How Best to Manage Treatment-Resistant Depression

“When I use a word…it means just what I choose it to mean—neither more nor less.” Humpty Dumpty¹

Framing TRD

In the past 2 decades, treatment-resistant depression (TRD) has commonly been viewed as capturing scenarios when a patient meeting criteria for major depressive disorder (MDD) fails to respond to nuanced antidepressant medication administration. While variably defined à la Humpty Dumpty—with Brown et al² identifying 155 TRD definitions—McIntyre et al³ noted that both the US Food and Drug Administration and the European Medicines Agency have “adopted the most widely used definition of TRD (ie, inadequate response to a minimum of [2] antidepressants despite adequacy of the treatment trial and adherence to treatment).”

Limitations to the TRD Framework

Such a framework invites 2 caveats. The base condition of MDD is commonly viewed as an entity but is, I suggest, no more a homogeneous entity than (say) “major breathlessness” or, at the disease level, “major cancer.” Entity status risks inviting singular linear-pathway management models for TRD and thus contrasts with a cancer model, whereby differing cancers generally invite quite differing and varying sequences of treatments.

Further, the common positioning of MDD as a “disease” leads to TRD management models weighting physical treatments such as switching or combining antidepressants, augmenting drugs, electroconvulsive therapy (ECT), other neurostimulatory strategies, and ketamine.

Consider 2 examples illustrating limitations to such modeling: first, a mother who has persistent clinical depression as a consequence of a physically abusive and demeaning husband but she (and her children) cannot leave him, as he manages the family money; second, a lawyer with a profoundly perfectionist personality who is criticized publicly by his superior, with such shaming inducing a prolonged depressive state. Both meet the criteria for MDD. Despite their managing doctors prescribing 3 antidepressants of different classes and at adequate doses for many weeks, their depressive episodes are nonresponsive. Are we not surprised? Yes, they meet the common definitional criteria for TRD, but, as their depression has no primary biological basis, and as an antidepressant medication (initially or subsequently) is unlikely to be therapeutic, why should such scenarios be deemed TRD as opposed to reflecting limitations in treatment modality selection?

History

The concept of TRD and an attendant staging model was introduced by Thase and Rush in 19974, but with their definition capturing patients with depression who were antidepressant drug nonresponders or unable to tolerate such drugs, and thus suggesting a weighting to depressive conditions with biological rather than social or psychological determinants. Subsequently, operational management models shifted to impose a base diagnosis of MDD as a disease category and, thus, predictably, the weighting of biological treatments. Psychotherapeutic interventions—which may be salient initial treatments or augmenting treatments in some TRD scenarios—are generally either not included or positioned as adjunctive. McIntyre et al considered, however, that they may be a preferred treatment option over pharmacotherapy “amongst persons with lived depressive experience,” may assist residual symptoms and comorbidities, may synergize with antidepressant medications, and may be salient for those with persistent depression and trauma—but then added that “data supporting them as standalone interventions in TRD are limited.”³

Quo Vadis, Clinicians?

If MDD is a nonspecific diagnosis capturing a heterogeneous set of clinical depressive conditions with variable biological, psychological, and/or social origins, then all its constituent conditions have the potential to be resistant to treatment, whether reflecting intrinsic or application limitations of the therapy or a paradigm failure in choosing management strategies lacking salience for that condition.

Management Suggestions

Allowing that clinical depressive conditions comprise some specific biological depressive conditions (ie, psychotic and melancholic depression) and a residual heterogeneous “fuzzy set” of nonmelancholic conditions having psychological and social determinants, then sequencing models of prioritized management options can be assembled for each—and such a model is now set out.⁵

For psychotic depression, the evidence favors the prioritization of a combination of an antidepressant drug and an antipsychotic drug as superior to each alone and comparable with ECT, although the latter would generally not be first ranked in light of the risk of cognitive impairment. Failure to respond to that 2-stage model would warrant a diagnosis of TRD and consideration of all physical treatment options noted earlier.

For melancholia, the evidence favors an antidepressant drug over nondrug therapies, as well as favoring a broad-spectrum antidepressant drug (such as a tricyclic antidepressant [TCA]) over a narrow-action drug (such as a selective serotonin reuptake inhibitor [SSRI]), although the latter might reasonably be first trialed. Here, the operative definition of TRD may well hold sway—allowing that there might be many reasons for any such failure to respond (eg, inadequate dosing, SSRIs being narrow action antidepressants, the patient being a rapid metabolizer, an undetected bipolar condition in play). While next-stage recommendations generally favor lithium augmentation or several second-generation antipsychotics, I offer 2 alternatives weighting dopaminergic transmission. Firstly, olanzapine augmentation of the base antidepressant for 1 to 2 weeks only, reflecting a “jump leads to the battery model” and thus not necessitating an ongoing role.⁶ Secondly, while I long judged trialing a monoamine oxidase inhibitor (MAOI) if a TCA failed to be the best course of action, I now instead augment the base antidepressant with a stimulant drug such as methylphenidate.⁷ Other options considered by McIntyre (for TRD in general) also come into play. These include combining certain antidepressants, using other dopaminergic antipsychotic drugs as augmenting agents, neurostimulatory treatments (eg, ECT, repetitive transcranial magnetic stimulation, direct current stimulation), and ketamine/esketamine.

For some nonmelancholic depressive conditions, TRD may be an operative diagnosis when a seemingly salient (nondrug) therapy has failed, subject to it being provided by a skilled practitioner. Thus, for situational (stress-induced) depressive conditions, the clinician might provide strategies for the patient to neutralize or overcome the stressor and/or assist the individual to come to terms with it. For those whose personality style is the primary driver of the depression, then a psychotherapy might be the preferred primary treatment—and which might be (say) “supportive psychotherapy” in some instances, or a psychotherapy with specific capacities—for example, cognitive behavioral therapy if faulty cognitive schema are operative, or dialectical behavior therapy if the individual’s depression is driven by a borderline personality style. For those whose depression is driven by overwhelming shyness and who are affected by the consequences, behavioral therapy (such as assertiveness training) might be the most appropriate first-line treatment. For such nonmelancholic states, condition-specific recommendations for TRD await development.

Recommendations favoring medication over nonmedication strategies remain unsettled for some nonmelancholic depressive episodes. For example, individuals whose depression has been driven by high trait internalizing or externalizing anxiety may benefit from SSRI antidepressants (as SSRIs mute anxious worrying), but they may also benefit from nondrug anxiolytic strategies (eg, relaxation training, hypnotherapy, psychotherapy).

TRD or DTTD?

For the nonmelancholic depressive scenarios, imposing a point where TRD might be an operative descriptor is clearly problematic when blocks to treatment responsiveness may reflect multiple factors, including the choice of therapy, background factors in the patient’s life, or the therapist providing suboptimal therapy. Here, the construct of “difficult-to-treat depression” (DTTD), as introduced by Rush et al,⁸ is a more useful descriptor or sobriquet and strikes me as having cogency. They so judged this heuristic’s advantage over TRD as it sought to overcome limitations to TRD and “stimulate the timely identification and personalized management of patients for whom our current treatments…are unlikely to either initiate or sustain symptomatic remission.”

Concluding Thoughts

If we are to improve our management of TRD, we first need to respect rather than constrain the heterogeneity of substantive depressive disorders, and thus not generate sequential management options on a base diagnosis of major depression alone. For some conditions (eg, psychotic depression, melancholic depression), operationalizing cutoff criteria and sequential management options for TRD strike me as achievable. For the residual nonmelancholic depressive disorders, we could also proceed with defining their most substantive conditions and developing condition-specific sequencing options, but respect that the current commonly nominated neurostimulatory treatments for TRD are quite unlikely to have salience for these conditions. For most of the residual nonmelancholic conditions, the heuristic DTTD is likely to have greater cogency than invoking TRD.

Dr Parker is an emeritus professor at the University of New South Wales in Sydney, Australia.

References

1. Carroll L. Lewis Carroll’s Alice: Alice’s Adventures in Wonderland and Through the Looking Glass – A Special Centenary Edition. MacMillan Children’s Books; 1998.

2. Brown S, Rittenbach K, Cheung S, McKean G, MacMaster FP, Clement F. Current and common definitions of treatment-resistant depression: findings from a systematic review and quantitative interviews. Can J Psychiatry. 2019;64(6):380-387.

3. McIntyre RS, Alsuwaidan M, Baune BT, et al. Treatment-resistant depression: definition, prevalence, detection, management, and investigational interventions. World Psychiatry. 2023;22(3):394-412.

4. Thase ME, Rush AJ. When at first you don’t succeed: sequential strategies of antidepressant nonresponders. J Clin Psychiatry. 1997;58(suppl 13):23-29.

5. Parker G. A revisionist model for treatment-resistant and difficult-to-treat depression. Aust N Z J Psychiatry. 2024;58(6):460-466.

6. Parker G. Olanzapine augmentation in the treatment of melancholia: the trajectory of improvement in rapid responders. Int Clin Psychopharmacol. 2002;17(2):87-89.

7. Parker G, Brotchie H. Do the old psychostimulant drugs have a role in managing treatment-resistant depression? Acta Psychiatr Scand. 2010;121(4):308-314.

8. Rush AJ, Aaronson ST, Demyttenaere K. Difficult-to-treat depression: a clinical and research roadmap for when remission is elusive. Aust N Z J Psychiatry. 2019;53(2):109-118.