Evaluation of diagnostic accuracy of serum calcium channel α2δ1 subunit in hepatocellular carcinoma-related cirrhosis

Hepatocellular carcinoma (HCC) is one of the commonest malignancies worldwide that carries a bad prognosis particularly in Egypt due to the high prevalence of HCV burden. Late diagnosis of HCC especially in cirrhosis suffering-liver is one of the causes that worsen HCC outcome. Identification of molecular pathways of HCC will open the gate for early diagnosis and effective management. Oscillation of calcium controlled by the α2δ1 subunit has been proposed as one of the mechanisms in tumor-initiating cell properties of HCC. In this study, we aim to evaluate the serum α2δ1 subunit level as a biological marker for HCC. A total of 90 participants were enrolled, 40 patients with HCC, 40 patients with cirrhosis, and 10 healthy volunteers; serum level of α2δ1 was assessed in all participants with ELISA The mean serum levels of α2δ1 were significantly higher in HCC group (19.53 ± 6.87 ng/dL) than cirrhotic (6.24 ± 2.64 ng/dL) and control groups (0.67 ± 0.48 ng/dL) (P = 0.001). There was no significance between α2δ1 and etiology of liver disease as viral (HCV, HBV) or non-viral (P = 0.14). α2δ1 subunit may serve as a potential non-invasive marker with excellent sensitivity for diagnosis of HCC regardless of the etiology of liver disease.


Background
Liver malignancy is the sixth commonest malignancy worldwide with an incidence of approximately 841,000 new cases in 2018 [1], of which hepatocellular carcinoma (HCC) represents 80% [2], and considered the fourth leading cause of cancer-related death worldwide [3]. In Egypt, HCC is the fourth common cancer [4].
HCC has molecular subgroups reflecting the biological background and might have prognostic potential and selection criteria for therapy; among the subgroups there are two major subtypes each encompassing almost 50% of patients: one, proliferation class, which is poorly differentiated, had worse clinical outcome, expressing high alpha-fetoprotein (AFP) levels with more vascular invasion, and commonly associated with HBV infection; the other class, non-proliferation, more commonly associated with alcohol-related HCC or HCV infection, had better outcome with moderate to well differentiation [5].
In 1994, Lapidot and his team introduced the concept of cancer stem cell (CSC) through transplantation of acute myeloid leukemia (AML) cells from recently diagnosed patients into severe combined immune deficient (SCID) mice [6]. This CSC can self-renew, differentiate into all tumor cell lines giving the criteria of tumor hierarchy and intra-tumoral heterogeneity, have the unique tumorinitiating capability, and own high resistance to chemotherapy [7]. They possess many names as CSCs, tumorinitiating cells (TICs), or cancer stem-like cells [8].
Many CSCs have been identified in HCC, one of them was isoform 5 of the cell surface voltage-gated calcium channel α2δ1 described by Zhao et al. [9], they inject α2δ1+ cells subcutaneously in (non-obese diabetic) NOD/SCID mice, there was higher tumorigenic potential in comparison with α2δ1− cells. Also, they showed that α2δ1 regulates calcium signaling and intracellular calcium levels, which is vital for activating signaling cascades that regulate gene transcription and various cell functions; one of them is the phosphorylation of ERK1/2 which prevents cell apoptosis [9].
Calcium oscillations not only occur in excitable tissues spontaneously, such as muscle, SAN, and neuronal tissues [10], but also occur in pluripotent cells (embryonic stem cells), multipotent cells (mesenchymal stem cells), immature dendritic cells, and G0/G1-phase cells [11]. Zhao et al.'s study proposed that α2δ1 was involved in amplitude-encoding signals that maintain the properties of HCC TICs and showed that inhibition of this calcium signaling could be a therapeutic strategy for HCC [9].
Other authors showed that there were other types of cancer that contain α2δ1+ cells such as small-cell lung cancer, laryngeal squamous cell carcinoma, and gastric cancer, with nearly the same stem cell-like properties [12], resistant to chemotherapy [13], and overexpression of ERK1/2 [14]. Also, they showed that inhibition of α2δ1 may serve as a new promising therapeutic target for these CSCs [15,16].
The standard of care for surveillance of HCC is by ultrasound (US) with or without AFP every 6 (4-8) months in a cirrhotic patient [17] and the cornerstone in diagnosis is multiphasic CT and/or MRI. The problem of US is its low sensitivity (46%) especially in detecting small lesions [18] dropping to 33% in obese patients with BMI > 30 [19]; also, AFP had a sensitivity of around 60% [17]. However, CT and MRI had higher sensitivity (69% and 84%, respectively) and specificity (94%), yet not recommended to be used as surveillance tools [17,20]. That necessitates finding a new tool that can detect HCC early with high sensitivity and specificity non-invasively at low cost. From the aforementioned biological properties of α2δ1subunit in tumor-initiating cells in HCC, we assume that it may have also a good diagnostic value.

Methods
The present study was conducted at the Ain Shams University Hospitals, Cairo, in the duration between December 2018 and December 2019. A case-control study was designed with a total of 90 subjects enrolled and divided into 3 groups as follow: 40 patients with HCC, 40 patients with cirrhosis with normal AFP and no radiological evidence of HCC, and 10 apparently healthy individuals as a control group, with no previous history of liver and other chronic disease or malignant diseases and negative for hepatitis viral markers. HCC patients were diagnosed according to definitive criteria in multiphasic CT/MRI showing arterial enhancement and delayed venous washout; any subject that had a history of malignancy of other organs, inflammatory diseases, or hematological diseases was excluded. The study protocol was approved by the local ethics committee of Ain Shams University. Informed consent was taken from both the patients and control group subjects after explaining the aim and concerns of the study. For all subjects, the following were done: a collection of relevant clinical data, basic laboratory tests including alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, serum albumin, AFP, INR, CBC, kidney function test, and hepatitis serology (HBsAg and HCV Ab) levels were measured using commercially available kits.

Blood sample collection, storage, and analysis
The serum sample was collected from all study subjects. Collected serum was allowed to clot for 10-20 min at room temperature. Centrifuge was at 2000-3000 rpm for 20 min. When the analysis was not performed immediately, the samples were frozen and stored at 80°C until use. The serum concentration of the α2δ1 subunit was measured using an ELISA kit according to the manufacturer's guidelines (MyBioSource, USA).

Statistical analysis
IBM SPSS Advanced Statistics version 21 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. The quantitative variables were presented as the mean ± standard deviation or median percentile of the interquartile range (25th to 75th) for non-parametric data analysis done by the Mann-Whitney test. For parametric comparison between two/all study groups, Student t test or ANOVA was used. Spearman rho was used for correlation. The receiver operating characteristic (ROC) curve was applied to identify the best cut-off values for, the α2δ1 subunit. P values of < 0.05 were considered significant.

Baseline characteristics of the studied subjects
The mean age of the whole studied sample was 50.5 years (± 11.75) and the number distribution of gender among all subjects was 60 (66.6%) male and 30 (33.3%) female. The etiology of liver disease in the studied sample: 12 patients (15%) were HBV positive, where 45 patients (56.25%) were HCV positive, the remaining 23 patients (28.75%) were non-viral causes of liver disease (Table 1). Liver fibrosis was assessed by FIB-4 with a median value of 3.01 (IQR 2.03) across both groups. The number of smokers across both HCC and cirrhotic groups was 26 (32.5%); 54 (67.5%) were non-smokers. Regarding associated comorbidities in both groups, 16 patients (20%) had hypertension and 29 (36.25%) had diabetes (Table 3). Child-Pugh A in this studied sample was 3 patients (7.5%) in the cirrhotic group and 12 patients (30%) in the HCC group. Child-Pugh B was 36 patients (90%) and 22 patients (55%) in cirrhotic and HCC groups, respectively. Child-Pugh C in the cirrhotic group was 1 patient (2.5%) while in the HCC group were 6 patients (15%). The median value AFP in the cirrhotic group was 4.9 IU/mL (IQR 2.7) while its median value in HCC group was 485 IU/mL (IQR 463). The mean value of the MELD score in cirrhotic groups was 18.5 ± 2.3, while in the HCC group the mean value was 16.6 ± 5.3.
There were statistically significant differences between the cirrhotic group and HCC group as regard ALT,  AST, serum albumin, and INR, while there were no statistically significant differences between the two groups as regard total bilirubin and platelet (Table 2).
Evaluation of serum α2δ1 subunit in-clinic-laboratory feature and etiology of liver disease The mean levels of α2δ1 in males were 11.04 ± 8.45 ng/ dL while in females were 12.51 ± 9.67 ng/dL; this showed no statistically significant differences across all study subjects. Also, there was no correlation with age or FIB-4 ( Table 1). The mean value of α2δ1in HCV patients was 11.57 ± 8.96 ng/dL and in HBV patients was 16.93 ± 6.04 ng/dL, while in the non-viral group was 13.35 ± 8.09 ng/dL; this was nonsignificant (Table 1). There were no statistically significant differences between the mean levels of α2δ1 as regarding smoking status, diabetes, and hypertension within each group (cirrhotic and HCC) ( Table 3). AFP in both the HCC and cirrhotic groups showed strong positive correlation with α2δ1 (r = 0.85, P < 0.001). However, there was no statistically significant difference between the α2δ1 subunit as regards laboratory tests (ALT, AST, total bilirubin, albumin, and INR) within each group (Table 2). Despite platelets showing significant correlation in the cirrhotic group, this correlation was a very weak negative correlation (r = − 0.31, P = 0.05).

Evaluation of serum α2δ1 subunit as a potential diagnostic marker for HCC
To further investigate the diagnostic value of serum α2δ1 subunit in HCC, ROC curves were constructed revealed excellent diagnostic value with AUC = 0.974 and P < 0.0001. The serum level of the α2δ1 subunit at the cut-off value ≥ 8.75 ng/dL showed a sensitivity of 95%, a specificity of 80%, a positive predictive value (PPV) of 82.6%, and a negative predictive value (NPV) of 94.1% with accuracy = 87.5%. However, its level at a cut-off 12 ng/dL showed a sensitivity of 85%, a specificity of 100%, PPV of 100%, and NPV of 87% with accuracy = 92.5%

Discussion
Many studies showed an expression of α2δ1 subunit in different types of malignancies such as pancreatic [21], ovarian [22], and lung tumors [14] and demonstrated a potential resistance to chemotherapy and may carry bad prognosis. Han et al. as well identified a subpopulation of TICs expressing α2δ1 subunit, which is essential for the activation of calcium influx that controls the TIC ability of HCC by an antibody against Hep-12 cells [23].
Since the liver cancer is one of the highest malignancies in Egypt and carries a bad prognosis, early diagnosis and management are essential [24].
In the present study, the serum level of α2δ1 subunit was significantly higher in the HCC group than both the cirrhotic and control group, in agreement with the study conducted by Zhao et al. who first reported α2δ1 subunit in HCC; on 86 freeze resected HCC tumor sample compared to the nearby para-cancerous area for the presence of 1B50-1 positive staining (a monoclonal Ab against α2δ1 subunit), they a found significantly higher percent in tumor cells (72.1%) than non-tumor cells (46.5%), P = 0.0006 [9].
The study conducted by Badr et al. was in agreement with our results as they showed significantly higher levels of serum α2δ1 subunit in HCC group than both the cirrhotic and control (P < 0.05) with mean values as 20.12 ± 3.7 ng/mL, 10.41 ± 3.4 ng/mL, and 10.2 ± 2.98 ng/mL, respectively. However, they did not find a significant difference between the cirrhosis and control groups and this was against our finding as we found significantly lower levels in the control than cirrhosis (P = 0.002); this may be attributed by the sample size used in their study [25].
In work-related to Zhao et al.'s study, Han and his team performed the expression level of α2δ1 mRNA in 85 of the resected HCC and para-cancerous area found that there were no significant differences between its level and clinicopathological features (age, gender, and cirrhotic vs non-cirrhotic) [23]; this was as our findings.
Serum α2δ1 subunit level was not significantly different in associated comorbidity as smoking, DM, and HTN or viral (HBV and HCV) or non-viral etiology of liver disease.
Badr et al. who first describe serum level of α2δ1 subunit as a potential marker for HCC found that the sensitivity and specificity at level 14.22 ng/dL are 100% and 96%, respectively, with PPV 98%, NPV 100%, and accuracy 98.7% [25]. Per that, we found a sensitivity of 95% and specificity of 80% at a level of ≥ 8.75 ng/dL with PPV of 82.6%, NPV of 94.1%, and accuracy of 87.5%; this indicates that serum level of α2δ1 subunit may be used as a good biomarker for diagnosis of HCC with high sensitivity regardless of the different etiologies of liver disease as shown in our results. Besides, we demonstrated a higher level that yields more specificity and higher diagnostic potentially of α2δ1 subunit was that at a cut-off of 12 ng/dL; sensitivity and specificity were 85% and 100%, respectively, with 100% PPV, 87% NPP, and 92.5% accuracy.
Some authors found that the α2δ1 subunit may have a role in prognosis and recurrence of HCC. Zhao and his team showed that the presence of positive staining of anti α2δ1 subunit antibodies cells in the para-cancerous tissues did correlate significantly with hepatic very rapid recurrence, and a lower rate of 4-year overall survival post-surgery (P = 0.00004 and 0.00005, respectively). Also, Han et al. identified that high level of α2δ1 mRNA is an independent risk factor of poor survival for HCC patients (relative risk = 2.66, P = 0.005) [9,23]

Conclusion
The study showed serum α2δ1 subunit as a potential non-invasive marker with excellent sensitivity for diagnosis of HCC regardless of the etiology of liver disease and not affected by common comorbidity as smoking, diabetes, and hypertension.