Hong Kong Journal of Psychiatry (1999) 9 (1) 3-11

REVIEW

pdf Full Paper in PDF

INTRODUCTION

Inadequate response to antidepressants is not uncommon. The U .S. CoIIaborative Study on the Psychobiology of Depression one of the major follow-up studies of depressed patients - has demonstrated significant rates of both chronicity and recuηence. In early reports from that studyKeller et aI. (1984) found that 21% of pa ents had not recovered after two years. Aer five years 12% 1I remained depressed (KeIIer et aI.1992). Longer follow-up studies over 15 to 20 years by other groups (e.g. Kiloh et aI. 1988; Lee and Murray1988) have found similarly high rates of both chronic and recurrent depression.

Although the term "treatment-resistant depression" is used commonlythere is Iittle agreement over how such a term should be defined (Wilhelm et 1994). Fawceìt and Kravitz (1985)for examplehave described seven levels of treatment resistancewith each level being defined in terms of duration of i1Inessas weII as type and dose of particular somatic therapies.

This article will be based upon the premise that there are different severities of eatment-resistancewhich ry from non­ response to a single ade late cαrrse of an antidepressant to failure to benefit from a number of antidepressant therapies including ECT ermore it will arbi rily define the "benchmark" level of treatment resistance as failure to respond to an adequate course of SSRI medications. Such a pragma c definition will be used as SSRIs have now become the rst-Iine antidepressant therapy in both general and psychiatric practice in many countries.

Which patients are Iikely to be refractory to antidepressant medications? RecentlyNelson et aI. (1994) found that patients who failed to respond to the tricyclic desipramine were more Iikely to have a personalidisorder prior treatment failure"near delusional" statusand be less than 35 years of age. Other less significant characteristics of non-responders were longer dura on of the current episode recurrent depressiondysthymia and secondary depression. Bonner and Howard (1995), studying an elderly populationfound resistance to be associated with onset of depression before 50hypochondriacal features and the presence of cognitive impairent.

The most effective means of treating depression incorporate both biological and psychologicaI therapies (Wilhelm et 1994; MitcheII1997a). There has been a tendencyhoweverfor psychological therapies to be under­ valued in the treatment of depressed patients particularly those with a non-melancholic subtype. While this article wilI focus upon the role of physical therapiesthis should not be interpreted as minimising the importance of non­ pharmacological options.

Prior to detailing specific therapies for treatment-resistant depression it is necessary to comment upon e consequences of such failure to respond. For those suffering from treatment-resistant depression there is severe demoralisationas the hopes and expectations of recovery with each succeeding course of antidepressants are repeatedly dashed. Low self-esteem and self-confidence lead to a gradual but inevitable withdrawaI from familyfriends and previous interests. For those close to the patient there is a gradual ₁₀₅₅ of empathy for their loved ₀ e. Irritation with their failure to improve leads to a tendency to "blame" the patient. While it is difficult for any family to maintain compassion for a relative with short term depression this becomes even more burdensome when the depression is unremitting. Such difficulties not infrequently lead to separa ondivorce orat the leastemotional withdrawal.

For the cnnician the frus ation is none the less. The continued failure of the patient to respond engenders a sense of therapeutic impotence. lt is not infrequent for the docìor to change the diagnosis to one of "personan disorder" a strategy which often reflecìs therapist frustration rather than cnnical reality. The diagnosis of personanty disorder may "absolve" the docìor from further responsibility for active treatment

ASSESSMENT

A detailed review of the paent with treatment-resistant depression is a critical first step before embarking upon "yet another" round of medications.

ADEQUACY OF PREVlOUS ANTIDEPRESSANTS

Duration

What is an adequate duration of a course of antidepressants? A number of recent studies have provided some empirical evidence on which to base this difficult decision. Quitkin et al. (1996) studied 593 pa ents who had been randomised to either active an depressant or placebo hose patients who showed no evidence of improvement by week four evinced no addi onal benefit from the antidepressant over placebo by e study end at six weeks

On the basis of s findingthe authors recommended at pa ents who have not had at least a minimal response to medications by week four should have their treatment regimen altered. Similarly those who have had only minimal improvement by week five should have their treatment changed.

The other study pertinent to this issue was an eight-week open trial of fluoxetine by Nierenberg et al. (1995). Of those patients who showed no improvement at weeks four or sixonly 19% and 7% respectively finally responded by week eight. Nierenberg et al. concluded that a lack of improvement by weeks four to six indicated negligible chance of later response

In view of these findingsit is reasonable to conclude that four weeks probably comprises an adequate trial of an antidepressant. Only if there is partial improvement by that time should the an depressant be continued.

Dose

What comprises an adequate dose of SSRIs? Most of the SSRI nterature (Boyer and Feighner1991) indicates at these antidepressants demonstrate a flat dose-response curvei.e. there is no definite benefit in increasing the dose beyond the starting dose (20 mg for fluoxetine or paroxetine and 50 mg for sertraline). Furthermore Dornseif et al. (1989) demons ated that for patients who had fai!ed to respond to fluoxe ne 20 mg aer three weeks randomisation to 60 mg or continuation of the original 20 mg resulted in similar rates of improvement over the subsequent five weeks.

The studies of Fava et al. (1992; 1994)however, appear to contradict those ndings. First, in an open study, Fava et al. (1992) found that in 15 patients initially unresponsive to fluoxetine 20 mg dailythere was a signicant improvement aer the dose was then increased to 60 to 80 mg for a further 4 weeks.

Fava et al. (1994) later studied 41 pa ents who had failed to respond to fluoxe ne 20 mg after eight weeks. Pa ents were randomly assigned to either high dose fluoxetine (40 to 60 mg) fluoxetine 20 mg and desipramineor fluoxetine 20 mg and lithium. They found that high dose fluoxetine was the most effective treatment among those who had only par ally responded to the previous fluoxe ne therapy. Moreoverey also found that high dose fluoxetine was effective for previous non­ responders.

In general, howevermost of the SSRI literature indicates minimal benefit in increasing the dose beyorthe usual therapeutic range. This author would recommend that a doubling of the initial SSRI dose may be a useful first step for failure to respond (e.g. increasing sertraline from 50 mg to 100 mg)but there is minimal benet in increasing the dosage beyond that. Drop-out rates due to severe side effecìs increase significantly at higher doses of all the SSRIs (Boyer and Feigner1991).

PSYCHOSOCIAL ASSESSMENT

The psychosocial contribuon to resistant depression is frequently ignored or under-recognised (Wilhelm et al. 1994; Thase and Howland 1994). Premorbid personality functioning (e.g. dependent or bordernne personan characìeristics or high trait anxiety) and coping styles should be assessedin addition to family or relationship issuesand per nent life events (either acute or chronic). While the presence of such psychosocial facìors do not preclude the need for antidepressant medicationsthey may indicate the importance of concurrent psychosocial treatments such as cognile ehavioural therapy asser veness training or family therapy (Fava et 1997).

BIOLOGICAL ASSESSMENT

Non-psychotropic medications

Medications including steroids dopaminergic anti­ Parkinsonian agents (L-dopabromocrip ne) and a number of older antihypertensive agents (such as beta-blockers alpha-methyl-dopa or reserpine) may contribute to or exacerbate depression (Mitchell, 1994).

Substance abuse

One of the commonest physical causes of treatment­ resistant depression in clinical practice is abuse of alcohol or other substances. Maruana in particular is potent in exacerbating a y pre-existing tendency to depression

Physical illness

Although very uncommon in clinical practice the clinician must consider the possibility of an unrecognised underlying physical disorder (Mitchell 1994). Endocrino­ pathies such as hypothyroidism Cushing's syndrome or Addison's disease should be considered. There have been a number of reports of either cllnical or subclinical hypothyroidism being more common in patients with chronic or trea lent-resistant depression (Howland 1993; Hickie et al. 1996). Neurological conditions such as Parkinson's disease or multiple sclerosis may produce secondary depressionbut u lally signs or symptoms of such disorders are apparent.

Vascular depression

Another relatively common physical cause of depression is cerebrovascular disease. There has been an increased awareness of the aetiological contribution of cerebrovascular illness to depression since reports of a strong association between late-onset depression and "hyperintense regions" demonstrable on MRI !anning (Brown et al.1992). These regions probably represent disease of small subcortical vessels.

Recently, two denitions of vascular depression - one clinicaJ (A1exopoulos et 1997)the other radioJogical (using MRI scans) (Krishnan et al. 1997) - have been described. Furthermorethere is preliminary evidence that such "vascular depression" may predict poor response to antidepressant medica ns or ECT (Hickie et al.1995). In the latler study 39 severely depressed inpa ents were evaluated by MRI scanning with the findings being correlated with response to standard antidepressant therapies. The severity of white matler hyperintensities (WMHS) predicted poorer response to treatment (r ₌ -0.44 p < 0.01). There were strong nega ve correla ons of a similar order for responses to both antidepressant medications and ECT.A fu her longitudinal study on that same depressed sample (Hickie et al.1997) has found that extensive WMHS predispose to chronic depression ar progressive cognitive decline. In another report, Simpson et al. (1997) described neuropsychologicaIneurological and radiological evidence of subcortical vascular pathology in the majority of 14 patents referred to a cerebral function unit for assessment of treatment-resistant depression. 

Other biological processes

Recently Mayberg et al. (1997) reported that hypometabolism in the rostral anterior cingulate region (demonstrated on PET scanning) was associated with poor response to an depressants. Hypermetabolism in that region was characteristic of responders.

TREATMENT STRATEGIES

DOSE INCREASE

As discussed above there is conflicting data concerning the usefulness of increasing the dosage of SSRIs in patients who fail to respond to the usual therapeutic doses. As the majority of studies indicate a f1at dose-response curvethere would appear to be minimal benet in more than doubling the initial dose.

AUGMENTATION OR COMBINATION OF ANTIDEPRESSANTS

The best controlled research data for managing treatment-resistant depression pharmacologically is for augmentation of a failed course of 'antidepressants. While most studies have tested lithium augmentationthere are a number of studies of T₃and in more recent yearspindolol. Despite this literatureaugmentation appears to be used uncommonly in cllnical prac ceeven amongst psychiatrists. In a recent survey of United Kingdom psychiatristsShergill and Katona (1997) surveyed the most commonly used strategies for dealing with resistant depression. he most popular treatment choices were increasing the dosage of tricyclic an depressants (CAs) or changing from TCAs to an SSRI. Augmenta on with lithium or '3 was used only occasionally. About 39% of psychiatrists stated that they were not confident in treating refractory depression

Lithium

The best researched strategy for resistant depression is augmentation of a failed course of antidepressants by lithium (de Mon gny 1994). Originally described by de Montigny et al. (1981, 1983) for lithium augmentation of MAOIs or tricyclic antidepressantsthis approach has been demonstrated to be effective with all an depressants including SSRIs (e.g. Katona et al.1995). Austin et al. (1991) calculated in a meta-analysisthat approximately 40% of pa ents resistant to an adequate course of antidepressants responded to lithium augmentation.  

This strategy is effective for both unipolar and bipolar depressed pa ents. Although originally described by de Montigny et al. (1981) as occurring within 48 hours response is more commonly found to develop within 1 to 3 weeks (Price and Heninger1994). While no relationship has been found between lithium levels and response to augmenta on most studies demonstra ng effec veness have used levels of at Jeast 0.4 mmol/L (Nelson & Price 1995).

There were early concerns about both pharmacodynamic and pharmacokinetic (SaJama and Shaff 1989) interactions between lithium and the SSRIswith particular anx18about the development of the serotonin syndrome. Such complica ons havehowevernot been substantiated in controlled studies (Mitchell 1997b). In the elderlythough there is a rela vely high rate of adverse effects with Iithium augmentation (Rint and Rifat 1994). In that latler open study of lithium augmenta on of a variety of antidepressants in eJderly patients approximateJy 50% of patients developed dose-limiting side effectsmainly neuromuscuJar or neurologic These adverse effects occurred more commonly in patients on f1uoxetine than in those prescribed nortriptyline or phenelzine.

Triiodothyronine (f3)

As recently reported in a meta-analysis by Aronson et al. (1996)there have now been eight controlled clinical studies of T₃ augmentation in euthyroid patients with refractory depression. That meta analysis of 292 pa ents found that those treated with T₃ augmentation were twice as likeJy to respond as controls. Of the eight studiesonly four used randomised doubJe-blind designs. When those four studies were pooled there was no significant benefit of T3 although one strongly negative study may have accounted for this overall negative finding.

There have as yet been no published reports of T₃ augmentation of SSRIswith aIl studies being undertaken with the TCAs imipramine or desipramine. The dosage of T₃ used in the studies was 25 o mcg dai!y. There has been one comparison of T3 and thyroxine (T4) augmenta on (Joffe and Singer1990)but the duration of the trial (3 weeks) was too short to demonstrate any possible effec veness of the longer half-life T_{4 .}

Moreover, there has only been one 5 dy comparing e relative effects of T₃ and lithium augmenta on (Joffe et al. 1993a). That placebo-con oIled randomised study foul1<no signicant difference between these two augmen ng agents. In a later paper based on the original study dataJoffe et al. (1993b) examined for poten al predictors of response to lithium and T₃ augmenta on in tricyclic non-responders. Significantly they did not find at thyroid func on predicted differential responsiveness. Lithium non­ responders werehowevermore severely depressed (wi more insomnia and weight 1055) than lithium respol1<ers. T ₃ responders had experienced greater weight 1055 than lithium responders. There were no differences between T₃ responders and non-responders.

In summarywhile there is some evidence for the benefit of T₃ augmentation5 is by no means as substantial as that for lithium augmentation. As yetthere have been no studies of T₃ augmenta on of SSRIs.

Tricyclic antidepressants (TCAs)

There have been a smaIl number of studies suggesting the benefit of combined SSRIs and TCAs in treatment­ resistant depression. In generalhowevermost of these studies have been open in design (e.g. Zajecka et al. 1995). Perhaps the most interesting study was the trial of Fava et al. (1994) discussed previously. ln that study of patients who had fai!ed to respond to eight weeks of treatment with fluoxe ne 20 mg daisubjects were ral1<omised to either high dose fluoxetinefluoxetine and desipramine 25-50 mg dai!yor fluoxetine and lithium. Ruoxetine combined with desipramine was a relatively ineffective treatment. The low dosage and inadequate blood levels of desipramine used in at study have however been subsequently criticised (Nelson & Price1995). Nelson et al. (1991) had previously demonstrated - in an open historical comparison study -at e combination of fluoxetine and desipramine produced a more rapid rate of response than desipramine alone.

There have been numerous reports of serious pharmacokinetic interactions in the combination of SSRIs and CAs due to inhibition of CYP enzymes by SSRIs (Mitchell, 1997b). The most common clinical complications have been deliriumcardiac arrhhmias and seizures.

In view of the minilllal evidence for the enhanced efficacy of this combinationand its high rate of adverse effects, this combination should rarely be utilised.

Pindolol

In 1994 Artigas et al.rst reported on the effectiveness of auglllenting failed courses of SSRIs with the beta-blocker pindolol. The initiaI report described a rapid response in 9 of the 12 patients so treated. The rationale for using pindolol was thatin addition to its beta-blocking properties pindolol is an antagonist of somato-dendritic 5-HT1a receptors. The theory was that blocking those inhibitory receptors wouk:l further enhance serotonin translllission above and beyond that due to serotonin uptake inhibition alone. This eect of pindolol augmenta n after fai!ure to respond to SSRIs has since been confirmed by a Canadian research group (Blier and Bergeron 1995).

It had been recognised in basic pharmacological studies (Rumero et al. 1997) that the early increase in extraneuronal serotonin levels due to SSRI-induced reuptake inhibition leads to stilllula on of the inhibitory SOlllato­ dendri C 5-HT1a receptorsas weIl as increasing levels at the presumed site of therapeu c ac on in e synaptic cleft. This stilllula on of 5-HT1a receptors produces inhibition of neuronal serotonin translllission which ini aIly predolllinates over the effect of increased synaptic serotonin. However as the SSRI-induced increased serotonin levels graduaIly cause a desensitisa on of the 5HT1a receptorsthis inhibitory effect is eventuaIly lost naIly leading to enhanced serotonin translllission after 1 to 2 weeks.

In view of this it was hypothesised at if 5-HT1a receptors were antagonised early in therapythis could block the ini al inhibitory effect on serotonin flow. Recently, Perez et al. (1997) reported (in a randolllised placebo­ controIled study) the use of pindolol adlllinistered concurrently with fluoxetine frolll the initiation of therapy. They described an increased rate of response in those pa ents treated with pindolol. There was also SOllle sugges on of an earlier response in those on pindolol.

Both strands of pindolol research (auglllenta on of a fai!ed courseρr concurrent COllllllencelllent with SSs) indicate that e combination of pindolol and SSRIs leads to an increased rate of responsiveness. As yetthere have been no placebo-con oIled trials of pindolol auglllenta on.

While research on the efficacy of pindolol is of great excitelllentthere are 10 areas of relllaining uncertainty. Firstthere have been no kine c studies of the effect of pindolol on SSRI levels; second there have been no cOlllparative studies of the relative efficacies of pindolol and lithiulll auglllentation.

Other augmentation or combination sírategies

Moclobemide There have been a slllaIl nUlllber of reports of cOlllbining SSRIs with the reversible selective MAOI _1ll0clobelllide (Hawley et al.1996a1996b). While there has been SOllle suggestion of greater effectiveness with this cOlllbinationthe high rate of adverse effects (and in particular the occurence of the serotonin syndrollle in a nUlllber of patients) precludes this as a safe clinical option (MitcheIl 1997b).

Methylphenidate There has been a tradition of lllethylphenidate auglllenta on of antidepressants in a nUlllber of eastern U.S. seaboard universi psychiatry deparìments (Gwirtslllan et al.1994). Recently, Stoll et al (1996) reported on an open series of patients for whom SSRIs were auglllented with lllethylphenidate. That study claimed such augmenta on to be saferapid and efficacious. There hashoweverbeen concern about both the poten al for addic onand the dura on of effect of such stimulants.

Yohimbine An intriging recent single-blind report described augmenta on of the SSRl fIuvoxamine with the alpha-2 adrenoceptor antagonist yohimbine to be effective in a smaIl number of pa ents (CappieIlo et aI.1995).

CHANGE TO A DIFFERENf ANTIDEPRESSANT

To another SSRI

There have been no controIled trials of this strategy which appears to be widespread in practice. here has only been one prospective study of the switch from one SS to another (Thase et 1997). In that open study of rates of response to fIuoxetine in sertraline non-responders40-70% responded (according to the Hamilton Depression Scale) with the specific rate depending on the previous dose of sertraline. Two retrospective case note reviews (Zarate et aI.1996; Stassen et aI.1993) have reported response rates (as defined by the clinician-rated CGI) of 42% and 51% respectively.

To a tricyclic antidepressant (TCA)

There has been one double-blind study of the switch from SSRIs to TCAs (Peselow et aI.1989)with a response rate of 50%. That studydidhoweverinclude both trea 1ent­ resistant and -intolerant patientssome of whom were switched after only 2 weeks of trea nent. The potenal benefit of such switching has also been suggested by a series of studies from the Danish University Antidepressant Group (DUAG1986; 1990) in which inpa ents with melancholic depression were more likely to respond to the TCA clomipramine than the SSRIs parox ne or citalopram.

To a monoamine oxidase inhibitor (MAOI)

There have been a number of studies demonstrating the effectiveness of the irreversible MAOIs phenelzine and tranylcypromine in patients failing to respond to TCAs (Nolen et aI.1988; Thase et 1992). While there have been no such data for the effectiveness of the irreversible MAOIs in SSRI-resistant patientsthe probability of such an eéct could be extrapolated from the TCA studies. It needs be emphasisedhoweverat because of seriousness of the serotonin syndrome due to the interaction of SSRIs and MAOIsthese antidepressants should not be co-prescribed in close temporal proximity.

To venlafaxine

There are a number of studies suggesting that the serotonin-noradrenaline reuptake inhibitor venlafaxine may be more effective than the SSRIs in both major depression and melancholia (Clerc et aI.1994; Dierick et 1996), though a randomised controIled comparative trial of venlafaxine and SSRIs in SSRI non-responders has yet to be published.

In an open study of weIl-defined treatment-resistant depressed patients (including patients who had failed to respond to SSRIs other antidepressants and ECT) Nierenberg et aI. (1994) reported that 30% to 40% responded to venlafaxine at an average dose of 245 mg.

To miriazapine

Mir zapi is an antagonist at both alpha-adrenergic d po naptic serotonergic ceptors(Preskom1997).here has been a recent repofrom a randornised con olled study of rnirtazapine beir more effecle than fIuoxetine in depressed pa ents with melancholia (presented at SateIlite Symposium 6th Work:l Congress of Biological Psychiatry, Nice 1997 by Nutt). AIthough there have been no formal reports of rnirtazapine in SSRI-resis nt patientsthe report of Nutt wouk:l suggeste potential benet of this strategy.

To other novel therapies

An intriguing recent strand of research in resistant depression has been the use of medica ons which reduce steroid synthesis. The rationale for such therapy has been that the increased cortisol Ievels found in severe depression may be contributing to maintenance of the depressive process. In 1995, Anand et al. reported the effectiveness of the antiglucocorticoid ketoconazole in a patient with refractory depression. In that same year, Ghadirian et al. (1995) reported the benefit of medications that decrease corticosteroid biosynthesis in 20 refractory depressed pa ents. Of 17 patients who completed the treatm 13 pa ents responded including five with psychotic depression. ParadoxicaIlyArana et aI. (1995) and Dinan et aI. (1997) have reported an antidepressant effect of the synthetic steroid dexamethasone.

Other less orthodox therapies recently reported for refractory depression include buprenorphine (Bodkin et aI. 1995) and intrathecal protirelin (MarangeIl et 1997)

ELECTROCONVULSIVE THERAPY (ECT), TRANSCRANIAL MAGNEfIC STIMULATION (TMS) OR PSYCHOSURGERY

Probably the best studied treatment for an depres nt­ resistant depression is stiIl ECT.It is now clear that 5 15 most effective when suprathreshold doses are employed (Sackeim et aI.1993). High dose unilateral treatment is more eective than low dose unilateral therapy while bilateral ECT is even more eéctive though it produces more cogni ve impai nent. A recent study of the relationship between resistance to antidepressants and response to ECT (Prudic et aI. 1996) found a poorer response to ECT in TCA non-responderswhile resistance to SSRIs did not predict re onse to ECT at aIl.he authors speculated at these findings may suggest at there are shared mechanisms of action between ECT and TCAs.

In the last few yearsa number of groups have reported the effectiven 5 of repetitive transcranial magnetic stimulation (rTMS) in patents with refractory depression (eg. George et aI.1995; Pascual-Leone et aI.1996). As yet there has only been one published sham-controlled study (Pascaule-Leone et aI.1996). While this technique appears very promising, there have been no comparative studies against ECT.

Though rarely used in contemporary psychiatric prac ce there is stiII a role (albeit smaIl) for stereotactic psychosurgery in those who do not respond to other therapies. Approximately two-thirds of patients benefit to some extent from this treatment (Bridges et al., 1994).

PSYCHOTIC DEPRESSION

The most effective treatment for psychotic depression is ECT (Parker et al., 1992). Bilateral ECT is more effective than either unilateral ECT, combined antidepressants and antipsychotics, or antidepressants alone. There have been some occasional reports of the effectiveness of antidepressants (such as fluvoxamine) alone in psychotic depression (Delgado et al., 1988). A recent case history described the benefit of clozapine in a psychotic depressed patient refractory to ECT and a large number of antidepressant medications (Dassa et al., 1993).

ATYPICAL DEPRESSION

Atypical depression has been defined by Liebowitz et al. (1988) as characterised by hypersomnia, hyperphagia, increased interpersonal sensitivity to rejection and reactivity of mood. This form of depression responds better to irreversible MAOis than TCAs (Quitkin et al., 1991). While SSRIs have been reported to be effective for atypical depression in clinical practice, there have been no formal studies of this.

CONCLUSION

In the late 1990s there are a wide range of somatic treatment strategies available for patients with resistant depression. The main alternatives are either augmenting, switching or combining antidepressants; or ECT. Unfortunately, in spite of this wealth of options, there is currently no research data to guide the clinician in which order these should be used. The clinician should also bear in mind that many patients do eventually remit or improve due to the natural course of their illness, even after years of unsuccessful treatment (Stoudemire et al., 1993).

ACKNOWLEDGEMENTS

The writing of this paper was supported in part by the Australian National Health and Medical Research Council Mood Disorders Unit Program Grant (no. 953208). The author is grateful to Mrs. Kerrie Eyers and Mrs. Zora Vuckovic for their assistance in the preparation of this article.

REFERENCES

Alexopoulos, G.S., Meyers, B.S., Young, R C., Kakuma, T., Silbersweig, D., Charlson, M. (1997) Clinically defined vascular depression. American Journal of Psychiatry, 154: 562- 565.

Anand, A., Malison, R , McDougle, C.J., Price, L.H. (1995) Antiglucocorticoid treatment of refractory depression with ketoconazole: a case report. Biological Psychiatry, 37: 338- 340.

Arana, G.W., Santos, AB., Laraia, M.T. (1995) Dexamethasone for the treatment of depression: a randomised, placebo- controlled, double-blind trial. American Journal of Psychiatry, 152: 265-267.

Aronson, R, Off man, H.J., Joffe, RT., Naylor, C.D. (1996) Triiodothyronine augmentation in the treatment of refractory depression: a meta-analysis. Archives of General Psychiatry, 53: 842-848.

Artigas, F., Perez, V., Alvarez, E. (1994) Pindolol induces a rapid improvement of depressed patients treated with serotonin reuptake inhibitors. Archives of General Psychiatry, 51: 248- 251.

Austin, M.P., Souza, F.G., Goodwin, G.M. (1992) Lithium augmentation in antidepressant-resistant patients: a quantitative analysis. British Journal of Psychiatry, 159: 510-514.

Blier, P., Bergeron, R (1995) Effectiveness of pindolol with selected antidepressant drugs in the treatment of major depression. Journal of Clinical Psychopharmacology, 15: 217- 222.

Bodkin, J.A., Zomberg, G.L., Lukas, S.E., Cole, J.O. (1995) Buprenorphine treatment of refractory depression. Journal of Clinical Psychopharmacology, 15: 49-57.

Bonner, D., Howard, R (1995) Clinical characteristics of resistant depression in the elderly. International Journal of Geriatric Psychiatry, 10: 1023-1027.

Boyer, W.F., Feighner, J.P. (1991) The efficacy of selective serotonin re-uptake inhibitors in depression. In: Perspectives in Psychiatry. Vol 1. Selective serotonin re-uptake inhibitors. (Eds. Feighner, J.P., Boyer, W.F.) Wiley & Sons, Chichester, England.

Bridges, P.K , Bartlett, J.R , Hale, A.S., Poynton, A.M., Malizia, A.L., Hodgkiss, A.O. (1994) Psychosurgery: Stereotactic subcaudate tractotomy. An indispensable treatment. British Journal of Psychiatry, 165: 599-611.

Brown, F.W., Lewine, RJ., Hudgins, P.A., Risch, S.C. (1992) White matter hyperintensity signals in psychiatric and nonpsychiatric subjects. American Journal of Psychiatry, 149: 620-625.

Cappiello, A., McDougle, C.J., Malison, RT., Heninger, G.R , Price, L.H. (1995) Yohimbine augmentation of fluvoxamine in refractory depression: a single-blind study. Biological Psychiatry, 38: 765-767.

Clerc, G.E., Ruimy, P., Verdeau-Pailles, J. (1994) A double-blind comparison of venlafaxine and fluoxetine in patients hospitalised for major depression and melancholia. International Clinical Psychopharmacology, 9: 138-143.

Danish University Antidepressant Group (1986) Citalopram: clinical effect profile in comparison with clomipramine. A controlled multicentre study. Psychopharmacology, 90: 141-138.

Danish University Antidepressant Group (1990) Paroxetine: a selective serotonin reuptake inhibitor showing better tolerance, but weaker antidepressant effect than clomipramine in a controlled multicentre study. Journal of Affective Disorders, 18: 289-299.

Dassa, D., Kaladjian, A., Zarin, J.M. , Giudicelli, S. (1993) Clozapine in the treatment of psychotic refractory depression. British Journal of Psychiatry, 163: 822-824.

De Montigny, C., Grunberg, F., Mayer, A., Deschenes, J.P. (1981) Lithium induces rapid relief of depression in tricyclic antidepressant drug non-responders. British Journal of Psychiatry, 138: 252-256.

De Montigny, C., Cournoyer, G., Morissette, R , Langlois, R , Caille, G. (1983) Lithium carbonate addition in tricyclic antidepressant-resistant unipolar depression. Archives of General Psychiatry40: 1327-1334.

De MontignyC. (1994) Lithium addition in treatment-resistant depression. International Clinical Psychopharmacology, 19: 31-35

Delgado P . Price L.H. Charney 0.5. (1988) Efficacy of fluvoxamine in treatment-refractory depression. Journal of Affective Disorders, 15: 55-60.

Dierick, M.RavizzaL.RealiniR.MartinA.(1996) A double­ blind comparison of venlafaxine and fluoxetine for treatment of major depression in outpatients. Progress in Neurpsycho pharmacology and Biological Psνchiatry 20: 57-71.

Dinan, T.G., Lavelle, E., Cooney, J. Burnet, F. Scott, L., Dash, A., Thakore, J., Berti, C.(1997) Dexamethasone augmentation in treatment-resistant depression. Acta Psychiatrica Scandinavica 95: 58-61.

Dornseif, B.E., Dunlop, S. R., Potvin, J.H. (1989) Effect of dose escalation after low-dose fluoxetine thera. Psychopharma­ cological Bulletin, 25: 71-79

Fava, G.A., Savron, G., Grandi, S. Rafanelli, C.(1997) Cognitive behavioural management of drug-resistant major depressive disorder. Journal of Clinical Psychiatry, 58: 278-282.

Fava, M., Rosenbaur, J.F., Cohen, L., Reiter, S., McCarthy, M. Steingard, R., ClancyK. (1992) High-dose fluoxetine in the treatment of depressed patients not responsive to a standard dose of fluoxetine. Journal of Affective Disorders, 25: 229- 234

Fava M. Rosenbaur J.F., McGrath P.J., Stewart J.W., Amsterdar, J.D., Quitkin, F.M. (1994) Lithium and tricyclic augmentation of fluoxetine treatment for resistant major depression: a double-blindcontrolled study. American Journal of Psychiatry, 151: 1372-1374.

Fawccett, J., Kravitz, H.M. (1985) Treatment refractory depression. In: Common treatment problems in depression. (Ed Schatzberg AJ.) American Psychiatric Press, Washington, OC

Flint A.J., Rifat, S.L.(1994) A prospective study of lithium augmentation in antidepressant-resistant geriatric depression. Journal of Clinical Psνchopharmacolo, 14: 353-356.

George, M.S., Wassermann, E.M., WilliamsW. A., Callahan, A.,  Ketler, T.A., Basser, P., Halletl, M. Post, R.M. (1995) Daily repetitive transcranial magnetic stimulation (rTM5) improves mood in depression. Neuroreport, 6: 1853-1856.

Ghadirian, A.M., Engelsmann, F., Dhar, V., Filipini, D., Keller, R., Chouinard, G., Pearson Murphy, B. E. (1995) The psychotropic effects of inhibitors of steroid biosynthesis in depressed patients refradory to treatment. Biological Psychiatry, 37: 369-375

Gwirtsman, H.E., Szuba, M.P., Toren, L., Feist, M. (1994) The antidepressant response to tricyclics in major depressives is accelerated with adjunctive use of methylphenidate Psychopharmacology Bulletin, 30: 157-164

Hawleyc.J., Ratnam, S., Patlinson, H.A. (1996a) Safety and tolerability of combined treatment with moclobemide and 55RIs: a preliminary study of 19 patients. Journal of Psychopharmacology, 10: 241-245

Hawley C. J.,  Quick, S.J., Ratnam S. (1996b) Safety and tolerability of combined treatment with moclobemide and SSRIs: a systematic study of 50 patients. International Clinical Psychopharmacology 11: 187-191.

Hickie, I., Scott, E. Mitchell P., Wilhelm. K., Austin MP., Bennet, B. (1995) Subcortical hyperintensities on magnetic resonance imaging: clinical correlates and prognostic significance in patients with severe depression. Biological Psychiaty, 37: 151-160

Hickie, I., Bennet, B., Mitchell, P., Wilhelm, K., Orlay, W.(1996) Clinical and subclinical hypothyroidism in patients with chronic and treatment-resistant depression. Australian and New Zealand Journal of Psychiatry, 30: 246-252.

Hickie, I., Scott, E. Wilhelm, K., Brodatym H. (1997) Subcortical hyperintensities on magnetic resonance imaging in patients with severe depression: a longitudinal evaluation. Biological Psychiatry, 42: 367-374

Howland R.H. (1993) Thyroid dysfunction in refractory depression: Implications of pathophysiology and treatment. Joumal of Clinical Psychiatry, 54: 47-54.

Joffe, R.T.,  Singer, W. (1990) comparison of triiodothyronine and thyroxine in the poteniation of tricyclic antidepressants Psνchiatry Research, 32: 241-251

Joffe, R.T.,  Singer, W., Levitt, A.J., MacOonald, C. (1993a) A placebo-controlled comparison of lithium and triiodothyronine augmentation of tricyclic antidepressants in unipolar refractory depression. Archives of General Psychiatry, 50: 387-393.

Joffe, R.T.,  Singer, W., Bagby, R.M., MacOonald, C.(1993b) Predictors of response to lithium and triiodothyronine augmentation of antidepressants in tricyclic non-responders. British Journal of Psychiatry, 163: 574-578.

Joffe, R.T., Levitt, A.J., Sokolov, S.T.H., Young, L.T. (1996) Response to an open trial of a second 55Rl in major depression Joumal of Clinical Psychiat 57:114-115.

Katona, C.L.E,  Abou-Saleh, M.T., Harrison, D.A., Nairac, B.A, Edwards O. R.L. Lock T. Burns R.A. Robertson M.M (1995) Placebo-controlled trial of lithium augmentation of fluoxetine and lofepramine. British Journal of Psychiatry 166: 80-86

Keller M.B., Klerman, G.L., Lavori, P.W., Coryell, W., Endicott, J.TaylorJ. (1984) Long-term outcome of episodes of major depression: clinical and public health significance. JAMA, 252: 788-792

KellerM.B., Lavori, P.W., Mueller, T.I., Endicotl, J., Coryell, W., Hirschfeld,  R.M.A. Shea T. (1992) Time to recovery chronicityand levels of psychopathology in major depression: A 5-year prospective follow-up of 431 subjeds. Archives of General Psychiatry, 49: 809-816.

Kiloh L.G.Andrews G. Neilson M. (1988) The long-term outcome of depressive illness. British Journal of Psychiatry 153: 752-757.

KrishnanK.R.R.HaysJ.c.BlazerO.G. (1997) MRl-defined vascular depression. American Journal of Psychiatry, 154: 497-501

Lee A.5. Murray R.M. (1988) The long-term outcome of Maudsley depressives. British Journal of Psychiat,  153: 741-751

Liebowitz M. R. Quitkin F.M. 5tewart J .w. (1988) Antidepressant specificity in atypical depression. Archives of General Psychiatry, 45: 128-137.

Marangell L.B., George M.S., Callahan A.M., Ketter,A., Pazzaglia, P.J. L'Herrou, T.A., Leverich G.S., Post R.M. (1997) Effects of intrathecal thyrotropin-releasing hormone (Protirelin) in refractory depressed patients. Archives of General Psychiatry, 54: 214-222.

Mayberg, H.S. Brannan, S.K, MahurinR.K. Jerabek P. A. Brickman, J.S., Tekell, J.L., Silva, J.A., Meginnis, S., Glass, T.G., Martin c.c. Fox P.T. (1997) Cingulate function in depression: a potential predictor of treatment response Neoreport, 8: 1057-1061

Mitchell P.B. (1994) Depression and other physical illnesses. Joumal of the Australian National Association for Mental Health, 6: 18-23.

Mitchell P.B. (1997a) Managing depression in a community setting. Medical Journal of Australia, 167: 383-388.

MitchellP.B. (1997b) Drug interactions of clinical significance with selective serotonin re-uptake inhibitors. Drug Safety17: 390 406

NelsonJ. c.MazureC.M.BowersM.B. (1991) A preliminary open study of the combination of fluoxetine and desipramine for rapid treatment of major depression. Archives of General Psychiatry48: 303-307

NelsonJ.c., Mazure, C.M., Jatlow, P.I. (1994) Characteristics of desipramine-refractory depression. Joumal of Clinical Psychiatry,55: 12-19.

Nelson J.c., Price L.H. (1995) Lithium or desipramine augmentation of fluoxetine treatment. American Journal of Psychiatry,152: 1538-1539.

Nierenberg, A.,  Feighner, J.P., Rudolph, R., Cole, J.O., Sullivan (1994) Venlafaxine for treatment-resistant unipolar depression. Joumal of Clinical Psychopharmacology,4: 419-423

Nierenberg, A.A., McLean, N.E., Alpert, J.E., Worhington, J.J. Rosenbaum J.F. Fava M. (1995) Early non-response to fluoxetine as a predictor of 8-week outcome. American Joumal of Psychiat, 152: 1500-1503.

Nolan W.A. Van De PutteJ.J. Dijken W.A. Kamp J.S., Blanaar, B. A., Kramer, H.J., Haffmans, J. (1988) Treatment strategy in depression: 11: MAO inhibitors in depression resistant to cyclic antidepressants: two controlled crossover studies with tranylcypromine vs L-5-hydroxytryptophan and nomifensine Acta Psychiatrica Scandinavica, 78: 676-683.

Parker, G., Roy, K., Hadzi-Pavlovic, D., Pedic, F. (1992) Psychotic (delusional) depression: a meta-analysis of physical treatments. Joumal of Affective Disorders, 24: 17-24

Pascual-Leone A. Rubio B. Pallardo Catala D. (1996) Beneficial effect of rapid-rate transcranial magnetic stimulation of the left dorsolateral prefrontal cortex in drug-resistant depression. Lancet, 348: 233-238

Peselow, E.D., Filippi, A.M., Goodnick, P., Barouche, F., Fieve, R.R.(1989) The short and long-term efficacy of paroxetine HCl: B. data from a double-blind crossover study and from a year­ long trial vs. imipramine and placebo. Psychopharmaco­ logical Bulletin, 25: 272-276

Perez, V., Gilaberte, I., Farles, D., Alvarez, E., Artigas, F. (1997) Randomiseddouble-blindplacebo-controlled trial of pindolol in combination with fluoxetine antidepressant t atment. The Lancet, 349: 1594-1597.

Preskorn, S.H. (1997) Selection of an antidepressant: mirtazapine. Joumal of Clinical Psychiatry, 58: 3-8.

Price, L.H., Heninger, G.R. (1994) Lithium in the treatment of mood disorders. New England Joumal of Medicine, 331: 591-598

Prudic, J., Haske. R.F., Mulsant, B., Malone, K.M., Pettinati, H.M., Stephens, S., Greenberg, R., Rifas, S.L. Sackeim, H.A. (1996) Resistance to antidepressant medications and short-term clinical response to ECT. American Joumal of Psychiatry, 153: 985-992.

Quitkin, F.M., Harrison, W., Stewart,  J.W. (1991) Response to phenelzine and imipramine in placebo non-responders with atypical depression. Archives of General Psychiatry 48: 319-323

Quitkin, F.M., McGrath, P.J., Stewart, J.W., Ocepek-Welikson, K., Taylor, P.B., Nunes. E., Deliyannides, D., Agosti, V., Donovan,  S.J. Petkova, E., Klein, D.F. (1996) Chronological milestones to guide drug change: When should clinicians switch antidepressants? Archives of General Psychiatry, 53: 785 792.

Romero, L., Ma Casanovas, J.., Hervás, Cortes, R.Artigas, F. (1997) 5trategies to optimize the antidepressant acìion of selecìive serotonin reuptake inhibitors. In Antidepressants: New Pharmacological Strategies (Ed. Phil Skolmik) National Institute of Health Bethesda MC. Humana Press USA

5ackeim H.A. Prudic J. Devanand D.P. Kiersky J.E. Fitzsimons, L., Moody, B.J., McElhiney, M.C., Coleman, E.A., Settembrino J.M. (1993) Effects of stimulus intensity and elecìrode placement on the efficacy and cognitive effects of electroconvulsive therapy. New England Journal of Medicine328: 839-846

Salama, A.A., Shafey, M. (1989) A case of severe lithium toxicity induced by combined fluoxetine and lithium carbonate. American Journal of Psychiat146: 278

Shergill, S.S., Katona, C.L.E. (1"997) Pharmacological choices after one antidepressant fails: a survey of UK psychiatrists. Joumal of Affective Disorders43: 19-25.

Simpson, S., Talbot, P.R., Snowden, J. Neary, D. (1997) 5ubcortical vascular disease in elderly patients with treatment resistant depression. Joumal of Neurology, Neurosurgery and Psychiatry, 62: 196-203.

Stoll., A.L.Pillay, S.S., Diamond, L., Workum, S.B., Cole, J.O. (1996) Methylphenidate augmentation of serotonin selecìive reuptake inhibitors: ! case series. Joumal of Clinical Psychiatry, 57: 72-76.

Stoudemire, A., Hill, C.D., Morris, R., Lewison, B.J. (1993) Long­ term outcome of treatment-resistant depression in older adults. American Journal of Psychiatry, 150: 1539-1540

Thase, M.E., Mallinger,A.G.Frank, E.Hamer, T., Kupfer, D.J. (1992) Treatment of imipramine resistant recurrent depression: III. Efficacy of monoamine oxidase inhibitor. Joumal of Clinical Psychiatry 51: 5-11.

Thase M.E. Howland R.H. (1994) Refractory depression: management. relevance of psychosical factors and therapies. Psychiatric Annals, 24: 232-240. 

Thase M.E., Blomgren, S.L., Birke, M.A., Apter, J.T., Tepner, R.G. (1997) Fluoxetine treatment of patients with major depressive disorder who failed initial treatment with sertraline. Journal of Clinical Psychiatry, 58: 16-21.

Wihelm, K., Mitchell, P., Boyce, P., Hickie, I., Brodaty, H., Austin, M-P., Parker, G (1994) Treatment resistant depression in an Australian context I: The utility of the term and approaches to management. Australian and New Zealand Journal of  Psychiatric Psνchiatry, 28: 14-22.

Zajecka, J.M., Jeffries, H., Fawcett, J. (1995) The efficacy of fluoxetine combined with he_.terocydic antidepressant in treatment-resistant depression: a retrospective analysis. Journal of Clinical Psychiatry, 56: 338-343.

Zarate, C.A., Kando, J.C., Tohen, M.K., Cole, J.O. (1996) Does intolerance or lack of response with fluoxetine predict the same will happen with sertraline? Journal of Clinical Psychiatry, 57: 67-71. 

Philip B. MitchelI MB BS, MD, FRANZCP, FRCPsych Associate Professor School of PsychiatryUniversity of NSW; Administrative Director, Mood Disorders  Unit, Prince Henry Hospital, Little Bay, Sydney, NSW  2036, Australia.