ABSTRACT

Delaying disease progression and reducing the risk of mortality are key goals in the treatment of chronic kidney disease (CKD). New drug classes to augment renin–angiotensin–aldosterone system (RAAS) inhibitors as the standard of care have scarcely met their primary endpoints until recently. This systematic literature review explored treatments evaluated in patients with CKD since 1990 to understand what contemporary data add to the treatment landscape. Eighty-nine clinical trials were identified that had enrolled patients with estimated glomerular filtration rate 13.9–102.8 mL/ min/1.73 m2 and urinary albumin-to-creatinine ratio (UACR) 29.9–2911.0 mg/g, with (75.5%) and without (20.6%) type 2 diabetes (T2D). Clinically objective outcomes of kidney failure and all-cause mortality (ACM) were reported in 32 and 64 trials, respectively. Significant reductions (P \ 0.05) in the risk of kidney failure were observed in seven trials: five small trials published before 2008 had evaluated the RAAS inhibitors losartan, benazepril, or ramipril in patients with (n = 751) or without (n = 84–436) T2D; two larger trials (n = 2152–2202) published onwards of 2019 had evaluated the sodium-glucose co-transporter 2 (SGLT2) inhibitors canagliflozin (in patients with T2D and UACR [ 300–5000 mg/g) and dapagliflozin (in patients with or without T2D and UACR 200–5000 mg/g) added to a background of RAAS inhibition. Significant reductions in ACM were observed with dapagliflozin in the DAPA-CKD trial. Contemporary data therefore suggest that augmenting RAAS inhibitors with new drug classes has the potential to improve clinical outcomes in a broad range of patients with CKD.

Key Summary Points

Why carry out this study?

Morbidity, mortality, and the economic burden from chronic kidney disease (CKD) are growing worldwide.

This systematic literature review examined contemporary clinical trial data relative to the overall CKD treatment landscape to view the impact of novel drug classes following 20 years of little to no innovation.

What was learned from the study?

Augmenting the standard of care with canagliflozin or finerenone could significantly improve clinical outcomes in patients with type 2 diabetes (T2D).

Augmenting the standard of care with dapagliflozin could significantly improve clinical outcomes regardless of T2D status and is the only agent that has been shown to significantly reduce all-cause mortality risk.

Composite and surrogate endpoints in clinical trials have varied widely over time, likely due to changing guidelines, and may benefit from standardization.

INTRODUCTION

An estimated 840 million people worldwide have chronic kidney disease (CKD) [1], which was responsible for 1.2 million deaths and 35.8 million disability-adjusted life years in 2017 [2]. However, only 12% of sufferers are aware of their condition [3]. CKD is diagnosed when the estimated glomerular filtration rate (eGFR) declines below 60 mL/min/1.73 m2 or the urinary albumin-to-creatinine ratio (UACR) equals or exceeds 30 mg/g for 3 months or longer [4]. As CKD progresses, healthcare costs increase and health-related quality of life (HRQoL) diminishes, with the greatest costs and HRQoL burden associated with kidney failure (eGFR \15 mL/min/1.73 m2 ) [5, 6]. Adverse clinical outcomes, healthcare utilization and costs, and disease burden also increase as albuminuria worsens [7–9], and UACR 30–300 mg/g (moderately increased) and even [300 mg/g (severely increased) are now considered important predictors of risk for CKD progression, cardiovascular events, and mortality [4]. Early identification and pharmacologic intervention could therefore delay or prevent CKD progression.

Current guidelines recommend using renin–angiotensin–aldosterone system (RAAS) inhibitors (either an angiotensin-converting enzyme inhibitor or an angiotensin receptor blocker) to delay or prevent CKD progression [4]. Clinical trials of other drug classes to augment RAAS inhibitors, delay progression, and improve outcomes have scarcely met their primary endpoints [10], except for sodium-glucose co-transporter 2 (SGLT2) inhibitors. Initially developed as blood glucose-lowering agents, reports of renal and cardiovascular benefits in patients with type 2 diabetes (T2D) [11–14] as well as cardiovascular benefits in patients with heart failure (HF) [15–17] have prompted the evaluation of SGLT2 inhibitors in patients with CKD who are already receiving standard of care treatment with RAAS inhibitors.

This systematic literature review explored the treatments evaluated in patients with CKD since 1990 to allow an assessment of contemporary data relative to the overall treatment landscape.

METHODS

This systematic literature review was conducted according to the recommendations of Cochrane [18], the Centre for Reviews and Dissemination [19], and the National Institute for Health and Care Excellence [20]. The protocol has been registered on PROSPERO (CRD42020190152).

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Data Sources and Searches

Using the terms listed in the Supplementary Material, we searched MEDLINE, Embase, and the Cochrane Library for peer-reviewed articles published between 1990 and November 2, 2020, that reported results from prospective, paralleldesign randomized controlled trials that evaluated pharmacologic treatments for patients aged 18 years or more with CKD and albuminuria. Search filters for MEDLINE and Embase were obtained from the Scottish Intercollegiate Guideline Network [21], and adapted for Embase by Cochrane [22]. In line with guidelines for the inclusion of gray literature [18–20, 23], the proceedings of key international conferences and trial registries were also searched (Supplementary Material).

Non-English-language publications, reviews, case studies, case reports, conference proceedings (other than those identified in the search described above), and animal studies were excluded.

Trial Selection

After removing duplicates from the combined search results, two independent reviewers screened the identified abstracts against prede- fined eligibility criteria (Table 1). Abstracts deemed eligible for inclusion were then compared and any discrepancies resolved mutually or by a third reviewer. This independent double-review process was repeated on the full-text articles to identify a final list of trials eligible for inclusion in this review.

Data Extraction and Quality Assessment

Data were extracted by one reviewer and validated by a second, with disagreements resolved by a third (Supplementary Material). Binary variables included trial population, number or proportion of patients experiencing an event, and incidence rates per population or persontime. Continuous and time-to-event variables included hazard ratio (HR), odds ratio, relative risk, mean, median, standard deviation, standard error, range, 95% confidence interval (CI), interquartile range, and P value. Outcomes reported without P values or 95% CIs were assumed not to be statistically significant. Outcomes reported with P \ 0.05 or with 95% CIs not crossing 1.0 for a HR or relative risk were assumed to be statistically significant.

Risk of bias and quality of reporting were assessed using eight questions from the PMG24 Company Evidence Submission Template (NICE single technology appraisal process) [24], developed based on previous recommendations [19]. Answers of ‘‘yes,’’ ‘‘no,’’ or ‘‘unclear due to inadequate reporting’’ were required. Depending on the question, answers of ‘‘yes’’ or ‘‘no’’ could indicate a higher or lower risk of bias (Supplementary Material).

Compliance with Ethics Guidelines

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

RESULTS

Search Results

Overall, 40,550 records were identified (Fig. 1).

After removal of 20,773 duplicates, 19,777 abstracts were reviewed against eligibility criteria, and 19,557 were excluded. The full texts of 220 articles were reviewed, and 121 were excluded (Table S1). The addition of one more article, identified during a search of conference proceedings, resulted in 100 eligible articles providing data for 89 randomized controlled trials (Table 2).

Trial Characteristics

Thirty-seven trials were multinational, 18 were conducted in Japan, and seven each were conducted in China and Italy, with the remaining trials conducted in a range of countries worldwide.

Antihypertensive agents were the most common intervention assessed overall, but were approximately twice as common in trials of patients without T2D (88%) than trials of patients with (42%), or with or without (45%) T2D. Blood glucose-lowering agents were also common in trials of patients with T2D (37%).

The most common comparators were placebo in trials of patients with T2D (53%) and active comparators in trials of patients without (53%), or with or without (38%) T2D. Placebo was also common in trials of patients without (35%), or with or without (34%) T2D (Fig. S3a, b).

Baseline Patient Characteristics

In more than 80% of trials, 50–100% of patients were male (Fig. S4). Mean age ranges were 51.0–72.1 years in trials of patients with or without T2D (except one trial with a mean age range of 34–35 years [82]), 53.8–70.2 years in trials of patients with T2D (except one trial with a mean age range of 34.0–35.0 years [66], and one trial with a median age of 33 years [65]), and 44.4–71.0 years in trials of patients without T2D.

While CKD etiologies other than diabetic nephropathy were infrequently reported in trials of patients with T2D, 13 trials (14.6%) of patients without T2D and 16 (18.0%) of patients with or without T2D reported glomerulonephritis as a key CKD etiology (Table S2a, b). Mean eGFR ranged between 13.9 and 102.8 mL/min/1.73 m2 , including two trials that enrolled patients with mean eGFR [ 90 mL/min/1.73 m2 (Table S3) [38, 59]. Trials most commonly reported albuminuria as UACR (50.6%), with mean UACR ranging between 29.9 and 2911.0 mg/g. Other trials reported UACR via categorization into normo-, micro-, or macroalbuminuria (16.9%), albumin excretion rate (12.4%), protein excretion rate (20.2%), protein-to-creatinine ratio (18.0%), or urinary albumin value (13.5%) (Table S4a–f). Thirty-one trials (34.8%) included patients with prior histories of cardiovascular disease, with the proportion of patients ranging from 1.7% to 92.0%, although cardiovascular disease history was either inconsistently defined or not defined at all (Table S5). Fourteen trials (15.7%) included patients with HF, with the proportion of patients ranging from 0.6% to 43.1% (Table S6). Eighty-two trials (92.1%) reported systolic and diastolic blood pressure (Table S7).

Composite Outcomes

Fifty-seven composite endpoints were identi- fied, only 13 of which were used in more than one trial (Fig. S5a, b). Composite outcomes are summarized in Table S8.

Twelve trials (13.5%) reported significant reductions in the risks of composites comprising kidney failure plus one or more of doubling of serum creatinine, eGFR reduction (C 40% or C 50%), mortality (all-cause, renal, or cardiovascular), myocardial infarction (MI), stroke, albuminuria progression, or other (Table 3). These included trials published before 2013 evaluating RAAS inhibitors losartan (RENAAL, ROAD) [61, 112], ramipril (REIN-1, AASK) [115, 123], irbesartan (IDNT) [62], valsartan (KVT) [83], and benazepril (ROAD, and an unnamed trial) [112, 113] in patients with, without, or with or without T2D. Also included were trials published onwards of 2019 evaluating dipeptidyl peptidase 4 inhibitor linagliptin (CARMELINA) [34], endothelin A receptor antagonist atrasentan (SONAR) [47], and the non-steroidal mineralocorticoid receptor antagonist finerenone (FIDELIO-DKD) [26] in patients with T2D, as well as the SGLT2 inhibitor canagliflozin (CREDENCE) [27] in patients with T2D and UACR [ 300–5000 mg/g.

Another SGLT2 inhibitor, dapagliflozin, signifi- cantly reduced the risk of composite endpoints comprising kidney failure and at least 50% eGFR reduction plus cardiovascular and/or renal mortality in patients with or without T2D and UACR 200–5000 mg/g (DAPA-CKD) [25]. Kid-ney failure as an independent outcome is reported below.

Four trials (4.5%) reported significant reductions in the risks of composites comprising cardiovascular mortality without kidney failure, plus at least one of doubling serum creatinine, renal mortality, MI, stroke, hospitalization for HF, or hospitalization for HF or unstable angina (Table 3). These included the CARMELINA [34], FIDELIO-DKD [26], and CREDENCE [27] trials, as well as the DAPA-CKD trial of dapagliflozin, which significantly reduced the risk of a composite endpoint comprising cardiovascular mortality and hospitalization for HF [25]. Conversely, the risk of a composite endpoint comprising cardiovascular mortality and hospitalization for HF or unstable angina increased in the BEACON trial of bardoxolone methyl, a nuclear 1 factor (erythroid-derived 2)- related factor 2 activator, although patients in this trial had CKD stage 4, T2D, and median UACR 320 mg/g [73].

Renal Outcomes

Kidney Failure

Kidney failure (previously end-stage kidney disease or end-stage renal disease [124]) ensues when eGFR declines below 15 mL/min/1.73 m2 (CKD stage 5) and the patient requires kidney replacement therapy (previously renal replace￾ment therapy [124]) in the form of a transplant or dialysis [4].

Thirty-two trials (36.0%) reported numbers of patients progressing to kidney failure (Table S9). Significant risk reductions were observed in seven trials (7.9%): the RENAAL trial of losartan in patients with T2D and UACR C 300 mg/g (P = 0.002) [61], the ROAD trial of optimal antiproteinuric doses of losartan (P = 0.046) and benazepril (P = 0.042) in patients without T2D [112], an unnamed trial of conventionally dosed benazepril in patients without T2D (P = 0.02) [113], the REIN-1 and AASK trials of ramipril in patients without T2D (both P = 0.01) [115, 118], the CREDENCE trial of canagliflozin (P = 0.002) [27], and the DAPA￾CKD trial of dapagliflozin (HR 0.64; 95% CI 0.50–0.82) [25].

Dialysis and Transplantation

Dialysis, kidney transplantation, or both were reported in 17 (19.1%), seven (7.9%), and two trials (2.2%), respectively (Table S10). Signifi- cant outcomes were limited to three trials (3.4%). The lipid-lowering agent probucol lengthened mean time to starting dialysis in a trial of patients with T2D and UACR [ 300 mg/ g (P = 0.009) [71], and the number of patients starting dialysis was significantly reduced in a trial of patients without T2D receiving the RAAS inhibitor captopril (P \0.005) [121], as well as patients receiving dapagliflozin in the DAPA￾CKD trial (HR 0.66; 95% CI 0.48–0.90) [25].

Kidney Function Decline

Percentage eGFR declines, mean eGFR declines, and final eGFR measurements at end of follow￾up were reported in 11 (12.4%), 30 (33.7%), and 25 (28.1%) trials, respectively (Table S11a–c).

The number of patients reaching an eGFR decline of 50% was significantly reduced in four trials (4.5%): the SONAR trial of atrasentan in patients with T2D and UACR 300–5000 mg/g (P = 0.038) [47], the LORD trial of lipid-lower￾ing agent atorvastatin in patients with or with￾out T2D (P = 0.023) [95], and the DAPA-CKD trial of dapagliflozin (HR 0.53; 95% CI 0.42–0.67) [25]. In the PREDICT trial of ery￾thropoiesis-stimulating agent darbepoetin alfa, the number of patients without T2D reaching an eGFR decline of 50% was also significantly reduced among those targeting a higher (11–13 g/dL) versus lower (9–11 g/dL) hemo￾globin level (P = 0.008); however, targeting a higher hemoglobin level did not improve kid￾ney outcomes overall [106]. The number of patients reaching an eGFR decline of at least 40% was significantly reduced in the FIDELIO￾DKD trial of finerenone (HR 0.81; 95% CI 0.72–0.92) [26].

Twenty trials (22.5%) reported numbers of patients doubling their serum creatinine (Table S12). Significant risk reductions were observed in seven trials (7.9%): the SONAR trial of atrasentan (P = 0.0055) [47], the FIDELIO￾DKD trial of finerenone (HR 0.68; 95% CI 0.55–0.82) [26], the RENAAL trial of losartan (P = 0.006) [61], the ROAD trial of optimal antiproteinuric doses of losartan (P = 0.040) and benazepril (P = 0.041) [112], an unnamed trial of conventional doses of benazepril (P = 0.02) [113], the IDNT trial of irbesartan (P \0.001 vs amlodipine, P = 0.003 vs placebo) [62], and the CREDENCE trial of canagliflozin (P \0.001) [27].

Cardiovascular Outcomes

Heart Failure

Fourteen trials (15.7%) reported incidences of HF (Table S13), with significant reductions observed in two trials (2.2%): the ASCEND trial of endothelin type A receptor antagonist avosentan in patients with T2D (P = 0.008 with a 25-mg dose, P = 0.05 with a 50-mg dose) [57] and the IDNT trial of irbesantan (P = 0.004 vs amlodipine, P = 0.048 vs placebo) [64].

Hospitalization for HF or Unstable Angina

Hospitalization for HF and hospitalization for unstable angina were reported in 10 (11.2%) and two trials (2.2%), respectively (Table S14). Significant reductions in hospitalization for HF were observed in two trials (2.2%): the RENAAL trial of losartan (P = 0.005) [61] and the CRE￾DENCE trial of canagliflozin (P \ 0.001) [27]. Conversely, bardoxolone methyl significantly increased hospitalization for HF in the BEACON trial (P \0.001) [73].

MI and Stroke

Twenty-four trials (27.0%) reported acute, non￾fatal, or fatal MI, and 25 trials (28.1%) reported non-fatal or fatal stroke (Tables S15 and S16). A significant reduction in MI was observed in patients receiving the calcium channel blocker amlodipine in the IDNT trial (P = 0.021 vs pla￾cebo) [64]. A significant reduction in non-fatal stroke was observed in the SONAR trial of atrasentan (P = 0.0021) [47], and significant reductions in ischemic (P = 0.0073) or any stroke (P = 0.01) were observed in the SHARP trial of a combination of lipid-lowering agents simvastatin and ezetimibe in patients with or without T2D [93]. Conversely, a significant increase in fatal or non-fatal stroke was observed in the TREAT trial of patients with CKD stages 3–4 and T2D receiving darbepoetin alfa (P \ 0.001) [75].

Mortality Outcomes

All-Cause Mortality

Sixty-three trials (70.8%) reported all-cause mortality (ACM) (Table S17), with a significant reduction observed in the DAPA-CKD trial of dapagliflozin (P = 0.004) [25].

Cardiovascular and Renal Mortality

Cardiovascular and renal mortality were repor￾ted in 18 (20.2%) and nine trials (10.1%), respectively, with no significant outcomes observed (Table S18).

Other Renal Outcomes

eGFR Slopes

eGFR slopes were reported in 15 trials (16.9%), with eGFR declines significantly reduced in three trials (3.4%): the RENAAL trial of losartan (P = 0.01) [61], an unnamed trial of benazepril (P = 0.006) [113], and the REIN-1 trial of rami￾pril (P = 0.036) [118] (Table S19).

Albuminuria

UACR changes from baseline and final UACR measurements at end of follow-up were reported in 20 (22.5%) and 17 (19.1%) trials, respectively (Table S20a, b). Significant UACR decreases from baseline were observed in eight trials (9.0%): the GUARD, ASCEND, AWARD-7 and EMPA-REG￾RENAL trials of dipeptidyl peptidase 4 inhibitor gemigliptin (P \0.001) [39], avosentan 25 or 50 mg (P \0.001) [57], glucagon-like peptide-1 receptor agonist dulaglutide 1.5 mg (P = 0.0024) [37], and the SGLT2 inhibitor empagliflozin 25 mg (P = 0.0257–0.0031) [42], respectively, in patients with T2D; unnamed tri￾als of calcium channel blocker benidipine (P \0.0001 vs amlodipine) [84] and xanthine oxidase inhibitor topiroxostat (P = 0.0092) [79] in patients with or without T2D; the ACCOM￾PLISH trial of a combination of benazepril and amlodipine (P = 0.0001 vs benazepril combined with hydrochlorothiazide) in patients with or without T2D [85]; and the EVALUATE trial of selective aldosterone antagonist eplenerone in patients without T2D (P = 0.0222) [107].

When final UACR measurements at end of follow-up were used, significant decreases in UACR from baseline were observed in four trials (4.5%): an unnamed trial of lipid-lowering agent rosuvastatin in patients with T2D (P \0.01 vs standard of care) [70], the AMADEO Adv Ther (2022) 39:193–220 211 trial of RAAS inhibitors telmisartan and losartan in patients with T2D (both P \0.0001) [58], the RENAAL trial of losartan (P \ 0.001) [61], and an unnamed trial of benidipine (P \ 0.01 vs amlodipine) in patients with or without T2D [84].

Health-Related Quality of Life

Five trials (5.6%) [75, 97, 99, 100] reported HRQoL during treatment. In one trial (1.1%), Kidney Disease and Quality of Life physical function score improved significantly from baseline (P \ 0.0001) in patients with CKD and metabolic acidosis treated with veverimer, a first-in-class hydrochloric acid binder [104].

Early Trial Discontinuation

Ten trials (11.8%) were stopped early due to low recruitment or low event rates (n = 2) [47, 100], safety concerns (n = 5) [53, 57, 73, 74, 115], negative results reported in a sister trial (n = 1) [72], other reasons (n = 1) [61], or for reasons not provided (n = 1) [113]. On the advice of independent data monitoring committees, the CREDENCE [27] and DAPA-CKD [25] trials were stopped early after meeting prespecified efficacy criteria for early cessation and after demon￾strating overwhelming efficacy, respectively.

Risk of Bias Assessment

For seven of eight questions, 65–100% of trials had a ‘‘lower’’ or ‘‘unclear’’ risk of bias, while 35% of trials were not double blind and there￾fore at a ‘‘higher’’ risk of bias. Potential conflicts of interest were identified in 57% of trials (Fig. S6a, b).

Safety

Key safety outcomes are provided in Table S21.

The highest overall incidence of treatment-related adverse events (AEs) was reported in a trial of phosphodiesterase type 5 inhibition for patients with diabetic nephropathy (active arm, 54.7%; placebo arm, 56.3%) [49]. In this trial, the most common treatment-related AEs occurred in the placebo arm, and included headache (7.8%), diarrhea (3.6%), dyspepsia (3.6%), and peripheral edema (1.6%) [49]. The highest overall incidence of serious AEs was reported in the TREAT trial of darbepoetin alfa (active arm, 61.6%; placebo arm, 60.4%), which was stopped early due to safety concerns [75]. The most common serious AE, reported in the placebo arm, was hypertension (24.5%) [75].

DISCUSSION

The 89 clinical trials identified by this system￾atic literature review included a broad range of patients with any stage of CKD (eGFR 13.9–102.8 mL/min/1.73 m2 ) and albuminuria (UACR 29.9–2911.0 mg/g), with (75.5%) or without (20.6%) T2D. Many trials evaluated the impact of treat￾ment on one or more composite endpoints, and 16 trials reported significant reductions in risks of composites comprising kidney failure (n = 12) or cardiovascular mortality without kidney failure (n = 4) while evaluating RAAS inhibitors, SGLT2 inhibitors, finerenone, or other drug classes. However, these composites were diverse and assessed in a broad range of patients, hindering comparisons.

Clinically objective independent outcomes, such as kidney failure and ACM, were more consistently defined. Of 32 trials reporting incidences of kidney failure, seven observed significant risk reductions following treatment. These included a small trial of losartan (n = 751) in patients with T2D [61] and four smaller trials of losartan, benazepril, and ramipril (n = 84–436) in patients without T2D [112, 113, 115, 118], all published before 2008. Consequently, RAAS inhibition became the standard of care for patients with CKD [4].

However, there had been a lack of success in developing new agents to augment RAAS inhi￾bitors, delay progression, and improve out￾comes, with trials of other drug classes scarcely meeting their primary endpoints until recently.

Two large trials (n = 2152 and 2202) published onwards of 2019 demonstrated significant 212 Adv Ther (2022) 39:193–220 reductions in the risk of kidney failure among patients with UACR C 200 mg/g treated with SGLT2 inhibitors [25, 27]. While the CRE￾DENCE trial of canagliflozin only enrolled patients with T2D, the DAPA-CKD trial of dapagliflozin showed that kidney-protective effects from SGLT2 inhibition could be exten￾ded to patients with or without T2D [25]. A significant reduction in ACM observed in the same trial of dapagliflozin is the only example of a marked prolongation of survival reported to date in patients with CKD [25], and evidence from a recent systematic review confirms that well-designed clinical trials are required to optimize existing treatments to meet this unmet need [125].

Kidney failure and other clinical outcomes develop late in CKD, requiring trials with rela￾tively long durations to enroll large patient populations [10]. Surrogate endpoints can be used to monitor disease progression and evalu￾ate treatments in earlier stages of CKD [10, 126–129]. However, this review identified a diverse range of surrogate endpoints, including specific eGFR changes from baseline (33.7%), final eGFR values at end of follow-up (28.1%), eGFR slopes (16.9%), and percentage eGFR declines from baseline (12.4%). Future clinical trials evaluating new treatments for patients in the earlier stages of CKD may therefore benefit from the standardization of surrogate endpoints.

While it has been shown elsewhere that HRQoL diminishes with progression of CKD [5, 6], this review highlights the paucity of data showing that improvements with treatment are accompanied by improvements in HRQoL. Only five trials (5.6%) were identified that assessed HRQoL during treatment, with signifi- cant improvements limited to a trial of a hydrochloric acid binder for patients with metabolic acidosis [104]. Difficulties capturing changes in HRQoL, including the number of instruments used and differences in their sen￾sitivities, have been highlighted recently [6]. This review has several limitations, including the exclusion of non-English-language publica￾tions and of trials enrolling patients without albuminuria. Phase was not reported in 51.7% of trials, and it is possible that some phase 2 trials were included against eligibility criteria. A ‘‘higher’’ risk of bias was identified for 35% of trials that were not double blind. Finally, eligi-bility criteria were broad and this review inclu￾ded patients with any stage of CKD, with or without T2D, and treated with any drug class since 1990. CKD etiologies differed markedly between patients with T2D and without T2D, and a diverse range of comparators was also identified. Surrogate and clinically objective measurements of declining kidney function and treatment efficacy have also evolved over time, and 57 different composite outcomes were identified. Given the breadth and diversity of the data acquired, the performance of a meta￾analysis was considered to be infeasible.

CONCLUSION

Until recently, only RAAS inhibitors had shown that they could delay CKD progression and reduce the risk of kidney failure; however, this evidence was generated in just one small trial of patients with T2D and four smaller trials of patients without T2D. Contemporary data from the CREDENCE, DAPA-CKD, and FIDELIO-DKD trials suggest that adding an appropriate SGLT2 inhibitor or finerenone on top of standard of care RAAS inhibition can significantly improve a range of both kidney and cardiovascular out￾comes in patients with or without T2D. More￾over, data from DAPA-CKD suggest that dapagliflozin added to standard of care RAAS inhibition can significantly decrease all-cause mortality in patients with or without T2D.

Given the morbidity and mortality burden of CKD, the impact of CKD progression on HRQoL and healthcare costs, and the increasing preva￾lence of risk factors such as hypertension and diabetes in aging populations, these new drug classes potentially have an important role in the future treatment and management of CKD.

ACKNOWLEDGMENTS

Funding. Development of this manuscript and all associated publication costs, including Adv Ther (2022) 39:193–220 213 the journal’s Rapid Service and Open Access Fees, were supported by AstraZeneca. Medical Writing, Editorial, and Other Assistance. Medical writing and editorial sup￾port was provided by Matthew Young, DPhil, and Rachael Cazaly (Core, London, UK), sup￾ported by AstraZeneca according to Good Publication Practice guidelines (https://www. acpjournals.org/doi/10.7326/M15-0288). Authorship. All named authors meet the International Committee of Medical Journal Editors (ICMJE) criteria for authorship for this article, take responsibility for the integrity of the work as a whole, and have given their approval for this version to be published.

Author Contributions. All authors con￾tributed to the data interpretation, critically reviewed the manuscript, approved the final version, and accept accountability for the overall work. Study design was performed by Juan Jose Garcia Sanchez, Juliette Thompson, Glen James, Stephen Nolan, Naveen Rao, Bergur V. Stefansson, Alyshah Abdul Sultan, and Eric T.

Wittbrodt. Data analyses were performed by Juan Jose Garcia Sanchez and Juliette Thomp￾son. Ultimate responsibility for opinions, con￾clusions, and data interpretation lies with the authors.

Prior Presentation. Data presented in this article were also presented at the American Society of Nephrology Kidney Week meeting, October 22–25, 2020 (poster: PO0570).

Disclosures. Juan Jose Garcia Sanchez, Nav-een Rao, Elisabeth So¨rstadius, Glen James, Ste￾phen Nolan, Eric T. Wittbrodt, Alyshah Abdul Sultan, Bergur V. Stefansson, and Dan Jackson are employees and shareholders of AstraZeneca. Juliette Thompson, David A. Scott, Rachel Evans, and Keith R. Abrams are partners/em￾ployees of Visible Analytics Ltd, which con￾ducted this systematic review and received consultancy fees and expenses from AstraZeneca.

Compliance with Ethics Guidelines. This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Open Access. This article is licensed under a Creative Commons Attribution-Non-Commercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permis-sion directly from the copyright holder. To view a copy of this licence, visit http:// creativecommons.org/licenses/by-nc/4.0/.