div class=”ChapterContextInformation”>
5. Secondary Schizophrenia
Keywords
Differential diagnosisPrimary psychosisSecondary psychosisClinical presentationDiagnosisMedical work-upSecondary schizophrenia(s)TreatmentEssential Concepts
Before diagnosing (primary) schizophrenia, you must rule out one of the secondary schizophrenias (i.e., schizophrenic symptoms are secondary to a nonpsychiatric medical disorder, either from a systemic disorder that affects the brain or from demonstrable neuropathology in the brain).
Only a very small number of patients with first-episode schizophrenia (diagnosed with modern criteria), around 5%, will have an identifiable, medical etiology of their psychosis.
Psychopathology is of no help to differentiate between primary and secondary causes of schizophrenia. The cross-sectional psychopathology is non-specific with regard to the etiology of psychosis. Do not give undue weight to one particular type of hallucination alone (e.g., visual hallucinations or Schneiderian voices) when making your diagnosis.
Merely detecting a medical condition does not establish causality. Establishing causality between an identified medical condition and psychosis relies on atypical features for schizophrenia (age of onset, symptoms in aggregate, treatment response, course), temporality, and biological plausibility.
Apart from some basic screening tests to exclude obvious medical organ disease and specific, highly treatable disorders, the further extent of the work-up is determined by your clinical suspicion for an underlying medical condition. Indiscriminate screening in psychotic but otherwise unremarkable patients is ill-advised as any positive result is likely false-positive.
The value of neuroimaging in first-episode patients is disputed as the yield is low. Expect non-specific (and in almost all cases clinically irrelevant) MRI findings in about 1 in 5 patients.
A wide variety of medical/neurologic disorders and some toxins are associated with psychosis, mimicking schizophrenia. Their diagnoses if there are no ancillary signs and symptoms require a combination of screening, a high index of suspicion, and clinical follow-up.
Even if psychosis is the result of identifiable pathology, symptomatic treatment with an antipsychotic in addition to medically treating the underlying disease is often necessary. Antipsychotics may be less effective than in schizophrenia and poorly tolerated, particularly in neurological conditions.
“Crude exogenous organic damage of the most varying kind can produce acute psychotic clinical pictures of a basically uniform kind.” (cited in [1])
–Karl Bonhoeffer, 1868–1948
(The father of “organic” psychiatry [2], also the father of the theologian Dietrich Bonhoeffer [3])
Many medical disorders can potentially mimic schizophrenia. In this chapter, I outline a clinical approach for a patient with psychosis of unknown etiology in order to not miss a medical disease that could account for the psychosis. For those schizophrenia-like psychoses that are the result of medical illness, I follow the suggestion of Spitzer and colleagues [4] to abandon the old term, “organic mental disorder,” and use “secondary schizophrenia” or “secondary psychosis” instead, terminology consistent with ICD-11 that uses “secondary psychotic syndrome” [5]. For brevities sake, I still occasionally use the adjective “organic” to indicate nonpsychiatric causes. In the older literature, you may encounter “idiopathic” schizophrenia to indicate primary schizophrenia. The distinction between primary and secondary disorders is familiar to physicians; it helpfully does not imply that schizophrenia is not brain-based (a wrong conclusions fostered by calling it a “functional” disorder). In this article, I focus on secondary schizophrenia due to a medical illness. Psychosis associated with substance use and psychosis in the context of a delirium are covered in the preceding Chaps. 3 and 4, respectively. It is important, however, to not forget that substance use disorders and delirium are both commonly associated with psychosis.
Differential Diagnosis
Differential diagnosis of new-onset psychosis
Psychosis could be attributable to a combination of factors (like a delirium which is often multifactorial), and all of them must be systematically examined. Determining causation is not always easy once you detect a medical condition (see the case in the Additional Resources). After its discovery, you need to determine if the condition is an incidental finding (fortuitous), relevant at some etiological level (“triggering” schizophrenia), or solely responsible for the psychopathology (if you remove it, the psychosis resolves). Criteria to use are atypicality (i.e., the presentation does not quite look like schizophrenia vis-à-vis age of onset; there is no prodrome; there are symptoms including other physical findings that are unusual; the symptoms in aggregate point toward organicity (visual hallucinations, olfactory hallucinations, lack of Schneiderian first-rank symptoms); the treatment response to standard treatment is poor), temporality (the time course of psychosis parallels time course of the medical condition), and biological plausibility (the medical condition is known to cause psychosis) [7].
Clinical Vignette
A young man developed a textbook case of paranoid psychosis. During the routine work-up for first-episode psychosis, a pituitary tumor was detected by magnetic resonance imaging (MRI) and partially resected. The patient now receives maintenance treatment for schizophrenia, as well as a dopamine agonist to reduce prolactin levels.
This case exemplifies the occasional scenario of an incidental discovery of a medical condition during the work-up for first-episode psychosis. You will need to judge if the discovered medical condition is etiologically related to the psychosis, etiologically unrelated yet important for management, or etiologically unrelated and not important for management. In this case, the finding is etiologically unrelated to schizophrenia but complicates the management of schizophrenia (treatment with a dopamine agonist).
Psychosis in otherwise healthy appearing first-episode patients diagnosed according to modern criteria is rarely due to an unrecognized medical condition. Johnstone and colleagues [8] found organic disease (judged to be relevant for the psychiatric presentation) in 15 out of 258 (less than 6%) patients with first-episode schizophrenia. Specific conditions identified were syphilis, lung cancer, autoimmune multisystem disease, cerebral cysticercosis, thyroid disease, and previous head injury. In a different sample, 22% of patients with first-episode psychosis who underwent clinical MRI scanning had an unsuspected finding [9]. Notably, first-episode patients and scans from healthy controls had similar rates of non-specific abnormal findings. However, only three patients (2%) of the first-episode group required an urgent referral (vascular lesion, arachnoid cyst, possible Huntington’s disease). Given the sensitivity of MRI imaging, non-specific findings are going to be detected in a significant minority of patients (about 1 in 5 patients).
Tip
A nonhierarchical approach (i.e., simply listing schizophrenia and medical conditions (including drugs and toxins) without making causal assumptions about their relationship) is often most appropriate unless the psychosis is clearly the result of the medical disease. How do you make this judgment? If you think that controlling the medical disease will eventually resolve the psychosis and not require long-term treatment with an antipsychotic, you are probably dealing with a secondary psychosis.
Clinical Presentation
Unfortunately, there is no easy way to differentiate primary from secondary psychoses phenomenologically. As noted in the epigraph, the father of organic psychiatry, Karl Bonhoeffer recognized 100 years ago that the psychiatric clinical picture produced by a medical condition was rather uniform and unspecific, regardless of etiology [10]. Later authors have similarly failed to differentiate primary from secondary psychosis based on psychopathology alone [11, 12]. While certain symptoms suggest a medical or toxic etiology (e.g., visual hallucinations, olfactory hallucinations, dream-like quality of delusions (patient as observer), lack of Schneiderian first-rank symptoms) [1], there are no pathognomonic signs or symptoms that unequivocally point clinicians either way. Schneiderian first-rank symptoms are common in schizophrenia, quasi per definition (50%), and less common but not unusual in secondary psychoses, particularly if there is a clear sensorium (20%) [13]. We often attribute visual hallucinations to drug use and olfactory hallucinations, but we need to be cautious since visual hallucinations also occur in 25% of schizophrenia patients [14]. Olfactory hallucinations, while probably relatively uncommon in schizophrenia, are less carefully assessed; they are often unpleasant smells or stenches (feces) and co-occur with tactile experiences [15]. To complicate matters, some psychiatric presentations can include acute confusion and perplexity which usually indicates a more medical etiology [16].
Key Point
You cannot determine from cross-sectional psychopathology if psychosis is due to schizophrenia or due to a medical condition. There are no pathognomonic symptoms that would steer you either way. Do not give undue weight to one particular type of hallucination alone (e.g., visual hallucinations or Schneiderian voices) when making your diagnosis [17].
Diagnosis
Diagnostic dilemma of psychosis in rare or atypical diseases
All patients with new-onset psychosis (and all chronic patients with a symptom exacerbation for that matter) therefore need a medical work-up. There is no generally agreed-upon work-up that every patient with psychosis must have. Some clinicians will be rather skeptical about expansive work-ups, others more enthusiastic. The neurologist Vladimir Hachinski noted that “without therapeutic enthusiasm there would be no innovation, and without skepticism there would be no proof” which captures the tension between those two poles nicely [19].
Initial work-up for first-episode schizophreniaa
Laboratory tests |
Screen broadly |
Complete blood count Electrolytes including calcium Renal function tests (BUN/creatinine) Liver function tests Erythrocyte sedimentation rate (ESR) Antinuclear antibodies (ANA) Glucose Urinalysis Urine drug screen |
Exclude specifically |
TSH Vitamin B12 and folate HIV screeningb FTA-Abs for syphilis (RPR not sufficient) Ceruloplasminc |
Neuroimaging |
MRI to rule out demyelinating disease, brain tumor, or stroked |
Ancillary tests, as clinically indicated |
EEG CXR, lumbar puncture, blood cultures, arterial blood gases (in infections) Autoantibodies (CSF!) Karyotype (early onset schizophrenia) Serum cortisol Medication drug levels Toxin search |
Put together a screening test battery to exclude common and a few selected yet very treatable diseases.
Epidemiology counts, both literally and figuratively: The extent of your work-up is determined by an emphasis on treatable condition from your neighborhood or brought back to your neighborhood (i.e., travel).
More tests are not necessarily better: Indiscriminate screening for rare diseases without clinical suspicion for the disease is inadvisable because of false-positive (and false-negative) test results [25]. Order specific tests to rule in or rule out a disease you suspect clinically.
An MRI will provide reassurance that a silent brain tumor (e.g., frontal lobe meningeoma) is not missed, although the clinical yield of ordering an MRI in this context will be low [9]. Expect to detect mostly innocuous, incidental MRI abnormalities seen in 20% of the normal population [26].
Electroencephalograms (EEGs) can be difficult to interpret since almost half of patients with first-episode schizophrenia will have EEG abnormalities of unclear significance [27]. Moreover, a normal (surface) EEG does not exclude epilepsy. You need to pursue a diagnosis by repeat (serial) testing under optimal conditions (special lead placement, sleep-deprived) [28]. As my late mentor, George Murray would point out, “If you don’t catch a fish in the ocean, that does not mean there are no fish.”
Order the correct test: For example, to exclude neurosyphilis, your patient needs to have a highly sensitive and specific serum treponemal-specific test (fluorescent treponemal antibody absorption test (FTA-Abs)) and, if positive, a lumbar puncture, not the commonly (and incorrectly) ordered rapid plasma reagin (RPR) blood test [29].
In poorly responsive psychosis, expand your search to exclude paraneoplastic syndromes or similar autoimmune inflammatory brain diseases, epilepsy, and sarcoidosis.
Tip
Longitudinal follow-up by the same person is probably the best safeguard against missing secondary schizophrenia, assuming that the medical disease “declares itself” with new and nonpsychiatric findings. Thus, any change in symptoms or new symptoms should lead you to revisit your initial impression.
Secondary Schizophrenias
Genetic Disorders
Several genetic syndromes have an increased risk for schizophrenia, particularly Klinefelter’s syndrome, fragile X syndrome, and velo-cardio-facial syndrome (VCFS). VCFS, which stems from a deletion on the long arm of chromosome 22 (22q11), is one of the strongest genetic risk factors for schizophrenia-like presentations, possibly in up to 25% of VCFS cases [30]. Consider VCFS testing in children with mild cognitive impairments [31].
Tip
Currently, genetic testing is not routinely recommended for patients with psychosis unless there is the clinical suspicion that a genetic syndrome is present (e.g., family history of Huntington’s disease). Even if there is no treatment, making a syndromal diagnosis of a genetic disease is important for counseling and to look for other syndromal features that might be treatable (e.g., cardiac problems that are part of a syndrome). I suspect that psychiatrists will need to keep up with genetics when whole exome and whole genome sequencing becomes part of the work-up for neuropsychiatric disorders [32].
Endocrine Diseases
One of the easiest-to-correct endocrine conditions associated with psychosis is hypoglycemia. Screen for thyroid disease with a thyroid-stimulating hormone test (TSH), to exclude both hyperthyroidism [33] and hypothyroidism (myxedema madness) [34]. In addition, consider Addison’s disease, Cushing’s disease [35], and hyperparathyroidism, as well as hypoparathyroidism. A pheochromocytoma is a very rare cause of psychosis [36].
Metabolic Diseases
Many inborn errors of metabolism include psychosis in the list of possible symptoms. Almost all are diagnosed during childhood, but atypical, adult onset is possible (e.g., Niemann-Pick type C [37]). Only acute intermittent porphyria is sufficiently common that you should suspect it if abdominal pain and peripheral motor neuropathy are present in a patient with psychosis [38].
Tip
It is probably almost impossible to diagnosis a rare presentation (psychosis alone) of a rare disorder (metabolic diseases) that presents at an atypical age (adulthood). You need other clues and a high index of suspicion that would suggest a metabolic disease.
Autoimmune Diseases
The most important disorder to consider is systemic lupus erythematosus (SLE) [39]. In the case of SLE, treatment with steroids greatly complicates diagnostic issues [40]. A highly steroid-sensitive form of autoimmune thyroiditis is Hashimoto’s encephalopathy [41]. Patients with schizophrenia have higher than expected titers of antibodies related to celiac disease [42]; however, their pathogenic role remains to be better understood before routine screening for celiac disease can be recommended. Rarely, classic paraneoplastic limbic syndromes can be a cause of psychosis [43]. The more recently discovered NMDA receptor encephalitis is described in more detail under Neurologic Conditions.
Narcolepsy
The hypnagogic hallucinations that are part of the narcolepsy tetrad (in addition to cataplexy, sleep paralysis, and excessive daytime sleepiness) can lead to a mistaken diagnosis of schizophrenia [44]. In one series of state hospital patients diagnosed with schizophrenia, 7% were found to suffer from narcolepsy [45]. In some, psychosis is probably related to the treatment of narcolepsy with stimulants.
Tip
Screen all patients with psychosis for the narcolepsy tetrad; however, only 10% of patients have the full tetrad. Consider a nocturnal sleep study, followed by a multiple sleep latency test (MSLT) to identify reduced daytime sleep latency and sleep onset rapid eye movement (SOREM) periods. Human leukocyte antigen (HLA)-typing and cerebrospinal fluid levels of hypocretin-1 can further assist in making the correct diagnosis if psychosis appears to be limited to the sleep-wake transitions, even in the absence of other symptoms of narcolepsy [46].
Neurologic Conditions
Stroke
Poststroke psychosis is a rare but possible complication of stroke [47]. To complicate matters, in many cases, seizures obfuscate the picture and might be related to the emergence of psychosis. Peduncular hallucinosis is a syndrome of vivid, colorful formed visual hallucinations, often of animals (e.g., of a parrot in full plumage), with localizing value as a focal lesion of the midbrain cause them, hence “peduncular.”
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
