Blood markers are needed to help guide anti-angiogenic therapy. miRs are emerging as important biomarkers, and several miRs have been implicated in regulating tumor angiogenesis . Whether measurement of these miRs in the circulation may be useful clinically has not been established. We found that plasma levels of specific angioregulatory miRs may have utility. There was a positive correlation between levels of the pro-angiogenic miR-199a and a negative correlation between the anti-angiogenic miR-106a with CEC, both moderate, in patients with uveal melanoma prior to receiving systemic adjuvant therapy. miR-199a promotes the proliferation of endothelial cells though effects on caveolin-2 . miR-106a, a paralog of miRs of the miR-17-92 cluster, is upregulated during hypoxia and is predicted to target VEGF [24, 25]. With treatment with sequential dacarbazine and interferon-alfa-2b, drugs with anti-angiogenic activities, decreases in miR-199a and in another pro-angiogenic miR, miR-126, and increases in miR-106 and another anti-angiogenic miR, miR-16, were observed. miR-126, an endothelial cell-restricted miR, regulates vascular integrity and angiogenesis. It enhances the actions of VEGF and bFGF by repressing the Spred-1, an inhibitor of angiogenic signaling . miR-16 is implicated in suppressing VEGF and bFGF . miR-16 and 199a have been shown to be produced by human endothelial cells . miR-106a and 199a, as well as miR-146, were among seven miRs found to be increased in the neo-vascularization response to ischemia within the eye in a mouse model .
Correlations and changes were not observed in other angioregulatory miRs. These included miR-20a, a pro-angiogenic miR in the 17–92 cluster that represses thrombospondin-1 and connective tissue growth factor ; miR-155, which is pro-angiogenic through effects on angiotensin signaling , miR-125b, which is anti-angiogenic through effects on placenta growth factor ; miR-146a, which is anti-angiogenic though effects of NFκB and suppression of IL-8 and epidermal growth factor receptor signaling ; and miR-221, which impairs stem-cell-factor-induced angiogenesis . miR-20a and 221 have also been shown to be produced by human endothelial cells . The relationship between a miR and a target is not exclusive, and depending on what genes are suppressed, a given miR can have either a positive or negative role in the regulation of angiogenesis. Although a member of the pro-angiogenic miR-17-92 cluster, miR-20a can also suppress angiogenesis by suppressing VEGF . miR-199a also affects hypoxia-inducible factor 1  and is predicted to target VEGF . Furthermore, all the miRs tested target multiple genes not directly or indirectly involved in angiogenesis, and several miRs, such as miR-20a and 106a, have been reported to be upregulated in uveal melanoma .
Correlations were not observed between levels of miRs and levels of VEGF, bFGF, and IL-8, angiogenic proteins that have been shown to be produced by uveal melanoma . No patient had clinical evidence of cancer at any time during the assessments. In lung cancer, tumor miR-20 was significantly associated with tumor VEGF, and tumor miR-155, with tumor bFGF [38, 39]. We also did not observe consistent effects of treatment with dacarbazine or interferon-alfa-2b on the levels of angiogenic proteins. Although changes in the blood levels of angiogenic proteins have been observed in patients administered a variety of anti-angiogenic treatments, they have not been consistent . In one study in patients with cutaneous melanoma administered interferon-alfa-2b, IL-8 levels increased while VEGF and bFGF did not change ; in another, levels of VEGF decreased, while levels of bFGF and IL-8 did not change .
Significant changes in levels of miRs and also of CEC were observed after treatment with interferon-alfa-2b, but not after treatment with dacarbazine. In contrast to interferon-alfa-2b, the dose and scheduling of dacarbazine were likely not optimal to demonstrate angioregulatory effects, more usually achieved with more repetitive, i.e., “metronomic,” dosing . The increases in CEC were most apparent when assayed at the week 17 time point, after 8 weeks of interferon, and persisted. The increases in pro-angiogenic and the decreases in anti-angiogenic miRs observed also were most apparent at week 17, but were transient. Interferon-α, which has immune modulatory, antiproliferative, and antiviral effects, has been shown to increase specific miRs by melanoma cells . Whether angioregulatory miRs are modulated has not been reported. Increases in several miRs were observed in patients with chronic hepatitis C virus infection treated with pegylated interferon-alfa-2b but did not correlate with viral load or liver function tests .
Whether blood levels of the miRs studied as well other miRs may be useful in monitoring anti-angiogenic therapy merits further investigation. The relative prognostic and predictive value of blood miRs compared to CEC, measurement of which has yielded conflicting results, also merits further evaluation. Of note, levels of angioregulatory miRs during treatment were not significantly correlated with CEC. The clinical situation, the anti-angiogenic approach, and the molecular targets will need to be considered. Given the complexity of the angiogenic process, a combination of several types of biomarkers may be necessary. Studies wherein miR levels and CEC are being assessed in conjunction with liver function tests and imaging studies as part of systemic surveillance for metastases in patients with uveal melanoma are underway.