#NephJC Chat
Tuesday, January 9th, 2024, at 9 pm Eastern
Wednesday, January 10th, 2023, at 9 pm Indian Standard Time and 3:30 GMT
Lancet. 2023 Dec 15 doi: 10.1016/S0140-6736(23)02408-X.
Efficacy and safety of aldosterone synthase inhibition with and without empagliflozin for chronic kidney disease: a randomized, controlled, phase 2 trial
Katherine R Tuttle , Sibylle J Hauske, Maria Eugenia Canziani , Maria Luiza Caramori , David Cherney , Lisa Cronin , Hiddo J L Heerspink, Christian Hugo, Masaomi Nangaku, Ricardo Correa Rotter , Arnold Silva , Shimoli V Shah , Zhichao Sun , Dorothea Urbach, Dick de Zeeuw , Peter Rossing ; ASi in CKD group
PMID: 38109916
Introduction
Inhibitors of the renin angiotensin aldosterone (RASi) system are time-tested, major players in delaying the progression of proteinuric chronic kidney disease (CKD). Although angiotensin receptor blockers (ARBs) and the angiotensin converting enzyme inhibitors (ACEi) are the venerable showstoppers, finerenone made a triumphant splash with the FIDELIO (and subsequently FIGARO) in the diabetic kidney disease space (NephJC summary, podcast). Notably, only a few hundred patients in the entire FIDELITY combined research program were on a flozin (since the flozin studies were still ongoing when they were planned). Also, finerenone did not include proteinuric but non-diabetic CKD (which DAPA CKD, and EMPA KIDNEY did). While endothelin antagonists and GLP1RAs are now muscling into this space as well, it is time to see if finerenone is enough, or we need more evidence to establish the second A firmly to change the RASi to RAASi. This time we have on stage BI 690517, a selective aldosterone synthase inhibitor. (as an aside: BI 690517 is quite a mouthful - it definitely needs a better name, perhaps B-SASI aka be sassy)?
Infographic courtesy Brian Rifkin, MD (X handle @brian_rifkin)
“Aldosterone breakthrough”, first observed (Staessen J et al, J Endocr 1981) in the 1980s, is the unusual phenomenon of increased aldosterone levels despite chronic inhibition of RAAS. The incidence varies between 10-53% depending on the definition (Bomback AS et al, Nat Clin Prac Nephrol 2007). “Aldosterone breakthrough” is different from “aldosterone escape” (Schrier RW,Nat Rev Neph 2010). Aldosterone escape is the increase in urinary sodium excretion, after an initial phase of sodium retention, despite high levels of aldosterone, which has been described in primary hyperaldosteronism. In the breakthrough phenomenon, the increase in aldosterone levels may lead to resistant hypertension and progression of cardiovascular and kidney diseases.
The deleterious effects of aldosterone are illustrated in figure below.
Mechanisms of cardiac and Kidney damage induced by Aldosterone excess from Chung EYM et al, Cochrane Library
The synthesis of aldosterone, and the breakthrough phenomenon, can be blocked by aldosterone synthase inhibitors (ASI). Since aldosterone synthase (AS) has a 95% amino acid homology with cortisol synthase (Kolkhof P et al. Pharmacol Res 2021), ASIs need to have a high selectivity towards AS to prevent hypocortisolism. The first orally active ASI, Osilodrostat, had a selectivity of 10:1 and was repurposed for treatment of Cushing’s syndrome (Bertagna X et al, JCEM 2014) due to the side effect of adrenal insufficiency. Baxdrostat, an ASI with >100 fold selectivity for AS was shown to be effective and safe in resistant hypertension in the phase 2 BrigHTN trial (Freeman, NEJM 2023) without any significant decrease in cortisol levels. Finally, Lorundrostat showed its efficacy in uncontrolled hypertension in the Target HTN trial (Laffin LJ et al, JAMA 2023) without any significant hyperkalemia. This newest molecule, BI 690517, has an even higher selectivity for AS (250:1) and a shorter half-life than baxdrostat or lorundrostat (Tuttle KR , Am J Neph 2023), possibly decreasing the risk of hyperkalemia (phase 1 data is as yet unpublished).
But why use ASIs when we already have potent MRAs? What are the potential advantages of ASIs over MRAs?
Blunting of the aldosterone breakthrough phenomenon
Mineralocorticoid receptors (MR) are activated by aldosterone but also by cortisol, oxidative stress, high salt diet and hyperglycemia (Figure 2 below). ASIs spare the cortisol mediated activation of the MR
Non-genomic effects of aldosterone, which are not inhibited by all MRAs, may be inhibited by ASIs
Off target estrogenic side-effects of MRAs do not occur with ASIs
Different mechanisms of MR antagonists and ASI. (Ando H, Hypertens Res,2023)
Let us find out if these theoretical aspects panned out practically in a clinical trial.
The Study
Methods
This multinational, phase II, double-blind study (NCT05182840) investigated the efficacy and safety of three doses of oral BI 690517, with or without empagliflozin 10 mg, in participants with CKD with or without T2DM.
Inclusion criteria
Consenting adults age ≥18 years
Diagnosed CKD with or without concurrent T2DM
eGFR ≥30 and <90 mL/min/1.73 m2 at screening and ≥20 mL/min before the treatment period
UACR ≥200 and <5,000 mg/g at screening
Serum potassium ≤4.8 mmol/L at screening and before the treatment period
Mean systolic BP ≥110 and ≤160 mmHg and mean diastolic BP ≥65 and ≤110 mmHg at screening
On stable dose of RASi (ACEi or ARB) for ≥4 weeks
On stable dose of NSAIDs/endothelin receptor agonist antagonist/low-dose systemic steroid treatment (up to 10 mg/day) for ≥ 4 weeks
Using contraceptive methods if child bearing potential
Key Exclusion criteria
Mineralocorticoid receptor antagonists (MRA) or potassium-sparing diuretics ≤7 days before randomization
Planned or current SGLT2 inhibitor or SGLT1/2 inhibitor use
Planned or current use of biotin at doses ≥5 mg/day within 72 h of screening (likely due to biotin interference with immunoassays measuring renin and aldosterone)
Immunotherapy or systemic steroid use (except prednisolone ≤10 mg) ≤3 months before screening
Acute kidney injury ≤30 days before screening or randomization
T1DM or diabetes of other autoimmune causes
Heart failure class III/IV NYHA, or coronary heart disease not compensated by medical treatment (Cases with indication of MRA)
Absolute cortisol <18 μg/dL 30 min after injection of ACTH at screening
Serum cortisol <5 μg/dL at screening
Any other major comorbidity or if pregnant or lactating
Study design and intervention
After the screening process was completed participants were randomized 1:1 to a run-in period to receive empagliflozin 10 mg once daily, or placebo, for the remainder of the study. Following 8 weeks of empagliflozin or placebo run-in, participants in both cohorts were randomized 1:1:1:1 to the treatment period during which they received either 3 different doses of BI 690517 (3 mg, 10 mg, or 20 mg) or placebo once daily for 14 weeks, with a follow-up visit at week 18.
Study design and intervention from Tuttle KR et al, Am J Neph 2023
Randomization
A two step randomization process (as described above) was followed using interactive response technology. The randomisation list was generated by the sponsor using a random number generating system. The schedules were independently verified by a trial-independent statistician. Randomization was stratified for eGFR (<45 and ≥45 mL/min) and UACR (≤750 and >750 mg/g). A capping of 70% patients with diabetes overall was used but diabetes status was not used for stratifying randomization.
Discontinuation
Participants permanently discontinued all allocated interventions for the following reasons:
Acute kidney injury
Decrease in baseline eGFR of ≥30% prior to visit 6 or of ≥40% at any time
Kidney failure ( eGFR of <15 mL/min/1.73 m2 or requirement for kidney replacement therapy).
Serum potassium levels ≥6.0 mmol/L (or ≥5.6 mmol/L if down-titration was not appropriate)- Down titration protocols were not mentioned in detail
Cortisol levels <82.8 nmol/L, or if they developed adrenal insufficiency (including cortisol levels <496.6 nmol/L within 30 min of adrenocorticotropic hormone stimulation) or Cushing’s syndrome during the study
Outcomes for efficacy
Primary outcome: Change from baseline in UACR measured in first morning void (FMV) urine after 14 weeks of treatment
Secondary outcome: Proportions of participants with ≥15% and ≥30% in UACR FMV at week 14
UACR FMV baseline was defined as the mean of all non-missing measurements collected from week -2 until the pre-dose measurement. UACR values for the measurement of the primary and secondary outcomes were based on six measurements from three collections on two consecutive days. Subgroup and sensitivity analyses of the primary and secondary endpoints were performed using age, sex, race, ethnicity, geographical region, BMI, presence of diabetes, eGFR, UACR, glycated hemoglobin, blood pressure, and presence of cardiovascular disease.
Safety outcomes
Adverse events of special interest in this study were drug-induced liver injury, Cushing’s syndrome, adrenal insufficiency, ketoacidosis, and events leading to lower extremity amputation. Blood samples for plasma aldosterone and cortisol levels were taken throughout the treatment and follow-up periods. In addition, aldosterone levels were measured at week −1 (run-in period), and cortisol levels were measured at screening. An adrenocorticotropic hormone stimulation test was performed for all participants at screening, in those with serum cortisol level ≥82.8 and <303.5 nmol/L with symptoms suggestive of hypocortisolism, and at the end of the treatment visit (or following premature study discontinuation) for participants with morning serum cortisol level <496.6 nmol/L.
Funding
Boehringer Ingelheim, maker of both empagliflozin as well as BI 690517, funded the study, and provided the randomization schedule. Though it is not clear, it is likely they also did or funded the statistical analysis. Four of the authors are BI employees. A medical writer funded by BI (Terri Penfold of Callisto, OPEN Health Communications, London, UK) was involved in drafting the manuscript.
Results
Of the 714 run-in participants, 586 were randomly assigned to receive BI 690517 or placebo. During the run-in period, 51 (7.2%) discontinued background therapy with empagliflozin or placebo (AEs, n = 10 [1.4%]; burden of study procedures, n = 4 [0.6%]; unknown reason, n = 4 [0.6%]; change of residence, n = 1 [0.1%]; other [not sponsor termination], n = 32 [4.5%]). Of the 662 participants who completed the run-in, 78 (11.8%) did not meet the additional selection criteria for randomization to the BI 690517 treatment period, thus leaving 586 who were randomized.
Figure 1 from Tuttle et al Lancet 2023
Additionally, during the treatment period, 105 (23.9%) participants discontinued BI 690517 treatment, 53 (17.8%) participants discontinued empagliflozin treatment, 23 (15.6%) participants discontinued BI 690517-matched placebo, and 37 (12.8%) participants discontinued empagliflozin-matched placebo. AEs were the commonest reason for discontinuation.
Table 1 shows the baseline characteristics. Of these, 33% were women and 42% had a racial identity other than White. The mean participant age was 64 years. A total of 13% (36/288) of participants who were on BI 690517 alone and 17% (51/298) with BI 690517 in combination with empagliflozin had a systolic blood pressure of less than 120 mmHg at baseline. ARB use was more common (n=405, 69%) than ACE inhibitor (n=178, 30%). About 71% (n=414) of participants had type 2 diabetes, though the number varied from 62 to 88% in the individual groups. The mean GFR was ~ 55, and albuminuria was ~ 400 mg/g.
Table 1 from Tuttle et al Lancet 2023
Change in Albuminuria
The percentage change in UACR (FMV) from baseline to week 14 was -3%, -22% , -39% and -37% respectively for placebo, 3mg, 10 mg and 20 mg arms without empagliflozin. In the arms with background flozin therapy the albuminuria reduction was greater (as expected) of the order of -11%, -19%, -46% and -40% respectively in placebo, 3mg, 10 mg and 20 mg arms. The CI pointed towards a significant decrease in UACR at 14 weeks in 3mg, 10 mg and 20 mg arms both with and without flozins with a probable ceiling dose of 10 mg. The reduction in UACR was documented to consistently occur from week 6 to week 14.
Figure 2. Change in UACR from baseline to week 14 from Tuttle et al Lancet 2023
Results of the subgroup analyses showed consistent decrease in both diabetic and non diabetic CKD, GFR <45 and baseline UACR between 300-3000 mg/g in the flozin group. The sensitivity analyses for missing data was congruent with the main results, with BI 690517 10 mg and 20 mg being consistently superior to placebo
Table S6: Participants achieving decreases of at least 30% absolute change in FMV UACR from baseline after 14 weeks of trial treatment (background flozin therapy) − Full analysis set
A higher than 30% reduction in UACR was observed in 51% of patients on 10 mg BI 690517 alone and 70% patients on 10 mg BI 690517 dose with flozin. The largest response rate for a 15% or higher reduction in UACR from baseline was observed in the BI 690517 10 mg dose group also receiving empagliflozin, with an OR versus placebo of 6·08 (2·73–13·57; appendix p 55).
Consistent with its mechanism of action, a BI 690517 reduced plasma aldosterone exposure (area under the curve) dose-dependently up to –62% (95% CI –76 to –41) without empagliflozin and –66% (95% CI –75 to –53) with empagliflozin compared with placebo at the highest dose of BI 690517 (20 mg) by week 14. Despite a greater decrease in aldosterone levels with 20 mg dose as compared to 10 mg, this did not translate to a significant decrease in albuminuria as compared to 10 mg arm.
Change in GFR and BP
Small decreases in eGFR from baseline to week 14 of study treatment were observed in response to BI 690517 with or without concurrent empagliflozin. The proportion of participants with a decrease in eGFR of 30% or more from baseline to week 14 was 3% each with placebo, 3 mg and 10 mg and 6% with BI 690517 20 mg. Among those that received BI 690517 combined with empagliflozin, the proportion of participants with a decrease in eGFR of 30% or higher from baseline to week 14 was 3%, 1%, 3% and 4% with placebo, 3 mg, 10 mg and 20 mg respectively. BI 690517 with empagliflozin resulted in larger decreases in placebo-corrected mean systolic and diastolic blood pressures compared with BI 690517 monotherapy across most dose groups. Investigator reported hypotension and orthostatic hypotension were infrequent adverse events
Figure S10 from Tuttle et al, Lancet 2023- Effect of BI 690517 on eGFR
Figure S11 from Tuttle et al, Lancet 2023- Effect of BI 690517 on systolic BP
Safety endpoints
In both the BI 690517 monotherapy and BI 690517 combined with flozin groups, there were no significant decreases in morning cortisol levels compared to placebo at any of the doses. BI 690517 was well tolerated with essentially no Cushing’s syndrome and very few (but not zero) patients with adrenal insufficiency (1-2%). Hyperkalemia events did increase in line with increasing doses of BI 690517. Hypotension and AKI were rare (1-3%).
Table 3 from Tuttle et al Lancet 2023
Discussion
This was the first phase 2 study of an aldosterone synthase inhibitor (ASI) in chronic kidney disease. The RCT was designed to study different doses of BI 690517 in proteinuric chronic kidney disease with or without empagliflozin background therapy. The population included CKD across a wide spectrum of eGFR (30-90 ml/min/1.73m² ), albuminuria (200-5000 mg/g) and CKD etiology, with about 30% not having diabetes. Overall it demonstrated a significant decrease in proteinuria in the 10 mg and 20 mg subgroups of BI 690517 at 14 weeks, both with and without empaglifozin background therapy. Across the doses of BI 690517, a large albuminuria reduction was seen with 10 mg arm if we consider >15% decrease in UACR (OR=6.08), and the effect on albuminuria seems to level off with further dose escalation to 20 mg. The change in albuminuria was seen in both populations with and without diabetes, and those with and without baseline empagliflozin therapy. Other subgroups were difficult to interpret with small sample sizes in each of the 8 arms.
The trial should be lauded for including people with diverse ethnic backgrounds, but failed to include a fair proportion of women. The trial started recruitment in 2022 when flozins had already proved their mettle in CKD (Heerspink HJL et al, NEJM 2020)– surprisingly, they still had a trial arm without flozins in patients with proteinuric CKD.
The sample size was calculated to accommodate for a drop out rate of 15% to be able to demonstrate a > 30% decrease in albuminuria. Despite a high rate of discontinuation (22.8%- unclear reasons for such high discontinuation rate), the 10 mg and 20 mg subgroups were still able to show a much larger decrease in proteinuria at 14 weeks than expected. Some holdouts, such as KDIGO, still consider that the gold standard for albuminuria estimation is a 24-hour urine sample (see this editorial, Rodby AJKD 2016, for why the 24 hour urine holdouts are misguided). The trial instead used an average of 3 readings of the urine albumin first morning void (UACR FMV) which mitigates possible error.
Safety concerns were of prime importance in this phase 2 trial. The trial had a robust mechanism to explicitly look for and detect adrenal insufficiency. As a result, the trial did find and report an adrenal insufficiency rate of 1-3%. The other safety concerns in this trial centered around the effects of a combination of RAAS blockers, which have been implicated in the past for causing hyperkalemia and acute kidney injury (ONTARGET trial, Yusuf et al, NEJM 2008; VA Nephron D trial, Fried et al, NEJM 2013 ; ALTITUDE trial, Parving et al NEJM 2012). The mean rise in potassium in the groups was miniscule ( 0. 30 meq/l and 0.20 meq/l with and without flozins at max dose of 20 mg) but investigators reported hyperkalemia rates between 10-24% in the high dose arms. Unfortunately, the percentage of patients requiring dose down titration or drug discontinuation due to hyperkalemia was not mentioned, nor the protocol for how to decrease the dose and at what threshold to discontinue. The hyperkalemia rates seem comparable to finerenone, a nonsteroidal MRA, which was seen to cause hyperkalemia in 18.3% of patients in FIDELIO DKD study with a 2.3% discontinuation rate and a mean increase of 0.23 meq/l (Bakris et al, NEJM 2020; Waitzman et al AJKD 2021).
Comparing BI 690517 to other ASIs in phase 2 development, the BrigHTN trial (Freeman et al, NEJM 2023) showed a larger decrease in BP with baxdrostat, with a lower rate of hyperkalemia (2-3% at highest doses) in a population with higher baseline eGFRs (<6% had eGFR <60 ml/min). Lorundrostat was reported to improve systolic BP in patients with uncontrolled hypertension with or without suppressed plasma renin activity, with low rates of hyperkalemia (Lauffin et al, JAMA 2023). Data on the long term effect of ASIs on eGFR and kidney failure is yet to be seen. As the nephrology community is embracing flozins with open arms, it is also waiting with bated breath to see if the newer refined ASIs are going to share center stage in this act, or just be another supporting actor in the show.
Conclusion
In this phase 2 trial, the newest ASI (BI 690517), with a high affinity for aldosterone synthase, shows great promise for decreasing proteinuria with limited adverse effects of adrenal insufficiency, hypotension, hyperkalemia and AKI. The ongoing EASiKIDNEY trial will include 11,000 patients and will conclusively demonstrate whether ASIs should be added to the RASi + Flozin standard of care in proteinuric CKD. Other trials in a similar space include FIND-CKD (N = 1584, finerenone versus placebo in non-diabetic proteinuric CKD) and the phase 3 trials with baxdrostat (N = 720) and lorundrostat (N = 1000), both in hypertension. Will this be the golden age for aldosterone, finally getting its place in the RAAS pantheon?