How do we select an antipsychotic for those with schizophrenia?

KEYWORDS:

• Schizophrenia
• antipsychotics
• long acting injectable
• clozapine
• medication selection
• first episode psychosis

1. Introduction

Schizophrenia is a severe progressive and debilitating condition, comprised of three primary symptom domains: positive, negative, and cognitive. Individuals with schizophrenia have a decreased life expectancy and suffer from a wide range of somatic conditions [1]. The etiology of schizophrenia is unclear, and there are likely many different pathophysiological pathways involving the interaction between dopamine and glutamate [2]. The mainstay of treatment for schizophrenia is antipsychotic medications, which function by antagonism or partial agonism at the dopamine D₂ receptor. Second generation antipsychotics (SGAs) also function at other receptors, including antagonizing serotonin at the 5-HT_2a receptor. Although these medications may lessen positive symptoms, negative and cognitive symptoms are less likely to improve. Antipsychotic medications can also precipitate movement disorders, such as extrapyramidal symptoms (EPS) and tardive dyskinesia (TD), metabolic syndrome, and other distressing side effects. Despite this, antipsychotic treatment adherence has been shown to prolong life [3]. Balancing the treatment of schizophrenia symptoms while concurrently managing antipsychotic induced side effects is a significant obstacle in treating schizophrenia. This paper will explore factors that go into selecting a medication for the treatment of schizophrenia, including antipsychotic efficacy and tolerability, and emphasize the critical importance of treatment adherence, especially during initial psychotic episodes.

2. Schizophrenia: to treat or not to treat

Compared with the general population, the life expectancy of those with schizophrenia is approximately 15 years less, with the primary cause of death being cardiovascular disease [1]. Individuals with schizophrenia are more likely to have a sedentary lifestyle, obesity, and tobacco-related disorders, but have lower rates of health care service utilization [1]. As a class, SGAs and low potency first generation antipsychotics (FGAs) are known to precipitate metabolic disturbances, such as insulin resistance, diabetes, dyslipidemia, and ketoacidosis, which can be precursors to cardiovascular disease. This raises the question, is the premature death experienced by those with schizophrenia attributable to the illness and its sequelae, or the treatment? In a recent systematic review and meta-analysis exploring the association between long-term mortality and antipsychotic use in patients with schizophrenia, a lower rate of all-cause mortality was found for patients with any antipsychotic exposure compared with patients who did not receive antipsychotic medication (pooled risk ratio 0.57, p value <.001) [3]. Although chronic antipsychotic use can be associated with metabolic side effects, the available evidence indicates a favorable risk-to-benefit ratio [1].

3. Factors used to select a medication

3.1. Second-generation antipsychotic or first-generation antipsychotic?

SGAs are marketed as more effective and tolerable than FGAs, and newer antipsychotics (cariprazine, brexpiprazole, lurasidone, asenapine, iloperidone) are considered to be SGAs. One might infer this to support that SGAs as a class are more advantageous than FGAs. However, the evidence supporting this premise is lacking, and initial optimism has faded [4–6]. Quality of life measures between SGAs and FGAs present conflicting results, and although SGAs are less likely to precipitate EPS and TD than FGAs, the risk has not been eliminated [7,8]. In comparison to FGAs, a meta-analysis of 57 studies showed that SGAs reduce the annualized incidence of TD to one-third [9]. The problem may lie in the classification system of FGAs and SGAs, implying that each class is homogenous. There is significant variability amongst individual antipsychotics. Rather than being a ‘me too’ drug, each individual antipsychotic has a unique binding and side effect profile, resulting in a distinct pattern of efficacy and tolerability [10]. The majority of studies comparing FGAs to SGAs utilize haloperidol as the comparator, a high-potency antipsychotic with an increased risk for EPS and TD, especially when given at high dosages. This limits the applicability of comparison trials, as mid or low-potency FGAs could result in different findings [5]. A meta-analysis concluded that there is insufficient evidence to support or reject the premise that all FGAs are equally efficacious [11]. Similarly, SGAs are not a homogenous group. Although little distinguishes the efficacy between SGAs and FGAs as a class, individual SGAs may confer unique advantages. A meta-analysis of 150 double-blind studies found that clozapine, olanzapine, risperidone, and amisulpride (not available in the US) were significantly more efficacious than FGAs, whereas other SGAs showed no significant differences [5]. Tolerability differences can help to dictate treatment in balancing the risk for metabolic concerns against the risk of EPS and TD. In general, SGAs as a class cause greater metabolic adverse effects with a reduced risk for movement disorders, though there is substantial variability between individual medications [6]. The heterogeneity of response to treatment, coupled with pharmacological variation, indicates that antipsychotic selection should focus on the individual patient and medication, using an ‘N of 1’ approach, rather than class membership of the drug.

3.2. Efficacy

The bulk of evidence from meta-analyses on antipsychotic medications have established three ‘tiers’ of efficacy: 1) clozapine has consistently demonstrated superiority over other treatment options, 2) olanzapine, risperidone, and amisulpride are significantly more efficacious than FGAs and most other SGAs, and 3) the remaining antipsychotics are similarly efficacious [5,6,12]. Some evidence suggests that quetiapine and ziprasidone may be inferior to some other SGAs, though quetiapine is often under dosed due to sedation [12].

Although SGAs and FGAs have been shown to have statistically significant improvements in the treatment of negative symptoms relative to placebo (effect size −0.579 and −0.531 respectively), these are not felt to be clinically meaningful as measured by the Clinical Global Impression Scales (CGI) [13]. Similarly, a separate meta-analysis found little evidence to support individual antipsychotic medications for the treatment of negative symptoms [14]. Cariprazine may show promise as it was found to improve negative symptoms (CI −0.48 to −0.11), but not positive symptoms, greater than risperidone in an industry sponsored trial, though more research is needed [15].

Before determining lack of treatment efficacy and changing medication, ensure that the dosage of the medication has been optimized and used for a sufficient duration. This can mean upwards of 6 weeks or 12 weeks for clozapine [16]. Efforts should also be made to limit confounding factors, such as medication non-adherence, substance abuse, and pharmacokinetic or pharmacodynamic failures. Premature medication discontinuation can result in polypharmacy and complicated medication regimens which can lead to side effects and adherence difficulty.

3.3. Tolerability

For a medication to be effective in the ‘real-world’, it needs to be tolerable in addition to efficacious. Focusing exclusively on efficacy can ignore considerable differences in individual antipsychotic side effect profiles, which can impact adherence. Tolerability issues vary from antipsychotic to antipsychotic. The rise of SGAs at the expense of FGAs has resulted in an increase in cardiometabolic side effects and a reciprocal decrease in EPS and TD [8]. At times, there is an inverse relationship between efficacy and cardiometabolic tolerability. While being highly efficacious, clozapine and olanzapine may cause the greatest increase in weight and blood glucose, possibly due to having the greatest affinity for 5-HT2c and histamine (H)1 receptors. The most dramatic weight gain is found during the first six weeks of treatment in antipsychotic naïve patients. Once it has occurred, switching to a more metabolically neutral antipsychotic fails to cause weight loss, underscoring the importance of initial medication selection [17]. Lurasidone and ziprasidone, followed by aripiprazole, cause the least increase in weight and blood glucose [6,18]. Ziprasidone, amisulpride, and sertindole cause the most QTc prolongation, although amisulpride is otherwise very well tolerated [6]. Risperidone and paliperidone, the active metabolite of risperidone, have high rates of dose-dependent EPS and cause the greatest increase in prolactin [6].

Haloperidol is not synonymous with all FGAs. Although it is relatively weight neutral, it has a high rate of EPS, TD, and all-cause discontinuation [6]. This increased risk of movement disorders is not uniform for all FGAs as illustrated in the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Schizophrenia Trial, where a mid-potency FGA did not differ significantly from four tested SGAs [4,8]. Some have suggested reclassifying FGAs and SGAs into a new classification system based on the risk of EPS [8]. In the same vein, perhaps we should consider a classification system based on cardiometabolic effects, in which case, meta-analyses that use indirect and direct antipsychotic comparisons can be referenced to aid in medication selection [6,18].

3.4. Effectiveness

All-cause time to treatment discontinuation has been accepted as a proxy measure of effectiveness, integrating aspects of antipsychotic adherence, efficacy, and tolerability. Using direct and indirect comparisons, a meta-analysis of 53,463 patients with chronic schizophrenia found that more patients drop out due to inefficacy than tolerability [6]. Of 20 antipsychotic medications that significantly lowered discontinuation rates compared with placebo, amisulpride, olanzapine, paliperidone, and clozapine were amongst those with the lowest rates of all-cause discontinuation, whereas haloperidol was the highest [6]. Focusing exclusively on direct comparisons of SGAs, a separate meta-analysis supported these findings, identifying clozapine, olanzapine, and risperidone as superior to several other SGAs, whereas quetiapine was inferior to several [12].

4. Adherence

Approximately 50% of patients with schizophrenia do not fully adhere to their antipsychotic treatment, and upwards of 75% become non-adherent within two years of hospital discharge [19]. Reasons for non-adherence can include: side-effect burden, lack of insight, inefficacy, poor therapeutic alliance, limited social support, a complicated medication regimen, amongst others. Consequences of non-adherence can include symptom decompensation, rehospitalization, suicide, and a worse prognosis [19]. Effectively treating the first episode of psychosis (FEP) is of particular importance because response rates fall dramatically during subsequent psychotic episodes. Medication non-adherence is associated with a 4-fold increase for relapse after a FEP, so optimizing and individualizing treatment to improve adherence should be a primary goal [20]. Pseudo-adherence, falsely believing one is taking their medication, is a common problem. One strategy to address this is long acting injectable (LAI) antipsychotics.

5. Long acting injectables

Long acting injectable antipsychotics are typically utilized to improve treatment adherence in a chronically non-adherent patient population. However, the benefits of LAIs may extend beyond this patient population. Depending on the medication selected, injection frequency currently ranges from 2–12 weeks. Despite the potential benefit, LAIs are underutilized [21]. Intuitively, LAIs should improve adherence to treatment, a prognosticator for a better outcome, however, meta-analyses and systematic reviews show inconsistent findings related to adherence and all-cause discontinuation [21,22]. Different study designs often recruit different patient populations (cohort vs randomized control trial (RCT) vs mirror-image), and some information suggests that patients on LAIs were more severely and/or chronically ill than patients on oral antipsychotics, which could bias the outcomes and explain differing results [21].

Studies focusing on early stage illness show outcomes favoring LAIs over oral antipsychotics [21]. However, there is inconsistent evidence to support overall differences in efficacy or tolerability between LAI and oral antipsychotics outside of this subgroup [22]. LAIs are similar to each other in relapse prevention but differ in their side effect profiles [21]. These tolerability differences can be used to individualize treatment.

6. Subgroups

6.1. First episode psychosis (FEP)

A prolonged initial duration of untreated psychosis (DUP) is associated with poor symptomatic and functional recovery, and poor overall long-term outcomes. A short DUP has been associated with a greater response to antipsychotic treatment, reinforcing the importance of prompt disease recognition and treatment [23]. The first episode of psychosis (FEP) has a high response rate to initial treatment (80% achieved at least a 20% Positive and Negative Syndrome Score [PANSS] or Brief Psychotic Rating Scale [BPRS] score reduction from baseline) [23]. These individuals are typically younger, and usually present with prominent positive symptoms, rather than negative or cognitive symptoms which are more difficult to treat.

A pairwise network meta-analysis of 19 RCTs (n = 2669) evaluated antipsychotic medications in the first episode of schizophrenia [24]. Few differences in efficacy were found between individual SGAs, though haloperidol was found to be a sub-optimal initial treatment. The authors recommend selecting treatment based on side effects.

In a randomized, open-label effectiveness trial, 202 drug-naïve patients with FEP were randomized to aripiprazole (n = 78), ziprasidone (n = 62), or quetiapine (n = 62), and followed for three years to measure all-cause treatment discontinuation. The overall discontinuation rate reached 81.7% by 3 years (similar to the CATIE study). However, more than three-quarters (76.4%) of patients who discontinued treatment did so during the first year of follow-up. The median time to all-cause discontinuation for quetiapine was significantly shorter (60 days), than for ziprasidone (251 days) or aripiprazole (452 days), primarily motivated by insufficient efficacy [20].

When the FEP occurs prior to age of 18, it is considered early-onset schizophrenia (EOS), and predicts a potentially worse prognosis. Common findings from guidelines and meta-analyses on EOS include: 1) antipsychotics being more efficacious than placebo (possibly excluding ziprasidone), 2) SGA’s being equally efficacious to FGA’s, 3) tolerability of SGA’s being superior to FGA’s, 4) adolescents experience more side effects from antipsychotics than adults; EPS from FGA’s, metabolic side effects from SGA’s, especially olanzapine [25]. Based on these findings, it is recommended to individualize antipsychotic medication selection in EOS, factoring in medication side effect profiles, medical comorbidities, and patient preference.

6.2. Treatment resistant schizophrenia

Approximately one-third of individuals with schizophrenia will show an inadequate response to antipsychotic medications, and are considered to have treatment resistant schizophrenia (TRS) [16]. Clozapine has been shown to have the greatest efficacy in treating schizophrenia, and is considered the gold standard treatment for TRS. However, there is a lack of a uniform diagnostic criterion for TRS. Recognizing this diagnostic ambiguity, a working group concluded that comparing studies on TRS is akin to ‘comparing apples to oranges’, and created a consensus guideline on the minimum and optimum diagnostic criteria [16]. Despite diagnostic uncertainty, a meta-analysis (n = 240,564) showed that patients on long-term treatment with clozapine demonstrated a 2.9-fold (95% CI = 1.5–5.7) overall risk reduction of completed suicide [26]. Additionally, despite the considerable side effect burden, a separate meta-analysis found that continuous clozapine treatment in schizophrenia patients resulted in a significantly lower long-term mortality rate than those who were continuously treated with other antipsychotics (mortality rate ratio = 0.56, 95% CI = 0.36–0.85,p = .007) [27]. Taking all factors into account, clozapine still shows the greatest efficacy in this difficult-to-treatment patient population, and is underutilized.

7. Expert opinion

Schizophrenia is likely a heterogenous condition with multiple different biologic subtypes. This is an important distinction because all persons with schizophrenia should not be conceptualized uniformly. The dilemma of antipsychotic medication selection in the ‘real world’ centers around choosing a medication that balances efficacy with tolerability on a case-by-case basis. Although all-cause discontinuation is considered a proxy for overall effectiveness, it should not be considered a substitute for individualized treatment. The priorities of each patient should be of utmost consideration. Is their primary concern, 1) relief from symptom psychopathology, in which case a medication should be selected with a favorable efficacy profile, such as olanzapine or risperidone, or, 2) tolerability concerns, in which case, lurasidone, ziprasidone, or aripiprazole can be considered.

Individuals with schizophrenia often present with numerous medical comorbidities and yet have chronically low rates of health care service utilization. The etiology of these comorbidities are likely multifactorial, but reinforce the importance of selecting a well-tolerated medication in persons with cardiometabolic disease. Lurasidone and ziprasidone (both need to be administered with food), followed by aripiprazole, may be the most cardiometabolically neutral antipsychotics. Lurasidone can also be efficacious in persons with concurrent mood disorders, whereas ziprasidone requires twice daily dosing and can significantly prolong the QTc interval, potentially limiting its utility. Aripiprazole has the additional benefit of being available as a long acting injectable formulation with different dosing frequencies. Starting an oral medication with a LAI formulation can make for an easy future medication conversion.

The stakes of medication adherence in schizophrenia are high, especially in FEP. These individuals have the most to gain with adherence, and the most to lose with non-adherence. Improving adherence is, in essence, prolonging life, and should be a critical focus of treatment. Although LAI antipsychotics have the potential to improve adherence, they are underutilized [21]. There is good evidence that outcomes favor LAIs over oral antipsychotics in early stage illness [21]. The FEP has a high treatment response rate regardless of which antipsychotic medication is selected, though response rates fall with each psychotic relapse [23]. This is an ideal time to select an antipsychotic with a favorable side effect profile, minimizing possible tolerability concerns that could trigger antipsychotic discontinuation.

Although side effect differences between individual antipsychotics are very apparent, efficacy differences are more subtle. Tiers of efficacy have emerged, with clozapine showing the most efficacy, followed by olanzapine, risperidone, and amisulpride, followed then by all remaining antipsychotic medications [5,6,12]. To maximize quality of life, some patients may be willing to live with varying side effects if the trade-off is a greater reduction in psychopathology. In patients with chronic schizophrenia, medications with added efficacy could be needed. Efficacy can be further enhanced in special situations, such as clozapine for TRS or agitation, or perhaps cariprazine for negative symptoms [15]. Ask the patient, or patient’s family, which symptom is causing the most distress.

In the future, pharmacogenetics may aid in personalizing treatment in schizophrenia, however, until then, clinicians should select treatment based on individual patient priorities. Special consideration should be given to clozapine and LAIs. Clozapine has consistently demonstrated superior efficacy to other medications, limits relapses due to inefficacy, and reduces both suicidality and all-cause mortality [26,27]. LAIs can improve adherence, especially when initiated with family or community support. There are several available LAIs, allowing for medication selection based on a desired side effect profile. Adherence is possibly the greatest correctable issue in the treatment of schizophrenia. Limiting relapse, especially in early stage illness, can result in a more favorable disease course.

Declaration of interest

J Faden has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Reviewer Disclosures

One referee declares having received honoraria and travel support from Otsuka-Lundbeck and Janssen Pharmaceuticals. They have also received research support from the Fundación Alicia Koplowitz and grants from the Spanish Ministry of Health and the Instituto de Salud Carlos III. Peer reviewers on this manuscript have no other relevant financial relationships or otherwise to disclose.

Additional information

Funding

This manuscript was not funded.