Clinical severity, treatment resistance, and recurrence of depression

Emeritus Professor of Psychiatry
Imperial College University of London

Beijing Anding Hospital Capital Medical University
Beijing, CHINA

Clinical severity, treatment resistance, and recurrence of depression

by S. Montgomery and G. Wang, United Kingdom and China

Severe depression is themost challenging subgroup of depression to treat; it is associated with greater suffering, increased disability, and increased morbidity and mortality. Treatment studies have found that some antidepressants are significantly more effective in treating severe depression than other, more conventional, antidepressants. Significant advantages are reported with escitalopram and with agomelatine both in individual studies and in meta-analyses of the data that define these antidepressants as superior. The issue of poor compliance with treatment in depression highlights the need for antidepressants that are not just superior in efficacy, but also superior in tolerability. Severe depression appears to bemore prone to relapse and recurrence, and this finding emphasizes the need to prioritize the use of superior treatments in the long-term treatment of depression. The finding that severity of depression is one of the predictors of the development of resistant depression supports the view that severe depression should be aggressively treated with appropriate superior antidepressants to avoid the consequences of the depression becoming resistant.

Medicographia. 2011;33:132-137 (see French abstract on page 137)

Depression, or rather major depressive disorder (MDD), is a common disorder associated with a high level of morbidity and mortality. It is one of the commonest causes of disability worldwide and its impact is particularly striking in the West. Depression affects 1 in 4 women in their lifetime and 1 in 8 men. The onset of the disorder occurs in the late teens to early twenties; in almost all cases it is a chronic or recurring disorder that has major impact on work and productivity as well as the personal and social life of the sufferer. For these reasons effective treatment is a high priority.

The importance of severe depression

The diagnostic criteria used to define a number of psychiatric disorders, eg, the anxiety disorders, have changed substantially over the last 25 years. This draws attention to an underlying instability of the diagnostic process and raises concern that the diagnostic criteria may be imprecise. This is not, however, the case with the diagnostic criteria for MDD, which have hardly varied over the same period. There is substantial agreement on the defining features and core symptoms of the disorder, which are well defined in the widely used and recognized symptomatic diagnostic criteria. The same pivotal rating scales that were used 25 years ago to assess the efficacy of treatment are still in use today. What has changed is the recognitionof the importance of establishing the severity of depression in order to identify the risk of disability, the risk of relapse, and the risk of the high levels of morbidity of the disorder. Severe depression is associated with increased mortality, higher morbidity, and greater disability than moderate or mild depression; it is also the group where the relapse rate in responders or remitters is highest.1 The treatment of severe depression is therefore one of the most challenging areas encountered by the physician. The assessment of severity also turns out to be a major tool in quantifying the proportion of individuals who respond well and achieve a reduction in symptom score to a low level in line with that when well. The concept of remission and the importance of remission as a target for treatment have been confirmed in recent studies.

Specific investigations of treatment and outcome in subgroups of patients with differing severity levels indicate that severe depression lies on a continuum of severity from mild and moderate depression with no obvious cutoff separating moderate from severe depression; rather there appears to be an increasing dimensional risk of disability with increasing severity of depression from moderate to severe illness. There is, however, no evidence to suggest that severe depression is a separate disorder although available treatments have been shown to have increased efficacy in the severe subgroup.

Definitions of severe depression

Severe depression is comparatively easy to recognize. Reasonable concordance between assessors can be achieved as to whether depression is mild, moderate, or severe, although this may be affected by the inherent variability due to differing levels of experience of the assessors. Quite marked cultural differences in assessment occur and these may be exacerbated by insensitive translation of global rating scales. A well-known example is the mismatch in meaning between the English word “moderate” and the French word “moderé,” which has led to difficulties in merging the data from studies carried out in different cultures when poorly translated global scales have been used.

Precise ratings of the severity of the individual symptoms of depression have helped to derive more easily defended definitions of severe depression identified by a score on an accepted depression rating scale. Currently, the most widely used criteria to define severe depression are a score of 30 or more on the Montgomery-Åsberg Depression Rating Scale (MADRS),2,3 or a score of 25 or more on the 17-item Hamilton Depression Rating Scale (HAMD-17),4,5 which are the two most widely used pivotal scales. These definitions used in clinical treatment studies have the advantage that they appear to be in accord with clinical judgment. Other definitions have been used, for example, hospitalization or suicidal risk, but they are less specific and vary widely in different treating settings. Other factors unrelated to severity can play a role, for example the availability of hospital beds varies widely from country to country and even within a country from one region to another. Similarly, the risk of suicide is not necessarily an accurate marker of the severity of depression since suicidality is substantially altered by personality factors, by concomitant drug or alcohol abuse, by concomitant physical illness, by religious conviction, and by family support. The concept of melancholia, used in some countries to indicate severity, has been found to represent a loosely defined diagnostic subgroup of depression with varying levels of severity. Although melancholia as currently defined appears more common in severe depression, it can be found also in those with mild depression, moderate depression, and severe depression, and therefore the use of this term to define severity is inappropriate.

Differential efficacy in severe depression

Recognition of the importance of severe depression, in terms of associated increased morbidity, disability, and increased mortality, has stimulated research into the possible differential efficacy of various antidepressants. Initially, the analyses carried out found that selective serotonin reuptake inhibitors (SSRIs) were effective in both moderate and severe depression with no obvious difference in efficacy between the two groups. For example, in an elegant analysis of the placebo controlled data with citalopram the treatment effect, that is, the difference in response between citalopram and placebo, was constant in moderate and severe depression and there was no signal of any change in efficacy with increasing severity.6,7 In contrast, the treatment effect increased with increasing baseline severity with escitalopram and with agomelatine.8

The greatest response to placebo is observed in mild depression and the least in severe depression. Indeed, in mild depression the placebo response is so high that it is very difficult in clinical studies to identify a significant treatment effect of antidepressants. In the large placebo-controlled efficacy study of fluoxetine given in doses of 20, 40, or 60 mg in mild depression, defined by a baseline score of less than 20 on the HAMD, there was no perceptible difference between placebo and any dose of fluoxetine.9 It is probable that any pharmacological effect in thesemildly depressed patients wasmarginal and the close attention and concern to be expected in the conduct of a clinical treatment trial had the consequence of improving all groups during the short duration of the study and obscured any difference. Factors such as the careful assessment and rating of depression in the context of a clinical study carry a nonspecific benefit that appears to be more marked in mild depression and less in severe depression.

Some antidepressants that are effective in moderate-to-severe depression have, however, been shown to have increased efficacy in the severely depressed subgroup. Analyses of the escitalopram clinical trial data were carried out to examine the possible influence of severity at the start of treatment on patients’ outcomes.When patients were categorized into cohorts of severity, a clear increase in efficacy was observedwith increasing severity.7,10 If some, but not all, antidepressants are associated with this important feature of extra response in severe depression it is of great interest for the clinician to clarify in direct comparisons whether a significant difference between antidepressants can be established. The most scientifically rigorous way to investigate potential differences between antidepressants is to conduct individual studies with random double-blind assignment to treatment to reduce the risk of bias in the population studied. Providing that the studies are carried out under conditions of fair comparison, the results can show whether there are differences in efficacy and whether there is clear-cut superiority for some antidepressants.

A number of antidepressants have been compared and significant differences reported. Escitalopram was found to be more effective than citalopram in treating severe depression defined as a baseline MADRS score of 30 or more11,12 and also more effective than paroxetine.13 More recently, agomelatine was found to show increasing treatment effect (difference from placebo) with increasing severity of baseline depression in an analysis of the pivotal placebo-controlled studies.8 The greater treatment effect observed in the more severe patients suggested that agomelatine might also have advantages compared with other antidepressants in treating more severe depression. This has been confirmed on the pivotal efficacy measure in a prospective randomized double-blind study in severe depression comparing agomelatine with fluoxetine.

The severity of depression in this study was defined as having a baseline score 25 or more on the HAMD 17-item scale, and agomelatine was shown to be significantly superior to fluoxetine. Previous studies in moderate and severe depression have shown the superiority of agomelatine in comparison with sertraline14 and venlafaxine.15 Venlafaxine has itself a reputation of being more effective in severe depression16 though the evidence from formal studies is sparse. A large meta-analysis reported a modest, but significant, increase in the percentage of remissions (6%) with increasing baseline severity.17 This finding is supported by the superiority of venlafaxine demonstrated in a double-blind randomized comparison with fluoxetine in a largely severe depression population.18 Extra efficacy in severe depression is now regarded by many as the hallmark of the superior antidepressant.

The results from comparisons of individual antidepressants are supported by a retrospectivemeta-analysis of several studies, though in this analysis the agomelatine studies were excluded.19 The use of meta-analyses has its drawbacks since they are almost always conducted post hoc and they generally make the assumption that every study is equally fair, valid, and useful. Attempts to exclude biased or unfair comparison studies such as when the dose of one antidepressant can be raised to a maximum while the other remains low20 or when the population is already resistant to one of the treatments21 are rare. Nevertheless, despite their weaknesses, meta-analyses can provide support for the findings from individual comparisons that antidepressants differ in efficacy and tolerability. Identifying the antidepressants that have superior efficacy is an important step in guiding treatment choice. We need to concentrate our treatment efforts, giving priority to the use of antidepressants that are superior, and also, since compliance with treatment is such a problem in depression, to those that additionally have a better tolerability profile close to that seen with placebo.

Discontinuation symptoms

Abrupt termination of antidepressant treatment is in almost all cases associated with the appearance of discontinuation symptoms.22 Discontinuation symptoms have been reported with all classes of antidepressants since the early studies with imipramine. In some cases, it appears that these discontinuation symptoms may destabilize the clinical status and induce relapses. The discontinuation symptoms, which usually appear in the first few days after stopping treatment and reach maximal severity in the first week, may well be mistaken for relapses. This phenomenon complicates the design of the relapse prevention studies, which require that responders or remitters to the treatment under investigation, for example an antidepressant, are then randomized to receive treatment with either the same antidepressant or placebo under double-blind randomized conditions.

Some antidepressants appear to be more prone to produce discontinuation symptoms than others, eg, paroxetine22 or venlafaxine.23 Some antidepressants have a lower propensity to cause discontinuation symptoms, eg, fluoxetine24 or escitalopram,25 but only agomelatine has been found to be completely free of discontinuation symptoms. The clear difference between antidepressants in the rate of associated discontinuation symptoms is illustrated by a placebo-controlled discontinuation study that investigated rates in agomelatine and paroxetine. Stopping agomelatine treatment was associated with no signal of discontinuation symptoms compared with continuing agomelatine. In contrast, in the same study, after stopping paroxetine, given in a low dose of 20 mg, up to 45% of patients showed discontinuation symptoms.26

Discontinuation symptoms and relapse

The need for long-term treatment of depression to prevent relapse following response to treatment and to reduce the risk of recurrence has long been established and is universally recommended. Improving long-term outcome is of particular concern for those who suffer severe depression, given the potential serious consequences. The clinically significant efficacy of antidepressants in long-term treatment is supported by an overwhelming weight of evidence in the literature. One example is the long-term placebo-controlled study of agomelatine, carried out largely in severe depression, which found that agomelatine has a powerful ability to reduce the risk of subsequent relapses or recurrences. This ability is reflected inthe strikingly low number of patients needed to treat (NNT) in order to show a significant advantage compared with placebo.27 There is also some indication in the agomelatine database that the risk of relapse is low in the mild depression subgroup suggesting that mild depression may not be the target for long-term antidepressant treatment. In placebo-controlled studies mild depression has a high spontaneous response rate, and evidence to support the need for long-termtreatment would therefore be difficult to obtain. Mild depression might be a lower priority for both short- and long-term treatment.

We can have less confidence in the evidence of efficacy in relapse prevention derived from studies where the abruptly discontinued treatment is associated with frequent or troublesome discontinuation symptoms. For example, in the relapse prevention study of fluvoxamine there was a high and rapid relapse rate, which was associated with high levels of discontinuation symptoms.28 To overcome the problem it is now routine procedure in a secondary analysis to exclude from the analysis data from the first month after stopping treatment to try to discount the discontinuation symptoms. A relapse prevention study with agomelatine is perhaps the first where the results are uncontaminated by a possible increase in the relapse rate associated with discontinuation symptoms, since such symptoms are not a problem with agomelatine. The efficacy of agomelatine in relapse prevention is therefore more securely based than other antidepressants that are prone to discontinuation symptoms and were tested using the same design.

Resistant depression

Nonresponse to antidepressants is common and is reported in between 30% and 50% of patients treated with antidepressants in randomized double-blind placebo-controlled clinical studies. Surprisingly few studies have investigated what happens to these patients and as a result the evidence base on which advice can be given is limited. The short duration of the acute studies, normally 6 to 8 weeks, leaves unanswered the question as to whether there may be a further response if treatment were contained longer. This could be the case, but the longer 12-week placebo-controlled studies, which also report similar high levels of nonresponse, suggest otherwise.

The treatment of resistant depression is a pressing clinical problem that dominates the practice of almost all specialists. In the US, a range of treatments are licensed as adjuncts in resistant depression, but, in the EU, regulatory recognition of the problem is more limited. This stems partly from the complicated definition of resistant depression used by the EU,29 which requires both a prospective demonstration of resistance and demonstration that patients have not responded to two different antidepressants from different pharmacological classes, each given in an adequate dosage for sufficient duration. Moreover, a treatment proposed for resistant depression has to show persistence of efficacy compared with placebo for at least 6 months. The evidence supporting the validity of these complicated requirements is trivial and appears to be based on a single meta-analysis, which included unfair comparisons, that showed a modest advantage in remission, though not responders, using these criteria.30

Other much larger studies have failed to find this advantage.31,32 In a very large survey of resistant depression across Europe the Treatment Resistant Depression Group found that switching classes of antidepressants was not helpful and that continuing with the same antidepressant for longer was associated with a significantly better response. In a prospective study in resistant depression, switching nonresponders from citalopram (a SSRI), to desipramine (a tricyclic norepinephrine reuptake inhibitor) was associated with significantly fewer responders than staying on the same medication. The problem with switching antidepressants comes partly from the discontinuation symptoms on stopping the previous antidepressant and partly from the slow onset of action of the new antidepressant, normally around 6 weeks. The effect of these findings should be to encourage prescribers to use antidepressants that have low or no discontinuation symptoms and also to use antidepressants with superior antidepressant efficacy including a faster onset of antidepressant action.

The criteria for defining treatment resistance have led to studies that have shown that the key criterion is the failure to respond to a single course of treatment with an approved dose for adequate duration.33,34 The extra requirement of demonstrating a failure to respond to a second treatment adds little and delays the diagnosis. The suggestion that an antidepressant from a different class must be tried to establish treatment- resistant depression is patently inaccurate and merely delays the correct diagnosis even further.

The predictors of resistant depression are well known.35 Most of these factors relate to severity of depression and also include comorbidity with any anxiety disorder (which of course increases the severity), the presence of suicidality, which is frequently severity related, and failure to respond to treatment of the first episode of depression. This suggests that resistance to treatment is an enduring feature that persists from one episode of depression to the next and emphasizes the need to use the more effective, superior antidepressants early, before resistance becomes established.

The same finding provides a rationale for seeking a possible genetic basis for nonresponse and resistance. The finding that polymorphism of the short arm of the serotonin transporter increases vulnerability to stress and increases levels of nonresponse36 has been confirmed in a number of studies. It now appears from the studies of the Treatment-Resistant Depression Group and the STAR*D studies (Sequenced Treatment Alternatives to Relieve Depression) that a numberof other genetic factors such as COMT (catechol-O-methyltransferase) also are also associated with increased resistance.

The finding that severity of depression and associated features is a predictor of treatment-resistant depression is provocative, and this raises the uncomfortable question that initial treatment with weaker conventional antidepressants may in some way induce resistance. In this context, current guidelines that are influenced by economic considerations, such as those of the National Institute for Health and Clinical Excellence (NICE) committee, and suggest using the weakest treatments first and without discussion of the superior efficacy of some antidepressants, now appear unacceptable.

Severe depression is associated with increased morbidity and mortality as well as increased treatment resistance and should be prioritized for accurate diagnosis and special treatment. We need to rethink our treatment recommendations and treat severe depression aggressively to achieve response and remission with those antidepressants that have been shown to be particularly effective. _

1. Prien RF, Kupfer DJ, Manskey PA, et al. Drug therapy in the prevention of recurrences in unipolar and bipolar affective disorders: Report of the NIMH Collaborative Study Group comparing lithium carbonate, imipramine and a lithium carbonate-imipramine carbonate combination. Arch Gen Psychiatry. 1984;41: 1096-1104.
2. Montgomery SA, Asberg M. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979;134:382-389.
3. Burke WJ, Gergel I, Bose A. Fixed-dose trial of the single isomer SSRI escitalopram in depressed outpatients. J Clin Psychiatry. 2002;63:331-336.
4. Hamilton M. A rating scale for depression. J Neurolog Neurosurg Psychiatry. 1960;23:56-62.
5. Montgomery SA, Lecrubier Y. Is severe depression a separate indication? European Neuropsychopharmacology. 1999;9:259-264.
6. Montgomery SA, Moller HJ. Is the significant superiority of escitalopram compared with other antidepressants clinically relevant? Int Clin Psychopharmacol. 2009;24:111-118.
7. Lam RW, Andersen M. The influence of baseline severity on efficacy of escitalopram and citalopram in the treatment of major depressive disorder: an extended analysis. Pharmacopsychiatry. 2006;39:180-184.
8. Montgomery SA, Kasper S. Severe depression and antidepressants: focus on a pooled analysis of placebo-controlled studies on agomelatine. Int Clin Psychopharmacol. 2007;22:283-291.
9. Dunlop SR, Dornseif BE, Wernicke JF, Potvin JH. Pattern analysis shows beneficial effect of fluoxetine treatment in mild depression. Psychopharmacol Bull. 1990;26:173-180.
10. Kennedy SH, Andersen M, Thase ME. Escitalopram in the treatment of major depressive disorder: a meta-analysis. Curr Med Res Opin. 2009;25:161-175.
11. Moore N, Verdoux H, Fantino B. Prospective, multicentre, randomized, doubleblind study of the efficacy of escitalopram versus citalopram in outpatient treatment of major depressive disorder. Int Clin Psychopharmacol. 2005;20:131-137.
12. Yevtushenko VY, Belous AI, Yevtushenko YG, Gusinin SE, Buzik OJ, Agibalova TV. Efficacy and tolerability of escitalopram versus citalopram in major depressive disorder: a 6-week, multicenter, prospective, randomized, double-blind, active-controlled study in adult outpatients. Clin Ther. 2007;29:19-32.
13. Boulenger J-P, Huusom AKT, Florea I, Baekdal T, Sarchiapone M. A comparative study of the efficacy of long-term treatment with escitalopram and paroxetine in severely depressed patients. Curr Med Res Opin. 2006; 22:1331-1341.
14. Kasper S, Hajak G, Wullf K, et al. Efficacy of the novel antidepressant agomelatine on the circadian rest-activity cycle and depressive and anxiety symptoms in patients with major depressive disorder: a randomized, double-blind comparison with sertraline. J Clin Psychiatry. 2010;71:109-120.
15. Lemoine P, Guilleminault C, Alvarez E. Improvement in subjective sleep in major depressive disorder with a novel antidepressant, agomelatine: randomized, double-blind comparison with venlafaxine. J Clin Psychiatry. 2007;68: 1723-1732.
16. Montgomery SA, Baldwin DS, Blier P, et al. Which antidepressants have demonstrated superior efficacy? A review of the evidence. Int Clin Psychopharmacol. 2007;22:323-329.
17. Nemeroff CB, Entsuah AR, Benattia I, Demitrack MA, Sloan DM, Thase ME. Comprehensive analysis of remission (COMPARE) with venlafaxine versus SSRIs. Biol Psychiatry. 2007;63:424-434.
18. Clerc GE, Ruimy P, Verdeau Pailles J. A double-blind comparison of venlafaxine and fluoxetine in patients hospitalized for major depression and melancholia. Int Clin Psychopharmacol. 1996;9:139-143.
19. Cipriani A, Furukawa TA, Salanti G, et al. Comparative efficacy and acceptability of 12 new-generation antidepressants: a multiple-treatments meta-analysis. Lancet. 2009;373:746-758.
20. Poirier MF, Boyer P. Venlafaxine and paroxetine in treatment-resistant depression. Double-blind, randomised comparison. Br J Psychiatry. 1999;175:12-16;
21. Hoehn-Saric R, Ninan PT, Black DW, Stahl S, Greist JH, Lydiard RB et al. Multicenter double-blind comparison of sertraline and desipramine for concurrent obsessive-compulsive and major depressive disorders. Arch Gen Psychiatry. 2000;57:76-82.
22. Rosenbaum JF, Fava M, Hoog SL, Ascroft RC, Krebs WB. Selective serotonin reuptake inhibitor discontinuation syndrome: a randomized clinical trial. Biol Psychiatry. 1998;2:77-87.
23. Montgomery SA, Fava M, Padmanabhan SK, Guico-Pabia CJ, Tourian KA. Discontinuation symptoms and taper/poststudy-emergent adverse events with desvenlafaxine treatment for major depressive disorder. Int Clin Psychopharmacol. 2009;24:296-305.
24. Judge R, Parry MG, Quail D, Jacobson JG. Discontinuation symptoms: comparison of brief interruption in fluoxetine and paroxetine treatment. Int Clin Psychopharmacol. 2002;17:217-225.
25. Baldwin DS, Cooper JA, Huusom AK, Hindmarch I. A double-blind, randomized, parallel-group, flexible-dose study to evaluate the tolerability, efficacy and effects of treatment discontinuation with escitalopram and paroxetine in patients with major depressive disorder. Int Clin Psychopharmacol. 2006;21:159-169.
26. Montgomery SA, Kennedy SH, Burrows GD, Lejoyeux M, Hindmarch I. Absence of discontinuation symptoms with agomelatine and occurrence of discontinuation symptoms with paroxetine: a randomized, double-blind, placebocontrolled discontinuation study. Int Clin Psychopharmacol. 2004;19:271-280.
27. Goodwin GM, Emsley R, Rembry S, Rouillon F. Agomelatine prevents relapse in patients with major depressive disorder without evidence of a discontinuation syndrome: a 24-week randomized, double-blind, placebo-controlled trial. J Clin Psychiatry. 2009;70:1128-1137.
28. Terra JL, Montgomery SA. Fluvoxamine prevents recurrence of depression: results of a long-term, double-blind placebo-controlled study. Int Clin Psychopharmacol. 1998;13:55-62.
29. Committee for Proprietary Medicinal Products. Note for Guidance on Clinical Investigation of Medicinal Products in the Treatment of Depression. The European Agency for the Evaluation of Medicinal Products Evaluation of Medicines for Human Use, CPMP/EWP/518-97-Rev.1. 2002.
30. Papakostas GI, Fava M, Thase ME. Treatment of SSRI-resistant depression: a meta-analysis comparing within- versus across-class switches. Biol Psychiatry. 2008;63:699-704.
31. Bschor T, Baethge C. No evidence for switching the antidepressant: systematic review and meta-analysis of RCTs of a common therapeutic strategy. Acta Psychiatr Scand. 2010;121:174-179.
32. Ruhe HG, Huyser J, Swinkels JA, et al. Switching antidepressants after a first selective serotonin reuptake inhibitor in major depressive disorder: a systematic review. J Clin Psychiatry. 2006;67:1836-1855.
33. Fava M. Diagnosis and definition of treatment-resistant depression. Biol Psychiatry. 2003;53:649-659.
34. Thase ME, Rush AJ. When at first you don’t succeed: sequential strategies for antidepressant nonresponders. J Clin Psychiatry Suppl. 1997;13:23-29.
35. Souery D, Oswald P, Massat I, et al. Clinical factors associated with treatment resistance in major depressive disorder: results from a European multicenter study. J Clin Psychiatry. 2007;68:1062-1070.
36. Caspi A, Sugden K, Moffitt TE, et al. Influence of life stress on depression: moderation by a polymorphism in the 5-HTT gene. Science. 2003;301:291-293.

Keywords: depression; relapse; recurrence; morbidity; mortality; antidepressant; treatment compliance; treatment resistance; escitalopram; agomelatine