Antidepressant drugs for older patients on polypharmacy: a systematic review reveals best evidence for sertraline

Clinical depression (e.g., major depressive disorder) is common in older adults. Its prevalence in the community is around 15%, and milder forms of depression are more common. The number of people over the age of 60 years is expected to double by 2050, and hence, interventions for this often long-term and recurrent condition are increasingly important (Wilkinson and Izmeth 2016). In a meta-analysis, the authors found that antidepressants (tricyclic, selective serotonin-reuptake inhibitors, and a mixed group of other drugs) were significantly better than placebo, with an average reduction in symptom severity of 48.0% versus 31.3% (Gerson et al. 1999). Based on drug-to-drug comparisons, the selective serotonin-reuptake inhibitors appeared to be as effective as heterocyclic drugs (McCusker et al. 1998). In real clinical practice, often one antidepressant may seem to be more effective than another, although there is little advantage for one antidepressant class over another that is supported by the literature. In another meta-analysis, the authors compared different antidepressant groups in depressed older patients. They found that older patients showed no differences in antidepressant class outcomes in terms of efficacy and tolerability (Mittmann et al. 1997). Although antidepressant treatment efficacy has been shown, compared with younger individuals, older patients may be more likely to experience relapses and less likely to achieve a full response to treatment with antidepressant medications (Dew et al. 2007, Driscoll et al. 2005, Driscoll et al. 2007). Although these results are important, patients treated with polypharmacy have not been included in these trials, which means that most patients in ‘real clinical practice’ are not included in these important papers; hence, an evidence-based approach based on well-designed network meta-analyses and clinical trials should be used to inform the choice of pharmacotherapy in this population.

Polypharmacy is a world-wide problem and is often considered in a negative way in relation to good clinical practice although many patients indeed need many of their medications. Four out of 5 people aged 75 years and above are taking medications, and 36% receive 4 or more medications at the same time, which means that most patients in this population are not considered in the existing treatment guidelines (Quality and Outwork framework 2012). The probability of a drug-drug interaction (DDI) increases with the number of medications (polypharmacy). Specifically, a patient taking 5–9 medications had a 50% probability, whereas the risk increased to 100% when a patient was found to be taking 20 or more medications (Doan et al. 2013). The probability of a DDI increased with the number of medications. Addition of each medication to a 5-drug regimen conferred a 12% increased risk of a potential CYP-mediated DDI (Doan et al. 2013). In terms of polypharmacy, antidepressants are often prescribed inappropriately, as was shown in an Austrian cross-sectional study that included 48 out of 50 nursing homes and 1,844 out of 2,005 residents. The authors found high prevalence of potentially inappropriate prescriptions (PIPs). The prevalence of residents with at least one PIP was 70.3% (95% CI 67.2–73.4) and the antidepressants amitriptyline and maprotiline as PIM were prescribed very often (Mann et al. 2013).

In addition, most older depressed adults are excluded from the existing treatment guidelines and well-designed RCTs and meta-analyses, which means that there is a great lack of evidence to manage these patients in real clinical practice. Thus, the main aim of this paper was to identify the data on antidepressant treatment in depressed older adults on polypharmacy using a systematic review approach.

Results of this systematic review show that there are almost no trials in this patient population, as we only found one trial that met the inclusion criteria. The trial found that sertraline has the highest evidence for its use in older adults on polypharmacy, primarily because of its very low impact on potential DDIs as a consequence of polypharmacy (Arranz et al. 1997). These results are in line with a network meta-analysis published by Cipriani et al. (Cipriani et al. 2009), where sertraline was the best choice when starting a treatment for moderate to severe major depression in adults because it had the most favourable balance between benefits, acceptability, and acquisition cost. Although this network meta-analysis excluded older adult patients on polypharmacy, results are comparable with those of our systematic review, which means that sertraline in this population has been supported by the highest evidence level obtained from a well-designed meta-analysis and a more ‘real clinical practice’ systematic review. These results show that there is some evidence to support the use of sertraline in this population, although this evidence is very weak, because of the fact that long-term treated patients and different types of polypharmacy were not included in these reviews. Thus, well-designed long term head-to-head trials should be conducted to obtain more relevant results, and these should be followed by a network meta-analysis of these trials. These results could be connected with a network meta-analysis of older adults (60 years and older) where for partial response, sertraline (RR = 1.28), paroxetine (RR = 1.48), and duloxetine (RR = 1.62) were significantly better than placebo (Thorlund et al. 2015). The remaining interventions yielded RRs lower than 1.20. For dizziness, duloxetine (RR = 3.18) and venlafaxine (RR = 2.94) were statistically significantly worse than placebo. Compared with placebo, sertraline had the lowest RR for dizziness (1.14) and fluoxetine being the second lowest (1.31). Citalopram, escitalopram, and paroxetine, all had RRs between 1.4 and 1.7. These results are in line with our results, that sertraline has the best risk-benefit ratio, although there was no data about polypharmacy in the included trials (Thorlund et al. 2015). In addition to this weak evidence in this important population, a solid knowledge of basic clinical pharmacology in pharmacotherapy planning in older adults would be the best approach to avoid serious DDIs and adverse events as was shown in various important papers (Holt et al. 2010, American Geriatrics Society 2015, O’Mahony et al. 2015). The use of sertraline in patients on polypharmacy could be explained pharmacologically. Sertraline has a low potential for DDIs because it is not a strong inductor nor inhibitor of cytochrome 450 isoforms. According to in vitro studies, sertraline is metabolized by multiple cytochrome 450 isoforms: CYP2D6, CYP2C9, CYP2B6, CYP2C19 and CYP3A4. It is important to add that there are few possibilities that the inhibition of any single isoform could cause clinically significant changes in the pharmacokinetics of sertraline (Obach et al. 2005). These important pharmacokinetic properties of sertraline show that it has a small number of possible clinically important DDIs in real clinical practice. Despite the fact that these pharmacokinetic properties are important, it is clear that sertraline cannot be used in all older patients on polypharmacy because antidepressants have different mechanisms of action that are very important when prescribing. These important issues are often excluded from clinical trials, in spite of their importance in daily practice. It is also important to add that sertraline has also been involved in some important DDIs in elderly patients. These mostly included DDIs via the CYP P450 3A4 system, where strong inhibitors/inductors could have an impact on sertraline adverse events or lack of efficacy (Khan et al. 2000). Thus, in addition to the results of this systematic review, a pharmacotherapy review before prescribing should be an important option in order to avoid serious DDIs and treatment failure in this population.

We should emphasise that in addition to the number of potential DDIs, the type of DDI is also very important for pharmacotherapy planning. Pharmacodynamic DDIs (e.g., QTc prolongation) are often associated with antidepressant use, especially in case reports (e.g., citalopram, duloxetine, venlafaxine, and bupropion), although these effects are excluded from treatment guidelines and often do not include older patients on polypharmacy (American Psychiatric Association 2010, Jasiak and Bostwick 2014, Štuhec 2013, Tampi et al. 2015). This has not been observed in a network meta-analysis in which mirtazapine, escitalopram, venlafaxine, and sertraline were significantly more efficacious than duloxetine (odds ratios [OR] 1.39, 1.33, 1.30 and 1.27, respectively), fluoxetine (1.37, 1.32, 1.28, and 1.25, respectively), fluvoxamine (1.41, 1.35, 1.30, and 1.27, respectively), paroxetine (1.35, 1.30, 1.27, and 1.22, respectively), and reboxetine (2.03, 1.95, 1.89, and 1.85, respectively). Escitalopram and sertraline showed the best profile of acceptability, leading to significantly fewer discontinuations than did duloxetine, fluvoxamine, paroxetine, reboxetine, and venlafaxine (Cipriani et al. 2009). Although both meta-analyses are interesting, there is no data on polypharmacy use including antidepressants, and therefore, these meta-analyses including the famous Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial are not representative for this patient population (Cipriani et al. 2009, Sinyor et al. 2010).

Although there is plenty of evidence described for the general population, there is a lack of supportive trials in older patients treated with polypharmacy in relation to the choice of antidepressant (no RCTs). A cautious psychopharmacological approach is necessary when prescribing in order to avoid serious adverse events, DDIs, and lack of treatment efficacy. Although these patients are often treated in real clinical practice they are actually not covered completely (or even not covered at all) in available treatment guidelines. This means that practitioners should use a cautious psychopharmacological approach in this patient group. One possible approach to reduce adverse events in this population is to use different lists of inappropriate medications (e.g., the STOPP/START criteria, 2015; the American Geriatrics Society Beers Criteria for Potentially Inappropriate Medication Use in Older Adults, the PRISCUS list), which would help minimising the use of inappropriate antidepressants in this population (American Geriatrics Society 2015, Holt et al. 2010, O’Mahony et al. 2015). According to the PRISCUS criteria fluoxetine, the MAO inhibitor tranylcypromine, and TCAs including amitriptyline, doxepine, imipramine, clomipramine, maprotiline and trimipramine, should be avoided in this population. Trazodone, other SSRIs (e.g., escitalopram, sertraline), and mirtazapine could be used. A reduced dose of TCAs is recommended for elderly patients because elderly patients taking TCAs have been reported to have a higher frequency of confusion and other manifestations relating to the central nervous system (Holt et al. 2010). However, although the PRISCUS list is useful in daily practice, it has many limitations. The PRISCUS list for the elderly was developed specifically for use in Germany, which means that many medications are not included (i.e., those which are not available in Germany). Another important limitation is the lack of comparative evidence for each medication group, which is not the case in the 2015 American Geriatrics Society Beers Criteria for Potentially Inappropriate Medication Use in Older Adults (American Geriatrics Society 2015). In the Beers criteria, a strength of recommendation has been added, which helps clinicians and clinical pharmacists to categorise different medication groups within the list. Amitriptyline, amoxapine, clomipramine, desipramine, doxepin > 6 mg/d, imipramine, nortriptyline, paroxetine, protriptyline, and trimipramine have been included in the Beers criteria 2015 with a high quality of evidence and a strong recommendation to avoid in clinical practice in this population (American Geriatrics Society 2015).

This systematic review also has many important limitations. We also included non RCTs, which increased the bias, although these trials are closer to the real clinical patients. It also only included patients on polypharmacy, which could be defined differently in different countries and we included only those trials with antidepressants, although other medications are also used for depression treatment in older adults (e.g., antipsychotics). Another very important limitation is the lack of long term trials because patients treated with antidepressants are treated for a long period of time (e.g., 12 months). Another limitation is the lack of comparative polypharmacy trials because different medications can have a different impact on DDIs. In addition, patients’ habits are also very important (e.g., smoking), as they can have an impact on drug metabolism (e.g., clozapine, duloxetine) (Mangerud et al. 2014). The next limitation is the lack of dose comparison. Some adverse events could be dose-dependent. Antidepressants also have different half-lives, which means that short-term trials including medications with long halflife including polypharmacy can lead to inappropriate results because many DDIs can occur after many weeks of treatment (e.g., DDIs with fluoxetine). Despite these limitations, this is the first systematic review including patients treated with polypharmacy without treatment-resistant depression, and therefore, the results are important in the field of evidence-based medicine.

Sertraline has the highest level of evidence for use in older adults with depression and on polypharmacy. An individual treatment approach including different PIM lists and basic pharmacology considerations is one of the possible tools to manage these patients. More ‘real clinical setting’ trials and RCTs are needed to provide further evidence on the important topic of optimisation of the treatment of depression in older patients on polypharmacy.