Background/Purpose: Non-adherence to medication has been shown to impact mortality, morbidity, and health care utilization in SLE and ranges from 40-80% depending on the methods used to assess compliance. Hydroxychloroquine (HCQ) is a key component of SLE therapy. HCQ levels are a sensitive method to assess medication adherence in lupus. Our study assessed the feasibility of HCQ measurement and its impact on SLE activity in a multiethnic, predominantly Hispanic, population.

Methods : SLE patients from the Columbia Lupus Center meeting ACR/SLICC criteria, treated with HCQ for >6 months, reporting adherence to HCQ were included.  HCQ testing was done by high performance liquid chromatography on whole blood by the Exagen Laboratory.  Non-adherence was defined as a HCQ level of < 500ng/ml. The association between HCQ and disease activity measured by SLEDAI-2K was evaluated in a multivariate model that included all variables showing an association at p<0.2 in univariate analysis (demographics, SLE variables, age, sex, race/ethnicity). Statistical significance was considered p£0.05

Results : 108 patients, average age 38 (range 19-66) and 91% female, 62% Hispanic, and average SLEDAI of 4.3 (range 0-20) were included. 41% of patients had HCQ<500 demonstrating non-adherence; 19% had undetectable levels. Of the patient demographics and SLE characteristics, Hispanic ethnicity, being primarily a Spanish speaker, having higher eGFR, organ involvement and a diagnosis of depression had a trend towards significance in univariate analysis (Table 1). There was a significant association between having had CNS involvement and higher HCQ levels. Higher SLEDAI was significantly associated with lower HCQ levels (Table 2) (p<0.001). After adjusting for age, sex, ethnicity, education, depression, smoking, organ involvement, current steroid use, and immunosuppression, HCQ levels were significantly associated with SLE disease activity measured by SLEDAI (p<0.004). Average adjusted SLEDAI in non-adherent patients was 5.8 (IQR 4.8, 6.8), and 3.3 in adherent patients (IQR 2.5, 4.2).

Conclusion : These data suggest that HCQ levels <500 ng/ml are significantly correlated with higher disease activity scores and explain 32% percent of the SLEDAI variability. Testing for HCQ levels is an easy and reliable way to evaluate medication adherence in SLE. Levels <500ng/ml should be discussed with the patients and the reasons for non-adherence should be further explored. A diagnosis of depression had a strong trend towards associating with non-compliance emphasizing the importance of screening and treating this co-morbidity.