1. Extrapyramidal side-effects.
a) Dystonia: This side-effect can be effectively treated with anticholinergics (e.g., biperiden or benztropin) (C). The effects of intravenous treatment are usually prompt. (C) Risks, especially of intraveneous administration, are exacerbation of schizophrenia and anticholinergic side-effects including delirium. Pragmatically, anticholinergic medication should be continued for some time after the dystonia or the antipsychotic drug can be changed to avoid a reoccurrence (C).
We are not aware of systematic reviews or randomized trials on the treatment of this side-effect, but there is no doubt about the effectiveness of anticholinergics. The risk for acute dystonia is highest for high-potency conventional antipsychotics such as haloperidol. Other risk factors are high dose and rapid dose increase, young men seem to be more frequently affected.
b) Parkinsonism (rigidity, tremor, akinesia): The therapy consists of anticholinergics (e.g., biperiden, benztropin) or dopamine agonists (e.g., amantadine), dose reduction or change of the antipsychotic drug (C). In case of antipsychotics with a high EPS risk, prophylactic anticholinergic medication should be considered (A).
We are not aware of a systematic review on the recommended treatment strategies. Prophylactic anticholinergic medication should be considered only for antipsychotics with a high EPS risk, because such medication has side-effects, as well. In 1991 Lavin and Rifkin produced a review on prophylactic anticholinergic medication (Lavin und Rifkin 1991a) and on the withdrawal of such medication (Lavin und Rifkin 1991b). While they found more than 30 RCTs for the latter question, only few studies had examined prophylactic administration right from the start. After careful consideration of the benefits and risk, they opted for prophylactic administration with high potency first-generation antipsychotics. If anticholinergics are used prophylactically, we recommend attempting to reduce and stop this medication after maintenance dose and therapy have been implemented. This is also in line with a WHO recommendation that cautious against using anticholinergics prophylactically (World Health Organisation, 2010).
c) Akathisia: Therapeutic options are ß-blockers (e.g. propranolol), benzodiazepines (e.g., lorazepam, clonazepam, diazepam) or anticholinergics (e.g. biperiden, benztropin ), as well as mirtazapine but often the antipsychotic dose has to be reduced or the antipsychotic changed (C).
Although ß-blockers, benzodiazepines, anticholinergics and mirtazapine (Poyurovsky, 2010) are frequently recommended, Cochrane reviews found that the evidence for all four strategies is limited if any (Lima et al., 2004a; Lima et al., 2002; Lima et al., 2004b). In a review, Miller and Fleischhacker, 2000 argue that the best available evidence points towards propranolol as the treatment of first choice. In practice, it will be frequently necessary to decrease the antipsychotic dose or to switch to another antipsychotic.
d) Tardive dyskinesia: Switch the antipsychotic to a less EPS prone one (B), reduce the dose (C). Consider tetrabenazine (B). No pharmacologic intervention for TD has been found to be convincingly effective (A-).
Soares und McGrath 1999 summarized the results of their Cochrane reviews on the treatment of tardive dyskinesia and found very small numbers of RCTs and participants for the various interventions (among others tiapride, tetrabenazine, vitamine E, vitamine B6, switching to clozapine and many others). In our opinion, none of them can be generally recommended. The randomized evidence for pragmatic solutions such as switching to a less EPS prone antipsychotic (especially clozapine) is also limited to open switch studies and case series, but it is the best recommendation we can make at this point. It is based on the assumption that those antipsychotics with a high risk for acute EPS are also associated with a higher risk for tardive dyskinesia. The American Academy of Neurology also suggests to consider tetrabenazine (Bhidayasiri et al., 2013).
e) Neuroleptic malignant syndrome (NMS): Stop antipsychotic, refer to intensive care unit, hydration, treat with dopa-agonists (e.g. amantadine, bromocriptine) or muscle relaxants such as dantrolene, consider ECT (C).
As the phenomenon is rare, the evidence is based only on case reports and case series. The problem is that it is difficult to entangle MNS from severe catatonia. In the case of doubt antipsychotics should be stopped. Benzodiazepines, dopaminergic agents such as bromocriptin, dantrolene as well as ECT have been described as helpful, usually in case reports or small case series (Strawn et al., 2007).
2. Sedation: Wait to see whether sedation is transient (C). Use most of the dose in the evening (C). Reduce the dose (C). Switch to another, less sedating antipsychotic (A).
Antipsychotic drugs differ in their sedating properties (Leucht et al., 2013b), therefore switching to a less sedating one is an obvious option. Low-potency first-generation antipsychotics, such as chlorpromazine, but also some second-generation antipsychotics, such as clozapine, zotepine, quetiapine or olanzapine, are more sedating than others. However, to change the antipsychotic is not always possible. As sedation can be transient, it can be worthwhile to wait for some time whether it remits spontaneously. Other pragmatic strategies are to reduce the dose or to give most of the dose before sleep to avoid plasma-level peaks during the day.
3. Weight gain and associated metabolic side-effects (glucose and lipid abnormalities): Prevent by choosing an antipsychotic with little weight gain (A). Facilitate life-style changes, diet and physical exercise (A). Switch to a lower-risk antipsychotic (A). Try adding metformin or topiramate to reduce weight (A). Use antihypertensives, lipid lowering drugs (e.g., statins) and antidiabetic agents if indicated (A).
Although various strategies to reduce overweight and related metabolic side-effects have been developed and for several ones positive evidence is available, questions about the magnitude of their effects, their long-term effectiveness and their side-effects remain. Therefore, prevention in terms of choosing an antipsychotic with a low weight gain risk plays a key role (Faulkner et al., 2007, see Table 1). Non-pharmacological strategies are not the focus of this guideline, but patients and their relatives should definitely be educated about and motivated for therapeutic life-style changes, including adequate physical activity and a healthy diet, as weight management interventions have shown significant benefits over control conditions (Caemmerer et al., 2012). One beneficial option with randomized controlled evidence is the switch to an antipsychotic with significantly lower weight gain potential (Stroup et al., 2011). Among the many drugs that have been tried to reduce weight as adjunctive treatments to antipsychotics in schizophrenia, the best evidence is currently available for metformin (Mizuno et al., 2014). Although we do not endorse antipsychotic polypharmacy (see above), for clozapine treated patients another option may be adding aripiprazole, which in an adequately powered RCT was associated with significantly more weight loss than adding placebo (Fleischhacker et al., 2010). Guidelines also emphasize that drugs like antihypertensives, lipid lowering drugs (e.g., statins) or antidiabetic agents should be used if indicated, because this is not sufficiently done in people with severe mental illness (De Hert et al., 2011a; de Hert et al., 2011b).
4. Cardiovascular side-effects.
a) Orthostatic hypotension and reflex tachycardia: advise to stand up slowly, slow dose increase, dose reduction, divide into several doses per day, change the substance (C).
Orthostatic hypotension (drop of blood-pressure when standing up) and resulting compensatory reflex tachycardia are explained by alpha-1 receptor blockade of certain antipsychotics. The listed recommendations are of pragmatic nature.
b) Tachycardia: Dose reduction (C), change the antipsychotic (C), add a ß-blocker (C).
Tachycardia can also be the result of anticholinergic effects of antipsychotics (e.g., clozapine). In addition to pragmatic measures, the addition of a ß-blocker can be useful.
c) QTc prolongation, higher grade arrhythmias: reduce the dose (C), change the antipsychotic (A), monitor electrolytes, comorbidities and co-medication (C).
The treatment with antipsychotic drugs can also be associated with arrhythmias such as bundle branch blocks, changes of the QRS complex, etc, and, in the worst case, potentially fatal torsades de pointe. Prolongation of the QTc interval, which can predispose to arrhythmias, occurs with many antipsychotics, but few prolong the QTc to a significant degree (see Table 1). QTc intervals >500 msec are associated with a higher frequency of torsades de pointe and sudden death, and should therefore lead to switch to an antipsychotic with a lower QTc prolonging potential. QTc >450msec is usually considered to be potentially dangerous, but the association between QTc and arrhythmias at that level is less clear. Among first-generation antipsychotics, the risk for QTc prolongation seems to be highest with thioridazine and pimozide; among second-generation antipsychotics sertindole, amisulpride and ziprasidone seem to be most affected (see Table 1, and Leucht et al., 2013b). Obviously, QTc prolonging drugs are especially problematic in patients with known heart disease, congenital long QT syndrome, a history of syncope and other risk factors. As the risk is often dose related, dose reduction may be tried, if possible. Electrolyte imbalances (e.g., hypokalemia or hypomagnesemia) can increase the risk further and should therefore be monitored and corrected. Moreover, the combination with other psychotropic or non-psychiatric medication with a risk of QT prolongation should be avoided (for a list see for example http://www.torsades.org).
5. Hyperprolactinemia: asymptomatic hyperprolactinemia does not necessarily need treatment (C). Strategies to reduce hyperprolactinemia are dose reduction (C), change to a prolactin sparing antipsychotic (A), adding a partial dopamine agonist (A).
Hyperprolactinemia can be associated with sexual side-effects (e.g., amenorrhea, galactorrhea, loss of libido), osteoporosis (Kishimoto et al., 2012), and there is also an unresolved discussion about an increased risk for breast cancer (Haddad and Wieck, 2004). Some of these side-effects are, however, multifactorial. E.g. certain sexual side-effects, such as lack of libido, may be due to the negative symptoms of schizophrenia, and osteoporosis has also been associated with lack of physical activity due to the same symptoms. The experts feel that as long as hyperprolactinemia is asymptomatic, changes in antipsychotic drug treatment are not mandatory. In case of very high levels (i.e., >200 ng/dL), an endocrinologist may, however, be consulted. The risk for hyperprolactinemia of the various drugs differs enormously (see Table 1, and Leucht et al., 2013b). In RCTs, it has been shown that adding the partial dopamine agonist aripiprazole can reduce prolactin levels (Gallego et al., 2012).
6. Seizures: reduce the dose (C), use an antipsychotic with less epileptogenic potential (C), add an antiepileptic (C).
Antipsychotic drugs differ in their propensity to reduce the seizure threshold. The literature stems mostly from case reports and retrospective analyses. These effects are dose dependent. Hedges et al., 2003 concluded that chlorpromazine has the greatest risk among the first-generation antipsychotic drugs and clozapine among the second-generation antipsychotics, while the risk of fluphenazine, haloperidol, molindone, pimozide, risperidone and trifluoperazine appears to be low. Other factors that should be considered are a history of seizure activity, combination with other seizure threshold lowering drugs, rapid dose increase and drug-drug interactions (Hedges et al., 2003).
7. Liver enzyme elevation: monitor (C), if persistent, obtain medical check-up (C), reduce the dose (C), use an antipsychotic which is metabolized little by the liver and (mainly) excreted via the kidney (C).
Liver enzyme elevation is frequent and often transient. If it persists or if levels are high, a medical check-up should be obtained. Dose reduction can be tried, and it can be necessary to switch to a drug which is only very little metabolized by the liver, such as paliperidone, amisulpride and (less so) ziprasidone.
8. Photosensitivity and other dermatologic side-effects: sun protection (limit sun exposure, use sun cream, protect skin with clothes) (C). Allergic reactions often make a switch of the antipsychotic necessary (C).
Antipsychotic drugs and other psychotropic agents can lead to photosensitivity and allergic reactions. The strategies to prevent or treat these conditions are pragmatic.
9. Blood count changes: monitor (C), consider adding lithium, and monitor white count in the afternoon (C), change or stop medication (C), follow rules for agranulocytosis.
Transient leukopenia, leukocytosis, lymphocytosis, thrombocytopenia and eosinophilia do not require a change of treatment. Agranulocytosis is a life-threatening side-effect which often requires hematologic intensive care (definition: granulocytes < 1000/mm3, complications must be expected with values < 500/mm3). The incidence of this side-effect is highest for clozapine; therefore, when this drug is used, regular differential blood-counts must be taken, and if the white blood-count or granulocyte count falls below a certain threshold clozapine needs to be stopped. Patients also need to be informed about the symptoms of agranulocytosis, which often first manifests as a fever or infections such as a sore throat. Rechallenge after simple leucopenia is possible and often successful. Since leucocyte and granulocyte levels have diurnal variation, blood draws in the afternoon yield higher levels. As the exact recommendations differ locally, we need to refer to the recommendations in the different countries.