SPRINT – A Kidney-Centric Narrative Review

Abstract

Hypertension is a potent cardiovascular risk factor with deleterious end-organ effects and is especially prevalent among patients with chronic kidney disease. The Systolic Blood Pressure Intervention Trial enrolled patients at an elevated cardiac risk including patients with mild to moderate chronic kidney disease and found that an intensive systolic blood pressure goal of < 120 mm Hg significantly reduced the rates of adverse cardiovascular events and all-cause mortality and non-significantly reduced the rates of probable dementia; these results were consistent whether one had chronic kidney disease or not. On the other hand, results of intensive blood pressure therapy on chronic kidney disease progression were inconclusive and there was an increased risk of incident chronic kidney disease and acute kidney injury, but the declines in kidney function appear to be hemodynamically driven and reversible. Overall, an intensive blood pressure target is effective in reducing cardiovascular disease and all-cause mortality and may reduce the risk of probable dementia in patients with mild to moderate CKD. More studies are needed to determine its long-term effects on kidney function.

Introduction

High blood pressure (BP) is one of the leading causes of preventable cardiovascular (CV) disease and death. Chronic kidney disease (CKD) is a risk factor for CV disease and many patients have concomitant high BP. High BP, in turn, is a risk-factor for kidney disease¹. Therefore, successful management of high BP is a top priority for optimizing CV health². The Systolic Blood Pressure Intervention Trial (SPRINT)³ continues to generate a wealth of information regarding how an intensive BP target affects a wide spectrum of health outcomes. The primary goal of SPRINT was to test whether an intensive systolic blood pressure (SBP) goal of < 120 mm Hg compared to a standard SBP goal of < 140 mm Hg improved CV outcomes in patients at high risk for CV disease. SPRINT pre-specified patients with CKD as a subgroup of interest, and SPRINT remains the largest randomized trial of BP targets in CKD to date. This paper provides a narrative review of what we have learned from SPRINT through a kidney-centric lens.

Cohort characteristics by baseline CKD status

SPRINT enrolled patients aged 50 years or older with high BP and an elevated risk of CV events. SPRINT had a particular focus on recruiting patients with CKD, defined as an estimated glomerular filtration rate (eGFR) of 20 to less than 60ml/min per 1.73m², calculated using the four-variable Modification of Diet in Renal Disease (MDRD) equation⁴. Notable exclusions included patients with type 2 diabetes mellitus (DM), as they were recently studied in the recent Action to Control Cardiovascular Risk in Diabetes BP trial (ACCORD BP) trial⁵; patients with polycystic kidney disease, as they were recently studied in the Halt Progression of Polycystic Kidney Disease (HALT-PKD) trial⁶; patients with glomerulonephritis requiring immunosuppressive medications; and patients with proteinuria >1g per day, as prior post hoc analyses suggested slower CKD progression with more intensive BP targets in this subgroup^{3, 7, 8}.

SPRINT enrolled 9361 participants, of which 2646 (28.3%) had baseline CKD. Patients in the CKD subgroup compared to the non-CKD subgroup were generally older, and had a larger proportion of women, participants of non-Hispanic white race and a higher prevalence of clinical cardiovascular disease (Table 1)^{3, 9, 10}. The average eGFR in the CKD subgroup was 47.9 ml/min per 1.73m² and had relatively low levels of albuminuria (mean urinary albumin-to-creatinine ratio [UACR] 80.6 mg/g). Baseline mean BPs between the CKD and non-CKD cohort were similar (139.2/74.9 vs. 139.2 / 79.4 mmHg), although patients with CKD required more medications to achieve the same BP than patients without CKD (2.1 vs. 1.7 agents)^{9, 10}.

Table 1–

Selected Baseline Characteristics of Overall SPRINT cohort and the CKD and non-CKD subgroups^{3, 9–10}

Characteristic	Overall Cohort (N – 9361)	CKD (N - 2646)	Non-CKD (N – 6715)
Age, y, mean ± SD	67.9 ± 9.4	71.9 ± 9.3	66.3 ± 9.0
- Age ≥ 75yr, %	28.2	43.9	21.9
Women, %	35.6	40.0	33.7
Race, %
- non-Hispanic White	57.7	67.2	54.0
- Black Race	29.9	24.1	32.2
- Hispanic Race	10.5	7.2	11.8
- Other	1.9	1.6	2.0
Clinical Cardiovascular Disease*, %	16.7	19.7	14.0
Never Smoked, %	44	45.6	43.5
Serum Creatinine, mg/dL, mean ± SD	1.07 ± 0.34	1.43 ± 0.39	0.93 ± 0.17
eGFR, ml/min per 1.73m2, mean ± SD	71.8 ± 20.6	47.9 ± 9.5	81.2 ± 15.5
- eGFR 45 – 59, %	18.8	66.0	n/a
- eGFR < 45, %	9.5	34.0	n/a
Urinary ACR, mg/g
- mean (SD)	42.6 ± 166.3	80.6 ± 243.4
- median (IQR)		13.3 (6.4 – 43.1)	8.6 (5.5 – 17.1)
BP, mm Hg, mean ± SD
- Systolic	139.7 ± 15.6	139.2 ± 16.1	139.9 ± 15.4
- Diastolic	78.1 ± 11.9	74.9 ± 12.2	79.4 ± 11.6
No. of antihypertensives agents, mean ± SD	1.8 ± 1.0	2.1 ± 1.0	1.7 ± 1.0

^*Clinical CV Disease:

- Previous MI, percutaneous coronary intervention, coronary artery bypass, carotid endarterectomy, carotid stenting.

- Peripheral arterial disease with revascularization.

- Acute coronary syndrome with or without resting ECG changes, ECG changes on a graded exercise test, or positive cardiac imaging study.

- ≥ 50% diameter stenosis of a coronary, carotid, or lower extremity artery.

- Abdominal aortic aneurysm (AAA) ≥ 5cm with or without repair.

CKD – chronic kidney disease, SD – standard deviation, eGFR – estimated glomerular filtration rate, ACR – albumin to creatinine ratio, BP – blood pressure.

During the course of the trial, considerable separation was achieved between the two BP treatments arms. In the CKD subgroup, the mean achieved SBP was 123 mm Hg and 135 mm Hg in the intensive and standard treatment arms, respectively⁹; in the non-CKD subgroup the achieved SBP was 119 and 135 mm Hg, respectively¹⁰. Patients in the intensive arm required an average of 1 additional BP medication (3 versus 2) compared to the standard arm, and more patients used angiotensin converting enzyme inhibitors /angiotensin receptor blockers (ACEI/ARBs) and diuretics in the intensive arm (76.7%, 67.0%, respectively) than in the standard arm (55.2%, 42.9%, respectively)³. These patterns were similar among patients with and without baseline CKD^{9, 10}.

Cardiovascular and death outcomes by baseline CKD status

The primary CV outcome was a composite of non-fatal myocardial infarction (MI), other acute coronary syndromes (ACS), non-fatal stroke, non-fatal acute decompensated congestive heart failure, and CV death³. SPRINT was stopped early after a median of 3.26 of the planned mean 5 years of follow-up when an interim analysis showed an overwhelming benefit of the intensive versus standard BP target on the primary CV outcome: 1.65% per year versus 2.19% per year, HR 0.75, 95% CI: 0.64 – 0.89 (p < 0.001). The benefit of the intensive BP target was also seen for several key secondary outcomes, including death from any cause and the composite of the primary CV outcome or death from any cause (Figure 1)³. Outcomes occurred more frequently in participants with baseline CKD, but rates were lower in the intensive versus the standard BP arm and the benefits of the intensive BP target did not vary by baseline CKD status (Figure 1)^{9, 10}. Yet, the separation in the cumulative event curves for the primary CV outcome occurred approximately 2 years into follow-up among participants with CKD, compared with < 6 months among participants without CKD^{9, 10}. Reasons for this discrepancy are unclear.

Figure 1.

Forest plots of hazard ratios and 95% confidence intervals of intensive vs. standard blood pressure control and their outcomes in SPRINT, stratified by baseline CKD status. Panel A depicts the primary cardiovascular outcome (defined as the composite of myocardial infarction, acute coronary syndrome, stroke, acute decompensated heart failure, and cardiovascular death), all-cause mortality, and the composite of the primary cardiovascular outcome and all-cause mortality. Panel B shows the composite kidney outcome for the CKD cohort (defined as a ≥ 50% decline in eGFR from baseline confirmed on testing 90 days later or reaching end-stage kidney disease), the risk of developing incident chronic kidney disease for the non-CKD cohort, and the risk of developing incident albuminuria. Panel C displays the cognitive outcomes assessed in SPRINT.^{3, 9–10, 33–34}

BP – blood pressure, CV – cardiovascular, HR – hazard ratio, CI – confidence interval, CKD – chronic kidney disease, MCI – mild cognitive impairment.

While the benefits of the intensive BP target did not vary by baseline CKD status when it was modeled as a dichotomous variable (i.e. eGFR < 60 vs. eGFR ≥ 60 ml/min per 1.73m²), a post-hoc analysis stratified the SPRINT cohort into four categories of kidney function: ≥ 90ml/min per 1.73m², 60 - 89 ml/min per 1.73m², 45 – 59 ml/min per 1.73m², and < 45 ml/min per 1.73m^{2 11}. The authors found a progressive attenuation in the benefit of intensive BP lowering on the primary CV outcome with each successive lower category of baseline eGFR; an interaction term between the intervention group and eGFR suggested a potential interaction of the intervention by eGFR modeled as continuous variable (P_int = 0.019). Given the post hoc nature of the analysis and relatively small number of patients in the lowest eGFR category, these results should be considered hypothesis-generating¹². To date, the benefits of intensive BP lowering among patients with more advanced CKD are uncertain and underscores the need for a future trial to specifically target this population.

Patients with CKD also have increased vascular stiffness¹³ which can lead to lower diastolic blood pressures (DBP); aggressively lowering BP could reduce coronary perfusion and lead to an increase in adverse events. A recent post hoc analysis of baseline DBP in the SPRINT CKD subgroup found that patients with lower DBP had higher risks for the primary CV outcome and all-cause mortality than those with higher DBP¹⁴. Yet the beneficial effect of the intensive versus standard BP target was still observed across tertiles of baseline DBP (P_int > 0.2), suggesting that lower DBP should not necessarily deter physicians from pursuing more intensive SBP lowering in these patients.

Beyond eGFR considerations, albuminuria is also an independent risk factor for CV disease¹⁵. A post hoc analysis by Chang et. al.¹⁶ stratified SPRINT participants by baseline albuminuria (UACR < or ≥ 30 mg/g). A UACR of 30mg/g is a well-established cut-off above which increases in albuminuria is directly correlated with an increase in the risk of cardiovascular death and all-cause mortality¹⁵. The SPRINT cohort had 1723 participants with baseline albuminuria (median UACR 70, inter-quartile range (IQR) 42 – 162 mg/g); at baseline, they were generally older, had a lower eGFR, higher SBP, and a higher prevalence of CV disease than participants without baseline albuminuria (median UACR 8, IQR 5 – 13 mg/g). While patients with baseline albuminuria had higher rates of the primary CV outcome and all-cause mortality, the benefit of an intensive versus standard BP target remained consistent regardless of baseline albuminuria status (P_int > 0.05)¹⁶. The consistent benefit of the intensive BP target by baseline albuminuria status was also similar among the CKD and non-CKD subgroups.

SPRINT and CV / death outcomes in context

In contrast to SPRINT, prior studies in CKD did not detect a benefit of more intensive BP lowering on CV outcomes or death. These studies, which included the MDRD¹⁷, African American Study in Kidney Disease (AASK)¹⁸, and the Ramipril Efficacy in Nephropathy – 2 (REIN-2)¹⁹, differed from SPRINT with regard to baseline characteristics, severity of CKD and degree of albuminuria or proteinuria (Table 2). Importantly, these prior trials were all primarily designed to assess the effect of more intensive BP lowering on CKD progression, and therefore were not powered to detect differences in CV outcomes^17–19. There has been longer-term follow-up examining the risk of death for two of these trials. In an extended observational follow-up analysis of AASK (median follow-up 14.4 years) participants in the intensive BP arm had lower adjusted risk of death than participants in the standard BP arm (HR 0.81, 95% CI 0.68 to 0.98; p - 0.03)²⁰. Similarly, an observational long-term follow-up of the MDRD (median follow-up 19.3 years) showed lower risk of death in the intensive versus standard BP arm (HR 0.82, 95% CI 0.68 to 0.98, P - 0.03)²¹.

Table 2–

Baseline Characteristics and Blood Pressure Targets of Trials in CKD^{3, 5, 17–19, 22}

Characteristic	MDRD	AASK	REIN-2	ACCORD (CKD cohort)	SPRINT (CKD cohort)
Sample Size, n	840 - Mod. CKD Cohort – 585 - Severe CKD Cohort – 255	1094 - UPCR ≤ 0.22g/g – 733 - UPCR ≥ 0.22g/g – 357	335	CKD – 1726 - CKD I – 693 - CKD II – 632 - CKD III – 401	CKD* – 2646 - eGFR 45 - <60 – 1759 - eGFR <45 – 891
Follow-up, y	Mean – 2.2 (0 to 3.7)	Median – 3.8 (3 to 6.4)	Med. (IQR) – 1.6 (1.0 – 2.9)	Mean – 4.7	Med. (IQR) – 3.3 (2.8 – 3.8)
Study Pop.	Non-Diabetic‡	Hypertensive CKD††	Non-Diabetic	Diabetic	Non-Diabetic
Age, y (mean ± SD)	Int. BP arm – 51.8 ± 12.4 Std. BP arm – 52.0 ± 12.2	Int. BP arm – 54.5 ± 10.9 Std. BP arm – 54.7 ± 10.4	Int. BP arm – 54.6 ± 14.7 Std. BP arm – 53.1 ± 15.8	Overall - 63.2 ± 7.3	Overall - 71.9 ± 9.3
GFR or eGFR Criteria†	Mod. CKD Cohort – 25 – 55 Severe CKD Cohort – 13 – < 25	20 – 65	< 75	N/A	20 - < 60
GFR, eGFR†, mean ± SD	Int. BP arm – 32.7 ± 12.1 Std. BP arm – 32.3 ± 11.9 Mod. CKD Cohort – 38.6 ± 8.9 Severe CKD Cohort – 18.5 ± 3.4	Int. BP arm – 46.0 ± 12.9 Std. BP arm – 45.3 ± 13.2	Int. BP arm – 38.6 ± 17.9 Std. BP arm – 38.9 ± 21.7	91.6 ± 28.8 (overall cohort)	Overall - 47.9 ± 9.5
sCr, mg/dL, mean ± SD	Mod. CKD Cohort – 1.9 ± 0.5 Severe CKD Cohort – 3.4 ± 0.9	2.00 ± 0.70	2.7 ± 1.1	1.0 ± 0.3	1.43 ± 0.39
Proteinuria Criteria	> 10g/d excluded	> 2.5g/d excluded	< 1g/d excluded	> 1g/d excluded	> 1g/d excluded
Proteinuria, median (IQR)	Overall: (g/d) 0.32 (0.07 – 1.5) Mod. CKD Cohort - 0.2 (0.0 – 2.9) Severe CKD Cohort – 0.8 (0.1 – 3.9)	UPCR, Overall: (g/g) 0.08 (0.03–0.37) UPCR ≥ 0.22g/g subgroup: (g/g) - Int. BP arm – 0.96 (0.35–1.99) - Std. BP arm – 0.73 (0.42–1.37)	(Mean ± SD, g/d) Int. BP arm – 2.8 ± 1.9 Std. BP arm – 2.9 ± 1.9	UACR (mg/g) 14.3 (6.9 – 44.8) Data for entire cohort	UACR - 12.8 (6.5 – 42.6) mg/g UACR, mean ± SD - 80.9 ± 236.2 mg/g
Target BP, mmHg	Int. BP arm – MAP ≤ 92 (~125/75) - Age >60 - MAP ≤ 98 (~135/80) Std. BP arm – MAP ≤ 107 (~140/90) - Age >60 - MAP ≤ 113 (~150/95)	Int. BP arm – MAP ≤ 92 (~125/75) Std. BP arm – MAP 102–107 (~135/85 –140/90)	Int. BP arm - < 130/80 Std. BP arm – DBP < 90	Int. BP arm – SBP < 120 Std. BP arm – SBP < 140	Int. BP arm – SBP < 120 Std. BP arm – SBP < 140
Achieved BP, mmHg, mean ± SD	Int. BP arm – MAP 93.0 ± 7.3 (~125/77) Std. BP arm – MAP 97.7 ± 7.7 (~133/80) Diff, mean – MAP 4.7 (~8/3)	Int. BP arm – MAP 95 (~128/78) Std. BP arm – MAP 104 (~141/85) Diff, mean – MAP 9 (~13/7)	Int. BP arm – 130/80 Std. BP – 134/82 Diff, mean – 4/2	Int. BP arm – SBP - 122 Std. BP arm – SBP - 134 SBP Diff, mean – 12	Int. BP arm – 123/66 Std. BP arm – 135.3/72.4 SBP Diff, mean – 12.3
ACEI / ARB use (Int. vs. Std.)	51 vs. 32%	Similar in both arms	100% in both arms	41 vs. 30% (Data for entire cohort)	71.7% vs. 57%

^*eGFR < 60ml/min per 1.73m²

^†All GFR, CrCl units are ml/min, and eGFR units are ml/min per 1.73m²

^‡The study population was mostly non-diabetic. Diabetics that did not require insulin therapy were allowed but only 26 of 840 participants had diabetic nephropathy.

^††Presumed hypertensive CKD. Patients were only included if they had hypertension, were non-diabetic, and had no other identified causes of renal dysfunction.

sCr – Serum Creatinine Concentration, UPCR – Urine protein-creatinine ratio, BP – blood pressure. SBP – systolic blood pressure, Int. – Intensive, Std. – Standard, Diff. – Difference. Mod. – Moderate, Med. – Median, IQR – inter-quartile range

The BP intervention in SPRINT mirrors that of ACCORD-BP, which also randomized patients to an intensive SBP goal of <120mmHg or a standard SBP goal of < 140mmHg⁵. The primary difference is that ACCORD-BP included only patients with type 2 DM, which was an exclusion criterion for SPRINT. In contrast to SPRINT, ACCORD-BP failed to find a significant benefit of intensive BP lowering on the primary CV outcome (HR 0.88, 95% CI: 0.73 – 1.06, P - 0.20) and even a non-significant increase in the risk for all-cause mortality (HR 1.07, 95% CI: 0.85 – 1.35, P - 0.55). Given that ACCORD-BP excluded participants with a serum creatinine concentration > 1.5 mg/dL, the cohort had a mean eGFR of 91ml/min per 1.73m² and few participants had an eGFR < 60 ml/min per 1.73m² (Table 2)^{5, 22}. Despite the milder phenotype, the ACCORD BP CKD subgroup experienced 1.5 to 2 times the number of CV events as compared to the non-CKD subgroup, and no differences in the effect (or lack thereof) of intensive BP control were observed among participants with and without baseline CKD²².

One postulated reason for ACCORD-BP’s discordant results with SPRINT is that ACCORD-BP had less statistical power²³ given its smaller sample size and lower than expected primary event rate [ACCORD 2010]. Second, ACCORD had a 2x2 factorial design that also randomized participants to intensive versus standard glycemic control. A clever post hoc analysis²⁴ found evidence of interaction between intensive glycemic control and the intensive BP target: taking only participants in the standard glycemic arm, intensive BP control did reduce the risk of the primary CV outcome, with an effect size similar to that which was seen in SPRINT (HR 0.77, 95% CI: 0.63 – 0.95)^{3, 24}.

Summary: SPRINT and CV / death outcomes in CKD

Targeting a lower SBP is beneficial for patients with mild to moderate CKD for reducing the risk of CV events and death. These conclusions are reflected in the recent 2021 KDIGO Clinical Practice Guidelines for the Management of BP in CKD, which recommends that adults with high BP and CKD be treated with a target SBP of < 120 mm Hg, when tolerated, and using standardized office BP measurement²⁵. However, there are certain subgroups where the benefit is less certain, including in patients with heavy proteinuria, more advanced CKD, and any other population that was not well represented in SPRINT.

Kidney outcomes by baseline CKD status

Kidney outcomes in the CKD cohort

SPRINT pre-specified a composite kidney outcome as one of its secondary outcomes, and its definition depended on baseline CKD status³. For participants with baseline CKD, the composite kidney outcome was defined as a sustained ≥50% decrease in eGFR from baseline (confirmed with follow-up testing 90 days later) or end-stage kidney disease (ESKD) requiring dialysis or kidney transplant. Participants in the intensive BP arm had a decline in eGFR observed by the first month of follow-up that stabilized by the sixth month, while participants in the standard BP arm had an increase in eGFR and subsequent stabilization during this same time period⁹. Despite this rapid and early divergence in eGFR of the two treatment arms, analyses of eGFR decline from the sixth month and onwards did not show any clinically significant differences between the two arms, and there was no significant difference in the composite kidney outcome (Figure 1)⁹. However, SPRINT participants in the CKD subgroup had relatively mild CKD and minimal albuminuria and therefore had relatively slow CKD progression. Adding in the early trial termination, it is not surprising that only 31 participants (1.2%) in the CKD subgroup experienced the composite renal outcome, with 16 requiring long-term dialysis and none receiving kidney transplantation⁹.

Incident albuminuria within the CKD cohort tended to be lower in the intensive BP arm, although the difference did not quite reach statistical significance (HR 0.72, 95% CI: 0.48 – 1.07, P = 0.11). Yet, levels of urinary albuminuria were significantly lower in the intensive BP group throughout the entire trial⁹.

Kidney outcomes in the non-CKD cohort

For participants without baseline CKD, the kidney outcome, incident CKD, was defined as a confirmed ≥ 30% decrease in eGFR to < 60 ml/min per 1.73m². In this subgroup, a similar rapid early divergence in the eGFR followed by stabilization at month 6 was seen between the two treatment arms, with an average eGFR difference of 4-5ml/min per 1.73m² after 6 months¹⁰. In contrast to the CKD subgroup, however, in this non-CKD subgroup, 140 participants (4.2%) in the intensive BP arm and 40 (1.2%) in the standard BP arm developed incident CKD (HR 3.54, 95% CI 2.50-5.02; Figure 1)¹⁰. While the majority of the extra kidney outcomes in the intensive arm were due to the acute eGFR decline in the first 6 months, rates of a subsequent 30% eGFR decline from 6 months to the end of the trial were still significantly higher in the intensive arm (HR 2.50, 95% CI: 1.57 – 4.11; P < 0.001)¹⁰. Yet among the 140 participants in the intensive BP arm with incident CKD, by the end of the trial, 40% improved to a <30% decline in eGFR. Among the 40 participants in the standard BP arm with incident CKD, 25% improved to a <30% decline in eGFR¹⁰. No patients in the non-CKD cohort developed ESKD requiring dialysis or kidney transplant. Again, this suggests that the decrease in eGFR was predominantly hemodynamically related and reversible, as incident albuminuria, a marker of kidney damage, was actually lower in the intensive arm, although the results did not quite reach statistical significance (Figure 1).

Kidney outcomes and baseline albuminuria

Just as Chang et. al.’s stratified analysis by baseline albuminuria status showed higher rates of the primary CV outcome and death among patients with albuminuria at baseline, patients with pre-existing albuminuria also experienced higher rates of the SPRINT kidney outcome¹⁶. While the absolute number of kidney events were low, patients with prior albuminuria played an out-sized role contributing to the respective kidney outcomes, especially within the CKD subgroup.

When stratified by baseline albuminuria status, intensive BP control significantly increased the relative risk of a ≥ 40% decline in eGFR in patients without baseline albuminuria (HR 4.55, 95% CI: 2.37 – 8.75), but this relative effect was significantly attenuated in patients with albuminuria (HR 1.48, 95% CI: 0.91 – 2.39) (P_int < 0.001)¹⁶. This pattern was observed even more so in the CKD subgroup as the number of patients with baseline albuminuria who experienced a ≥ 40% eGFR decline was similar between the intensive and standard BP arms at 25 and 22 patients respectively (HR 1.07, 95% CI: 0.60 – 1.90)¹⁶. Given that patients in the intensive arm experienced an acute drop in eGFR while patients in the standard BP arm did not, the fact that rates of eGFR decline were similar between the two arms suggest that patients with baseline albuminuria experience faster rates of CKD progression with the standard BP target and derive a relatively greater benefit with intensive BP control.

Acute Kidney Injury events in SPRINT

Related to the secondary kidney outcomes is the occurrence of acute kidney injury (AKI), a serious adverse event in SPRINT. While rates of serious adverse events overall were similar among the intensive and standard BP arms (38.3% versus 37.1%, P=0.25), AKI occurred more often in the intensive compared with the standard BP arm (4.4% versus 2.6%, P<0.001)³. A subsequent analysis of AKI in SPRINT examined whether there was recovery of kidney function after an AKI event, which was determined by comparing the peak recorded serum creatinine value with the lowest outpatient SPRINT creatinine value obtained in the subsequent 365 days. Kidney recovery was categorized as complete (within 20% of pre-AKI values) or partial (recovery to within 30% of pre-AKI values)²⁶. Reassuringly, they found complete resolution of the AKI for 90.4% events in the intensive arm and 86.9% of events in the standard arm.

SPRINT and kidney outcomes in context

The results of the kidney outcomes in SPRINT were largely inconclusive within the CKD cohort and even showed a possible harm with an intensive BP target in patients without CKD due to the rapid initial drop in eGFR. Yet, the rapid initial fall in eGFR with intensive BP lowering mirrors the results seen in other trials testing interventions that ultimately proved to be beneficial, including the Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan (RENAAL) study, one of the major trials that led to ARBs becoming the standard or care for treatment of proteinuric kidney disease²⁷, and the more recent sodium-glucose cotransporter 2 inhibitors (SGLT2is) trials^{28, 29}. These results highlight the inherent challenges of conducting a CKD progression study population that is at a lower risk of CKD progression such as in SPRINT, with only mildly reduced eGFR and amounts of albuminuria, where a prohibitively long follow-up time would be required to make a more definitive conclusion about the effect of intensive BP control on kidney outcomes.

As noted above, the MDRD, AASK and REIN-2 studies were designed to examine the effect of more intensive BP control on CKD progression. Like SPRINT, all three trials enrolled participants with non-diabetic kidney disease. Yet, they also had notable differences in their baseline characteristics, severity of kidney disease, and study-design (Table 2)^{8, 17, 19}. Compared to the SPRINT CKD subgroup, the AASK population was younger, all self-identified as African American, and only included patients with presumed hypertensive CKD¹⁸ Like SPRINT, AASK enrolled patients with mild to moderate CKD, with mean eGFR of 45 to 50 ml/min per 1.73m² and low levels of albuminuria. The MDRD and REIN-2 participants were younger than the SPRINT CKD cohort but had more advanced CKD, with mean eGFR of 30 to 40ml/min per 1.73m² and heavier albuminuria.^{17, 19} Not surprisingly, 13 percent of MDRD participants developed ESKD during the trial while only 0.6% of the SPRINT CKD cohort reached ESKD.^{9, 17} The MDRD and AASK studies also used mean arterial pressure (MAP) targets while more recent BP targeting trials including SPRINT focused on SBP targets.

Despite these differences, like SPRINT, none of these previous trials showed a significant benefit of a more intensive BP target on their respective kidney outcomes. However, observational follow-up studies of the MDRD and AASK trials showed a benefit of intensive BP lowering for reducing the rate of CKD progression^{8, 30}. A subsequent meta-analysis of the MDRD and AASK studies with a median follow-up of 14.9yrs continued to find a marginal benefit of intensive BP lowering on the outcome of ESKD requiring dialysis or kidney transplantation (HR 0.88, 95% CI 0.78-1.00)²⁰.

Furthermore, both the MDRD and AASK trials and subsequent follow-up studies demonstrated that patients with higher levels of proteinuria experienced fasters rates of kidney function decline and derived a greater benefit from an intensive BP target^{7, 8, 18, 30, 31}. While the overall low kidney event rates in SPRINT preclude making a definitive verdict, there was also a trend within SPRINT towards a similar conclusion.

Summary: SPRINT and kidney outcomes

Taken together, these results suggest that the initial decline in eGFR with more intensive BP lowering observed in SPRINT is likely due to the reversible effects of hemodynamic changes rather than true intrinsic kidney damage. Further support for a hemodynamic etiology is provided by a longitudinal urinary biomarker study³². A random sample of 978 SPRINT participants with baseline CKD had stored urine samples analyzed for biomarkers of tubular injury at baseline, year 1 and year 4 of follow-up. The results showed no increase for any of the urine biomarkers in the intensive versus standard BP arm for the duration of the study, despite having a lower eGFR. While the aggregate data favors a lower BP target in patients with CKD and proteinuria, it remains to be seen whether intensive BP lowering overall will ultimately prove beneficial, detrimental, or neutral with regard to kidney outcomes.

Cognitive outcomes by baseline CKD status

A key secondary outcome in SPRINT assessed whether intensive vs. standard BP lowering reduced the incidence of probable dementia, mild cognitive impairment (MCI) or the composite of both outcomes³³. After a median follow-up of 5.1 years, the results showed that an intensive BP target non-significantly lowered the risk of probable dementia and significantly reduced the incidence of MCI and the composite of probable dementia and MCI (Figure 1)³³. As with the other outcomes, the incidence of probable dementia, MCI, and their composite were higher in patients with CKD compared with the non-CKD subgroup. Yet, results were generally similar for participants with and without baseline CKD defined by eGFR criteria, although intensive vs. standard BP control may have a larger benefit on reducing rates of MCI in patients without CKD than with CKD (P_{interaction =} 0.04) (Figure 1). Exploratory modeling of eGFR as a continuous variable suggested that the benefit of intensive vs. standard BP control may be reduced or even harmful as eGFR declines, although these analyses are only hypothesis-generating³⁴. Tests for treatment effect heterogeneity between patients with and without albuminuria were not significant for cognitive outcomes, although the effect estimates were more attenuated among participants with baseline albuminuria.

Summary: SPRINT and cognitive outcomes

Because the trial was terminated early and cognitive events occurred less frequently than anticipated, the trial was likely underpowered to detect significant treatment effects and treatment heterogeneity between patients with CKD and non-CKD for dementia as ascertaining cognitive outcomes usually requires longer periods of follow-up³⁴. Yet, the aggregate data brought forth by SPRINT³³ and the kidney-focused secondary analysis³⁴ highlights the following: patients with baseline CKD and albuminuria have higher crude rates of adverse cognitive outcomes than those without baseline CKD and albuminuria; an intensive vs. standard BP target tends to reduce the rates of dementia and MCI in patients with and without CKD. Questions remain about the benefits of intensive BP control in patients with more advanced CKD and higher degrees of albuminuria. Importantly, concerns of cerebral hypoperfusion that could potentially arise from a lower BP target and lead to faster cognitive decline³⁵ were not observed in SPRINT.

Conclusions

Since the publication of the main trial results in 2015³, SPRINT has continued to provide a wealth of information to help guide BP management and advance our understanding of how intensive BP targets affect a wide variety of important clinical outcomes. In this kidney-centric narrative review of SPRINT, we conclude that the data support more intensive SBP targets in patients with CKD to reduce the occurrence of CV events and death and possibly to reduce the occurrence of cognitive events. Intensive BP lowering appears to confer short-term reductions in eGFR, particularly in patients without CKD, but its effect on longer-term kidney outcomes remains to be determined in future trials. As with all things in medicine, treatment decisions must be individualized for each particular patient.