Available online at www.sciencedirect.com
ScienceDirect EJSO 40 (2014) 567e575
www.ejso.com
Hepatolithiasis-associated cholangiocarcinoma: Results from a multi-institutional national database on a case series of 23 patients A. Guglielmi a, A. Ruzzenente a, A. Valdegamberi a, F. Bagante a, S. Conci a, A.D. Pinna b, G. Ercolani b, F. Giuliante c, L. Capussotti d, L. Aldrighetti e, C. Iacono a,* a
Department of Surgery, Division of General Surgery “A”, “GB Rossi” University Hospital, University of Verona Medical School, Verona, Italy b Department of Surgery and Organ Transplantation, Ospedale Sant’Orsola-Malpighi, University of Bologna, Bologna, Italy c Hepatobiliary Surgery Unit, Department of Surgery, Catholic University of the Sacred Heart School of Medicine, Roma, Italy d Department of Hepato-Biliary-Pancreatic and Digestive Surgery, Ospedale Mauriziano ‘Umberto I’, Torino, Italy e Department of Surgery-Liver Unit, Scientific Institute San Raffaele, Milano, Italy Accepted 6 December 2013 Available online 18 December 2013
Abstract Aims: Few papers focused on association between hepatolithiasis (HL) and cholangiocarcinoma (CCC) in Western countries. The aims of this paper are to describe the clinical presentation, treatment, and postoperative outcomes of CCC with HL in a cohort of Western patients and to compare the surgical outcomes of these patients with patients with CCC without HL. Materials and methods: Among 161 patients with HL from five Italian tertiary hepato-biliary centers, 23 (14.3%) patients with concomitant CCC were analyzed. The results of surgery in these patients were compared with patients with CCC without HL. Results: The 60.9% of patients with HL received the diagnosis of CCC intra- or postoperatively, with a resectability rate of 91.3%. The postoperative morbidity was 61.6%. The 1- and 3-year survival rates were 78.6% and 21.0%, respectively. The recurrence rate was 44.4% and the 3year disease-free survival rates were 18.8%. The comparison with patients with CCC without HL showed a higher resectability rate ( p ¼ 0.02) and a higher frequency of earlier stage ( p ¼ 0.04) in CCC with HL. Biliary leakage was more frequent in CCC with HL group ( p ¼ 0.01) compared to CCC without HL group. We found no differences in overall and disease-free survival between the two groups. Conclusions: Patients with HL and CCC showed a high resectability rate but a higher morbidity. Nevertheless, overall and disease-free survival of patients with CCC and HL showed no differences compared to those of patients with CCC without HL. Also in Western countries, HL needs a careful management for the possible presence of CCC. Ó 2013 Elsevier Ltd. All rights reserved. Keywords: Cholangiocarcinoma; Hepatolithiasis; Liver surgery
Introduction
Abbreviations: HL, hepatolithiasis; CCC, cholangiocarcinoma. * Corresponding author. Department of Surgery, Division of General Surgery “A”, Unit of HPB Surgery, “GB Rossi” University Hospital, University of Verona Medical School, Piazzale L.A. Scuro 10, 37134 Verona, Italy. Tel.: þ39 (0) 458124412. E-mail address: [email protected] (C. Iacono). 0748-7983/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejso.2013.12.006
Cholangiocarcinoma (CCC) is the second most common primary liver cancer.1 Several risk factors have been identified for the development of CCC, including primary sclerosing cholangitis, choledochal cysts, Caroli’s disease, and liver fluke infection. Hepatolithiasis (HL) is associated with a high incidence of CCC and it is one of the major risk factors for the development of CCC.1e3 The incidence of
568
A. Guglielmi et al. / EJSO 40 (2014) 567e575
HL shows a regional distribution, with a greater prevalence in South-East Asian countries. In contrast, HL is fairly uncommon in Western countries with an incidence to 1.3%.4,5 Epidemiological, pathological, and genetic studies have seemed to prove the relationship between HL and CCC. It has been suggested that biliary carcinogenesis is caused by prolonged inflammation of the biliary epithelium secondary to biliary stones, bile stasis, and bacterial infections.6 In the literature, few papers have focused specifically on the association between HL and CCC in Western countries.7e9 No data are available on the clinical presentation, treatment, and outcome of patients with both conditions. The aim of the present paper is to describe the clinical presentation, treatment, and postoperative outcomes of CCC associated to HL in a cohort of patients recruited from a multi-institutional database from tertiary referral centers. Moreover, the long-term results of surgical resection in these patients were compared with the results in patients with CCC without HL. Material and methods From January 1st, 1997 to December 31st, 2011, 161 patients affected by HL were recruited prospectively in five Italian tertiary hepato-biliary centers (Department of Surgery, Division of General Surgery ‘A’, GB Rossi Hospital, University of Verona, Verona; Department of Hepato-BiliaryePancreatic and Digestive Surgery, Ospedale Mauriziano ‘Umberto I’, Turin; Department of Surgery e Liver Unit, Scientific Institute San Raffaele, Milan; Hepatobiliary Surgery Unit, Department of Surgery, Catholic University of the Sacred Heart School of Medicine, Rome; Department of Surgery and Organ Transplantation, Ospedale Sant’Orsola-Malpighi, University of Bologna, Bologna). All the institutions are members of the Italian Chapter of the International Hepato-Pancreato-Biliary Association (IHPBA). From this dataset, 23 (14.3%) patients with concomitant CCC were included into this study. The diagnosis of CCC was pathologically confirmed in all patients. We defined HL as concretions existing in the intrahepatic bile ducts. Although the confluence of the hepatic ducts is situated outside of the parenchyma of the liver, for convenience we included in this study HL of any bile duct proximal to the confluence of the right and left hepatic ducts, as reported by other Authors in literature.10 The following clinical variables were analyzed: age; gender; symptoms; medical history (previous operations on the bile duct, liver fluke infestation, underlying hepatic disease); clinical characteristics of HL (location, time between diagnosis and treatment); laboratory data including serum bilirubin, gamma-glutamyl transferase , preoperative liver function test (indiocyanine green retention rate), presence of tumor markers, histological differentiation, lymph node involvement, and the presence of metastasis. The imaging modalities that were used for the diagnosis of CCC were recorded and analyzed.
After surgery, a computed tomography or magnetic resonance scan was performed every three or six months for follow-up examination of the patients. One patient was lost to follow-up, and the median duration of follow-up was 15 months (range 3e121 months). To evaluate the long-term results of surgical resection of CCC associated with HL, we compared the results of surgery on our cohort with those for a cohort of 414 patients submitted to surgical resection for CCC without HL (peri-hilar or intrahepatic). The data of patients affected by CCC without HL were obtained from a multi-institutional database held by two of the five hepato-biliary tertiary centers (Verona and Bologna) during the same time period. The median follow-up period in these patients was 18 months (range 3e275 months). Statistical analysis The data were collected and analyzed with the statistical software SPSS (version 18.0; SPSS, Chicago, Ill., USA). Continuous variables were compared using an independent two sample t-test. The differences between categorical variables were analyzed with Pearson’s chi-squared test. Survival analyses and survival plots were carried out using the KaplaneMeier method. We considered the day of treatment to be time zero, and data from patients alive at the end of the follow-up period were considered to be censored. Univariate analyses of survival were performed with the KaplaneMeier method, using the log-rank test to verify the significance of differences. Multivariate analysis of survival was carried out with Cox’s regression model. A value of p < 0.05 was considered significant. A propensity score match analysis (1:3) was performed between patients with HL and CCC and patients with CCC alone in order to better define the prognostic value of HL in patients with CCC. Results Demographic data Among the 161 patients with HL observed during the study period 23 with CCC (14.3%) were included into the study. Among these, 14 were men and 9 were women with a mean age of 63.9 years (range 42e82 years). The patients’ characteristics are described in detailed in Table 1. The clinical features are summarized in Table 2. No differences were found between these patients and patients with CCC alone. Symptoms at presentation The mean time from the diagnosis of HL and of malignancy was 61.2 months (range 1e372 months). Seventeen patients (73.1%) developed symptoms due to HL before the diagnosis of CCC. The mean time span between the onset of symptoms and diagnosis of CCC was 17.3 months (range 0e72 months). In most cases HL patients were
Table 1 Clinical characteristics of 23 patients with CCC and HL. Previous procedures involving bile ducts
Symptoms
Time between HL and CCC (months)
HL location
Diagnosis
Type of surgery
T stage
N stage
M stage
Follow up (months)
Alive
Pain, Cholangitis Cholangitis
2
Right liver
Intraop.
Right hepatectomy
3
x
0
42
Yes
24
Bilateral
Postop.
3
0
0
24
No
Liver failure
0
Bilateral
Preop.
3
x
0
10
no
Pd
48
Right liver
Postop.
Bisegmentectomy S2e3 and bilio-digestive anastomosis (further segmentectomy S4e1) Left hepatectomy, segmentectomy S8 Right hepatectomy
2
0
0
36
No
96 6
Left liver Right liver
Postop. Intraop.
Left hepatectomy Right hepatectomy
1 3
0 1
0 0
120 15
Yes No
Ictus cerebri e Pd
372
Left liver
Postop.
Bisegmentectomy
1
0
0
121
No
Pd
Pain, Cholangitis
4
Left liver
Intraop.
4
1
1
6
No
Pd
Pain, Cholangitis Cholangitis, weight loss
180
Right liver
Intraop.
4
0
1
Lost
–
–
252
Left liver
Intraop.
Explorative laparotomy and T-tube placement Explorative laparotomy Left hepatectomy (þS1)
2a
0
0
7
No
Pd
Cholangitis, jaundice Cholangitis None None
1
Right liver
Preop.
Right hepatectomy
1
0
0
18
Yes
47 40 2
Left liver Left liver Left liver
Preop. Preop. Preop.
4 2b 3
0 0 1
0 0 0
12 6 25
Yes Yes No
108 48 2 2 180
Right liver Right liver Right liver Bilateral Right liver
Preop. Postop. Preop. Preop. Postop.
Left hepatectomy Left hepatectomy Left hepatectomy þ S1 þ EBD resection Right hepatectomy Segmentectomy S6e7 Right trisectionectomy Right hepatectomy Bisegmentectomy
2 1 1 3 1
1 x 0 1 x
0 0 0 0 0
3 18 18 4 4
Yes Yes No Yes Yes
Pd
0
Right liver
Intraop.
Right trisectionectomy
4
0
0
4.4
No
Pd
1
M
65
None
2
M
66
ERCP
3
M
75
None
4
M
82
5 6
M M
62 58
Bile duct lithotomy None Cholecystectomy
7
F
71
8
M
66
9
F
42
10
F
54
11
F
51
Cholecystectomy and PTC Cholecystectomy and bilio-digestive anastomosis Multiple ERCPs
12 13 14
M M M
72 75 66
ERCP None None
15 16 17 18 19
M F F M F
72 44 56 74 47
Cholecystectomy Cholecystectomy Cholecystectomy None ERCP
20
F
45
Cholecystectomy and bilio-digestive anastomosis
Trans-duodenal papillotomy and bilio-digestive anastomosis None
Pain, Cholangitis Recurrent cholangitis Cholangitis Recurrent cholangitis Pain , Cholangitis
None None Cholangitis None Cholangitis Pain Cholangitis
Cause of death
A. Guglielmi et al. / EJSO 40 (2014) 567e575
Age
Pd
569
(continued on next page)
Postop. Postop. Left liver Left liver 4 2 Cholangitis Jaundice 81 71 F M 22 23
74 M 21
ERCP: Endoscopic retrograde cholangiopancreatography; EBD: extrahepatic bile duct; Pd: progression disease; Preop.: preoperative; Intraop.: intraoperative; Postop.: postoperative; PTC: percutaneous transhepatic cholangiography.
Pd No Yes 19.4 60 2 1
0
0 0
Pd No 8.7 0 Cholangitis
Cholecystectomy and bilio-digestive anastomosis None None
0
Right liver
Preop.
Right hepatectomy and bilio-digestive anastomosis Bisegmentectomy EBD resection
3
0
Follow up (months) Symptoms Previous procedures involving bile ducts Age
Table 1 (continued )
Time between HL and CCC (months)
HL location
Diagnosis
Type of surgery
T stage
N stage
M stage
Cause of death
A. Guglielmi et al. / EJSO 40 (2014) 567e575
Alive
570
asymptomatic and no specific treatment was performed. The symptomatic patients were submitted to endoscopic or radiological procedures, as such as endoscopic retrograde cholangiopancreatography (ERPC) or percutaneous transhepatic cholangiography (PTC). Surgery was chosen when non operative treatment failed (see Table 1). Laboratory data and tumor markers On laboratory testing, 14 patients (75%) showed elevated serum GGT (>45 U/L; mean value of 179 U/L; range 24e517). The median values of the tumor markers CA19.9 and CEA were 85.45 U/L and 3.7 U/L, respectively (range 6e12153 and 0.8e7, respectively). Diagnosis of malignancy All patients underwent preoperative abdominal ultrasonography. Magnetic-resonance cholangio-pancreatography Table 2 Demographic data. Features
N (%)
M:F Age
14:9 63.9 (range 42e82)
Underlying liver disease Yes No
3 (13) 20 (87)
Previous biliary surgery Yes No
14 (60.9) 9 (39.1)
Liver fluke infestation Yes No
0 23
Symptoms None Cholangitis Pain Jaundice alone Weight loss
5 (21.7) 17 (73.9) 6 (26.1) 1 (4.3) 1 (4.3)
Duration of symptoms GGT (U/L, median value) Total serum bilirubin (mg/dL, median value) CEA (U/L, median value) Ca19.9 (kU/L, median value) Location of HL Right liver Left liver Bilateral Time between diagnosis of HL and CCC Diagnosis of CCC Preoperative Intraoperative Postoperative a
Inter-quartile range.
17.3 months (1e48) 105 (36.75e346.75)a 0.95 (0.8e1.1)a 3.7 (1.9e5.1)a 85.45 (6e12153)a 11 (47.9) 9 (39.1) 3 (13) 5.1 years (range 1e372 months)
9 (39.1) 6 (26.1) 8 (34.8)
A. Guglielmi et al. / EJSO 40 (2014) 567e575
571
Figure 1. AeBeC: Case n. 15 in Table 1. Preoperative diagnosis of CCC with HL. Preoperative magnetic resonance cholangio-pancreatography revealed an intrahepatic mass-forming CCC (arrow in A) in the right lobe of liver with strictures caused by HL (arrow in B). Right hepatectomy was performed. Control CT scan showed no intrahepatic recurrence (C).
was performed in 19 patients (82.6%, Fig. 1). A computed tomography scan was performed in 18 patients (78.3%) and a positron emission tomography-computed tomography scan in two patients (8.7%). A preoperative diagnosis of CCC was obtained in 39.1% of the cases (nine patients). Intraoperative diagnosis was performed in six patients (26.1%), using intraoperative ultrasonography (one case) and intraoperative frozen sections (five cases). Postoperative pathological examination of the specimen led to a diagnosis of cancer in eight patients (34.8%). According to the CCC macroscopic type 20 of the 23 patients with CCC and HL (86.9%) had an intrahepatic massforming CCC. Only three cases were diagnosed as perihilar CCC (13.1%). Resectability, morbidity, and mortality rates Resectability was 91.3% (21 patients), two patients were not submitted for resection because of the presence of metastases along the diaphragmatic peritoneum. Radical resection (R0) was performed in 17 cases (80.9%). One patient underwent a second operation to achieve R0 after the histological diagnosis of CCC: the patient underwent bisegmentectomy of S2e3 for HL without any suspicion of cancer: definitive pathological examination revealed the diagnosis of CCC associated to HL with positive resection margins, for this reason further bisegmentectomy S4e1 was performed to achieve curative resection. The rate of resectability in the group without HL was 68.8%. This appeared to be significantly lower than the rate of resectability in patients with both CCC and HL (91.3%, p ¼ 0.02). Among patients who underwent surgery for CCC without HL, radical resection was achieved in 179 patients (75.5%), which showed no significant difference from patients with both CCC and HL ( p ¼ 0.1). In patients with both CCC and HL the overall morbidity rate was 61.9%, and all postoperative complications were of grade II according to the DindoeClavien classification.11
The most frequent complication was biliary leakage (23.8%), followed by intra-abdominal collection (19.5%), wound infection (14.3%), and pleural effusion (4.3%). There was no perioperative mortality. When the patients with both CCC and HL were compared with these patients without HL, we identified a significant difference in the rate of postoperative complications (61.9% vs. 41.2% respectively; p ¼ 0.02; Table 3). In particular, patients with HL showed a higher rate of biliary leakage (23.8% vs. 11.3%, respectively; p < 0.01). The rate of other postoperative complications (such as pneumonia, abdominal abscess, pleural effusion, etc.) did not show a statistically significant difference (38.1% vs. 29.9%, respectively; p > 0.05; Table 3). Postoperative mortality in CCC without HL was 6.3% (18 patients), this difference was not statistically significant compared to CCC with HL ( p ¼ 0.37). TNM staging and histological grading On the basis of the tumorenodeemetastasis (TNM) classification, 7th edition,12 seven patients (33.3%) were classified as T1, eight patients (38.1%) as T2, five patients (23.8%) as T3, and only one patient (4.8%) as T4. The median number of lymph-nodes harvested was 4 (range 0e21), but the median number of positive lymph-nodes was 0 (range 0e10). Lymphadenectomy was not performed in four patients because no suspicion of cancer nor lymph node enlargement was found during surgical exploration. Fourteen patients were classified as N0 and five patients as N1 (61.9% and 19%, respectively). Two patients (8.7%) had distant metastasis, in the form of diaphragmatic peritoneum in both cases. In total, 47.8% of patients were in stages IeII according to the TNM system (11 patients), and 52.2% (12 patients) were in stages IIIeIVaeb. Comparison of the tumor stage and histological grading between patients with CCC and HL and those with CCC without HL is reported in Table 3. The histological
572
A. Guglielmi et al. / EJSO 40 (2014) 567e575
Table 3 Comparison of intraoperative and pathological features, early and longterm results of patients affected by CCC with and alone. Feature
Number of patients with CCC and HL (%)
Number of patients with CCC alone (%)
p
Resectability
21 (91.3)
284 (68.8)
0.02
Radicality R0 R1
17 (80.9) 2 (9.5)
179 (75.5) 79 (24.5)
Tumor location Intrahepatic Peri-hilar
20 (86.9) 3 (13.1)
171 (41.4) 242 (58.6)
Macro vascular invasion Portal resection/ reconstruction
7 (39.1)
94 (38.2)
0.41
1 (4.3)
28 (12)
0.86
Overall morbidity Biliary leakage Others
13 (61.9) 5 (23.8) 8 (38.1)
117 (41.2) 32 (11.3) 85 (29.9)
0.02 0.01 0.23
Mortality
0
18 (6.3)
0.37
Histological grade G1e2 G3e4
16 (69.6) 7 (30.4)
174 (61.2) 110 (38.8)
Stage T according TNM 7th edition T1 T2 T3 T4 Stage N according TNM 7th edition N0 N1 Stage M according TNM 7th edition M0 M1 Stage TNM 7th edition Stage IeII Stage IIIeIV Overall median survival time Overall survival 1-year 3-year 5-year Recurrence rate Disease-free median survival time Disease-free survival 1-year 3-year
0.10
0.33
0.30
0.04 7 8 5 1
(33.3) (38.1) (23.8) (4.8)
27 (10.5) 165 (64.1) 62 (24.2) 3 (1.2) 0.59
14 (61.9) 5 (19)
142 (67.3) 69 (32.7) 0.60
21 (91.3) 2 (8.7)
261 (91.9) 23 (8.1) 0.14
11 (47.8) 12 (52.2)
125 (30.3) 286 (69.7)
24 mo.
25.2 mo.
78.6% 21.0% 21.0%
78.1% 50.8% 30.4%
47.6% 6.0 mo.
51.0% 20.7 mo.
18.8% 18.8%
43.4% 21.6%
0.78
0.61
0.14
evaluation revealed that 30.3% of the latter group of patients were in stage IeII according to the TNM system, and 69.7% were in stage IIIeIV. The difference in the distribution of tumor stage between patients with both HL and CCC and those with CCC alone did not reach statistical significance ( p ¼ 0.14). The subgroup analysis showed that 10.5% of the patients with CCC alone were in stage T1, compared with 33.3% of the patients with both CCC and HL ( p < 0.05). However, the analysis of other histological characteristics revealed no significant differences between the groups. Overall and disease-free survival The overall median survival time was 24 months, with 1, 3-, and 5-year actuarial survival rates of 78.6%, 21.0%, and 21.0%, respectively. In patients with CCC alone, the overall median survival time was 25.2 months, with 1-, 3, and 5-year actuarial survival rates of 78.1%, 50.8%, and 30.4%, respectively. Overall survival was not significantly different in patients with CCC and HL from that in patients with CCC alone ( p ¼ 0.78, Fig. 2). The rate of recurrence in patients with HL and CCC was 47.6% (10 out of 21 patients) and the 1- and 3-year diseasefree survival rates were 18.8% and 18.8%, respectively. Regarding the site of recurrence three patients had intrahepatic recurrence, two patients had lymph-nodal recurrence and 5 patients had distant recurrence. In patients with CCC alone, the 1-, 3-, and 5-year disease-free survival rates were 43.4%, 21.6%, and 19.2%, respectively. The 1- and 3-year disease-free survival in patients with CCC and HL appeared to be lower than that in patients with CCC without HL although the difference did not reach statistical significance ( p ¼ 0.14, Fig. 2). We performed a sub-analysis that takes into account the stage of tumor, namely T stage. In stages T1 and T2, median OS was 18.00 months (CI 95%, 3.59e32.40) for patients with CCC and HL and 42.63 months (CI 95%, 33.62e51.64) for patients with CCC without HL (HR 2.75, CI 95% 1.18e6.42, p ¼ 0.019). In stages T3 and T4, median OS was 24.33 months (CI 95%, 5.20e43.46) for patients with CCC and HL and 20.06 months (CI 95%, 13.12e27.00) for patients with CCC without HL ( p ¼ 0.878). To confirm our results a propensity matched analysis was performed. Before matching the minimum p value was computed among T stage ( p ¼ 0.006), after matching the balance analysis does not reveal difference among covariates in groups of patients with or without HL. HL resulted an important prognostic factor in terms of OS for patients with CCC with an HR of 2.566 (CI 95%, 1.25e5.28, p ¼ 0.01). Discussion Hepatolithiasis is frequent in Eastern countries, in which its risk factors are endemic. Consequently, most of the
A. Guglielmi et al. / EJSO 40 (2014) 567e575
573
Figure 2. Comparison of overall survival and disease-free survival of CCC patients with (N ¼ 21) and without (N ¼ 414) HL.
knowledge about this disease and its correlation with CCC comes from studies of Eastern patients.13e16 This disease is rare in Western countries, and the literature contains few data about the presentation, features, treatment, and outcome of HL associated with CCC in Western populations. Although the symptoms, blood tests, and imaging results might suggest a diagnosis of CCC in patients with HL, the preoperative diagnostic accuracy varies from 20% to 70%.16,17 This variation in accuracy might be due to the pattern of growth of CCC, which may infiltrate the bile duct without mass formation, or because the stricture or stenosis that is caused by CCC can be difficult to differentiate from that caused by stones.18 We report herein that preoperative diagnosis was achieved in only 39.1% of our patients, which is lower than the rate of 68.4% reported by Eastern authors.15 The reason for this might be that in Western countries, preoperative cholangioscopy is not used routinely in the diagnosis and treatment of HL. The reported resectability rate14,15,19 for CCC and HL is higher than that for CCC without HL. Lee et al.15 showed a resectability rate of 52.6% in patients treated for HLassociated CCC and of 39.2% in patients with CCC without HL ( p < 0.001). In addition, Chen et al.14 showed a higher rate of resectability for tumors in patients with CCC and HL than in patients who had CCC without HL (31.1% vs. 26.8%). In accordance with the literature, we report herein that patients affected by CCC with HL had a higher rate of resectability than patients with CCC alone (91.3% vs. 68.8%; p ¼ 0.02). The higher rate of resectability in the present data in comparison with other studies probably reflects the improvements in intraoperative and postoperative management that have occurred during the last ten years.20e25 In the present study, more patients affected by both HL and CCC were in early T stage than those with CCC alone (33.3% vs. 10.5%, respectively; p ¼ 0.04). This confirms the data from the Eastern literature. Chen et al. showed
that 33.0% of patients with HL and CCC presented at TNM stages IeII compared with 12.5% of patients with CCC alone ( p < 0.05). A possible explanation for that is that the presence of symptoms due to HL could lead to an earlier diagnosis of CCC.14 In literature, surgical treatment is associated with a higher postoperative morbidity in CCC in patients with HL than in patients who undergo surgery for CCC without HL: wound infection and bile leakage are three times more likely in the former patients, probably because of the previous cholangitis and the abnormalities of the bile ducts that are related to HL.14,15,19 The postoperative outcomes in the present study were in agreement with these results. The rate of morbidity was significantly higher, especially biliary leakage, in patients with CCC and HL than in patients with CCC only, though all the complications were of stage II according to the Dindo classification. Nevertheless, the rate of postoperative mortality reported in the literature is low and no differences were shown between the two groups of patients in this respect. Chen et al.14 reported a surgical mortality rate of 3.8% in 106 patients affected by CCC with HL and 3.6% in 56 patients alone. Our findings confirmed these data: the mortality rate was comparable between patients with both CCC and HL and those with CCC alone (0.0% and 6.3% respectively, p ¼ 0.37). Despite a reported higher rate of resectability and the earlier stage at diagnosis, recent data from the literature seem to show no differences in overall survival between patients with CCC and HL and patients with CCC alone. Although Su et al.19 reported significantly poorer survival in patients with CCC and HL, Chen et al.14 stated that the overall survival of patients with resected CCC and HL did not differ from that of patients with CCC but alone, with a 5-year survival rate of 16.5% in the former and 7.8% in the latter. These data were not statistically different also when stratified by different tumor stages, or by the type of treatment received (hepatic resection or palliative). In
574
A. Guglielmi et al. / EJSO 40 (2014) 567e575
addition, Lee et al.,15 in a study of 38 patients with CCC with HL and 102 patients affected only by CCC, showed a 5-year survival rate of 20.0% for patients both with and alone. We also did not find statistically significant differences between the two groups with respect to survival: the 5year survival rates in patients affected by CCC with HL and those alone were 21.0% and 30.4%, respectively ( p ¼ 0.78). In agreement with the reports in the Eastern literature, we conclude that, in patients with CCC from Western countries, HL does not impair long-term survival, nevertheless it allows an earlier presentation and greater resectability rate. One possible explanation of this could be the inadequate lymph-node harvesting: in patients with HL and CCC, the median number of lymph-nodes harvested was 4, but the median number of positive lymph-nodes was 0 (see Table 1). Moreover, in four patients in which CCC was discovered postoperatively, lymphadenectomy was not performed because no suspicion of cancer nor lymph node enlargement was found during surgical exploration. We believe that the lymph-node involvement was underestimated in these patients. This could explain the low rates of overall and disease-free survival. Furthermore, a recent Eastern paper26 demonstrated that CCC arising in HL medically treated prior to surgery has significantly worse outcomes than concomitant CCC in patients with HL surgically treated. Most patients with CCC and HL had advanced disease. This may explain why in our study patients with CCC and HL showed similar results to patients with only CCC, despite the earlier presentation and the greater rate of resectability in the former group. The most (60.9%) of patients with CCC and HL in our case series, in fact, has been previously submitted to some kind of treatment (endoscopic or surgical) for benign biliary diseases (42.8% of whom for HL), thus decreasing the overall survival in this group. Unfortunately, the small number of patients prevents to perform a subgroup analysis about this issue. Finally, the literature contains few data on disease-free survival in patients affected by CCC and HL. Han et al.16 reported that the 1- and 3-year rates of disease-free survival in these patients were 30.4% and 8.7%, respectively. In the present study, the 1-year disease-free survival rate in patients affected by HL with CCC was lower than that of patients with CCC alone (18.8% vs. 43.4%, respectively), although this difference did not reach statistical significance ( p ¼ 0.144). Our paper had several limitations. Though to our knowledge this is the largest Western case series, this is a small cohort of patients respectively collected from multiinstitutional dataset in a long period of time. The preoperative and perioperative protocols were different through the centers and they changed over the time, reflecting the improvement occurred in hepato-biliary surgery in the last years.23 The short-term and long-term results could be affected and no definitive conclusions can be drawn from this study. However, this is the largest Western cohort
of patients affected by HL and CCC and it can be useful to better define this disease in Western countries. Further studies are also needed to better understand the carcinogenesis pathways that lead cancer from HL. Particularly, studies about biliary stone composition and molecular mechanisms involved in the genetic alterations in this disease are desirable also in Western patients, as such as in Eastern patients.27 Conclusions The patients with HL and CCC showed a high rate of resectability and an early TNM presentation. The risk of postoperative complications, particularly biliary leakage, is higher in these patients compared to patients with CCC alone. Our study confirms the data from series of cases from Eastern countries, in which despite the higher rate of resectability and the earlier presentation of CCC in cases with HL the overall and disease-free survival were similar to those obtained for patients affected by CCC alone. Even in Western countries, HL needs careful management because of the possible presence of concomitant biliary cancer. Conflict of interest statement All authors declare no conflict of interest.
References 1. Shaib Y, El-Serag HB. The epidemiology of cholangiocarcinoma. Semin Liver Dis 2004;24(2):115–25. 2. Okuda K, Nakanuma Y, Miyazaki M. Cholangiocarcinoma: recent progress. Part 1: epidemiology and etiology. J Gastroenterol Hepatol 2002;17(10):1049–55. 3. Tyson GL, El-Serag HB. Risk factors for cholangiocarcinoma. Hepatology 2011;54(1):173–84. 4. Tabrizian P, Jibara G, Shrager B, Schwartz ME, Roayaie S. Hepatic resection for primary hepatolithiasis: a single-center Western experience. J Am Coll Surg 2012;215(5):622–6. 5. Endo I, Gonen M, Yopp AC, et al. Intrahepatic cholangiocarcinoma: rising frequency, improved survival, and determinants of outcome after resection. Ann Surg 2008;248(1):84–96. 6. Zen Y, Sasaki M, Fujii T, et al. Different expression patterns of mucin core proteins and cytokeratins during intrahepatic cholangiocarcinogenesis from biliary intraepithelial neoplasia and intraductal papillary neoplasm of the bile duct e an immunohistochemical study of 110 cases of hepatolithiasis. J Hepatol 2006;44(2):350–8. 7. Falchuk KR, Lesser PB, Galdabini JJ, Isselbacher KJ. Cholangiocarcinoma as related to chronic intrahepatic cholangitis and hepatolithiasis. Case report and review of the literature. Am J Gastroenterol 1976; 66(1):57–61. 8. Donato F, Gelatti U, Tagger A, et al. Intrahepatic cholangiocarcinoma and hepatitis C and B virus infection, alcohol intake, and hepatolithiasis: a case-control study in Italy. Cancer Causes Control 2001;12(10): 959–64. 9. Lesurtel M, Regimbeau JM, Farges O, Colombat M, Sauvanet A, Belghiti J. Intrahepatic cholangiocarcinoma and hepatolithiasis: an
A. Guglielmi et al. / EJSO 40 (2014) 567e575
10.
11.
12.
13. 14.
15.
16.
17. 18.
unusual association in Western countries. Eur J Gastroenterol Hepatol 2002;14(9):1025–7. Nimura Y, Kitagawa Y. Intrahepatic stones. In: Jarnagin WR, editor. Blumgart’s surgery of the liver, pancreas and biliary tract, vol. 1. Saunders; 2005. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240(2):205–13. Sobin LH, Gospodarowicz MK, Wittekind C. International Union against cancer. TNM classification of malignant tumours. 7th ed. Chichester, West Sussex, UK; Hoboken, NJ: Wiley-Blackwell; 2010, p. 309. Kubo S, Kinoshita H, Hirohashi K, Hamba H. Hepatolithiasis associated with cholangiocarcinoma. World J Surg 1995;19(4):637–41. Chen MF, Jan YY, Hwang TL, Jeng LB, Yeh TS. Impact of concomitant hepatolithiasis on patients with peripheral cholangiocarcinoma. Dig Dis Sci 2000;45(2):312–6. Lee CC, Wu CY, Chen GH. What is the impact of coexistence of hepatolithiasis on cholangiocarcinoma? J Gastroenterol Hepatol 2002; 17(9):1015–20. Han SL, Zhou HZ, Cheng J, et al. Diagnosis and surgical treatment of intrahepatic hepatolithiasis associated cholangiocarcinoma. Asian J Surg/Asian Surg Assoc 2009;32(1):1–6. Uenishi T, Hamba H, Takemura S, et al. Outcomes of hepatic resection for hepatolithiasis. Am J Surg 2009;198(2):199–202. Mori T, Sugiyama M, Atomi Y. Gallstone disease: management of intrahepatic stones. Best Pract Res Clin Gastroenterol 2006;20(6): 1117–37.
575
19. Su CH, Shyr YM, Lui WY, P’Eng FK. Hepatolithiasis associated with cholangiocarcinoma. Br J Surg 1997;84(7):969–73. 20. Garai C, editor. Surgical treatment of hilar and intrahepatic cholangiocarcinoma. Milan, Italy: Springer; 2008. 21. Guglielmi A, Ruzzenente A, Campagnaro T, et al. Intrahepatic cholangiocarcinoma: prognostic factors after surgical resection. World J Surg 2009;33(6):1247–54. 22. Guglielmi A, Ruzzenente A, Campagnaro T, et al. Does intrahepatic cholangiocarcinoma have better prognosis compared to perihilar cholangiocarcinoma? J Surg Oncol 2010;101(2):111–5. 23. Ercolani G, Zanello M, Grazi GL, et al. Changes in the surgical approach to hilar cholangiocarcinoma during an 18-year period in a Western single center. J Hepato-biliary-pancreatic Sci 2010;17(3): 329–37. 24. Guglielmi A, Ruzzenente A, Campagnaro T, et al. Prognostic significance of lymph node ratio after resection of peri-hilar cholangiocarcinoma. HPB (Oxf) 2011;13(4):240–5. 25. Nuzzo G, Giuliante F, Ardito F, et al. Improvement in perioperative and long-term outcome after surgical treatment of hilar cholangiocarcinoma: results of an Italian multicenter analysis of 440 patients. Arch Surg 2012;147(1):26–34. 26. Lin CC, Lin PY, Chen YL. Comparison of concomitant and subsequent cholangiocarcinomas associated with hepatolithiasis: clinical implications. World J Gastroenterol 2013;19(3):375–80. 27. Kuroki T, Tajima Y, Kanematsu T. Hepatolithiasis and intrahepatic cholangiocarcinoma: carcinogenesis based on molecular mechanisms. J Hepato-biliary-pancreatic Surg 2005;12(6):463–6.
ScienceDirect EJSO 40 (2014) 567e575
www.ejso.com
Hepatolithiasis-associated cholangiocarcinoma: Results from a multi-institutional national database on a case series of 23 patients A. Guglielmi a, A. Ruzzenente a, A. Valdegamberi a, F. Bagante a, S. Conci a, A.D. Pinna b, G. Ercolani b, F. Giuliante c, L. Capussotti d, L. Aldrighetti e, C. Iacono a,* a
Department of Surgery, Division of General Surgery “A”, “GB Rossi” University Hospital, University of Verona Medical School, Verona, Italy b Department of Surgery and Organ Transplantation, Ospedale Sant’Orsola-Malpighi, University of Bologna, Bologna, Italy c Hepatobiliary Surgery Unit, Department of Surgery, Catholic University of the Sacred Heart School of Medicine, Roma, Italy d Department of Hepato-Biliary-Pancreatic and Digestive Surgery, Ospedale Mauriziano ‘Umberto I’, Torino, Italy e Department of Surgery-Liver Unit, Scientific Institute San Raffaele, Milano, Italy Accepted 6 December 2013 Available online 18 December 2013
Abstract Aims: Few papers focused on association between hepatolithiasis (HL) and cholangiocarcinoma (CCC) in Western countries. The aims of this paper are to describe the clinical presentation, treatment, and postoperative outcomes of CCC with HL in a cohort of Western patients and to compare the surgical outcomes of these patients with patients with CCC without HL. Materials and methods: Among 161 patients with HL from five Italian tertiary hepato-biliary centers, 23 (14.3%) patients with concomitant CCC were analyzed. The results of surgery in these patients were compared with patients with CCC without HL. Results: The 60.9% of patients with HL received the diagnosis of CCC intra- or postoperatively, with a resectability rate of 91.3%. The postoperative morbidity was 61.6%. The 1- and 3-year survival rates were 78.6% and 21.0%, respectively. The recurrence rate was 44.4% and the 3year disease-free survival rates were 18.8%. The comparison with patients with CCC without HL showed a higher resectability rate ( p ¼ 0.02) and a higher frequency of earlier stage ( p ¼ 0.04) in CCC with HL. Biliary leakage was more frequent in CCC with HL group ( p ¼ 0.01) compared to CCC without HL group. We found no differences in overall and disease-free survival between the two groups. Conclusions: Patients with HL and CCC showed a high resectability rate but a higher morbidity. Nevertheless, overall and disease-free survival of patients with CCC and HL showed no differences compared to those of patients with CCC without HL. Also in Western countries, HL needs a careful management for the possible presence of CCC. Ó 2013 Elsevier Ltd. All rights reserved. Keywords: Cholangiocarcinoma; Hepatolithiasis; Liver surgery
Introduction
Abbreviations: HL, hepatolithiasis; CCC, cholangiocarcinoma. * Corresponding author. Department of Surgery, Division of General Surgery “A”, Unit of HPB Surgery, “GB Rossi” University Hospital, University of Verona Medical School, Piazzale L.A. Scuro 10, 37134 Verona, Italy. Tel.: þ39 (0) 458124412. E-mail address: [email protected] (C. Iacono). 0748-7983/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejso.2013.12.006
Cholangiocarcinoma (CCC) is the second most common primary liver cancer.1 Several risk factors have been identified for the development of CCC, including primary sclerosing cholangitis, choledochal cysts, Caroli’s disease, and liver fluke infection. Hepatolithiasis (HL) is associated with a high incidence of CCC and it is one of the major risk factors for the development of CCC.1e3 The incidence of
568
A. Guglielmi et al. / EJSO 40 (2014) 567e575
HL shows a regional distribution, with a greater prevalence in South-East Asian countries. In contrast, HL is fairly uncommon in Western countries with an incidence to 1.3%.4,5 Epidemiological, pathological, and genetic studies have seemed to prove the relationship between HL and CCC. It has been suggested that biliary carcinogenesis is caused by prolonged inflammation of the biliary epithelium secondary to biliary stones, bile stasis, and bacterial infections.6 In the literature, few papers have focused specifically on the association between HL and CCC in Western countries.7e9 No data are available on the clinical presentation, treatment, and outcome of patients with both conditions. The aim of the present paper is to describe the clinical presentation, treatment, and postoperative outcomes of CCC associated to HL in a cohort of patients recruited from a multi-institutional database from tertiary referral centers. Moreover, the long-term results of surgical resection in these patients were compared with the results in patients with CCC without HL. Material and methods From January 1st, 1997 to December 31st, 2011, 161 patients affected by HL were recruited prospectively in five Italian tertiary hepato-biliary centers (Department of Surgery, Division of General Surgery ‘A’, GB Rossi Hospital, University of Verona, Verona; Department of Hepato-BiliaryePancreatic and Digestive Surgery, Ospedale Mauriziano ‘Umberto I’, Turin; Department of Surgery e Liver Unit, Scientific Institute San Raffaele, Milan; Hepatobiliary Surgery Unit, Department of Surgery, Catholic University of the Sacred Heart School of Medicine, Rome; Department of Surgery and Organ Transplantation, Ospedale Sant’Orsola-Malpighi, University of Bologna, Bologna). All the institutions are members of the Italian Chapter of the International Hepato-Pancreato-Biliary Association (IHPBA). From this dataset, 23 (14.3%) patients with concomitant CCC were included into this study. The diagnosis of CCC was pathologically confirmed in all patients. We defined HL as concretions existing in the intrahepatic bile ducts. Although the confluence of the hepatic ducts is situated outside of the parenchyma of the liver, for convenience we included in this study HL of any bile duct proximal to the confluence of the right and left hepatic ducts, as reported by other Authors in literature.10 The following clinical variables were analyzed: age; gender; symptoms; medical history (previous operations on the bile duct, liver fluke infestation, underlying hepatic disease); clinical characteristics of HL (location, time between diagnosis and treatment); laboratory data including serum bilirubin, gamma-glutamyl transferase , preoperative liver function test (indiocyanine green retention rate), presence of tumor markers, histological differentiation, lymph node involvement, and the presence of metastasis. The imaging modalities that were used for the diagnosis of CCC were recorded and analyzed.
After surgery, a computed tomography or magnetic resonance scan was performed every three or six months for follow-up examination of the patients. One patient was lost to follow-up, and the median duration of follow-up was 15 months (range 3e121 months). To evaluate the long-term results of surgical resection of CCC associated with HL, we compared the results of surgery on our cohort with those for a cohort of 414 patients submitted to surgical resection for CCC without HL (peri-hilar or intrahepatic). The data of patients affected by CCC without HL were obtained from a multi-institutional database held by two of the five hepato-biliary tertiary centers (Verona and Bologna) during the same time period. The median follow-up period in these patients was 18 months (range 3e275 months). Statistical analysis The data were collected and analyzed with the statistical software SPSS (version 18.0; SPSS, Chicago, Ill., USA). Continuous variables were compared using an independent two sample t-test. The differences between categorical variables were analyzed with Pearson’s chi-squared test. Survival analyses and survival plots were carried out using the KaplaneMeier method. We considered the day of treatment to be time zero, and data from patients alive at the end of the follow-up period were considered to be censored. Univariate analyses of survival were performed with the KaplaneMeier method, using the log-rank test to verify the significance of differences. Multivariate analysis of survival was carried out with Cox’s regression model. A value of p < 0.05 was considered significant. A propensity score match analysis (1:3) was performed between patients with HL and CCC and patients with CCC alone in order to better define the prognostic value of HL in patients with CCC. Results Demographic data Among the 161 patients with HL observed during the study period 23 with CCC (14.3%) were included into the study. Among these, 14 were men and 9 were women with a mean age of 63.9 years (range 42e82 years). The patients’ characteristics are described in detailed in Table 1. The clinical features are summarized in Table 2. No differences were found between these patients and patients with CCC alone. Symptoms at presentation The mean time from the diagnosis of HL and of malignancy was 61.2 months (range 1e372 months). Seventeen patients (73.1%) developed symptoms due to HL before the diagnosis of CCC. The mean time span between the onset of symptoms and diagnosis of CCC was 17.3 months (range 0e72 months). In most cases HL patients were
Table 1 Clinical characteristics of 23 patients with CCC and HL. Previous procedures involving bile ducts
Symptoms
Time between HL and CCC (months)
HL location
Diagnosis
Type of surgery
T stage
N stage
M stage
Follow up (months)
Alive
Pain, Cholangitis Cholangitis
2
Right liver
Intraop.
Right hepatectomy
3
x
0
42
Yes
24
Bilateral
Postop.
3
0
0
24
No
Liver failure
0
Bilateral
Preop.
3
x
0
10
no
Pd
48
Right liver
Postop.
Bisegmentectomy S2e3 and bilio-digestive anastomosis (further segmentectomy S4e1) Left hepatectomy, segmentectomy S8 Right hepatectomy
2
0
0
36
No
96 6
Left liver Right liver
Postop. Intraop.
Left hepatectomy Right hepatectomy
1 3
0 1
0 0
120 15
Yes No
Ictus cerebri e Pd
372
Left liver
Postop.
Bisegmentectomy
1
0
0
121
No
Pd
Pain, Cholangitis
4
Left liver
Intraop.
4
1
1
6
No
Pd
Pain, Cholangitis Cholangitis, weight loss
180
Right liver
Intraop.
4
0
1
Lost
–
–
252
Left liver
Intraop.
Explorative laparotomy and T-tube placement Explorative laparotomy Left hepatectomy (þS1)
2a
0
0
7
No
Pd
Cholangitis, jaundice Cholangitis None None
1
Right liver
Preop.
Right hepatectomy
1
0
0
18
Yes
47 40 2
Left liver Left liver Left liver
Preop. Preop. Preop.
4 2b 3
0 0 1
0 0 0
12 6 25
Yes Yes No
108 48 2 2 180
Right liver Right liver Right liver Bilateral Right liver
Preop. Postop. Preop. Preop. Postop.
Left hepatectomy Left hepatectomy Left hepatectomy þ S1 þ EBD resection Right hepatectomy Segmentectomy S6e7 Right trisectionectomy Right hepatectomy Bisegmentectomy
2 1 1 3 1
1 x 0 1 x
0 0 0 0 0
3 18 18 4 4
Yes Yes No Yes Yes
Pd
0
Right liver
Intraop.
Right trisectionectomy
4
0
0
4.4
No
Pd
1
M
65
None
2
M
66
ERCP
3
M
75
None
4
M
82
5 6
M M
62 58
Bile duct lithotomy None Cholecystectomy
7
F
71
8
M
66
9
F
42
10
F
54
11
F
51
Cholecystectomy and PTC Cholecystectomy and bilio-digestive anastomosis Multiple ERCPs
12 13 14
M M M
72 75 66
ERCP None None
15 16 17 18 19
M F F M F
72 44 56 74 47
Cholecystectomy Cholecystectomy Cholecystectomy None ERCP
20
F
45
Cholecystectomy and bilio-digestive anastomosis
Trans-duodenal papillotomy and bilio-digestive anastomosis None
Pain, Cholangitis Recurrent cholangitis Cholangitis Recurrent cholangitis Pain , Cholangitis
None None Cholangitis None Cholangitis Pain Cholangitis
Cause of death
A. Guglielmi et al. / EJSO 40 (2014) 567e575
Age
Pd
569
(continued on next page)
Postop. Postop. Left liver Left liver 4 2 Cholangitis Jaundice 81 71 F M 22 23
74 M 21
ERCP: Endoscopic retrograde cholangiopancreatography; EBD: extrahepatic bile duct; Pd: progression disease; Preop.: preoperative; Intraop.: intraoperative; Postop.: postoperative; PTC: percutaneous transhepatic cholangiography.
Pd No Yes 19.4 60 2 1
0
0 0
Pd No 8.7 0 Cholangitis
Cholecystectomy and bilio-digestive anastomosis None None
0
Right liver
Preop.
Right hepatectomy and bilio-digestive anastomosis Bisegmentectomy EBD resection
3
0
Follow up (months) Symptoms Previous procedures involving bile ducts Age
Table 1 (continued )
Time between HL and CCC (months)
HL location
Diagnosis
Type of surgery
T stage
N stage
M stage
Cause of death
A. Guglielmi et al. / EJSO 40 (2014) 567e575
Alive
570
asymptomatic and no specific treatment was performed. The symptomatic patients were submitted to endoscopic or radiological procedures, as such as endoscopic retrograde cholangiopancreatography (ERPC) or percutaneous transhepatic cholangiography (PTC). Surgery was chosen when non operative treatment failed (see Table 1). Laboratory data and tumor markers On laboratory testing, 14 patients (75%) showed elevated serum GGT (>45 U/L; mean value of 179 U/L; range 24e517). The median values of the tumor markers CA19.9 and CEA were 85.45 U/L and 3.7 U/L, respectively (range 6e12153 and 0.8e7, respectively). Diagnosis of malignancy All patients underwent preoperative abdominal ultrasonography. Magnetic-resonance cholangio-pancreatography Table 2 Demographic data. Features
N (%)
M:F Age
14:9 63.9 (range 42e82)
Underlying liver disease Yes No
3 (13) 20 (87)
Previous biliary surgery Yes No
14 (60.9) 9 (39.1)
Liver fluke infestation Yes No
0 23
Symptoms None Cholangitis Pain Jaundice alone Weight loss
5 (21.7) 17 (73.9) 6 (26.1) 1 (4.3) 1 (4.3)
Duration of symptoms GGT (U/L, median value) Total serum bilirubin (mg/dL, median value) CEA (U/L, median value) Ca19.9 (kU/L, median value) Location of HL Right liver Left liver Bilateral Time between diagnosis of HL and CCC Diagnosis of CCC Preoperative Intraoperative Postoperative a
Inter-quartile range.
17.3 months (1e48) 105 (36.75e346.75)a 0.95 (0.8e1.1)a 3.7 (1.9e5.1)a 85.45 (6e12153)a 11 (47.9) 9 (39.1) 3 (13) 5.1 years (range 1e372 months)
9 (39.1) 6 (26.1) 8 (34.8)
A. Guglielmi et al. / EJSO 40 (2014) 567e575
571
Figure 1. AeBeC: Case n. 15 in Table 1. Preoperative diagnosis of CCC with HL. Preoperative magnetic resonance cholangio-pancreatography revealed an intrahepatic mass-forming CCC (arrow in A) in the right lobe of liver with strictures caused by HL (arrow in B). Right hepatectomy was performed. Control CT scan showed no intrahepatic recurrence (C).
was performed in 19 patients (82.6%, Fig. 1). A computed tomography scan was performed in 18 patients (78.3%) and a positron emission tomography-computed tomography scan in two patients (8.7%). A preoperative diagnosis of CCC was obtained in 39.1% of the cases (nine patients). Intraoperative diagnosis was performed in six patients (26.1%), using intraoperative ultrasonography (one case) and intraoperative frozen sections (five cases). Postoperative pathological examination of the specimen led to a diagnosis of cancer in eight patients (34.8%). According to the CCC macroscopic type 20 of the 23 patients with CCC and HL (86.9%) had an intrahepatic massforming CCC. Only three cases were diagnosed as perihilar CCC (13.1%). Resectability, morbidity, and mortality rates Resectability was 91.3% (21 patients), two patients were not submitted for resection because of the presence of metastases along the diaphragmatic peritoneum. Radical resection (R0) was performed in 17 cases (80.9%). One patient underwent a second operation to achieve R0 after the histological diagnosis of CCC: the patient underwent bisegmentectomy of S2e3 for HL without any suspicion of cancer: definitive pathological examination revealed the diagnosis of CCC associated to HL with positive resection margins, for this reason further bisegmentectomy S4e1 was performed to achieve curative resection. The rate of resectability in the group without HL was 68.8%. This appeared to be significantly lower than the rate of resectability in patients with both CCC and HL (91.3%, p ¼ 0.02). Among patients who underwent surgery for CCC without HL, radical resection was achieved in 179 patients (75.5%), which showed no significant difference from patients with both CCC and HL ( p ¼ 0.1). In patients with both CCC and HL the overall morbidity rate was 61.9%, and all postoperative complications were of grade II according to the DindoeClavien classification.11
The most frequent complication was biliary leakage (23.8%), followed by intra-abdominal collection (19.5%), wound infection (14.3%), and pleural effusion (4.3%). There was no perioperative mortality. When the patients with both CCC and HL were compared with these patients without HL, we identified a significant difference in the rate of postoperative complications (61.9% vs. 41.2% respectively; p ¼ 0.02; Table 3). In particular, patients with HL showed a higher rate of biliary leakage (23.8% vs. 11.3%, respectively; p < 0.01). The rate of other postoperative complications (such as pneumonia, abdominal abscess, pleural effusion, etc.) did not show a statistically significant difference (38.1% vs. 29.9%, respectively; p > 0.05; Table 3). Postoperative mortality in CCC without HL was 6.3% (18 patients), this difference was not statistically significant compared to CCC with HL ( p ¼ 0.37). TNM staging and histological grading On the basis of the tumorenodeemetastasis (TNM) classification, 7th edition,12 seven patients (33.3%) were classified as T1, eight patients (38.1%) as T2, five patients (23.8%) as T3, and only one patient (4.8%) as T4. The median number of lymph-nodes harvested was 4 (range 0e21), but the median number of positive lymph-nodes was 0 (range 0e10). Lymphadenectomy was not performed in four patients because no suspicion of cancer nor lymph node enlargement was found during surgical exploration. Fourteen patients were classified as N0 and five patients as N1 (61.9% and 19%, respectively). Two patients (8.7%) had distant metastasis, in the form of diaphragmatic peritoneum in both cases. In total, 47.8% of patients were in stages IeII according to the TNM system (11 patients), and 52.2% (12 patients) were in stages IIIeIVaeb. Comparison of the tumor stage and histological grading between patients with CCC and HL and those with CCC without HL is reported in Table 3. The histological
572
A. Guglielmi et al. / EJSO 40 (2014) 567e575
Table 3 Comparison of intraoperative and pathological features, early and longterm results of patients affected by CCC with and alone. Feature
Number of patients with CCC and HL (%)
Number of patients with CCC alone (%)
p
Resectability
21 (91.3)
284 (68.8)
0.02
Radicality R0 R1
17 (80.9) 2 (9.5)
179 (75.5) 79 (24.5)
Tumor location Intrahepatic Peri-hilar
20 (86.9) 3 (13.1)
171 (41.4) 242 (58.6)
Macro vascular invasion Portal resection/ reconstruction
7 (39.1)
94 (38.2)
0.41
1 (4.3)
28 (12)
0.86
Overall morbidity Biliary leakage Others
13 (61.9) 5 (23.8) 8 (38.1)
117 (41.2) 32 (11.3) 85 (29.9)
0.02 0.01 0.23
Mortality
0
18 (6.3)
0.37
Histological grade G1e2 G3e4
16 (69.6) 7 (30.4)
174 (61.2) 110 (38.8)
Stage T according TNM 7th edition T1 T2 T3 T4 Stage N according TNM 7th edition N0 N1 Stage M according TNM 7th edition M0 M1 Stage TNM 7th edition Stage IeII Stage IIIeIV Overall median survival time Overall survival 1-year 3-year 5-year Recurrence rate Disease-free median survival time Disease-free survival 1-year 3-year
0.10
0.33
0.30
0.04 7 8 5 1
(33.3) (38.1) (23.8) (4.8)
27 (10.5) 165 (64.1) 62 (24.2) 3 (1.2) 0.59
14 (61.9) 5 (19)
142 (67.3) 69 (32.7) 0.60
21 (91.3) 2 (8.7)
261 (91.9) 23 (8.1) 0.14
11 (47.8) 12 (52.2)
125 (30.3) 286 (69.7)
24 mo.
25.2 mo.
78.6% 21.0% 21.0%
78.1% 50.8% 30.4%
47.6% 6.0 mo.
51.0% 20.7 mo.
18.8% 18.8%
43.4% 21.6%
0.78
0.61
0.14
evaluation revealed that 30.3% of the latter group of patients were in stage IeII according to the TNM system, and 69.7% were in stage IIIeIV. The difference in the distribution of tumor stage between patients with both HL and CCC and those with CCC alone did not reach statistical significance ( p ¼ 0.14). The subgroup analysis showed that 10.5% of the patients with CCC alone were in stage T1, compared with 33.3% of the patients with both CCC and HL ( p < 0.05). However, the analysis of other histological characteristics revealed no significant differences between the groups. Overall and disease-free survival The overall median survival time was 24 months, with 1, 3-, and 5-year actuarial survival rates of 78.6%, 21.0%, and 21.0%, respectively. In patients with CCC alone, the overall median survival time was 25.2 months, with 1-, 3, and 5-year actuarial survival rates of 78.1%, 50.8%, and 30.4%, respectively. Overall survival was not significantly different in patients with CCC and HL from that in patients with CCC alone ( p ¼ 0.78, Fig. 2). The rate of recurrence in patients with HL and CCC was 47.6% (10 out of 21 patients) and the 1- and 3-year diseasefree survival rates were 18.8% and 18.8%, respectively. Regarding the site of recurrence three patients had intrahepatic recurrence, two patients had lymph-nodal recurrence and 5 patients had distant recurrence. In patients with CCC alone, the 1-, 3-, and 5-year disease-free survival rates were 43.4%, 21.6%, and 19.2%, respectively. The 1- and 3-year disease-free survival in patients with CCC and HL appeared to be lower than that in patients with CCC without HL although the difference did not reach statistical significance ( p ¼ 0.14, Fig. 2). We performed a sub-analysis that takes into account the stage of tumor, namely T stage. In stages T1 and T2, median OS was 18.00 months (CI 95%, 3.59e32.40) for patients with CCC and HL and 42.63 months (CI 95%, 33.62e51.64) for patients with CCC without HL (HR 2.75, CI 95% 1.18e6.42, p ¼ 0.019). In stages T3 and T4, median OS was 24.33 months (CI 95%, 5.20e43.46) for patients with CCC and HL and 20.06 months (CI 95%, 13.12e27.00) for patients with CCC without HL ( p ¼ 0.878). To confirm our results a propensity matched analysis was performed. Before matching the minimum p value was computed among T stage ( p ¼ 0.006), after matching the balance analysis does not reveal difference among covariates in groups of patients with or without HL. HL resulted an important prognostic factor in terms of OS for patients with CCC with an HR of 2.566 (CI 95%, 1.25e5.28, p ¼ 0.01). Discussion Hepatolithiasis is frequent in Eastern countries, in which its risk factors are endemic. Consequently, most of the
A. Guglielmi et al. / EJSO 40 (2014) 567e575
573
Figure 2. Comparison of overall survival and disease-free survival of CCC patients with (N ¼ 21) and without (N ¼ 414) HL.
knowledge about this disease and its correlation with CCC comes from studies of Eastern patients.13e16 This disease is rare in Western countries, and the literature contains few data about the presentation, features, treatment, and outcome of HL associated with CCC in Western populations. Although the symptoms, blood tests, and imaging results might suggest a diagnosis of CCC in patients with HL, the preoperative diagnostic accuracy varies from 20% to 70%.16,17 This variation in accuracy might be due to the pattern of growth of CCC, which may infiltrate the bile duct without mass formation, or because the stricture or stenosis that is caused by CCC can be difficult to differentiate from that caused by stones.18 We report herein that preoperative diagnosis was achieved in only 39.1% of our patients, which is lower than the rate of 68.4% reported by Eastern authors.15 The reason for this might be that in Western countries, preoperative cholangioscopy is not used routinely in the diagnosis and treatment of HL. The reported resectability rate14,15,19 for CCC and HL is higher than that for CCC without HL. Lee et al.15 showed a resectability rate of 52.6% in patients treated for HLassociated CCC and of 39.2% in patients with CCC without HL ( p < 0.001). In addition, Chen et al.14 showed a higher rate of resectability for tumors in patients with CCC and HL than in patients who had CCC without HL (31.1% vs. 26.8%). In accordance with the literature, we report herein that patients affected by CCC with HL had a higher rate of resectability than patients with CCC alone (91.3% vs. 68.8%; p ¼ 0.02). The higher rate of resectability in the present data in comparison with other studies probably reflects the improvements in intraoperative and postoperative management that have occurred during the last ten years.20e25 In the present study, more patients affected by both HL and CCC were in early T stage than those with CCC alone (33.3% vs. 10.5%, respectively; p ¼ 0.04). This confirms the data from the Eastern literature. Chen et al. showed
that 33.0% of patients with HL and CCC presented at TNM stages IeII compared with 12.5% of patients with CCC alone ( p < 0.05). A possible explanation for that is that the presence of symptoms due to HL could lead to an earlier diagnosis of CCC.14 In literature, surgical treatment is associated with a higher postoperative morbidity in CCC in patients with HL than in patients who undergo surgery for CCC without HL: wound infection and bile leakage are three times more likely in the former patients, probably because of the previous cholangitis and the abnormalities of the bile ducts that are related to HL.14,15,19 The postoperative outcomes in the present study were in agreement with these results. The rate of morbidity was significantly higher, especially biliary leakage, in patients with CCC and HL than in patients with CCC only, though all the complications were of stage II according to the Dindo classification. Nevertheless, the rate of postoperative mortality reported in the literature is low and no differences were shown between the two groups of patients in this respect. Chen et al.14 reported a surgical mortality rate of 3.8% in 106 patients affected by CCC with HL and 3.6% in 56 patients alone. Our findings confirmed these data: the mortality rate was comparable between patients with both CCC and HL and those with CCC alone (0.0% and 6.3% respectively, p ¼ 0.37). Despite a reported higher rate of resectability and the earlier stage at diagnosis, recent data from the literature seem to show no differences in overall survival between patients with CCC and HL and patients with CCC alone. Although Su et al.19 reported significantly poorer survival in patients with CCC and HL, Chen et al.14 stated that the overall survival of patients with resected CCC and HL did not differ from that of patients with CCC but alone, with a 5-year survival rate of 16.5% in the former and 7.8% in the latter. These data were not statistically different also when stratified by different tumor stages, or by the type of treatment received (hepatic resection or palliative). In
574
A. Guglielmi et al. / EJSO 40 (2014) 567e575
addition, Lee et al.,15 in a study of 38 patients with CCC with HL and 102 patients affected only by CCC, showed a 5-year survival rate of 20.0% for patients both with and alone. We also did not find statistically significant differences between the two groups with respect to survival: the 5year survival rates in patients affected by CCC with HL and those alone were 21.0% and 30.4%, respectively ( p ¼ 0.78). In agreement with the reports in the Eastern literature, we conclude that, in patients with CCC from Western countries, HL does not impair long-term survival, nevertheless it allows an earlier presentation and greater resectability rate. One possible explanation of this could be the inadequate lymph-node harvesting: in patients with HL and CCC, the median number of lymph-nodes harvested was 4, but the median number of positive lymph-nodes was 0 (see Table 1). Moreover, in four patients in which CCC was discovered postoperatively, lymphadenectomy was not performed because no suspicion of cancer nor lymph node enlargement was found during surgical exploration. We believe that the lymph-node involvement was underestimated in these patients. This could explain the low rates of overall and disease-free survival. Furthermore, a recent Eastern paper26 demonstrated that CCC arising in HL medically treated prior to surgery has significantly worse outcomes than concomitant CCC in patients with HL surgically treated. Most patients with CCC and HL had advanced disease. This may explain why in our study patients with CCC and HL showed similar results to patients with only CCC, despite the earlier presentation and the greater rate of resectability in the former group. The most (60.9%) of patients with CCC and HL in our case series, in fact, has been previously submitted to some kind of treatment (endoscopic or surgical) for benign biliary diseases (42.8% of whom for HL), thus decreasing the overall survival in this group. Unfortunately, the small number of patients prevents to perform a subgroup analysis about this issue. Finally, the literature contains few data on disease-free survival in patients affected by CCC and HL. Han et al.16 reported that the 1- and 3-year rates of disease-free survival in these patients were 30.4% and 8.7%, respectively. In the present study, the 1-year disease-free survival rate in patients affected by HL with CCC was lower than that of patients with CCC alone (18.8% vs. 43.4%, respectively), although this difference did not reach statistical significance ( p ¼ 0.144). Our paper had several limitations. Though to our knowledge this is the largest Western case series, this is a small cohort of patients respectively collected from multiinstitutional dataset in a long period of time. The preoperative and perioperative protocols were different through the centers and they changed over the time, reflecting the improvement occurred in hepato-biliary surgery in the last years.23 The short-term and long-term results could be affected and no definitive conclusions can be drawn from this study. However, this is the largest Western cohort
of patients affected by HL and CCC and it can be useful to better define this disease in Western countries. Further studies are also needed to better understand the carcinogenesis pathways that lead cancer from HL. Particularly, studies about biliary stone composition and molecular mechanisms involved in the genetic alterations in this disease are desirable also in Western patients, as such as in Eastern patients.27 Conclusions The patients with HL and CCC showed a high rate of resectability and an early TNM presentation. The risk of postoperative complications, particularly biliary leakage, is higher in these patients compared to patients with CCC alone. Our study confirms the data from series of cases from Eastern countries, in which despite the higher rate of resectability and the earlier presentation of CCC in cases with HL the overall and disease-free survival were similar to those obtained for patients affected by CCC alone. Even in Western countries, HL needs careful management because of the possible presence of concomitant biliary cancer. Conflict of interest statement All authors declare no conflict of interest.
References 1. Shaib Y, El-Serag HB. The epidemiology of cholangiocarcinoma. Semin Liver Dis 2004;24(2):115–25. 2. Okuda K, Nakanuma Y, Miyazaki M. Cholangiocarcinoma: recent progress. Part 1: epidemiology and etiology. J Gastroenterol Hepatol 2002;17(10):1049–55. 3. Tyson GL, El-Serag HB. Risk factors for cholangiocarcinoma. Hepatology 2011;54(1):173–84. 4. Tabrizian P, Jibara G, Shrager B, Schwartz ME, Roayaie S. Hepatic resection for primary hepatolithiasis: a single-center Western experience. J Am Coll Surg 2012;215(5):622–6. 5. Endo I, Gonen M, Yopp AC, et al. Intrahepatic cholangiocarcinoma: rising frequency, improved survival, and determinants of outcome after resection. Ann Surg 2008;248(1):84–96. 6. Zen Y, Sasaki M, Fujii T, et al. Different expression patterns of mucin core proteins and cytokeratins during intrahepatic cholangiocarcinogenesis from biliary intraepithelial neoplasia and intraductal papillary neoplasm of the bile duct e an immunohistochemical study of 110 cases of hepatolithiasis. J Hepatol 2006;44(2):350–8. 7. Falchuk KR, Lesser PB, Galdabini JJ, Isselbacher KJ. Cholangiocarcinoma as related to chronic intrahepatic cholangitis and hepatolithiasis. Case report and review of the literature. Am J Gastroenterol 1976; 66(1):57–61. 8. Donato F, Gelatti U, Tagger A, et al. Intrahepatic cholangiocarcinoma and hepatitis C and B virus infection, alcohol intake, and hepatolithiasis: a case-control study in Italy. Cancer Causes Control 2001;12(10): 959–64. 9. Lesurtel M, Regimbeau JM, Farges O, Colombat M, Sauvanet A, Belghiti J. Intrahepatic cholangiocarcinoma and hepatolithiasis: an
A. Guglielmi et al. / EJSO 40 (2014) 567e575
10.
11.
12.
13. 14.
15.
16.
17. 18.
unusual association in Western countries. Eur J Gastroenterol Hepatol 2002;14(9):1025–7. Nimura Y, Kitagawa Y. Intrahepatic stones. In: Jarnagin WR, editor. Blumgart’s surgery of the liver, pancreas and biliary tract, vol. 1. Saunders; 2005. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240(2):205–13. Sobin LH, Gospodarowicz MK, Wittekind C. International Union against cancer. TNM classification of malignant tumours. 7th ed. Chichester, West Sussex, UK; Hoboken, NJ: Wiley-Blackwell; 2010, p. 309. Kubo S, Kinoshita H, Hirohashi K, Hamba H. Hepatolithiasis associated with cholangiocarcinoma. World J Surg 1995;19(4):637–41. Chen MF, Jan YY, Hwang TL, Jeng LB, Yeh TS. Impact of concomitant hepatolithiasis on patients with peripheral cholangiocarcinoma. Dig Dis Sci 2000;45(2):312–6. Lee CC, Wu CY, Chen GH. What is the impact of coexistence of hepatolithiasis on cholangiocarcinoma? J Gastroenterol Hepatol 2002; 17(9):1015–20. Han SL, Zhou HZ, Cheng J, et al. Diagnosis and surgical treatment of intrahepatic hepatolithiasis associated cholangiocarcinoma. Asian J Surg/Asian Surg Assoc 2009;32(1):1–6. Uenishi T, Hamba H, Takemura S, et al. Outcomes of hepatic resection for hepatolithiasis. Am J Surg 2009;198(2):199–202. Mori T, Sugiyama M, Atomi Y. Gallstone disease: management of intrahepatic stones. Best Pract Res Clin Gastroenterol 2006;20(6): 1117–37.
575
19. Su CH, Shyr YM, Lui WY, P’Eng FK. Hepatolithiasis associated with cholangiocarcinoma. Br J Surg 1997;84(7):969–73. 20. Garai C, editor. Surgical treatment of hilar and intrahepatic cholangiocarcinoma. Milan, Italy: Springer; 2008. 21. Guglielmi A, Ruzzenente A, Campagnaro T, et al. Intrahepatic cholangiocarcinoma: prognostic factors after surgical resection. World J Surg 2009;33(6):1247–54. 22. Guglielmi A, Ruzzenente A, Campagnaro T, et al. Does intrahepatic cholangiocarcinoma have better prognosis compared to perihilar cholangiocarcinoma? J Surg Oncol 2010;101(2):111–5. 23. Ercolani G, Zanello M, Grazi GL, et al. Changes in the surgical approach to hilar cholangiocarcinoma during an 18-year period in a Western single center. J Hepato-biliary-pancreatic Sci 2010;17(3): 329–37. 24. Guglielmi A, Ruzzenente A, Campagnaro T, et al. Prognostic significance of lymph node ratio after resection of peri-hilar cholangiocarcinoma. HPB (Oxf) 2011;13(4):240–5. 25. Nuzzo G, Giuliante F, Ardito F, et al. Improvement in perioperative and long-term outcome after surgical treatment of hilar cholangiocarcinoma: results of an Italian multicenter analysis of 440 patients. Arch Surg 2012;147(1):26–34. 26. Lin CC, Lin PY, Chen YL. Comparison of concomitant and subsequent cholangiocarcinomas associated with hepatolithiasis: clinical implications. World J Gastroenterol 2013;19(3):375–80. 27. Kuroki T, Tajima Y, Kanematsu T. Hepatolithiasis and intrahepatic cholangiocarcinoma: carcinogenesis based on molecular mechanisms. J Hepato-biliary-pancreatic Surg 2005;12(6):463–6.
