Limbic Encephalitis and Late-Onset Psychosis

The patient with late-onset psychosis described in a recent Clinical Case Conference (1) had a constellation of signs and symptoms suggestive of the syndrome of paraneoplastic encephalomyelitis and sensory neuropathy. One of its signs is limbic encephalitis. This may manifest in hallucinations, paranoia, or mood disorders with or without memory loss and complex partial seizures. Paraneoplastic encephalomyelitis is often associated with sensory neuropathy with distal symmetric sensory loss, as suggested by the results of this patient’s neurological examination. Paraneoplastic encephalomyelitis/sensory neuropathy is often associated with the presence of the anti-Hu antibody in serum and CSF. The syndrome may be progressive, or it may stabilize. It has in some cases been reported to remit with the treatment of an underlying malignancy (2).

Of 71 patients with paraneoplastic encephalomyelitis/sensory neuropathy and anti-Hu antibodies, 77.5% (N=55) were eventually diagnosed with small-cell lung cancer. In 12.7% (N=9), no tumor was found. The remainder of the patients (N=7) had a variety of other tumors, including prostate cancer, one case of which was discovered only at autopsy (3).

I would not agree that the neurologic findings as described were generally consistent with age. Tremor, rigidity, and postural instability are likely attributable to examination after treatment with neuroleptics (although symptoms of parkinsonism have also been reported in paraneoplastic encephalomyelitis/sensory neuropathy) (3). However, decreased distal sensation and lower extremity reflexes in the setting of “sensory ataxia” and a “slapping gait” are highly suggestive of peripheral neuropathy and are not attributable to neuroleptics or to normal aging. Also, the snout reflex is not a normal reflex in an adult. The temporal characteristics of symptom onset are not explicit in the report, but a clear subacute onset would favor the diagnosis of paraneoplastic encephalomyelitis/sensory neuropathy. The duration of symptoms is neither supportive nor incompatible with such a diagnosis.

Abnormalities may be found from magnetic resonance imaging (in a minority of cases), EEG, electromyogram, and CSF studies and serum tests for the anti-Hu antibody (3). (Since the submission of my original letter, several other antineuronal antibodies have been identified, and they are described in the second 2000 edition of Neurobase.) Syndromic presentation should prompt a workup for malignancy. Treatment should be directed at any underlying malignancy, seizures, or psychiatric symptoms. Also, some patients, particularly those with limbic encephalitis, may benefit from steroids, plasmapheresis, or intravenous immunoglobulin.

In this patient’s case, it is likely that he would decline invasive diagnostic evaluation. The capacity to decline evaluation and treatment requires patient insight into his or her condition and an understanding of the consequences of refusing treatment. The patient is described as understanding the risks and benefits of declining evaluation of his prostate nodule. However, the delusional nature of his paranoia might be considered to preclude capacity for evaluation of his abnormal mental status. Obviously, these ethical issues are complex.