Year: 2023
Type: article
Abstract: As admirably detailed by McIntyre et al1, most definitions of treatment-resistant depression (TRD) weight failure to respond to a set of antidepressant medications and other physical treatment options. Is this the optimal paradigm when major depressive episodes may be caused not only by biological factors, but also by social and psychological ones? Wouldn't we expect, for example, that a woman depressed due to abuse by a coercive husband, or a man whose procrastinating perfectionism leads his manager to heap opprobrium and deep depression on him, fails to respond to two or three antidepressants of different classes and at appropriate doses? If the incorrect treatment paradigm is employed in the latter instances (i.e., antidepressant medication rather than respective social and psychological interventions being prioritized) is non-response better viewed as "treatment failure" (and accorded TRD status) or more as "paradigm failure"? Similarly, delivery of multiple psychotherapy approaches alone to an individual with severe melancholic depression might generate TRD status but more correctly reflect paradigm failure. In concept and operation, most TRD models and definitions constrain the heterogeneity of depressive disorders and then effect a Procrustean management model weighted to physical treatments. McIntyre et al1 consider an alternate "difficult-to-treat" framework, which they regard as relying more on a biopsychosocial approach. Such an approach theoretically allows alternative definitional and management strategies. The authors note that psychotherapeutic interventions are recommended as first-line strategies for those with mild or moderately severe depressions. This might allow TRD definition to be weighted by baseline depression severity. For example, for those with "severe" depression, TRD status would be assigned by failure to respond to a sequential and operationalized set of drug and physical treatment strategies, thus weighting their management to biological treatments, while those with less severe depression would achieve TRD status by failure to respond to cogent psychotherapies. However, immediate concerns about such a model include severity of depression not being linked with depressive subtype, while "failure to respond" to a cogent psychotherapy would be limited by judging of "cogency". A second option would emphasize a biopsychosocial definition (e.g., failure to respond to a therapeutic intervention salient to cause – including psychological and social determinants and not biological ones alone) and weight a subtyping model as against a severity-based one. Such a model would aim for TRD status to be operationalized in relation to specified biologically-weighted depressive conditions (i.e., melancholia, psychotic depression, bipolar I and II depression, depression caused by medical conditions) and a set of residual non-melancholic depressive conditions with presumed social and/or psychological causes. Differing type-specific criteria for according TRD status, and logical type-specific sequential management strategies, would be developed. Rather than seeking a TRD definition that has universal application, do we not need one that has disorder-specific nuances as is observable in general medicine? Turning to narrower issues, McIntyre et al importantly note that some people meeting TRD status do so as a consequence of being "rapid metabolizers". Such a state affects up to 30% of some races, but is a TRD determinant that may be overlooked by practitioners. The clinical clues are that the individual rarely has side effects from any antidepressant medication, is seemingly unaffected by moderate amounts of alcohol, and fails to obtain analgesic benefit from paracetamol. In my pursuit of this possibility, I do not find genomic testing informative but, less anecdotally, note a report by De Leon et al2 quantifying that up to 80% of ultra-metabolizers are missed by genetic testing. I generally initiate a tricyclic (TCA), increase its dose to 150 mg/day and then obtain a "tricyclic level", finding serum TCA level results to have high sensitivity and specificity3. The impact on management is that some people whose TRD status is so established then become "responders" when taking a higher dose of the TCA (although TCA metabolites may produce serious side effects). Further, some people with rapid metabolizing status can be "converted" to normal metabolizing status by adding paroxetine or fluoxetine, with De Leon et al2 noting that such a strategy increases the concentration of the active drug and reduces the concentration of the hydroxylated metabolites. Some TRD criteria include failure to respond to two or more antidepressants of differing classes – a non-specific model which allows random progression from narrow-action to broader-action antidepressant classes, but also allows the converse and other sequences. Unspecified progression might be appropriate if all antidepressant classes have comparable efficacy, which is the general finding in relation to treating "major depression". However, broad-action tricyclics have been quantified as superior to narrow-action selective serotonin reuptake inhibitors (SSRIs) in those with melancholic depression, with differential effectiveness increasing with older age of the patient4, while the dual-action antidepressants appear to have intermediate efficacy. Any such gradient might then logically argue for a TRD management model whereby those with melancholia not responding to an SSRI then receive a broader-action antidepressant – perhaps a dual-action antidepressant, followed by a TCA, and then possibly a monoamine oxidase inhibitor (MAOI) – with MAOIs long positioned as strong candidates for managing treatment-resistant melancholia5. While I have previously employed MAOIs when narrow- or dual-action antidepressants failed in those with a melancholic depression, I now trial augmentation of a baseline antidepressant drug with a psychostimulant (e.g., methylphenidate or dexamphetamine) prior to any trialling of an MAOI – in light of dietary and other concerns related to the MAOIs. While earlier meta-analyses of psychostimulants failed to indicate distinct efficacy, a more recent meta-analysis by McIntyre et al6 supported their efficacy, but with these authors suggesting that their benefits might operate across "select domains". I view melancholia as being one such key domain, finding such psychostimulant augmentation commonly beneficial in those with melancholia who have failed to respond to several antidepressants, and with tolerance rarely emerging. A recent review7 summarized two previous reports of people with unipolar and bipolar melancholic depression meeting common TRD criteria (subjects having received a mean of five previous antidepressants in both samples, and seven and eight psychotropic drugs in total across the two samples). Overall, some 20% reported the psychostimulant augmentation as "very effective", while an additional one-half reported the strategy as "somewhat effective". Some 45% described the psychostimulant as the best or equal best to previously prescribed antidepressants, and 48% reported a sustained benefit. Such effectiveness rates in those with TRD are impressive, and argue for such a strategy being trialled – in those with melancholia – before more demanding treatments such as electroconvulsive therapy are enlisted. While major depression is a diagnostic entity, it is not a clinical entity8. To the extent that TRD status is defined by failure of people with major depression to respond to multiple antidepressants, then TRD is also not an entity. Then, it must be expected that its specific and fuzzy-set constituent conditions will show differential response gradients to differing interventions. McIntyre et al provide an extraordinarily rich review of current definitions and management options, and thus set up a base camp for future exploratory studies. These will benefit from the recognition that heterogeneity reigns and needs to be constrained.
Source: World Psychiatry
DOI: 10.1002/wps.21135
Author: Gordon Parker
Institution: UNSW Sydney
Language: English
Subfield: Pharmacology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4386765456