 205 patients were randomized between upfront nephrectomy and
preoperative chemotherapy. Tumor rupture rate was higher in the upfront nephrectomy
group (14.6% vs. 0%) (p < 0.05). However, they did not find statistically significant differences in terms
of OS and EFS in both groups (10%). They also found a higher incidence of locoregional
relapse in the preoperative chemotherapy group (11.0% vs. 5.4%); however, the differences
were not statistically significant.[10] Our rate of tumor rupture (4.4%), is higher than reported by the SIOP group with
preoperative chemotherapy (SIOP-93–01, 2.2%) but lower than other groups with upfront
nephrectomy (NWTS-5, 15.3%; UKWT3, 14.6%). Along with the data reported in the literature
by the major cooperative groups, in our experience, tumor rupture—whether it is spontaneous
or intraoperative—did not correlate well with worsening the prognosis. We also analyzed
whether there was any relationship between surgical complications and tumor or patient
characteristics. We found that blood transfusion requirements were increased in cases
with higher stage disease. Tumor size larger than 12 cm positively correlated with
the need for blood transfusions. In other publications, the importance of tumor size
had been observed but not correlated with disease stage.[4]
Previous studies have noted that administration of preoperative chemotherapy induces
tumor necrosis and regression of the tumor that may result in downstaging. Therefore,
there are more stage I patients and less stage III patients in SIOP studies than NWTS
(SIOP-93–01, 14.2% vs. NWTS-5, 30.4%).[11] Nevertheless, the staging system in SIOP and COG is not readily comparable since
they reflect different moments during treatment. In our series, the stage III rate
was 33%, similar to what is reported in the NWTS.
The presence of anaplasia (mainly reflecting p53 mutations) is one, if not the only
one, of the important prognostic factors in WT. It increases the risk of relapse,
metastasis, and death.[12]
[13]
[14]
[15]
[16] In our series, it was the only prognostic factor related to worsen EFS (50% vs.
5%, p < 0.05). Survival in WT has been shown over the years not to be related to surgical
aspects but mainly to the biology of the disease. Our results using surgery upfront
are in line with the major cooperative groups with 5 years OS and EFS of 100 and 91%,
respectively. According to our experience, with current surgical technology and in
skilled multidisciplinary teams, the arguments to not surgically remove WT upfront
are not justified by operative complications.
