Nonadherence by Serum Drug Analyses in Resistant Hypertension: 7-Year Follow-Up of Patients Considered Adherent by Directly Observed Therapy

Abstract

Background Measurement of serum concentrations of drugs is a novelty found useful in detecting poor drug adherence in patients taking ≥2 antihypertensive agents. Regarding patients with treatment-resistant hypertension, we previously based our assessment on directly observed therapy. The present study aimed to investigate whether serum drug measurements in patients with resistant hypertension offer additional information regarding drug adherence, beyond that of initial assessment with directly observed therapy. Methods and Results Nineteen patients assumed to have true treatment-resistant hypertension and adherence to antihypertensive drugs based on directly observed therapy were investigated repeatedly through 7 years. Serum concentrations of antihypertensive drugs were measured by ultra-high-performance liquid chromatography-tandem mass spectrometry from blood samples taken at baseline, 6-month, 3-year, and 7-year visits. Cytochrome P450 polymorphisms, self-reported adherence and beliefs about medicine were performed as supplement investigations. Seven patients (37%) were redefined as nonadherent based on their serum concentrations during follow-up. All patients reported high adherence to medications. Nonadherent patients expressed lower necessity and higher concerns regarding intake of antihypertensive medication (P=0.003). Cytochrome P450 polymorphisms affecting metabolism of antihypertensive drugs were found in 16 patients (84%), 21% were poor metabolizers, and none were ultra-rapid metabolizers. Six of 7 patients redefined as nonadherent had cytochrome P450 polymorphisms, however, not explaining the low serum drug concentrations measured in these patients. Conclusions Our data suggest that repeated measurements of serum concentrations of antihypertensive drugs revealed nonadherence in one-third of patients previously evaluated as adherent and treatment resistant by directly observed therapy, thereby improving the accuracy of adherence evaluation. Registration URL: https://www.clinicaltrials.gov; unique identifier: NCT01673516.

Keywords: antihypertensive drugs; blood pressure; directly observed therapy; nonadherence; resistant hypertension; serum drug concentrations.

Figure 1. Flowchart showing number of patients and relevant procedures at the study visits.

ABPM indicates ambulatory blood pressure measurement; BMQ, Beliefs About Medicines Questionnaire; BP, blood pressure; CYP, cytochrome P450; DOT, directly observed treatment; HT, hypertension; and RDN, renal denervation.

Figure 2. Mean daytime ambulatory blood pressure (ABPM) in the 7 patients who revealed nonadherence during follow‐up compared with the adherent patients (n=12).

SDs are marked with bars (±1 SD). The y axis is truncated. Generalized linear mixed models showed that daytime systolic ABPMs were significantly higher in the nonadherent patients (P=0.042) and diastolic ABPMs did not statistically differ between the groups (P=0.11). Fractions of nonadherent patients at the different visits on the x axis. DBP indicates diastolic blood pressure; and SBP, systolic blood pressure.

Figure 3. Adherence over time in 7 patients defined as nonadherent on the basis of serum drug concentrations at ≥1 visits during 7‐year follow‐up.

The y axis shows the mean systolic daytime ambulatory blood pressure for each visit (0=baseline, 6m=6‐month visit, 3y=3‐year visit, 7y=7‐year visit). Light gray, dark gray, and black columns indicate the adherence status. *Refused to ingest any medications, and serum drug measurements were not performed. ^†Admitted nonadherence to 1 of 2 antihypertensive drugs, but a method to analyze this particular drug was not available. ^‡Admitted to not having ingested antihypertensive medication during the past 48 hours, consistent with serum drug measurements (drugs with rapid elimination were below the detection limits). ABPM indicates ambulatory blood pressure measurement; RDN, renal denervation; and SBP, systolic blood pressure.

Figure 4. Difference in Beliefs About Medicines Questionnaire scores between nonadherent and adherent patients.

Results are reported as boxplots of all 4 dimensions at baseline and boxplot at baseline and change over time for the necessity–concern difference.

Figure 5. Correlation between serum concentration of metoprolol and heart rate from another ongoing study.

Patients (n=104) using metoprolol prolonged release are classified into different cytochrome P450 2D6 phenotypes. Their mean daytime ambulatory heart rate (beats/min) are given on the y axis, and s‐[metoprolol] corrected for daily dose of metoprolol (nmol/L per mg) on the x axis. Nonadherent patients on the basis of their s‐[metoprolol] (n=4) were all normal metabolizers.