Keywords
gallbladder cancer - retroperitoneal lymph node metastasis - interaortocaval lymph
node metastasis - para-aortic lymph-node metastasis - distant metastasis
Introduction
In gallbladder cancer (GBC), there is conflicting evidence in the literature whether
retroperitoneal lymph nodal metastases (RLNM) should be considered as regional nodal
metastases or as distant metastases (DMs) and the jury is out on radical curative
surgery in presence of RLNM. This is an analysis of GBC patients, to see the effect
of RLNM on survival and to compare with that of patients with DMs, where curative
surgery was abandoned due to RLNM or DM, found intraoperatively.
Material and Methods
This study was performed at the department of surgical gastroenterology at a tertiary
care center in North India where GBC is rife. All patients of GBC, where curative
resection was abandoned between January 2013 and December 2018, on account of positive
retroperitoneal lymph node (RLN) or DM on frozen section biopsy, were studied for
survival. The study design is depicted in [Fig. 1].
Fig. 1 Study design.
All patients were staged with a triple-phase computed tomography (CT) scan. In case
of locally advanced disease where major hepatectomy or hepaticopancreaticoduodenectomy
was contemplated, neoadjuvant treatment was instituted after a staging laparoscopy
(SL). Patients with obstructive jaundice underwent preoperative or preneoadjuvant
treatment biliary drainage. Those who were nonmetastatic and possibly resectable on
imaging were considered for curative surgery. Most of the patients underwent SL. In
case a liver, peritoneal, or omental nodule was seen on SL, the lesion was biopsied
and sent for frozen section examination. In the absence of dissemination (liver, peritoneal,
omental nodule, or ascites) on SL, the findings were confirmed at laparotomy when
the RLNs were sampled for frozen section biopsy. The planned curative resection was
abandoned if the biopsy report suggested metastatic disease and the patients were
offered palliative care. The study population was divided into two groups based on
the site of metastatic disease—RLNM and DM, and outcome was compared. Demographic
profile, preoperative blood parameters, neoadjuvant treatment, biliary drainage, and
postoperative palliative treatment were recorded. Survival was calculated from the
day of surgery. Patients were followed up through hospital visits and telephonically.
Data was analyzed using the Statistical Package for the Social Sciences software (version
22.0). Continuous variables were compared with independent t-test and categorical variables were compared with chi-square test. A p-value of < 0.05 was considered significant. In case of skewed variables, that is,
bilirubin and survival, median and interquartile range (IQR) were used. Survival among
both the groups (RLNM and DM) was compared with log-rank test.
Results
A total of 235 patients with ostensibly resectable GBC underwent surgical exploration
between January 2013 and December 2018. The planned curative resection was executed
in 144 (61%) patients and abandoned in 91 (39%) patients because of RLNM (n = 20, 9%) or DM (n = 71, 30%) on frozen section biopsy. Demographic profile including age, gender, comorbidities,
preoperative clinical features, blood parameters (hemoglobin, bilirubin, albumin,
and international normalized ratio), requirement of biliary drainage, or neoadjuvant
treatment were comparable between the two groups as shown in [Table 1]. SL was performed in 214 out of 235 patients (21 patients underwent laparotomy without
SL for techno-logistical reasons). In the 71 patients, where the curative resection
was aborted, metastases were detected on laparoscopy in 54 patients and at laparotomy
in 17 patients (15 of these were missed at SL, while 2 were found in patients who
did not undergo a prelaparotomy SL). Metastases missed at laparoscopy were peritoneal
(n = 6), omental (n = 5), liver (n = 3), and gastric serosal nodule (n = 1). Overall, SL changed the management in 54/214 (25%) patients by averting a laparotomy
and abandoning curative resection. The RLNM included interaortocaval nodes (n = 18) and paraaortic nodes (n = 2) in patients without any DM. The DM group consisted of liver nodules (n = 27), peritoneal nodules (n = 34), omental metastasis (n = 15), and malignant ascites on fluid cytology (n = 1). Six patients had polymetastatic disease.
Table 1
Comparison between retroperitoneal lymph node metastasis (RLNM) and distant metastasis
(DM) groups
|
Parameters
|
RLNM (n = 20)
|
DM (n = 71)
|
p-Value
|
|
Age in years, mean ± SD [range]
|
53.6 ± 8.5 [35–71]
|
51.2 ± 10.8 [30–80]
|
0.14
|
|
Sex, M:F
|
7:13
|
26:45
|
1.0
|
|
BMI in kg/m2, mean ± SD [range]
|
23.5 ± 3.9 [17.7–32.3]
|
22.4 ± 3.6 [18.5–40.2]
|
0.29
|
|
Comorbidity, n (%)
|
8 (40)
|
20 (28.1)
|
0.14
|
|
Diabetes, n (%)
|
6 (30)
|
9 (12.6)
|
0.08
|
|
Hypertension, n (%)
|
4 (20)
|
10 (14)
|
0.5
|
|
Coronary artery disease (CAD), n (%)
|
0
|
3 (4.2)
|
1.0
|
|
Jaundice, n (%)
|
4 (20)
|
20 (28.1)
|
0.57
|
|
LOA, n (%)
|
11 (55)
|
38 (53.5)
|
0.55
|
|
LOW, n (%)
|
11 (55)
|
39 (54.9)
|
1.0
|
|
Incidental, n (%)
|
1 (5)
|
12 (16.9)
|
0.28
|
|
Node involvement (N + ) on preoperative images (CECT/USG), n (%)
|
17 (85)
|
52 (73.2)
|
0.38
|
|
Hemoglobin in g/dL mean ± SD
|
11.3 ± 1.4
|
11.5 ± 1.8
|
0.29
|
|
Bilirubin in g/dL median (IQR)
|
0.8 (0.45–1.4)
|
0.8 (0.5–1.5)
|
0.07
|
|
Albumin in g/dL mean ± SD
|
3.9 ± 0.57
|
3.9 ± 0.51
|
0.6
|
|
INR mean ± SD
|
1.01 ± 0.13
|
1.03 ± 0.11
|
0.74
|
|
Preoperative biliary drainage, n (%)
|
4 (20)
|
11 (15.4)
|
0.73
|
|
Neoadjuvant treatment, n (%)
|
2 (10)
|
3 (4.2)
|
0.3
|
|
Palliative treatment, n (%)
|
10 (50)
|
16 (22.5)
|
0.04
|
|
Survival in months median (IQR) [range]
|
5 (3–11) [2–26]
|
6 (4–10) [2–24]
|
0.08
|
|
Six-month survival, n (%)
|
9 (45)
|
37 (52.1)
|
0.62
|
|
1-year survival, n (%)
|
4 (20)
|
13 (18.3)
|
1.0
|
|
2-year survival, n (%)
|
2 (10)
|
1 (1.4)
|
0.12
|
Abbreviations: BMI, body mass index; CECT, contrast-enhanced computed tomography;
INR, international normalized ratio; IQR, interquartile range; LOA, loss of appetite;
LOW, loss of weight; SD, standard deviation; USG, ultrasonography.
Note that 50% patients in the RLNM group received palliative treatment as compared
with 22% in the DM group; the difference was statistically significant (p = 0.04). In majority of the cases, the patients opted against chemotherapy either
due to logistic issues or the nihilism associated with disseminated disease. All patients
were followed up. The median survival for RLNM and DM groups were 5 months (range
2–26; IQR 3–11) and 6 months (range 2–24; IQR 4–10), respectively, without any significant
difference on log-rank test (p-value = 0.64) ([Fig. 2]).
Fig. 2 Comparison of survival between retroperitoneal lymph node metastasis (RLNM) and distant
metastasis (DM) groups.
There was no 3-year survivor in either group. In the studied population, only one
patient with liver metastasis and stable disease is still alive (18 months postsurgery),
with no evidence of disease progression after 6 cycles of chemotherapy.
When survival was compared between the RLNM and DM groups receiving palliative chemotherapy,
there was no significant survival difference (median 11 months in both the groups
and on log-rank test p-value = 0.89) ([Fig. 3]).
Fig. 3 Survival comparison among retroperitoneal lymph node metastasis (RLNM) and distant
metastasis (DM) groups who received palliative chemotherapy.
Discussion
GBC is the most common malignancy in the biliary tract.[1] In GBC lymph node involvement is associated with poor survival.[2]
[3]
[4] The extent of lymphadenectomy and the level of lymph node involvement which precludes
curative surgery is still a matter of controversy. A study by Murakami et al on the
survival of patients with paraaortic lymph node (n = 17) metastasis in resected biliary carcinoma documented absence of any survival
advantage.[5] Similarly, other authors have also documented poor outcomes in the presence of RLNM
in GBC akin to DM.[3]
[4]
[6]
[7]
[8]
[9] On the contrary, Nishio et al have shown that in presence of RLN involvement, the
survival is better after resection than nonoperated similar patients with RLNM (p-value 0.014).[10] There are also few case reports showing anecdotal long-term survival in the presence
of distant lymph node metastasis.[11]
[12] To the best of our knowledge, this report is the largest reported experience on
survival of GBC patients with RLNM. In this population, there was no significant difference
in survival among both the groups indicating that the prognosis of the RLNM group
was as poor as of the DM group. All patients were counseled about the disease stage
and prognosis and they were offered palliative chemotherapy. In the subgroup of patients
receiving palliative chemotherapy again the survival was similar for patients with
RLNM or DM. Our study validates poor prognosis of the RLNM, as with the DM groups.
Because of poor prognosis, RLNM should actively be sought for in the preoperative
imaging, so as to avoid unnecessary laparotomy. Presently, imaging modalities are
less accurate in evaluation of RLNM unless the nodes are large in size with obvious
signs of involvement. CT criteria for evaluation for metastatic lymph node[13] have been found to have poor sensitivity (14.7%) and positive predictive value (33.3%),[14] similarly positron emission tomography scan also has limited role.[15] If an enlarged RLN is suspected to be involved at imaging, it should be target for
fine-needle aspiration cytology (FNAC) under ultrasound, CT, or endoscopic ultrasound
guidance. But the preoperative sampling has limitations because of difficult location
and sampling error leading to a false negative rate of almost 30%.[14] Hence, in all cases of GBC, retroperitoneal (interaortocaval and paraaortic) lymph
nodes should be sampled as a routine for frozen section histological examination before
starting the curative resection to avert a futile extensive surgery.
Conclusion
GBC patients with RLNM have poor survival similar to DMs and should be considered
as the equivalent of DM. This study strengthens evidence to avoid curative resectional
surgery in patients with RLNM. These lymph nodes should be sampled, preoperatively
if suspicious on imaging for FNAC and at surgery, as a routine for frozen section
histological examination before initiating curative resection to avert a futile exercise.
