Schizophrenia: A Telltale of the Brain’s Complexity

At a recent meeting, a colleague asked me: “I am treating a person with schizophrenia that is still psychotic on 100 mg of haloperidol. What should I add to this?” My metaphorical jaw dropped. I pictured this person, assuming adherence to the haloperidol, having a frontal cortex being blasted by dopamine-2 receptor (D₂R) blockade, a prolactin level that was likely through the roof, and at risk of movement disorders (if not already present)—all while not receiving any benefit of the primary symptom that this extreme dose of haloperidol was being used for: treating psychosis.

Regrettably, questions like this are not all that uncommon in the frontline treatment of individuals with schizophrenia. Neuropsychiatric research over the past 50 years has exploded with specialties and subspecialties that continue to sculpt an ever-more-complex model of the human brain. Unsurprisingly, there have been parallel advances in our understanding of the complex interplay of wide-ranging brain vulnerabilities that increase the risk of a neurodevelopmental process resulting in a first psychotic episode of schizophrenia. It is challenging for the busy clinician to keep up.

How Much D2R Blockade Is Enough?

I thought back to 2 of my favorite publications from the 1990s by Kapur and colleagues,^1,2 the pioneers who radically altered our understanding of dosages of haloperidol that provided the percentage D₂R occupancy in the live human brain necessary to effectively treat psychotic symptoms. In their 1996 study,¹ 7 individuals with schizophrenia were treated for 14 days with 2 mg/day of oral haloperidol. Dopamine D₂R occupancy on day 14 was determined using PET scan brain imaging with the radioligand [¹¹C]raclopride. The percentage D₂R occupancy in these individuals ranged from 53% to 74%, with 5 subjects demonstrating significant clinical improvement, and none reporting significant adverse effects.

In their 1997 study,² they investigated the relationship between D₂R occupancy and haloperidol plasma levels. Twelve individuals with schizophrenia were treated with from 1 to 5 mg/day of haloperidol and D₂R occupancy was again measured with [¹¹C]raclopride-PET while haloperidol plasma levels were determined with gas chromatography–mass spectrophotometry. D₂R occupancy ranged from
53% to 88% with corresponding plasma levels ranging from 0.5 to 5.8 ng/mL. The authors concluded²:

“Our findings demonstrate that 2-5 mg/day of haloperidol, which usually leads to plasma levels of 1-2 ng/ml, would be expected to induce 60-80% dopamine D2 receptor occupancy. If, as has been claimed, 70% D2 occupancy is adequate for typical neuroleptic response, then the conventional use of > 10 mg/day may have been too high, since 70% occupancy can be achieved in most patients by 2-5 mg/day.”

I recall the excitement of these results in the late 1990s and the revised recommendation that haloperidol ideally be dosed in the range of 5 to 10 mg/day. Remarkably, when I looked at the National Library of Medicine’s StatPearls review of haloperidol, which was updated in September 2023, it concluded that for schizophrenia “the maximum recommended oral dosage should not exceed 100 mg/d.”³

Psychiatry’s preoccupation, arguably its irrational addiction, to the belief that the only way to treat schizophrenia is through D2R antagonism needs to evolve as the science already has. D2R antagonism improves psychotic symptoms in only a subset of individuals with schizophrenia; in others there is little or no benefit, but the individual does still receive the unwanted D2R antagonism adverse effects of worsening of negative and cognitive symptoms, prolactin elevation, sedation, movement disorders, and weight gain to name a few of the adverse effects.

It is well accepted that approximately a third of individuals diagnosed with schizophrenia will meet criteria for treatment resistance from usual antipsychotics. The only medication Food and Drug Administration (FDA) approved for these individuals is clozapine, which antagonizes D₂Rs much more weakly than most antipsychotics, but whose mechanism of action remains unknown. Siskind et al reviewed data from a clozapine systematic review and meta-analysis that concluded that only 40% of treatment-resistant individuals respond to clozapine.⁴ Unsurprisingly the voluminous research into the neuropathophysiology of schizophrenia implicates numerous neurotransmitters beyond dopamine. So, let’s take a look at our current understanding of schizophrenia.

Neurotransmitters Associated With Schizophrenia⁵

Dopamine

When chlorpromazine arrived in the United States in 1954, it revolutionized the treatment of schizophrenia, leading to a wave of deinstitutionalization. Its mechanism for improving the positive symptoms in schizophrenia was initially unknown but was subsequently demonstrated to be the result of antagonizing D₂R. This was a paradigm changer but resulted in a distraction from addressing the negative and cognitive symptoms of the schizophrenia triad.

Acetylcholine

Only 3 years later, in 1957, Pfeiffer and Jenney, 2 researchers at Emory University in Georgia, published a provocative study recording 23 trials of injecting arecoline (the pan–muscarinic acetylcholine receptor [mAChR] agonist) mixed with methyl atropine (the pan-mAChR antagonist that minimally crosses the blood-brain barrier) subcutaneously in the same syringe in untreated, institutionalized individuals with schizophrenia.⁶ They reported the “antischizophrenic action of arecoline begins 1 to 2 minutes after injection, reaches a peak in 5 minutes, and disappears completely in 15 to 20 minutes.” The authors concluded: “Research studies should be extended to include the examination of naturally occurring metabolites that have a strong central muscarinic action, since these metabolites may be more antischizophrenic than the presently available tranquilizing drugs.” Xanomeline, a selective M1/M4 muscarinic cholinergic agonist which is a synthetic analogue of arecoline, was finally FDA approved for the treatment of adults with schizophrenia 67 years later in 2024.

Glutamate

Glutamate has strong evidence of contributing significantly to the pathophysiology of schizophrenia.⁷ In 1991, Javitt and Zukin reported that the dissociative anesthetics phencyclidine and ketamine, both sharing the property of antagonism of the N-methyl-_D-aspartate glutamate receptor (NMDAR), will mimic the positive, negative, and cognitive symptoms of schizophrenia in healthy volunteers, as well as exacerbate these symptoms in individuals with schizophrenia if dosed high enough and administered regularly.⁸ This finding gave rise to the ongoing hypothesis that inherent dysfunction of NMDARs, which could occur due to wide-ranging etiologies, was involved in the pathophysiology of schizophrenia. In contrast to this, amphetamines, which increase dopamine and norepinephrine in the brain, only exacerbate the positive symptoms.

Supporting NMDAR dysfunction as a possible contributing etiology of schizophrenia, in 2007, Josep Dalmau et al described anti-NMDAR autoimmune encephalitis, which could present with the abrupt onset of symptoms that mimic schizophrenia and other psychiatric disorders.9,10 Pharmacological research over the past several decades has studied target molecules that interface with the NMDAR, a complex receptor with multiple subunits and genetic heterogeneity. These classes of medications that have been extensively studied but ultimately failed for a range of reasons include metabotropic glutamatergic receptor agonists, glycine transporter type 1 inhibitors, and D-amino acid oxidase inhibitors.⁷

Other Neurotransmitters

Additional neurotransmitter systems are implicated in the etiology of schizophrenia, and merit further study by the interested reader. These include serotonin (especially the 5-HT_2A receptor), γ-aminobutyric acid (GABA) receptors (related to optimizing the synergistic excitatory-inhibitory balance of glutamate and GABA for healthy circuitry function), endocannabinoids, and TAAR1 agonists, as well as others. This vast untapped reservoir of potential drug targets generates optimism as to what the future may bring.

Genetic Associations

Schizophrenia is known to be highly heritable. Monozygotic twin studies consistently report a 50% concordance for schizophrenia.¹¹ Currently approximately 300 common gene variants have been identified by genome-wide association studies, as well as over 20 rare variants, that collectively increase the risk for schizophrenia.¹² This finding supports the hypothesis that the genetic risk for developing schizophrenia involves wide-ranging complex mechanisms and is highly polygenic.

Singh et al¹³ meta-analyzed the whole exomes of 24,248 individuals with schizophrenia and 97,322 controls and identified 10 genes that likely contribute a significant risk for the development of schizophrenia, with ORs ranging from 3 to 50. Notably, these genes were quite rare in the individuals with schizophrenia and were designated as ultrarare coding variants (URVs). Significantly, these 10 URVs were highly expressed in neurons in the central nervous system and were involved in diverse processes related to the structure, formation, and function of synapses. Two of these URVs, the gene GRIN2A, which codes for an NMDAR subunit and the gene GRIA3, which codes for an α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor subunit, support involvement of the glutamate system in the pathogenesis of schizophrenia.

Genetic research to date has supported a shared genetic vulnerability for schizophrenia, bipolar disorder, and childhood neurodevelopmental disorders.¹⁴

Structural Brain Abnormalities

In 1976, Johnstone et al published the seminal paper reporting the finding that CT scan imaging of the brains of individuals with schizophrenia who were institutionalized had increased volumes of their cerebral ventricles compared with age-matched controls. Within the subjects with schizophrenia, the increased ventricular size highly correlated with the degree of cognitive impairment.15 As neuroimaging has expanded to include a wide range of imaging devices, the database of brain structures correlated with disease states has exploded. The Enhancing Neuro Imaging Genetics through Meta-Analysis (ENIGMA) Consortium began in 2009 and has developed into an international collaborative committed to combining many global datasets with the goal of establishing reliable large-scale biomarkers. The Schizophrenia Working Group has published multiple articles since 2016 summarizing their findings comparing structural brain differences in individuals with schizophrenia and healthy controls. Table 1 summarizes the findings from 3 of these publications.^16-18 Collectively, clear structural changes are present, but none of these structural findings have risen to the level of clinical actionability as biomarkers.

Factors Increasing the Risk

In addition to neurotransmitters, genetics, and brain structure variability, there are many developmental and environmental events, exposures, and factors that can converge to increase the likelihood of a particular individual to develop schizophrenia. Table 2 highlights a partial list of possible risk factors in the development of schizophrenia.

Concluding Thoughts

Schizophrenia is simultaneously a neurodevelopmental and neurodegenerative disorder, appearing to result from a large heterogeneity of accumulated risk factors that ultimately can converge to cross the threshold onto a first psychotic episode. It should be no surprise that an individual’s treatment plan will be unique and include trial and error, as well as vigilance to emerging information and treatments.

My recommendations to my colleague whose question began this article were as follows:

• Obtain a 12-hour postdose haloperidol level. This will provide information about possible adherence issues, could document a haloperidol level that confirms an adequate dose for an adequate time with no benefit, and leaves a bookmark in the chart as to the haloperidol level at that time in treatment.
• Obtain a serum prolactin level. With such a high dose of haloperidol, I would like to know what this level is and assess for possible medical complications if elevated.
• Obtain other appropriate blood work and a urine toxicology screen.
• Order a brain MRI or CT if one was never done.
• Review the history, records, labs, and recent physical evaluation, medication history, and substance use history, and speak with family or friends for more information if possible.
• If the diagnosis remains consistent with schizophrenia and the haloperidol level was in the accepted treatment range, consider a trial of clozapine or KarXT (Cobenfy).

We have much to learn about the pathophysiology and novel treatments for the brain disorder that we call schizophrenia. Its current complex biography is a small telltale of the larger story of the remaining mysteries awaiting to be discovered about the structure and function of the healthy human brain.

References

1. Kapur S, Remington G, Jones C, et al. High levels of dopamine D2 receptor occupancy with low-dose haloperidol treatment: a PET study. Am J Psychiatry. 1996;153(7):948-950.

2. Kapur S, Zipursky R, Roy P, et al. The relationship between D2 receptor occupancy and plasma levels on low dose oral haloperidol: a PET study. Psychopharmacology (Berl). 1997;131(2):148-152.

3. Rahman S, Marwaha R. Haloperidol. StatPearls. Updated September 1, 2023. Accessed October 21, 2026. https://www.ncbi.nlm.nih.gov/sites/books/NBK560892/

4. Siskind D, Siskind V, Kisely S. Clozapine response rates among people with treatment-resistant schizophrenia: data from a systematic review and meta-analysis. Can J Psychiatry. 2017;62(11):772-777.

5. Howes OD, Bukala BR, Beck K. Schizophrenia: from neurochemistry to circuits, symptoms and treatments. Nat Rev Neurol. 2024;20(1):22-35.

6. Pfeiffer CC, Jenney EH. The inhibition of the conditioned response and the counteraction of schizophrenia by muscarinic stimulation of the brain. Ann N Y Acad Sci. 1957;66(3):753-764.

7. Egerton A, Grace AA, Stone J, et al. Glutamate in schizophrenia: neurodevelopmental perspectives and drug development. Schizophr Res. 2020;223:59-70.

8. Javitt D, Zukin S. Recent advances in the phencyclidine model of schizophrenia. Am J Psychiatry.1991;148(10):1301-1308.

9. Barry H, Byrne S, Barrett E, et al. Anti-N-methyl-d-aspartate receptor encephalitis: review of clinical presentation, diagnosis and treatment. BJPsych Bull. 2015;39(1):19-23.

10. Al-Diwani A, Handel A, Townsend L, et al. The psychopathology of NMDAR antibody encephalitis in adults: a systematic review and phenotypic analysis of individual patient data. Lancet Psychiatry. 2019;6(3):235-246.

11. Zhang H, Khan A, Kushner SA, Rzhetsky A. Dissecting schizophrenia phenotypic variation: the contribution of genetic variation, environmental exposures, and gene-environment interactions. Schizophrenia (Heidelb). 2022;8(1):51.

12. Sullivan PF, Yao S, Hjerling-Leffler J. Schizophrenia genomics: genetic complexity and functional insights. Nat Rev Neurosci. 2024;25(9):611-624.

13. Singh T, Poterba T, Curtis D, et al. Rare coding variants in ten genes confer substantial risk for schizophrenia. Nature. 2022;604(7906):509-516.

14. Owen MJ, Legge SE, Rees E, et al. Genomic findings in schizophrenia and their implications. Mol Psychiatry. 2023;28(9):3638-3647.

15. Johnstone EC, Crow TJ, Frith CD, et al. Cerebral ventricular size and cognitive impairment in chronic schizophrenia. Lancet. 1976;2(7992):924-926.

16. van Erp TG, Hibar DP, Rasmussen JM, et al. Subcortical brain volume abnormalities in 2028 individuals with schizophrenia and 2540 healthy controls via the ENIGMA consortium. Mol Psychiatry. 2016;21(4):547-553.

17. van Erp TGM, Walton E, Hibar DP, et al. Cortical brain abnormalities in 4474 individuals with schizophrenia and 5098 control subjects via the Enhancing Neuro Imaging Genetics Through Meta Analysis (ENIGMA) consortium. Biol Psychiatry. 2018;84(9):644-654.

18. Omlor W, Rabe F, Fuchs S, et al. Estimating multimodal structural brain variability in schizophrenia spectrum disorders: a worldwide ENIGMA study. Am J Psychiatry. 2025;182(4):373-388.