Renal tissue hypoperfusion and hypoxia are key elements in the pathophysiology of acute kidney injury and its progression to chronic kidney disease. Yet, in vivo assessment of renal haemodynamics and tissue oxygenation remains a challenge. Many of the established approaches are invasive, hence not applicable in humans. Blood oxygenation level dependent (BOLD) magnetic resonance imaging (MRI) offers an alternative. BOLD-MRI is non-invasive and indicative of renal tissue oxygenation. Nonetheless recent (pre-)clinical studies revived the question as to how bold renal BOLD-MRI really is. This presentation aims to deliver some answers. It is designed to inspire the preclinical and clinical imaging, renal physiology, and nephrology communities to foster explorations into the assessment of renal oxygenation and haemodynamics by exploiting the powers of MRI. The work presented in this talk takes advantage of and heavily uses a recent review paper published by a inter-disciplinary team (1).

First, the specifics of renal oxygenation and perfusion are outlined. The fundamentals of BOLD-MRI are briefly summarized. The link between tissue oxygenation and the oxygenation sensitive MR biomarker T₂* is outlined. The merits and limitations of renal BOLD-MRI in animal and human studies are surveyed together with their clinical implications. Explorations into detailing the relation between renal T₂* and renal tissue partial pressure of oxygen (pO₂) are discussed with a focus on factors confounding the T₂* versus tissue pO₂ relation. Multi-modality in vivo approaches suitable for detailing the role of the confounding factors that govern T₂* are considered. A schematic approach describing the link between renal perfusion, oxygenation, tissue compartments and renal T₂* is proposed. Future directions of MRI assessment of renal oxygenation and perfusion are explored.

Renal oxygenation constitutes a delicate balance between oxygen delivery, as determined by renal blood flow and arterial O₂ content, and O₂ consumption, for which energy-dependent tubular reabsorption is the major determinant. Increased renal blood flow (RBF) is, in general, accompanied by increased glomerular filtration rate (GFR), and therefore necessitates increased tubular reabsorption. As compared to most non-renal tissue, whole-kidney blood flow is high and whole-kidney arterio-venous difference in O₂ content is small. Yet renal tissue perfusion and oxygenation is highly heterogeneous. Virtually all of the blood flowing into the kidney perfuses the cortex. The medulla is perfused by a small fraction (about 10% of total RBF) of blood that had traversed the cortex. Even intra-layer perfusion is quite heterogeneous. In accordance, tissue pO₂ is low in the medulla and also varies within the respective layer.

The kidney is equipped with efficient mechanisms of autoregulation, i.e., the ability to maintain RBF and GFR relatively constant in the face of moderate changes in renal perfusion pressure (RPP). It has been suggested that autoregulatory mechanisms also serve the purpose of balancing O₂ delivery with O₂ demand that arises from tubular reabsorption. It is conceivable that in the setting of renal disorders autoregulatory mechanisms create a vicious circle in which low perfusion results in tissue hypoxia that in turn further reduces perfusion. The differential perfusion and oxygenation of renal tissue is subject to changes induced by a variety of (patho)physiological factors.

The assumption that T₂* provides a surrogate of renal tissue oxygenation is based upon the T₂* dependence on O₂ saturation of Hb (StO₂) and motivated by the link between StO₂, blood pO₂, and tissue pO₂. T₂* or its reciprocal value (R₂*=1/T₂*) have been employed in numerous (pre)clinical studies as an MR based marker and surrogate of intra-renal oxygenation. T₂* sensitized MRI was employed in a broad spectrum of experimental research applications designed to detail renal haemodynamics during renal ischaemia/reperfusion (2), following administration of x-ray contrast media (3), during reversible interventions including hypoxia, hyperoxia, and short-lasting aortic occlusion (Figure 1).

It is sensible to state that T₂* is directly related to the amount of deoxyHb per tissue volume and hence correlated with tissue pO₂ via blood pO₂ and the oxyHb dissociation curve. Recent findings on the correlation between T₂* and tissue pO₂ revealed discrepancies that point at factors other than the known shifts of the oxyHb dissociation curve and changes in haematocrit, which may also confound the renal T₂*/tissue pO₂ relationship (1). The T₂* to tissue pO₂ correlation differences between interventions of hyperoxia, hypoxia and aortic occlusion together with the renal vascular conductance and kidney size data obtained by an integrated MR-PHYSIOL approach (4,5) indicate that changes in the blood volume fraction considerably influence renal T₂*. Changes in renal vascular conductance point at changes in intrarenal blood volume. This occurs via passive circular distension of vessels following changes in the transmural pressure gradient, or by active vasomotion.

Alterations in kidney size may be induced by volume changes in any of the renal fluid compartments. Besides the vasculature, the interstitial and the tubular compartments could also experience rapid volume changes and could therefore modulate the blood volume fraction. The tubular volume fraction is a unique feature of the kidney; it is quite large and can rapidly change due to (i) changes in filtration, (ii) alterations in tubular outflow towards the pelvis, (iii) modulation of the transmural pressure gradient, and (iv) changes in resorption.

En route towards a comprehensive understanding of renal haemodynamics and perfusion Figure 2 provides a schematic survey on the potential contributions of several confounders to T₂*. These include changes in (i) the tubular compartment, (ii) the intrarenal vascular compartment, and (iii) Hb concentration per blood volume (haematocrit). Aortic occlusion (or that of the renal artery) results only in moderate reduction of renal T₂* as illustrated in the empirical model shown in Figure 2. Figure 2 also illustrates a massive T₂* reduction following occlusion of the renal vein. Simultaneous occlusion of the renal vein and artery causes an intermediate T₂* reduction as pointed out in Figure 2. In each of these three cases, tissue pO₂, blood pO₂, and StO₂ approach zero. Yet the vascular volume fraction is reduced in case of arterial occlusion, increased in case of venous occlusion, and unchanged in case of the common arterio-venous occlusion. As also illustrated in Figure 2, vasodilation causes a T₂* decrease due to an increase in the vascular volume fraction, although renal tissue pO₂, blood pO₂, and StO₂ increase due to improved O₂ delivery. Vasoconstriction induces an increase in T₂* despite reduced O₂ delivery due to reduction in vascular volume fraction. Distension of tubules (e.g. induced by x-ray contrast media) results in increase in T₂* in the face of primarily unchanged StO₂, blood and tissue pO₂ due to reduced vascular volume fraction (Figure 2). Anaemia or an increase in plasma skimming evoke an increase in T₂* despite the drop in O₂ delivery and tissue pO₂ (Figure 2). Finally, changes in the inspiratory O₂ fraction change renal T₂*. Hypoxia decreases T₂* in parallel with StO₂, blood and tissue pO₂. The effect of hyperoxia on T₂* is small.

The recognition that renal T₂* does not quantitatively mirror renal tissue oxygenation in several (patho)physiological conditions and that T₂* may not reflect blood oxygenation quantitatively in some scenarios should induce due caution for quantitative interpretation of BOLD-MRI.

Deciphering the relation between regional renal T₂* and tissue pO₂ – including the role of the T₂* confounders vascular volume fraction, tubular volume fraction and oxyHb dissociation curve – requires further research. To be rendered a quantitative physiological approach, BOLD-MRI needs to be calibrated with established quantitative physiological measurements (marked in light blue in Figure 3). With this in mind Figure 3 attempts to provide a basic scheme of an integrative approach that makes good use of the capabilities of magnetic resonance, physiological measurements and near infrared spectroscopy. Productive engagement in this area carries on to drive further developments including validation and calibration of MR with the ultimate goal to provide quantitative means for interpretation of renal hemodynamics, renal oxygenation, renal blood volume fraction and tubular volume fraction related parametric MRI (Figure 3). This requires that renal blood oxygenation level associated T₂* changes need to be differentiated from T₂* changes induced by changes in the T₂* confounders tubular and vasculature volume fraction (marked in red in Figure 3). To meet this goal the renal MRI portfolio needs to entail T₂* mapping to probe for changes in blood oxygenation level but also T₂ mapping and proton density imaging to monitor net water changes and fluid shifts (marked in dark blue in Figure 3). T₂ mapping would also help to discriminate the susceptibility induced contributions (T₂') from the T₂ relaxation contributions to T₂*. For this purpose it is also appealing to pursue quantitative susceptibility mapping (QSM). QSM provides a novel MR contrast mechanism to determine apparent magnetic susceptibility in tissue, which is useful for identification of BOLD contributions. For a comprehensive characterization of renal oxygenation and perfusion it is prudent to include MR assessment of renal blood volume (RBV) using intravascular contrast agents such as ultrasmall superparamagnetic iron oxide (USPIO) based agents to probe for vasodilation, vasoconstriction and other changes in the blood volume fraction as illustrated in Figure 3.

The development and application of multi-modality approaches continues to be in a state of creative flux. It is to be expected that future hybrid implementations may include optical imaging techniques such as NIRS. Notwithstanding its depth penetration constraints which at present limits its application to the renal cortex, NIRS has the potential to help characterize renal oxygenation (6). NIRS can be applied to monitor changes of StO₂ and Hb concentration per tissue volume in the kidney. With advanced NIRS approaches it is even possible to measure absolute values of Hb concentration and StO₂.

To conclude, further explorations are essential before the quantitative capabilities of parametric MRI can be translated from experimental research to the clinic to improve our understanding of hemodynamics/oxygenation in kidney disorders. Moving T₂* sensitized MRI from the research area into the clinic remains challenging but has spurred the drive towards a comprehensive stand-alone MR protocol for assessment of renal haemodynamics and oxygenation. As parametric mapping of T₂* and other MR biomarkers become increasingly used in preclinical research and clinical science, they should help to further advance the potentials of MR for assessing kidney diseases.