2025 年 7 巻 3 号 p. 183-190
Background: Studies have shown an increased risk of cardiovascular events during treatment with febuxostat vs. allopurinol, but comparative data with another xanthine oxidoreductase inhibitor (XORi), topiroxostat, are lacking. In this retrospective study we compared the incidence of cardiovascular/renal events in Japanese patients with newly diagnosed hyperuricemia and/or gout treated with allopurinol, febuxostat or topiroxostat.
Methods and Results: Data came from the JMDC Claims Database from September 2013–September 2019. Participants (n=24,112, age ≥20 years, ≥93% male) were diagnosed with hyperuricemia and/or gout and prescribed XORi treatment in the same month or the following month. Using a Poisson regression model, the adjusted risk (rate ratio [RR]; 95% confidence interval [CI]) of major adverse cardiovascular events was slightly lower with topiroxostat vs. allopurinol (0.63; 0.28–1.41) and febuxostat (0.64; 0.31–1.30). Adjusted risks (RR [95% CI]) for events during treatment with topiroxostat vs. febuxostat and allopurinol were 0.22 [0.10–0.48] and 0.26 [0.11–0.63], respectively, for heart failure, 0.43 [0.27–0.67] and 0.51 [0.31–0.86], respectively, for total cardiovascular events, and 0.46 [0.30–0.69] and 0.62 [0.39–0.98], respectively, for total cardiovascular + renal events. Adjusted risks of atrial fibrillation, heart failure, dialysis, total cardiovascular events, and total cardiovascular + renal events were significantly higher with febuxostat vs. allopurinol.
Conclusions: Topiroxostat may provide a better tolerated option for the treatment of hyperuricemia and/or gout in Japanese patients with respect to cardiovascular events.
Uric acid is the end product of purine metabolism in humans, and xanthine oxidoreductase (XOR) is a crucial enzyme for the catabolism of purines. The main activities of XOR are xanthine dehydrogenase activity, which performs the last 2 steps of purine catabolism, and xanthine oxidase activity, which (in addition to purine catabolism) produces reactive oxygen species (ROS).1,2 High XOR activity can contribute to elevated serum uric acid levels and is also associated with elevated levels of ROS, leading to tissue damage. It is therefore not surprising that hyperuricemia is a risk factor for vascular complications, including hypertension, chronic kidney disease (CKD) and cardiovascular disease (CVD).3–8
There are 2 main groups of uric acid-lowering drugs (ULDs): uric acid excretion stimulants and uric acid production inhibitors. There are 3 commercially available XOR inhibitors (XORi) that are used as uric acid production inhibitors: allopurinol, febuxostat and topiroxostat. Of these, allopurinol and febuxostat are widely available globally, but topiroxostat is currently only used in Japan. Although these XORi agents have the same general mechanism of action, there are some differences between them. For example, animal data suggest that topiroxostat may be superior to febuxostat for suppressing plasma XOR activity in some settings.9,10
There also appear to be some differences in the cardiovascular safety of febuxostat vs. allopurinol based on clinical trial data in patients with gout,11–13 and in patients undergoing hemodialysis.14 In addition, meta-analysis data indicate an increased risk of adverse cardiovascular events with febuxostat compared with allopurinol, particularly during long-term follow-up.15,16 These differences have also been documented in the US-based multicenter, double-blind, noninferiority Cardiovascular Safety of Febuxostat and Allopurinol in Patients with Gout and Cardiovascular Morbidities (CARES) trial, where rates of cardiovascular and all-cause death were significantly higher in patients with gout and CVD who were treated with febuxostat vs. allopurinol.17 However, the UK-based multicenter, open-label noninferiority Febuxostat vs. Allopurinol Streamlined Trial (FAST) showed that febuxostat was non-inferior to allopurinol with respect to the primary cardiovascular endpoint (a composite of hospitalization for non-fatal myocardial infarction or biomarker-positive acute coronary syndrome, non-fatal stroke or cardiovascular death), and that there was no long-term risk increased risk of death or serious adverse events during treatment of patients with gout and ≥1 additional cardiovascular risk factor with febuxostat vs. allopurinol.18
In Japan, the availability of topiroxostat as a third XORi option for the management of hyperuricemia and gout means that treatment patterns differ from those in the USA and Europe, and there is a lack of data regarding the comparative cardiovascular risk profile of the 3 XORi agents. Furthermore, rates of coronary artery disease are generally lower in Japanese vs. American and European populations,19 and the proportion of individuals of Asian ethnicity enrolled in large clinical trials to date has been very small.17,18 Therefore, it is unclear whether data obtained in those regions can be extrapolated to Japan.
For this study, we used the JMDC Claims Database, the largest commercially available receipts database in Japan, to compare the incidence of cardiovascular and renal events in Japanese patients with newly diagnosed hyperuricemia and/or gout who were started on treatment with allopurinol, febuxostat or topiroxostat.
The JMDC Claims Database, which is an epidemiological receipt database that has accumulated receipts (inpatient, outpatient, dispensing) and medical examination data received from multiple health insurance associations in Japan since 2005, was used for this retrospective analysis. Data were obtained for the period from September 2013 to September 2019, and included at least the 12-month period before initiation of XORi treatment for hyperuricemia and/or gout. The index date, defined as the time of initiation of XORi therapy with allopurinol, febuxostat or topiroxostat after the first diagnosis of hyperuricemia and/or gout, was between September 2014 and October 2017. An index diagnosis of hyperuricemia was identified using the ICD-10 code E790, and an index diagnosis of gout was identified using the ICD-10 code M10.
The study was approved by the Research Institute of Healthcare Data Science (RI2021013) and as an epidemiological study we used existing anonymized processed information and data containing personal information were not accessed. Therefore, under Japanese laws and regulations, patient consent was not required for the secondary use of this anonymized data.
Patients and ProceduresEligible participants were aged ≥20 years, had been diagnosed with hyperuricemia and/or gout and prescribed treatment with a XORi in the same month or the following month, and had ≥12 months of medical data available before the index date. Individuals who had a prescription for a diuretic on the index date, had a diagnosis of heart failure, and those with a prescription of urate excretion-promoting drugs during the follow-up period (after the diagnosis of hyperuricemia or gout) were excluded.
TreatmentsATC code definitions for ULDs are shown in Supplementary Table 1. All participants were prescribed allopurinol, febuxostat or topiroxostat for the treatment of hyperuricemia and/or gout. Continuation of a prescription for a XORi was defined as the period of time from the date of XORi prescription (i.e., date of prescription + number of days of treatment provided by the prescription − 1 day), and the next XORi prescription had to occur within 90 days of the end of the scheduled period of the previous prescription. The date of XORi therapy discontinuation was defined as the last scheduled day of existing prescription. If a different XORi was prescribed, the prescription discontinuation date for the previous agent was taken as the day before the start of a new XORi prescription. The duration of prescription was defined as the period from the index date to the date of discontinuation.
OutcomesDefinitions of each type of endpoint event are shown in Table 1 and more detailed lists of the diagnostic codes are provided in Supplementary Tables 2–4. All of these events required hospitalization. The composite outcomes of interest were the incidence of total cardiovascular events (major adverse cardiovascular events [MACE: angina pectoris, acute myocardial infarction, intracerebral hemorrhage, cerebral infarction] + aortic aneurysm/dissection + heart failure + atrial fibrillation/flutter) and the incidence of total cardiovascular/renal events (total cardiovascular events [as above] + transition to dialysis).
Outcome Event Definitions
| Outcome | ICD-10 code | Prescriptions/procedures on admission |
|---|---|---|
| MACE | ||
| Angina pectoris | I20 | PCI/CABG/angiography/IABP/ventricular assist |
| Acute myocardial infarction | I21 | PCI/CABG/angiography/IABP/ventricular assist |
| Intracerebral hemorrhage | I61 | Antihypertensive drug prescriptions (injection)/systemic hemostatic drug prescriptions (injection) [ATC code: B02G] / antifibrinolytic agent prescriptions (injection) [ATC code: B02A] / antiedema drug prescriptions (injection) [ATC code: K01F] / catheter intervention/hematoma evacuation |
| Cerebral infarction | I63 | Tissue plasminogen activator prescriptions (injection)/brain protectant drug prescriptions (injection) [ATC code: N07X] / heparin prescriptions (injection)/antiplatelet drug prescriptions (injection) [ATC code: B01C] / direct thrombin inhibitor (injection) [ATC code: B01E] / antiedema drug prescriptions (injection) [ATC code: K01F] / catheter intervention |
| Aortic aneurysm/dissection | I71, I72 | |
| Heart failure | I50 | Diuretic drug prescriptions (injection for outpatients) |
| Atrial fibrillation/flutter | I48 | |
| Transition to dialysis | J038, J042 (medical fee code) |
|
MACE, major adverse cardiovascular events; PCI, percutaneous coronary intervention; CABG, coronary artery bypass grafting; IABP, intra-aortic balloon pump.
Statistical Analysis
Clinical and demographic data for each drug group are presented using descriptive statistics (mean±standard deviation or number of participants and percentage). Event rates in each drug group are reported as the incidence per 1,000 person-years with the 95% confidence interval (CI).
The adjusted incidence rate ratio (RR) for each pair of drugs was calculated using a Poisson regression model, with the following factors included as covariates: age group (<40, 40–64 and ≥65 years); sex (male/female); hypertension (yes/no); diabetes (yes/no); malignancy (yes/no); renal disease (yes/no); and history of MACE (yes/no).
Matched cohorts in each drug pair group (topiroxostat vs. allopurinol, topiroxostat vs. febuxostat, and febuxostat vs. allopurinol) were creating using propensity score matching based on the following variables: age group (<40, 40–64 and ≥65 years); sex (male/female); prescription for an angiotensin-converting enzyme inhibitor or angiotensin-receptor blocker (yes/no); prescription for a calcium-channel blocker (yes/no); prescription for diuretics (yes/no); prescription for a β-blocker (yes/no); prescription for other antihypertensive drugs (yes/no); prescription for a dipeptidyl peptidase 4 inhibitor (yes/no); prescription for a sodium-glucose cotransporter 2 inhibitor (yes/no); prescription for a biguanide (yes/no); prescription for an α-glucosidase inhibitor (yes/no); prescription for a thiazolidinedione (yes/no); prescription for a sulfonylurea (yes/no); prescription for a glinide (yes/no); prescription for insulin (yes/no); prescription for a glucagon-like peptide 1 receptor agonist (yes/no); history of malignancy (yes/no); and history of renal disease (yes/no). A Cox proportional hazards model for the matched cohorts was used to estimate the adjusted hazard ratio (aHR) or crude hazard ratio (cHR) and 95% CI values for the occurrence of events in the drug group of interest compared with the control drug group (reference). The following factors were included as covariates in the adjusted model: age group (<40, 40–64 and ≥65 years); sex (male/female); hypertension (yes/no); diabetes (yes/no); history of MACE (yes/no). The Kaplan-Meir plot was analyzed using the log-rank test.
All statistical analyses were performed using SAS statistical software (version 9.4; SAS Institute).
Of the 152,974 individuals with a first diagnosis of hyperuricemia and/or gout during the study inclusion period, 24,112 met all the inclusion criteria and were included in the analysis (Figure). In the primary evaluable population, 6,401 individuals were treated with allopurinol, 16,021 with febuxostat and 1,690 with topiroxostat. The majority of patients (≥93%) were male, and the most common diagnosis in all treatment groups was hyperuricemia only, hypertension was the most common comorbidity, and approximately one-third of the total population were receiving antihypertensive drug therapy (Table 2). The patient flowchart and baseline clinical and demographic characteristics in the 3 matching cohorts (topiroxostat vs. allopurinol, topiroxostat vs. febuxostat, and febuxostat vs. allopurinol) are shown in Supplementary Figure 1 and Supplementary Table 5.
Flow chart of participant selection and inclusion. ULD, uric acid-lowering drug.
Baseline Clinical and Demographic Characteristics of the Study Population
| Parameter | Allopurinol (n=6,401) |
Febuxostat (n=16,021) |
Topiroxostat (n=1,690) |
|---|---|---|---|
| Age, years | 48.4±10.8 | 48.9±11.0 | 48.4±11.1 |
| Age group, n (%) | |||
| 20–39 years | 1,325 (20.7) | 3,153 (19.7) | 372 (22.0) |
| 40–64 years | 4,721 (73.8) | 11,762 (73.4) | 1,204 (71.2) |
| ≥65 years | 355 (5.5) | 1,106 (6.9) | 114 (6.7) |
| Male sex, n (%) | 6,053 (94.6) | 14,918 (93.1) | 1,572 (93.0) |
| Index diagnosis, n (%) | |||
| Hyperuricemia only | 4,292 (67.1) | 11,026 (68.8) | 1,150 (68.0) |
| Gout only | 1,430 (22.3) | 3,489 (21.8) | 326 (19.3) |
| Hyperuricemia and gout | 679 (10.6) | 1,506 (9.4) | 214 (12.7) |
| Comorbidities, n (%) | |||
| Hypertensiona | 1,714 (26.8) | 5,418 (33.8) | 517 (30.6) |
| Dyslipidemiab | 1,025 (16.0) | 2,984 (18.6) | 358 (21.2) |
| Diabetes mellitusc | 339 (5.3) | 1,354 (8.5) | 143 (8.5) |
| Angina pectoris | 259 (4.0) | 984 (6.1) | 79 (4.7) |
| Acute myocardial infarction | 35 (0.5) | 159 (1.0) | 7 (0.4) |
| Intracerebral hemorrhage | 25 (0.4) | 54 (0.3) | 4 (0.2) |
| Cerebral infarction | 110 (1.7) | 372 (2.3) | 42 (2.5) |
| Heart failure | 266 (4.2) | 1,319 (8.2) | 88 (5.2) |
| Atrial fibrillation/flutter | 99 (1.5) | 454 (2.8) | 30 (1.8) |
| Malignancy | 276 (4.3) | 779 (4.9) | 50 (3.0) |
| Chronic kidney disease | 393 (6.1) | 1,838 (11.5) | 194 (11.5) |
| Liver disease | 1,438 (22.5) | 3,307 (20.6) | 367 (21.7) |
| Concomitant medications, n (%) | |||
| Antihypertensive drugs | 1,835 (28.7) | 5,776 (36.1) | 547 (32.4) |
| Antihyperlipidemic drugs | 1,050 (16.4) | 3,087 (19.3) | 365 (21.6) |
| Antidiabetic drugs | 361 (5.6) | 1,453 (9.1) | 147 (8.7) |
Values are mean±standard deviation or number of patients (%). aHypertension was defined as diagnoses accompanied by prescription containing an angiotensin-converting enzyme inhibitor, angiotensin-receptor blocker, calcium-channel blocker, diuretic, β-blocker, or other antihypertensive drug (see Supplementary Table 3 for details) for treatment in an outpatient setting at the time of the index diagnosis. bDyslipidemia was defined as diagnoses accompanied by prescription containing a statin, fibrate or other cholesterol- or triglyceride-regulating drugs (see Supplementary Table 3 for details) for treatment in an outpatient setting at the time of the index diagnosis. cDiabetes mellitus was defined as diagnoses accompanied by prescription containing antidiabetic drug(s) (see Supplementary Table 3 for details) in an outpatient setting at the time of the index diagnosis.
Outcome Events
In the analysis calculated using a Poisson regression model, the adjusted risk (RR [95% CI]) of MACE was slightly lower with topiroxostat vs. allopurinol (0.63 [0.28–1.41]) and febuxostat (0.64 [0.31–1.30]). Furthermore, the adjusted risk of heart failure (RR 0.26, 95% CI 0.11–0.63), total cardiovascular events was significantly lower in the topiroxostat vs. allopurinol group (RR 0.51, 95% CI 0.31–0.86), and total cardiovascular and renal events (RR 0.62 95% CI 0.39–0.98) (Table 3). Results were similar for the comparison between topiroxostat and febuxostat, with a significantly lower adjusted risk of heart failure (RR 0.22, 95% CI 0.10–0.48), total cardiovascular events (RR 0.43, 95% CI 0.27–0.67) and total cardiovascular and renal events (RR 0.46, 95% CI 0.30–0.69) in the topiroxostat group (Table 4). In contrast, the adjusted risks of atrial fibrillation (RR 1.80, 95% CI 1.17–2.78), heart failure (RR 1.46, 95% CI 1.08–1.99), dialysis (RR 4.71, 95% CI 1.73–12.83), total cardiovascular events (RR 1.33, 95% CI 1.06–1.66), and total cardiovascular and renal events (RR 1.43, 95% CI 1.15–1.78) were significantly higher in the febuxostat vs. allopurinol group (Table 5). The Kaplan-Meir plot of the adjusted cumulative incidence rate of outcomes calculated based on the Cox regression model is shown as Supplementary Figure 2. The same trend was observed in the matching cohorts (Supplementary Table 6).
Incidence of Endpoint Events in the Topiroxostat vs. Allopurinol Group
| Event | Allopurinol | Topiroxostat | Incidence rate ratio* (Ref=allopurinol) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time at risk (person- years) |
Number of events |
Incidence rate (per 1,000 person- years) |
95% CI | Time at risk (person- years) |
Number of events |
Incidence rate (per 1,000 person- years) |
95% CI | Non-adjusted | Adjusted** | |||
| Point estimate |
95% CI | Point estimate |
95% CI | |||||||||
| MACE | 8,058 | 32 | 4.0 | 2.7–5.6 | 2,275 | 8 | 3.5 | 1.5–6.9 | 0.89 | 0.41– 1.92 |
0.63 | 0.28– 1.41 |
| Angina pectoris |
8,077 | 17 | 2.1 | 1.2–3.4 | 2,278 | 3 | 1.3 | 0.3–3.8 | 0.63 | 0.18– 2.14 |
0.50 | 0.14– 1.75 |
| Acute myocardial infarction |
8,078 | 6 | 0.7 | 0.3–1.6 | 2,282 | 1 | 0.4 | 0.0–2.4 | 0.59 | 0.07– 4.90 |
0.46 | 0.05– 4.13 |
| Intracerebral hemorrhage |
8,090 | 2 | 0.2 | 0.0–0.9 | 2,282 | 2 | 0.9 | 0.1–3.2 | 3.55 | 0.50– 25.18 |
1.94 | 0.24– 15.76 |
| Cerebral infarction |
8,081 | 9 | 1.1 | 0.5–2.1 | 2,279 | 2 | 0.9 | 0.1–3.2 | 0.79 | 0.17– 3.65 |
0.56 | 0.12– 2.72 |
| Atrial fibrillation/ flutter |
8,065 | 24 | 3.0 | 1.9–4.4 | 2,272 | 6 | 2.6 | 1.0–5.7 | 0.89 | 0.36– 2.17 |
0.76 | 0.31– 1.90 |
| Heart failure | 8,035 | 49 | 6.1 | 4.5–8.1 | 2,278 | 6 | 2.6 | 1.0–5.7 | 0.43 | 0.19– 1.01 |
0.26 | 0.11– 0.63 |
| Aortic aneurysm/ dissection |
8,078 | 12 | 1.5 | 0.8–2.6 | 2,282 | 2 | 0.9 | 0.1–3.2 | 0.59 | 0.13– 2.64 |
0.62 | 0.14– 2.78 |
| Dialysis | 8,091 | 4 | 0.5 | 0.1–1.3 | 2,281 | 5 | 2.2 | 0.7–5.1 | 4.43 | 1.19– 16.51 |
2.27 | 0.57– 9.09 |
| Total CVE | 7,983 | 94 | 11.8 | 9.5–14.4 | 2,260 | 19 | 8.4 | 5.1–13.1 | 0.71 | 0.44– 1.17 |
0.51 | 0.31– 0.86 |
| CVRE | 7,983 | 96 | 12.0 | 9.7–14.7 | 2,260 | 24 | 10.6 | 6.8–15.8 | 0.88 | 0.56– 1.38 |
0.62 | 0.39– 0.98 |
*Exponentially transformed values of parameter estimates by Poisson regression model with number of events as the dependent variable and drug as the explanatory variable. **Adjusted for age group (<40, 40–64 and ≥65 years), sex (male, female), hypertension (no, yes), diabetes (no, yes), malignancy (no, yes), CKD (no, yes), and history of MACE (no, yes). CI, confidence interval; MACE, major adverse cardiovascular events; CVE, cardiovascular events; CVRE, cardiovascular and renal events.
Incidence of Endpoint Events in the Topiroxostat vs. Febuxostat Group
| Event | Febuxostat | Topiroxostat | Incidence rate ratio* (Ref=febuxostat) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time at risk (person- years) |
Number of events |
Incidence rate (per 1,000 person- years) |
95% CI | Time at risk (person- year) |
Number of events |
Incidence rate (per 1,000 person- years) |
95% CI | Non-adjusted | Adjusted** | |||
| Point estimate |
95% CI | Point estimate |
95% CI | |||||||||
| MACE | 25,023 | 144 | 5.8 | 4.9–6.8 | 2,275 | 8 | 3.5 | 1.5–6.9 | 0.61 | 0.30– 1.25 |
0.64 | 0.31– 1.30 |
| Angina pectoris |
25,119 | 63 | 2.5 | 1.9–3.2 | 2,278 | 3 | 1.3 | 0.3–3.8 | 0.52 | 0.16– 1.67 |
0.58 | 0.18– 1.84 |
| Acute myocardial infarction |
25,172 | 31 | 1.2 | 0.8–1.7 | 2,282 | 1 | 0.4 | 0.0–2.4 | 0.36 | 0.05– 2.61 |
0.37 | 0.05– 2.69 |
| Intracerebral hemorrhage |
25,204 | 9 | 0.4 | 0.2–0.7 | 2,282 | 2 | 0.9 | 0.1–3.2 | 2.45 | 0.53– 11.36 |
2.26 | 0.49– 10.55 |
| Cerebral infarction |
25,139 | 56 | 2.2 | 1.7–2.9 | 2,279 | 2 | 0.9 | 0.1–3.2 | 0.39 | 0.10– 1.61 |
0.39 | 0.10– 1.61 |
| Atrial fibrillation/ flutter |
25,016 | 149 | 6.0 | 5.0–7.0 | 2,272 | 6 | 2.6 | 1.0–5.7 | 0.44 | 0.20– 1.00 |
0.45 | 0.20– 1.02 |
| Heart failure | 24,808 | 294 | 11.9 | 10.5–13.3 | 2,278 | 6 | 2.6 | 1.0–5.7 | 0.22 | 0.10– 0.50 |
0.22 | 0.10– 0.48 |
| Aortic aneurysm/ dissection |
25,155 | 37 | 1.5 | 1.0–2.0 | 2,282 | 2 | 0.9 | 0.1–3.2 | 0.60 | 0.14– 2.47 |
0.59 | 0.14– 2.44 |
| Dialysis | 25,143 | 110 | 4.4 | 3.6–5.3 | 2,281 | 5 | 2.2 | 0.7–5.1 | 0.50 | 0.20– 1.23 |
0.47 | 0.19– 1.16 |
| Total CVE | 24,573 | 473 | 19.2 | 17.6–21.1 | 2,260 | 19 | 8.4 | 5.1–13.1 | 0.44 | 0.28– 0.69 |
0.43 | 0.27– 0.67 |
| CVRE | 24,528 | 544 | 22.2 | 20.4–24.1 | 2,260 | 24 | 10.6 | 6.8–15.8 | 0.48 | 0.32– 0.72 |
0.46 | 0.30– 0.69 |
*Exponentially transformed values of parameter estimates by Poisson regression model with number of events as the dependent variable and drug as the explanatory variable. **Adjusted for age group (<40, 40–64 and ≥65 years), sex (male, female), hypertension (no, yes), diabetes (no, yes), malignancy (no, yes), CKD (no, yes), and history of MACE (no, yes). Abbreviations as in Table 3.
Incidence of Endpoint Events in the Febuxostat vs. Allopurinol Group
| Event | Allopurinol | Febuxostat | Incidence rate ratio* (Ref=allopurinol) |
|||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Time at risk (person- years) |
No. of events |
Incidence rate (per 1,000 person- years) |
95% CI | Time at risk (person- year) |
No. of events |
Incidence rate (per 1,000 person- years) |
95% CI | Non-adjusted | Adjusted** | |||
| Point estimate |
95% CI | Point estimate |
95% CI | |||||||||
| MACE | 8,058 | 32 | 4.0 | 2.7–5.6 | 25,023 | 144 | 5.8 | 4.9–6.8 | 1.45 | 0.99– 2.13 |
1.18 | 0.80– 1.74 |
| Angina pectoris |
8,077 | 17 | 2.1 | 1.2–3.4 | 25,119 | 63 | 2.5 | 1.9–3.2 | 1.19 | 0.70– 2.04 |
0.92 | 0.54– 1.59 |
| Acute myocardial infarction |
8,078 | 6 | 0.7 | 0.3–1.6 | 25,172 | 31 | 1.2 | 0.8–1.7 | 1.66 | 0.69– 3.98 |
1.38 | 0.57– 3.34 |
| Intracerebral hemorrhage |
8,090 | 2 | 0.2 | 0.0–0.9 | 25,204 | 9 | 0.4 | 0.2–0.7 | 1.45 | 0.31– 6.69 |
1.16 | 0.24– 5.46 |
| Cerebral infarction |
8,081 | 9 | 1.1 | 0.5–2.1 | 25,139 | 56 | 2.2 | 1.7–2.9 | 2.00 | 0.99– 4.05 |
1.72 | 0.85– 3.50 |
| Atrial fibrillation/ flutter |
8,065 | 24 | 3.0 | 1.9–4.4 | 25,016 | 149 | 6.0 | 5.0–7.0 | 2.00 | 1.30– 3.08 |
1.80 | 1.17– 2.78 |
| Heart failure | 8,035 | 49 | 6.1 | 4.5–8.1 | 24,808 | 294 | 11.9 | 10.5–13.3 | 1.94 | 1.44– 2.63 |
1.46 | 1.08– 1.99 |
| Aortic aneurysm/ dissection |
8,078 | 12 | 1.5 | 0.8–2.6 | 25,155 | 37 | 1.5 | 1.0–2.0 | 0.99 | 0.52– 1.90 |
0.88 | 0.46– 1.71 |
| Dialysis | 8,091 | 4 | 0.5 | 0.1–1.3 | 25,143 | 110 | 4.4 | 3.6–5.3 | 8.85 | 3.26– 24.01 |
4.71 | 1.73– 12.83 |
| Total CVE | 7,983 | 94 | 11.8 | 9.5–14.4 | 24,573 | 473 | 19.2 | 17.6–21.1 | 1.64 | 1.31– 2.04 |
1.33 | 1.06– 1.66 |
| CVRE | 7,983 | 96 | 12.0 | 9.7–14.7 | 24,528 | 544 | 22.2 | 20.4–24.1 | 1.84 | 1.49– 2.29 |
1.43 | 1.15– 1.78 |
*Exponentially transformed values of parameter estimates by Poisson regression model with number of events as the dependent variable and drug as the explanatory variable. **Adjusted for age group (<40, 40–64 and ≥65 years), sex (male, female), hypertension (no, yes), diabetes (no, yes), malignancy (no, yes), CKD (no, yes), and history of MACE (no, yes). Abbreviations as in Table 3.
The key finding of this study was that treatment of hyperuricemia and/or gout with topiroxostat was associated with a lower risk of cardiovascular and renal events than treatment with allopurinol or febuxostat. In addition, the risk of several important cardiovascular events and renal events were significantly higher in individuals with hyperuricemia and/or gout who were treated with febuxostat compared with allopurinol.
There are very limited existing data on the cardiovascular safety of topiroxostat, including how this compares with the other currently available XORi agents. A previous Japanese study utilized the National Database of Health Insurance Claims and Specific Health Checkups to compare the rate of cardiovascular events in people treated with febuxostat or topiroxostat vs. allopurinol.20 Compared with allopurinol, the 10-year cardiovascular event risk was lower with both febuxostat (aHR 0.97, 95% CI 0.95–0.98) and topiroxostat (aHR 0.84, 95% CI 0.78–0.90). However, febuxostat and topiroxostat were not directly compared. Our data confirm the previous findings of significantly lower cardiovascular risk with topiroxostat vs. allopurinol and extend this to also include the risk of renal events, and directly compare topiroxostat and febuxostat for the first time.
The mechanisms underlying the higher rates of MACE and other cardiovascular events with febuxostat seen in our study and the CARES trial (especially after discontinuation of febuxostat in CARES21,22) remain unclear. In CARES, the rate of various events increased early after discontinuation of the study drug, suggesting that the rapid increase in serum uric acid levels after therapy ended may have caused an acute inflammatory reaction (cardiovascular gout attack) due to the formation of free uric acid crystals in the blood vessels and cardiovascular system.21,22 Serum uric acid levels before study drug administration were similar in the 2 groups in the CARES study, but more patients in the febuxostat vs. allopurinol group had a serum uric acid level <5.0 mg/dL,18 and it is speculated that there was greater rebound hyperuricemia after drug discontinuation in the febuxostat group. In the current study, the duration of prescription was defined as the period from the prescription date of the uric acid production inhibitor to the prescription date + prescription days − 1 day, which may have captured events in the early phase after drug discontinuation, but this cannot be guaranteed. Also, inherent differences between the agents, including the frequency of drug administration, the mode of binding to the XOR and the effect on ABCG2,23,24 may have contributed to the differing cardiovascular event rates with topiroxostat, febuxostat and allopurinol in our study.
Another potential mechanism for the lower risk of cardiovascular events in individuals treated with topiroxostat could be its beneficial effects on vascular endothelial function. A Japanese study showed that topiroxostat may improve vascular endothelial function based on a significant improvement in flow-mediated dilatation in patients with hyperuricemia.25 Topiroxostat has also been shown to improve arterial stiffness (determined using the cardio-ankle vascular index) in hyperuricemic individuals with liver dysfunction.26 In contrast, a previous study of people with hyperuricemia not limited to those with liver dysfunction failed to detect any significant effect of topiroxostat on arterial stiffness, although topiroxostat (but not febuxostat) did significantly reduce plasma XOR activity compared with baseline.27 On the other hand, in basic studies, it has also been reported that topiroxostat has a stronger inhibitory effect on plasma XOR activity than other drugs.9,10 Clinically, it has been reported that XOR-mediated oxidative stress, as well as uric acid, is involved in the development of CKD and CVD.25 These findings suggest that the inhibition of plasma XOR activity by topiroxostat may have led to the results of this study through improvement of vascular endothelial function.
Our findings of a higher rate of adverse cardiovascular events in individuals treated with febuxostat compared with allopurinol confirm the majority of existing data.11–13,15–17 In addition to the large randomized trial from the USA,17 studies in Asian populations have also shown greater cardiovascular risk with febuxostat. A Taiwanese cohort study using a health insurance database reported a significantly higher risk of adverse cardiovascular events in patients with hyperuricemia or gout who were treated with febuxostat vs. allopurinol.28 Furthermore, a meta-analysis of data from Asian individuals also found that the use of febuxostat compared with allopurinol was associated with a significantly greater risk of adverse cardiovascular events.16 In contrast, the previous Japanese database study did not find any evidence of an increase in cardiovascular risk with febuxostat vs. allopurinol.20
Study Strengths and LimitationsBeing the first comparison of topiroxostat and febuxostat with respect to cardiovascular events, having a large sample size, and using an excellent patient follow-up database are key strengths of this study. Although the Japanese Society of Gout and Uric & Nucleic acids 2019 guidelines for management of hyperuricemia and gout29 define hyperuricemia as a serum uric acid level >7.0 mg/dL, and suggest drug treatment depending on the presence or absence of gout arthritis or tophi, plus the presence or absence of complications, they do not specify the drug treatment flow. Conversely, US30 and European Union31 guidelines include a flowchart of drug treatment, and the results of this study may be useful to inform a similar flowchart for the Japanese guidelines.
Nevertheless, several limitations need to be considered when interpreting our findings. Some of these relate to the use of the receipts database, whereby the diagnosis stated on the receipt at the time of the medical visit may not fully represent an individual’s health status. In addition, medication usage was based on prescription receipt data and we cannot conclusively state that treatments were used as prescribed, or know the level of compliance with drug therapy. Furthermore, we do not have any data on laboratory test results and therefore the existence of residual confounding (despite population matching) cannot be ruled out, including other unmeasured factors. Outcome determination was also based on receipts and there was no external clinical validation of any of the cardiovascular or renal events that occurred during the study. There is also the possibility that events that occurred on multiple occasions were counted each time rather than being categorized as a single event. Furthermore, because deaths could not be accurately identified in this study, it was not possible to assess total and cardiovascular deaths. Although comparisons were made between drug pair groups, an untreated control group was not included, meaning that the background rate of cardiovascular events in this population was unknown. In addition, the number of patients in the topiroxostat group was much smaller than those in the other two groups. Finally, the findings are only generalizable to the subset of the Japanese population covered by the receipt database, which may not be representative of the overall population in Japan; for example, no individuals aged ≥75 years were included in the analysis and the proportion of individuals aged ≥65 and women was also small. The fact that topiroxostat is not yet internationally available is also a limitation in terms of the generalizability of our study findings.
Our findings require validation in randomized trials, including those that enroll older individuals, both sexes and those from different ethnic groups.
Topiroxostat may provide a better tolerated option for the treatment of hyperuricemia and/or gout in Japanese patients under the age of 65 years with respect to cardiovascular event risk.
The authors thank JMDC Inc. independent data analysis. Medical writing assistance was provided by Nicola Ryan, independent medical writer, funded by Sanwa Kagaku Kenkyusho Co., Ltd.
K.K. has received research grants from Otsuka Pharmaceutical Co., Ltd., Daiichi Sankyo Company, Limited, Sumitomo Pharma Co., Ltd., and Nippon Boehringer Ingelheim Co., Ltd., consulting fee from Sanwa Kagaku Kenkyusho Co., Ltd. and honoraria from Otsuka Pharmaceutical Co., Ltd. Novartis AG, Daiichi Sankyo Company, Limited, and Viatris Inc. K.K. has participated in an Advisory Board for Novartis AG and Daiichi Sankyo Company, Ltd. S.A. is a full-time employee of Sanwa Kagaku Kenkyusho Co., Ltd. H.K. has no conflicts of interest to declare.
This study was funded by Sanwa Kagaku Kenkyusho Co., Ltd.
This study was approved by the Research Institute of Healthcare Data Science (RI2021013).
The data that support the findings of this study are available from JMDC but were used under license for the current study; therefore, restrictions apply and the data are not publicly available. For inquiries about access to the dataset used in this study, please contact JMDC (https://www.jmdc.co.jp).
Please find supplementary file(s);
https://doi.org/10.1253/circrep.CR-24-0178
