Dapagliflozin Utilization in Chronic Kidney Disease and Its Real-World Effectiveness Among Patients with Lower Levels of Albuminuria in the USA and Japan
verfasst von:
Navdeep Tangri, Anjay Rastogi, Cassandra Nekeman-Nan, Lai San Hong, Asuka Ozaki, Stefan Franzén, Tadashi Sofue
Sodium–glucose cotransporter 2 inhibitors such as dapagliflozin have been proven effective for slowing chronic kidney disease (CKD) progression in large outcomes trials that mainly included patients with higher levels of albuminuria. Understanding the real-world utilization and effectiveness of these drugs among patients with CKD with lower levels of albuminuria can inform clinical decision-making in this population.
Methods
Claims data from the USA and Japan were used to describe patients with CKD and urinary albumin-to-creatinine ratio (UACR) < 200 mg/g who were eligible for dapagliflozin 10 mg treatment (initiators and untreated) following its approval for CKD. A quantile regression analysis was performed to evaluate the effect of dapagliflozin 10 mg initiation versus no initiation on estimated glomerular filtration rate (eGFR) slope in a propensity score-matched cohort, using a prevalent new-user design.
Results
Dapagliflozin initiators (n = 20,407) mostly had stage 3–4 CKD (69–81% across databases). The most common comorbidities were type 2 diabetes, hypertension and cardiovascular disease. At baseline, a renin–angiotensin system inhibitor was prescribed in 53–81% of patients. Eligible but untreated patients were older and had a higher eGFR and lower comorbidity burden than initiators. Following dapagliflozin initiation, the differences in median eGFR slope between initiators and matched non-initiators were 1.07 mL/min/1.73 m2/year (95% confidence interval [CI] 0.40–1.74) in all patients with UACR < 200 mg/g and 1.28 mL/min/1.73 m2/year (95% CI − 1.56 to 4.12) in patients with UACR < 200 mg/g without type 2 diabetes.
Conclusions
Dapagliflozin 10 mg was prescribed to a broad range of patients with CKD. In patients with UACR < 200 mg/g, dapagliflozin initiation was associated with a clinically meaningful attenuation of eGFR slope compared with non-initiation. These findings supplement available clinical efficacy evidence and suggest that dapagliflozin effectiveness may extend to patients with CKD and UACR < 200 mg/g. Graphical Abstract and Video Abstract available for this article.
Prior Presentation: A descriptive analysis of dapagliflozin use for CKD in Japan from the OPTIMISE-CKD study was presented as a focussed oral at the 60th European Renal Association Congress, Milan, Italy, 15–18 June 2023. Abstract number #4420. A descriptive analysis of dapagliflozin use for CKD in the USA from the OPTIMISE-CKD study was presented as a poster (TH-PO855) at the American Society of Nephrology Kidney Week 2022, Orlando, Florida, USA, 3–6 November 2022.
Key Summary Points
Why carry out this study?
Chronic kidney disease (CKD) is a progressive condition that affects approximately 10% of people worldwide, with later stages associated with significant effects on health-related quality of life and substantial economic costs.
Understanding real-world dapagliflozin utilization and effectiveness among patients with CKD is important to inform healthcare decision-making, particularly with regards to patients with urinary albumin-to-creatinine ratio (UACR) < 200 mg/g and patients without type 2 diabetes, in whom studies are limited.
This study aimed to describe real-world utilization of dapagliflozin 10 mg following its approval for the CKD indication in the USA and Japan, and to assess the effect of initiating versus not initiating dapagliflozin 10 mg on kidney function decline in patients with UACR < 200 mg/g.
What was learned from the study?
Dapagliflozin 10 mg was prescribed to a broad range of patients following its approval to treat CKD in the USA and Japan.
Among patients with UACR < 200 mg/g, initiating dapagliflozin 10 mg was associated with a clinically meaningful attenuation of estimated glomerular filtration rate (eGFR) slope compared with not initiating: average eGFR difference of 1.07 mL/min/1.73 m2/year (95% confidence interval [CI] 0.40–1.74) and 1.28 mL/min/1.73 m2/year (95% CI −1.56 to 4.12) among patients with UACR < 200 mg/g and without type 2 diabetes.
These findings supplement available clinical efficacy evidence and suggest that dapagliflozin effectiveness in patients with CKD may extend to those with UACR < 200 mg/g.
Digital Features
This article is published with digital features, including a Graphical Abstract and a Video Abstract, to facilitate understanding of the article. To view digital features for this article, go to https://doi.org/10.6084/m9.figshare.24808380.
Anzeige
Introduction
Chronic kidney disease (CKD) is a progressive condition that is estimated to affect approximately 10% of the global population [1, 2]. As kidney function continues to deteriorate, the later stages of CKD are typically associated with poor health-related quality of life and adverse outcomes, such as cardiovascular disease, end-stage kidney failure and death [3‐5].
In the USA, CKD is a significant public health concern, with an estimated prevalence of approximately 14% among adults [6]. The disease is also associated with a substantial cost burden [7], with Medicare spending for beneficiaries with CKD and end-stage renal disease totalling more than US$129 billion in 2021 [8, 9]. A similar trend can be seen in Japan, where the estimated prevalence of CKD is 13% [10] and the annual expenditure on kidney disease-related healthcare was approximately ¥1.66 trillion (US$11.2 billion) in 2019 (3.7% of all medical costs) [11]. Furthermore, the cost for haemodialysis therapy accounted for 5% of the total medical costs in Japan in 2017 [12]. In 2019, in response to the increasing prevalence of CKD and associated healthcare costs, government-backed initiatives to identify strategies to prevent the development and progression of CKD were announced in Japan and the USA [13, 14].
Until recently, renin–angiotensin system inhibitors (RASis) were the only medications recommended for reducing kidney function decline in patients with CKD. However, a growing body of evidence supports the use of sodium–glucose cotransporter 2 inhibitors (SGLT2i) in the management of CKD [15‐18]. In 2020, results from the Dapagliflozin and Prevention of Adverse Outcomes in Chronic Kidney Disease (DAPA-CKD) trial showed that dapagliflozin 10 mg once daily significantly reduced the risk of a composite endpoint comprising sustained decline in estimated glomerular filtration rate (eGFR), end-stage renal disease and death from renal or cardiovascular causes among patients with CKD, regardless of type 2 diabetes status [17]. Following these results, dapagliflozin 10 mg was approved for the treatment of CKD in patients with or without type 2 diabetes in 2021, making it the first SGLT2i approved for use in a broad population of patients with CKD [19].
Despite its broad indication, use of dapagliflozin 10 mg for CKD among patients with urinary albumin-to-creatinine ratio (UACR) < 200 mg/g may be hindered by the limited number of studies performed in this group of patients, particularly among those without type 2 diabetes.
Anzeige
OPTIMISE-CKD is an observational study programme using claims data sources describing dapagliflozin treatment for CKD in routine clinical practice. The present analysis comprises two components: a descriptive analysis of the real-world utilization of dapagliflozin 10 mg following its approval for CKD treatment and an analysis of the real-world effectiveness of initiating dapagliflozin 10 mg among patients with CKD with UACR < 200 mg/g.
Methods
Data Sources and Study Period
Optum’s de-identified Clinformatics® Data Mart Database is an administrative claims database for a privately insured population in the USA with commercial or Medicare Advantage coverage. The dataset used in the current study included data from over 76 million people collected between 1 January 2007 and 31 March 2023. The Medical Data Vision Co. Ltd database (MDV; Tokyo, Japan) is a de-identified hospital claims database, which covers over 39 million Japanese inhabitants [20]. The dataset in the current study included 3.5 million patients with at least one eGFR measurement of < 90 mL/min/1.73 m2 or a CKD diagnosis between 1 April 2008 and 31 December 2022. The Real World Data Co. Ltd database (RWD; Kyoto, Japan) contains inpatient and outpatient data from approximately 26 million patients from approximately 229 Japanese medical institutions [20]. This database is maintained by the Health, Clinic, and Education Information Evaluation Institute (HCEI; Kyoto, Japan) with support from Real World Data Co. Ltd (Kyoto, Japan). The dataset used in the current study included over 4 million adult patients with eGFR < 90 mL/min/1.73 m2 between 10 June 1985 and 31 December 2022.
The study period for the descriptive dapagliflozin 10 mg utilization analysis was between the CKD indication approval (30 April 2021 [USA]; 25 August 2021 [Japan]) and 30 September 2022 (USA) or 31 December 2022 (Japan). The study period for the effectiveness analysis was from 30 August 2020 (first release of DAPA-CKD trial results) until the end of data available in each dataset.
Descriptive Analysis of Dapagliflozin 10 mg Utilization After CKD Indication Approval
All adult patients in the Clinformatics Data Mart, RWD and MDV databases were included in the descriptive analysis if they initiated or were eligible for dapagliflozin 10 mg during the study period. A list of eligibility criteria for this descriptive analysis can be found in Table S1.
Summary statistics were used to describe patient demographics, clinical characteristics, comorbidities, medications and laboratory tests during the baseline period or on the index date. The study index date was defined as the date of first dapagliflozin 10 mg prescription (in the case of dapagliflozin initiators) or the first date on which patients met all eligibility criteria during the study period (in the case of eligible but untreated patients). Missing data were not imputed but are reported as a separate category.
Effectiveness of Dapagliflozin 10 mg Among Patients with UACR < 200 mg/g
Study Population and Index Date
The base cohort for this analysis included all patients who met the eligibility criteria (Table 1) on any day (possible index date) during the study period in the two databases that contained UACR records (Clinformatics Data Mart and RWD). To increase the number of patients in this effectiveness analysis, patients were also included if they had a urinary protein-to-creatinine ratio (UPCR) measurement that corresponded to the UACR thresholds used as inclusion and exclusion criteria for the study. UPCR values were converted using the crude model PCR equation for predicted UACR, as described by Sumida et al. [21]. To allow a pooled analysis of both countries and thereby increase the study sample size, this analysis applied the more conservative dapagliflozin 10 mg indication for CKD treatment from the USA. Therefore, all included patients were eligible for dapagliflozin 10 mg according to the approved CKD indication. All eligible patients were considered potential comparators until they initiated dapagliflozin 10 mg. Patients who subsequently initiated dapagliflozin 10 mg and met the indication for the treatment of CKD were then categorized as ‘dapagliflozin initiators’. Dapagliflozin initiators could therefore be a matched comparator in the period before initiation.
Table 1
Effectiveness analysis: eligibility criteria to identify patients with a possible index date
Inclusion criteria
Age
≥ 18 years on index date
Meets CKD definition on or within 2 years before index date
UACR ≥ 30 mg/ga
UPCR ≥ 150 mg/g
CKD diagnosis code
Two eGFR measurements ≥ 90 days apart, both < 60 mL/min/1.73 m2
Exclusion criteria
Continuous enrolment before index date
< 730 days
eGFR below threshold (on or within 1 year before index date)
25 mL/min/1.73 m2
UACRa above threshold (on or within 1 year before index date) or UPCR equivalent
aAlso included quantitative UPCR converted to UACR
The index date of dapagliflozin initiators was defined as the date of the first dapagliflozin 10 mg prescription. To avoid immortal time bias and selection bias when setting an index date for the comparators [22] and to ensure a manageable size for each country-specific analysis cohort, up to five potential comparators were randomly sampled for each dapagliflozin initiator in chronological order of their index dates. These comparators were matched on the basis of age, sex, heart failure diagnosis, type 2 diabetes diagnosis and RASi prescription (defined as having a prescription for angiotensin-converting enzyme inhibitors or angiotensin II receptor antagonists). Matching was performed without replacement: once comparators had been matched a dapagliflozin initiator, these comparators were no longer able to be chosen for matching with other dapagliflozin initiators. Index dates for the randomly sampled comparators were set to the index dates of their matched dapagliflozin initiators; the datasets were then merged. To maximize retention of dapagliflozin initiators and minimize potential unmeasured bias, only patients with sufficient post-index data for eGFR slope estimation were retained. Each patient could only contribute once as an initiator and/or once as a comparator.
Follow-up Period
Patients were followed from index date until the earliest of the following: loss to follow-up, death or end of the study period. Specifically in the case of comparators who became dapagliflozin initiators, the follow-up period as a comparator ended on the day that these patients initiated dapagliflozin 10 mg.
Outcomes
The primary outcome was eGFR slope between index and the end of follow-up. No difference was made between total, acute or chronic slope owing to the variability in the timing of eGFR measurements. The Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation (2021) was used to calculate eGFR [23] for all patients to ensure that a consistent and up-to-date method was used across patients. In clinical practice, an eGFR test may be redone shortly after it is first conducted in case of an unusual eGFR test result. This could lead to extreme and inaccurate slopes if based on measurements within a short time interval. Therefore, individual eGFR slopes based on only two measurements that were less than 30 days apart were excluded from this analysis.
Statistical Analyses
The statistical methodology described in this section applies only to the effectiveness analysis, using the prevalent new user design as described by Suissa et al. [22]. Each dapagliflozin initiator was matched 1:1 with a potential comparator who had not initiated treatment on the same date and had the closest matching propensity score. Nearest neighbour matching was performed using logistic regression within the MatchIt package (version 4.5.4) in R (version 4.0.2) [24]. A propensity score model, which included all variables in the full baseline table (Table S4) and an interaction between angiotensin receptor/neprilysin inhibitor (ARNI) and heart failure for matching patients, was developed to maximize the balance between groups. Covariates were considered well balanced if standardized mean differences were less than 0.1 after propensity score matching.
Individual post-index eGFR slopes were extracted from individual unadjusted linear regression models; one for each time a patient appeared in the analysis cohort. The difference in median eGFR slopes between dapagliflozin initiators and matched comparators was assessed with a quantile regression model with the Quantreg package version 5.95 in R [25], adjusted for age, sex, baseline eGFR and UACR, RASi and diagnoses for cardiovascular disease, anaemia, heart failure and type 2 diabetes.
Subgroup and Sensitivity Analysis
A subgroup analysis was performed among patients without type 2 diabetes, by including only patients who did not have a recorded diagnosis of type 2 diabetes, diabetic kidney disease or a prescription for glucose-lowering drugs in their history. Up to five potential comparators for each dapagliflozin initiator were randomly sampled using the approach mentioned above with an additional matching on any diagnosis for cardiovascular disease. Patients with a sufficient number of post-index eGFR measurements were then propensity score matched at a 1:1 ratio. A quantile regression was performed to assess differences in eGFR slopes between groups, using the same covariates except type 2 diabetes.
To compensate for the expected small sample size in the subgroup analysis of patients without type 2 diabetes, a post hoc analysis was performed using results from the total cohort to inform the estimated treatment effect in the cohort of patients without type 2 diabetes [26]. Further details of the methodology used in this analysis can be found in the Supplementary Material.
A sensitivity analysis was also performed on propensity score-matched patients with UACR < 200 mg/g, regardless of post-index date eGFR data availability.
Compliance with Ethics Guidelines
OPTIMISE-CKD used de-identified data from existing databases and did not require data collection beyond that of routine clinical care. No identifiable information was collected or examined as part of the study. All externally conducted analyses were completed in line with local ethics regulations/legislation. De-identified, internally licensed databases were shared with AstraZeneca by the licensee; therefore, ethics review and approval were not required for the use of these databases for this study.
Results
Descriptive Analysis of Dapagliflozin 10 mg Utilization After CKD Indication Approval
Across all databases, 20,407 dapagliflozin initiators were included in the analysis (Fig. S1), with a median age of 73 (USA), 77 (JapanRWD) and 71 years (JapanMDV). Dapagliflozin 10 mg was initiated in patients across all CKD stages, but mostly in those with stage 3–4 CKD. In general, the most commonly recorded comorbidities anytime in the patient history were type 2 diabetes (USA, 85%; JapanRWD, 39%; JapanMDV, 36%), hypertension (USA, 98%; JapanRWD, 92%; JapanMDV, 86%), heart failure (USA, 47%; JapanRWD, 71%; JapanMDV, 45%) and other cardiovascular diseases (USA, 75%; JapanRWD, 60%; JapanMDV, 33%). The majority of patients were prescribed a cardiovascular medication (USA, 96%; JapanRWD, 82%; JapanMDV, 63%) or a RASi (USA, 81%; JapanRWD, 63%; JapanMDV, 53%) at or within the year before the index date. A numerical UACR value at baseline was available for 46% and 42% of patients in the Clinformatics Data Mart and RWD databases, respectively; the majority of these UACRs were below 300 mg/g. Moreover, among dapagliflozin initiators with an available UACR measurement, the proportion with UACR < 200 mg/g was 75% and 89% in the Clinformatics Data Mart and RWD databases, respectively.
Of all dapagliflozin-eligible but untreated patients, average age was slightly higher than the dapagliflozin initiators, and eGFR was higher. The most common comorbidities and medications were similar to those of the dapagliflozin initiators, although the proportion of patients with comorbidities or medications was generally lower in dapagliflozin-eligible but untreated patients than in dapagliflozin initiators. In particular, fewer dapagliflozin-eligible but untreated patients had a RASi prescription (USA, 52%; JapanRWD, 27%; JapanMDV, 19%) than initiators at or within the year before the index date. A numerical UACR value at baseline was available for 26% and 25% of patients in the Clinformatics Data Mart and RWD databases, respectively. Descriptive patient characteristics of dapagliflozin initiators and dapagliflozin eligible but untreated patients in each database can be found in Table S2.
Anzeige
Effectiveness of Dapagliflozin 10 mg Among Patients with UACR < 200 mg/g
Baseline Characteristics
Figure 1 describes the patient flow diagram for the dapagliflozin initiators and comparator groups. In total, 3029 dapagliflozin 10 mg initiators with UACR < 200 mg/g were included during the study period. Additionally, 13,813 comparators who did not initiate dapagliflozin 10 mg were randomly sampled from 444,000 potential comparators. Before propensity score matching, dapagliflozin initiators had a lower median baseline eGFR than the comparator group (USA, 53.5 vs 61.8 mL/min/1.73 m2; Japan, 55.6 vs 68.1 mL/min/1.73 m2). The proportion of patients with comorbidities was largely similar in both groups in the USA and Japan, except for some numerical differences, whereby higher proportions were seen among the dapagliflozin initiators. Similarly, the proportion of dapagliflozin initiators with comedications was equal to or larger than that of the comparators (Tables S3a, b). After propensity score matching 2972 dapagliflozin 10 mg initiators at a 1:1 ratio with a comparator, baseline characteristics were well balanced between the groups, with a standardized mean difference of less than 0.1 for each variable (Tables 2 and S4). The US dataset contributed most patients (92%). In both groups, 44% of patients were female. In dapagliflozin 10 mg initiators and matched comparators, respectively, the median age was 74 versus 73 years. The most common comorbidities at any time in the patient history were hypertension (97% in both groups), type 2 diabetes (89% vs 90%), anaemia (54% vs 53%) and heart failure (43% vs 41%). The most common comedications prescribed at or within 2 years before index were RASi (85% vs 84%), statins (85% vs 83%) and antidiabetic treatments (79% vs 77%). All included patients had a baseline eGFR value (median 53.7 vs 54.8 mL/min/1.73 m2) and UACR value (median 20.1 vs 19.0 mg/g). The proportion of patients with UACR < 30 mg/g was 59% in dapagliflozin initiators and 61% in comparators. The maximum intervals between two eGFR measurements among dapagliflozin initiators and comparators, respectively, were 913 days and 861 days.
Fig. 1
Patient flow diagram: effectiveness of dapagliflozin 10 mg initiation versus no initiation in the USA and Japan. aIncludes both dapagliflozin initiators and non-initiators; dapagliflozin 10 mg initiators could be selected as potential comparators in the period before dapagliflozin initiation if they met all the other study criteria. bCKD defined as any of the following: CKD diagnosis code, UACR > 30 mg/g or two eGFR values ≥ 90 days apart (both < 60 mL/min/1.73 m2). cStratified sampling of up to five comparators for each dapagliflozin initiator; covariates: age, gender, heart failure, type 2 diabetes, RASi prescription. dPool of potential comparators (with sufficient post-index eGFR data) in the USA: overall n = 416,616; no type 2 diabetes n = 138,340. ePool of potential comparators (with sufficient post-index eGFR data) in Japan: overall n = 27,512; no type 2 diabetes n = 10,131. CKD chronic kidney disease, eGFR estimated glomerular filtration rate, RASi renin–angiotensin system inhibitor, UACR urinary albumin-to-creatinine ratio
Table 2
Effectiveness analysis: key baseline characteristics of propensity score-matched patients with CKD and UACR < 200 mg/g who initiated dapagliflozin 10 mg and who did not
UACR < 200 mg/g
UACR < 200 mg/g AND no type 2 diabetes
Initiated dapagliflozin 10 mg
Did not initiate dapagliflozin 10 mg
SMDa
Initiated dapagliflozin 10 mg
Did not initiate dapagliflozin 10 mg
SMDa
Number of patients
2972
2972
275
275
Female, n (%)
1296 (44)
1305 (44)
0.0061
107 (39)
107 (39)
< 0.0001
Age, years, median (IQR)
74 (69–79)
73 (69–78)
0.0222
75 (68–81)
76 (69–81)
0.0479
BMI available, n (%)
1071 (36)
939 (32)
85 (31)
100 (36)
BMI, kg/m2, mean (SD)
31.5 (7.0)
31.0 (7.1)
0.0669
27.4 (6)
27.5 (6.4)
0.0313
BMI category, kg/m2, n (%)
0–18.4
< 5b
5 (1)
< 5b
< 5b
18.5–24.9
178 (17)
191 (20)
26 (31)
33 (33)
25.0–29.9
318 (30)
257 (27)
36 (42)
40 (40)
≥ 30
571 (53)
486 (52)
20 (24)
25 (25)
Comorbidities, n (%)
Atrial fibrillation
802 (27)
797 (27)
0.0038
99 (36)
107 (39)
0.0600
Heart failure
1278 (43)
1232 (41)
0.0313
148 (54)
155 (56)
0.0511
Hypertension
2890 (97)
2885 (97)
0.0101
259 (94)
262 (95)
0.0487
Myocardial infarction
391 (13)
368 (12)
0.0232
29 (11)
32 (12)
0.0347
Stroke
942 (32)
892 (30)
0.0364
76 (28)
88 (32)
0.0953
Anaemia
1602 (54)
1582 (53)
0.0135
135 (49)
133 (48)
0.0145
Type 2 diabetes
2658 (89)
2667 (90)
0.0099
N/A
N/A
Medications, n (%)
RASi
2521 (85)
2503 (84)
0.0258
217 (79)
225 (82)
0.0731
ARNI
304 (10)
260 (9)
0.0110
32 (12)
25 (9)
0.0834
Calcium channel blockers
1269 (43)
1249 (42)
0.0027
124 (45)
123 (45)
0.0073
Diuretics
1503 (51)
1377 (46)
0.0143
159 (58)
155 (56)
0.0293
Statins
2536 (85)
2452 (83)
0.0138
173 (63)
174 (63)
0.0075
Antidiabetic treatments
2333 (79)
2283 (77)
0.0395
N/A
N/A
eGFR available, n (%)
2972 (100)
2972 (100)
275 (100)
275 (100)
eGFR, mL/min/1.73 m2, median (IQR)
53.7 (42.8–71.1)
54.8 (43.2–69.1)
0.0133
50.1 (40.3–60.2)
49.5 (38.4–61.4)
0.0121
UACR available, n (%)
2972 (100)
2972 (100)
275 (100)
275 (100)
UACR, mg/g, median (IQR)
20.1 (7.0–55.8)
19.0 (7.0–52.0)
0.0092
8.6 (1.6–32.4)
14.0 (2.2–39.4)
0.0552
UACR category, n (%)
0–29
1739 (59)
1813 (61)
202 (73)
189 (69)
30–200
1233 (41)
1159 (39)
73 (27)
86 (31)
ARNI angiotensin receptor/neprilysin inhibitor, BMI body mass index, CKD chronic kidney disease, eGFR estimated glomerular filtration rate, IQR interquartile range, N/A not applicable, RASi renin–angiotensin system inhibitor, SD standard deviation, SMD standardized mean difference, UACR urinary albumin-to-creatinine ratio
aAn SMD of less than 0.1 was considered good balance between covariates
bExact n numbers for cohorts with n < 5 not shown in accordance with Clinformatics Data Mart patient privacy guidelines
×
In the subgroup without type 2 diabetes, 300 dapagliflozin 10 mg initiators and 1284 randomly sampled comparators from 148,000 potential comparators who did not initiate dapagliflozin 10 mg were included. In total, 275 dapagliflozin initiators were matched 1:1 using propensity scores. The groups were well balanced and similar to the UACR < 200 mg/g group in terms of age and sex, but fewer patients (65%) were from the USA. Although comorbidity profiles also looked similar to those in the UACR < 200 mg/g group, a larger proportion of patients in the subgroup had comorbid heart failure (54% of dapagliflozin initiators and 56% of comparators). There was also a larger proportion of patients (58% of dapagliflozin initiators and 56% of comparators) with a prescription for diuretics at baseline than in the UACR < 200 mg/g group. The proportion of patients with UACR < 30 mg/g was higher in this subgroup than the overall UACR < 200 mg/g group (73% and 69% for dapagliflozin initiators and comparators, respectively).
Effectiveness of Initiating Versus Not Initiating Dapagliflozin 10 mg
An overview of the post-index eGFR slopes per group for each analysis can be found in Table 3. Among dapagliflozin initiators with UACR < 200 mg/g, the median eGFR slope was 1.07 mL/min/1.73 m2/year (95% confidence interval [CI] 0.40–1.74) better than in patients who did not initiate treatment. The benefit of dapagliflozin 10 mg initiation was observed across the whole eGFR slope distribution among patients with UACR < 200 mg/g (Fig. S2). The sensitivity analysis, in which post-index eGFR slope data availability was not accounted for before propensity score matching, showed a median eGFR slope difference of 1.19 mL/min/1.73 m2/year (95% CI 0.55–1.83) in favour of dapagliflozin 10 mg initiation.
Table 3
Median eGFR slopes and median difference in eGFR slopes among patients who initiated dapagliflozin 10 mg compared with those who did not
Median eGFR slope, mL/min/1.73 m2/year (95% CI)
Median difference (quantile regression), mL/min/1.73 m2/year (95% CI)
Initiated dapagliflozin 10 mg
Did not initiate dapagliflozin 10 mg
UACR < 200 mg/g
0.18 (−0.50 to 0.99)
−0.83 (−2.23 to 0.39)
1.07 (0.40–1.74)
UACR < 200 mg/g without type 2 diabetes
−0.86 (−6.36 to 4.35)
−2.02 (−5.74 to 0.88)
1.28 (−1.56 to 4.12)
CI confidence interval, eGFR estimated glomerular filtration rate, UACR urinary albumin-to-creatinine ratio
In the subgroup of patients with CKD without type 2 diabetes, the difference was 1.28 mL/min/1.73 m2/year (95% CI −1.56 to 4.12) in favour of dapagliflozin 10 mg initiation.
Given the small sample size for patients with UACR < 200 mg/g without type 2 diabetes, a post hoc analysis was performed which used information from the total cohort of patients with UACR < 200 mg/g to inform estimates of treatment effect among these patients. Results from this analysis found that a weight of 30% on the information from the total cohort was sufficient to result in a significant effect of dapagliflozin initiation (vs non-initiation) on eGFR slope in patients without type 2 diabetes (1.09 mL/min/1.73 m2/year [95% credibility interval 0.02–2.51], see Supplementary Statistics in the Supplementary Materials).
Discussion
SGLT2i are a cornerstone of treatment for CKD and have been proven to reduce cardiorenal outcomes in patients with diabetes and heart failure [16]. In patients with non-diabetic CKD and in those with minimal albuminuria, a need for additional evidence has been highlighted prior to broader implementation. To our knowledge, OPTIMISE-CKD fills this gap as the first study to assess the real-world effectiveness of dapagliflozin 10 mg for the treatment of patients with CKD who have UACR < 200 mg/g.
Our effectiveness analysis showed that dapagliflozin 10 mg initiation was associated with a clinically meaningful attenuation in eGFR slope (1.07 mL/min/1.73 m2/year) among patients with CKD and a UACR < 200 mg/g.
It is important to note that the estimated difference of initiating dapagliflozin 10 mg on eGFR slope in the present analysis was similar to the benefit observed with dapagliflozin (vs placebo) on total slope among patients enrolled in DAPA-CKD (0.95 mL/min/1.73 m2/year [95% CI 0.63–1.27]) [27]. Furthermore, the effect observed in the present study is directionally consistent with findings from EMPA-KIDNEY, a placebo-controlled clinical trial of the SGLT2i empagliflozin in patients with CKD. In EMPA-KIDNEY, empagliflozin (vs placebo) was associated with a difference in annual rate of change in total eGFR slopes of 0.17 mL/min/1.73 m2/year in patients with UACR < 30 mg/g and 0.46 mL/min/1.73 m2/year in patients with UACR 30–300 mg/g [18]. In the subgroup of patients without type 2 diabetes, our results show an effect that is directionally similar to the observed effect of dapagliflozin and empagliflozin on the total eGFR slope in the respective trial population without type 2 diabetes (dapagliflozin, 0.46 mL/min/1.73 m2/year; empagliflozin, 0.62 mL/min/1.73 m2/year) [18, 27]. Our results also trend in the same direction as a post hoc analysis of 24 patients from DAPA-CKD without type 2 diabetes and with baseline UACR 30–200 mg/g (difference in total slope of 0.24 mL/min/1.73 m2/year with dapagliflozin vs placebo) [28]. Our findings complement the results from the clinical trials that mainly included patients with UACR > 200 mg/g and suggest that the reduced rate of CKD progression seen with SGLT2i such as dapagliflozin extends to patients with lower levels of albuminuria. The consistent real-world benefit on eGFR slope observed with dapagliflozin 10 mg in patients with UACR < 200 mg/g is likely to be clinically meaningful because several studies have suggested that a 0.5–1.0 mL/min/1.73 m2/year attenuation of the eGFR slope is associated with a decreased risk of kidney disease progression [29‐31].
Although a positive effect was observed in this patient population, the sample size for dapagliflozin initiators with UACR < 200 mg/g and without type 2 diabetes was small. This low sample size was likely primarily driven by a lack of routine UACR testing in clinical practice. Indeed, our descriptive analysis of dapagliflozin utilization showed that up to 58% of dapagliflozin initiators and 75% of eligible but untreated patients did not have a recorded UACR measure at baseline. These findings are in line with previous studies that have highlighted low UACR testing rates in clinical practice [32, 33]. For example, a recently published observational study of US adults found that only 17.5% of patients at risk for CKD had been tested for albuminuria [34]. It may be expected that UACR testing is performed more often in patients with known or suspected albuminuria, for example among patients with type 2 diabetes. Moreover, UACR testing in Japan is only reimbursed for patients with type 2 diabetes. UPCR testing may therefore be a more common method of assessing kidney damage among patients without type 2 diabetes in Japan. However, the inclusion of UPCR measures converted to UACR equivalents in the present study only minimally increased the sample size, indicating that neither UACR nor UPCR testing is currently widely performed. Recent evidence shows that even patients with stage 3a–4 CKD without type 2 diabetes are at increased risk of adverse cardiorenal outcomes [32], highlighting the importance of early and accurate risk stratification of patients with CKD. In routine clinical practice, healthcare professionals should be encouraged to perform regular UACR testing not just on patients with diagnosed CKD but on all patients at risk of CKD. This change will help to ensure accurate and early risk stratification and timely treatment decisions. In addition, although algorithms to predict UACR values among patients without a measurement are currently being developed and validated for use in research and clinical practice [35], increased testing rates would enable larger studies investigating the burden of CKD in patient groups for which data are currently limited.
Anzeige
Strengths and Limitations
A key strength of this study is the analysis of contemporary data in two countries, allowing insights into early dapagliflozin 10 mg prescribing practices. However, there was some heterogeneity between the US and Japanese populations included in the analysis. For example, BMI and prevalence of type 2 diabetes were lower in the Japanese cohort than in the US cohort. The databases used in this study collected information on a wide range of clinical variables, allowing for the inclusion of multiple potential confounding variables when propensity score matching. It should be noted that these databases were not primarily designed for research purposes. As such, key variables may be incomplete or missing, and the reason for dapagliflozin initiation is not included. The threshold for low albuminuria in this study was set as < 200 mg/g based on the inclusion criteria for DAPA-CKD, which enrolled only patients with a UACR between 200 mg and 5000 mg [17]. This threshold was chosen to investigate the effectiveness of dapagliflozin in a patient group for which there are few existing data from clinical trials.
It is possible that patients who initiate a new treatment are followed up more intensely shortly after initiation, thereby increasing the chance of having a recorded event of interest, while comparators may not have this opportunity. Therefore, it was not possible to assess other outcomes such as hospitalizations that require longer follow-up time or a higher event frequency. Given the relatively recent approval of dapagliflozin for CKD and the resulting short observation window, it was also difficult to determine the long-term effect on eGFR slope. Safety data, such as side effects and reasons for discontinuation, were not examined as part of this study because these are difficult to discern from the claims data used. However, the safety profile of dapagliflozin has been well established in clinical trials for multiple diseases [36]. Furthermore, other factors such as social determinants of health and treatment adherence could not be investigated from the claims data used in this study. Additionally, it was not possible to follow patients if they attended different hospitals in Japan or switched healthcare insurance in the USA, which may have affected the availability of eGFR measurements. The point estimates for the effects of dapagliflozin on eGFR decline observed in the present study appear larger than those observed in clinical trials. However, there is still overlap in the CIs between the two sets of results. Our methodology was designed to adjust for known confounders, although there will be differences between clinical trial populations and real-world patient cohorts. In addition, there may still be unmeasured confounding that could not be accounted for. Differences in baseline characteristics of the patients in the present cohorts compared with clinical trial populations, such as the prevalence of different comorbidities, may have in part accounted for the difference in estimated effect on eGFR slope. Because dapagliflozin 10 mg has multiple indications, patients in the present study may have initiated dapagliflozin for heart failure or uncontrolled type 2 diabetes. However, given that there is still overlap in the CIs between the two sets of results, and all patients included in the present study had clinical CKD, the effect of dapagliflozin 10 mg on the eGFR slope would likely remain. In the case of patients without type 2 diabetes, the prohibitive sample size resulted in large CIs, making it difficult to determine conclusions from this cohort. However, results from a post hoc analysis suggest that only minimal borrowing of information from the total cohort was required to indicate at least a 97.5% posterior probability of a significant treatment effect for dapagliflozin initiation in patients with UACR < 200 mg/g without type 2 diabetes. This suggests that it is reasonable to assume a treatment effect for dapagliflozin initiation in these patients, although these findings should be confirmed with larger studies in this population.
Conclusion
In a real-world cohort of patients with CKD and UACR < 200 mg/g, initiating dapagliflozin 10 mg was associated with a clinically meaningful attenuation of eGFR slope, supplementing available clinical efficacy evidence and suggesting that dapagliflozin effectiveness may extend to patients with lower levels of albuminuria. In line with previous studies, a similar benefit was observed in a small group of patients with UACR < 200 mg/g without type 2 diabetes, although larger studies should be performed in this population.
Acknowledgements
We would like to thank the wider OPTIMISE-CKD study team for their support with the data, definitions and real-world context used in this analysis. The authors would also like to acknowledge the Health, Clinic, and Education Information Evaluation Institute for the database development for the study.
Anzeige
Medical Writing.
Under the direction of the authors, medical writing support for this manuscript was provided by Bobby Thompson, MSc (Res), of Oxford PharmaGenesis, Oxford, UK, and was funded by AstraZeneca.
Declarations
Conflict of Interest
Cassandra Nekeman-Nan, Asuka Ozaki and Stefan Franzén are employees and shareholders of AstraZeneca. Cassandra Nan holds shares in GlaxoSmithKline. Navdeep Tangri has received grants from AstraZeneca, Boehringer Ingelheim/Eli Lilly and Company, Janssen Pharmaceuticals, Otsuka Pharmaceutical Co., Ltd and Tricida, Inc., has received honoraria from AstraZeneca, Boehringer Ingelheim/Eli Lilly and Company, Janssen Pharmaceuticals, Otsuka Pharmaceutical Co., Ltd and Tricida, Inc. and holds stock options from Mesentech, Inc., pulseData, Rénibus Therapeutics, Inc. and Tricida, Inc. Anjay Rastogi has received advisory board and speakers’ bureau fees from AstraZeneca and Relypsa Inc., has received research grants from AstraZeneca and Bayer, and has received consulting fees from Bayer. Lai San Hong was an independent contractor for AstraZeneca at the time that this study was conducted. Tadashi Sofue has received personal fees for lectures from AstraZeneca and Mitsubishi Tanabe Pharma.
Ethical Approval
OPTIMISE-CKD used de-identified data from existing databases and did not require data collection beyond that of routine clinical care. No identifiable information was collected or examined as part of the study. All externally conducted analyses were completed in line with local ethics regulations/legislation. De-identified, internally licensed databases were shared with AstraZeneca by the licensee; therefore, ethics review and approval were not required for the use of these databases for this study.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 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 permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc/4.0/.
Mit e.Med Innere Medizin erhalten Sie Zugang zu CME-Fortbildungen des Fachgebietes Innere Medizin, den Premium-Inhalten der internistischen Fachzeitschriften, inklusive einer gedruckten internistischen Zeitschrift Ihrer Wahl.
Mit e.Med Allgemeinmedizin erhalten Sie Zugang zu allen CME-Fortbildungen und Premium-Inhalten der allgemeinmedizinischen Zeitschriften, inklusive einer gedruckten Allgemeinmedizin-Zeitschrift Ihrer Wahl.
Dapagliflozin Utilization in Chronic Kidney Disease and Its Real-World Effectiveness Among Patients with Lower Levels of Albuminuria in the USA and Japan
verfasst von
Navdeep Tangri Anjay Rastogi Cassandra Nekeman-Nan Lai San Hong Asuka Ozaki Stefan Franzén Tadashi Sofue
Bei ausgeprägter Nierenfunktionsstörung steigen unter der Einnahme von Renin-Angiotensin-System(RAS)-Hemmstoffen nicht selten die Serumkaliumspiegel. Was in diesem Fall zu tun ist, erklärte Prof. Jürgen Floege beim diesjährigen Allgemeinmedizin-Update-Seminar.
Das Risiko, nach einem Gestationsdiabetes einen Typ-2-Diabetes zu entwickeln, hängt nicht nur von der Zahl, sondern auch von der Reihenfolge der betroffenen Schwangerschaften ab.
Gerade bei aggressiven Malignomen wie dem duktalen Adenokarzinom des Pankreas könnte Früherkennung die Therapiechancen verbessern. Noch jedoch klafft hier eine Lücke. Ein Studienteam hat einen Weg gesucht, sie zu schließen.
Im Battle of Experts traten zwei Experten auf dem Diabeteskongress gegeneinander an: Die eine vertrat die Auffassung „Sport statt Spritze“ bei Adipositas und Typ-2-Diabetes, der andere forderte „Spritze statt Sport!“ Am Ende waren sie sich aber einig: Die Kombination aus beidem erzielt die besten Ergebnisse.
Update Innere Medizin
Bestellen Sie unseren Fach-Newsletter und bleiben Sie gut informiert.
