CNI Trough Variability Does Not Reliably Reflect Medication Adherence: Insights From a 3-Year Follow-Up Study

Calcineurin inhibitor (CNI) trough concentrations and their variability are frequently used as adherence proxies, despite limited validation. We evaluated the association between self-reported adherence and CNI exposure during the first year after kidney transplantation. We included 619 patients from two prospective French cohorts (14,829 C₀ values). Adherence was assessed using the MMAS-4 questionnaire. CNI exposure was evaluated via C₀ levels, intra-patient variability (IPV; CV threshold = 30%), and underexposure rates. Cross-sectional and longitudinal analyses were performed. No significant differences in C₀, IPV, or underexposure were observed between adherent and non-adherent patients, regardless of the CNI used or analytical approach. In longitudinal analysis, IPV was similar (31.3% [25.5-38.2] vs. 31.6% [23.6-38.9], p = 0.68), as was the proportion of patients with high IPV (55.5% vs. 51.5%, p = 0.5). At 3 years, high IPV was not significantly associated with rejection (HR 1.02 [0.67-1.55], p = 0.93). Self-reported adherence was not associated with CNI C₀ levels, IPV, or underexposure. CNI C₀ variability alone cannot reliably detect non-adherence and should not be interpreted as a standalone adherence marker. Multimodal strategies combining pharmacokinetics with validated self-report tools are needed to evaluate adherence.