The prognostic relationship between histopathological and immunohistochemical features of hepatocellular carcinoma, intrahepatic cholangiocarcinoma and mixed type

Introduction

Hepatocellular carcinoma (HCC) is the fifth most common cancer and the third leading cause of cancer-related death worldwide [1]. Intrahepatic cholangiocarcinoma (ICC) is the second most common cause of primary liver malignancies after HCC and constitutes approximately 10-15% of all primary liver cancers [2, 3, 4]. Combined hepatocellular-cholangiocarcinoma (cHCC-ICC) is a rare tumour, with variations reported between 1.0 and 4.7% of all primary hepatic tumours in a series of patients undergoing hepatic resection [5]. HCC is derived from hepatocytes, ICC from intrahepatic bile duct, and cHCC-ICC both from hepatocytes and biliary duct epithelial cells [6]. The most common aetiological factor for HCC is hepatitis B or hepatitis C or both infections [1]. Causes for liver damage, such as alcoholic liver disease, HCV, HDV, HIV, diabetes, nonspecific cirrhosis, and parasitic infections, are considered common risk factors in the development of ICC [7]. The risk factors for cHCC-ICC are similar to those for pure HCC [8]. Although the pathogenesis is uncertain, there is evidence that progenitor or stem cells play a role in the cause of these lesions [2, 8]. Trabecular pattern is the most common in well and moderately differentiated HCCs [2]. HCCs could show a gland-like pattern (pseudo acinar), usually admixed with the trabecular pattern [2]. The most common pattern in poorly differentiated HCCs is solid or compact pattern [2]. In classical morphology, ICCs appear in small or wide glandular patterns, with very few micropapillaries, HCC-like trabecular pattern, and all mixed-mixed types. Most ICCs have a tubular pattern of growth with variable size lumina, although micropapillary, acinar or cord like patterns also occur [2, 4, 9]. The most typical form of classic cHCC-ICC demonstrates morphological and histochemical evidence of a mixture of hepatocellular and ductular (or glandular) elements throughout the tumour [2, 8]. Also, a new type of cHCC-ICC, showing stem cell properties, has been described [2]. There is no need for immunohistochemical markers in the diagnosis of classical HCC. However, in the presence of different patterns, it may sometimes be necessary to distinguish between different types. Confirmation of hepatocellular differentiation is easily provided by immunohistochemical staining with Hep Par (granular cytoplasmic staining) [2]. ICC is negative with Hep Par. CK 19 can be positive both in HCC and ICC [2, 8]. CK7 is mainly positive in ICC [8]. Combined HCC-ICC with stem cells has features of mature-appearing hepatocytes with peripheral clusters of small cells positive for stem cell markers such as CK7, CK19, and CD133, epithelial cell adhesion molecule (EpCAM), and CD56/ neural cell adhesion molecule (NCAM) [2, 7, 8, 10, 11]. A classical combined type is CD56 negative. In the literature, CD56, CK19, and p53 are found to be significant in terms of prognostic parameters [12, 13]. In all three tumours, the histological grade, tumour diameter, lymphovascular invasion, perineural invasion, portal venous or hepatic venous invasion, regional lymph node metastasis, tumour grade, and tumour size should be indicated as prognostic parameters in the report, according to the 8th edition of the TNM, AJCC Staging Manual classification [14]. In terms of prognosis, classical HCC is generally very poor, particularly for cases in which partial or complete portal vein thrombosis and ICC are associated with a high rate of fatality because of early invasion and widespread metastasis. The prognosis for patients with combined HCC-ICC without stem cell features is thought to be worse than for pure HCC [2]. The prognosis of combined HCC-ICC with stem cell features is unknown [2]. In this study, the contribution of immunohistochemical markers and histological patterns in differential diagnosis of HCC, ICC, and cHCC-ICC types and the relationship between clinicopathologic features and prognostic parameters in all three tumour types, were investigated.

Material and methods

A total of 61 cases, with clinical follow-up of 50 cases with HCC diagnosis, 9 cases with ICC diagnosis, and 2 cases with cHCC-ICC diagnosis, between 2009 and 2013 were included in the study. All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008. The clinicopathologic features of these cases were determined. Haematoxylin-eosin preparations were re-evaluated by two pathologists in terms of histopathologic features. Then, formalin-fixed paraffin-embedded tissue samples, cut at 3 µm thickness, were deparaffinized with xylene and rehydrated with graded ethanol. Antigen retrieval was performed by boiling at 98°C for 40 minutes in 0.01 mol/l sodium citrate buffer (pH 6.0), or in Tris/EDTA buffer (pH 9.0). Immunohistochemistry was performed using a Leica Bond Max Automated IHC/ISH Stainer. Primary antibody CK7 (Leica Novocastra, OV-TL, Mouse monoclonal; 1/50), CK19 (Leica Novocastra, b170, Mouse monoclonal; 1/50), CD56 (Leica Novocastra, CD56), Mouse monoclonal; 1/100), Hep par-1 (Leica Novocastra, 0CH1E5, Mouse monoclonal; 1/50), and ve P53 (Leica Novocastra, DO-7, Mouse monoclonal; 1/100). As controls, gastric mucosa for CK7, skin for CK19, colon mucosa for CD56, liver tissue for Hep par-1, and hepatocellular carcinoma for p53 were used. The cases were evaluated as negative, with focal staining ≤ 50% and strong staining > 50%, according to staining patterns. Changes in the diagnosis with IHC were evaluated. The differences in prognostic parameters between HCC and other types, after diagnosis with IHC, were evaluated. For evaluation of prognostic parameters, the tumours were divided into two groups, HCC and ICC. Because of the low number of cHCC-ICC, they were excluded from further analysis. The significant difference between overall survival (OS) and disease free survival (DFS) for two types of tumours were also evaluated. The effect of prognostic parameters on OS and DFS were evaluated independently of tumour type. The SPSS 21 (Armonk, New York, USA) software was used for statistical calculations. OS and DFS were compared using the Kaplan-Meier method, and the survival differences between the two groups were compared using the log-rank test. The distribution was analysed with the help of the Kolmagorov-Smirnov test. The categorical variables were assessed using the χ2 test. The quantitative variables were analysed by means of the independent simple T test. The OS and DFS rates among the patients were calculated by the Kaplan-Meier method and compared with the log rank test. The statistical significance was indicated by p values less than 0.05.

Results

The diagnosis changed in four cases after immunohistochemical (IHC) examinations. After IHC staining, the diagnosis was changed from HCC to ICC in one case and to cHCC-ICC in three cases. In 46 cases (75.4%), the HCC diagnosis remained unchanged. The histological features were trabecular in 36 cases (78%) and mixed (trabecular, tubular/pseudoacinar, solid) in 10 cases (22%) among the cases with unchanged diagnosis of HCC (Fig. 1). The staining characteristics of HCC cases were all positive with Hep par-1, all negative with CK19 and CD56, and focal positively with CK7 and p53 in a few cases (Fig. 2). A mixed histological pattern was observed where the diagnosis changed to ICC. The histologic features of the 10 cases with ICC were 5 small glandular (50%), 3 large glandular (30%), and 2 mixed (large glandular, trabecular) type (20%). The staining properties of ICC were all negative with Hep par-1, all positive with CK7 and CK19, and positive with CD56 and p53 in some cases. A total of five cases, with three cases diagnosed after IHC, and two cases previously diagnosed with and before IHC, were accepted as cHCC-ICC. All cHCC-ICC cases were in mixed pattern histologically. All of the cases with Hep par-1, CK7, and CK 19 were positive; however, positive staining with p53 were seen in some of them. When the CD56 staining characteristics were examined, it was determined that staining was negative in two cases of classical cHCC-ICC and focal positive in 3 cases of cHCC-ICC with stem cell properties; however, due to their small number, they were collected within a group of cHCC-ICC. The cases with changed diagnosis and staining characteristics of all tumours are shown in Tables I and II, reespectively. In terms of their clinicopathological features, the analysis was conducted by dividing them into three groups: HCC, ICC and cHCC-ICC types, at the final diagnosis. All HCC cases were hepatectomized for the purpose of transplantation due to cirrhosis. Four cases of ICC and mixed types were developed due to cirrhosis. While most of the ICCs underwent partial hepatectomy, most of the mixed types underwent total hepatectomy. Other clinicopathological and histopathological features are summarised in Tables III and IV. The OS of patients was 35.3 months (0 to 87, ±27.176) and DFS was 33 months (0 to 87, ±25.880). Patients who died and recurred during this period are identified in Table IV. HCC and ICC types were evaluated for statistical significance in recurrence, death, grade, angiolymphatic invasion (ALI), perineural invasion (PNI), lymph node metastasis, necrosis, grade, tumour diameter, and OS and DFS values. The grade and stage were significantly higher in HCC patients (p = 0.000 and p = 0.003, respectively). Lymph node metastasis and perineural invasion was common in tumour types other than HCC (p = 0.032 and p = 0.000, respectively). There were no significant differences between tumour types in terms of diameter (p = 0.064). The OS and DFS values were significantly higher in HCC than in others (p = 0.008 and p = 0.011, respectively; Table V). However, no significant differences were found in the Kaplan-Meier life analysis between the tumour types and OS and DFS values (0.705 and 0.200 respectively; Figs. 3 and 4). In the Kaplan-Meier survival analysis of OS and DFS with independent prognostic parameters, such as angiolymphatic invasion independent of tumour type, PNI, hepatoportal venous invasion (HPVI), lymph node metastasis, necrosis, tumour diameter, tumour foci, stage and grade, no significant results could be obtained. Only the relationship between angiolymphatic invasion with OS trended towards significance (p = 0.052). P53 and CD 56 staining rates were evaluated in terms of prognostic parameters. In all tumour types, no significant relationships were found between CD 56 and death (p = 0.433), recurrence (p = 0.314), ALI (p = 0.576), PNI (p = 0.703), necrosis (p = 0.637), HPVI (p = 0.791), and lymph node metastasis (p = 0.791). While analysing the relation between p53 and prognostic parameters, only necrosis (p = 0.027) and lymph node metastasis (p = 0.001) were significant. No relationship was found between p53 and death (p = 0.955), recurrence (p = 0.340), ALI (p = 0.381), PNI (p = 0.550), and HPVI (p = 0.550). Also, no significant relationship was observed in the Kaplan-Meier life analysis between p53 and OS (p = 0.419) and DFS (p = 0.348).

Discussion

The three most common malign tumour types in the liver are HCC, ICC, and cHCC-ICC. Because the cHCC-ICC type is less common, the prognosis is not clear. ICC is known to have a poor prognosis with early invasion and metastasis. It has been reported that ICC and cHCC-ICC have worse prognosis than HCC [8, 15, 16, 17, 18]. In our study, HCC and ICC types were compared in terms of prognostic parameters. Statistically significant results were obtained between grade, lymph node metastasis, focality, perineural invasion, stage, overall survival, and disease-free survival. In our study, lymph node metastasis in HCC-diagnosed cases was as low as mentioned in the literature [15, 19]. The presence of lymph node metastasis and perineural invasion suggests poor prognosis in the non-HCC group [10]. Grade, tumour foci, and stage were higher in HCC than for other tumour types. In accordance with the literature, in our study, most of the cases with HCC and cHCC-ICC were grade 2 [3, 15, 17]. Unlike other studies in the literature, multifocality of HCC type was found more frequently in our study [3]. In the literature, HCC is generally reported at early stages (stage I/II), whereas in our study the patients were at more advanced stages (stage III-IV) [3, 10]. Patients in cHCC-ICC were diagnosed mainly at stage II instead of stage I [17]. Important parameters affecting stage in HCC were tumour diameter and multifocality. Patients with one HCC of 5 cm or smaller, or up to three nodules of 3 cm or smaller, without vascular invasion or extrahepatic spread, had a 4-year survival rate of 75% with a recurrence rate of below 15% [20]. Multifocality is associated with high recurrence and impaired survival [20]. The mean HCC diameter was higher in our study compared to the other tumour types. In our study, the average tumour diameter in HCC was found to be 7.7 cm. In the literature, this size was between 3.1 and 6.4 cm on average [6, 15, 19, 21]. In ICC and cHCC-ICC types, our average tumour diameter was smaller in magnitude than the tumour diameters in the literature [18, 21]; the stage was lower in these tumours. We found no significant correlation between diameter and tumour type. OS and DFS were statistically significant in two tumour groups. In our study, OS in HCC was 40.7 months, OS in cHCC-ICC was 20 months, and OS in ICC was 17.6 months. These findings support a higher OS in HCC than ICC [6, 15, 16, 17, 18]. OS was longer in HCC patients, but not significant in the Kaplan-Meier life analysis. This may be because HCC cases were at more advanced stages than the cases in other studies in the literature. When we examined the clinicopathological parameters, the male sex was more pronounced in all tumour types, which is consistent with the incidence in Asian countries in terms of mean age and sex [2, 6, 19]. HBV was detected most frequently in HCC as well, which is consistent with the fact that HBV is the most common cause in developing and Asian countries [8, 19]. When we evaluated the histological pattern with IHC, Hep par-1, which stains hepatocyte-derived cells, was positive in the classic trabecular and mixed pattern in HCC [22]. With CK19 and CK7 staining, which are for biliary epithelial cells, only two cases of focal positivity with CK7 was observed. These two cases demonstrated a mixed pattern. In typical trabecular pattern, CK7 and CK19 was negative. However, there are also publications in the literature that find CK19 positive in HCC [21]. In our study, when the IHC was re-evaluated in the previously diagnosed tumours, the diagnosis was changed only in four patients, who underwent HCC diagnosis and showed were more mixed type areas. The IHC study should be conducted as required after pattern analysis. In particular, tumours of mixed pattern should receive support from IHC. Studies have shown that CK 19 positivity is associated with a poor prognosis in HCC [2, 23]. There are studies that associate CK19 positivity with a grade of HCC, but they did not show a significant relationship [20, 23]. All HCC cases were negatively stained with CK19. Because CD56 was positive in stem cell, it can be used for diagnosis of stem cells, subtype of cHCC-ICC [5, 10]. In our study, prognostic parameters were also compared with CD56, but no significant results were obtained. This could be due to the lower number of cases both in our study and in the literature [13, 21]. The literature suggested that p53 is a prognostic marker between stage, metastasis, and OS [24]. In our study, we obtained significant results only with tumour necrosis and lymph node metastasis with p53. The vascular pattern is important in the prognosis of HCC [25, 26]. In particular, sinusoidal neovascularisation pattern has worse prognosis than capillary pattern in terms of DFS and OS [25]. For evaluation of vascular pattern CD34 and CD31 could be used [25, 26]. Also, Glypican-3 (GPC-3) which plays an important role in regulating cells growth, could also be used for prognosis [27, 28]. It is postulated that GPC-3 is an important molecule for differentiation of dysplastic nodules from HCC and determination of poor prognosis but more studies are needed to confirm these results [27, 28]. In our study, although the sample size was small, the most common types of liver primary carcinomas were examined in a wide spectrum of clinicopathological and prognostic parameters. When HCC and other tumour types were compared in terms of prognostic parameters, perineural invasion and lymph node metastasis were higher in the non-HCC group, whereas the stage, grade, and multifocality were higher in the HCC group. There was no significant difference in tumour types between DSF and OS. The HCC patients in our study were considered to have worse OS than expected because they were in an advanced stage compared to many studies in the literature. When tumour types were diagnosed, we suggested that IHC should be studied especially in morphologically mixed cases. As a prognostic marker, p53 was more significant than CK 19 and CD 56.
The authors declare no conflict of interest.

References

1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer J Clin 2011; 61: 69-90.
2. Bosman FT, Carneiro F, Hruban RH, et al. WHO Classification of Tumours of the Digestive System. 4th ed. IARC Press, Geneva 2010.
3. Nakanuma Y, Sato Y, Harada K, et al. Pathological classification of intrahepatic cholangiocarcinoma based on a new concept. World J Hepatol 2010; 2: 419-427.
4. Nakanuma Y, Sasaki M, Ikeda H, et al. Pathology of peripheral intrahepatic cholangiocarcinoma with reference to tumorigenesis. Hepatol Res 2008; 38: 325-334.
5. Gera S, Ettel M, Acosta-Gonzalez G, Xu R. Clinical features, histology, and histogenesis of combined hepatocellular-cholangiocarcinoma. World J Hepatol 2017; 9: 300-309.
6. Yin X, Zhang BH, Qiu SJ, et al. Combined hepatocellular carcinoma and cholangiocarcinoma: clinical features, treatment modalities, and prognosis. Ann Surg Oncol 2012; 19: 2869-2876.
7. Shaib YH, El-Serag HB, Davila JA, et al. Risk factors of intrahepatic cholangiocarcinoma in the United States: a case-control study. Gastroenterology 2005; 128: 620-626.
8. Odze RD, Goldblum JR (eds.). Surgical Pathology of the GI tract, Liver, Biliary tract and pancreas. In: Ferrell LD (ed.). Benign and malignant tumors of the liver. 3rd ed. Elsevier, Philadelphia 2015; 1540-1562.
9. Aishima S, Kurado Y, Nishihara Y, et al. Proposal of progression model for intrahepatic cholangiocarcinoma: clinicopathologic differences between hilar type and peripheral type. Am J Surg Pathol 2007; 31: 1059-1067.
10. Akiba J, Nakashima O, Hattori S, et al. Clinicopathologic analysis of combined hepatocellular-cholangiocarcinoma according to the latest WHO classification. Am J Surg Pathol 2013; 37: 496-505.
11. Chan AW, Tong JH, Chan SL, et al. Expression of stemness markers (CD133 and EpCAM) in prognostication of hepatocellular carcinoma. Histopathology 2014; 64: 935-950.
12. Sempoux C, Fan C, Singh P, et al. Cholangiolocellular carcinoma: an innocent-looking malignant liver tumor mimicking ductular reaction. Semin Liver Dis 2011; 31: 104-110.
13. Sung JJ, Noh SJ, Bae JS, et al. Immunohistochemical Expression and Clinical Significance of Suggested Stem Cell Markers in Hepatocellular Carcinoma. J Pathol Transl Med 2016; 50: 52-57.
14. Gospodarowicz MK, Wittekind C (eds.). TNM classification of malignant tumours. 8th ed. In: Brierley JD, Mary K. Digestive system tumours. 2016; 55-105.
15. Shibahara J, Hayashi A, Misumi K, et al. Clinicopathologic Characteristics of Hepatocellular Carcinoma With Reactive Ductule-like Components, a Subset of Liver Cancer Currently Classified as Combined Hepatocellular-Cholangiocarcinoma With Stem-Cell Features, Typical Subtype. Am J Surg Pathol 2016; 40: 608-616.
16. Koh KC, Lee H, Choi MS, et al. Clinicopathologic features and prognosis of combined hepatocellular cholangiocarcinoma. Am J Surg 2005; 189: 120-125.
17. Chang CC, Chen YJ, Huang TH, et al. Living Donor Liver Transplantation for Combined Hepatocellular Carcinoma and Cholangiocarcinoma: Experience of a Single Center. Ann Transplant 2017; 22: 115-120.
18. Ariizumi S, Kotera Y, Katagiri S, et al. Long-term survival of patients with cholangiolocellular carcinoma after curative hepatectomy. Ann Surg Oncol 2014; 21 Suppl 3: S451-S458.
19. Tao CY, Liu WR, Jin L, et al. Surgical Treatment of Combined Hepatocellular-Cholangiocarcinoma is as Effective in Elderly Patients as it is in Younger Patients: A Propensity Score Matching Analysis. J Cancer 2018; 9: 1106-1112.
20. Forner A, Llovet JM, Bruix J. Hepatocellular carcinoma. Lancet 2012; 379: 1245-1255.
21. Kumagai A, Kondo F, Sano K, et al. Immunohistochemical study of hepatocyte, cholangiocyte and stem cell markers of hepatocellular carcinoma: the second report: relationship with tumor size and cell differentiation. J Hepatobiliary Pancreat Sci 2016; 23: 414-421.
22. Jia SQ, Ren JJ, Dong PD, Meng XK. Probing the hepatic progenitor cell in human hepatocellular carcinoma. Gastroenterol Res Pract 2013; 2013: 145253.
23. Sempoux C, Paradis V, Saxena R. Variant differentiation patterns in primary liver carcinoma. Semin Diagn Pathol 2017; 34: 176-182.
24. Li Z, Han C, Feng J. Relationship of the expression levels of XIAP and p53 genes in hepatocellular carcinoma and the prognosis of patients. Oncol Lett 2017; 14: 4037-4042.
25. Park JO, Yeh MM. Clinical significance and implication of neoangiogenesis in hepatocellular carcinoma. J Gastroenterol Hepatol 2011; 26: 792-793.
26. Fodor D, Jung I, Turdean S, et al. Angiogenesis of hepatocellular carcinoma: An immunohistochemistry study. World J Hepatol 2019; 11: 294-304.
27. Wang L, Yao M, Pan LH, et al. Glypican-3 is a biomarker and a therapeutic target of hepatocellular carcinoma. Hepatobiliary Pancreat Dis Int 2015; 14: 361-366.
28. Moudi B, Heidari Z, Mahmoudzadeh-Sagheb H. Meta-analysis and systematic review of prognostic significance of Glypican-3 in patients with hepatitis B-related hepatocellular carcinoma. Virusdisease 2019; 30: 193-200.