By definition, medical researchers test their hypotheses because they are uncertain whether a procedure or medication will actually help patients combat illness. So when a promising technique fails to meet expectations, doctors then know that a particular treatment may not deliver, and that it is time to test other methods. But when an eight-year, multicenter study that cost nearly $30 million failed to show that cognitive-behavioral therapy (CBT) and enhanced social support following a heart attack had an effect on survival or the likelihood of recurrent cardiac events, it left investigators puzzling over what went wrong and what to do next.
Even the study's acronym, ENRICHD, which stands for Enhancing Recovery in Coronary Heart Disease patients, evoked the confidence teams of researchers brought to a project that ultimately enrolled nearly 2,500 subjects within 28 days of a heart attack. Those patients, recruited from 1996 to 1999, came from 73 hospitals affiliated with eight clinical centers across the country. Once evaluated with depression, low social support or both, patients were randomized to either receive usual care or a CBT-based psychosocial intervention. The researchers hypothesized that members of the treatment group would live longer and suffer fewer future cardiac events.
They didn't. In June, researchers published the results of the ENRICHD randomized trial in JAMA (289:3106-3116), concluding that the therapy-based intervention had no effect on mortality and did not reduce cardiac events during an average 29-month follow-up period. That result, however, was tempered by a finding that those patients who took a selective serotonin reuptake inhibitor antidepressant did have a lower risk of dying or non-fatal myocardial infarction. The antidepressant use, which was not randomized and was provided to patients who needed them in both groups, has now raised new questions about the way heart attack victims should be treated in a system that already shortchanges their mental health care needs.
In the aftermath of the study's negative results, researchers are shrugging off their disappointment and attempting to understand why patients who received treatment for depression failed to respond with longer lives or fewer heart attacks. With prior studies establishing a link between depression and low perceived social support (LPSS) and cardiac mortality and morbidity, it seemed likely that providing a mental health intervention would yield improvement. The treatment did help alleviate the group's depression and lack of social supports, raising the question of whether quality-of-life enhancement should have been the appropriate endpoint of the study.
"The questions of survival and cardiac events are still important questions," said Susan Czajkowski, Ph.D., ENRICHD's project officer and a research psychologist with the National Heart, Lung and Blood Institute (NHLBI), the agency that funded the study. She told Psychiatric Times, "One way of looking at ENRICHD is that it was a specific type of intervention for a specific group patients and a very specific time point in the course of disease, which was immediately after the myocardial infarction ... Therefore, what we are doing right now is taking a look and thinking about the parameters of that and whether there might be better or longer interventions at different time points in cardiovascular disease, and whether there might be different kinds of patients that might benefit."
That scientific soul-searching could be important because of the data that indicated treatment with an SSRI antidepressant could generate the survival and anti-recurrence effect sought by investigators. For the time being, the finding has raised more questions about what care heart attack patients should receive, since the use of antidepressants may have been a confounding factor that yielded the negative results in the ENRICHD study.
The antidepressant medication became a confounding factor by possibly masking the benefits of therapy, said Ranga Krishnan, M.D., chair of psychiatry at Duke University School of Medicine, one of the clinical centers involved in the study. Meanwhile, because administration of antidepressants was not randomized, the findings relating to the benefits of SSRI use could not be confirmed with a high degree of confidence.
"It is fair to conclude that [talk therapy] will treat depression but not the problem," Krishnan told PT. "At this point in time, probably the most positive way of putting is to say you can't tell one way or another whether it will affect survival and there's no data to support that it will affect survival." Since the CBT-based intervention helped treat the depression, however, it should not be rejected as a possible therapeutic intervention, particularly because there are individuals who will not take medication, he added.
The larger issue still remains the lack of integration of mental health care as part of the post-heart attack rehabilitation process, though Krishnan doubts that this study will crimp efforts to provide improved access to treatment. "They're not reimbursing for it anyway," Krishnan said, because once a diagnosis of depression is made "carved" out, benefits levels often do not measure up to the ones provided for the physical elements of cardiac care. So part of the problem is not only the uncertainty regarding the method of treatment, but the systemic barriers to care that have traditionally plagued mental health.
That issue plagues cardiologists who must consider the mental health component of their patients' care. The systemic issues combine with the pragmatic elements of acute cardiac care to muddy the waters.
"Clinicians most often focus on the strictly medical during the acute phase of the heart attack. We have very little time and so we focus on those things that can impact acutely on patient morbidity and mortality during the acute event," Allan S. Jaffe, M.D., an ENRICHD study co-chair, cardiologist and professor of medicine at the Mayo Clinic, told PT. With cardiac patients now only spending between four to six days in the hospital, a physician has a limited time to interact and there is a tendency to give priority to acute rather than chronic problems. Cardiologists also recognize that secondary prevention problems are important and have implemented programs to address them for their patients. "But we need more definitive research findings showing mortality and morbidity benefits to mental health care. Even though I am an advocate, without hard data, it is hard to say that treatment of depression belongs in secondary prevention," Jaffe said. "We haven't shown that therapy changes outcomes. So for now I think the treatment of depression is indicated predominantly for its psychological benefit which can be substantial."
Even though Jaffe considers himself an advocate for integrating mental health care into cardiac rehabilitation programs, he also said that inadequate resources, low-level reimbursements and the unavailability of mental health care programs stand in the way. In addition, he conceded, "From the hard medical standpoint, therapy is not necessarily associated with improvements in hard medical events."
The statistics for heart patients under these circumstances are grim. According to the NHLBI, what justified the study were data showing that heart disease patients with depression or social isolation faced a risk of death three to four times higher. Despite the increased exposure faced by the 25% who fall into these categories, only an estimated 25% received treatment.
For Thomas N. Wise, M.D., the editor-in-chief of the consultation-liaison psychiatry journal Psychosomatics and professor of psychiatry at Johns Hopkins School of Medicine, the next step is to determine whether SSRI antidepressant medication can be proven to have the salutary effects sought but not attained in the ENRICHD study.
"The big question is then would we ever consider an SSRI as a prophylaxsis for heart disease," Wise told PT. "We're certainly not there yet, but you wonder whether in the future we might be. That's the research question now. Would it be useful to study the prophylactic effects of an SSRI with high affinity for serotonin reuptake inhibition?"
For Wise, the ENRICHD study finding that SSRI antidepressants did have an effect on lessening mortality and recurrence after a myocardial infarction, along with other research, suggests that there's not only a psychological effect, but also a possible physiologic influence. How that translates into practice guidelines for patients is still not clear, however.
"To date, all we know is that CBT in that [ENRICHD] sample did not improve mortality, except if you're on sertraline [Zoloft] [or other SSRI] and you were significantly depressed," Wise said.
"It's probably because the SSRI does improve clotting mechanisms," Wise added. "On the other hand, the physiology of depression affects the heart as well. ... When you're depressed, your heart rate isn't as elastic, and that can be bad." As a result, combined approaches that include therapy and medication may be the way to treat depression while at the same time reducing death and cardiac events.
The premise also makes sense to George I. Viamontes, M.D., Ph.D., a psychiatrist and cell biologist who is the regional medical director of United Behavioral Health in St. Louis. He said that the ENRICHD study results should not discourage health care systems from incorporating proper mental health care into their cardiac rehabilitation programs although he, too, acknowledged that there is still resistance to its inclusion. But whether psychotherapy alone can make the difference is a function of a number of psychological and physiological forces.
"It's not that I wouldn't support adjunctive psychotherapy for people who may benefit from that in some other way, say to improve quality of life, but I would certainly urge people to be on a medication," Viamontes told PT. With the research indicating that the risk of mortality and morbidity is highest within the first few months after a heart attack, it is possible that medication works faster than CBT.
It's a question of approaching the problem from a top-down or bottom-up approach, Viamontes said. In the case of CBT, it may show behavioral changes, but it will take longer for it to generate a physiologic response. In the case of SSRI antidepressants, which he agreed may have an effect on platelets and clotting, the drugs begin to work immediately on the cellular level and may have a positive impact even before noticeable behavioral changes occur in the depressed patient. Ultimately, what the study may be showing is that talk therapy takes too long to create the types of changes at the molecular level that will reduce mortality and morbidity in heart attack patients whose risk is highest after an event.
Despite what appears to be a setback, NHLBI's Czajkowski said that this is just the beginning. "What will happen now is that all the investigators are madly analyzing data in various aspects of the trial. Many publications are in the works that will be looking specifically at the depression outcomes, at the social support outcomes, at the subgroups of women and minorities, the pharmacology outcomes and those will hopefully be published in the coming year."