Adherence to Antihypertensive Drugs: Insights from the SPYRAL HTN Trials and Implications for Hypertension Trial Design.

Background: Variable adherence to prescribed antihypertensive drugs may have confounded prior renal denervation trials. Adherence to prescribed drugs was measured in the SPYRAL HTN-OFF MED and ON MED trials of renal denervation for uncontrolled hypertension when no drugs, and when up to 3 anti-hypertensive drug classes were prescribed, respectively. Methods: Patients were enrolled with uncontrolled hypertension defined as office systolic blood pressure (SBP) ≥150 and <180 mmHg with diastolic blood pressure >90 mmHg and 24-hour mean SBP ≥140 and <170 mmHg. Patients were randomized 1:1 to renal denervation or sham control. OFF MED trial patients were either drug naïve or had antihypertensive drugs discontinued at enrollment. ON MED trial patients had to maintain baseline prescribed antihypertensive medications. Both groups were informed on enrollment that urine and blood samples would be obtained for drug adherence testing. This testing utilized tandem HPLC and mass spectrometry. Following enrollment, both groups were followed for 4-8 weeks prior to randomization. Results: In OFF MED, antihypertensive drug or drug metabolites were detected in 10% of patients just prior to randomization, 11.2% at 3 months following randomization, and in 15% of patients at either time point, despite protocol-required absence of all antihypertensive medications following enrollment. In ON MED, 62.5% were fully adherent to prescribed medications (i.e., all prescribed medications identified) at baseline, 55% at 3 months and 62.5% at 6 months. Drug adherence was not consistent: 24% of baseline adherent patients were non-adherent at 3 months, and 16% of patients adherent at 3 months were no longer adherent at 6 months. Conclusions: Despite being aware that drug adherence testing was being conducted, almost half did not take all medication when they were prescribed and a minority took anti-hypertensive medications when instructed to abstain. Furthermore, non-adherence was inconsistent, with individual behavior at one time point not predictive of drug adherence at other time points. Attention to protocol adherence in device-based clinical hypertension trials is critical to determine non-drug treatment effects.