Extracellular vesicles as a new hope for diagnosis and therapeutic intervention for hepatocellular carcinoma

Abstract Hepatocellular carcinoma (HCC) is the sixth most common cancer with a high mortality rate. Early diagnosis and treatment before tumor progression into an advanced stage is ideal. The current diagnosis of HCC is mainly based on imaging modalities such as ultrasound, computed tomography, and magnetic resonance imaging. These methods have some limitations including diagnosis in the case of very small tumors with atypical imaging patterns. Extracellular vesicles (EVs) are nanosized vesicles which have been shown to act as an important vector for cell‐to‐cell communication. In the past decade, EVs have been investigated with regard to their roles in HCC formation. Since these EVs contain biomolecular cargo such as nucleic acid, lipids, and proteins, it has been proposed that they could be a potential source of tumor biomarkers and a vector for therapeutic cargo. In this review, reports on the roles of HCC‐derived EVs in tumorigenesis, and clinical investigations using circulating EVs as a biomarker for HCC and their potential diagnostic roles have been comprehensively summarized and discussed. In addition, findings from in vitro and in vivo reports investigating the potential roles of EVs as therapeutic interventions are also presented. These findings regarding the potential benefits of EVs will encourage further investigations and may allow us to devise novel strategies using EVs in the early diagnosis as well as for treatment of HCC in the future.

on tumor size, tumor location, and liver function. These interventions include tumor resection, local ablation, transarterial chemoembolization, and liver transplantation which are shown to be effective for HCC treatment. 4,5 However, some patients cannot be treated by these methods due to an advanced tumor stage and poor liver function. 6 Currently, the overall 5-year survival rate for HCC patients is still low despite the availability of multiple treatment modalities. 7 Because most HCC patients are asymptomatic in the early stages, the early detection of small HCC before it progresses to the later stages is very important. Several pieces of evidence show that early detection of small HCC by screening with a liver ultrasound can significantly improve survival time. 8 However, data from a meta-analysis report showed that the sensitivity of ultrasound screening for detection of an early HCC is only 45%, and the sensitivity of combined ultrasound screening with serum alpha-fetoprotein (AFP) for early HCC detection is only 63%. 9 Currently, imaging such as CT and MRI are primary tools for the diagnosis of HCC in most guidelines. 10,11 With typical arterial enhancement and contrast washout in the delayed phase, HCC can be diagnosed without the need for tissue diagnosis. 12 However, since approximately 40% of HCC patients showed atypical imaging features, this group of patients still need a liver biopsy for a definite diagnosis. 13 All these findings indicate the need for improved tools to detect HCC earlier with increased sensitivity and specificity while being less invasive.
Liquid biopsy is one of the noninvasive methods for cancer diagnosis. Detection of circulating tumor DNA, 14 circulating tumor cell, 15 or extracellular vesicle (EV) in patients' blood has been shown to be useful for cancer diagnosis. However, as the amount of circulating tumor cells and DNA are usually low in number with their short survival time in circulation, these factors, therefore, limit the use of these methods. 16 The extracellular vesicle (EV) is a nano-sized vesicle which can be produced and secreted by all cells. 17 EVs can deliver molecules such as protein, mRNA, microRNA (miRNA), circularRNA (circRNA), long noncoding RNA (lncRNA), and DNA and play an important role in cell-to-cell communication. 17 These vesicles can be found in almost all body fluid types such as blood, urine, ascites, effusion, and breast milk. 18 There is increasing evidence to demonstrate that certain tumor cells can secrete EVs to modify the microenvironment of tumors and have a major role in tumor progression. 19,20 A previous study demonstrated that some liver-derived transcripts can be found in the circulation of an HCC patient. 21 The lipid raft domain structure of EVs can protect these transcripts against degradation by circulating RNase, thus allowing the EV-RNA to be stable in the blood. 22 As a result, these EVs have been proposed as a potential diagnostic biomarker. [22][23][24] Furthermore, it has been shown that EVs could be used to stimulate an immune response as well as to carry the tumor suppressor gene, all of which could have therapeutic potential in cancer treatment. 25,26 In this review, the biologic roles of HCC-derived EVs obtained from in vitro and in vivo reports are comprehensively summarized. Reports on their potential roles as diagnostic biomarkers and possible therapeutic interventions are also presented and discussed.
Literature was searched on the PubMed database from its inception until August 30, 2021 using the following search terms: hepatocellular carcinoma, extracellular vesicle, and exosome. In vitro, in vivo, and clinical studies were included in this review.

CELL-DERIVED EVS: REPORTS FROM IN VITRO STUDIES
Tumor microenvironments play an essential role in HCC development and progression. 27,28 Recent evidence showed the dysregulation of multiple signaling pathways and cell-to-cell communication in HCC. 29 HCC-derived EVs are one of the major mediators of cell-to-cell communication. 30 They were found to contain multiple miR-NAs, lncRNA, and circRNA which regulate tumor cell proliferation, 31-37 migration, 33 and decrease tumor cell apoptosis and chemoresistance. 32,36,37 Several RNAs can regulate the tumor microenvironment by increasing angiogenesis and decreasing cell adhesion. 31,33,34 In addition, serum-free miRNA in cancer patients has been shown to correlate with tumor cell proliferation, metabolism, invasion, and metastasis. 38 Recent evidence showed that other noncoding RNA (lncRNA and circRNA) are also related to tumorigenesis. CircRNA has been demonstrated to act as an miRNA sponge, 39 whereas the dysregulation of lncRNA was associated with cancer. 40 High metastatic HCC cell (LM3) had increased expression of EV-circ-PTGR1 which has been shown to increase cell migration and apoptosis. 41 HepG2 and HuH7 cell lines also had increased EV-lnc 544, 239, 959, 171, and 85 and were shown to increase cell proliferation. 42 As these molecules in HCC-derived EVs promoted the survival of the tumor cells, they could be a target for HCC treatment. A summary of these reports from in vitro studies is shown in Table 1 [43][44][45][46][47][48][49][50][51][52][53][54] Recent evidence has also shown the superiority of EV-miRNA over serum biomarkers in accurate diagnosis. [55][56][57] Since EVs selectively pack and carry specific cargo (i.e., protein, DNA, RNA, and tumor-specific transcripts) from their cell of origin, it is proposed that the cargos in EVs are potential diagnostic biomarkers. [58][59][60] EVs can be secreted by all cell types including HCC cells 61 and can be detected in bodily fluids, especially in the blood. 18 Unlike RNA in the serum, RNA inside the EV can be protected from RNase and other adverse conditions in circulation making EV-RNA more stable in the blood. EV-RNA can also be used as a diagnostic biomarker for tumors. 23 Moreover, EV-RNA can also be quantitated by qRT-PCR, thus making their analysis comparable with conventional proteomic methods. 62 Currently, there are multiple methods to extract EVs from plasma by using their physical properties including ultracentrifugation, filtration, size exclusion chromatography, and precipitation. Ultracentrifugation (UC) is the commonly used method; however, it is burdensome and is unlikely to be used in clinical practice. 63 The filtration method uses less time compared with the UC, however, some EVs may be lost due to the jamming of EVs on the filter. 64 Size exclusion chromatography has been shown to effectively isolate protein contaminants from EVs; however, it has relatively low throughput. 65 Precipitation with UC is a time-saver and can be used to test multiple concurrent samples simultaneously. 66 However, some non-EV contaminants will also be obtained. Recently, Sun et al. had developed the novel EV purification system, EV Click chips. 67 This method combined covalent chemistry-mediated EV capture/release, multimarker antibody cocktails, nanostructured substrates, and microfluidic chaotic mixers to selectively extract tumor-derived EV from total circulating EV. This method has been shown to overcome the limitation of prior methods for EV extraction and was proposed to increase the diagnostic power of circulating EV. 67 In the majority of reports included in this review, a combination of a UC method and other methods were frequently used. However, despite almost all studies used the qRT-PCR technique to measure the EV content (Table 2), one used the TLN biochip combined with TIRF microscopy to detect small fragments of mRNA. 55 This method might be more effective than qRT-PCR which can detect only intact large fragments of mRNA.
Many clinical studies have shown alterations in circulating EV-RNA in HCC patients. Wang et al. and Sohn et al. investigated both EV-miRNA and serum-free miRNA. They proved that miRNA detection in EVs can result in increased sensitivity compared with serum-free miRNA. 68,69 Seven reports found a significant difference in circulating EV-RNA in HCC patients compared with LC or CHB patients which represent real-life situations. 56,[67][68][69][70][71][72] Moreover, multiple EV-mRNA, miRNA, lncRNA, and circRNA have been shown to correlate with tumor burden. 35,62,67,68,72,73 In addition, EV surface antigens (Annexin V, EpCAM, and ASGPR1) were also detected in the serum of HCC and CCA patients. 74 Reports on these findings are summarized in Table 2.

| EV AS A DIAGNOSTIC TEST: EVIDENCE FROM CLINICAL REPORTS
In the past decade, there are several studies which used a combination of multiple circulating EV-RNA or EVsurface antigens to develop diagnostic tests for HCC. There are five reports of high diagnostic performance model scores for diagnosing HCC in liver cirrhosis patients. 42,56,57,67,70 Combined Z-score of 10 EV-mRNA (ALB, GPC3, AFP, AHSG, APOH, FABP1, FGB, FGG, RBP4, and TF) showed high diagnostic performance for HCC diagnosis from at-risk patients (sensitivity 93.8%, specificity 74.5%, AUC 0.87) and other primary cancer (sensitivity 95.7%, specificity 89.5%, AUC 0.95). 67 High ability for early HCC detection from liver cirrhosis patients was also demonstrated (sensitivity 94.4%, specificity 88.5%, AUC 0.93). 67 One report used a combination of serum EV-miR-21-5p, EV-miR-92a-3p, and AFP which showed a sensitivity of 95% with a specificity of 50%, and an AUC of 0.85. 56,57 The other report used a combination of serum EV-miR-122, EV-miR-148a, and AFP which resulted in a sensitivity of 86%, specificity of 87.5%, and AUC of 0.93. 56,57 A better diagnostic performance using a combination of model scores was demonstrated compared with conventional biomarkers (AFP and GPC3) in HCC patients. 55,56 Recently, two reports using the EV-lncRNA method have demonstrated that it could provide higher diagnostic performance than serum AFP. 42,70 EV-LINC00853 demonstrated the sensitivity of 93.75% with a specificity of 89.77% and AUC of 0.97 for the diagnosis of early HCC in patients at-risk group (liver cirrhosis and hepatitis B). 70 In addition, EV-lnc85 showed a high diagnostic performance for differentiating both AFP-positive (AUC 0.90) and AFP-negative HCC (AUC 0.88) from liver cirrhosis patients. 42 In cirrhosis patients, in addition to serum EV-RNA, the EV surface antigen has been investigated for its diagnostic potential of HCC and CCA. 74 Although an increase in serum EV surface antigens was found in HCC and CCA patients, 74 the results indicated that EV surface antigens could not be used to differentiate HCC from CCA patients. 74 In HCC patients, another report using tumorspecific transcript-1 (TST-1) demonstrated its very high specificity (100%) for HCC detection but low sensitivity (28%). 60 These reports are summarized in Table 3.

INTERVENTION FOR HCC: EVIDENCE FROM IN VITRO REPORTS
EVs can be loaded with therapeutic cargo such as miRNA enabling transference into the target tumor cell. EVs may therefore be used as a personalized cancer treatment. 75 There are several in vitro studies using EVs for HCC cell treatment.  80 In another report, sodium iodide symporter (NIS) genes were transfected to donor HCC cells and cocultured with recipient HCC cells. 83 The NIS protein increased the I-131 toxicity in recipient HCC cells, thus allowing the HCC cells to be susceptible to I-131 ablation. 83 Most studies use EVs as a therapeutic cargo. [78][79][80][81] In one study, EVs from tumor cells were used to activate bone marrow stem cells. 82 Then, these activated bone marrow stem cells were cocultured with tumor cells. The results showed that tumor proliferation was significantly decreased. This method is known as the "Exosome-based vaccine". 82 All these in vitro reports are summarized in Table 4.

INTERVENTION FOR HCC: EVIDENCE FROM IN VIVO REPORTS
In this context, most in vivo studies were done in an immunocompromised mouse model such as NOD SCID mice or nude mice in which an HCC xenograft tumor was subcutaneously implanted to these mice (Table 5). Therapeutic molecules which had antitumoral effects from in vitro studies were endogenously loaded into EVs. The route for EVs treatment introduction was mainly either intratumoral injections or intravenous injections via tail veins. The therapeutic EVs resulted in decreased tumor sizes, 32,[78][79][80][84][85][86][87] increased tumor cell apoptosis, 32,78-80 and increased chemoresistance. 79,84 In addition to using EVs as a therapeutic cargo, indirect use regarding the efficacy of EVs for HCC treatment has been reported. Intraperitoneal injection of propofol was shown to stimulate tumor-associated macrophages to produce miR-142-3p EVs, which was associated with decreased tumor growth. 86 In one study, an exosome-based tumor vaccine (dendritic cells stimulated by tumor EVs) was used to treat xenograft HCC in mice. 87 The result showed that intravenous injection of exosome-based tumor vaccine effectively decreased xenograft tumor volume. 87 All these in vivo studies demonstrated the benefit of EVs as a potential tool for HCC treatment including the use of EVs as a therapeutic cargo or a tumor vaccine since EV could transfer therapeutic molecules into the xenograft HCC, leading to decreased tumor proliferation. These reports are summarized in Table 5.

PROSPECTIVE
HCC is common cancer with a high mortality rate. Clinical outcomes of HCC patients have been improved in the past decade due to early tumor detection and the availability of multiple treatment modalities. Currently, although HCC surveillance using ultrasound and serum AFP, imaging-based diagnosis (CT, MRI), and the new therapeutic methods such as local ablation and transarterial chemoembolization have been shown to be beneficial in increasing survival time, recent data show that the sensitivity of HCC surveillance is still low (about 60%) and that some HCC patients had impaired liver function contradictory to undergoing TACE. 11 Accumulating evidence shows that HCC cells can use EVs for cell-to-cell communication to promote their growth. Clinical studies have already demonstrated that HCC derived-EV containing RNA cargo can be potential serum biomarkers which may help in the diagnosis of HCC. Some of this EV-RNA is also associated with tumor burden, indicating that it might be used as a prognostic indicator. In addition, reports from in vitro and in vivo studies indicated that EVs could be used to decrease tumor growth and tumor size. A summary of potential roles of EVs is shown in Figure 1.
This increasing evidence suggests the potential translational use of EVs in HCC patients in the near future. In the case of diagnostic roles, circulating EVs might be used as a serum biomarker for HCC or as an added value for the diagnosis of liver mass with uncertain imaging. Correlating the EVs with tumor staging, tumor size, and metastasis may help to determine tumor burden and prognosis. However, as regards the therapeutic role, there are no clinical reports available at this time. Further studies are needed to discover the best way to use EV for HCC patient care. Investigations aiming to deliver therapeutic EVs via endovascular methods such as using transarterial chemoembolization (TACE) will also provide important information and warrants further clinical investigation.

ETHICS STATEMENT
Not applicable.

F I G U R E 1
The roles of HCC-derived EVs, HCC-derived EVs as a biomarker, and the evidence of therapeutic EVs from currently available reports. HCC cells secrete EVs that can lead to increased tumor cell proliferation, migration, chemoresistance, and decreased tumor cell apoptosis. They can also affect tumor microenvironments such as increased angiogenesis. Some of these HCC-derived EVs can be detected in circulation, making them available for use as a diagnostic biomarker. Moreover, it is possible for EVs to be used as a therapeutic cargo to transfer therapeutic molecules into tumor cells. Several in vitro and in vivo reports have demonstrated antitumoral effects using this method. miR: microRNA; Circ: circular RNA; siRNA: signal interference RNA; ANGPT2: angiopoietin2; linc: long interceding/intergenic noncoding RNA; TUC: tumor ultraconserved RNA.