**3. MicroRNAs as novel biomarkers for hepatocarcinoma diagnosis**

#### **3.1. Diagnostic potential of single microRNA for hepatocarcinoma**

With increasing incidence and death rate of hepatocarcinoma, it is very expected to identify one or several diagnostic biomarkers (with both high sensitivity and specificity) such as microRNAs for this malignancy. Growing evidence has shown that the expression change of all microRNAs in the peripheral blood may have a unique advantage because they exhibit tissue specificity and relative stability and can also provide some specific cues for early and small hepatocarcinoma [8–14]. Until now, more than 30 circulating microRNAs have been identified to have diagnostic potential for hepatocarcinoma (**Table 1**). For example, microRNA-122 has been reported as a hepatic-specific microRNA, accounting for 70% of the total microR-NAs in hepatic tissues. This microRNA, a high conservative microRNA between vertebrate species, is indicative of a regulator of fatty acid metabolism and playing a critical role in liver homeostasis and tumorigenesis [19, 33, 34]. Increasing evidence has shown that elevated serum amount of microRNA-122 is positively associated with the severity of hepatic diseases including hepatitis, fatty- and alcohol-related liver damage, and drug-induced hepatotoxicity [35–39]. Interestingly, this increasing serum expression of microRNA-122 is noticeable and indicated that it could serve as a potential biomarker for the detection of patients with hepatocarcinoma from healthy controls with about 85% of the area under the receiver operating characteristic curve (AUC), 80% of sensitivity, and 80% of specificity [40, 41]. These results indicate that the dysregulated miR-122 in the peripheral blood may be used as a potential marker for hepatocarcinoma diagnosis. Results from retrospective studies have suggested that the microRNA-200 family (consisting of microRNA-200a and microRNA -200b) is also a promising biochemical biomarker for hepatocarcinoma diagnosis because of its deregulation during the development of both hepatic fibrosis and hepatocarcinoma [42, 43]. The elevated plasma levels of microRNA-21 can distinguish patients with hepatocarcinoma from cases with chronic hepatitis (with 61.1% of sensitivity and 83.3% of specificity) or healthy controls (the corresponding sensitivity and specificity are 87.3 and 92.0%, respectively) [44]. This suggests that this biomarker may have higher diagnostic potential than AFP. Some


targeting genes via recognizing and integrating into the 3′-untranslated region of these genes' mRNAs. On the basis of perfect or imperfect base-base complementarity of microRNAs-their targeting mRNA binding, one microRNA specifically regulates the expression of multiple mRNAs, and at the same time, one mRNA might be inhibited by multiple microRNAs. This indicates the specificity and diversity of microRNAs regulating gene expression. In the past decades, microR-NAs are emerged as important players in a very wide range of physiological processes including cell differentiation, cell proliferation and apoptosis, cycle regulation, survival, detoxification, physiological timing, metabolism, angiogenesis, hormone secretion, and DNA damage repair (**Figure 1**). Furthermore, growing evidence has shown that microRNAs can also display a role in the etiology and pathogenesis of various cancers by targeting many oncogenes or tumor inhibitive genes (**Figure 1**) [24, 27, 29–32]. Recent several reports have exhibited that some microRNAs involve in the tumorigenesis and procession of hepatocarcinoma and may become new potential

markers for hepatocarcinoma diagnosis and prognosis [24, 27, 29, 31, 32].

106 Hepatocellular Carcinoma - Advances in Diagnosis and Treatment

**3.1. Diagnostic potential of single microRNA for hepatocarcinoma**

**3. MicroRNAs as novel biomarkers for hepatocarcinoma diagnosis**

With increasing incidence and death rate of hepatocarcinoma, it is very expected to identify one or several diagnostic biomarkers (with both high sensitivity and specificity) such as microRNAs for this malignancy. Growing evidence has shown that the expression change of all microRNAs in the peripheral blood may have a unique advantage because they exhibit tissue specificity and relative stability and can also provide some specific cues for early and small hepatocarcinoma [8–14]. Until now, more than 30 circulating microRNAs have been identified to have diagnostic potential for hepatocarcinoma (**Table 1**). For example, microRNA-122 has been reported as a hepatic-specific microRNA, accounting for 70% of the total microR-NAs in hepatic tissues. This microRNA, a high conservative microRNA between vertebrate species, is indicative of a regulator of fatty acid metabolism and playing a critical role in liver homeostasis and tumorigenesis [19, 33, 34]. Increasing evidence has shown that elevated serum amount of microRNA-122 is positively associated with the severity of hepatic diseases including hepatitis, fatty- and alcohol-related liver damage, and drug-induced hepatotoxicity [35–39]. Interestingly, this increasing serum expression of microRNA-122 is noticeable and indicated that it could serve as a potential biomarker for the detection of patients with hepatocarcinoma from healthy controls with about 85% of the area under the receiver operating characteristic curve (AUC), 80% of sensitivity, and 80% of specificity [40, 41]. These results indicate that the dysregulated miR-122 in the peripheral blood may be used as a potential marker for hepatocarcinoma diagnosis. Results from retrospective studies have suggested that the microRNA-200 family (consisting of microRNA-200a and microRNA -200b) is also a promising biochemical biomarker for hepatocarcinoma diagnosis because of its deregulation during the development of both hepatic fibrosis and hepatocarcinoma [42, 43]. The elevated plasma levels of microRNA-21 can distinguish patients with hepatocarcinoma from cases with chronic hepatitis (with 61.1% of sensitivity and 83.3% of specificity) or healthy controls (the corresponding sensitivity and specificity are 87.3 and 92.0%, respectively) [44]. This suggests that this biomarker may have higher diagnostic potential than AFP. Some


**MicroRNAs Source Diagnostic relevance Expression level AUC**

miR-4651 Serum AHCCs (n = 279) vs. CTLs

miR-143 Serum HCCs (n = 131) vs. HCs

miR-125b Plasma HCCs (n = 64) vs. HCs

miR-125b Plasma HCCs (n = 64) vs. CHBs

miR-125b Plasma HCCs (n = 64) vs. CCs

miR-150 Serum HCCs (n = 120) vs. CHBs

miR-150 Serum HCCs (n = 120) vs. HCs

miR-106b Plasma HCCs (n = 47) vs. CTLs

miR-200a Serum HCCs (n = 22) vs. HCs

miR-200a Serum HCCs (n = 22) vs. CCs

miR-143 Serum HCCs (n = 95) vs. CHCs

miR-215 Serum HCCs (n = 95) vs. CHCs

miR-143 Serum HCCs (n = 95) vs. HCs

miR-215 Serum HCCs (n = 95) vs. HCs

miR-101 Serum HCCs (n = 67) vs. HCs

miR-483-5p Serum HCCs (n = 49) vs. HCs

miR-122a Plasma HCCs (n = 85) vs. HCs

miR-618 Urine HCCs (n = 32) vs. CTLs

miR-650 Urine HCCs (n = 32) vs. CTLs

tissue

miR-126 tumor

(n = 662)

(n = 122)

(n = 56)

(n = 63)

(n = 59)

(n = 110)

(n = 120)

(n = 61)

(n = 15)

(n = 22)

(n = 118)

(n = 118)

(n = 127)

(n = 127)

(n = 30)

(n = 49)

(n = 85)

(n = 74)

(n = 74)

receiver operating characteristic curve; CI, confidence interval; Refs, references.

**Table 1.** The microRNAs as diagnostic biomarkers for hepatocarcinoma.

**(95% CI)**

(0.82–0.88)

Downregulated 0.83 80.3 82.4 [56]

Downregulated 0.89 90.0 80.0 [57]

Downregulated 0.96 90.0 90.0 [57]

Downregulated 0.96 90.0 90.0 [57]

(0.84–0.93)

(0.90–0.96)

(0.69–0.97)

(0.56–0.89)

(0.51–0.76)

(0.67–0.95)

(0.68–0.92)

(0.72–0.97)

(0.69–0.87)

Upregulated 0.91 75.5 89.8 [62]

Downregulated 0.71 70.6 67.1 [63]

Upregulated 0.66 64.0 68.0 [47]

Downregulated 0.65 72.0 58.0 [47]

Upregulated 0.81 0.7 0.8 [59]

Upregulated 0.85

The Diagnostic and Prognostic Potential of MicroRNAs for Hepatocellular Carcinoma

Downregulated 0.88

Downregulated 0.93

Downregulated 0.82

Downregulated 0.73

Upregulated 0.62

Upregulated 0.80

Upregulated 0.80

Upregulated 0.82

Downregulated 0.79

HCCs vs. CAs Upregulated / / / [48]

Abbreviation: miR, microRNA; HCCs, cases with hepatocarcinoma; CTLs, non-HCC controls (including healthy control and other nontumor controls); HCs, healthy controls; CCs, controls with liver cirrhosis; CHBs, patients with chronic hepatitis B; CLDs, cases with nontumor chronic liver diseases; Sen, sensitivity; Spe, specificity; AUC, the area under the

**Sen (%)**

http://dx.doi.org/10.5772/intechopen.72276

**Spe (%)**

78.1 92.1 [55]

79.1 76.5 [58]

82.5 83.7 [58]

/ / [60]

/ / [60]

78.0 64.0 [46]

78.0 89.0 [46]

78.0 89.0 [46]

80.0 91.0 [46]

76.1 70.0 [61]

**Refs**

109


Abbreviation: miR, microRNA; HCCs, cases with hepatocarcinoma; CTLs, non-HCC controls (including healthy control and other nontumor controls); HCs, healthy controls; CCs, controls with liver cirrhosis; CHBs, patients with chronic hepatitis B; CLDs, cases with nontumor chronic liver diseases; Sen, sensitivity; Spe, specificity; AUC, the area under the receiver operating characteristic curve; CI, confidence interval; Refs, references.

**Table 1.** The microRNAs as diagnostic biomarkers for hepatocarcinoma.

**MicroRNAs Source Diagnostic relevance Expression level AUC**

miR-122-5p Serum HCCs (n = 261) vs. HCs

108 Hepatocellular Carcinoma - Advances in Diagnosis and Treatment

miR-192-5p Serum HCCs (n = 261) vs. HCs

miR-26a-5p Serum HCCs (n = 261) vs. HCs

miR-206 Serum HCCs (n = 261) vs. CCs

miR-143-3p Serum HCCs (n = 261) vs. CCs

miR-433-3p Serum HCCs (n = 261) vs. CCs

miR-1228-5p Serum HCCs (n = 261) vs. CCs

miR-199a-5p Serum HCCs (n = 261) vs. CCs

miR-122-5p Serum HCCs (n = 261) vs. CCs

miR-192-5p Serum HCCs (n = 261) vs. CCs

miR-26a-5p Serum HCCs (n = 261) vs. CCs

miR-16 Serum HCCs (n = 105) vs. CTLs

miR-199 Serum HCCs (n = 105) vs. CTLs

miR-199a Serum HCCs (n = 105) vs. CTLs

miR-375 Serum HCCs (n = 78) vs. HCs

miR-199a-3p Serum HCCs (n = 78) vs. HCs

miR-30c-5p Plasma HCCs (n = 8) vs. CTLs

miR-223-3p Plasma HCCs (n = 8) vs. CTLs

miR-202c-3p Plasma HCCs (n = 8) vs. CTLs

miR-17-57 Plasma HCCs (n = 8) vs. CTLs

miR-4651 Serum AHCCs (n = 279) vs. HCs

miR-4651 Serum AHCCs (n = 279) vs. AHCs

miR-4651 Serum AHCCs (n = 279) vs. ALCs

(n = 173)

(n = 173)

(n = 173)

(n = 233)

(n = 233)

(n = 233)

(n = 233)

(n = 233)

(n = 233)

(n = 233)

(n = 233)

(n = 188)

(n = 188)

(n = 188)

(n = 156)

(n = 156)

(n = 86)

(n = 86)

(n = 86)

(n = 86)

(n = 338)

(n = 292)

(n = 32)

**(95% CI)**

(0.63–0.77)

(0.62–0.77)

(0.70–0.82)

(0.62–0.77)

(0.60–0.73)

(0.58–0.71)

(0.47–0.61)

(0.52–0.66)

(0.69–0.81)

(0.62–0.75)

(0.68–0.81)

Downregulated / 72.1 88.8 [14]

Downregulated / 62.9 93.5 [14]

Downregulated / 78.1 64.5 [14]

(0.56–0.74)

(0.83–0.94

Upregulated / / / [54]

Downregulated / / / [54]

Upregulated / / / [54]

Upregulated / / / [54]

(0.86–0.92)

(0.78–0.85)

(0.71–0.88)

Downregulated 0.70

Downregulated 0.70

Downregulated 0.76

Upregulated 0.69

Upregulated 0.66

Upregulated 0.64

Upregulated 0.54

Downregulated 0.59

Downregulated 0.75

Downregulated 0.69

Downregulated 0.74

Downregulated 0.64

Downregulated 0.88

Upregulated 0.89

Upregulated 0.82

Upregulated 0.80

**Sen (%)**

**Spe (%)**

48.9 82.2 [52]

71.9 75.6 [52]

68.9 74.4 [52]

77.8 68.9 [52]

60.7 72.7 [52]

56.4 67.4 [52]

66.7 47 [52]

59.3 57.6 [52]

48.9 90.2 [52]

54.8 83.3 [52]

60.7 90.9 [52]

/ / [53]

/ / [53]

78.1 99.1 [55]

78.1 85.3 [55]

78.1 81.2 [55]

**Refs**

other serum microRNAs, such as microRNA-15b, microRNA-130b, miR-143, and miR-215, are additional potential biomarkers that are significantly dysregulated in hepatocarcinoma [45, 46]. Noticeably, these biomarkers also exhibit their diagnostic potential for patients with early-stage hepatocarcinoma and/or negative-status AFP [45, 46].

Recently, some evidence has also exhibited that microRNAs in urine samples and liver tissues have screening potential for hepatocarcinoma (**Table 1**). Actually, the detection of five deregulated microRNAs, including microR-618, microRNA-625, microRNA-650, microRNA-532, and miR-516-5P, in the urine samples has already been used for screening patients with the early and small hepatocarcinoma from these with risk factors such as chronic virus hepatitis, liver cirrhosis, and dysplasia [47]. Barshack et al. [48] investigated differential diagnosis potential of microR-NAs for discriminating hepatocarcinoma from metastatic tumors in the liver using custom microarray expression technique. In their study, they tested the distributed features of microR-NAs among 144 tumor samples with or without metastatic adenocarcinoma and similar hepatocarcinoma in the morphology and immune types and found that microR-141 and microR-200c can promote non-hepatic epithelial phenotypes while microRNA-126 displays hepatic epithelial phenotypes. Higher expression of microRNA-126 is further shown in these tissue samples with hepatocarcinoma. Therefore, the change profiles of microRNAs in body fluids (such as urine) and tumor tissues may represent a kind of gold biomarkers for such cancers as liver carcinoma.

However, the specificity of a single microRNA identifying hepatocarcinoma is relatively poor. For example, the serum level of aforementioned liver-specific microRNA-122 is upregulated not only among cases with hepatocarcinoma but also among these with chronic virus hepatitis, liver cirrhosis, and fatty liver diseases caused by alcohol or non-alcohol [49, 50]. Evidence has shown that serum microRNA-122 does not discriminate patients with hepatocarcinoma from these with chronic hepatitis, although higher expression is observed among cancer cases [40, 41]. This indicates that more investigations on the basis of large size of samples and the prospective randomized controlled trials should help us for addressing these concerns.

### **3.2. Diagnostic potential of microRNA panel for hepatocarcinoma**

Because hepatocarcinoma is a multifactor-induced highly complex malignant disease with heterogeneous feature, a combination of multiple microRNAs in place of a single microRNA may have higher accuracy for hepatocarcinoma discrimination. Several circulating microRNA panels have been reported to have higher early diagnostic value for hepatocarcinoma (**Table 2**) [47, 51, 52, 64–70]. For example, Lin et al. [70] preformed a three-stage study consisting of the discovery stage (including 6 cases with hepatocarcinoma and 8 cases with chronic hepatitis B), the training stage (including 108 cases with hepatocarcinoma, 51 cases with chronic hepatitis B, 47 cases with liver cirrhosis, and 51 healthy controls), and the validation stage (including 229 patients with hepatocarcinoma and 424 controls with or with nontumor liver diseases). In the first stage, they identified 31 different serum microRNAs between individuals with hepatocarcinoma and those with chronic hepatitis B using the TaqMan Array technique. Next, they validated these different microRNAs and constructed diagnostic panel containing miR-29a, miR-29c, miR-133a, miR-143, miR-145, miR-192, and miR-505 on the basis of logistic regression model. Finally, the established serum microRNA panel was tested among

individuals from the training and validation cohorts. These data identified seven microR-NAs and constructed a serum microRNA panel with an increasing diagnostic accuracy for hepatocarcinoma [AUC = 0.826 (0.771–0.880) for training set and 0.817 (0.769–0.865) and 0.884 (0.818–0.951) for two different validation sets, respectively]. Interestingly, a nest case-control study has further proved that this panel could be used to detect preclinical hepatocarcinoma

as well as small-size, early-stage, and α-fetoprotein-negative disease.

receiver operating characteristic curve; CI, confidence interval; Refs, references.

**Table 2.** The microRNA panel as diagnostic biomarkers for hepatocarcinoma.

**MicroRNA panel Source AUC (95%** 

miR-10b + miR-106b + miR-181a Serum 0.94

miR-10b + miR-106b + miR-181a Serum 0.91

miR-27b-3p + miR-192-5p Serum 0.84

miR-92-3p + miR-107 + miR-3126-5p Serum 0.97

88-miRNA Serum 1.00

miR-206 + miR-143-3p + miR-433-3p + miR-1228-5p + miR-199a-5p + miR-122-5p + miR-192-5p + miR-26a-5p

miR-206 + miR-143-3p + miR-433-3p + miR-1228-5p + miR-199a-5p + miR-122-5p + miR-192-5p + miR-26a-5p

miR-122 + miR-192 + miR-21 + miR-223 + miR-26a + miR-27a + miR-801

miR-122 + miR-192 + miR-21 + miR-223 + miR-26a + miR-27a + miR-801

miR214-5p + miR-125b + miR-1269 + miR-

miR-29a + miR-29c + miR-133a + miR-143 + miR-145 + miR-192 + miR-505

miR-29a + miR-29c + miR-133a + miR-143 + miR-145 + miR-192 + miR-505

375

**CI)**

Serum 0.89

Serum 0.89

Plasma 0.86

Plasma 0.89

miR-122 + miR-885-5p + miR-29b Serum 1.00 / / HCCs (n = 192) vs.

Serum 0.82

Serum 0.88

miR-618 + miR-650 Urine 0.69 58.0 75.0 HCCs (n = 32) vs.

(0.89–0.99)

(0.80–0.97)

(0.85–0.94)

(0.84–0.94)

(0.83–0.90)

(0.85–0.92)

(0.78–0.89)

(0.95–0.99)

(0.97–1.00)

(0.77–0.87)

(0.82–0.95)

Abbreviation: miR, microRNA; HCCs, cases with hepatocarcinoma; CTLs, non-HCC controls (including healthy control and other nontumor controls); HCs, healthy controls; CCs, controls with liver cirrhosis; CHBs, patients with chronic hepatitis B; CLDs, cases with nontumor chronic liver diseases; Sen, sensitivity; Spe, specificity; AUC, the area under the

**Sen (%) Spe (%)**

The Diagnostic and Prognostic Potential of MicroRNAs for Hepatocellular Carcinoma

**Diagnostic relevance**

http://dx.doi.org/10.5772/intechopen.72276

HCs (n = 50)

CLDs (n = 31)

HCs (n = 173)

CCs (n = 233)

CTLs (n = 303)

CTLs (n = 194)

CTLs (n = 91)

HCs (n = 40)

CCs (n = 233)

HCs (n = 84)

HCs (n = 96)

CTLs (n = 199)

CTLs (n = 90)

CTLs (n = 74)

/ / HCCs (n = 27) vs.

/ / HCCs (n = 27) vs.

82.8 83.3 HCCs (n = 261) vs.

81.6 84.6 HCCs (n = 261) vs.

68.6 90.1 HCCs (n = 204) vs.

81.8 83.5 HCCs (n = 196) vs.

0.7 0.9 HCCs (n = 91) vs.

/ / HCCs (n = 115) vs.

100.0 99.2 HCCs (n = 261) vs.

74.5 89.9 HCCs (n = 153) vs.

85.7 91.1 HCCs (n = 49) vs.

Serum 0.95 96.9 83.2 HCCs (n = 224) vs.

**Refs**

111

[51]

[51]

[52]

[52]

[64]

[64]

[65]

[66]

[67]

[68]

[69]

[70]

[70]

[47]


Abbreviation: miR, microRNA; HCCs, cases with hepatocarcinoma; CTLs, non-HCC controls (including healthy control and other nontumor controls); HCs, healthy controls; CCs, controls with liver cirrhosis; CHBs, patients with chronic hepatitis B; CLDs, cases with nontumor chronic liver diseases; Sen, sensitivity; Spe, specificity; AUC, the area under the receiver operating characteristic curve; CI, confidence interval; Refs, references.

**Table 2.** The microRNA panel as diagnostic biomarkers for hepatocarcinoma.

other serum microRNAs, such as microRNA-15b, microRNA-130b, miR-143, and miR-215, are additional potential biomarkers that are significantly dysregulated in hepatocarcinoma [45, 46]. Noticeably, these biomarkers also exhibit their diagnostic potential for patients with

Recently, some evidence has also exhibited that microRNAs in urine samples and liver tissues have screening potential for hepatocarcinoma (**Table 1**). Actually, the detection of five deregulated microRNAs, including microR-618, microRNA-625, microRNA-650, microRNA-532, and miR-516-5P, in the urine samples has already been used for screening patients with the early and small hepatocarcinoma from these with risk factors such as chronic virus hepatitis, liver cirrhosis, and dysplasia [47]. Barshack et al. [48] investigated differential diagnosis potential of microR-NAs for discriminating hepatocarcinoma from metastatic tumors in the liver using custom microarray expression technique. In their study, they tested the distributed features of microR-NAs among 144 tumor samples with or without metastatic adenocarcinoma and similar hepatocarcinoma in the morphology and immune types and found that microR-141 and microR-200c can promote non-hepatic epithelial phenotypes while microRNA-126 displays hepatic epithelial phenotypes. Higher expression of microRNA-126 is further shown in these tissue samples with hepatocarcinoma. Therefore, the change profiles of microRNAs in body fluids (such as urine) and tumor tissues may represent a kind of gold biomarkers for such cancers as liver carcinoma. However, the specificity of a single microRNA identifying hepatocarcinoma is relatively poor. For example, the serum level of aforementioned liver-specific microRNA-122 is upregulated not only among cases with hepatocarcinoma but also among these with chronic virus hepatitis, liver cirrhosis, and fatty liver diseases caused by alcohol or non-alcohol [49, 50]. Evidence has shown that serum microRNA-122 does not discriminate patients with hepatocarcinoma from these with chronic hepatitis, although higher expression is observed among cancer cases [40, 41]. This indicates that more investigations on the basis of large size of samples and the prospective randomized controlled trials should help us for addressing these concerns.

early-stage hepatocarcinoma and/or negative-status AFP [45, 46].

110 Hepatocellular Carcinoma - Advances in Diagnosis and Treatment

**3.2. Diagnostic potential of microRNA panel for hepatocarcinoma**

Because hepatocarcinoma is a multifactor-induced highly complex malignant disease with heterogeneous feature, a combination of multiple microRNAs in place of a single microRNA may have higher accuracy for hepatocarcinoma discrimination. Several circulating microRNA panels have been reported to have higher early diagnostic value for hepatocarcinoma (**Table 2**) [47, 51, 52, 64–70]. For example, Lin et al. [70] preformed a three-stage study consisting of the discovery stage (including 6 cases with hepatocarcinoma and 8 cases with chronic hepatitis B), the training stage (including 108 cases with hepatocarcinoma, 51 cases with chronic hepatitis B, 47 cases with liver cirrhosis, and 51 healthy controls), and the validation stage (including 229 patients with hepatocarcinoma and 424 controls with or with nontumor liver diseases). In the first stage, they identified 31 different serum microRNAs between individuals with hepatocarcinoma and those with chronic hepatitis B using the TaqMan Array technique. Next, they validated these different microRNAs and constructed diagnostic panel containing miR-29a, miR-29c, miR-133a, miR-143, miR-145, miR-192, and miR-505 on the basis of logistic regression model. Finally, the established serum microRNA panel was tested among

individuals from the training and validation cohorts. These data identified seven microR-NAs and constructed a serum microRNA panel with an increasing diagnostic accuracy for hepatocarcinoma [AUC = 0.826 (0.771–0.880) for training set and 0.817 (0.769–0.865) and 0.884 (0.818–0.951) for two different validation sets, respectively]. Interestingly, a nest case-control study has further proved that this panel could be used to detect preclinical hepatocarcinoma as well as small-size, early-stage, and α-fetoprotein-negative disease.

Similarly, Jiang et al. [59] and Zhou et al. [64] also attempted to identify possible combination of different microRNAs for increasing diagnostic accuracy of hepatocarcinoma on the basis of different controls with or without liver diseases. They found that the panel consisting of miR-10b, miR-106b, and miR-181a as well as the combination of miR-122, miR-192, miR-21, miR-223, miR-26a, miR-27a, and miR-801 can improve detection of hepatocarcinoma. These reports indicate that the panel of microRNAs may have better performance than a single-microRNA assay.

**MicroRNAs Source Expression level Prognostic significance HR** 

poor OS

poor OS

chemistry parameters of hepatic necroinflammation, liver function,

The Diagnostic and Prognostic Potential of MicroRNAs for Hepatocellular Carcinoma

cirrhosis, and tumor stage; (2) increasing levels decreased

and synthetic capacity

survival rate

Upregulated Increasing levels correlate with

Upregulated Increasing levels correlate with

Upregulated Increasing levels correlate with

Upregulated Increasing levels correlate with

Upregulated Increasing levels promoting tumor metastasis

poor OS

poor OS

Upregulated Increasing expression promotes

protective factors

metastasis-free survival

6 were risk factors and 14 were

poor OS and RFS

poor OS and RFS

poor OS and RFS

poor OS and RFS

poor OS and RFS

poor OS and RFS

miR-122 Serum Upregulated Increasing levels correlate with

miR-1 Serum Upregulated Increasing levels correlate with

miR-221 Serum Upregulated (1) Correlated with tumor size,

miR-4651 Serum Upregulated Increasing levels correlate with

miR-143 Serum Downregulated Decreasing levels correlate with

miR-150 Serum Downregulated Increasing levels correlate with

miR-21 Serum Upregulated Increasing levels correlate with

10 downregulated and 20 upregulated

miRNAs

miR-1268a Tumor

miR-24 Tumor

miR-429 Tumor

miR-9 Tumor

miR-92b Tumor

miR-221 Tumor

20-miRNA signature

tissues

tissues

tissues

tissues

tissues, serum

Tumor tissues

tissues

miR-122 Serum Upregulated Correlated with clinical

**(95%CI)**

http://dx.doi.org/10.5772/intechopen.72276

OS, 0.08 (0.03–0.22)

OS, 0.45 (0.23–0.86)

OS, 2.67 (1.61–4.42) RFS, 3.62 (1.49–8.81)

OS, 2.44 (1.82–3.23) RFS, 2.86 (2.08–3.85)

OS, 3.58 (2.34–5.46) RFS, 4.75 (2.66–8.47)

OS, 4.64 (2.56–8.41) RFS, 6.94 (3.19–15.08)

/ [56]

/ [123]

/ [102]

0.45 (0.23–0.85)

2.23 (1.33–3.74)

OS, 2.75 (1.58–4.79)

/ [125]

/ [121]

/ [122]

**Refs**

113

[120]

[121]

[55]

[115]

[77]

[78]

[58]

[76]

[124]

### **3.3. Diagnostic potential of microRNAs binding with AFP for hepatocarcinoma**

AFP has been regarded as the most important marker for hepatocarcinoma screening and diagnosis, ever since it was identified in the peripheral blood samples from patients with hepatocarcinoma in 1964 [8, 71, 72]. However, this marker is relatively unsatisfactory because of its low sensitivity and specificity. This is mainly because only 60–80% of cases with hepatocarcinoma show positive AFP, whereas about 40% of cirrhotic patients also exhibit different degree increasing level of serum AFP [73, 74]. Thus, AFP may not be a reliable hepatocarcinoma marker, especially for early-stage and/or AFP-negative hepatocarcinoma. On the basis of low sensitivity and specificity of AFP for hepatocarcinoma diagnosis, the American Association for the Study of Liver Disease Practice Guidelines has thrown it away for prognostic surveillance and tumor diagnosis [75]. However, recent studies have displayed that the combination of AFP in the peripheral blood and microRNAs in body fluids may improve the sensitivity and specificity of hepatocarcinoma diagnosis and increase their diagnostic potential [14, 47, 55, 58, 65–67, 69, 70, 76].

For example, Wu et al. [55] investigated the joint diagnostic value of serum microRNA-4651 and AFP for hepatocarcinoma in 279 hepatocarcinoma patients, 324 controls with liver injury, and 338 healthy controls. Their results imply that serum microRNA-4651 has higher expression level among cases with hepatocarcinoma (AUC of 0.85; sensitivity of 78.1% and specificity of 92.1%); this increasing expression also displays higher diagnostic potential than AFP at cutoff of 20 ng/mL (AFP20) (AUC = 0.80, sensitivity = 61.3%, and specificity = 98.8%) and of 400 ng/mL (AFP400) (AUC = 0.72, sensitivity = 43.0%, and specificity = 100.0%). Noticeably, the combination of serum microRNA-4651 with AFP significantly improves the discrimination power between patients with hepatocarcinoma and with chronic nontumor liver injury (AUC = 0.90, sensitivity = 83.2%, and specificity = 97.1%). Similar findings have also been observed in the analyses of combination of serum AFP and other microRNAs, such as miR-29a, miR-29c, miR-133a, miR-143, miR-145, miR-192, miR-505, miR-16, miR-195, and miR-199a [14, 47, 58, 65–67, 69, 70, 76]. Altogether, these data suggest that the combination of microRNAs with AFP may improve diagnostic potential of hepatocarcinoma.
