More than half a million people are diagnosed each year with hepatocellular carcinoma (HCC), a malignant tumor of the liver associated with poor prognosis.1 Major risk factors for HCC include chronic infection by hepatitis B virus (HBV) or hepatitis C virus (HCV) and alcoholic liver cirrhosis. Although most HCC patients present with advanced and symptomatic disease not amenable to curative surgery, screening programs for high-risk populations have increased early detection and effective surgical treatment of HCC.1 Although surveillance of high-risk patients may be pursued by periodic ultrasonography of the liver, a definitive diagnosis of HCC can be made only based on concordant findings from liver biopsy, serum alpha-fetoprotein (AFP) levels, computed tomography, or magnetic resonance imaging.1 However, early-stage HCC is difficult to detect by noninvasive imaging, and AFP as a “surveillance biomarker” has been dropped in current guidelines because of low sensitivity and specificity.2 Thus, novel biomarkers for the early detection of HCC are greatly needed. In the current issue of HEPATOLOGY, Matsubara et al.3 report on the significance of circulating TIE2-expressing monocytes (TEMs) as biomarkers for the detection of both early- and late-stage HCC.
Different circulating bone marrow (BM)-derived cell (BMDC) types have been proposed as cancer biomarkers with diagnostic and/or prognostic value. Among BMDCs, CD133+/vascular endothelial growth factor receptor 2-positive (VEGFR2+) circulating endothelial progenitors (CEPs) were reported to have both diagnostic and prognostic value in HCC.4 CEP levels—inferred from the frequency of early-colony-forming units in ex vivo cultures of blood-derived mononuclear cells—were significantly higher in patients with HCC, compared to patients with cirrhosis and healthy controls, and positively correlated with serum AFP levels. Furthermore, patients with advanced HCC had higher CEP levels than patients with resectable tumors, and higher preoperative CEP levels were associated with higher recurrence rates.4 More recently, CEPs were found to predict HCC response to sorafenib (a multitarget small-molecule inhibitor approved for first-line treatment of advanced HCC) plus chemotherapy, with higher CEP levels at baseline correlating with worse progression-free and overall survival.5 Although CD133+VEGFR2+ CEPs likely represent rare circulating hematopoietic progenitors and not bona fide endothelial-lineage cells,6 the aforementioned clinical data support the potential of CEPs as biomarkers in HCC patients.4,5 Yet, it is not clear whether CEP levels are also increased in patients with very-early-stage HCC, thus inclusion of CEP measurements in surveillance protocols for high-risk patients may be warranted, along with sufficiently powered prospective trials that more clearly assess the diagnostic value of these cells.
Myeloid-derived suppressor cells (MDSCs) are another BMDC population whose diagnostic and prognostic significance has been recently investigated in HCC patients. MDSCs, which comprise both immature and mature elements of the mononuclear and polymorphonuclear myeloid lineages, are often expanded in the blood and lymphoid organs of cancer patients and are thought to promote tumor progression primarily by suppressing antitumor immunity and promoting tumor angiogenesis.7 Hoechst et al. found that the frequency of mononuclear CD14+/human leukocyte antigen/DR–/low (HLA-DR–/low) MDSCs, but not total CD14+ monocytes, was significantly increased in the blood of HCC patients, compared to healthy controls.8 CD14+HLA-DR–/low MDSCs were found to stimulate the expansion of immunosuppressive CD4+CD25+FOXP3+ regulatory T cells8 and to inhibit the activity of natural killer (NK) cells9 when cocultured ex vivo with T or NK cells, respectively. These data highlight the clinical significance of mononuclear MDSCs as a diagnostic biomarker of HCC and provide evidence for their involvement in immunosuppression associated with HCC progression.
HCC is a highly vascularized tumor, and the progression of early lesions to overt HCC is associated with the appearance of a prominent arterial blood supply.10,11 The antiangiogenic drug, sorafenib, which targets several kinases that may control VEGF-mediated angiogenesis, prolongs survival of patients with advanced HCC, likely reflecting HCC dependence on angiogenesis.10 Of note, several preclinical studies have shown that both CEPs and MDSCs (or subsets of these cells) contribute to tumor angiogenesis in mouse models of cancer.7,12 Therefore, increased CEP and MDSC numbers in the blood of HCC patients4,5,8 may also have the potential to foster HCC angiogenesis and progression to increased malignancy.
TEMs are another BMDC type with proangiogenic activity in mouse models of cancer.13 In humans, TIE2 expression is higher in nonclassical (CD14lowCD16+) than classical (CD14+CD16–) monocytes14–16 and identifies a subset of circulating monocytes endowed with proangiogenic activity.14 Infiltrating TEMs are also detected in a variety of human cancers.14,17
Matsubara et al.3 report on the frequency of circulating TEMs in a cohort of 168 HCV-infected patients, of which 89 were with HCC. The investigators found that the frequency of TEMs was significantly higher in patients with HCC than HCV-infected patients without HCC or healthy subjects, thus establishing TEMs as a stage-independent, diagnostic biomarker for HCC. Although TEM frequency did not correlate with tumor stage, HCC patients with higher TEM levels in their blood had a worse recurrence-free survival after HCC resection or radiofrequency ablation (RFA) therapy than patients with lower TEM levels, suggesting that TEMs may also represent a prognostic biomarker for this tumor type. Interestingly, circulating TEM levels declined in those patients who underwent HCC resection or RFA, possibly suggesting that yet unidentified tumor-secreted factor(s) regulate TEM mobilization from the BM and/or enhance TIE2 expression in nonclassical monocytes.3 Of note, a recent study documented significantly enhanced TIE2 expression in the circulating monocytes of colorectal cancer patients, compared to healthy subjects.17
Matsubara et al.3 also identified TEMs in HCC specimens and observed that these cells preferentially localize in perivascular tumor areas, in agreement with findings in mouse models of cancer.13 Furthermore, it was found that a higher TEM infiltration correlated with increased microvessel density in the tumors, possibly suggesting that HCC-infiltrating TEMs are proangiogenic.
Although the biological significance of the findings of Matsubara et al.3 need to be investigated in ad-hoc mouse models of hepatocellular carcinogenesis, the current study is the first to present evidence suggesting that circulating TEMs may be a diagnostic biomarker for both early- and late-stage HCC. Future studies should address several important issues raised by these observations.3 According to Matsubara et al.,3 high circulating and intratumoral TEM levels correlate with a more-advanced Child-Pugh stage, a finding that may suggest that TEM frequency correlates positively with the degree of liver inflammation/stage of cirrhosis and negatively with liver function. In this regard—and contrary to the findings of Matsubara et al.3— a recent study showed that circulating and intrahepatic TEMs are significantly increased in HCV-infected patients without HCC, compared to healthy subjects.18 In that study, HCV patients who responded to antiviral therapy had significantly lower TEM levels than naïve (untreated) or nonresponder patients.18 These interesting findings suggest that chronic liver inflammation may be a stimulus for TEM mobilization from the BM, their differentiation/expansion in the periphery, and/or the up-regulation of TIE2 in nonclassical monocytes. Although Rodriguez-Munoz et al.18 analyzed a relatively small cohort of HCV-infected patients, their data raise the concern that mobilization/expansion of TEMs may not be strictly HCC driven, but more generally associated with chronic liver infection.
Virtually nothing is known about the biology underlying TEM's involvement in human tumor angiogenesis and progression. What tumor-derived factors, if any, increase TEM frequency (or their expression of TIE2) in the peripheral blood and tumors of HCC patients? What factors tether extravasated TEMs to the tumor blood vessels? Do TEMs—similar to mononuclear MDSCs—exert immunosuppressive functions that foster HCC progression? Do TEMs promote HCC angiogenesis? Is the positive correlation between TEM infiltration and microvessel density truly indicative of a proangiogenic function of these cells? Furthermore, does sorafenib inhibit HCC angiogenesis also by limiting TEM recruitment to the tumors and/or impairing their proangiogenic functions? And, in turn, do TEMs express proangiogenic programs in HCC that limit the efficacy of sorafenib or other antiangiogenic treatments? Interestingly, elevated circulating levels of the TIE2 ligand, angiopoietin-2 (ANG2), correlate with a worse prognosis in untreated HCC patients and predict shorter survival in HCC patients treated with sorafenib.19 It is tempting to speculate that tumor-derived ANG2 and engagement of the TIE2 receptor on endothelial cells and monocytes may promote angiogenesis in HCC and, possibly, also limit HCC sensitivity to sorafenib. In mouse models of mammary carcinogenesis, systemic ANG2 neutralization or Tie2 gene knockdown in TEMs inhibited tumor angiogenesis, suggesting that the ANG2/TIE2 axis modulates the proangiogenic activity of TEMs, at least in mouse models of cancer.20
In summary, Matsubara et al.3 report interesting new findings that provide further evidence that BMDCs may serve as biomarkers for HCC. Furthermore, the data also suggest that BMDCs could be involved in the pathogenesis of HCC. Indeed, CEPs,4 mononuclear MDSCs,8 and TEMs3 may all have the potential to regulate HCC angiogenesis and progression, possibly by releasing proangiogenic growth factors or molecules that blunt the endothelial- or cancer-cell killing activity of cytotoxic T cells.7 Thus, inhibiting the proangiogenic and/or immunosuppressive functions of these BMDCs may represent a promising strategy to improve the efficacy of current treatments for HCC.