Are antidepressants a double-edged sword? Treatment emergent affective switch or antidepressant discontinuation syndrome

A treatment emergent affective switch (TEAS) is one prompted by pharmacotherapy (Malhi et al., 2015), and when prescribing antidepressants in bipolar disorder, a TEAS into mania is a common concern, particularly in the absence of concurrent mood-stabilizing/anti-manic agents. Paradoxically, mania can also emerge when withdrawing antidepressant therapy, though more commonly the cessation of antidepressant treatment produces somatic and psychiatric symptoms such as nausea, headache and low mood – collectively described as an ‘antidepressant discontinuation syndrome’ (ADS). Occasionally, ADS can include manic symptoms (Harvey and Slabbert, 2014) and this makes distinction of ADS from TEAS difficult. However, separating the two phenomena is important for the purposes of informing future management – as exemplified by the following case.

A patient treated for a first episode of major depressive disorder had not responded to venlafaxine XR (262.5 mg) augmented with lithium (250 mg) and so this combination of medications was withdrawn over a period of 10 days and cross-titrated to mirtazapine 45 mg – during which time her mood became labile. Over a subsequent similar period, she experienced racing thoughts and increased dysphoria and energy – leading to impulsive behaviour and absconsion from hospital. Mirtazapine was then withheld, and her mixed mood state was treated with olanzapine (10 mg daily). Her manic features gradually resolved and did not recur when antidepressant treatment was eventually reintroduced. However, the precise cause of the mixed features (ADS or TEAS) remained unclear as did how to prevent this recurring in the future.

Factors likely to be involved in the development of ADS

Following the abrupt cessation of antidepressant treatment, the rapid elimination of medication from the plasma appears to trigger ADS, especially when agents have a short half-life (Warner et al., 2006). Therefore, rapid CYP450 metabolisers are at greater risk of ADS. The duration of prior treatment is also important, with ADS usually occurring after duration of antidepressant treatment of at least 5–8 weeks. Similarly, higher doses of medications are associated with a greater incidence and severity of ADS. This association with dose and duration suggests the possibility of neuroplastic events which might resemble those that take place when an antidepressant is commenced to the time it reaches its therapeutic effect.

Despite these ‘known’ associations, our understanding of ADS remains limited, in particular in relation to its underlying neurobiological mechanisms. Cholinergic overdrive was first noticed following the withdrawal of tricyclic antidepressants (TCAs) and some selective serotonin reuptake inhibitors (SSRIs), and this is a concern given that excessive cholinergic drive may trigger relapse. Hyper-serotonergic states have also been implicated because of down regulation occurring during treatment and disinhibition of serotonin receptors occurring upon cessation of antidepressant treatment, possibly leading to neuropsychiatric changes responsible for the symptoms of ADS. Alternatively, a hypo-serotonergic state that is thought to emerge during the gradual recovery of serotonin receptor transporter activity may also play a precipitatory role. This could explain why ADS can instantly be alleviated by the reintroduction of an antidepressant. In addition, increased glutamate release as part of a stress response, with a possible neurotoxic effect, alongside dopaminergic and noradrenergic pathways, may contribute. All of these changes could be conceptualised as a ‘bio-behavioural stress response’, potentially instigating neuroplastic changes in the brain with long-term detrimental effects resulting in clinical sequelae such as treatment resistance, increased likelihood of relapse and poor outcome (Harvey et al., 2003). Clearly, ADS is potentially damaging and should be avoided where possible. Diagnostic criteria for ADS-induced manic states in the context of uninterrupted antidepressant treatment for at least 4 weeks specify that the manic state should appear within 1 week of antidepressant dose reduction or cessation. In addition, there should be no pharmacological confounds (Narayan and Haddad, 2010). However, in clinical practice, the latter stipulation is often difficult to satisfy, as shown in the aforementioned case. This is partly because cross-titration is a common strategy when switching antidepressants – making it difficult to determine if a mood switch is best attributed to TEAS or ADS (Harvey and Slabbert, 2014). Covert non-adherence to antidepressants inducing a withdrawal mania, could also be interpreted as a TEAS rather than ADS, altering the focus of treatment. Nonetheless, whether mania incipient upon withdrawal indicates a latent bipolar disorder or a forme fruste remains unknown.

Recommendations

Evidently further research is required to better understand the aetiology of TEAS and ADS, and this is likely to provide useful insights into the actions of antidepressants and the neurobiology of mood disorders.

In the interim, in practice, TEAS and ADS have important ramifications for clinical practice such as patients refusing future treatment (Warner et al., 2006), or receiving inappropriate medication. To minimise such risks, it is advisable to taper antidepressants over a period of at least 4 weeks, or to cross-titrate medications where possible. Furthermore, where there is a known past history of TEAS/ADS, it might be prudent to concurrently prescribe a mood stabiliser.