PO:12:179 | Cardiovascular outcomes of treat-to-target urate-lowering therapy in gout: emulated target trials

Edoardo Cipolletta¹, Davide Rozza², Abhishek Abhishek³ | ¹Azienda Ospedaliero Universitaria delle Marche, Ancona, Italy; ²Società Italiana di Reumatologia, Milano, Italy; ³University of Nottingham Nottingham, United Kingdom

Aim. To evaluate the effect of achieving SU <360 µmol/l within 12 months of first prescription of urate-lowering therapy (ULT) on the 5-year risk of major adverse cardiovascular events (MACE) among patients with gout.

Methods. English and Swedish primary-care data linked to hospitalisation and mortality records were obtained from CPRD Aurum (2007-2021) and VEGA (2007-2017). We included people newly prescribed ULT with the latest pre-treatment SU >360 µmol/l. We assigned patients to the T2T-ULT arm if they had a recorded SU measurement <360 µmol/l within 12 months after ULT initiation. We assigned patients to the fire-and-forget arm if they had not had any recorded SU measurements within 12 months after ULT initiation or they had persistent SU levels >360 µmol in the 12 months after ULT initiation. Patients were followed up from ULT initiation to up to 5 years. The first MACE after cohort entry (i.e., non-fatal myocardial infarction, non-fatal stroke, or cardiovascular death) was the primary outcome. Outcomes were ascertained using primary care, secondary care and mortality records. We evaluated the number of gout flares as a positive control outcome and acute bronchitis, cataract, and appendicitis as negative control outcomes. We performed an emulated target trial using a cloning, censoring, and weighting approach. We estimated the 5-year overall survival and survival difference using a non-parametric Kaplan-Meier estimator weighted for the inverse conditional probability of being censored (conditioned on baseline covariates). The 95%CIs were obtained using non-parametric bootstrap with 100 replicates. We calculated HRs and rate ratios and their 95%CIs using robust standard error estimators.

Results. We included data from 116,518 patients, 109,504 and 7,014 from CPRD Aurum and VEGA, respectively (Figure). Patients in CPRD were younger than those in VEGA (mean (SD) age 62.9 (15.2) vs. 70.0 (14.7) years), less often female (24,358 (22.2%) vs. 1,985 (28.3%)), had longer duration of gout since the diagnosis (mean (SD) 2.5 (3.6) vs. 1.1 (1.8) years) and lower pre-treatment serum urate level (mean (SD) 512.4 (83.3) vs. 547.0 (106.8)) µmol/l). In both populations, covariates were well-balanced after weighting (SMD<0.10). The 5-year weighted survival rates in T2T-ULT and fire-and-forget ULT arms were 89.43% (95%CI:88.97-89.89) and 88.03% (95%CI:87.73-88.33) (survival difference:1.40%, 95%CI:0.85-1.95, hazard ratio:0.88, 95%CI:0.86-0.89) in CPRD and 76.82% (95%CI:74.03-79.61) and 75.50% (95%CI:73.99-77.01) (survival difference:1.32%, 95%CI:-1.86 to 4.50, hazard ratio:0.94, 95%CI:0.84-1.04) in VEGA. Findings were similar in analyses exploring different MACE definitions, when follow-up was censored on ULT discontinuation, and when patients without available SU during the first year of follow-up were excluded. A significantly lower number of flares was recorded in those who achieved SU<360 µmol/L We did not observe significant differences for negative control outcomes.

Conclusions. T2T-ULT significantly lowered the risk of MACE in two independent European datasets.