The results of this study indicate that chronic oral furosemide administration attenuates the increase in renal medullary oxygenation in response to an acute 20 mg IV furosemide stimulus measured by BOLD MR, that higher daily doses of furosemide are associated with reduced increases in the medullary BOLD responses to a standard furosemide challenge, and that changes in RBF are correlated with changes in renal tissue oxygenation in the cortex but not in the medulla.
Renal artery stenosis is a complex pathophysiologic process that can result in malignant hypertension (HTN), renal dysfunction and exacerbation of congestive heart failure [7–9]. Renal artery revascularization procedures, either surgical or percutaneous, have been utilized to treat renovascular HTN in patients with malignant HTN, declining renal function, or those who develop pulmonary edema based on current guideline recommendations . However, less than half of patients who undergo these revascularization procedures actually receive significant clinical benefit defined as improvement in renal function and GFR, or blood pressure . To date, ideal strategies to identify patients who will derive a benefit from renal artery revascularization procedures remain elusive.
Recently, BOLD MR imaging has been evaluated as a potential tool to determine if a kidney with a renal artery lesion can increase oxygenation in response to a 20 mg IV furosemide stimulus [1, 2]. Small cross-sectional studies suggest that a kidney with the ability to improve tissue oxygenation (manifested by an increase in BOLD signal intensity) in response to a furosemide stimulus may be more likely to benefit from a revascularization procedure [1, 2, 11]. Furosemide, a loop diuretic, blocks transport of the sodium-potassium-2chloride (Na-K-2Cl) transporter located in the thick ascending limb of the loop of Henle (TALH) within the renal medulla . Luminal Na + uptake via the apical membrane Na-K-2Cl transporter is coupled to basolateral membrane transport which depends on Na-K-ATPase activity, an oxygen (O2)-dependent process [13, 14]. Therefore, blocking the Na-K-2Cl transporter reduces renal medullary O2 consumption and increases tissue oxygenation.
Since many patients with severe renovascular disease require loop diuretics to prevent hypervolemia, it is important to determine the effect of chronic loop diuretic use on the BOLD MR response to an acute stimulus with furosemide. Otherwise, patients receiving chronic loop diuretic therapy who demonstrate attenuated increases in T2* signal intensities after acute furosemide administration may be inappropriately deemed less likely to receive clinical benefit from renal artery revascularization procedures. Our results demonstrate that chronic oral furosemide consumption is associated with attenuated increases in medullary oxygenation after 20 mg of IV furosemide (Figure 2) despite abstaining from furosemide ingestion greater than 12 hours prior to BOLD MR examination. The half-life of furosemide is 30 to 60 minutes, therefore 12 hours should be more than adequate time to withhold furosemide. The results of our model indicate that a patient receiving 40 mg of furosemide twice daily or 80 mg once daily would be predicted to exhibit nearly a 10% attenuation of the T2* signal intensity response in the renal medulla to a 20 mg IV furosemide stimulus, assuming no differences in the other variables.
Despite worsened renal dysfunction and a higher prevalence of diabetes in the chronic furosemide group, pre-furosemide cortical and medullary T2* values were similar to furosemide naïve participants. This observation illustrates the point that unless patients have a critical renal arterial narrowing, the renal cortex and medulla are able to maintain tissue oxygenation, at least at rest. However, upon a 20 mg IV furosemide challenge, participants receiving chronic furosemide did not significantly increase T2* signal intensities in the cortex. Although these participants did exhibit a significant increase in T2* signal intensities after 20 mg IV furosemide in the renal medulla compared to pre-furosemide values, they exhibited attenuated increases in medullary T2* values compared to furosemide naïve participants (Figure 4).
After adjustment for covariates age, gender, race, GFR, diabetes, stenosis severity and BMI, number of antihypertensive medications, use of ACE inhibitors or ARBs, chronic daily furosemide dose was a significant negative predictor of the renal medullary T2* response to an acute furosemide stimulus. Although it did not quite meet statistical significance (p = 0.07) due to a relatively small number of participants in this study, total daily furosemide dose was associated with attenuated T2* signal intensity changes in response to an acute furosemide stimulus. However, chronic daily furosemide dose did not predict the response in the cortex. This finding is consistent with the site of action of loop diuretics such as furosemide which act on the TALH . In rodents, furosemide significantly increased O2 levels measured by O2-sensitive electrodes in the renal medulla . This increase in medullary oxygenation was due to decreased tubular O2 consumption as medullary blood flow was decreased by nearly 30%.
White race was also associated with significantly attenuated T2* responses in the medulla to an acute furosemide stimulus. Textor et al. found higher baseline BOLD MR renal medullary R2* values (inverse of T2*) in African Americans, compared to whites . After a furosemide stimulus, however, these R2* levels decreased to the same values as in whites. These findings suggest increased O2 consumption and TALH Na + reabsorption in African Americans compared to whites. If whites reabsorb less Na + in the TALH, they would be expected to have less inhibition of the Na-K-2Cl transporter in response to furosemide. Therefore, our findings corroborate those of Textor et al. suggesting that African Americans have increased Na + reabsorption in the TALH compared to whites. The mechanisms involved in the augmented TALH Na + reabsorption in African Americans are poorly understood and warrant further study.
Higher chronic furosemide doses were associated, albeit weakly, with attenuated T2* responses to an acute furosemide stimulus. Although this association was not statistically significant, it would likely achieve significance with a larger study size. This finding is intuitive, given that with most classes of medications, higher chronic daily doses reflect altered pharmacokinetics requiring higher acute doses to achieve a desired physiologic response.
Renal tissue oxygenation is at least partly dependent on oxygen delivery via RBF [18, 19]. One limitation of many studies evaluating renal oxygenation with BOLD MR is the reliance on either iodinated or gadolinium based contrast administration for determination of RBF. The former would require a separate imaging examination (computed tomography [CT]), possibly introducing different physiological conditions at separate time points. While RBF and perfusion can be determined during one single imaging examination with the administration of gadolinium contrast, this agent is contraindicated in patients with GFR < 30 ml/min/1.73 m2 due to the risk of nephrogenic systemic fibrosis , thereby limiting the potential application of this diagnostic strategy in individuals with progressed renal dysfunction.
We employed phase contrast techniques to quantify RBF in each renal artery before and after IV furosemide to determine whether RBF influenced O2 utilization. In furosemide naïve participants, RBF increased in response to 20 mg IV furosemide suggesting vasodilation which has been reported previously in human studies . In participants chronically administered furosemide, RBF tended to be lower at baseline and did not increase after 20 mg IV furosemide. This may reflect the higher prevalence of diabetes and/or renal insufficiency in these participants. There was no difference in the severity of stenoses between the two groups.
Change in cortical T2* intensity was correlated with RBF measures, while there was no correlation observed in the medulla. Compared to the cortex, the medulla receives much less blood flow relative to its high metabolic workload and functions in near-hypoxic conditions at baseline [22, 23]. Therefore, alterations in RBF and subsequently O2 delivery should be better tolerated in the cortex due to its abundance of blood flow . Our findings suggest that the increase in tissue oxygenation observed in the cortex is likely related to increases in total RBF and O2 delivery, at least in those not receiving chronic furosemide therapy. However, we observed increases in medullary tissue oxygenation in response to an acute furosemide stimulus although this response was reduced compared to those not receiving chronic furosemide. The relationship between RBF and oxygenation is complex, particularly in the medulla. Increases in RBF do increase renal O2 delivery; however, increases in RBF also increase tubular Na + delivery and transport resulting in increased metabolic workload . Our observations may reflect compensatory mechanisms in the medulla to maintain homeostasis and potentially downregulate non-essential metabolic functions in times of reduced RBF and O2 delivery that may occur with renal arterial stenoses. Alternatively, RBF may be shunted towards the medulla during stress conditions to maintain metabolic demands as the cortex is relatively overperfused and is likely able to better tolerate reductions in total RBF compared to the medulla. Hence, there is no significant correlation between total RBF and renal medullary oxygenation unless RBF is reduced below a critical point which may be observed in severe renal arterial narrowings.
Our study has some limitations. We were unable to obtain measurements of renal oxygenation by BOLD in 6 of the 76 kidneys due to artifacts (motion or bowel gas) and poor image quality. RBF measures were not performed in 12 of the 76 renal arteries due to inadequate visualization of the renal artery on phase contrast velocity maps due to extremely low blood flow or motion artifact. Second, Na + intake or reabsorption and water loading conditions were not monitored in our study and could potentially confound our results. Another potential confounder is the higher prevalence of participants on ACE inhibitors or ARBs in the furosemide naïve group. We attempted to adjust for these factors in our multivariable analyses. Although these agents theoretically may influence renal tissue oxygenation , they did not increase renal issue oxygenation in type 2 diabetic hypertensive patients in a recent study using BOLD MR imaging with a furosemide stimulus . Because the GFRs were different between our two groups, we performed additional analyses with exclusion of five of the furosemide naïve group with GFR >90 to try to minimize some of the between group differences which and the same variables (ethnicity, stenosis severity and chronic daily furosemide dose) remained as significant independent predictors of changes in T2* intensity in response to an acute furosemide stimulus. Our findings are consistent with findings published in a study of 280 patients demonstrating no differences in T2* intensities between different stages of CKD . However, our findings build upon this work by demonstrating no relationship of GFR to changes in T2* signal intensities in response to an acute furosemide stimulus. Because many of these factors are closely linked, future studies specifically aimed at determining the effects of chronic furosemide therapy on acute T2* responses to a furosemide stimulus should be evaluated in better matched cohorts, possibly with differing doses of furosemide stimuli. It would be reasonable to investigate administering a larger intravenous dose of furosemide comparable to the chronic furosemide dose that patients’ receive chronically to be able to detect an appropriate increase in renal oxygenation before determining if a kidney downstream from a severe renal artery stenosis would benefit from a revascularization procedure.