Renal and Vascular Effects of Uric Acid Lowering in Normouricemic Patients With Uncomplicated Type 1 Diabetes

Plasma uric acid (PUA) levels are associated with the pathogenesis of diabetic complications, including cardiovascular disease and kidney injury (1). Interestingly, extracellular PUA levels are lower in young adults and adolescents with type 1 diabetes (T1D) compared with healthy control subjects (HCs) (2–4), likely due to a stimulatory effect of urinary glucose on the proximal tubular GLUT9 transporter, which induces uricosuria (2). Thus, PUA-mediated target organ injury may be mediated by the intracellular uric acid effects, uricosuria-related tubular cell exposure (5), sequestration of PUA along the vascular endothelium, or PUA-mediated inflammation and activation of the renin-angiotensin (ANG)-aldosterone system (RAAS) (1).

Consequently, even within the normal range, PUA is associated with impaired renal function (6,7), early glomerular filtration rate (GFR) loss (8), and an increased risk of the development of proteinuria in patients with T1D (9). Even in young adults and adolescents with T1D without complications, higher PUA levels are associated with lower GFR (2–4), which may be driven by a PUA-mediated increase in afferent renal arteriole resistance potentially promoting ischemia to the renal microcirculation (4). Additionally, accumulating evidence suggests that PUA levels are independently associated with increased intimal medial thickness, endothelial dysfunction, and vascular stiffness (10), promoting the development of hypertension, cardiovascular disease, and chronic kidney disease (6,11). Although not observed in adolescents with T1D, the association between higher PUA levels within the normal range and higher blood pressure (BP) has been reported in young adults with uncomplicated T1D (2). Such relationships between PUA and early renal and cardiovascular risk factors in young patients with T1D suggest that lowering PUA may be an important strategy to reduce renal and cardiovascular complications related to diabetes.

From the limited data available in patients with diabetes, lowering of PUA levels may improve endothelial function (12), lower BP (13–15), slow GFR decline, reduce proteinuria, and suppress renal inflammation (1,16–18). Given the lack of effective therapy that protects against initiation and progression of diabetic complications, it is of the utmost importance to evaluate the renal and vascular effects of pharmacological PUA lowering in patients with T1D. Accordingly, the goals of our physiological study were to determine whether PUA lowering with the xanthine oxidase inhibitor febuxostat (FBX) modifies 1) the effect of hyperglycemia and infused ANG II on renal hemodynamic function, 2) systemic BP, and 3) arterial stiffness and endothelial function during clamped euglycemia and hyperglycemia in an even earlier cohort of young T1D adults without any complications.

Using a study design with both euglycemic and hyperglycemic clamp conditions, we wanted to assess the effects of uricosuria-related tubular exposure that is augmented by hyperglycemia and the consequent glycosuria. As we have previously shown, glycosuria stimulated by hyperglycemia leads to uricosuria (2). It was therefore important to compare PUA-lowering effects during euglycemic conditions with those during hyperglycemic conditions. In addition, unlike other PUA-lowering agents, FBX lowers PUA levels sequestered along the vascular endothelium (1), allowing us to potentially target PUA-mediated injury mechanisms that are due to either paracellular or even intracellular uric acid levels. Finally, because of interactions among PUA, the RAAS, and inflammation, we measured plasma aldosterone and renin levels as well as urinary inflammatory marker levels and hemodynamic responses to an exogenous ANG II infusion to better elucidate PUA lowering as a modulator of these traditional renal and cardiovascular injury pathways.

Twenty-four HCs and 49 patients with T1D completed this open-label, proof-of-principle, 8-week FBX treatment study (Fig. 1) (ClinicalTrials.gov identifier NCT02344602). T1D study participants included 42 patients with T1D with normofiltration (T1D-N; GFR <135 mL/min/1.73 m²) and 7 patients with T1D with hyperfiltration (T1D-H; GFR ≥135 mL/min/1.73 m²). Detailed inclusion criteria were as follows: 1) male and female participants 18–40 years old; 2) normoalbuminuria on a 24-h urine collection; 3) BMI of 18–35 kg/m²; 4) normal renal and liver function; 5) normal electrocardiogram; 6) clinic BP <140/90 mmHg; 7) T1D duration >5 years; 8) normal PUA levels <450 µmol/L based on Clinical Reference Laboratory Guidelines at the time of study initialization; and 9) no history of renal or cardiovascular complications and no intake of concomitant medications that would alter BP or cardiovascular outcomes or interfere with purine metabolism. The study was approved by the University Health Network Research Ethics Board (Toronto, ON, Canada), and all subjects gave written informed consent.

Patients with T1D were studied at baseline (day 1, euglycemic; day 2, hyperglycemic) and after 8 weeks of FBX therapy 80-mg daily (day 3, euglycemic; day 4, hyperglycemic). Medication compliance was assessed by pill counting and was >90% in all participants. During clamped euglycemic study days, the blood glucose level was maintained between 4 and 6 mmol/L, and during clamped hyperglycemic study days the blood glucose level was maintained between 9 and 11 mmol/L. Studies were performed after 7 days on a controlled diet consisting of ≥150 mmol/day sodium and ≤1.5 g/kg/day protein. The sodium-replete diet was used to avoid circulating volume contraction, RAAS activation, and between-subject heterogeneity and in an attempt to keep study conditions similar to typical North American dietary patterns. Prestudy protein intake was modest to avoid the hyperfiltration effect of high-protein diets (19). Compliance was ascertained by the measurement of 24-h urine sodium and urea excretion on the seventh day prior to the studies. All study participants were instructed to avoid caffeine-containing products and to have the same light breakfast on the morning of each study visit. HCs were studied during normoglycemic conditions only.

Subjects presented to the Renal Physiology Laboratory on days 1 and 2 for the baseline euglycemic and hyperglycemic studies. After the respective clamped glycemic level was achieved for 5 h, blood samples were collected for measuring levels of inulin and p-aminohippuric acid (PAH) and for baseline circulating levels of RAAS mediators (plasma renin concentration and aldosterone). Oscillometric brachial artery BP measurements were obtained in duplicates in a reclining position at 30-min intervals throughout the study (Critikon, Tampa, FL). Subjects remained supine at all times. Baseline renal hemodynamic function (GFR and effective renal plasma flow [ERPF]) was measured using inulin and PAH clearance according to the plasma disappearance technique (20,21). The mean of the final 2 clearance periods represented baseline GFR and ERPF, expressed per 1.73 m².

The following parameters were calculated as follows:

where MAP is mean arterial pressure. Indirect intraglomerular hemodynamic parameters were estimated using equations of Gomez (22) based on data from animal studies. These equations were successfully used in a similar manner to evaluate patients with conditions such as hypertension, endocrine disorders, and T1D (4,23,24). The following assumptions were imposed by the Gomez equations: 1) intrarenal vascular resistances are divided into afferent, postglomerular, and efferent; 2) hydrostatic pressures within the renal tubules, venules, and Bowman space and interstitium (P_BOW) are in an equilibrium of 10 mmHg; 3) the glomerulus is in filtration disequilibrium; and 4) the gross filtration coefficient (K_fg) is 0.1733 mL/s/mmHg (glomerular hydrostatic pressure [P_GLO] = 47.5 mmHg) for HCs and K_fg = 0.1012 mL/s/mmHg (P_GLO = 56.4 mmHg) for patients with T1D to reflect the different P_GLO values in diabetic and control conditions observed in previous micropuncture studies in Munich-Wistar rats (25). MAP (in mmHg), ERPF (in mL/s), GFR (in mL/s), and total protein (in g/dL) measurements were used to calculate efferent resistance (R_E; in dyne ⋅ s ⋅ cm⁻⁵), afferent resistance (R_A; in dyne ⋅ s ⋅ cm⁻⁵), P_GLO (in mmHg), change in filtration pressure across glomerular capillaries (ΔP_F; in mmHg), and glomerular oncotic pressure (in mmHg).

The ΔP_F was calculated as follows:

The glomerular oncotic pressure from the plasma protein mean concentration (C_M) within the capillaries:

P_GLO:

R_A and R_E were estimated using the principles of Ohm’s law, where 1,328 is the conversion factor to dyne ⋅ s ⋅ cm⁻⁵ (22):

On euglycemic day 1 only, after baseline clearance periods were complete, ANG II (Clinalfa; Bachem AG, Bubendorf, Switzerland) was administered at incremental doses of 1 and 3 ng/kg/min, each over 30 min, followed by a 30-min recovery phase (26). Oscillometric brachial artery BP measurements were obtained in duplicate in a reclining position every 5 min during each ANG II infusion (Critikon). Blood was collected during each ANG II infusion period and after a 30-min recovery for hematocrit, inulin, and PAH to assess renal hemodynamic parameters.

Assessments of nitrite and nitrate levels, metabolites of nitric oxide (NO), were performed on midstream urine samples collected on each of the 4 study days after respective euglycemic and hyperglycemic clamps were stabilized for 3 h. Nitrite assessments were performed based on the ELISA kit instructions provided by the manufacturer (R&D Systems, Minneapolis, MN). Similarly, urine specimens were used to measure levels of cytokines/chemokines using an established Cytokine/Chemokine Panel Luminex Assay (Eve Technologies, Calgary, Alberta, Canada). The following urinary cytokine/chemokines were measured: epidermal growth factor, fibroblast growth factor-2, Eotaxin 1, transforming growth factor-α, granulocyte colony-stimulating factor, Flt-3 ligand, granulocyte-monocyte colony-stimulating factor, Fractalkine, interferon (IFN)-α2, IFN-γ, growth-related oncogene-α, interleukin (IL)-10, MCP-3, IL-12p40, macrophage-derived chemokine, IL-12p70, platelet-derived growth factor-AA, IL-13, platelet-derived growth factor-BB, IL-15, soluble CD40L, IL-17A, IL-1RA, IL-1α, IL-9, IL-1β, IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-8, IFN-inducible protein-10, MCP-1, macrophage inflammatory protein-1α, macrophage inflammatory protein-1β, regulated on activation normal T-cell expressed and secreted, tumor necrosis factor-α, tumor necrosis factor-β, vascular endothelial growth factor A, and IL-18. All markers in urine were corrected for urinary creatinine concentration.

Vascular assessments were performed on each of the 4 study days after ambient glycemia had been stabilized and before renal hemodynamic function testing. In brief, arterial compliance was measured noninvasively using a SphygmoCor device (AtCor Medical Systems Inc., Sydney, New South Wales, Australia). Right carotid artery waveforms were recorded with a high-fidelity micromanometer (SPC-301; Millar Instruments) and using the validated transfer function, and corresponding central aortic pressure waveform data were generated. MAP and heart rate (HR) were determined using the integral software. The augmentation index (AIX), an estimate of arterial stiffness, was calculated as the difference between the second systolic peak and the inflection point, which was expressed as a percentage of the central pulse pressure corrected to a HR of 75 beats per minute (bpm). The aortic pulse wave velocity (PWV) was measured using the same device.

Endothelial function of the brachial artery was determined by the diameter change in response to an increased blood flow generated by reactive hyperemia (flow-mediated dilation [FMD]). Reactive hyperemia was stimulated by a 5-min inflation of a pneumatic cuff placed distal to the antecubital fossa followed by deflation. Endothelial-independent function of the brachial artery was measured by the diameter change in response to a sublingual nitroglycerin spray (400 µg; nitroglycerin-mediated dilation [GMD]). A high-resolution B-mode ultrasound device (Vivid i [7- to 15-MHz linear-array transducer]; GE/Vingmed, Waukesha, WI) was used to capture longitudinal electrocardiogram-gated end-diastolic images of the brachial artery before and after cuff inflation. The diameter was determined using an automated edge-detection algorithm that has been described previously (27), and blood flow was measured from the velocity-time integral of the Doppler signal. FMD was defined as the maximal percentage change in vessel diameter after reactive hyperemia and was also reported as FMD/flow, which was defined as the maximal percentage change in vessel diameter divided by the percentage change in flow to create a stimulus-adjusted response measure (28). GMD was defined as the maximal percentage change in vessel diameter within 5 min after the sublingual nitroglycerin spray was administered.

The primary end point of this study was the change from baseline in GFR after 8 weeks of FBX treatment under stable euglycemic and hyperglycemic conditions. Our previous data have shown that the SD of the ΔGFR in response to RAAS modulation is ∼19 mL/min/1.73 m² (26,29). To detect a 10 mL/min/1.73 m² between-group difference in the GFR response to PUA lowering, for a two-sided test with P = 0.01 (to correct for multiple comparisons) and with Z_α = 2.58, the sample size equals 24 in each group (T1D-H and T1D-N) for a total of 48 patients with T1D (29). Because only seven patients with T1D exhibited hyperfiltration in our cohort, the analysis based on hyperfiltration status was only exploratory.

The difference between renal hemodynamic parameters at euglycemic clamp and hyperglycemic clamp was used to analyze the hyperglycemic response before and after FBX treatment. The difference between renal hemodynamic parameters at baseline euglycemic clamp and 30 min after the 1 and 3 ng/kg/min ANG II infusions were used to analyze the ANG II response before and after FBX treatment.

One T1D-H patient was excluded from the analysis examining the effect of FBX on renal hemodynamic function during the euglycemic clamp only due to issues with the blood sample collected for the inulin and PAH measurements. For similar reasons, one patient with T1D-N was excluded from the analysis of ANG II infusion responses before and after FBX treatment. One patient with T1D-N was excluded from the analysis of radial AIX during hyperglycemia because of an inability to obtain measurements before and after FBX treatment.

Data are presented as the mean ± SD. Within-group differences and responses to FBX treatment were analyzed using paired t tests. To assess for between-group differences, ANOVA with a post hoc Tukey test was used. Linear regression analysis was used to determine correlations between renal hemodynamic responses and PUA levels. Statistical significance was defined as P < 0.05. All statistical analyses were performed using SAS version 9.1.3 and GraphPad Prism software (version 5.0).

The study population comprised 49 patients with T1D (42 patients with T1D-N, 7 patients with T1D-H) and 28 HC participants (Table 1). All patients had a mean T1D duration of 14.3 ± 7.2 years. Overall, baseline characteristics were similar between the T1D and HC groups, whereas 24-h protein intake tended to be lower and hemoglobin A_1c level was higher in the T1D group.

As expected, patients with T1D at baseline under euglycemic conditions had lower PUA levels compared with HCs (240 ± 62 vs. 303 ± 71 μmol/L, respectively; P = 0.0002), and PUA levels were further lowered in patients with T1D during hyperglycemic conditions (240 ± 62 vs. 221 ± 61 µmol/L; P < 0.0001). FBX treatment significantly decreased PUA levels by ∼50% in each group, as follows: HC group, 303 ± 71 to 131 ± 55 μmol/L (P < 0.0001); T1D group during euglycemic conditions, 240 ± 62 to 124 ± 53 μmol/L (P < 0.0001); and T1D group during hyperglycemic conditions, 221 ± 61 to 108 ± 42 μmol/L (P < 0.0001) (Table 2).

In the HC group, FBX treatment did not influence renal hemodynamic function (ERPF, GFR, FF, RBF, and RVR), intraglomerular hemodynamics (P_GLO, R_A, R_E, and R_A/R_E ratio), BP (systolic BP [SBP], diastolic BP [DBP], and HR), or vascular parameters (aortic AIX, carotid AIX, carotid femoral and carotid radial PWVs, FMD, and GMD) (Table 3). FBX treatment increased the root mean square of successive differences (RMSSD; 64.1 ± 29.1 to 80.9 ± 46.9 ms; P = 0.0293) but not the SD of all normal R-R intervals (SDNN; 77.4 ± 22.8 to 88.2 ± 37.8 ms, P = 0.0769) in HC participants.

In the overall T1D group, FBX treatment led to a modest decrease in SBP (112 ± 9 to 109 ± 9 mmHg; P = 0.0491) )Table 3) but not in DBP (67 ± 6 to 66 ± 7 mmHg; P = 0.2823) or HR (67 ± 11 to 66 ± 10 bpm; P = 0.8605). Although there were no differences observed in renal hemodynamic function under euglycemic conditions in response to FBX treatment, P_GLO decreased (54.4 ± 4.1 to 53.4 ± 3.2 mmHg; P = 0.0497) without significant changes in R_A, R_E, and R_A/R_E ratio in patients with T1D (Table 3). FBX treatment did not alter vascular stiffness parameters (aortic AIX, carotid AIX, and carotid femoral and carotid radial PWVs), measures of endothelial-dependent and endothelial-independent vascular function (FMD and GMD, respectively), or HR variability measures (RMSSD and SDNN, Table 3).

Further analysis revealed normalization of the GFR in the six patients with T1D-H (150 ± 13 to 129 ± 10 mL/min/1.73 m²; P = 0.0113) (Fig. 2) but not in patients with T1D-N (111 ± 14 to 111 ± 16 mL/min/1.73 m²; P = 0.9227).

In patients with T1D during hyperglycemic conditions, FBX treatment did not affect renal hemodynamic function, intraglomerular hemodynamics, BP, vascular parameters, or measures of endothelial-dependent and endothelial-independent vascular function (Table 3) (Fig. 3).

As observed in our previous studies in separate cohorts (2–4), a higher PUA level was correlated with a lower GFR (during euglycemia, r = −0.37, P = 0.009; during hyperglycemia, r = −0.46, P = 0.0009) and lower ERPF (during euglycemia, r = −0.29, P = 0.047; during hyperglycemia, r = −0.39, P = 0.006 during hyperglycemia). After FBX treatment, a higher PUA level was still correlated with a lower GFR during euglycemia (r = −0.30, P = 0.038) but not during hyperglycemia (r = −0.18, P = 0.22), and associations with ERPF were not significant (during euglycemia, r = −0.087, P = 0.56; during hyperglycemia, r = −0.15, P = 0.29). No correlations were observed between PUA and GFR or ERPF in HCs before or after FBX treatment.

FBX treatment led to an augmented FF response to hyperglycemia (0.01 ± 0.04 to 0.03 ± 0.04, P = 0.0296), which was accompanied by an exaggerated increase in R_E (195 ± 384 to 400 ± 525 dyne ⋅ s ⋅ cm⁻⁵, P = 0.0271) and P_GLO (0.5 ± 3.6 to 2.7 ± 3.7 mmHg, P = 0.0053) but not in R_A, GFR, or ERPF (Fig. 3). No significant differences were observed in BP or vascular parameters responses to hyperglycemia before versus after FBX treatment.

FBX treatment did not alter plasma aldosterone or renin levels in HCs or patients with T1D under euglycemic and hyperglycemic conditions (Table 2) and did not alter the renal hemodynamic response to 1 or 3 ng/kg/min ANG II infusions (Fig. 4), nor was there an impact on changes in intraglomerular hemodynamic parameters or BP responses to ANG II in either group before versus after FBX treatment.

FBX treatment decreased hemoglobin A_1c levels (5.05 ± 0.22 to 4.97 ± 0.23%, P = 0.0167) in HC participants but not in patients with T1D during euglycemic or hyperglycemic conditions. FBX treatment did not alter 24-h protein intake, 24-h urine sodium excretion, BMI, estradiol or progesterone levels (females only), or any other clinically relevant biochemical and hematological parameters assessed (including serum sodium, potassium, calcium, magnesium, chloride, phosphate, and liver enzymes) in T1D or HC study participants. No adverse events were reported, aside from some mild nausea after FBX intake in several patients, which resolved after taking the agent with food.

FBX treatment did not significantly affect levels of urinary nitrite and nitrates in either HC or T1D study participants during euglycemic or hyperglycemic conditions. In patients with T1D, PUA lowering with FBX caused an exaggerated increased response to hyperglycemia in urinary IL-5 (P = 0.0029), IL-9 (P = 0.0067), and IL-18 (P = 0.0136) levels (Fig. 5). No other changes in the measured urinary inflammatory markers were observed.

In patients with T1D, PUA levels are linked with early renal function loss (3,7). Although the association between PUA levels and renal risk is compelling, studies measuring the effects of PUA-lowering therapies on renal outcomes focused on older, hypertensive patients with type 2 diabetes and established chronic kidney disease. Our study examined the effect of PUA lowering on renal hemodynamic and vascular function in young, normotensive patients with T1D who had normal renal function and normal baseline PUA levels. Our overall aim was to determine whether there is a physiological rationale for PUA lowering in the presence of uncomplicated T1D. We used FBX rather than allopurinol as a physiological probe because of the greater potency and better safety profile of this agent compared with older agents. Furthermore, because of its mechanism of action, FBX does not stimulate the production of reactive oxygen species, which may limit the vascular protective effects of allopurinol (30).

Although our first major observation suggests that PUA lowering does not affect renal function measured by GFR and ERPF in patients with T1D or HCs, we observed a significant reduction in GFR in each of the six patients with hyperfiltration at baseline. Because of the lack of a placebo group in this physiological analysis, it is unclear whether such GFR normalization occurred due to a regression to the mean or to a physiological effect (31). Given the role of hyperfiltration in predicting the subsequent development of microalbuminuria and nephropathy in patients with T1D (32), placebo-controlled studies with larger T1D-H cohorts are needed to determine whether PUA lowering can normalize hyperfiltration, thereby promoting renal protection.

Similar to our previous reports (2,4), in our current T1D cohort, higher baseline PUA levels within the normal range correlated with lower GFR and ERPF during euglycemia and hyperglycemia. After FBX treatment, the correlation between PUA and renal hemodynamic function remained significant only for ERPF during euglycemic conditions—likely as a result of the decreased PUA range after FBX treatment, which limited our power to detect significant correlations. Existing data therefore suggest that long-term PUA exposure does not cause hyperfiltration but that short-term lowering of PUA may normalize hyperfiltration through as yet undefined mechanisms. Our study does not clearly indicate whether PUA is a cause or a consequence of altered renal hemodynamic function in patients with T1D and thus requires further investigation.

Consistent with previous studies reporting allopurinol-related effects on BP (13–15), our second observation was the modest decrease in SBP observed in T1D patients under euglycemic conditions in response to FBX treatment. In our cohort, FBX treatment did not alter the following previously reported FBX-mediated BP-lowering effects (33–35): arterial stiffness, HR variability, levels of plasma RAAS markers, or the hemodynamic responses to an exogenous infusion of ANG II, endothelial-dependent, or endothelial-independent function. Given that FMD is a measure of NO bioavailability and that PUA lowering has been shown not to have an impact on FMD in other cohorts, it is unlikely that the BP-lowering effect of FBX is mediated by significant changes in NO bioavailability (36). Future studies may consider examining the role of inflammation and other mechanisms in FBX-mediated BP-lowering effects.

Whether PUA influences the vascular response to clamped hyperglycemia is unknown in patients with uncomplicated T1D. Our third major observation, contrary to what we expected, was that PUA lowering enhanced the renal FF response to clamped hyperglycemia through an increase in R_E. Constriction of the efferent renal arteriole is a known mechanism of hyperfiltration mediated by intrarenal RAAS activation in T1D (10). Moreover, PUA is positively associated with plasma renin activity in humans, even when PUA levels are within the normal range (<450 μmol/L) (34,37). The association between PUA and RAAS activation has been further strengthened by previous findings demonstrating a negative correlation between PUA and the renal hemodynamic response to an ANG II infusion in a cohort of individuals with a wide range of PUA values (100–595 μmol/L) (38). Contrary to these previous observations, ANG II infusion hemodynamic responses and plasma renin and aldosterone levels did not change after FBX treatment in our cohort. Interestingly, PUA lowering with FBX in patients with T1D led to an increase in levels of a proinflammatory cytokine, IL-18, which has been reported to play a pivotal role in the pathogenesis of diabetic nephropathy (39). Therefore, PUA lowering may enhance the renal FF response to clamped hyperglycemia through an increase in R_E and IL-18 but without impacting the RAAS or NO levels, as measured by the FMD and urinary nitrate and nitrite levels. PUA may therefore augment hemodynamic mechanisms that mediate the renal response to hyperglycemia at the efferent arteriole and may be an important anti-inflammatory mechanism required to maintain the renal hemodynamic response to hyperglycemia. Our findings suggest the need for an optimal PUA balance, whereby PUA levels are high enough to limit the renal vasoconstrictive response to hyperglycemia but are not excessively elevated in a range that lowers renal function and increases BP. Such observation are consistent with recent findings by Uedono et al. (40), where a U-shaped relationship between PUA and renal hemodynamic parameters was observed, with mild hyperuricemia and mild hypouricemia being significantly associated with lower GFR and ERPF in healthy Japanese subjects. Hypouricemia has also been associated with endothelial dysfunction in otherwise healthy patients with a URAT1 loss-of-function mutation (41). Therefore, future studies should examine a range of FBX doses to find an optimal level of PUA that maximizes cardiorenal protection.

Our study has limitations. To minimize the effects of the relatively small sample size, we recruited homogenous study groups with careful prestudy preparation. As a result, our data are limited in generalizability to other populations, such as T1D patients with longer diabetes duration, degree of proteinuria, nephropathy, and others. Moreover, all study participants were normouricemic, and, thus, our findings cannot be generalized to hyperuricemic patients. Therefore, it is possible that although the hemodynamic changes we observed in normouricemic patients were modest, lowering PUA in patients with hyperuricemia might still yield benefits around renal and cardiovascular protection. Although endogenous production or exogenous consumption of uric acid in the form of purines and fructose were not recorded, we observed similar levels of urinary urea excretion (as a surrogate marker of protein intake) before FBX and after FBX treatment. Each study participant acted as his/her own control to decrease possible intraindividual variability. Additionally, we were not able to study intracellular uric acid mechanisms because of the inability to perform such measurements in human studies. Next, we recognize that the Gomez equations are indirect estimates of intraglomerular hemodynamic parameters and are based on a few physiological assumptions, which have nevertheless been carefully studied and successfully used in various healthy populations and populations with disease over the last 60 years, including in patients with uncomplicated T1D, and do appear to reflect dynamic changes in renal hemodynamic function (4,24).

In conclusion, treatment with FBX in patients with uncomplicated T1D for 8 weeks had a modest BP-lowering effect, appeared to normalize hyperfiltration, and enhanced the renal FF response to clamped hyperglycemia but without impacting the RAAS or NO. Our findings suggest that PUA modulates the renal response to hyperglycemia at the efferent arteriole. Future placebo-controlled trials in patients with a wider distribution of PUA levels at various stages of kidney and vascular disease progression are required to determine the groups of patients that could ultimately benefit from PUA lowering. Ongoing longitudinal outcome trials, such as the ongoing Protecting Early Renal Function Loss study (ClinicalTrials.gov identifier NCT02017171) (42), will determine whether our physiological findings can be applied to long-term PUA-lowering effects on renal and cardiovascular outcomes in patients with T1D.

Acknowledgments. The authors thank the study participants, whose time and effort are critical to the success of our research program. The authors also thank Takeda for providing the study medications without charge. In addition, the authors thank M.L. Fritzler and Eve Technologies Corporation for performing the cytokine assays for this project.

Funding. Y.L. was supported by a Heart & Stroke/Richard Lewar Centre of Excellence Studentship, a Javenthey Soobiah Scholarship, a Queen Elizabeth II/Dr. Arnie Aberman Scholarship in Science and Technology, a University of Toronto Fellowship in the Department of Pharmacology and Toxicology, and a Canadian Diabetes Association Postdoctoral Fellowship. R.H. was supported by Banting & Best Diabetes Centre Graduate Studentships (University of Toronto), a Hilda and William Courtney Clayton Paediatric Research Fund Award, and an Institute of Medical Science Graduate Student Award. D.Z.I.C. was supported by a Kidney Foundation of Canada Scholarship, a Canadian Diabetes Association-KRESCENT Program Joint New Investigator Award, and funding from the Canadian Institutes of Health Research and the Kidney Foundation of Canada.

Duality of Interest. B.A.P. has received honoraria for continuing medical education events from Boehringer Ingelheim, Janssen, Medtronic, and Abbott. His research institute has received operating funds on his behalf from Medtronic, Novo Nordisk, and Boehringer Ingelheim. He has served as a consultant for Boehringer Ingelheim. D.Z.I.C. has received speaker/consultant honoraria from Boehringer Ingelheim, Eli Lilly, AstraZeneca, Sanofi, Merck, Mitsubishi Tanabe, Abbvie, and Janssen and has received operational funding for clinical trials from Boehringer Ingelheim, Merck, Janssen, and AstraZeneca. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. Y.L., R.H., A.L., V.L., D.F., A.A., B.A.P., and D.Z.I.C. researched data, wrote the manuscript, contributed to discussion, and reviewed and edited the manuscript. All authors have approved the final version of the manuscript. D.Z.I.C. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Prior Presentation. Parts of this study were presented in abstract form at Kidney Week 2016, Chicago, IL, 15–20 November 2016.