Secondary psychoses: an update

Abstract

Psychotic disorders due to a known medical illness or substance use are collectively termed secondary psychoses. In this paper, we first review the historic evolution of the concept of secondary versus primary psychosis and how this distinction supplanted the earlier misleading classification of psychoses into organic and functional. We then outline the clinical features and approach to the diagnosis of secondary psychotic disorders. Features such as atypical presentation, temporal relation to detectable medical cause, evidence of direct physiological causal relationship to the etiological agent, and the absence of evidence of a primary psychotic illness that may better explain the presentation suggest consideration of a secondary psychosis. Finally, we discuss how careful studies of secondary psychotic disorders can help elucidate the pathophysiology of primary, or idiopathic, psychotic disorders such as schizophrenia. We illustrate this issue through a discussion of three secondary psychotic disorders — psychoses associated with temporal lobe epilepsy, velocardiofacial syndrome, and N-methyl D-aspartate (NMDA) receptor encephalitis — that can, respectively, provide neuroanatomical, genetic, and neurochemical models of schizophrenia pathogenesis.

Classification of psychotic disorders has continued to generate debate. The idea that psychoses may be categorized into those with detectable anatomic pathology (organic) and those without such pathology (functional) has been prevalent for more than a century 1. Unfortunately, this distinction sometimes misled the field, particularly during the early decades of the 20th century, when functional was equated with psychogenic, and the etiology of schizophrenia was ascribed to psychologic factors such as parental upbringing.

Recent decades have seen a growing realization that it is more useful to categorize psychiatric disorders as secondary, when the symptoms are due to a known medical illness or substance use, and as primary (or idiopathic), if the symptoms cannot be explained by another cause. In this context, the DSM-IV dropped the terms organic and functional used in the earlier editions 2. The emphasis was shifted from the presence or absence of discernible brain pathology (which is often difficult to identify, even in many neurological disorders) to etiology (known, presumed, or unknown). The DSM-IV also made the distinction between psychotic disorders secondary to medical illness versus those secondary to substance use.

In this paper, the primary versus secondary approach has been taken to understand psychotic disorders, since the term primary has the advantage of not ruling out a neurobiological basis. While a variety of secondary psychotic disorders are described throughout this text, it is important to have an appreciation of how such understanding can elucidate the neurobiological substrates that might underlie primary (or idiopathic) psychotic illness. We seek to briefly provide an overview of what is known in this regard and offer an approach to the differential diagnosis between primary and secondary psychotic disorders.

Disorders presenting with secondary psychotic symptoms

Virtually any substance, prescribed drug, or medical condition affecting nervous system function can present with psychiatric symptoms, including psychosis (Table 1). The mnemonic TACTICS MDS USE may offer the clinician an easy way to remember the main groups of disorders to consider in the differential diagnosis.

Table 1.

Possible causes of secondary psychoses

	Examples	Investigations
Trauma	Traumatic head injury	CT, MRI
Autoimmune disorders	Systemic lupus erythematosus, NMDA receptor encephalitis	Autoantibody titers
Cytogenetic/congenital disorders	Velocardiofacial syndrome, agenesis of corpus callosum	Karyotyping, MRI
Toxic/substance-induced disorders	PCP, MDMA, LSD, cannabis, alcohol Lead, mercury or arsenic poisoning	Careful medication history; urine screen for drugs, heavy metal screen; trial off the offending agent
Iatrogenic disorders	Antimalarials, steroids, isoniazid	Careful medication history; trial off the offending agent
Cerebrovascular disorders	Stroke, subdural hematomas	CT, MRI
Space-occupying disorders	Cerebral tumors	CT, MRI
Metabolic disorders	Phaeochromocytoma, metachromatic leukodystrophy, Wilson's disease	Urinary catecholamines; arylsulphatase-A levels, copper and ceruloplasmin levels
Dietary disorders	Pellagra, B12 deficiency; vitamin D deficiency	B12, Folate, D3 levels
Sepsis/infectious disorders	Neurosyphilis, toxoplasmosis, HIV disease	RPR to rule out syphilis; HIV antibody titers; glucose, protein in CSF
Unknown cause/degenerative/demyelinating disorders	Lewy body dementia, Parkinson's disease, Huntington's disease, multiple sclerosis, Fredreich's ataxia	MRI, CT, EEG, evoked potentials
Seizure disorders	Partial complex seizures, temporal lobe epilepsy	EEG, including sleep deprivation; telemetric EEG as indicated
Endocrine disorders	Hyperthyroidism, hypothyroidism, hyperparathyroidism	Serum calcium, thyroid/parathyroid hormone levels

CT – computed tomography; MRI – magnetic resonance imaging; NMDA – N-methyl D-aspartate; PCP – phencyclidine; MDMA – 3,4-methylenedioxy-N-methylamphetamine; LSD – lysergic acid diethylamide; RPR – rapid plasma reagin; HIV – human immunodeficiency virus; CSF – cerebrospinal fluid; EEG – electroencephalography

Traumatic brain injury

Traumatic brain injury (TBI) has been proposed as a risk factor for schizophrenia-like psychosis, though there have been few systematic studies of the relationship between these two conditions.

Fujii and Ahmed 3 conducted a review of case reports in which they retroactively applied DSM-IV criteria to a total of 69 cases and concluded that TBI can be either a primary cause of psychosis or contribute to the development of psychosis through inducing seizures, though this study was limited by the heterogeneity of the case reports. A cohort study of 3552 Finnish World War II veterans 4 reported a rate of psychosis of 8.9% following TBI, but the proportion of open injuries, potential comorbidities, and lack of standard diagnostic tools call into question the generalizability of these findings.

Evidence from additional studies, including epidemiologic and case control studies, indicates that TBI may marginally increase the risk of psychosis, though the increased risk may be significantly higher for individuals with a genetic predisposition to psychosis 5–7. The risk of psychosis appears to be elevated when TBI is severe, diffuse, involves the frontal and temporal lobes, and is associated with abnormal findings on electroencephalography and neuroimaging investigations 8–10. When psychosis does occur in the setting of TBI, it is characterized by prominence of persecutory delusions and auditory hallucinations, with a relative paucity of negative symptoms 3,8. Given this evidence, further research examining the interaction between TBI and genetic risk factors for schizophrenia is warranted.

Autoimmune disorders

Psychotic symptoms are a rare, but known, manifestation of systemic lupus erythematosus (SLE), with a reported prevalence ranging from 1 to 11% 11–13. Retrospective analyses reveal that, when psychosis does occur in SLE, it tends to appear early in the disease course. Approximately 30–60% of the patients found to have psychosis were psychotic when diagnosed with SLE and up to 80% developed these symptoms within 1 year of SLE diagnosis 13–15. While psychosis can be the presenting symptom in SLE, it occurs more frequently in the context of other lupus symptoms, most commonly cutaneous involvement and arthritis 13.

Symptomatically, lupus psychosis is frequently characterized by paranoia, auditory and visual hallucinations, and delusions of grandeur 14. Immunosuppressive treatment, including steroids and antimalarials, typically results in resolution of the psychotic symptoms, although these medications themselves can induce psychotic symptoms and contribute to diagnostic difficulty 14.

Autoimmune disease as a cause of psychotic symptoms is not limited to SLE. Psychosis may be the presenting symptom in multiple sclerosis 16,17 and has been documented to occur in Hashimoto's disease 18,19.

Congenital/cytogenetic disorders

Velocardiofacial syndrome (VCFS), discussed in further detail below, is caused by a heterozygous chromosome 22q11.2 deletion and is associated with the strongest link yet identified between a genetic condition and psychosis, with 25–30% of VCFS patients developing symptoms analogous to schizophrenia 20–23.

VCFS is of particular interest because the neurocognitive findings, brain structural abnormalities, psychotic symptoms, and patterns of treatment response seen in this disorder are all highly similar to those seen in schizophrenia 24.

Other genetic disorders in which psychosis can occur include Prader-Willi syndrome (in which the rate of schizophrenia-like psychosis is approximately 16%) 25,26, Huntington's disease (rate of 5–16%) 27,28, and Fahr's disease/syndrome 29,30.

Toxic/drug-induced disorders

Individuals using psychoactive substances can experience psychotic symptoms in various contexts, including acute intoxication, withdrawal, delirium induced by either intoxication or withdrawal, substance-induced mood disorder with psychotic features, and substance-induced psychosis (SIP). SIPs are best conceptualized as those conditions where the psychosis begins in the context of substance use but persists for days to weeks in the absence of continued substance use.

The substances with the clearest psychotogenic properties include stimulants (amphetamine, cocaine) 31–33, cannabis 34,35, and the psychotomimetics (phencyclidine, ketamine) 36,37. Opiates and nicotine have not been clearly shown to produce psychosis, while alcohol and benzodiazepines may induce psychosis just in acute withdrawal states 33,38,39. Lysergic acid diethylamide and 3,4-methylenedioxy-N-methylamphetamine (MDMA) can produce hallucinations in acute intoxication, but there is no clear evidence that they induce an ongoing psychotic disorder 40,41.

The interaction between susceptibility to psychotic disorders and to substance abuse disorders is complex, with evidence suggesting both that genetic vulnerability to psychosis in combination with drug abuse can bring about psychosis and that individuals with such vulnerability may be more likely to abuse substances.

Toxic psychosis can also occur in the context of exposure to heavy metals, including lead 42, mercury, and arsenic 43.

Iatrogenic psychoses

Toxic psychosis can be caused by numerous centrally active medications and is often characterized by acutely impaired cognitive function in addition to psychotic symptoms.

The reported rates of psychotic symptoms in patients receiving glucocorticoids have varied widely, although there is evidence that mood disorders secondary to glucocorticoids are associated with psychotic symptoms more frequently than are primary mood disorders. While onset of psychotic symptoms usually occurs within several days of initiating glucocorticoid treatment, the symptoms can occur anytime from hours to weeks after the first dose is administered 44.

Elderly patients are at particular risk for developing toxic psychosis in reaction to medications with anticholinergic properties 45,46 and to benzodiazepines 47,48. The development of psychotic symptoms following initiation of isoniazid treatment has been documented in a growing number of case reports 49, and ascribed to alteration in catecholamine and serotonin levels via monoamine oxidase inhibition. Antimalarial drugs including chloroquine and mefloquine have been linked to psychotic symptoms, with evidence that the risk is higher in patients with a history of psychiatric illness 50,51. Additional classes of medications that have been linked to toxic psychosis include antidepressants 52, anticonvulsants 53, antiemetics 54, antiparkinsonian agents 55, antipsychotics 47, opioids 47, histamine antagonists 56, and antibiotics 57, particularly when toxic blood levels occur.

Cerebrovascular disorders

Post-stroke psychosis prevalence rates are estimated at 3–4% 58,59. While psychotic symptoms have occurred in association with strokes in numerous brain regions, they are most common in temporo-parietal-occipital lesions 59. The nature of the psychotic symptoms differs from schizophrenia, with post-stroke psychosis more likely to include visual, tactile, and olfactory hallucinations 60–62. Hallucinatory behavior in post-stroke patients has been further classified into frank hallucinations and hallucinosis, the latter distinguished by ego-dystonia and retained insight into the fact that the perceptions are not real 63.

There has been little systematic study of the link between vascular dementia and the development of psychotic symptoms, though the fact that psychosis occurs at similar rates in both vascular and Alzheimer's dementia argues against a psychotogenic mechanism specific to the cerebrovascular disturbance 64.

Studies that have searched for evidence of cerebrovascular injury in patients who develop late-onset psychotic symptoms have produced conflicting results 65. Based on the limited data currently available, it is difficult to conclude to what extent cerebrovascular disease predisposes a patient to the development of psychotic symptoms, and further research is needed to confirm or disprove this link.

Space-occupying intracranial disorders

Brain tumors are an uncommon, but important, cause of secondary psychosis, and there is evidence that the prevalence of intracranial tumors is increased in patients with psychiatric illness 66–68. Symptomatically, psychosis secondary to intracranial tumors can be indistinguishable from schizophrenia, although it is more commonly associated with visual hallucinations, simple unelaborated delusions, and absence of formal thought disorder 69–72.

Tumors located in the temporal lobes or limbic structures are the most likely to produce psychosis, with one study demonstrating that 20% of tumors in the temporal lobe resulted in psychotic symptoms 69,73. There has been no demonstrated link between the histological type of tumor and the frequency of psychosis, though low-grade, slow-growing tumors seem most likely to produce psychotic symptoms in the absence of neurological signs 74.

Neuroimaging with either computed tomography (CT) or magnetic resonance imaging (MRI) is recommended in older patients presenting with new-onset psychosis and in patients who present with focal neurological findings on exam 75.

Metabolic disorders

Disordered neural connectivity, the putative mechanism for the symptoms of schizophrenia, can also occur in metabolic disorders that result in disrupted neuronal function or neuronal death. For example, although lysosomal storage disorders typically produce early severe neurological deficits and often death, there are adolescent or adult forms of these disorders that are associated with secondary psychosis. These disorders, such as Niemann-Pick disease type C (NPC), Tay Sachs disease, and alpha mannosidosis, likely produce psychotic symptoms through the interaction of the neuropathological processes with neurodevelopmental changes including synaptic pruning, myelination, and changes in connectivity 76–78. A similarly disordered interplay of functional connectivity and neurodevelopment is seen in the leukodystrophies, with the late-onset form of metachromatic leukodystrophy (MLD) as the prototypical example of the link between this group of disorders and secondary psychosis 76,79.

Psychotic symptoms have been reported to occur in up to 50% of patients with late-onset NPC and MLD 76. In addition, there have been case reports of patients with mitochondrial disorders presenting with psychotic symptoms, and it is thought that the leukodystrophy that can occur in mitochondrial disorders is implicated in the psychosis 80–82.

Wilson's disease, involving abnormal deposition of copper in the liver and brain, has reported prevalence rates of psychosis ranging from 2 to 11% 83,84. The psychosis in Wilson's disease is characterized not only by typical symptoms of hallucinations, delusions, and thought disorders, but also by a myriad of additional symptoms, including euphoria, sexual preoccupation, hebephrenia, and catatonia 85.

Dietary disorders

The link between nutritional deficiencies and psychiatric symptoms has been investigated and debated for decades. There is evidence that psychiatric symptoms of cobalamin deficiency may occur without evidence of hematological or neurological abnormalities 86,87. Case reports of psychosis secondary to cobalamin deficiency describe symptoms of persecutory delusions, auditory and visual hallucinations, disorganized thought processes, and psychomotor agitation that were unresponsive to treatment with antipsychotic medications but resolved completely after parenteral cobalamin therapy 88–90. Additional evidence in support of this connection includes a study in which depressed patients with psychotic depression had significantly lower cobalamin levels than those with nonpsychotic depression 91.

Although the link between folate deficiency and psychosis is tenuous, there is some evidence that folate supplementation in patients with schizophrenia enhances recovery 92. Psychotic symptoms including auditory hallucinations, persecutory delusions, and delusional parasitosis can very rarely develop in the setting of pellagra (niacin deficiency), though this occurs almost exclusively in chronic alcohol abuse 93,94.

Sepsis/infectious diseases

Historically, neurosyphilis has been closely tied to psychiatric hospitalization. In 1900, approximately 5% of institutionalized mental patients were diagnosed with general paresis of the insane 95. Syphilis infection rates have increased since the appearance of human immunodeficiency virus (HIV), and patients presenting with exclusively psychiatric symptoms in the setting of neurosyphilis have been noted 96. Such patients can present with both affective and psychotic symptoms that are indistinguishable from primary psychiatric disorders 97. Treatment of neurosyphilis with antibiotics, and adjunctive antipsychotic medications as necessary, typically halts, but does not reverse, mental status deterioration due to neuronal loss 98, although there are case reports of significant clinical improvement 99,100.

Estimates of new-onset psychosis associated with HIV infection have ranged between 0.23% and 15%, with symptoms generally presenting either in late-stage HIV or when patients have transitioned to acquired immunodeficiency syndrome (AIDS) 101,102. Symptomatically, HIV-related psychosis is characterized by persecutory, grandiose, and somatic delusions, with hallucinations as a second prominent symptom cluster 103. The effect of treatment with highly active antiretroviral therapy on psychotic symptoms remains unclear, and is further complicated by the fact that side effects of antiretrovirals include hallucinations 102. Although antipsychotics are generally effective in HIV psychosis, patients with HIV are at increased risk of developing extrapyramidal symptoms and tardive dyskinesia, particularly with the use of typical antipsychotics 101,104.

Additional infectious causes of secondary psychosis have been postulated, including infection with Toxoplasma gondii. This link is based on the fact that studies have demonstrated an increased prevalence of antibodies to Toxoplasma gondii in patients with schizophrenia 105,106. Moreover, toxoplasmosis has been linked to psychotic symptoms including delusions and auditory hallucinations, even in the absence of concurrent AIDS 107,108.

Finally, there is increasing evidence to support prenatal infections as potential risk factors for schizophrenia. Although still inconclusive, prenatal infections with influenza, toxoplasmosis, rubella, herpes simplex virus, and syphilis have been associated with the development of secondary psychosis 109–113.

Unknown cause/degenerative/demyelinating disorders

Although the data on psychosis in multiple sclerosis (MS) have been inconsistent, recent evidence suggest that MS does increase the likelihood of developing psychotic symptoms 17. A large, population-based study from Canada reported that 2–4% of patients with MS became psychotic 114. Delusions are the primary psychotic symptom observed in MS, while hallucinations and negative symptoms are rarely seen 115. Reports of a temporal correlation between MS symptoms and psychosis, however, have been conflicting 74,115. Psychotic symptoms may be related to increased lesion burden in the periventricular white matter and temporal horns, though a precise mechanism remains to be elucidated 116. Chronic psychosis due to MS seems to be rare.

Estimates of the prevalence and incidence of psychotic symptoms in Alzheimer's disease (AD) have varied, with one review of 55 studies reporting an overall prevalence of 41%, consisting of 36% of patients with delusions and 18% with hallucinations 117. The first three years of AD are characterized by an increasing incidence of psychotic symptoms, after which there seems to be a plateau 117,118. Psychotic symptomatology in AD most often consists of delusions, typically of theft or suspicion, visual hallucinations more frequently than auditory hallucinations, and misidentifications 119. These symptoms are typically coincident with other psychiatric symptoms including aggression, agitation, apathy, and depression 120.

While psychotic symptoms appear to be less common in the fronto-temporal dementias, they have been noted in approximately 20% of patients, with higher rates in particular subtypes 121,122.

Seizure disorders

Schizophrenia-like psychosis has been associated with epilepsy for more than a century, and there is strong evidence that chronic psychosis occurs more frequently in patients with epilepsy than in the general population. Psychotic syndromes in epilepsy have traditionally been categorized based on their temporal association with clinical seizures. Ictal psychoses represent psychotic symptoms occurring in the context of an active nonconvulsive seizure, tend to last for minutes to hours, and consist of prominent hallucinations and paranoid delusions 65. Post-ictal psychoses are brief psychotic episodes that typically occur hours to days following a seizure cluster, consist of delusions, hallucinations, and affective symptoms, and generally resolve within several days 123,124. The development of a chronic psychosis associated with epilepsy is also well documented, with a recent study reporting a relative risk of 2.48 125,126.

Phenomenologically, psychosis associated with epilepsy is difficult to distinguish from schizophrenia, as suggestions that it is characterized by a more benign course and relative lack of negative symptoms have not been verified in the literature 65,125. Risk factors for developing secondary psychosis include a more severe form of epilepsy with multiple seizure types 127, history of status epilepticus 125, and resistance to medication treatment 65. The preferential, though not exclusive, association of temporal lobe epilepsy and the development of psychotic symptoms is discussed in further detail below.

Endocrine disorders

While psychotic symptoms secondary to abnormal thyroid functioning are rare, there are case reports of psychosis developing secondary to hyperthyroidism, hypothyroidism, and even rapid alteration of thyroid state. Psychosis has been described as the presenting symptom of thyrotoxicosis in Graves disease 128, thyroid storm 129, toxic nodular goiter 130, subacute thyroiditis 131, and painless thyroiditis 132. Similarly, hypothyroidism has been associated with psychosis, with one study reporting that as many as 5–15% of myxedematous patients have some form of psychotic symptoms 133–135.

Thyroid-associated psychosis does not contain a characteristic psychotic symptom cluster, as patients have presented with auditory and visual hallucinations, delusions, and paranoia. The majority of patients present with affective disturbance 130,136. Treatment of the underlying thyroid abnormality tends to result in resolution of the psychosis, such that antipsychotic medications are necessary only in the acute setting 128,132,136. Interestingly, however, rapid correction of abnormal serum thyroid hormone levels can both induce and exacerbate thyroid-associated psychosis 137,138.

The association of hyperparathyroidism and hypercalcemia with significant psychiatric symptoms is well known, although the specific prevalence of psychosis in patients with hyperparathyroidism remains unclear. Case reports describe patients presenting with auditory and visual hallucinations, persecutory delusions, and disorganized thought processes in the setting of hyperparathyroidism-induced hypercalcemia 139,140. The available evidence indicates that correction of the hypercalcemia, generally via parathyroidectomy, results in resolution of the psychotic symptoms with no subsequent recurrence 139–141.

Although less common and investigated, hypoparathyroidism can also present with psychosis. A review of 268 cases of hypoparathyroidism published in 1962 reported that 11% of the patients had psychotic symptoms, often in the setting of surgically induced hypoparathyroidism 142. Subsequent case reports have supported the occurrence of psychotic symptoms in hypoparathyroidism and emphasized the fact that symptomatic improvement requires normalization of magnesium and calcium levels, in addition to treatment with antipsychotics 141,143.

INVESTIGATING SECONDARY PSYCHOSIS VERSUS SCHIZOPHRENIA

Establishing a cause-effect relationship between substance use/medical illness and psychosis is not easy. Suspecting an underlying medical illness is a logical initial step when encountering psychosis in general medical settings. Comorbid medical illnesses are also quite common in patients presenting with psychotic symptoms, especially among the elderly. Suspecting and identifying an underlying medical illness in younger patients with psychosis in mental health settings is more challenging.

In making the distinction between primary and secondary psychosis, it is important to first establish the presence of the general medical condition. The next step — establishing the cause-effect relationship between the medical condition and psychosis — is often difficult, but can be helped by considering the following three key principles: atypicality, temporality, and explicability.

Is the presentation of the psychosis atypical?

An underlying medical cause for psychosis should be especially suspected if the presentation is atypical. An example is later age of onset: new onset of psychosis in a 70-year-old man should raise suspicion of an underlying medical illness. It is to be kept in mind that no single clinical feature or combination of symptoms reliably distinguishes between primary and secondary psychotic disorders. If, however, a particular feature predominates, that should raise a red flag. Thus, a strong component of severe disorientation and/or confusion must raise suspicion. Catatonic symptoms, altered states of consciousness (i.e., confusional or “dream like” states), and visual hallucinations are more frequent in secondary psychoses. Certain delusions, such as those involving beliefs of mistaken identity of others (i.e., Capgras delusions), are thought to be more common in secondary psychoses than in schizophrenia. The presence of multimodality hallucinations (e.g., visual and tactile) also increases the likelihood of a secondary psychotic disorder.

Accompanying symptoms that are disproportionate to what may be expected from a psychotic disorder should also make one consider a potential underlying medical disease. For example, a large weight loss that may not be easily explained by the mild depression in a psychotic patient may trigger suspicion of a medical illness.

Specific types of psychotic symptoms may also often point to regional alterations in brain function and raise suspicion of neurological disease. Thus, denial of blindness that may appear delusional should trigger suspicion of Anton's syndrome (cortical blindness, due to visual cortex lesions) and denial of paralysis should lead to a consideration of anosognosia (due to lesions in the nondominant parietal cortex). Likewise, both an isolated delusion of misidentification (Capgras syndrome) and olfactory hallucinations are suggestive of temporal lobe disease.

Is the medical condition or substance use temporally related to the psychosis?

Secondary psychosis is likely when the psychosis begins following the onset of the medical condition, varies in severity with the severity of the medical condition, and resolves when the medical condition improves. An example is the appearance of delusions when a patient with hypothyroidism stops taking thyroid treatment and the resolution of the symptoms after resuming medication. This rule, however, has many exceptions; for example, psychosis in temporal lobe epilepsy appears several years after the onset of the seizures. Conversely, a medical illness may simply worsen or trigger a relapse of schizophrenia without being the direct cause of the illness.

Is the psychosis not better explained by a primary psychotic disorder or another mental disorder?

Comorbid medical illness is very common in patients with chronic psychotic disorders such as schizophrenia. In some cases, even if a concomitant medical illness may raise suspicion of a secondary psychosis, the presence of a strong family history of schizophrenia and a premorbid schizoid personality point to a diagnosis of schizophrenia. Similarly, in patients with a known history of an affective illness, the appearance of psychotic symptoms is likely related to the affective illness rather than a medical condition.

Is psychosis a direct physiological consequence of a medical illness or substance use?

The answer to this question depends first on establishing the presence of an underlying medical condition or substance use that might be the etiological agent. This will require a careful history, physical, and neurological examination, along with appropriate laboratory investigations. Second, even if such a causal agent exists, it is often difficult to ascertain whether the symptoms of psychotic illness are a direct physiological consequence of that factor. Sometimes such a direct link may be obvious. For example, the presence of autonomic hyperactivity (e.g., dilated pupils, tachycardia) along with paranoid anxiety might suggest a sympathomimetic agent such as amphetamine, phencyclidine, or an adrenal tumor.

In every patient with a first episode of psychosis, it is critical to obtain a detailed history and complete physical, including neurological examination and laboratory evaluation to rule out common medical disorders (Table 2). Additional investigations such as brain imaging, cerebrospinal fluid and electrophysiologic studies may be needed, especially for those with atypical presentations and those in whom there is reason to rule in a primary disorder.

Table 2.

Approach to investigating patients to rule out secondary psychoses

First line assessments (to be routinely considered in all first psychotic episode patients)

Detailed medical and neurological/psychiatric history

Physical/neurological examination

Neuropsychological tests

Laboratory tests: complete and differential blood count, erythrocyte sedimentation rate, glucose, electrolytes, thyroid function tests, liver function tests, urinary drug screen

Second line assessments (to be considered when above assessments raise specific diagnostic possibilities)

Laboratory tests: rapid plasma reagin to rule out syphilis; HIV testing; serum heavy metals; copper and ceruloplasmin levels; serum calcium levels; autoantibody titres (e.g., antinuclear antibodies for lupus); B12, folate levels; arylsulfatase-A levels; urine: culture and toxicology, drug screen

Neuroimaging: computed tomography, magnetic resonance imaging, positron emission tomography, single proton emission tomography

Electroencephalography, polysomnography, evoked potentials

Cerebrospinal fluid investigations: glucose, protein, cultures, cryptococcal antigen

Karyotyping

The question of whether to conduct a brain scan in a patient with suspected schizophrenia on a routine basis is debatable. When brain scans, commonly structural (such as MRI or CT), have been used clinically as part of the work-up of a psychotic patient, the purpose has been to rule out a space occupying lesion or developmental malformation as potentially causative of the psychosis. Although incidental findings have been reported in MRI studies of patients who present with psychosis 144, and indeed occur even in healthy individuals 145, such findings are rare. Thus, in the absence of quantitative analysis, routine brain imaging cannot aid in the differential diagnosis of psychosis without considering the clinical presentation 146.

CAN SECONDARY PSYCHOSES ILLUMINATE PATHOPHYSIOLOGY OF SCHIZOPHRENIA?

Schizophrenia, a common and highly disabling disease with unclear causation, is a broad heterogeneous entity that may comprise several idiopathic psychotic disorders 147. To unravel its complexity, it is important to identify homogeneous subgroups within this illness to better characterize its pathophysiology. One approach to address this is to examine syndromes of known etiology that present with clinical manifestations similar to schizophrenia, that is, phenocopies of this illness.

To identify suitable phenocopies of schizophrenia to study, one needs to first define what we know about the pathophysiology of schizophrenia. At an anatomical level, schizophrenia is characterized by brain structural abnormalities in frontal, temporal, parietal, basal ganglia, thalamic, and limbic regions 148. Functional imaging studies suggest impaired function of prefrontal regions (hypofrontality) and impaired interhemispheric and intrahemispheric connectivity. At a neurochemical level, increasing evidence points to dopaminergic, glutamatergic, and GABAergic dysfunction in the pathophysiology of psychosis, with the N-methyl D-aspartate (NMDA) hypofunction hypothesis as a leading theory to describe the pathogenesis of schizophrenia. Several lines of evidence also point to alterations in immune and oxidative stress mechanisms in schizophrenia 149. From an etiological point of view, schizophrenia is thought to be highly heritable (heritability > 70%), but several environmental factors including viruses, drugs of abuse, head trauma, and obstetric complications have been implicated 150. It is thus likely that the symptoms of the illness arise through the combination of genetic vulnerability and environmental stressors.

Given the substantive heterogeneity of schizophrenia, it would be difficult to envision any individual secondary psychosis being an adequate model to explain all aspects of that disorder. Not surprisingly, several psychotic disorders stemming from the etiological factors listed in Table 2 demonstrate only partial similarity to what we clinically identify as schizophrenia. Some disorders, however, are more likely to show clinical, pathophysiological, or etiological parallels to schizophrenia; we discuss them briefly as they may provide clues to better understand that disease.

Temporal lobe epilepsy: an anatomical model of schizophrenia

Psychosis appears to occur in 7–11% of patients with epilepsy, a rate much higher than in the general population 151. Psychosis, when it occurs in temporal lobe epilepsy, has been thought to closely resemble schizophrenia, as described by Slater in a classic early paper 125. The emergence of psychoses in temporal lobe epilepsy has been associated with onset of epilepsy under the age of 20 years, a history of epilepsy lasting for more than 10 years, a history of complex partial seizures, and lesions on the left side 152.

The occurrence of psychoses in temporal lobe epilepsy is consistent with the medial temporal structural alterations reported in schizophrenia 148. In particular, positive symptoms such as auditory hallucinations and formal thought disorder have been linked to structural alterations in the auditory association areas in the superior temporal gyri 148. The striking prominence of positive symptoms in temporal lobe epilepsy has led to the question whether there might be a resemblance between temporal lobe epilepsy and the neurochemical models of schizophrenia. Ando et al 153 examined alterations of central dopaminergic systems in the kainate model of temporal lobe epilepsy using methamphetamine-induced locomotor activity as an index of dopaminergic sensitivity in adult rats. They found evidence of dopaminergic hypersensitivity, which can clearly explain the mechanisms underlying epileptic psychosis and can also indicate similar alterations in idiopathic psychoses.

Velocardiofacial syndrome (VCFS): a genetic model of schizophrenia

While the etiology of schizophrenia remains unclear, it is widely agreed that genetic factors have a substantial contribution, with heritability estimates of greater than 70%. Currently, schizophrenia is thought to be polygenic and multifactorial, with a small proportion of cases due to copy number variations, such as microdeletions or microduplications of chromosomal regions. A large portion of the genetic etiology of schizophrenia remains uncharted, necessitating the study of discrete genetic syndromes that present with schizophrenia-like features.

VCFS is characterized by a large deletion on one copy of chromosome 22 (comprising up to 30 genes), which can be detected with genetic testing. It is the most common chromosomal microdeletion in humans, and is characterized by congenital abnormalities of the heart, facial dysmorphism, and cognitive deficits in childhood. A substantial proportion of affected individuals develop major psychiatric illnesses in adolescence or early adulthood, with schizophrenia spectrum disorders occurring in 25–30% of affected individuals 20–23.

VCFS may offer a model of the relationship between genetic liability and risk of symptomatic manifestations of schizophrenia. The cognitive deficits in schizophrenia have been linked to a polymorphism of the gene coding for catechol O-methyltransferase (COMT), an enzyme involved in the degradation of dopamine. Individuals who have the Val/Val polymorphism have reduced prefrontal dopamine because of increased activity of COMT, and as a result, may have impaired cognitive function 154. VCFS is associated with similar cognitive impairments to those seen in schizophrenia, including deficits in executive control, memory, and attention. Individuals with VCFS who have the hemizygous COMT Val158Met genotype have been shown to have improved cognition associated with decreased enzymatic degradation of dopamine 154. Brain morphologic brain abnormalities, including prefrontal and cingulate gray matter loss, similar to those seen in schizophrenia, are also observed in VCFS, and such abnormalities appear to be correlated with cognitive impairments 155. Longitudinal studies suggest that brain structural alterations in VCFS appear to predict the emergence of psychotic symptoms 156, though further research is needed to identify the neurobiological and genotypic signatures of VCFS patients that go on to develop secondary psychoses. Studies, such as these, of the neurobiological basis of psychosis in VCFS will likely elucidate the pathophysiology of at least a subgroup of patients with schizophrenia.

NMDA encephalitis: a pathophysiological model of schizophrenia

While there has been considerable progress in our understanding of the pathophysiological substrate of schizophrenia, the core dysfunctions remain a matter of debate. A leading theory is that glutamatergic and GABAergic dysfunction may underlie the developmental pathophysiology of psychosis, perhaps through glutamatergic NMDA receptor hypofunction 157,158. This model is supported by clinical observations of a psychosis closely resembling schizophrenia that is caused by the NMDA receptor antagonist phencyclidine as well as by neuropathological observations of altered NMDA receptor binding and expression in postmortem brains of patients with schizophrenia 159. The NMDA receptor, therefore, provides a natural biological model for further study and may provide insight into the complex pathophysiological heterogeneity of schizophrenia.

Glutamatergic dysfunction may result from a failure of GABA-mediated regulation of predominantly glutamatergic pyramidal neurons. This theory is supported by observations of reduced GABA synthesis, as reflected by decreased neuronal activity of the 67-kDa isoform of glutamic acid decarboxylase (GAD67), in patients with schizophrenia 160. Several lines of evidence also point to alterations in immune mechanisms in schizophrenia 161. There is a need to draw connections between the NMDA/GABA alteration model and theories of immune system modulation in schizophrenia; studies of the appropriate forms of secondary psychoses might be one way to further elucidate this model.

An intriguing form of secondary psychosis that leads to a hypofunctioning NMDA receptor state may provide insight into the pathophysiology of schizophrenia. Since 2007, many case reports described a form of encephalitis in which patients presenting with neurological and psychiatric symptoms were found to have positive autoantibodies to the NR1/NR2 heteromers of the NMDA receptor. Dalmau and colleagues 162,163 examined 100 patients presenting with NMDA receptor encephalitis and showed that 77% presented with a variety of psychiatric symptoms that included anxiety, insomnia, fear, grandiosity, delusions, hyper-religiosity, mania, and paranoia.

A model that integrates NMDA receptor encephalitis with GABA/glutamate dysfunction includes the idea that an antibody-mediated decrease in NMDA receptors may inactivate GABAergic neurons, which normally serve to inhibit extracellular glutamate. Without the regulating presence of GABA, the resultant excessive glutamate may induce and exacerbate psychosis. A similar mechanism may be present in schizophrenia, though definitive evidence of alterations of brain glutamatergic neurotransmission in this disease has yet to be produced. The NMDA receptor provides a biological mechanism for the development of psychosis that warrants further study, with the further possibility that variations in the nature of NMDA receptor dysfunction could inform the complex pathophysiological heterogeneity of schizophrenia.

Conclusions

The historical distinction between organic (or structural) and functional (purely psychological) disease has not served the field and stunted systematic investigation in the disorders that were earlier deemed purely psychological. Classification of psychiatric disorders into those with or without identifiable etiology has been more clinically meaningful.

Identification of an underlying medical, toxic, or iatrogenic cause in a patient presenting with psychosis can be diagnostically challenging. Careful history taking, physical examination, and judicious use of modern medical testing, combined with an informed mind, can help the clinician to arrive at a timely diagnosis and optimum intervention, which can be quite gratifying.

In appreciating the causes of secondary psychoses, the clinician-scientist also gains potential insight into the puzzling pathophysiology and etiology of primary psychotic disorders such as schizophrenia.