Cognitive-behavioral therapy for treatment-resistant depression in adults and adolescents : a systematic review

Objetivos: Realizar uma revisão sistemática sobre o uso da terapia cognitivo-comportamental (TCC) e sua eficácia no tratamento da depressão resistente ao tratamento em adultos e adolescentes. Métodos: Realizamos uma revisão sistemática utilizando os critérios do Prisma Guidelines, nos seguintes bancos de dados: PubMed, SciELO, Psychiatry Online, Scopus, PsycArticles, Science Direct e Journal of Medical Case Reports. Estudos controlados randomizados, estudos abertos e relatos de casos foram incluídos neste estudo. Resultados: A pesquisa retornou um total de 1.580 artigos, publicados de 1985 até 2017. Após aplicarmos os critérios de inclusão, 17 artigos foram selecionados, seus textos completos foram lidos e 8 foram incluídos nesta revisão. Do total, quatro eram estudos controlados randomizados com adultos, tendo um incluído um período de seguimento pós-estudo; dois eram estudos controlados randomizados com adolescentes, tendo um apresentado dados de seguimento; um era um estudo aberto; e o último era um relato de caso. Os estudos apresentaram boa qualidade e evidências robustas sobre o tópico abordado. Conclusões: A combinação de TCC com tratamento medicamentoso para pacientes resistentes ao tratamento mostra uma diminuição dos sintomas depressivos. A TCC pode ser um tipo eficaz de terapia para adultos e adolescentes com depressão resistente ao tratamento. Descritores: Terapia cognitivo-comportamental, transtorno depressivo maior, transtorno depressivo resistente a tratamento. Abstract


Introduction
Treatment-resistant depression (TRD) is diagnosed when patients who suffer from major depressive disorder (MDD) are unable to achieve an adequate therapeutic response despite using one or more antidepressants. 1 TRD is a relatively common occurrence in clinical practice with adult populations, with up to 50 to 60% of patients not achieving adequate response following antidepressant treatment. 2 Moreover, only about 60% of adolescents with depression will show an adequate clinical response to antidepressant treatment. 3 In addition, 40% of adolescents will not respond to treatment and thus suffer from TRD. TRD is associated with greater severity than MDD and is associated with risk of suicide. A study by Bergfeld et al. 4 showed that 30% of patients with TRD attempt suicide at least once during their lifetime.
The usual treatment for moderate to severe depression (also called treatment as usual [TAU]) is based on use of medication such as selective serotonin reuptake inhibitor (SSRI) antidepressants. Although this treatment has shown efficacy in some cases, many will not respond to treatment even after several trials. 5 Therefore, treating TRD is a considerable challenge. Nakagawa et al. 6 conducted a randomized controlled trial with the aim of investigating the effectiveness of CBT in TRD patients. Their study included outpatients randomly assigned to CBT combined with TAU or to TAU alone and the primary outcome was alleviation of depressive symptoms. They found that supplementary CBT alleviated depressive symptoms at 16 weeks and that the treatment effect was maintained for at least 12 months.
Other authors have tested the efficacy of this strategy in an adolescent population. Brent et al. 3 conducted a randomized controlled study (TORDIA) with the objective of evaluating the best treatment option for adolescents aged 12 to 18 years (either switching medications or adding CBT to the treatment). Participants were treated for 12 weeks (TAU alone and TAU+CBT) and the results showed that for adolescents with depression that did not respond to TAU, the combination of CBT with switching to another medication resulted in a higher rate of clinical response than a medication switch alone.
A recent systematic review and meta-analysis by Li et al. 7 assessed the effectiveness of CBT and other related interventions for treatment of patients with TRD. The authors included randomized controlled trials performed with adults over 18 years old who were suffering from TRD and were allocated to CBT and/or other forms of interventions. Patients were assessed using validated designed to assess depression, such as the 17-item Hamilton Rating Scale for Depression (HRSD-17), the 21item Hamilton Rating Scale for Depression (HAMD-21), the Beck Depression Inventory-II (BDI-II) or the Patient Health Questionnaire-9 (PHQ-9). The authors selected six out of 526 studies identified: two trials that used CBT, two trials that applied mindfulness-based cognitive therapy (MBCT), one that adapted rumination-focused cognitive-behavioral therapy (RFCBT) and one that used smartphone CBT. They reported that these interventions were effective for treatment of TRD symptoms and that the effects were maintained at 6-month follow-up.
However, one of the most significant criticisms of this review is that inclusion of several different types of CBT and CBT-based interventions could have impacted the findings. There is a need to look at more homogeneous intervention groups such as CBT alone or MBCT alone.
The purpose of this review therefore differs from the one described above because our aim is to conduct a systematic review to examine the efficacy of CBT alone, with no other CBT based interventions, for treatment and alleviation of depressive symptoms in TRD.
Moreover, to examine whether CBT reduces depressive symptoms in TRD compared to TAU alone, our study included randomized controlled trials, open trials and case reports and is not restricted to adult populations, but also includes studies with adolescents (age < 18), since few studies have addressed this subset.

Methods
We followed the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) guidelines throughout this systematic review. 8

Eligibility criteria
Studies were selected and included in the present review according to the following criteria: 1) studies assessing the efficacy of CBT in TRD, 2) randomized controlled trials (RCTs), long-term follow-up studies derived from RCTs, open trials and case studies are included, 3) inclusion of participants that meet criteria for TRD defined in Trevino et al., 1 4) inclusion of studies that only use face-to-face CBT as a treatment model, 5) inclusion of both adolescent (12-18 years old) and adult populations, and 6) inclusion of studies that used validated instruments to assess depressive symptoms.

Information source and search strategy
Searches were run on PubMed, SciELO, Psychiatry Online, Scopus, PsycArticles, Science Direct, and the Journal of Medical Case Reports using combinations of the Medical Subject Headings (MeSH) terms "cognitive behavioral therapy," "cognitive behavior therapy," "cognitive psychotherapy", "cognitive therapy", "cognition therapy," "treatment-resistant depression", "refractory depression" and "therapy resistant depression". Additional records were identified from other sources by searching references lists of the studies found in databases. The search was last run on April 8, 2019. No restriction was placed on language of publication.

Study selection
The lead author of this study independently screened the articles identified for appropriateness for inclusion.
Non-randomized controlled trials, literature review studies, systematic reviews and studies that used types of therapy other than CBT were excluded from the systematic review.

Data collection process and data items
Data collection was conducted independently by the lead author of this systematic review. The titles and abstracts identified by the initial search were screened to determine their relevance to the review. Articles that did not meet the inclusion criteria were excluded at this stage and then the full texts of potentially relevant studies were examined. The lead author of this review selected articles independently.

Risk of bias in individual studies
Only the randomized controlled trials selected for the review study were assessed for quality and risk of bias. The Randomized Controlled Trial Psychotherapy Quality Rating Scale (RCT-PQRS) was used. 9

Data synthesis
This systematic review is presented as a narrative synthesis measuring the efficacy of adding CBT treatment for adults and adolescents who suffer from TRD.

Study selection
A total of 1,576 articles were found by database searches. Additional records (n = 4) were identified by searching the references lists of the studies found.
Hence, a total of 1,580 articles published from 1985 to 2017 were identified for the systematic review. After removing duplicate articles (n = 19) and excluding studies on the basis of title (n = 1535), 26 were retained for further consideration. Nine of these were excluded after reading the abstracts. Nine of the 17 remaining articles assessed for eligibility were excluded after reading the full texts, because they did not meet inclusion criteria and were non-randomized controlled studies. Finally, 8 studies were included in the systematic review: 4 randomized controlled trials, 6,10-12 one of which is a long-term follow-up RCT 11 assessing the efficacy of CBT in adults suffering from TRD; 2 randomized controlled trials 3,13 assessing the efficacy of CBT in adolescents suffering from TRD, one of which is a long-term followup RCT 13 ; 1 open trial 14 ; and 1 case report. 15 The search and selection process for articles is illustrated in a flow diagram (see Figure 1).

Methods
Six studies 3,6,10-13 selected for the review were randomized controlled trials published in English. The participants in these studies were randomly assigned to different types of intervention groups in order to compare them and detect results. One study 14 was an open trial using only one group of participants and administering the intervention to all participants. Finally, one study 15 was a case report that used the case of a woman receiving individual CBT therapy to illustrate whether CBT is effective for TRD.

Intervention characteristics
The randomized controlled trials in this systematic review analyzed two types of interventions: TAU alone or TAU + CBT. Participants in the open trial study 14 and the case study 15 received only CBT as treatment modality.
In all cases, TAU was prescribed by psychiatrists: TAU is treatment of these patients with medication only.
Psychiatrists usually use SSRI antidepressants for TAU.
In the studies included in this systematic review, CBT was facilitated by therapists with at least a masters degree in a mental health field and prior experience in CBT 3,15 ; by four psychiatrists, 1 clinical psychologist with a Masters degree and 1 psychiatric nurse 6 ; by two experienced and well-trained therapists under the supervision of a PhD-level clinical psychologist who had six years' experience in treating depressed patients with cognitive therapy 14 ; by therapists who received at least 1 day of training specific to the trial from an experienced CBT therapist and trainer and weekly supervision from skilled CBT supervisors at each center 11 ; and, finally, by therapists with a master's degree in a mental health field and a license permitting provision of therapy, trained for 35 hours in a face-to-face workshop. 12 All of the participants were assessed with validated scales that assess depressive symptoms and most of them were followed-up for at least 12 months after intervention.

Outcomes
In all studies, the primary outcome assessed was alleviation of depressive symptoms measured using validated scales and compared from baseline to at least 12-month follow-up and as much as 40-month follow-up.

Risk of bias within studies
We utilized the RCT-PQRS to assess the risk of bias in individual studies. 9 The results are summarized in Tables 2 and 3. Only the randomized controlled trials  were assessed for quality and risk of bias. All randomized controlled studies were rated as low risk of bias.

Randomized controlled trials
In a randomized controlled trial by Nakagawa et al., 6 a total of 80 patients aged 20-65 years old (mean = 39.5) were selected to participate in the study. All participants had at least a minimal degree of TRD and a score of ≥16 on the GRID-Hamilton Depression Rating Scale (GRID-HDRS). All eligible participants were randomly allocated to receive CBT plus TAU or TAU alone with 12-month follow-up. Those who received CBT plus TAU were offered 16 individual 50-minutes sessions scheduled weekly with up to 4 additional sessions. SSRIs were the most common antidepressant medication prescribed for the TAU only group. The primary outcome of this study was alleviation of depressive symptoms measured by  Participants with other disorders such as bipolar or psychotic disorder were excluded, as were participants who received CBT treatment in the previous 3 years.
Participants were taking antidepressants at the time of randomization and were expected to continue with the drugs. In addition, they were randomly allocated to receive CBT plus TAU or TAU alone. Those who received CBT + TAU received 12 sessions of individual CBT lasting 50-60 minutes with a further six sessions: 90% of patients were followed up at 6 months and 84%     Table 1). Measurements were made pre-therapy, mid-therapy and post-therapy. The results showed that following 12 sessions of CBT, given in combination with a course of venlafaxine 375 mg daily, there were clear improvements in mood and in levels of hopelessness as well as in overall social and occupational functioning. These improvements were reflected in the rating scales: pre-therapy (BDI score = 31, BHI score = 16), mid-therapy (BDI score = 25, BHI score = 13) and post-therapy (BDI score = 16, BHI score = 7). The patient benefited from the combination of CBT + TAU.

Summary of evidence
This review systematically evaluated the efficacy of CBT for TRD in participants aged at least 12 years, based upon evidence from controlled studies that used All of these results were based on use of validated scales to assess depressive symptoms over 12 months of followup and even as much as 46 months of follow-up.
Just one randomized controlled trial showed that the combination CBT + TAU was also effective in an adolescent population. It is known to be the first clinical trial to enroll adolescents with depression who were not responding to an evidence-based treatment. However, when a follow-up of the same study was conducted, the results showed that the effects of CBT did not last up to 24 weeks. The findings could be applicable to community samples, which, while often more ethnically diverse than the study sample, have comparable clinical complexity.
Results address a relevant and important issue in the clinical aspect of a highly prevalent disease. When depressed patients do not respond to evidence-based interventions, healthcare providers and mental health experts should think of other strategies to manage these patients. This systematic review will help mental health experts by providing robust evidence that combination treatment for TRD patients is a choice that should be considered in their management, since it shows promising results for alleviation and reduction of depressive symptoms.

Limitations
There are several limitations to consider when interpreting this review. The first is that only four randomized controlled trials addressed this topic. 6,10,12 Moreover, only one randomized controlled trial 3 studied an adolescent population. Furthermore, there were insufficient data to conduct a meta-analysis, hence the absence of such an analysis in our systematic review.

Conclusions
The difficulty of treating patients with TRD poses considerable challenges to healthcare providers and mental health experts. The challenge is greater because of the limited number of studies. Since the available findings are interesting and some demonstrate the efficacy of CBT for TRD patients, this study has provided further support for the contention that adding CBT to TRD treatment is a promising and effective approach.
Significantly more CBT patients had clinically meaningful reductions in symptoms, according to the validated scales of depression used in the studies reviewed. However, it should be noted that there is a need to conduct additional studies of this topic. More specifically, further research is needed to investigate in greater detail whether adding CBT to TRD is beneficial in adolescents, since 40% of them will not respond to TAU treatment.

Disclosure
No conflicts of interest declared concerning the publication of this article.