Biological, as opposed to classic antipsoriatic drug or apremilast, treatment mitigates the risk of death and cardiovascular disease in psoriasis

Abstract

Background: Cardiovascular comorbidity increases morbidity and mortality in psoriasis. Systemic treatments, particularly biologics, are effective in alleviating skin and joint inflammation. Conversely, the impact of systemic therapy on cardiovascular disease risk and mortality in psoriasis remains uncertain.

Methods: Impact of systemic treatments on all-cause mortality and cardiovascular disease risk in psoriasis patients' electronic health records (EHRs) from TriNetX was assessed. Treatment categories included apremilast, IL-17 inhibitors (IL-17i), IL23i, TNFi, and classic antipsoriatic drugs. Index event was the first prescription of each treatment, requiring a two-year continuous treatment with exclusion of other systemic antipsoriatic drugs. Propensity-score matching was used to improve comparability. Sensitivity analyses ensured study robustness.

Findings: In descriptive analysis, all-cause mortality rates were 0.61% (classic antipsoriatics, n = 7929), 0.91% (apremilast, n = 1101), 0.00% (IL17i, n = 677), 0.81% (IL23i, n = 1242), and 0.20% (TNFi, n = 6468). Major adverse cardiac events (MACE) were documented in 8.49% (classic antipsoriatics), 5.14% (apremilast), 2.99% (IL17i), 2.09% (IL23i), and 3.74% (TNFi) EHRs. Propensity-score matching showed all-cause mortality rates of 0.23% for any biologic vs. 0.49% for classic antipsoriatics or apremilast, resulting in an HR of 2.21 (95% CI 1.21-3.71, p = 0.0073). MACE risk was also higher with classic antipsoriatics or apremilast (HR 1.66, CI 1.43-1.93, p < 0.0001). The majority of findings were consistent across all four sensitivity analyses. No significant differences in all-cause mortality or MACE risk were observed among biologics.

Interpretation: Biological treatment, as opposed to classic antipsoriatic drugs or apremilast, reduces risk of death and cardiovascular disease in psoriasis. Prospective trials are required to validate these findings.

Funding: DFG: EXC 2167 and LU 877/25-1. State of Schleswig Holstein: Excellence-Chair Program.

Keywords: Cardiovascular; Comorbidity; Major adverse cardiac events (MACE); Mortality; Psoriasis; TriNetX.

Fig. 1

Flow chart of study design.

Fig. 2

Risk of major adverse cardiac events (MACE) in psoriasis exposed to biologic or non-biologic treatment. EHRs of psoriasis patients exposed to biologic or non-biologic systemic antipsoriatic treatment were retrieved from the TriNetX US Collaborative Research Network. Following propensity-score matching, the risk of MACE was contrasted among the groups. Exposure to the respective drug class was ensured during the two-year follow-up period. In 3.917% of EHRs indicating exposure to biologic antipsoriatic treatment (shown in blue) MACE were documented. In those EHRs with exposure to non-biological treatment (shown in red), this risk amounted to 6.426%, translating into a hazard ratio (solid line) of 1.661 (95% confidence interval [shaded area] 1.432–1.925, p < 0.0001).

Fig. 3

Risk of developing the indicated outcomes in psoriasis patients exposed to biologic or non-biologic systemic antipsoriatic treatment. EHRs of psoriasis patients exposed to biologic or non-biologic systemic antipsoriatic treatment were retrieved from the TriNetX US Collaborative Research Network. Following propensity score matching, the risks of the indicated outcomes was analyzed. In the primary analysis (black) exposure to the respective drug class was ensured for the duration of the two-year follow-up. In the first sensitivity analysis (dark red), outcomes within one month following the index event were excluded to mitigate the potential inclusion of prevalent cases of cardiovascular diseases. In the second sensitivity analysis (blue), a more lenient definition of psoriasis was applied to retrieve a larger number of EHRs. In the third sensitivity analysis (orange), follow-up was restricted to one year with ensured drug exposure during that time. In the fourth sensitivity analysis (green), outcomes prior to the index event were excluded at data retrival and follow up was for 5 years to allow for a better matching of the cohorts. Diamonds represent hazard ratios and error bars correspond to the 95% confidence intervals. Diamonds and error bars shown in gray indicate non-significant results. Please note that the confidence intervals displayed are at the 95% level. Due to the application of Bonferroni adjustment α(adjust) is 0.008. The adjusted confidence interval would be 99.2%.