Reduced HBV cccDNA and HBsAg in HBV-associated hepatocellular carcinoma tissues

Approximately 50% of hepatocellular carcinoma (HCC) is attributable to chronic infection with hepatitis B virus (HBV). Serum hepatitis B surface antigen (HBsAg) is an important diagnostic marker of HBV infection, whereas intrahepatic HBV covalently closed circular DNA (cccDNA) is a surrogate marker of HBV persistence. This study aimed to investigate relationships between serum HBsAg, intrahepatic HBsAg, and intrahepatic cccDNA in HBV-associated HCC. Intrahepatic HBsAg was determined by immunohistochemistry in matched non-cancerous and HCC tissues from 88 patients; 56 patients (63.64%) were serum HBsAg positive. In serum HBsAg-positive group, intrahepatic HBsAg was positive staining in 73.2% of non-cancerous tissues, but only in 10.7% of HCC tissues. Significant correlation between serum HBsAg and intrahepatic HBsAg was observed in non-cancerous tissues (p < 0.001), but not in HCC tissues (p = 0.415). Absolute quantification of intrahepatic cccDNA was performed by droplet digital PCR in tissues from 30 patients; 18 patients (60%) were serum HBsAg positive. In serum HBsAg-positive group, intrahepatic cccDNA was detected in 66.66% of non-cancerous tissues, but only in 5.55% of HCC tissue; intrahepatic cccDNA levels in non-cancerous tissues were significantly higher than those in HCC tissues (p < 0.001), and correlated with serum HBsAg (p < 0.01). Significant correlations between intrahepatic HBsAg and intrahepatic cccDNA were found in both non-cancerous tissues (p < 0.01) and HCC tissues (p < 0.05). We concluded that HBV cccDNA and intrahepatic HBsAg in HBV-associated HCC tissues were significantly reduced, as compared with matched non-cancerous tissues. This warrants further investigation into the impacts and the cause(s) of cccDNA reduction in HBV-associated HCC tissues, which might yield novel immune-related therapy for HBV-associated HCC.


Introduction
Hepatocellular carcinoma (HCC) is the fifth most common cancer in men and the ninth in women, as well as the second leading cause of cancer-related death globally [1,2]. Almost 50% of all cases of HCC are associated with chronic infection with hepatitis B virus (HBV) [3,4]. HBV belongs to a family of viruses known as Hepadnaviridae and encodes only four genes in a highly compact viral genome: the surface gene (S), the core gene (C), the X gene (X), and the polymerase gene (P) [5]. Hepatitis B surface antigen (HBsAg) is an important diagnostic marker of hepatitis , which has been proposed as a surrogate marker of HBV-infected hepatocytes [6,7]. Several studies have investigated the HBV cccDNA level in HCC and matched non-cancerous tissues but the results are still inconclusive. Wong et al. [8] reported that in HBsAg-positive patients, HBV cccDNA level of tumor tissues was significantly higher than the level of non-tumor tissues. In contrast, two studies reported that HBV cccDNA level of cancer tissues was significantly lower than the level of non-cancerous tissues [9,10]. In addition, Fu et al. [11] and Bai et al. [12] reported that there was no significant difference in intrahepatic HBV cccDNA levels between tumor and non-tumor liver tissues. Correlation between serum HBsAg levels and HBV cccDNA in tumor or nonneoplastic liver tissues of HBV-associated HCC patients was reported by Wang et al. [13]. In contrast, Wang et al. reported that serum HBsAg correlated poorly with intrahepatic HBV cccDNA in the tumor and non-tumor liver tissues [14]. Therefore, in this study, we investigated both HBsAg and HBV cccDNA in HCC and non-cancerous tissues so as to explore (1) the correlation between the serum HBsAg and the intrahepatic HBsAg; (2) the correlation between the serum HBsAg and the intrahepatic HBV cccDNA; and (3) the correlation between the intrahepatic HBsAg and the intrahepatic HBV cccDNA. Since HBV cccDNA is responsible for viral persistence, droplet digital PCR (ddPCR) technique, a highly sensitive PCR method, was applied for HBV cccDNA quantification in this study [15].  [15]. For each ddPCR reaction mixture, 70 µl droplet generation oil and 20 µl were added to the DG8 cartridge, then the droplets were produced by a droplet generator of the QX200 Droplet Digital PCR system (Bio-Rad). The droplets containing the PCR reaction mixture and droplet generation oil were then transferred to a 96-well PCR plate for amplification using the C1000 Touch Thermal Cycler (Bio-Rad) applying a thermalcycle beginning at 94 °C for 10 min, followed by 35 cycles of 94 °C for 30 s, 57 °C for 60 s, and a final cycle of 98 °C for 10 min. Subsequently, a droplet reader (Bio-Rad) was used to calculate the number of both positive and negative droplet events from each PCR reaction mixture. A PCR reaction mixture with no DNA template was used as a reference control for potential PCR contamination. Triplicate reactions were run for each sample. The ddPCR data were analyzed using the QuantaSoft analysis software (Bio-Rad), which calculates the total number of droplets (positive droplets plus negative droplets).

Serum hepatitis B surface antigen (HBsAg) assay
Chemiluminescent microparticle immunoassays (CMIA) for the qualitative detection of hepatitis B surface antigen (HBsAg) in serum from the patients were performed using ARCHITECT HBsAg Qualitative II assay (Abbot Laboratories, Illinois, USA).

Serum alpha-fetoprotein (AFP) assay
Electrochemiluminescence immunoassays (ECLIA) for the in vitro quantitative determination of alpha-fetoprotein (AFP) in serum from the patients were performed using the AFP kit with a cobas e601 analyzer (Roche Diagnostics Limited GmbH, Mannheim, Germany).
The p values less than 0.05 were considered statistically significant.

Correlation between serum HBsAg and intrahepatic HBsAg
Intrahepatic HBsAg was determined in 88 pairs of HCC and matched non-cancerous tissues by IHC. The data of serum HBsAg were collected from patient records. The clinicopathological features are illustrated in Table 1. Of 88 patients included in this study, serum HBsAg was determined as positive in 56 cases (63.64%) and negative in 32 cases (36.36%). In serum HBsAg-positive group, intrahepatic HBsAg was positive staining in 73.2% (41/56) of matched non-cancerous tissues but only in 10.7% (6/56) of HCC tissues, as shown in Fig. 1a. In serum HBsAg-negative group, intrahepatic HBsAg was positive staining in 18.7% (6/32) of matched non-cancerous tissues and only in 3.1% (1/32) of HCC tissues, as shown in Fig. 1b. The significant correlation between serum HBsAg and intrahepatic HBsAg in matched non-cancerous tissues was observed (p < 0.001) but there was no correlation between serum HBsAg and intrahepatic HBsAg in HCC tissues (p = 0.415), as shown in Table 2.
In serum HBsAg-negative group (12 samples), intrahepatic cccDNA was detected in 16.66% (2/12) of matched noncancerous tissues (12 and 17 copies/ng DNA) but not in all 12 HCC tissues, as shown in Table 4 and Fig. 2a. In serum HBsAg-positive group, cccDNA levels in matched noncancerous tissues were significantly higher than in HCC tissues, p = 0.0005 (Fig. 2b). In addition, there was a significant correlation between serum HBsAg and cccDNA in matched non-cancerous tissues (p < 0.01).

Correlation between intrahepatic HBsAg and intrahepatic HBV cccDNA
Both intrahepatic HBsAg and intrahepatic cccDNA were evaluated in 30 pairs of matched non-cancerous and HCC tissues. The clinicopathological features are illustrated in Table 3. In 15 matched non-cancerous tissues with positive staining for intrahepatic HBsAg, intrahepatic cccDNA was detected in 73.3% (11/15); in 15 matched non-cancerous tissues with negative staining for intrahepatic HBsAg, intrahepatic cccDNA was detected only in 20% (3/15), as shown in Table 5a. Interestingly, in HCC tissues, both intrahepatic HBsAg and intrahepatic cccDNA were detected only in the  Table 5b and Fig. 3. Representative of intrahepatic HBsAg and cccDNA is shown in Fig. 4. Significant correlations between intrahepatic HBsAg and cccDNA were found in both matched non-cancerous (p < 0.01) and HCC tissues (p < 0.05).

Discussion
Serum HBsAg is an important diagnostic marker for HBV infection. However, some studies suggest that HBV DNA can be detected in the liver tissue of HBsAg seroclearance or patients with occult HBV, mainly in the form of cccDNA, and these patients can continue to develop HCC [16,17].
In the present study, we investigated intrahepatic HBsAg in both matched non-cancerous and HCC tissues by IHC, and found that serum HBsAg was significantly associated with intrahepatic HBsAg only in matched non-cancerous tissues. In contrast to matched non-cancerous tissues, no association between serum HBsAg and intrahepatic HBsAg in HCC tissues was observed. In serum HBsAg-positive group, positive intrahepatic HBsAg was detected in matched noncancerous and HCC tissues at the frequency of 73.2 and 10.7%, respectively. Our results showed that intrahepatic HBsAg was reduced in most HCC tissues, which is consistent with the finding of Tian et al. [18], Jing et al. [19] and Wang [20], but the exact mechanism is still unclear. The finding that intrahepatic HBsAg could not be detected in HCC tissues by IHC from patients with positive serum HBsAg might be explained as follows, (1) HBV in HCC tissues could not express HBsAg, (2) HBsAg in HCC tissues could not be detected by the antibody used in this study, (3) HBV DNA is no longer present in HCC tissues. Therefore, in this study, so as to explore whether the decreased expression of HBsAg in HCC tissues is caused by HBV cccDNA reduction, intrahepatic HBV cccDNA was quantitated by ddPCR in both negative intrahepatic HBsAg and positive intrahepatic HBsAg tissues.
In the present study, the significant correlation between serum HBsAg and intrahepatic cccDNA was found only in matched non-cancerous tissues. This is different from   previous studies; as Fu et al. reported that there was no correlation between serum HBsAg and intrahepatic cccDNA in HBV-related hepatocellular carcinoma [11], but Wang et al. [13] reported correlation between serum HBsAg and intrahepatic cccDNA in both tumor and non-neoplastic liver tissues of HBV-associated HCC patients. The reasons for variety of results are inconclusive. But the possible contributing factor might be the size of tumors in tissue samples, in which 60% of specimen in this study was larger than 5 cm. Recently, Wang et al. [10] reported that cccDNA copy number per tumor in early tumors (smaller than median, ≤ 3 cm) was higher than those in larger tumors (> 3 cm); furthermore, in patients with small tumors, the cccDNA copy number did not differ between tumor and paired non-neoplastic liver, whereas in patients with larger tumors, the cccDNA copy in tumor was significantly lower than that of the paired nonneoplastic liver. Another striking finding of the present study was the significant correlation between intrahepatic HBsAg and intrahepatic cccDNA in both matched non-cancerous and HCC tissues. Furthermore, a significant reduction of intrahepatic cccDNA in HCC tissues was observed, which is consistent with two previous studies [6,7] but not consistent with some other studies [8,11,12]. The discordance of the research results might be due to the tissue sampling. Of 15 tissues with negative intrahepatic HBsAg, intrahepatic cccDNA could be detected in 3 (20%) tissues. On the other hand, of 15 tissues with positive intrahepatic HBsAg, intrahepatic cccDNA could not be detected in 4 (26.7%) tissues. These discordant findings might be due to heterogeneity of HBV infection in tissues, and tissue sampling for IHC and ddPCR might not be from the same area. Our finding demonstrated the followings: (1) In matched non-cancerous tissues, serum HBsAg significantly correlated with both intrahepatic HBsAg and cccDNA; (2) In HCC tissues, serum HBsAg did not correlate with both intrahepatic HBsAg and cccDNA; (3) Intrahepatic HBsAg significantly correlated with intrahepatic cccDNA in both matched non-cancerous and HCC tissues; (4) The intrahepatic cccDNA level in HCC tissues was significantly lower than that in matched non-cancerous tissues. We concluded that intrahepatic HBsAg reduction in HCC tissues was correlated with HBV cccDNA reduction. The reduction of cccDNA may be due to factors controlling proliferation of cancerous cells and/or factors inducing eradication of

Compliance with Ethical Standards
Conflict of interest The authors declare that they have no conflict of interest.
Ethical approval This study was approved by the Ethics Committee on Research involving Human Subjects of the Faculty of Medicine, Ramathibodi Hospital, Mahidol University.
Informed consent As this retrospective study was observational and link anonymized, no patient informed consent was required.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat iveco mmons .org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.