Tanshinone ⅡA Inhibits Metastasis After Palliative Resection Of Hepatocellular Carcinoma And Prolongs Survival Via Vascular Normalization

Liver cancer (mainly hepatocellular carcinoma) is the sixth most frequently diagnosed cancer worldwide but the third leading cause of cancer death. Surgical resection is the most promising strategy for hepatocellular carcinoma (HCC) in early stage. However, there is as high as70%of5-year risk of recurrence. Most of surgical resection is actually palliative resection (PR) owing to the existence of satellites and microvascular invasion. These residual tumor nests could be stimulated to reanimation by PR. Although several ways, such as interferon-alpha and sorafenib, have been proposed to diminish relapse, unfortunately, the pro-metastatic side-effects of these options were also noted.Residual tumor cells may increase the release of proangiogenic factors and decrease antiangiogenic factors after PR. Meanwhile, operative injury would induce recruitment of macrophages and bone-marrow derived endothelial progenitor cells around and infiltrating into residual tumor, and instigating angiogenesis. Residual tumor growth needs angiogenesis, however, this neovessels may be disorderly structured and inefficiently perfused, thereby creating hypoxia. Both abnormal endothelium and pericytes integrated into the capillary wall along with deficient coverage could be responsible for the vascular structure abnormalities. The resulting hypoxia creates a hostile tumor milieu from where tumor cells intra/extravade through a leaky and immature vessel. Van der Bilt and colleagues demonstrated that surgery-induced hypoxia unfavorably affected the prognosis of cancer patients by inducing angiogenesis. Therefore, restoring oxygen supply via vascular normalization may reduce metastasis, even tumor growth. In addition, remodeling vascular structure could mechanically prevent tumor cells invasion and intra/extravasation. Mazzone et al. showed that downregulating oxygen-sensing prolyl hydroxylase domain-2protein could restore tumor oxygenation and inhibit metastasis via endothelial normalization. This was because endothelial cells changed their shape and formed a phalanx to block metastasis. Accordingly, to explore agent that has a potential role of vascular normalization and easily available is of clinical significance. Rolny et al. found that a host-produced histidine-rich glycoprotein could inhibit tumor growth and metastasis by inducing vascular normalization. Although several ways has been showed to promote vascular remodeling, seldom was available for clinical usage.Tanshinone ⅡA (Tan ⅡA), which is widely used in patients with cardiovascular or cerebrovascular diseases, is an herbal monomer isolated from Salvia miltiorrhiza. In Chinese traditional medicine, Salvia miltiorrhiza is considered to promote blood circulation for removing blood stasis and improving microcirculation, which we speculated might include vascular normalization. We have shown an herbal formula "Songyou Yin" could attenuate HCC metastases that caused by PR or chemotherapy, and Salvia miltiorrhiza is one of the five constituents of "Songyou Yin", while the mechanism was indefinite. Tan ⅡA owned direct vasoactivity and a certain anti-tumor effects. It is still enigmatic if PR induces tumor vessel abnormality and Tan ⅡA promotes vascular normalization. Whether Tan ⅡA decreases residual tumor metastasis indirectly through vascular normalization has not been reported.We aimed to identify the inhibitory effects of Tan ⅡA on HCC metastasis after PR, and its effects on survival, with mainly focus on tumor vessel maturity as a potential marker for evaluating Tan IIA treatment responses and the possible mechanism.1. Establishment of the residual tumor nude mice model after palliative resection (PR) of hepatocellular carcinomaThe orthotopic inoculation of double-tumor in the same liver lobe was used to establish the MHCC97H orthotopic xenografts nude mice model. Then PR was performed to resect one tumor, mice were scheduled to observe5weeks. The results showed that the double-tumor formation rate, success rate of PR, and mice overall survival were95.6%,95.3%(operative mortality4.7%), and90.2%, which was significantly better than PR in single-tumor inoculation model; the tumor formation rate, operative mortality, and overall survival for the latter were100%,73.3%, and20%. Therefore, the residual tumor model was successfully established.2. Palliative resection (PR) instigated residual tumor growth and metastasisThe tumor volume of PR group was larger than Sham operation (Sham) group (P <0.05). Compared with Sham group, the lung metastasis (HCCLM3) of PR group significantly increased (110.11±28.01vs7.22±4.71, P<0.001); both the intrahepatic and abdomen metastasis also increased (P<0.05); and the circulating tumor cells elevated both in2and35d postresection (P<0.001).3. Tanshinone ⅡA (Tan ⅡA) did not directly inhibit tumor growth but reduced metastasis, and prolonged survival of tumor-bearing nude miceIn single-tumor model, there was no statistical significance among the tumor volume of normal saline (NS), Tan IIA1,5,10mg/kg/day groups. Compared with NS, the lung metastasis of Tan IIA5or10mg/kg/day group significantly decreased (P=0.046and P<0.001). Compared with Tan IIA1or5mg/kg/day, the lung metastasis of10mg/kg/day group also decreased (P=0.003and P=0.046). Both the intrahepatic metastasis rate and abdomen metastasis of Tan IIA5,10mg/kg/day groups significantly reduced (P<0.05). Abdomen metastasis of HepG2had not been found. The inhibitory effects of dose cohort10mg/kg/day was the best, thus was chosen as the intervention dosage in the next in vivo experiment.4. Tan ⅡA significantly inhibited the enhanced residual tumor metastasis after PR, and prolonged survival.Administration of Tan IIA10mg/kg/day after PR got a decreased manner in residual tumor volume compared with NS (P<0.05). Compared with PR+NS group, the lung metastasis (HCCLM3) of Tan IIA group significantly decreased (13.56±11.33vs94.56±31.47, P<0.001); both the intrahepatic and abdomen metastasis also decreased (P<0.01), and the circulating tumor cells relatively decreased (P<0.001). Most importantly, Tan IIA dramatically retarded the weight loss of mice after PR. The estimated survival of PR mice was significantly shorter than Sham mice (P<0.01). In single-tumor model, Tan IIA could prolong mice’s survival up to16d for HCCLM3(P=0.004) and19d for HepG2(P<0.001) compared with NS. Furthermore, it could prolong the declined survival after PR in double-tumor model (HCCLM3:47.833±3.280vs69.000±1.693; HepG2:52.167±2.496vs79.667±2.940, both P=0.001).5. Tan ⅡA did not inhibit proliferation while blocked invasion of tumor cellsCompared with0.1%DMSO, the OD value of Tan IIA0.01-100μM groups had no change; the number of invasive cells for Tan IIA5and10μM groups significantly reduced (P<0.01); there was no statistical difference for1μM group. 6. Tan IIA significantly decreased residual tumor hypoxia in vivo but did not downregulate HIF-la of tumor cells under hypoxia in vitroThe Pimonidazole and HIF-1α significantly increased after PR, and could be reduced by Tan IIA; consistently, the residual tumor epithelial-mesenchymal transition (EMT) enhanced (N-cadherin and Vimentin were upregulated, E-cadherin was downregulated), and could also be reversed by Tan IIA. These results indicated that PR aggravated residual tumor hypoxia and promoted EMT, and Tan IIA could alleviate hypoxia and prohibit EMT in vivo. We then found HIF-la, N-cadherin, and Vimentin of tumor cells were upregulated, and E-cadherin was downregulated under hypoxia; while except E-cadherin was upregulated, other molecules had not been downregulated as anticipated by Tan IIA in vitro. We further observed E-cadherin expression divorced from hypoxia and found it could indeed be upregulated by Tan IIA. In whole experiment, protein of all molecules was consistent with their corresponding mRNA.7. Tan IIA did not affect microvascular density (MVD) but significantly promoted microvascular integrity (MVI) of residual tumorThe MVD of mice was indicated by CD31. NG2, as a major marker of vascular pericyte, was adopted to evaluate MVI. The CD31of PR group was higher than Sham group (P<0.001), and there was no statistical significance between PR+NS and Tan IIA groups. Although the NG2had no change after PR, its level significantly elevated in PR+Tan IIA group (P<0.01). Immunohistochemistry of CD31, NG2, and Pimonidazole in serial sections showed that in PR+NS group, CD31was high, NG2was low, and hypoxia was serious (Pimonidazole was high). In PR+Tan IIA group, CD31was high, NG2was also high, thus hypoxia was slight. These results indicated the MVD of residual tumor increased after PR, but the MVI was still low. Tan IIA did not inhibit MVD but markedly improved MVI, promoting integrity of tumor vessels. Scanning electron microscopy of tumor vessels further revealed the vascular wall of PR+Tan IIA tumor was more integrated than NS tumor. Immunofluorescent results also confirmed Tan IIA did not change CD31, but elevated NG2, thus decreased hypoxia.8. Tan IIA significantly enhanced tube formation of endothelial cells (HUVECs, Human and Mice TECs) and closely associated with VEGFR1and PDGFR upregulationThe tube formation of HUVECs and human TECs cultured in both normoxia and hypoxia was enhanced by Tan IIA. And the tube formation of mice TECs from PR+Tan IIA (in vivo) group was much stronger than PR group. Using TECs from PR group, we further found the tube formation was significantly enhanced by Tan IIA in vitro, and was roughly equivalent to PR+Tan IIA (in vivo) group. Then the flow cytometric analysis indicated both the percentage of mice TECs bearing VEGFRl and PDGFR and the relative fluorescence intensity of the two receptors for PR+Tan IIA group were significantly higher than NS group (P<0.05), and there was no change to VEGFR2, EGFR, and FGFR1. Using a VEGFR1/PDGFR inhibitor SU6668, we found that the promotion of tube formation by Tan IIA was dramatically weakened; while SU6668did not appear to inhibit tube formation without Tan IIA.9. Combination of Tan ⅡA and interferon-alpha leaded to dose reduced one third for the latterBased on the above findings about Tan IIA, we continued to study the differences of efficacy between delivery with large dose of IFN-a (1.5×107U/kg/day) and combination of IFN-a dose reduced one third (1.0×107U/kg/next day delivery) plus Tan IIA (10mg/kg/day). The results showed that the tumor volume of IFN-a single group and combination group were0.391±0.411cm3and0.756±0.349cm3, no statistical significance was found between them (P=0.282), and both were smaller than3.160±0.405cm3of NS group (P<0.001). The number of lung metastatic lesion of IFN-a single and combination groups was12.83±6.85and1.50±2.51, the latter was less than the former (P=0.001), and both were less than29.33±6.95of NS group (P<0.001). These results indicated that there was no difference on inhibition of tumor growth between the two groups, the suppression of metastasis for combination with IFN-a dose reduced one third, administration time shortened by half and Tan IIA was stronger than large dose of IFN-a delivered alone.Conclusions:1. Tanshinone IIA can inhibit the enhanced hepatocellular carcinoma metastasis instigated by palliative resection through promoting VEGFR1/PDGFR-upregulation related vascular normalization and prolong survival of nude mice bearing residual tumor.2. Combination of tanshinone IIA and interferon-alpha produces a similar effect between dose reduced one third and administration time shortened by half of the latter and delivery with large dose of interferon-alpha alone.Originalities of this work:1. This work firstly demonstrated that tanshinone IIA could significantly inhibit the enhanced metastasis after palliative resection of hepatocellular carcinoma and prolong survival; the mechanism was associated with promotion of VEGFR1/PDGFR upregulated tumor vascular normalization.2. This work innovatively established a metastatic human orthotopic residual hepatoma xenografts nude mice model with double-tumor simultaneously inoculated into the same liver lobe, and then perfomed palliative resection.Clinical applications and scientific significance:1. This work indicates that tanshinone IIA inhibites the enhanced metastasis after palliative resection of hepatocellular carcinoma and prolongs survival via vascular normalization. It provides a simple and little side effects adjuvant therapy for preventing metastasis and recurrence after surgery; the reduction of interferon-alpha dose leads to patients tolerates the longer-term medication.2. This work provides us an important implication:the malignant phenotype of tumor can be relieved through a dredging way rather than blind eradication. They also highlight the potential of proangiogenic "vessel normalizing" strategies in silencing metastasis and prolonging survival.3. The most effective mice dose10mg/kg/day of tanshinone IIA used in the present study converts into human dose should be1.11mg/kg/day, which contained in0.67-1.33mg/kg/day, and suitable for clinical use.