Patient characteristics as effect modifiers for psoriasis biologic treatment response: an assessment using network...

Area of Science:

• Dermatology
• Clinical Epidemiology
• Biostatistics

Background:

• Network meta-analyses (NMAs) are used in psoriasis treatment appraisals.
• Heterogeneity in study populations can impact NMA validity.
• This study investigated the effect of patient characteristics on NMAs of psoriasis treatments.

Purpose of the Study:

• To explore the impact of including studies with heterogeneous patient characteristics on NMA results for psoriasis treatments.
• To assess how subgroup analyses affect treatment effect estimates and rankings in psoriasis NMAs.

Main Methods:

• Identified NMAs for psoriasis Single Technology Appraisals (STAs).
• Constructed five networks: one including all studies and four with restricted patient characteristics (prior biologic use, PASI score, weight, ethnicity).
• Utilized random effects models with log-normal prior distributions for subgroup NMAs.

Main Results:

• Reduced heterogeneity was observed in the four smaller, less heterogeneous networks.
• No significant differences in relative treatment effects (PASI 75 response) were found across the five NMAs, though noticeable differences existed.
• Treatment rankings were generally consistent, except for the network of biologic-naïve patients, where anti-TNF therapies ranked higher.

Conclusions:

• Heterogeneity in patient characteristics can influence NMA results, particularly for biologic-naïve populations.
• Clinically relevant subgroup analyses are crucial for assessing effect modification related to patient characteristics in NMAs.
• Careful assessment of trial heterogeneity is recommended for all NMAs.