With a Little Help from my Friends: The Role of the Renal Collateral Circulation in Atherosclerotic Renovascular Disease

Abstract

The collateral circulation can adapt to bypass major arteries with limited flow and serves a crucial protective role in coronary, cerebral, and peripheral arterial disease. Emerging evidence indicates that the renal collateral circulation can similarly adapt and thereby limit kidney ischemia in atherosclerotic renovascular disease. These adaptations predominantly include recruitment of pre-existing micro-vessels for arteriogenesis, with de-novo vessel formation playing a limited role. Yet, adaptations of the renal collateral circulation in renovascular disease are often insufficient to fully compensate for the limited flow within an obstructed renal artery and may be hampered by the severity of obstruction or patient comorbidities. Experimental strategies have attempted to circumvent limitations of collateral formation and improve the prognosis of patients with various ischemic vascular territories. These have included pharmacological approaches such as endothelial growth factors, renin-angiotensin-aldosterone system blockade, and If-channel-blockers, as well as interventions like preconditioning, exercise, enhanced external counter-pulsation, and low-energy shockwave therapy. However, few of these strategies have been implemented in atherosclerotic renovascular disease. This review summarizes current understanding regarding the development of renal collateral circulation in atherosclerotic renovascular disease. Studies are needed to apply lessons learned in other vascular beds in the setting of atherosclerotic renovascular disease in order to develop new treatment regimens for this patient group.

Introduction:

Atherosclerotic renovascular disease (ARVD) involves atheromatous narrowing of the renal arterial lumen, a progressive disease often leading to secondary hypertension and loss of kidney function. Its overall prevalence ranges between 6.8–54.1%, depending on the patient population.^{1, 2} Despite early promise, randomized controlled trials found endovascular revascularization to have limited prognostic significance in the treatment of ARVD, leaving precious few treatment options for most patients with ARVD.^{3, 4} This dearth of options prompts consideration of the therapeutic potential of promoting the peri-renal collateral circulation, a network of endogenous vessels that bypass obstructions in several vascular beds. The human coronary collateral circulation is far more extensive than in most species.^{5, 6} While a preformed collateral circulation exists in the kidney,^7–11 its abundance relative to other species is unknown. Computational modelling has shown that collateral arteries contribute to reductions in resistance and increases in flow distal to an arterial occlusion more than the distal microcirculation, and are considered to be protective in occlusive arterial disease in the heart, brain and peripheral circulations.^{9, 12–15} Less studied in the kidney, this review will examine the development, role, and therapeutic potential of these vessels in ARVD, and discuss lessons learned from studies in other vascular beds.

Collateral-formation in ARVD:

We have recently observed an association between increased severity of ARVD with development of collateral circulation in the stenotic human kidney,¹⁶ but the underlying mechanism is incompletely understood.¹⁷ According to established paradigms, development of collaterals occurs through sprouting and nonsprouting formation of arteries from existing vessels (angiogenesis)¹⁸ or through expansion of pre-existing collateral vessels (arteriogenesis). Angiogenesis plays an important role in kidney disease. We have demonstrated robust angiogenesis in a model of diet-induced hypercholesterolemia, involving increased density and sprouting of intra-renal micro-vessels, likely secondary to inflammation.¹⁹ Angiogenesis also contributes significantly to blood supply in the ischemic rat kidney²⁰, and hallmarks of atheromatous-related ischemia such as oxidative stress and inflammation promote kidney neovascularization, which may imply a comparable role for de-novo vessel formation in collateralization.^{21, 22} However, while small vessels generated through angiogenesis may ameliorate ischemia, they may be unable to fully restore blood flow in the ischemic kidney.²³ Moreover, collaterals to the kidney most commonly originate from the lumbar arteries but can also originate directly from the aorta (Figure 3), intercostal, hypogastric, ovarian, or testicular arteries^{9, 24, 25}, so their length and caliber argue against their generation primarily through angiogenesis.

Figure 3:

Multidetector computed-tomography 3D images acquired in patients with atherosclerotic renovascular disease depicting the renal collateral circulation (green arrow).

Perhaps a more likely origin of observable collaterals in ARVD is development and remodeling of pre-existing radiographically invisible micro-vessels to the kidney into fully fledged collateral arteries through arteriogenesis.^{26, 27} Microvascular collaterals found in most tissues are typically <150μm in diameter and tortuous, often with an unorganized pattern.^{16, 28, 29} These vessels are capable of increasing their diameter 10–20-fold without necessarily losing wall thickness in response to increased blood flow and thereby increased tangential vascular shear-stress caused by occlusion of adjacent major arteries.^30–33 This impact of shear-stress upon arteriogenesis is augmented by reversal of flow in arterioles.³³ According to a prevailing model, reduced blood flow through a major artery due to stenoses causes distal pressure to fall, increasing the pressure gradient over the associated pre-existing collateral circulation. Consequently, collateral artery flow increases and changes from bidirectional to unidirectional, triggering processes that magnify its caliber (Figure 1).^{27, 33}

Figure 1:

Schematic representation of shear-stress-induced collateral formation in renovascular disease. A. Non-atherosclerotic renal artery connected to a feeding artery through micro-vessels with limited bidirectional flow. B. With a recent atherosclerotic occlusion, collaterals in first stage of remodeling increase slightly in size and flow, which becomes unidirectional. C. Collaterals subsequently increase significantly in diameter to bypass the occluded vessel. E. A dominant vessel ultimately accounts for most of the collateral flow, while the other collateral has regressed. Created with Biorender.com.

Endothelial cells have complex sensing mechanisms that enable a change in function in response to a rise in shear-stress.^{26, 34} While still somewhat unclear³⁵, important mechanisms are believed to include ligand-independent activation of vascular endothelial growth-factor (VEGF) receptor-2, ion-channels, G-protein-coupled receptors, the plasma membrane, trimeric G-proteins, adhesion molecules³⁴, endothelial nitric-oxide synthase, endothelin-1²², the glycocalyx,^35–37 and most likely a complex interplay among these.^{35, 38} The processes that follow this activation have been discussed excellently elsewhere.³⁰

The contribution of peri-vascular structures to collateral development cannot be understated. We have shown expansion of the coronary vasa vasorum microcirculation during early atherogenesis³⁹, and these might possibly sprout collateral vessels bypassing a luminal obstruction. Furthermore, the peri-vascular fat tissue releases inflammatory and angiogenic factors as well as vasoactive adipokines, such as adventitia-derived relaxing factor, angiotensin-II, and NO, that may fuel this process.^40–44 Clearly, additional studies are needed to establish the impact of the vascular microenvironment on collateral development.

While current understanding is primarily derived from studies of other vascular beds, evidence obtained in both animals and humans supports a similar pattern of rapid and meaningful proliferation of collateral arteries in the kidney following renovascular arterial obstruction.^{16, 20, 28, 45–47} These processes may lead to development and expansion of several higher-caliber renal collaterals. Most likely, a mature renal collateral circulation with larger and smaller (perhaps radiographically invisible) collaterals develop due to what, according to Poiseuille’s law, is a more energy-efficient remodeling, as some collaterals become dominant while others regress (Figure 1).^{18, 48} While processes leading to collateral formation have been characterized in atherosclerotic vessels, their applicability to non-atherosclerotic vascular disease, like fibromuscular dysplasia (FMD), is unclear. While non-atherosclerotic obstructions likely increase shear stress in collateral micro-vessels, mutations in genes associated with arterial formation in FMD-patients, such as PHACTR1, may influence vascular responses.⁴⁹ Furthermore, regulation of collateral formation by ambient atherogenic factors in ARVD remains to be elucidated.

Clinical and functional significance of the collateral circulation:

Lessons from the coronary collateral circulation:

Studies on the functional role of the collateral circulation have historically been dominated by studies on coronary collaterals. These are generally considered instrumental in patients with coronary occlusive disease,⁶ given that a well-developed collateral circulation limits ischemia during acute myocardial infarction^50–52 and may extend the duration of myocardial viability for coronary reperfusion.⁵² Comparably, we have shown that increased density of intramyocardial coronary arterioles preserved cardiac function in hypertensive pigs.⁵³ Collateral circulation amounts to 18% of the flow in acutely occluded arteries, and protects some patients from myocardial ischemia during occlusion.⁵⁴ Hence, preexisting collateral arteries are useful during an acute coronary occlusion, preceding arteriogenesis-adaptations, and in a Meta-analysis a good collateral circulation was found to improve survival.⁵⁵ Contrarily, the Osaka Acute Coronary Insufficiency Study showed that low-moderate extent of collaterals (Rentrop Collateral Score 1–2) improved mortality compared to no visible collaterals. Yet, paradoxically the highest score was associated with a worse outcome,⁵⁶ likely due to greater cardiac morbidity.

Besides the coronary and cerebrovascular⁵⁷ circulations, collateral vessels are also fundamental in preservation of limb function in patients with peripheral arterial disease.⁵⁸ While beyond the scope of the current paper, an excellent review has recently discussed this issue.⁵⁹

Collaterals in ARVD:

The functional importance of the renal collateral circulation sustains scientific contention¹⁷ and extrapolating data from other vascular beds, even within the same species, may be misleading.⁷

In hypercholesterolemic pigs, we found that inflammation-induced intra-renal microvascular proliferation contributed to sustain renal perfusion.⁶⁰ Contrarily, similar to other organs,³⁰ hypercholesterolemia did not impact collateral-formation in RAS²⁸, underscoring the role of shear-stress relative to inflammation in promotion of arteriogenesis in RVD. Interestingly, in pigs collateral formation increased with severity of RAS beyond a hemodynamically significant 65%²⁸, but may endow limited benefit. In a recent study in 34 patients with ARVD, we found that patients with a collateral circulation in fact had greater kidney dysfunction, atrophy, hypoxia and inflammation than patients without, and a poorer blood pressure-response to revascularization.¹⁶ Similarly, in RAS pigs the extent of collateral circulation is inversely associated with distal cortical volume and perfusion.²⁸ Nonetheless, the collateral circulation may sustain the post-stenotic kidney viability until revascularization is feasible^61–63. Notably, proliferation of collaterals preserves swine kidney function in mild-moderate, but not in severe RAS.²⁸ While variable, innate maximum ability of the collateral circulation to compensate for arterial occlusion approximates 40%.^{64, 65} In all likelihood renal collaterals develop as a protective mechanism against the detrimental effects of RAS, but have limited adaptive capabilities sufficient to counterbalance only mild-moderate RAS, so that their association with kidney outcomes might be u-shaped as in the heart.

Furthermore, other elusive systemic factors may influence collateral formation. Patients with CKD possess cardiac collateral arteries of inferior quality and quantity⁶⁶, which may be exacerbated in patients with diabetes and hypertension.⁶⁷ Future studies of the effects of the comorbidities on the renal collateral circulation are needed to elucidate these issues.

Notably, while the association between blood pressure and collateral formation has been well-studied in the heart⁶⁸, its circulation is unlike the renal circulation in the sense that most of its perfusion takes place during diastole,⁶⁹ and less is known about this association in the renal circulation. Pigs with a well-developed renal collateral circulation have higher blood pressure than those without, independently of the degree of RAS,²⁸ whereas in dogs collateral development seems to reduce blood pressure.⁷⁰ However, in humans, there was no difference in the blood pressure of patients with and without collaterals, and the effect of revascularization on blood pressure was lower in patients with identifiable collaterals.¹⁶

Therapeutic manipulation of the collateral circulation:

Pharmacological interventions:

Most pharmacological manipulations of collaterals have aimed to promote angiogenesis (Figure 2). In experimental settings, monocyte-activating factors like granulocyte-macrophage and granulocyte-colony stimulating factors increase collateral-formation in the heart, brain and peripheral circulations,^{71, 72} as did VEGF⁷³ and FGF.⁷⁴ While no studies reported the effect on renal collateral formation, VEGF-infusion improves the intra-renal microvasculature in swine ARVD⁷⁵ and rats with ischemic renal disease.⁷⁶ Indeed, the angiogenic collateral-promoting potential of mesenchymal stem cells is likely partly mediated by release of these substances.⁷⁷ However, despite promising experimental results, randomized clinical trials involving these interventions have been largely disappointing.^{78, 79} Importantly, their safety in humans has been questioned, given their potential pro-atherogenic⁸⁰ and pro-oncogenic potential.⁸¹ Clearly, angiogenesis is orchestrated in-vivo by multiple concerted mechanisms that are difficult to replicate.

Figure 2:

Putative factors that may induce arteriogenesis (top-right) and angiogenesis (bottom-right) of collaterals. Dashed arrows indicate less established relationships. CSF: Colony stimulating factor. Created with Biorender.com.

Alternatively, the If-channel inhibitor Ivabradine increases pulse pressure, elevating shear-stress⁸² and avoiding inflammatory angiogenesis. Ivabradine increases collateral artery growth in the peripheral circulation of atherosclerotic mice⁸³ and coronary circulation of humans with coronary artery disease.⁸⁴ It could potentially also be effective in boosting collateral growth in ARVD, as renal vessels are exposed to blood pressure relatively similar to that in the peripheral circulation.

Other potentially collateral-promoting drugs include renin-angiotensin-aldosterone system (RAAS) blockers (Figure 2). The pro-angiogenic effect of enalapril is well-established,⁸⁵ captopril induces proliferation of collateral endothelial cells in RAS rats,⁸⁶ valsartan preserves the intra-renal microcirculation,⁸⁷ and irbesartan increases collateral blood flow in the canine heart.⁸⁸ The mechanism by which RAAS-inhibition is pro-angiogenic may involve increased availability of NO (which mediates VEGF-induced angiogenesis), H₂S⁸⁹, FGF, and/or VEGF.⁹⁰ Notably, treatment with an NO-donor increased arteriogenesis in rabbits with femoral artery ligature.⁹¹ Finally, while endothelin-A blockade improved renal microvascular density and function in pigs,²² no such effect was observed for endothelin-B.⁹² Nevertheless, long-term dual endothelin-A/B-blockade improves microvascular density in the renal cortex in RAS.⁹³ Notably, the association between mutations in the PHACTR1-gene and vascular conditions such as coronary artery disease and FMD may be mediated by changes in endothelin-receptor signaling.^{94, 95} However, the effects of the endothelin family on development of renal collateral circulation remains to be established.

Non-pharmacological interventions:

In the hindlimb of pigs with an arterio-venous shunt and increased tangential shear-stress, the collateral circulation successfully offset the lost blood flow of an occluded artery.⁹⁶ Furthermore, repeated occlusion of the murine femoral artery accelerated collateral adaptation and subsequent flow restoration, suggesting “preconditioning”.⁹⁷ Similarly, myocardial ischemic preconditioning in humans bolstered collateral supply and may endow partial protection from future episodes of myocardial ischemia.⁹⁸ While such maneuvers cannot readily be applied clinically, less invasive remote ischemic preconditioning may exert protective effects through alternative mechanisms.⁹⁹ Conceivably, patients with ARVD with renal artery re-stenosis after PTRA might be similarly “preconditioned” and acquire a higher tolerance to ischemia, although the inter-procedural time-course might be too prolonged.

Less invasive mechanical means of increasing shear-stress, like exercise, may also induce collateral formation in the human heart.¹⁰⁰ Alas, renal blood flow falls during exercise, which would naturally limit shear-stress related arteriogenesis.¹⁰¹ Hence, if exercise were to induce kidney collateral growth it would likely engage alternative mechanisms.

Enhanced external counter-pulsation (EECP) is reserved for patients with intractable chronic stable angina who are not candidates for invasive procedures. Inflation and deflation of pneumatic cuffs on the legs cause retrograde aortic flow during diastole and reduced vascular resistance during systole.¹⁰² This causes tangential vascular shear-stress, and increased coronary collateral formation.¹⁰³ Plausibly, EECP increases tangential shear-stress also in the renal vasculature. Indeed, kidney function improved after EECP in patients with stabile angina and/or congestive heart failure,¹⁰⁴ possibly secondary to improved heart function. EECP also improves renal function and blood flow in patients with liver cirrhosis.¹⁰⁵ Nonetheless, no evidence currently supports its use for induction of collateral formation in ARVD.¹⁰⁶

Recently, emerging evidence highlighted angiogenic and arteriogenic effects of extra-corporal low-energy shock-wave therapy (ESWT) in several vascular beds, including cardiac and peripheral.^{107, 108} Its effect is likely mediated mainly through mechanical stimulation of endothelial mechanoreceptors provoking release of pro-angiogenic factors like VEGF.^{109, 110} Our findings revealed effects of ESWT on swine renovascular disease with improved microcirculation and markers of inflammation.¹¹¹ However, ESWT-induced augmentation of collateral formation described in animal vasculatures¹¹² has not been replicated in humans with peripheral arterial disease.¹¹³ Nevertheless, given its mechanical stimulation of endothelial cells and release of factors implicated in shear-stress induced arteriogenesis, the potential effect of ESWT on collateral formation warrants additional investigation.

Lastly, studies in ischemic heart disease have demonstrated a pronounced effect of revascularization on coronary collaterals, where particularly smaller vessels undergo regression,¹¹⁴ although some persist, possibly consequent to systemic diseases like diabetes.¹¹⁵ We have also observed regression of collaterals after revascularization in patients with ARVD,¹⁶ but its long-term effects remain obscure.

Perspectives:

Collateral circulation is a compensatory mechanism engaged by many vascular beds, including the heart, peripheral circulation, and kidney, to adapt for obstruction of major arteries through angio- and arteriogenesis. In ARVD, experimental and clinical evidence suggests that these adaptations partially compensate for reduced perfusion of a vascular territory supplied by a stenotic vessel and sustain renal viability at least for a limited time. However, the extent of protection provided by the collateral circulation is limited and susceptible to many factors, including the severity of RAS and comorbidities. Growth factors and several pharmaceuticals may bolster collateral formation in non-renal vascular beds in experimental and/or clinical settings. Furthermore, non-pharmacological promising interventions like EECP, ESWT and ischemic preconditioning promote collateral growth in several tissues, with the renal circulation hitherto unstudied. Hence, there is an untapped potential for promotion of collateral-formation in ARVD, and theoretically to foster tissue viability in additional situations of renal hypoperfusion. Future studies are warranted to apply lessons learned from other vascular beds in patients with ARVD. Given the insidious remodeling of the intra-renal microcirculation distal to RAS, development of collateral vessels might achieve rejuvenation of the kidney parenchyma and microvasculature. Hopefully, these approaches establish new treatment opportunities for a patient group with few available therapeutic options.