Treatment-resistant depressive disorders: The when,how and what of augmentation therapy

Despite our best efforts, a significant proportion of patients suffering from depressive disorders do not fully recover from their illness. Established practice is to confirm the diagnosis and consider the presence of other mental and physical health problems (including substance abuse) and to modify treatment accordingly (Malhi et al., 2018). Before instituting any changes, it is important to verify that the treatment regimen is being adhered to as prescribed and to determine whether the use of psychotherapy, sleep management and lifestyle enhancements are adequate and being appropriately utilised. Typically separate trials of at least two antidepressant medications, preferably from different classes, and each having been taken for at least 3 weeks at an optimal dose, are essential before a depressive disorder can be regarded as treatment resistant (Cowen, 2017). However, these criteria bias the diagnosis of treatment-resistant depressive disorders (TRD) towards more biological causal factors and do not take into consideration failure to respond to psychotherapeutic interventions and the invariable contribution of psychosocial stressors.

The management of TRD thought to arise from biological causes includes the prescription of additional antidepressant medication (combination therapy), the administration of agents that may not have antidepressant properties in themselves but facilitate the effects of an antidepressant regime (augmentation therapy) or the use of electroconvulsive therapy (ECT). Ideally, strategy and choice of treatments should be based upon clinical judgement and the consideration of factors such as the prior treatment experience of the treatment-resistant patient and the tolerance and safety of specific treatments. Where feasible, patient preference should also be factored into formulating management, but it is important to note that many options may simply not be possible either because of limited availability/access or because the individual wishes not to pursue a particular therapy pathway.

In terms of pharmacotherapeutic strategies, the augmentation of medication sits alongside switching, and usually, either strategy is preceded by initially increasing the dosage of a partially effective agent. Augmentation may operate through a direct enhancement of the antidepressant in use or it may exert a synergistic pharmacodynamic action (e.g. delaying metabolism or enhancing therapeutic effects). In practice, the term augmentation is also used to describe a separate beneficial impact (e.g. enhancing sleep or relieving psychosis) or a positive effect upon activity (e.g. enhancing energy, motivation or cognitive functions), but technically these are separate effects that are more accurately described as consequences of combination therapy.

Three classes of medications are well established as augmenting agents when tackling TRD and particularly in the face of melancholic symptoms (where the benefits of augmentation are particularly evident), while a number of other agents have varying levels of supportive evidence (see Table 1).

Table 1.

Summary of augmenting agents.

Level I Evidence: Multiple positive trials, supported by extensive clinical observations
Atypical antipsychotics: Olanzapine, quetiapine, aripiprazole, ziprasidone, risperidone, lurasidone and brexpiprazole
Lithium
Level II Evidence: Trials mostly supportive, with positive clinical observations
Liothyronine
Oestradiol
Level III Evidence: Several positive trials with positive clinical observations
L-methyl folate
Pramipexole (notably bipolar depression)
Modafinil
Level IV Evidence: Positive clinical observations with limited positive trials
Methylphenidate
Anti-inflammatory medications (notably TNF inhibitors, e.g. infliximab, etanercept)
Level V Evidence: Emerging clinical observations and/or open trials
Ketamine (limitations of safety and administration)
S-adenosylmethionine
Buprenorphine
Acetyl-l-carnitine
Psilocybin

TNF: tumour necrosis factor.

Atypical antipsychotics

These are the most intensively studied group of augmenting medications (Zhou et al., 2015), and evidence exists (daily dose ranges from controlled studies and clinical experience) for olanzapine (1.25–15 mg), quetiapine (25–300 mg), aripiprazole (5–20 mg), ziprasidone (40–160 mg), risperidone (0.25–1 mg), lurasidone (20–120 mg) and brexpiprazole (1–3 mg). Clinical experience suggests that the response to augmentation can be equally unpredictable as is the response to antidepressants themselves, with patients experiencing variable improvement to individual medications. Remission rates as high as 57% have been quoted with atypical antipsychotics (APs) augmentation but, in practice, tend to be less impressive (20–30%). Clinically, APs are particularly useful at the outset when initiating therapy of TRD and may often be successfully withdrawn after 3–4 weeks. However, some patients may benefit from APs being prescribed for longer periods to maintain benefits or prevent relapse. But, it is important in these instances to monitor patients and ensure that they do not develop metabolic complications and/or extrapyramidal symptoms. Hence, these considerations are critically important and should inform the initial choice of antipsychotic augmentation agent.

Lithium

Lithium has been established as an effective augmenting agent for TRD, especially with tricyclic antidepressants and particularly for melancholic depression (Nelson et al., 2014). Clinically, it is often useful in this role in combination with a variety of antidepressant medications and is generally well tolerated. Blood levels from 0.3 to 0.8 mmol/L appear sufficient, with a dose of 450 mg/day (slow-release) being a reasonable starting dose for most adult patients of average build. Side-effects such as nausea, tremor, polyuria and weight gain may limit its use, but the risk of longer term side-effects such as hypothyroidism, skin/hair changes and renal impairment can be minimised by restricting its duration of use. Measures of efficacy have been as high as 41% but, in practice, response occurs less frequently in TRD (approximately 20–30%), and if there has been no response by 4 weeks (at therapeutic dosage), lithium should be ceased. As with APs, lithium augmentation may be withdrawn within weeks of initiation or extended longer according to clinical need. And just as there should be a willingness to consider lithium augmentation as a treatment option, there should be an equal readiness to stop treatment if it proves to be ineffective.

Thyroid hormones

Liothyronine (T3) has long been used as an augmenting treatment and clinical experience supports its efficacy in a small percentage of patients, while controlled trials have reported variable efficacy. Doses of 10–50 µg daily are appropriate, but tolerability (thyroid-stimulating hormone [TSH] levels remaining above 0.5 mU/L and the absence of cardiac arrhythmias) must be considered. The benefits are most evident in women and when thyroid function measures are at the lower end of the normal range. In contrast, unless hypothyroidism is clearly present, the benefits of thyroxine (T4) for TRD are much less evident. And while remission in TRD with T3 has been recorded to be as high as 25% (Touma et al., 2017), clinical experience suggests much lower levels of effectiveness – possibly as little as 5–10%.

Other potential augmenting medications

Many medications have been investigated as augmentation strategies for managing TRD, with varying levels of support. Controversy surrounds the use of stimulants in the management of TRD, and the benefits of agents such as methylphenidate and modafinil in the management of melancholic depression clearly require further investigation. Other agents also worthy of further examination are supplements such as L-methyl folate, which is purported to make folate available to the brain – bypassing the reduced effectiveness of dietary folate which comes about because of some polymorphisms of the MTHFR gene. In a similar vein, oestradiol has a beneficial effect in TRD that occurs during perimenopause, while the evidence supporting testosterone supplements for TRD in older males is less well established. More recently, ketamine has once again come to the fore because of its seemingly short-term antidepressant effects; however, the severity of its side-effects and the need for repeated administration limit its application to research settings at this time. Finally, research continues into the use of anti-inflammatory medications (notably tumour necrosis factor [TNF] inhibitors) and pramipexole, as well as a variety of other novel and experimental agents such as S-adenosyl methionine, buprenorphine, acetyl-l-carnitine and psilocybin (Cowen, 2017).

Conclusion

Pharmacological augmentation of antidepressant medications is a worthwhile consideration in the management of TRD especially if the illness presents with clear melancholic features indicating a likely biological loading. The APs, lithium and liothyronine, are the best-established augmenting agents. They can be used alongside most antidepressants. Newer and some novel agents, some of which appear to have short-term benefits, may also be considered – noting that they still require further investigation and their benefits are often tentative and short-lived at best.