1. Introduction
While pharmacotherapy of depression appears to have moved from the monoamine hypothesis to more fertile grounds of glutamatergic and GABAergic mechanisms, it has been challenging to shift the antipsychotic paradigm beyond the dopaminergic hypothesis of schizophrenia. Efforts to develop non-dopaminergic antipsychotic medications (APMs) have produced negative results; thus, there has been no effective APM without dopamine involvement during 70 years of antipsychotic drug development. The only exception has been the approval of pimavanserin, a selective 5HT2A receptor blocker, to treat Parkinson’s psychosis but not schizophrenic psychosis.
Although Federal Drug Agency (FDA) has approved multiple APMs in the last 70 years, most have been with modest variations in molecular targets to qualify for “me too” drugs. The lack of significant differences in mechanisms of action explains why current APMs are primarily effective for positive symptoms without any measurable improvements in negative and cognitive symptoms. The antipsychotic response has been defined as only a 20% reduction in total scores on the most commonly used scale to assess schizophrenia symptoms, the Positive and Negative Syndrome Scale (PANSS)
[1]. This reduction in response criterion accommodated the limited efficacy of current APMs for negative and general schizophrenia symptoms, including some cognitive symptoms, such as orientation, attention, insight, and judgment. However, a 20% reduction in total PANSS scores may be mediated by a completely different set of schizophrenia symptoms from one to the other APM, making it difficult to interpret and compare antipsychotic efficacy with precision.
Major psychiatric disorders are diagnosed at a syndromic level, with various symptom domains having potentially different neurobiological underpinnings. Still, most psychiatric studies have used atheoretical diagnoses based on the Diagnostic Statistical Manual
[2][3] or the International Statistical Classification of Diseases and Related Health Problems (
https://icd.who.int, accessed on 6 December 2022), which could explain the failure to develop comprehensive treatments. In addition, inadequate sample sizes have failed to account for heterogeneity across schizophrenia patients, not allowing for sub-group analyses to generate testable hypotheses and guide future research. These diagnostic limitations were addressed by restricting federal funding for prespecified Research Domain Criteria (RDoC)
[4], which were vague, poorly defined, and not scientifically rigorous. Unfortunately, this measure was also counterproductive in promoting neurobiological research that could help advance psychiatric treatments.
Furthermore, researchers have failed to develop early predictors or intermediate phenotypes for antipsychotic response or tolerability, which could guide treatment options for individual patients. In addition, there is almost no research on finding the maintenance dose and length of antipsychotic therapy in schizophrenia patients recovering from acute psychosis. Similarly, there is a lack of laboratory measures to optimize antipsychotic response or tolerability, except for hyperprolactinemia (indicating excessive D2 receptor occupancy), therapeutic drug monitoring (TDM) for clozapine, and pharmacogenetic testing for refractory patients. Even though these measures can enhance antipsychotic efficacy and tolerability, particularly in treatment-refractory schizophrenia, they are not frequently employed, resulting in suboptimal use of APMs
[5]. In addition, some providers use high antipsychotic doses, especially in severely and persistently ill patients in state hospitals and prison systems, compromising tolerability and medication adherence and resulting in repeated hospitalizations
[5]. Furthermore, the antipsychotic doses required to manage acute psychosis and behavioral dyscontrol are not reduced even after the patient stabilizes. Several clinicians do not employ therapeutic drug monitoring for clozapine before labeling patients as clozapine non-responders
[6].
Since NIMH does not fund clinical trials, developing new APMs is primarily owned by the pharmaceutical industry. Although this process involves billions of dollars and may not be fundable by the NIMH, post-marketing and repurposing trials for already approved agents can be affordable, providing a less biased approach than industry-sponsored trials. Another limitation is that the results from the preclinical trials for drug approval cannot be applied to the general population as they are completed in a near-ideal patient population without comorbidities. Thus, it takes years to gather post-marketing effectiveness data applicable to the general patient population. Furthermore, prescriptions of newly approved APMs are too expensive to be covered by Medicaid to be used in the severely and persistently ill, homeless, and forensic population that needs these medications the most.
Although a detailed discussion of the statistical limitations of clinical trials is beyond the scope of this re
svie
archw, the statistics using
p < 0.05 has failed to translate into clinically relevant results
[7]. However, this trend is changing to more clinically applicable measures, such as confidence interval, effect size, number needed to treat (NNT)
[8], and number needed to harm (NNH)
[9]. In addition, an increasing level of placebo response may have also failed several promising psychotropics, including APMs
[10]. In this context, a higher placebo response can lower the chances of finding a significant difference (i.e., signal) between placebo and active treatment. Several strategies have been proposed to reduce placebo response, but none have been foolproof
[11].
However, despite all these obstacles in developing more effective APMs, currently available treatments have made some progress in tolerability, if not efficacy. Thus, relatively higher dose thresholds for adverse effects are observed with second-generation APMs, partial agonists, and especially the recently approved multimodal antipsychotic lumateperone. It is also worth mentioning that long-acting injectables (LAIs) have significantly impacted long-term clinical and functional outcomes in schizophrenia patients worldwide. This resviearchw provides a synopsis of different mechanism-based classes of antipsychotic medications approved to manage schizophrenia and other psychotic disorders over the last 70 years.
While pharmacotherapy of depression appears to have moved from the monoamine hypothesis to more fertile grounds of glutamatergic and GABAergic mechanisms, it has been challenging to shift the antipsychotic paradigm beyond the dopaminergic hypothesis of schizophrenia. Efforts to develop non-dopaminergic antipsychotic medications (APMs) have been disappointing, with none of the currently available APMs moving beyond the dopamine system during 70 years of antipsychotic drug development. The only exception has been the approval of pimavanserin, a selective 5HT2A receptor blocker, to treat Parkinson’s psychosis but not schizophrenic psychosis.
Although Federal Drug Agency (FDA) has approved multiple APMs over the last seven decades, most have been with modest variations in molecular targets to qualify for “me too” drugs. The lack of significant differences in mechanisms of action explains why current APMs are primarily effective for positive symptoms without any clinically meaningful improvements in negative and cognitive symptoms. Therefore, the antipsychotic response is defined as only a 20% reduction in total scores on the Positive and Negative Syndrome Scale (PANSS)
[1] instead of a 50% reduction required for an antidepressant response. This reduction in response criterion accommodated the limited efficacy of current APMs for negative and some of the cognitive symptoms assessed with PANSS, such as orientation, attention, insight, and judgment. A 20% reduction in PANSS scores suggests an antipsychotic response may be defined by a completely different set of symptoms from one patient to the other, making it difficult to interpret and compare antipsychotic efficacy with precision.
Major psychiatric disorders are diagnosed at a syndromic level, with various symptom domains having potentially different neurobiological underpinnings. However, most diagnoses in psychiatric research are based on atheoretical diagnostic tools such as the Diagnostic Statistical Manual
[2][3] or the International Statistical Classification of Diseases and Related Health Problems (
https://icd.who.int, accessed on 6 December 2022). The lack of neurobiologically-oriented diagnoses could explain the failure to develop more effective and comprehensive treatments. In addition, relatively small sample sizes have failed to account for heterogeneity across schizophrenia patients, not allowing for sub-group analyses to generate testable hypotheses and guide future research. The relative failure of DSM-based research was addressed by restricting federal funding to prespecified Research Domain Criteria (RDoC)
[4], which were also vague, poorly defined, and not scientifically rigorous. Unfortunately, this measure was also counterproductive in promoting neurobiological research that could help advance psychiatric treatments.
Thus, second-generation APMs, including partial agonists, may have higher dosing thresholds to cause adverse effects, such as extrapyramidal symptoms (EPS) and hyperprolactinemia. It is also worth mentioning that long-acting injectables (LAIs) have significantly impacted long-term clinical and functional outcomes in schizophrenia patients worldwide.
2. Older Antipsychotic Medications
Despite limited efficacy and frequent adverse effects, older or conventional APMs provided the first effective treatment for schizophrenia patients, who did not have any chance to live outside asylums and the prison systems. These APMs are further subclassified into high potency and low potency. The low-potency conventional APMs are generally less tolerable than the high-potency conventional APMs, due to multiple sites of action representing a “shotgun” approach
[12]. These actions include muscarinic, histaminic, and alpha-1 adrenergic receptor blockade. Antimuscarinic effects include dry mouth, blurred vision, urinary retention, constipation, tachycardia, loss of sweating, confusion, and worsening of closed-angle glaucoma and cognitive function. Antihistaminic effects translate into sedation and short-term weight gain, and the alpha-1 adrenergic blockade mediates postural hypotension, dizziness, and sedation.
In contrast, high-potency APMs are relatively more selective for blocking dopamine-2 (D2) receptors, and the main adverse effects associated with them are due to D2 blockade, especially at high doses. These adverse effects include extrapyramidal symptoms (EPS), hyperprolactinemia (
Figure 1 and
Figure 2), and treatment-induced, or worsening of, pre-existing negative, affective or cognitive symptoms.
Figure 1.
Extrapyramidal Symptoms (EPS) with antipsychotic medications.
Figure 2.
Hyperprolactinemia with antipsychotic medications.
However, the most clinically-serious adverse effect of low-potency APMs is the effects on QTc prolongation, which can result in sudden cardiac death
[13]. Although the high-potency APMs also prolong QTc interval, the most clinically serious effects include tardive dyskinesia
[14] and neuroleptic malignant syndrome
[15]. These adverse effects are generally reported with long-term use of high-dose antipsychotic pharmacotherapy
[15]. However, some of the adverse effects of low-potency conventional APMs may be beneficial in few patients, including sedation and weight gain with the histamine-1 receptor blockade, reduction in pre-existing hypertension with the alpha-1 receptor blockade, and protection from EPS with the built-in anticholinergic effects. These adverse effects are also witnessed with newer APMs, such as olanzapine and clozapine
[16]. Of note, none of the APMs is indicted for managing dementia-related psychosis due to increased risk for mortality in older adults
[17].
There is one APM, loxapine, that is not only effective in schizophrenia but also in major depressive disorder because loxapine is demethylated to a tetracyclic metabolite, amoxapine, which has antidepressant properties
[18]. Another worth-mentioning APM, molindone, has a moderate affinity for D2 receptors, with potential benefits in patients not responding to high-or low-affinity APMs
[19]. It is also interesting that one of the older high-potency APM, pimozide is the only FDA-approved treatment for Tourette’s syndrome
[20].
One of the most clinically significant advances in antipsychotic treatments has been the development of long-acting injectables (LAI) with high-potency APMs, haloperidol, and fluphenazine, followed by newer APMs, risperidone, paliperidone, and aripiprazole. Since medication nonadherence is commonly observed in schizophrenia subjects, the LAIs have improved the maintenance of antipsychotic response and prevention of psychotic relapse and rehospitalizations
[21]. In addition, these LAIs are now also available with some of the newer generation APMs as discussed below.
In summary, serotonin and dopamine antagonism (SDA) with SGAPMs may have provided some cushion from dose-related adverse effects but without any significant benefits in efficacy. The only exception among SGAPMs is clozapine, which remains the gold standard for managing treatment-refractory schizophrenia (TRS). However, despite its unique efficacy, to
our
esearchers' knowledge, clozapine has not been compared with other SGAMs or studied in post-marketing effectiveness or efficacy trials because it is only approved for patients with TRS.
3. Second-Generation Antipsychotic Medications (SGAPMs)
This class includes every APM developed after the conventional APMs despite significant differences in the mechanisms of action (
Table 1), which has created confusion and misperceptions.
Table 1.
Molecular targets and binding affinities (Ki, nM) for antipsychotic medications.