Cardiovascular diseases remain the leading cause of death and a major contributor to medical morbidity and disability in the United States. Despite improved prognosis associated with the identification of clinical risk factors and the development of sophisticated interventions, there continues to be a 2.40-fold (1.80-3.20) mortality risk associated with clinical depression and a 1.76-fold (1.30-2.40) mortality risk associated with depression symptoms (c.f., 1
) for patients after acute coronary syndrome (ACS), which is defined to include acute myocardial infarction (AMI) and unstable angina (UAP). In an effort to reduce this risk, two large randomized clinical trials, M-HART (2
) and ENRICHD (3
), were initiated in the 1990’s. The success of prior trials that targeted psychosocial factors that have either a direct or indirect relationship to depression, such as anger, hostility, type A behavior pattern, and general distress (c.f., 4
), provided support for the overall notion of intervening on an emotion related factor to improve cardiovascular outcomes. In addition, a rich literature demonstrated the efficacy for both cognitive behavioral and pharmacologic therapies in the treatment of depression (c.f., 7
). Despite the logic of treating depression to improve cardiac event-free survival - and of the approaches chosen for such treatment - neither trial was successful at improving these outcomes, though statistically modest improvements were realized in the level of depressive symptoms and/or depression status for patients who received active treatment (e.g., for ENRICHD, a mean decrease in BDI score of 49% vs. 33%, intervention vs. usual care, p<.001; 3).
Several reasons have been discussed for the failure of ENRICHD and M-HART to affect medical outcomes (c.f., 3
). Among these is that while statistically significant, the modest improvements in depression for patients randomized to intervention may not have been sufficient to affect the pathophysiological pathway(s) (e.g., inflammation, medication adherence) tying depression to post-ACS prognosis. At the same time, patients randomized to usual cardiologic care often demonstrated early and substantial improvement in depression, suggesting that many of the patients randomized into the trial were experiencing a short-lived adjustment reaction rather than real depression, and hence did not carry the associated risk to survival that had been identified in observational studies (1
). Under such a scenario, the power to detect an effect would have been greatly reduced. This could have been a particular issue for the ENRICHD trial, which enrolled patients within days after their index ACS event.
Another reason discussed for the modest depression treatment effects and hence, lack of effect on event-free survival concerns the degree to which patients in these trials received a “full dose” of therapy. Analyses from ENRICHD found that because of missed appointments and treatment dropout, only 54% of patients randomized to intervention received the complete course of protocol defined individual treatment (3
), while only 31% received the group therapy that was an augmentation to this primary, individual therapy (9
). The underlying issue may have been the satisfaction with, and acceptability of the Trial intervention for those randomized to receive treatment. Psychiatric depression trials typically employ reactive recruitment methods, enrolling those who are highly motivated treatment seekers. In contrast, ENRICHD and M-HART employed proactive recruitment methods, approaching all patients in hospital with an admission diagnosis of AMI, screening them to ascertain whether they met depression criteria, and then enrolling them if they did. Hence, because the patients did not “self-present” for depression treatment, many randomized to treatment may have been less than fully motivated to engage in therapy, particularly the approach offered by the Trial.
Concerns regarding the cardiovascular safety of the psychosocial treatments used by M-HART and ENRICHD have also been raised. For example, there was a non-significant but elevated trend to harm (death, p=.06) in the post-ACS female sub-sample randomized to treatment in the M-HART study (11
). An exploratory, planned subgroup analysis of the ENRICHD cohort revealed a similar, non-significant but elevated trend to harm (death, p=.12) in post-ACS women randomized to the intervention condition of ENRICHD (3
). In both cases excess deaths were primarily cardiovascular and none were from suicide. Speculation for these findings - while post-hoc - includes that the women enrolled and randomized into the treatment condition(s) found the type of treatment provided neither welcome nor particularly helpful to them. As a result, they may have experienced increased distress rather than decreased distress. While this is a post hoc suggestion, it is supported by analyses from M-HART demonstrating a lower event rate among women whose distress symptoms abated as a function of treatment, and a higher event rate among those whose symptoms either increased or did not decrease (12
Thus, although depression as a clinical syndrome and symptom constellation are each detrimental to event-free survival for post-ACS patients, key questions remain regarding how best to safely and effectively treat depression and its symptoms in post-ACS patients and whether and how this treatment improves event-free survival. These key questions, obstacles to the mounting of any future clinical trials of depression treatment in ACS patients for which event-free survival is the primary outcome, concern: 1) the timing of treatment delivery to capture patients whose depression symptoms are neither transient nor merely reflective of adjustment issues, but rather indicative of depression related risk; 2) the selection of a treatment that is acceptable to post-ACS patients with depressive symptoms, and efficacious for these symptoms; and, 3) the safety of the delivered intervention with regard to medical morbidity and mortality. In addition, the effect of the treatment on hypothesized pathways tying depression to post-ACS survival – such as inflammation and medication adherence - is an important, intermediate issue, as these pathways can serve as surrogate endpoints in smaller, exploratory trials that are not powered for event-free survival. For example, since depression is associated with increased levels of circulating inflammatory markers and ACS is a disease of inflammation, an intervention that reduces both depression and inflammation would be seen as more promising to test in a large trial than an intervention that only decreases depression. Similarly, since depression is associated with reduced medical adherence and reduced medical adherence is in turn associated with reduced post-ACS event-free survival, an intervention that decreases depression and improves medication adherence would be seen as more promising to test in a large trial.
By its design and clinical approach, the Project COPES randomized clinical trial is designed to specifically address these questions. The Trial utilizes a 3-month observational “run-in” phase to insure that the depressive symptoms of immediate post-ACS patients are not merely reflective of transient adjustment issues and persist at 3-months. In addition, the treatment is based on a 6-month patient preference, stepped-care approach that has previously been found acceptable to medical patients with depression (10
). Adverse events are monitored to assess cardiovascular safety, and depressive symptoms are assessed prior to, and immediately after the 6-month treatment window to provide estimates of the effect size on depression symptoms that the intervention yields. Information concerning effect size will then contribute to power calculations for future, larger trials should this trial be demonstrated safe for, and acceptable to the patient population. Lastly, questions regarding the ability of this intervention to effect underlying pathophysiology are explored by monitoring of two key pathways – inflammation and medication adherence - hypothesized as tying depression to post-ACS prognosis.