From treatment resistance to sequential treatments of depression

McIntyre et al ¹ provide an excellent overview of “treatment‐resistant depression” (TRD) and of the ways future research can contribute to a better knowledge on how to handle the many patients with depression who do not respond to treatment. However, I argue here that the notion of TRD is based on a misconception of the effects of treatments in depression, and that it is much better to focus research on sequential treatments of depression in general. I also argue that the literature on TRD is biased towards pharmacological treatments and ignores several of the best available therapeutic interventions.

One major problem with TRD (defined as inadequate response to a minimum of two antidepressants despite adequacy of the treatment trial and adherence to treatment) is that it is very much based on a misconception of the effects of treatments of depression. On the one hand, many patients with depression recover without treatment; on the other, response rates of treatments are modest. For example, we found that 41% of patients receiving psychotherapy responded (50% symptom reduction), while 31% responded to placebo and 17% to usual care ² . If we assume, for the sake of argument, that a next treatment will have the same effects as the previous one, we would need on average 2.5 treatments per patient in order to realize response in 100% of patients. Many of them would need only one or two treatments (65%), but the other 35% would need more treatments (up to 10). In reality, the number of treatments needed to realize response in all patients is even larger, because the number of patients responding to a treatment is lower when they have received a treatment before ³ .

Remission rates following the first round of psychotherapy (26% after treatment with psychotherapy; 17% in placebo conditions; 12% in usual care) are even lower than response rates, and the total number of treatments needed to realize remission in all patients is even higher than the average 2.5 treatments that are needed for response in all patients (>3) ² . And many patients who respond or remit would have actually responded or remitted with pill placebo as well. This all means that TRD is simply the logical result of the limited effects of treatments. The concept of TRD suggests that there is a threshold that patients should pass (two unsuccessful treatments), while in fact there is no threshold. There is only a limited number of patients who will respond to the next treatment, just as only a limited number responds to the first treatment.

There are many pharmacological, psychological and other treatments of depression and they all have comparable, but limited effects ² , ⁴ . At the same time, hardly anything is known about who benefits from which treatment. Very little is known about the first treatment that should be offered to a patient. Any treatment is as good as another. That means that we very much need research on who benefits from which treatment. But we also need research on sequential treatments. If a patient does not respond to one treatment, what treatment should be offered next, and which one if the second treatment also does not work, and the third and the fourth? From this perspective, research on TRD is very useful, because that is exactly the focus of this research: what should we do if patients do not respond to several treatments? So, although the concept of TRD is based on a misunderstanding of the effects of treatments, the research on interventions for TRD is very much needed.

Unfortunately, there is another major problem with research on TRD: the almost complete absence of psychological treatments. In one systematic overview, a total of 148 different definitions of TRD were collected from the literature ⁵ . All definitions included at least one failed treatment with antidepressants, but only six definitions (4%) included one failed treatment with psychotherapy. This is remarkable, because there is not only much evidence that psychotherapy overall is as effective as antidepressants in the short term ⁶ , but also that psychotherapy is more effective in the longer term ⁶ , ⁷ , and that combined treatment is more effective than either pharmacotherapy or psychotherapy, in the short and longer term ⁷ . Also, almost all treatment guidelines for depression not only recommend antidepressants but also psychotherapy as first line treatment. This suggests that almost all people who meet one of the current definitions of TRD have not received the best available treatments. Fortunately, the review by McIntyre et al ¹ tries to repair this omission in the literature. But it still means that most of the other literature on TRD is flawed and biased towards pharmacological treatments of depression.

There is also some evidence that pharmacotherapy and psychotherapy work independently from each other, and that their effects are additive, without interfering with each other ⁸ . At the same time, there is some evidence that prior use of antidepressants results in lower response rates when another antidepressant is used ⁴ . This makes it even less understandable why previous definitions of TRD usually do not include psychotherapies. It further illustrates the biased nature of this research area, and that many patients with TRD just received suboptimal treatments before being defined as having TRD.

Taken together, one could argue that the concept of TRD should be abandoned, because it is based on a misconception of the effects of treatments of depression, and we should move towards an agenda for research on sequential treatments. The current research on TRD fits very well into this agenda, but also has serious limitations, especially the focus on antidepressants and the exclusion of psychological and combined treatments.

Such an agenda should also include other research questions. For example, there is very little research on sequential psychological treatments of depression. Although there are now almost 1,000 randomized controlled trials on these treatments, hardly anything is known about which treatment should be used when a patient does not respond. The same is true for combined treatments. Although it is clear from a considerable number of trials that combined treatment is more effective than either psychotherapy or pharmacotherapy alone ⁶ , ⁷ , very little is known about what to do when a patient does not respond to that treatment. Should we change the antidepressant, the psychotherapy, or both? We simply do not know, while these are the questions that need to be answered if we want to help as many patients as possible.

It is time that we move away from the concept of TRD and focus on research on sequential treatments, because that is what patients need most.