Liver transplantation in patients with incidental ... - ScienceDirect

Original Article / Transplantation

Hepatobiliary & Pancreatic Diseases International

Liver transplantation in patients with incidental hepatocellular carcinoma/cholangiocarcinoma and intrahepatic cholangiocarcinoma: a single-center experience Mohammed Elshamy, Naftali Presser, Abdulrahman Y Hammad, Daniel J Firl, Christopher Coppa, John Fung and Federico N Aucejo Cleveland, USA

BACKGROUND: Reports of liver transplantation (LT) in patients with mixed hepatocellular carcinoma/cholangiocarcinoma (HCC/CC) and intrahepatic cholangiocarcinoma (ICC) are modest and have been mostly retrospective after pathological categorization in the setting of presumed HCC. Some studies suggest that patients undergoing LT with small and unifocal ICC or mixed HCC/CC can achieve about 40%-60% 5-year post-transplant survival. The study aimed to report our experience in patients undergoing LT with explant pathology revealing HCC/CC and ICC. METHODS: From a prospectively maintained database, we performed cohort analysis. We identified 13 patients who underwent LT with explant pathology revealing HCC/CC or ICC.

date their pathology. Identified patients with early HCC/CC or ICC may benefit from LT if unresectable. Additionally, incorporating adjunctive perioperative therapies such as in the case of patients undergoing LT with hilar cholangiocarcinoma may improve outcomes but this warrants further investigation. (Hepatobiliary Pancreat Dis Int 2017;16:264-270) KEY WORDS: mixed hepatocellular carcinoma/ cholangiocarcinoma; liver transplantation; survival outcomes

RESULTS: The observed recurrence rate post-LT was 31% Introduction (4/13) and overall survival was 85%, 51%, and 51% at 1, 3 and ixed hepatocellular carcinoma/cholangiocar5 years, respectively. Disease-free survival was 68%, 51%, and cinoma (HCC/CC) and intrahepatic chol41% at 1, 3 and 5 years, respectively. In our cohort, four paangiocarcinoma (ICC) are two pathological tients would have qualified for exception points based on upsubgroups of primary liver tumors that can be observed dated HCC Organ Procurement and Transplantation Network in patients undergoing liver transplantation (LT).[1, 2] imaging guidelines.

M

Reports of LT in patients with these tumor subtypes

CONCLUSIONS: Lesions which lack complete imaging characteristics of HCC may warrant pre-LT biopsy to fully eluci- are modest and have been mostly retrospective after

Author Affiliations: Hepatobiliary and Transplant Surgery, Digestive Disease and Surgery Institute (Elshamy M, Presser N, Hammad AY, Firl DJ, Fung J and Aucejo FN), and Imaging Institute (Coppa C), Cleveland Clinic, Cleveland, OH, USA Corresponding Author: Federico N Aucejo, MD, Hepatobiliary and Transplant Surgery, Digestive Disease and Surgery Institute, Cleveland Clinic, 9500 Euclid Ave. /A100, Cleveland, OH 44195, USA (Tel: +216-445-8021; Email: [email protected]) © 2017, Hepatobiliary Pancreat Dis Int. All rights reserved. doi: 10.1016/S1499-3872(17)60016-X Published online May 3, 2017.

pathological categorization in the setting of presumed HCC.[3, 4] Although initially embraced as a mean of cure for a disease with a dismal prognosis, early experience with LT for ICC was hampered by high recurrence rates and low overall survival such that ICC became a contraindication for LT.[5, 6] Nevertheless, some experience would suggest that patients undergoing LT with small and unifocal ICC (presumed HCC) can have adequate post-transplant survival.[7] Similarly, the natural history of mixed HCC/CC is as yet incompletely defined with studies revealing conflicting results as to the behavior of these lesions.[8, 9] Since HCC/CC lies somewhere between pure HCC and ICC

264 • Hepatobiliary Pancreat Dis Int，Vol 16，No 3 • June 15，2017 • www.hbpdint.com

Liver transplantation for HCC/CC and ICC

from a tumor biology standpoint, LT could be considered as a surgical option in selected patients who are not candidates for resection. Scarce published data about LT in patients with mixed tumors appear to be supportive of this assumption.[5, 10] Of note, no standard adjunctive therapies in the setting of LT for HCC/CC or ICC have been established and whether that would improve survival in this subset of patients remains uncertain. Herein we report our center’s experience in patients undergoing LT with explant pathology revealing HCC/CC and ICC, and a review of the literature about this subject.

scribed in Tables 1 and 2. Eight patients were male (62%), mean age was 60.0±1.7 years, and hepatitis C virus was the most common etiology of cirrhosis (n=5). The four patients with ICC had underlying cirrhosis from primary sclerosing cholangitis (n=2), secondary sclerosing cholangitis (1) and autoimmune hepatitis (1) (Table 1).

Explant histopathology Thirteen patients had 20 tumors on explant histopathology. Four patients had just one HCC/CC tumor, Table 1. Demographic and tumor characteristics of patients receiving liver transplant with incidental findings of HCC/CC or ICC on explant pathology

Methods A Cleveland Clinic prospectively maintained and Institutional Review Board approved LT and liver tumor related database was queried. Medical records of patients undergoing LT from January 2005 to December 2015 were reviewed and patients who underwent LT with explant pathology of mixed HCC/CC or ICC were included. Histopathologic characteristics of mixed HCC/CC and ICC including number and size, degree of differentiation and presence of vascular invasion were analyzed. Patient demographics, underlying liver diseases, model for endstage liver disease (MELD) score, pre-transplant diagnosis of HCC and pre-transplant locoregional therapy were assessed. Radiologic pre-transplant tumor characteristics were addressed using computed tomography (CT) and/or magnetic resonance imaging (MRI), the results of which were reviewed by one experienced radiologist (Coppa C). Pre-transplant tumor markers including alpha-fetoprotein (AFP) and carbohydrate antigen 19-9 (CA19-9) levels were analyzed. Follow-up was calculated from time of LT to last clinic visit or death. Post-LT overall survival and disease-free survival were assessed. For patients who developed disease recurrence, time from LT to disease recurrence, site of recurrence, treatment and outcome were described. Categorical variables were expressed as number and continuous variables were expressed as mean±standard error measurement (SEM). Analysis of overall survival and disease-free survival was performed utilizing the Kaplan-Meier method. Statistical analysis was performed with SPSS (version 17, SPSS Inc., Chicago, IL, USA).

Results Table 1 summarizes the characteristics of 13 identified HCC/CC or ICC cases from our 1341 transplants over the interval examined. Nine patients had HCC/CC and four had ICC. Patients and tumor characteristics are de-

Variables

Mixed ICC HCC/CC (n=4) (n=9)

Age at Tx (yr) Gender (male/female) Underlying liver diseases HCV NASH HCV+ALD HBV PSC SSC AIH Diagnosed as HCC pre-Tx (Yes/No) MELD at Tx Waitlist time (mon) Number of lesions on CT scan Size of largest lesion on CT scan (cm)

58.8±1.4 7/2

AFP pre-Tx (IU/mL) CA19-9 pre-Tx (IU/mL)

63.8±17.7 2.1±1.7 44.8±14.6 42.8±39.2 159.3±299.3 101.0±207.2

Hepatic resection pre-Tx (Yes/No)

5 2 1 1 0 0 0

62.5±4.7 1/3

Total (n=13)

0 0 0 0 2 1 1

60.0±1.7 8/5

9/0

2/2

5 2 1 1 2 1 1 11/2

21.5±3.3 1.5±0.3 1.6±0.3 2.5±0.2

24.2±5.9 9.2±7.3 0.5±0.3 1.1±0.7

23.0±5.3 3.9±2.3 1.3±0.3 2.1±0.3

0/9

1/3

1/12

LRT pre-Tx (Yes/No) Tx type (OLT/LDLT) Pathology lesion number Pathology largest lesion (cm)

5/4 9/0 2.3±0.5 2.8±0.5

2/2 3/1 1.0±0 2.7±0.4

7/6 12/1 2.0±0.4 2.8±0.4

Tumor differentiation (Well/Moderate/Poor)

1/6/2

0/3/1

1/9/3

3/6 3/6

1/3 1/3

4/9 4/9

Vascular invasion (Yes/No) Recurrence (Yes/No)

HCC/CC: hepatocellular carcinoma/cholangiocarcinoma; ICC: intrahepatic cholangiocarcinoma; HCV: hepatitis C virus; NASH: nonalcoholic steatohepatitis; ALD: alcoholic liver disease; HBV: hepatitis B virus; PSC: primary sclerosing cholangitis; SSC: secondary sclerosing cholangitis; AIH: autoimmune hepatitis; AFP: alpha-fetoprotein; CA19-9: carbohydrate antigen 19-9; LRT: locoregional therapy; Tx: transplant; OLT: orthotopic liver transplantation; LDLT: living donor liver transplantation.

Hepatobiliary Pancreat Dis Int，Vol 16，No 3 • June 15，2017 • www.hbpdint.com • 265


three patients had just one ICC tumor and five patients had multifocal tumor including ICC, pure HCC and ICC, or pure HCC and HCC/CC. One patient had multiple microscopic foci of ICC.

reported in three patients (two patients with just one HCC/CC and one patient with two pure HCC and one ICC). Two patients with ICC did not have pre-LT imaging of a mass and underwent transplantation for chronic end-stage liver disease (Table 2). Seven (54%) patients underwent pre-LT locoregional Pre-LT imaging and pre-LT treatments therapy via chemoembolization. One patient underwent Preoperative imaging was examined. Four patients had lesions that fulfilled Organ Procurement and Trans- left lateral segmentectomy for recurrent cholangitis with plantation Network (OPTN) class 5 lesion characteristics imaging findings showing segmental biliary dilatation for four pure HCC lesions (two patients) and one HCC/ and left lateral segment atrophy. CC lesion (two patients). In five patients, there were lesions that did not meet OPTN class 5 criteria: three ICC, Post-LT follow-up one pure HCC and one HCC/CC. Due to insufficient Mean post-LT follow-up was 35.6±8.0 months. Overimaging technique, no OPTN class 5 designation was all survival was 85%, 51% and 51% at 1, 3 and 5 years, Table 2. Radiologic and pathologic features of liver tumors Cause of Patient liver disease

Lesion number/ location

1

NASH

1/Segment VIII PV

1.5-cm hypoenhancing N/A mass with upstream biliary ductal dilation

2

HCV

1/Segment VI

N/A

3

HCV

1/Segment VIII NC, PV

3-cm lesion with brisk N/A enhancement relative to background (hyperenhancing)

1. Fatty liver with cirrhosis 2. HCC vs metastasis

2/Segment II

1.1-cm lesion with brisk enhancement relative to background (hyperenhancing)

HCC vs metastasis

HCC

3/Segment V

0.9-cm lesion with brisk enhancement relative to background (hyperenhancing)

HCC vs metastasis

ICC

4

HCV+ ALD

1/Segment II

CT technique

N/A

NC, Art, PV

2/Segment VIII

5

HCV

1/Segment II

2/Segment V

NC, Art, PV, delayed (Chest CT)

CT features of lesion

2.7-cm hypervascular mass, isodense to background on venous phase

MRI features of lesion

Within Milan Pathology criteria

Radiologic impression

OPTN class 5

1. Cirrhosis 2. malignantappearing mass

insufficient Yes data

Mixed HCC/CC (4.2 cm)

Unable to retrieve: “2.7 1. Cirrhosis insufficient Yes ×2.4 cm hyperen2. Mass suspicious data hancing mass” for HCC

Mixed HCC/CC

2.5-cm hypervascular 1. Cirrhosis mass with heteroge- 2. HCC neous wash-out on delayed phase; STIR signal hyperintensity

insufficient Yes data

Yes

Yes

HCC

HCC/CC

1.5-cm hypervascular 1.5-cm hypervascular Indeterminate for lesion without wash- lesion without wash- HCC out out

No

HCC

2-cm hypervascular N/A lesion with wash-out and capsule

1. Cirrhosis 2. HCC

Yes

Yes

HCC

1.9-cm peripherally enhancing hypervascular lesion with delayed, progressive enhancement

HCC vs ICC

No

Yes

ICC

(To be continued)



Table 2. Radiologic and pathologic features of liver tumors Cause of Patient liver disease

Lesion number/ location

6

NASH

1/Segment IV

NC, Art, PV, 3-cm hypovascular delayed mass all phases (Chest CT)

7

PSC

1/Segment VII

8

AIH

9

Radiologic impression

N/A

1. Cirrhosis No 2. Lesion suspicious but not diagnostic of HCC

Yes

Mixed HCC/CC (HCC predominant)

NC, Art, PV, 4-cm poorly circum- 3.3-cm mass with delayed scribed hypoenhanc- T2 signal hyper(Chest CT) ing mass intensity nodular hypervascularity with progressive enhancement

1. PSC cirrhosis No 2. Indeterminate lesion - DDX sclerosedhemangioma, ICC, HCC

Yes

ICC

1/Segment VIII

NC, Art, PV 2.5-cm hypervascular mass with persistent venous phase enhancement

2.5-cm hypervascular mass with T2 signal hyperintensity and persistent brisk enhancement on delayed sequences

1. Fatty liver with No cirrhosis 2. Hypervascular mass - HCC vs ICC

Yes

ICC

HCV

1/Segment VII

NC, Art, PV 3-cm hypervascular mass with washout and capsule

2.8-cm mass with T2 1. Cirrhosis hyperintense rim; 2. HCC slightly hypoenhancing center and hypervascular rim which persists

No

Yes

Mixed HCC/CC

10

SSC

No identifi- NC, Art, PV No mass able lesion

No mass

1. Segmental intrahepatic biliary ductal dilation 2. No mass

N/A

N/A

Multiple microscopic foci of ICC

11

PSC

No identifi- NC, Art, PV No mass able lesion

No mass

1. PSC cirrhosis N/A 2. No mass

N/A

Periductal infiltrating CC of left lobe

12

HCV

1/Segment III

NC, Art, PV 1.1-cm hypervascular mass with capsule and no washout

2.2-cm hyperenhanc- 1. Cirrhosis ing mass with wash- 2. HCC out and capsule

Yes

Yes

Mixed HCC/CC

2/Segment IVb

1.5-cm hypervascular mass with capsule and no washout

2.4-cm hyperenhanc- 1. Cirrhosis ing mass with wash- 2. HCC out and capsule

Yes

Mixed HCC/CC

3/Segment VIII/VII

0.7-cm hypervascular 1.2-cm isointense 1. Cirrhosis hyperenhancing mass hypervascular mass 2. HCC with no capsule and no washout

Yes

HCC

HBV

1/Segment VIII

2/Segment V

CT features of lesion

OPTN class 5

Within Milan Pathology criteria

MRI features of lesion

13

CT technique

NC, Art, PV 4.1-cm hypervascular mass with washout

2.7-cm hypervascular HCC mass with washout

Yes

Yes, at Mixed HCC/CC time of MRI not CT

3.5-cm hypervascular mass with washout

2.2-cm hypervascular HCC mass with washout

Yes

Necrotic nodules with features consistent with prior embolization therapy

HCC: hepatocellular carcinoma; CC: cholangiocarcinoma; ICC: intrahepatic cholangiocarcinoma; HCV: hepatitis C virus; NASH: non-alcoholic steatohepatitis; ALD: alcoholic liver disease; PSC: primary sclerosing cholangitis; AIH: autoimmune hepatitis; SSC: secondary sclerosing cholangitis; HBV: hepatitis B virus; PV: portal venous phase; NC: non-contrast phase; Art: arterial phase; DDX: differential diagnosis; N/A: not-applicable.

Hepatobiliary Pancreat Dis Int，Vol 16，No 3 • June 15，2017 • www.hbpdint.com • 267


Fig. Kaplan-Meier curves showing the overall survival (A) and disease-free survival (B) in patients with HCC/CC and ICC.

respectively. Disease-free survival was 68%, 51% and 41% at 1, 3 and 5 years, respectively (Fig.). Seven (54%) patients died. Three patients died from tumor recurrence. One patient developed de novo gastric adenocarcinoma, one died from chronic rejection and sepsis, one died as a result of chronic kidney disease, and one died as a consequence of chronic neurological disease.

Treatment of tumor recurrence Overall tumor recurrence rate was 31% (4 patients). One patient (HCC/CC) had a recurrence 12.2 months after LT, deferred systemic chemotherapy and died 2.2 months later. One patient (ICC) developed pulmonary metastasis 2 months post-LT and died 18 months postLT after receiving systemic chemotherapy. One patient (HCC/CC) developed bone metastases 65 months postLT, and was undergoing chemo-radiation at the last date of follow-up. One patient (HCC/CC) developed pulmonary metastases 2.8 months post-LT and died 3 months later after receiving systemic chemotherapy.

Discussion About a third of patients undergoing LT carry the diagnosis of HCC; and in patients within Milan criteria (single lesion ≤5 cm, or up to 3 lesions ≤3 cm each) post-LT survival is equivalent to patients undergoing LT without HCC.[11] In contrast, as opposed to hilar cholangiocarcinoma for which LT is an accepted therapy in selected

patients, ICC has been considered a contraindication for LT due to poor outcomes.[12-14] There is however rather anecdotal experience demonstrating that small unifocal ICC or incidentally found ICC on explant histopathology can be associated with acceptable post-LT outcome.[6] Additionally, a small subgroup of patients with mixed HCC/CC underwent LT with presumed diagnosis of HCC.[3, 6, 13] Since mixed HCC/CC may have different biological behavior than HCC, and have demonstrated inferior outcomes post-LT compared to pure HCC patients, it is important to recognize these tumors pre-LT.[8] Several classifications for HCC/CC have been proposed. Allen and Lisa[15] described a type A variant with two distinctly separate lesions, each having histology of either HCC or CC, type B variant (collision lesions) with distinct sub-areas that develop into CC and HCC; and type C variant which are hybrid lesions where histological elements of HCC and CC are truly intermingled. The histopathology of combined hepatobiliary tumors is progressively being elucidated lying between classical HCC lesions and CC. These lesions often express cellular markers of both CC and HCC. Serum AFP and CA19-9 levels, typically associated with HCC and CC respectively can both be elevated. In line with this, these “mixed” lesions likely represent a spectrum between the two diseases.[16] Diagnosis of HCC relies mainly on characteristic imaging features and organ allocation hinges on these pathognomonic findings. Before the implementation of the new OPTN HCC radiologic criteria, the American Association for the Study of Liver Disease (AASLD) guidelines relied fundamentally on the presence of arterial phase hyper-vascularity or enhancement with subsequent portal venous phase or delayed phase “washout” on cross-sectional contrasted studies to presume diagnosis of HCC.[17] Washout is based on the premise that HCC contains mostly arterial blood and by the time of delayed imaging has relatively less observed compared to normal liver, thus appearing hypo-attenuating. In contrast, CC tend to demonstrate delayed enhancement. However, the imaging characteristics of HCC/CC are less well defined. In one recent study, sensitivity of both CT and MRI imaging for detecting HCC/CC was about 33% with lesions being mistaken for both HCC and CC; thus highlighting the diagnostic challenges associated with this diagnosis.[17] The current OPTN guidelines for the imaging diagnosis of HCC are as follows: for lesions 1-2 cm characteristic arterial enhancement with both portal/ venous phase contrast washout and pseudo-capsule enhancement are required while either washout or pseudocapsule enhancement suffice for lesions 2-5 cm. Alternately, if these conditions are not fulfilled, demonstration of >50% growth over 6 months allows for a diagnosis of HCC. These OPTN class 5 lesions qualify for MELD ex-



Table 3. Summary of published reports of LT for ICC and HCC/CC Studies Sapisochin et al[3] Groeschl et al[10] Panjala et al[25] Hu et al[26] Sapisochin et al[27] Takahashi et al[28]

Year published 2014 2013 2010 2011 2011 2016

Sample size 27 ICC and 15 HCC/CC 19 HCC/CC 12 HCC/CC 20 ICC 14 ICC and HCC/CC 13 ICC

Recurrence rate 21.4% 58% 60% (12/20) 57% (8/14) 53.8 % (7/13)

Survival 1-, 3- and 5- yr OS: 83%, 70%, 60% 3-yr OS: 48% 1-, 3- and 5- yr OS: 79%, 66%, 16% 1-, 3- and 5- yr OS: 84.2%, 32.7%, 21.8% 1-, 3- and 5- yr OS: 76%, 66%, 47% 1- and 3- yr DFS: 67%, 42%

OS: overall survival; DFS: disease-free survival.

ception points in current organ allocation schema. In our series with patients presenting with pre-transplant liver masses, after excluding one patient where pre-LT imaging could not be retrieved, three patients fulfilled none of the OPTN class 5 imaging criteria, two patients did not fully fulfill OPTN class 5 criteria and were reported to be indeterminate or atypical and only four fulfilled all criteria for OPTN class 5 HCC diagnosis. Thus, four lesions would have qualified for exception points based on updated HCC imaging guidelines.[17, 18] Of note, CT imaging protocols used at the time when several patients in our series were imaged did not incorporate delayed phase imaging, which is now mandated by OPTN and helps to appreciate wash-out and delayed enhancement used to differentiate HCC and CC. Some reports suggest that HCC/CC trend towards decreased survival compared with pure HCC.[10, 19] Recent studies[20, 21] have corroborated these findings with survival curves for mixed tumors falling between pure HCC and ICC, showing relatively poorer and better survival, respectively. While LT along with neoadjuvant therapy in selected patients with hilar cholangiocarcinoma has become promising, the applicability of LT for non-resectable small unifocal ICC or HCC/CC remains to be fully elucidated.[22-24] One of the largest prospective series to date describes 42 patients with 27 ICC and 15 HCC/CC. Recurrence rates for HCC/CC were similar to HCC controls (1-, 3- and 5-year recurrence rates of 7%, 7% and 7%, respectively vs 0%, 4% and 4% for controls; P=0.6), as was actuarial survival (HCC/CC 1-, 3- and 5-year overall survival 93%, 78% and 78% vs 97%, 86% and 86% for the HCC controls; P=0.9).[3] ICC tended to have higher recurrence rates than HCC controls (12%, 25% and 36% vs 0%, 2% and 2%, respectively; P