Jan 23, 2009
Lyme Disease: A Neuropsychiatric Illness
By Brian A. Fallon, M.D., M.P.H., and Jenifer A. Nields, M.D.
Am J Psychiatry 151:11, November 1994 pp.1571-1580
http://www.angelfire.com/biz/romarkaraoke/lymeart.html Go to this website to read entirely.
Objective : Lyme disease is a multisystemic illness that can affect the central nervous system (CNS), causing neurologic and psychiatric symptoms. The goal of this article is to familiarize psychiatrists with this spirochetal illness.
Method : Relevant books, articles, and abstracts from academic conferences were perused, and additional articles were located through computerized searches and reference sections from published articles.
Results : Up to 40% of patients with Lyme disease develop neurologic involvement of either the peripheral or central nervous system. Dissemination to the CNS can occur within the first few weeks after skin infection. Like syphilis, Lyme disease may have a latency period of months to years before symptoms of late infection emerge. Early signs include meningitis, encephalitis, cranial neuritis, and radiculoneuropathies. Later, encephalomyelitis and encephalopathy may occur. A broad range of psychiatric reactions have been associated with Lyme disease, including paranoia, dementia, schizophrenia, bipolar disorder, panic attacks, major depression, anorexia nervosa, and obsessive-compulsive disorder. Depressive states among patients with late Lyme disease are fairly common, ranging across studies from 26% to 66%. The microbiology of Borrelia burgdorferi sheds light on why Lyme disease can be relapsing and remitting and why it can be refractory to normal immune surveillance and standard antibiotic regimens.
Conclusions: Psychiatrists who work in endemic areas need to include Lyme disease in the differential diagnosis of any atypical psychiatric disorder. Further research is needed to identify better laboratory tests and to determine the appropriate manner (intravenous or oral) and length (weeks or months) of treatment among patients with neuropsychiatric involvement. (Am J Psychiatry 1994; 151:1571-1583)
Lyme Disease (Lyme borreliosis), caused by the tick-borne spirochete Borrelia burgdorferi, may progress from an initial skin infection to a disabling multisystemic illness. Now the most common vector-borne infection in the United States, Lyme disease is increasing in incidence and geographic spread (1). The disease has dermatologic, arthritic, ophthalmologic, cardiac, neurologic, and psychiatric manifestations (2). In its protean manifestations, in its spirochetal etiology, and in its course (early skin localization and rapid invasion of the central nervous system [CNS]), Lyme disease is similar to syphilis (3). Like syphilis, early recognition is important to prevent an acute, treatable illness from becoming a chronic or relapsing one. Because current diagnostic tests are not always reliable, physicians must rely on clinical presentation as the basis for diagnosis. Because many of the symptoms of Lyme disease involve the CNS, patients with Lyme disease may be referred to psychiatrists both before and after diagnosis.
In this article, we present an overview of Lyme disease with a particular emphasis on its neuropsychiatric features.
Lyme disease is transmitted by an infected nymphal or adult female Ixodes tick. Smaller than the dog tick, the Ixodes tick may easily be missed on casual inspection. The bite is usually not painful. Transmission of the spirochete appears to require the tick to feed at least 12-24 hours (4). The ticks are most commonly carried by deer and by the white-footed mouse, but other carriers have been described as well.
Lyme disease has been reported throughout the United States and in numerous countries around the world. The geographic spread and the incidence in the United States have been rapidly increasing. For example, during 1992, 45 states reported 9,677 cases, representing a 19-fold increase over the 492 cases reported by 11 states in 1982 (5). The State of Connecticut, which in 1992 had the highest rate of Lyme disease in the country, reported between 1991 and 1992 a threefold increase in the proportion of infected ticks in four communities and a one and a half-fold increase in reported cases throughout the state (6). The most heavily affected areas include the Northeast (New York, New Jersey, Connecticut, Massachusetts, Rhode Island, Pennsylvania), the upper Midwest (Minnesota, Wisconsin), and the Pacific coastal region (California, Oregon).
Although Lyme disease was first described in the United States as an arthritic illness preceded by a rash (7), early reports in Europe described a primarily neurologic illness without any arthritis (8,9). Psychiatric symptoms were described in some of these early reports.
In 1909, a Swedish physician described the classic Lyme rash, known as erythema chronicum migrans, noting that it developed at the site of an Ixodes tick bite (10). In 1922 two French doctors, Garin and Bujadoux, wrote a case entitled "Paralysis by Ticks," now thought to be the first report of Lyme meningoradiculitis (8). The patient developed erythema migrans after a tick bite, followed by radiculopathy, paralysis of a portion of one arm, anxiety, and meningitis. In 1930 Hellerstrom, of the Karolinska Institute, described a man who, 3 months after an erythema migrans rash, developed an encephalitis with psychotic symptoms, disorientation, and marked CSF abnormalities (11). In 1941 Bannwarth, a German neurologist, described the syndrome of chronic lymphocytic meningitis, which was characterized by radicular pains, lymphocytic meningitis, and peripheral nervous system involvement, especially facial palsy (9). The cases in all of these early reports, previously described as Garin-Bujadoux syndrome, Bannwarth's syndrome, and neuroborreliosis, are now considered to have been cases of Lyme disease.
In the United States, the first report of a tick-induced erythema migrans rash was in 1970 (12). In 1977 "Lyme arthritis" was described by Steere et al. (7); the article was based on an epidemiological investigation of an outbreak of presumed juvenile rheumatoid arthritis in Connecticut. In 1978 the link between Lyme arthritis and the bite of an Ixodes tick (13) was recognized. In 1982 Burgdorfer et al. isolated the etiologic agent of Lyme disease from an Ixodes tick--a spirochete now known as B. burgdorferi (14). Early in the history of Lyme disease, aspirin and nonsteroidal anti-inflammatory agents were used for symptoms that emerged after the erythema migrans rash (15). Subsequently, penicillin was shown to shorten the duration of illness, thus supporting an infectious etiology. While short courses (10 days) of oral or intravenous antibiotics were recommended at first, it is now recognized that some patients benefit from longer courses (6 weeks or longer) or repeated treatments (16-18).
Typical Clinical Manifestations of Lyme Disease
Within days or weeks after the bite of a tick infected with B. burgdorferi , a localized skin reaction may occur, consisting most typically of an erythematous annular rash (erythema migrans), which may enlarge to a size of 5 cm or greater. This early localized sign of infection may soon be followed by mild to severe flu-like symptoms.
Hematogenous dissemination may lead to early (weeks to months) heart, ophthalmologic, or nervous system involvement. Although second- or third-degree atrioventricular block is most common, rare reports of myopericarditis, left ventricle dysfunction, and cardiomegaly exist (19). Conjunctivitis can be an early manifestation of ocular involvement.
Within the first few weeks after skin infection, B. burgdorferi may disseminate to the CNS (20-22) where it may remain quiescent for months to years before producing symptoms (23). Because approximately one-third of infected patients do not recall the tick bite or rash and because the flu-like symptoms are nonspecific and may be mild (24), patients may not realize that they are infected until long after the initial bite. Neurologic problems, which occur in 15%-40% of patients (25), may be the presenting symptom. Early on, patients may experience headaches without any signs of inflammation in the CSF (18). Shortly thereafter, patients may develop meningitis, cranial neuritis, and motor or sensory radiculitis (26). With meningitis, symptoms may include recurrent severe headaches, stiff neck, photophobia, and, less commonly, nausea and vomiting. At this stage, objective signs are commonly present in the CSF (see the section on CNS laboratory tests). In less than half of the patients with meningitis, a mild encephalitis develops that is characterized by fluctuating disturbances of mood, concentration, memory, and sleep. Cranial neuritis, such as Bell's palsy, occurs in 5%-10% of patients with neurologic Lyme disease (27). Other signs of peripheral nerve involvement include sensory or motor radiculoneuropathies; objective abnormalities may be evident on nerve conduction studies. Symptoms of peripheral neuropathy typically include sharp shooting pains, areas of numbness, paresthesias, weakness, or fasciculations.
Later-stage illness (months to years after infection) generally affects the joints, eyes, skin, or CNS. Arthritic involvement begins with migratory arthralgias and, in 60% of untreated patients, develops into an inflammatory arthritis, typically affecting the large inflammatory arthritis, typically affecting the large joints, such as the knee (28). Ophthalmologic involvement may consist of localized inflammation such as uveitis, iritis, or optic neuritis (29, 30). A late dermatologic manifestation of Lyme disease, acrodermatitis chronica atrophicans, is seen almost exclusively in Europe (31).
Late neurologic involvement may be manifested by encephalomyelitis or encephalopathy (18, 23, 32). Encephalomyelitis, an uncommon late manifestation of Lyme disease, may have quite severe and diverse presentations, including spastic paraparesis, transverse myelitis, cerebellar syndromes, hemiparesis, and movement disorders (18, 32). More common in late Lyme disease is an encephalopathy characterized by subtle to severe cognitive changes and a polyradiculoneuropathy (23). In this stage of illness, the CSF may appear normal (18, 21) (see section on CNS laboratory tests). Other accompanying symptoms of later infection include profound fatigue, sleep disturbance, photophobia, auditory hyperacusis, extreme irritability or emotional lability, word-finding problems, dyslexic-like errors when speaking or writing, and spatial disorientation (23, 33). Disturbances in other sensory modalities, such as taste and smell, have been reported (33, 34). These symptoms may fluctuate in intensity so that symptoms are present on some days but not others. The profile of persistent, marked fatigue and cognitive deficits associated with late-stage Lyme disease is similar to the symptom profile of the chronic fatigue syndrome (35). Whether the late-stage symptoms of Lyme disease are due to persistent infection or to a postinfectious immune activation is an important question that requires further elucidation.
Because patients with encephalopathy or encephalitis may experience marked mood lability, irritability, and sleep disturbance, the distinction between an organic mood disorder and a concomitant primary major depression may be quite difficult to make.
A typical Neurologic Manifestations of Lyme Disease
Because the clinical spectrum of Lyme disease continues to expand, physicians who work in endemic areas should keep Lyme disease in the differential diagnosis of any atypical neurologic illness with multisystemic features. Case reports, for example, have linked a variety of neurologic syndromes to late Lyme disease; these include blindness (30), progressive dementias (32, 36, 37), seizure disorders (34, 38, 39), the Tullio phenomenon (40), strokes (41), extrapyramidal disorders (42), amyotrophic lateral sclerosis (43), Guillain-Barre syndrome (44), and progressive demyelinating-like syndromes mimicking multiple sclerosis (23).
Centers for Disease Control (CDC) Criteria for Diagnosis
Lyme disease surveillance by the CDC began in 1982, and in 1991 Lyme disease became nationally reportable. For epidemiologic surveillance studies (1), the CDC requires history of exposure in an endemic area and either 1) a physician-diagnosed erythema migrans rash of at least 5 cm in diameter or 2) laboratory confirmation of exposure to B. burgdorferi and at least one systemic manifestation. Systemic manifestations must be either musculoskeletal (arthritis), neurologic (lymphocytic meningitis, cranial neuritis, radiculopathy, encephalomyelitis with intrathecal antibody production), or cardiac (second- or third-degree atrioventricular conduction delays). Laboratory confirmation requires the isolation of B. burgdorferi , the demonstration of diagnostic levels of B. burgdorferi immunoglobulin (1g) M or IgG antibodies in serum or CSF, or a rising specific antibody titer on serum samples taken from acutely ill and convalescent patients.
These criteria have been useful for epidemiologic studies, but not all patients with Lyme disease will meet this case definition. About one-third of patients do not recall the erythema migrans rash; serologic testing may be unreliable (45); and the clinical spectrum of Lyme disease continues to expand beyond the manifestations currently included in the CDC case definition.
Because B. burgdorferi is difficult to culture, indirect methods are used to detect the presence of the spirochete. Currently available serologic tests, such as the enzyme-linked immunosorbent assay (ELISA) and the indirect immunofluorescence assay, rely on the immune response following exposure to B. burgdorferi , but they can be unreliable, with both false positive and false negative results (45). In a recent study, over half of the 45 laboratories studied reported falsely negative values in a known positive serum sample from a patient with Lyme disease (46). The Western blot is also often used to examine the serum for antibodies against epitopes that are specific for B. burgdorferi (e.g., 31 kD, 34 kD, or 39 kD bands). Other laboratory tools are emerging--such as urine antigen tests (47), cell-mediated immunoassay (48), immune complex assays (49), polymerase chain reaction assays (50, 51), and borreliacidal antibody tests (52)--but these are not yet well standardized across laboratories.
Several B. burgdorferi antigens are shared by other spirochetes. For example, both B. burgdorferi and the etiologic agent of syphilis, Treponema pallidum , may cause a positive finding on the fluorescent treponemal antibody absorption tests; results of nontreponemal tests, such as the rapid plasma reagin and Venereal Disease Research Laboratory tests, are usually negative in Lyme disease (18). Patients with syphilis or periodontal disease (oral spirochetes) may have falsely positive Lyme ELISA serologies and a common 41 kD antibody to flagellar antigen evident on Western blot.
Falsely negative test results may occur for a variety of reasons. If tested too soon after initial infection, the patient may not yet have mounted an antibody response (53). In addition, antibiotic treatment early in the infection may abrogate the humoral immune response (54). In some cases, free antibodies may not be detected because the borrelia antibodies are bound within circulating immune complexes (55). Finally, interlaboratory variability in antigenic standardization of Lyme assays may result in false negative as well as false positive results (46).
CNS Laboratory Tests for Lyme Disease
The results of laboratory testing among patients with neurologic Lyme disease vary depending on the stage of the illness. In very early CNS involvement (meningismus) or late-stage infection (encephalopathy), the CSF may appear normal (18). When clinical signs of meningitis or encephalitis are present, a spinal tap may reveal a mononuclear pleocytosis, mildly increased protein, and, in some cases, an elevated IgG index or oligoclonal immunoglobulins. Intrathecal anti- B. burgdorferi antibody production is present in 70%-90% of patients with Lyme meningitis (18). Magnetic resonance imaging (MRI) studies may demonstrate punctate white matter lesions of T 2 weighted images, similar to those seen in demyelinating disorders, such as multiple sclerosis. EEG studies may show diffuse slowing or epileptic discharges, but this is uncommon.
In patients with late Lyme encephalopathy, results of brain MRI and EEG studies are generally normal. Functional brain imaging using quantitative brain perfusion single photon emission computed tomography (SPECT), however, may reveal hypoperfusion, particularly in the cerebral white matter, even in patients with no CSF or MRI abnormalities (E.L. Logigian et al., unpublished data, 1994). Objective deficits may be seen on neuropsychological testing (23, 27, 56, 57) (table 1). In about half of these patients, typical markers of CSF infection (pleocytosis, elevated protein, intrathecal antibody production) cannot be found (18). Current experimental research using sensitive ELISA and Western blot techniques has demonstrated the continued presence of spirochetal antigens among many patients with encephalopathy whose CSF otherwise tests normal (21). In some of these patients, results of standard antibody testing of both the serum and the CSF have been negative, but the immune complex dissociation assay revealed bound B. burgdorferi -specific antibody (58).
Given the limitations of diagnostic tests, clinicians need to consider clinical factors that would aid in the diagnosis of Lyme disease. These include a history of an erthema migrans rash or Ixodes tick bite, exposure to a Lyme endemic area, and the combination of neuropsychiatric and extraneural symptoms. Because Lyme disease is a multisystemic illness, patients whose neuropsychiatric symptoms start after a flu-like illness should be asked about a history of other symptoms of Lyme disease, including rashes, joint pains, arthritis, cardiac problems, changes in vision, and radicular pains or cranial nerve palsies.
Most studies have found that patients with Lyme encephalopathy have subtle impairments in memory, concentration, learning, and conceptual ability. Typically, the deficits suggest frontal lobe involvement, affecting short-term memory, verbal fluency, or executive cognitive functions (table 1). Logigian et al. (23), in a study of 27 patients with chronic neurologic Lyme disease, found that 15 of the 27 had quantifiable memory deficits. In order to determine whether psychological factors might account for the memory impairment among Lyme disease patients, Kaplan et al. (56) compared 20 patients with Lyme encephalopathy with 11 fibromyalgia patients and 11 nonpsychotic depressed patients on a neuropsychological test battery. They found that the Lyme disease patients showed greater impairment on standardized memory tests than either of the comparison groups and that the impairment was independent of the number of somatic complaints and the presence of depression.
Cognitive impairments among patients with late Lyme encephalopathy often improve with antibiotic treatment (23, 27), suggesting that active spirochetal infection causes the encephalopathy. In Halperin et al.'s study (27) of patients with late Lyme borreliosis, serial neuropsychological testing before and after a course of intravenous antibiotics revealed marked improvement on tests of memory, attention and concentration, conceptual ability, and psychomotor and perceptual motor function. Noteworthy is that many patients with cognitive deficits did not have clinical evidence of focal CNS disease. Results of EEGs, CSF studies, and other laboratory investigations were often normal. MRI scans were abnormal in some of the patients with severe memory impairment, revealing hyperintense T 2 white matter lesions suggestive of edema or inflammation. Some patients with late Lyme encephalopathy continue to have residual neuropsychological deficits after antibiotic treatment.
Krupp et al. (57) compared 15 patients with Lyme disease who had complaints of persistent cognitive difficulty 6 months after antibiotic treatment to 10 healthy comparison subjects matched in age and education. Compared to the healthy subjects, the Lyme disease patients exhibited marked impairment on memory tests. In that study, the memory impairment was not correlated with serum of CSF anti- B. burgdorferi antibody titers and was not explained by MRI findings or depression. Fatigue, however, a nonspecific marker of chronic Lyme disease, was correlated with memory impairment; this suggested to the authors that the persistent encephalopathy could be an indirect effect of systemic infection elsewhere in the body. The authors noted that persistent neuropsychological deficits were somewhat more common among patients who had received only oral rather than intravenous therapy. In addition, of the six Lyme patients with no objective neuropsychological test deficits but subjective complaints of memory impairment, five had the highest depression scores of the entire group of 15, suggesting that depression in some Lyme patients may account for the subjective experience of cognitive dysfunction.
Psychiatric Manifestations of Lyme Disease
A limited but ever increasing literature is beginning to suggest that psychiatric disorders may be part of the clinical profile of Lyme disease. Before reviewing this literature, we present the following case:
Ms. A, a previously healthy 18-year-old college freshman, suddenly developed severe and sustained anxiety, depersonalization, and panic attacks associated with insomnia and appetite loss. She consulted the university health services. After evaluation by both a psychologist and an internist, rest was recommended, under the assumption that these symptoms represented an adjustment reaction to being away from home. As her symptoms worsened, Ms. A began to fear that she was going crazy.
Two weeks later, Ms. A returned home on a medical leave of absence. An extensive medical workup revealed no abnormalities except for a positive Lyme ELISA titer. A Western blot for B. burgdorferi also came back positive. Ms. A insisted on getting a spinal tap. Although the cell count and total protein were normal, the CSF revealed IgG antibodies to B. burgdorferi . The diagnosis of CNS Lyme disease was made. The patient was treated with a 6-week course of intravenously administered antibiotics, and over the course of the following 3 months she felt 80% better.
Noteworthy in this case is that a diagnosis of Lyme disease was never considered by the college's counselor and internist. The private community physician also did not suspect Lyme disease, but because the patient lived in an endemic area and because Lyme disease is well-known as the "new great imitator" (59), this doctor included a Lyme test in the battery of blood tests. After the ELISA results came back positive, the patient recalled a large annular rash several months earlier that had been followed by a brief period of moderate headaches and uncharacteristic fatigue. Ms. A did not have any joint pain, radiculopathy, cranial nerve palsies, or cardiac symptoms. Her primary manifestation of Lyme disease was psychiatric. Because ms. A's CSF studies demonstrated borrelial antibodies and because her psychiatric symptoms resolved so rapidly after intravenous antibiotic treatment, active CNS infection was presumed to have been the cause of the severe anxiety and depersonalization. Had Ms. A's serologic Lyme test results come back falsely negative, the diagnosis of Lyme disease would have been missed. It should be emphasized, however, that careful history taking by a clinician well-versed in the clinical spectrum of Lyme disease would have suggested the diagnosis even in the absence of positive serologic test results.
The psychiatrist's evaluation becomes complicated when psychiatric symptoms emerge after the patient has already received a standard course of antibiotic treatment. Such a situation developed with Ms. A.
One year later, Ms. A developed a return of anxiety with panic attacks and agoraphobia. In addition, she developed rare deja vu episodes, repetitive musical hallucinations, and intrusive obsession thoughts and images. Results of a repeat spinal tap were normal on routine testing with a nonreactive CSF ELISA for B. burgdorferi antibodies. An EEG after sleep deprivation revealed intermittent slowing in the right and left temporal areas with rare sharp waves. Ms. A was treated for 6 months with imipramine, with complete resolution of her panic attacks and agoraphobia. With time, the obsessional thoughts, deja vu experiences, and musical hallucinations also resolved.
Although the panic disorder and obsessive compulsive disorder may have been mere epiphenomena with the Lyme disease, the lack of a family history of these disorders and the normal premorbid history suggest that Lyme disease may have triggered these symptoms. Severe anxiety and panic attacks have been previously described in Lyme disease (60). Obsessive-compulsive disorder has been associated with CNS infections, such as the encephalitis epidemic of 1916-1922 (61), and has also been associated with Lyme disease (62); anti-neuronal antibodies triggered by systemic infection may induce certain subtypes of obsessive-compulsive disorder (63). Persistent infection could not be completely ruled out in Ms. A's case, given the fact that patients with late CNS Lyme disease may have no demonstrable CSF abnormalities on currently available testing. However, because of the absence of other systemic Lyme symptoms and the improvement with psychiatric medications, this patient has continued to be treated symptomatically and observed for the possible reemergence of signs of CNS infection.