Going beyond VEGF pathway inhibition for antiangiogenic cancer therapy: Is inhibition of the PP2A/B55α complex the answer?

In healthy organs, balance between pro-and anti-angiogenic signaling ensures that blood vessels lined with endothelial cells (EC) and surrounded by pericytes ensconced in vascular basement membrane are structurally normal and functionally optimal. In tumors, this balance is tipped towards pro-angiogenic signaling, due to overexpression of factors such as vascular endothelial growth factor (VEGF), and the vasculature becomes structurally and functionally abnormal thereby impairing oxygen delivery and fueling disease progression and treatment resistance ¹. Indeed, angiogenesis and hypoxia are negative prognostic biomarkers ².

Restoring the balance of pro-and anti-angiogenic signaling in cancer by inhibiting VEGF signaling to induce vascular normalization is a validated modality of cancer treatment that can enhance the efficacy of cytotoxic therapies and immunotherapies ^1-3. However, the benefits of anti-VEGF drugs are usually transient. In this issue of Circulation Research, Ehling et al. report a new strategy that could be complementary to current angiogenesis therapies (AATs) – targeting the PP2A-phosphatase subunit B55α, which regulates HIF/PHD2 oxygen-sensing machinery ⁴. Inhibition of the PP2A/B55α complex restores PHD2 function in tumor ECs enabling apoptosis in response to cellular stress thereby pruning nascent/immature vessels while sparing quiescent/mature vasculature ⁴.

Because angiogenesis accompanies disease progression, the initial therapeutic approaches were directed at blocking VEGF to limit tumor growth by “starving” them. This strategy limits tumor growth by extensively pruning the tumor vasculature, while increasing hypoxia. In contrast, “normalizing” doses of anti-VEGF agents create a time window of normalization by actively recruiting pericytes via Ang1/Tie2 signaling to fortify blood vessels and passively pruning immature vessels resulting in increased oxygenation ⁵. The window ends when the remaining vessels are insufficient to supply adequate amount of oxygen to the entire tumor ^{1-3, 5}. Therefore, fortifying immature tumor vessels to improve their function while avoiding excessive pruning can extend the window of normalization and increase oxygenation ^{6, 7}. A fortified vasculature and ameliorated hypoxia inhibit key steps of the metastatic cascade including reducing the shedding of cancer cells into the circulation ¹. Indeed, markers of functional normalization are emerging as potential biomarkers of response to antiangiogenic therapy (AAT) ^7-11. Thus, the balance of vessel pruning (characterizing passive component of vascular normalization) and fortification (characterizing active component of normalization) must be carefully controlled to alleviate hypoxia when AAT is used as a monotherapy or with cytotoxic therapies ^{2, 6, 8}.

Several challenges remain for more effective use of VEGF inhibitors, such as dose and timing, in extending the window of normalization. First, chronic use of VEGF inhibitors causes toxicity by affecting some vessels in healthy organs while inducing hypertension ¹. Second, extrinsic resistance to VEGF inhibitors after pruning-induced hypoxia can be caused by promotion of alternative angiogenic signaling pathways thereby allowing vascular regrowth ¹. Third, in some cancers the vessels are already collapsed by compressive forces generated by tumor growth, which could be responsible for impaired blood flow and hypoxia; currently approved AAT are not able to fully overcome this challenge ^{1, 2}.

Ehling et al. demonstrate that PP2A inhibitors possess features that could make them superior to and/or combinable with VEGF inhibitors ⁴. The PP2A-phosphatase subunit B55α stabilizes ECs when encountered with cell stress (e.g., the onset of blood flow) thereby protecting ECs in vessels undergoing remodeling from apoptosis and promoting pathological angiogenesis. In contrast, VEGF inhibitors target proliferative ECs, making them more likely to induce systemic toxicity and regression of more mature tumor vessels. Thus, PP2A inhibition could have fewer effects on healthy organs and mature intratumor vessels than VEGF inhibitors.

Ehling et al.’s work highlights the antiangiogenic and antitumor effects of PP2A inhibition in AAT-resistant murine breast and lung cancer models of spontaneous metastasis (Figure 1A) ⁴. AAT with VEGF pathway inhibitors alone has failed to show benefit in breast and lung cancer patients. PP2A inhibition as a monotherapy for human disease may face similar challenge. Nonetheless, as a vascular disruption strategy, PP2A inhibition in combination with VEGF inhibitors may be worth exploring for two reasons. First, PP2A inhibition might avoid VEGF-inhibitor induced resistance caused by activation of redundant angiogenic signaling pathways. Second, PP2A inhibition might allow lowering the dose of VEGF inhibitors to reduce toxicity while enabling a larger anti-vascular effect.

Figure 1 – Potential strategies for using PP2A inhibition for cancer treatment.

(A) In Ehling et al. ⁴, PP2A inhibition prunes vessels leading to vascular regression, hypoxia and prolonged survival in murine models of AAT-resistant breast and lung cancer. (B) Normalizing doses of anti-VEGF therapy induce recruitment of pericytes to vessels via Ang1/Tie2 signaling resulting in a normalization window of reduced hypoxia and metastasis leading to increased survival. However, anti-VEGF agents also passively prune immature vessels. Hence, the window of normalization ends when the remaining vessels are insufficient in number to supply adequate amount of oxygen to the entire tumor. Once the normalization window is closed hypoxia begins to increase again. (C) Future work is necessary to determine if and how AAT combined with PP2A inhibition could increase the window of normalization (left arrow). Following vessel decompression, PP2A inhibition could induce normalization leading to reduced hypoxia and increased efficacy of cytotoxics and immunotherapy (right arrows).

Vascular pruning-induced hypoxia is known to cause tumor progression and treatment resistance (Figure 1B) ^{1, 2}. While targeting PP2A might be an effective adjuvant strategy to reduce the growth of metastases after surgical removal of the primary tumor, resistance to angiogenesis inhibitors often occurs through vessel co-option especially in metastasis ^{12, 13}. If metastatic lesions become more hypoxic through PP2A inhibition-induced hypoxia, they could become more aggressive and seed new metastases. Of note, AAT with VEGF/VEGFR2 blockers has failed in adjuvant settings to prevent or delay metastasis, including in breast cancer ². This and compelling pre-clinical data indicate that vascularization of metastases could be independent of angiogenesis and instead mediated by co-option ¹³. Advancing this concept will benefit from demonstration that metastases are VEGF-independent but PP2A-dependent for new vessel formation.

Our interpretation is that inhibition of PP2A, while potentially effective as a monotherapy in anti-VEGF sensitive tumors, will have to be carefully evaluated in combination with cytotoxic, targeted and/or immunotherapy to increase survival of patients. Future work should investigate what role PP2A inhibition plays in active vascular normalization ⁴. These studies need to compare vascular morphology and function after PP2A inhibition and VEGF inhibition to differentiate these strategies and guide combination (Figure 1C). In certain tumor types and at certain doses, PP2A inhibition could enhance perfusion by reducing redundant blood flow into nascent, inefficient vessels. When combining PP2A with VEGF inhibitors in anti-VEGF-resistant tumors, we postulate that the most effective sequence will be PP2A inhibitors following low-dose non-pruning AAT such that AAT fortifies immature vessels before PP2A inhibition prunes the remaining nascent vasculature. PP2A inhibitors combined simultaneously with VEGF inhibitors could cause excessive pruning by targeting various tumor vessels simultaneously through complementary mechanisms. Administering VEGF inhibitors after PP2A inhibitors seems counterproductive, because AAT could prune vessels that PP2A inhibitors spared. Irrespective of approach, the combination would benefit from predictive biomarkers. While active normalization is preferable in poorly perfused cancers, passive normalization can be effective in patients with high baseline vessel density such that the remaining fortified vasculature is sufficient ^{1, 2, 8}.

Finally, like VEGF inhibitors, PP2A inhibition will likely be ineffective in tumors with many collapsed vessels. In these highly desmoplastic tumors, physical forces generated within tumors compress fragile vessels ^{1, 2}. Administering PP2A inhibiting agents will not overcome compression and their effect will be limited to the few perfused vessels, further aggravating hypoxia. Thus, alleviating compression to increase perfusion and the number of ECs targeted by subsequent administration of PP2A could further enhance perfusion than either strategy alone (Figure 1C) ¹⁴. While AAT and agents causing vessel “decompression” could be combined simultaneously ¹⁵, PP2A inhibitors would cause ECs in newly decompressed, re-perfused vessels to undergo apoptosis. Additionally, breast cancer patients with a higher density of non-compressed vessels responded better to VEGF inhibition ^{2, 8}, providing further rationale to administer vessel decompressing agents before PP2A inhibitors.

PP2A inhibition could provide further insights into the mechanism of benefit involving passive and active normalization of tumor vessels in patients. Indeed, the PP2A-inhibitor LB100 is currently in clinical trials. While vascular disruption has failed in clinical trials, PP2A inhibition’s complementary mechanism of action and toxicity profile suggests that enhanced normalization could be achieved depending on the tumor type, dose and schedule of combinations of PP2A inhibitor and other AATs. In terms of reducing hypoxia, PP2A inhibition might be most useful following vessel decompression to avoid pruning of reperfused vessels. Given the importance of alleviating hypoxia with AAT to anti-tumor immunity ⁷ and that LB100 enhances infiltration/activation of CD8+ T cells while inhibiting immunosuppression ¹⁶, Ehling et al. provide a foundation to develop strategies combining PP2A inhibition with other therapies that may improve vessel function towards enhancing immunotherapies.