• Diabetes Mellitus Clinical Trials
• Lipid Disorders

Patients who are at high risk for cardiovascular (CV) events face a vicious cycle in which treatment nonadherence increases event rates while events themselves reinforce nonadherence, according to new findings from the Ongoing Telmisartan Alone and in Combination With Ramipril Global Endpoint Trial (ONTARGET).

ONTARGET included 25,620 patients with coronary heart disease or diabetes plus two additional risk factors for vascular events. In the main study analysis, telmisartan was noninferior to ramipril for the primary composite endpoint of CV death, myocardial infarction (MI), stroke, or heart failure (HF) hospitalization (p=0.0045) and for the secondary endpoint of CV death, MI, or stroke (p=0.001) [Yusuf S et al. N Engl J Med 2008].

The current ONTARGET analysis focused on risk factors and outcomes that are associated with nonadherence, the latter defined as premature permanent discontinuation of all study medications. Michael Böhm, MD, University of the Saarland, Saarbrücken, Germany, presented findings from the study.

Throughout ONTARGET, there was a continuous increase in the proportion of patients who permanently discontinued treatment. After 72 months of follow-up, 4629 patients (18.6%) had discontinued treatment. Increasing age, female gender, black race, physical inactivity, smoking status, diabetes, and depression were significantly associated with non-adherence.

The strongest factor that was associated with premature permanent drug cessation was the occurrence of a non-fatal event during trial follow-up. Premature discontinuation significantly increased after nonfatal MI (HR, 2.24; 95% CI, 1.95 to 2.57; p<0.0001), stroke (HR, 2.28; 95% CI, 1.98 to 2.63; p<0.0001), and HF hospitalization (HR, 2.85; 95% CI, 2.49 to 3.27; p<0.0001). Other nonfatal events, such as revascularization, angina, end-stage renal disease, malignancy, and new-onset HF, diabetes, and atrial fibrillation also increased the risk of treatment discontinuation (p<0.0001 for each event). The risk of premature permanent cessation also increased with early events and with an increasing number of total events.

Over time, the risk of CV events increased sharply and soon after premature cessation of study medication. Compared with patients who remained on treatment, patients who stopped were 29% more likely to reach the primary endpoint of CV death, MI, stroke, or HF hospitalization (HR, 1.298; 95% CI, 1.181 to 1.427; p<0.0001) and 38% more likely to reach the secondary endpoint of CV death, MI, or stroke (HR, 1.385; 95% CI, 1.255 to 1.528; p<0.0001). Premature permanent cessation was also strongly associated with the individual endpoints of CV death (HR, 2.050; 95% CI, 1.824 to 2.303; p<0.0001) and HF hospitalization (HR, 1.464; 95% 1.228 to 1.745; p<0.0001).

The authors conclude that the ONTARGET trial revealed a deleterious “nonadherence-event” cycle, in which stopping study medications increased CV events, leading to increased morbidity, less trust in therapy, and further nonadherence. To break this cycle, interventions that promote treatment adherence are needed for patients who are at high risk of nonadherence, and for all patients immediately after a CV event, Prof. Böhm concluded.

Editors Note: While these observations raise an important hypothesis, it is important to note that the strongest predictors of premature cessation were the occurrences of clinical events during the trial. Patients who experience events (eg non-fatal MI) are at a heightened risk of a recurrent event and therefore it is unclear that treatment cessation led to recurrent events or was just associated with sicker patients. In addition, it is important to recognize that patients may stop therapy prematurely for many reasons in a trial, including related adverse events and acute illnesses for which treatment cessation may be medically indicated (eg, cessation of antihypertensive patients who are in shock). Thus while drug discontinuation may be followed shortly by a clinical event (eg, MI), therefore suggesting a temporal association, establishing a true causal relationship may be difficult. Additional analyses that show consistency in patients with no preceding hospitalization or CV event as well a description of reasons for treatment cessation would strengthen the overall findings.

• © 2011 MD Conference Express