34-11: Lyme Disease (Lyme Borreliosis)

This illness, named after the town of Old Lyme, Connecticut, is the most common tick-borne disease in the United States and Europe and is caused by genospecies of the spirochete B burgdorferi. In the United States, the causative genospecies is B burgdorferi senu strictu, whereas in Europe and Asia B garinii and B afzelii predominate. Most US cases are reported from the mid-Atlantic, northeastern, and north central regions of the country. The true incidence of Lyme disease is not known for a number of reasons: (1) serologic tests are not standardized (see below); (2) clinical manifestations are nonspecific; and (3) even with reliable testing, serology is insensitive in early disease.

The tick vector of Lyme disease varies geographically and is Ixodes scapularis in the northeastern, north central, and mid-Atlantic regions of the United States; Ixodes pacificus on the West Coast; Ixodes ricinus in Europe; and Ixodes persulcatus in Asia. The disease also occurs in Australia. Mice and deer make up the major animal reservoir of B burgdorferi, but other rodents and birds may also be infected. Domestic animals such as dogs, cattle, and horses can also develop clinical illness, usually manifested as arthritis.

Ticks feed once during each of their three stages of life. Larval ticks feed in late summer, nymphs in the following spring and early summer, and adults during the fall. In the northeastern United States and the mid-Atlantic states, the preferred host for the nymphs and larvae is the white-footed mouse (the black-striped mouse in Europe). This animal is tolerant of infection—a fact that is critical in maintaining infection, since the mouse can remain spirochetemic and transmit the agent to the larvae the following spring after being infected by the nymphal form in early summer. Adult ticks prefer the white-tailed deer as host. Although only 20–25% of nymphs harbor spirochetes compared with 50–65% of adults, most infections occur in the spring and summer (when nymphs are active), and fewer cases occur in the cooler months (October to April), when adults feed. This is probably due to the greater abundance of nymphs; greater human outdoor activity in spring and summer, when nymphs feed; and the fact that adult ticks are larger, easier to detect by the human host, and thus can be removed before disease is transmitted. Less than 1% of larvae are infected with spirochetes, and transmission of the disease through contact with larvae is unlikely. In the western United States, the I pacificus nymph prefers to feed on lizards, which are not susceptible to infection. It is only the few nymph and larval ticks that feed on wood rats, which can be infected, that are capable of transmitting disease.

Under experimental conditions, ticks must feed for 24–36 hours or longer to transmit infections. Most cases are reported in the spring and summer months. Human epidemiologic studies have indicated that the incidence of disease is significantly higher when tick attachment is for longer than 72 hours than if it is less than 72 hours, though rare cases have been documented with attachment of less than 24 hours. In addition, the percentage of ticks infected varies on a regional basis. In the northeastern and midwestern United States, 15–65% of I scapularis ticks are infected with the spirochete; in the west, less than 5% of I pacificus are infected. These are important epidemiologic features in assessing the likelihood that tick exposure will result in disease. Eliciting a history of brushing a tick off the skin (ie, the tick was not feeding) or removing a tick on the same day as exposure (ie, the tick did not feed long enough) decreases the likelihood that infection will develop.

Ixodes ticks are smaller than the more common dog ticks (Dermacentor variabilis). Larvae are smaller than 1 mm, and the adult female is 2–3 mm in size, with a red body and black legs. After a blood meal, ticks can reach two to three times their unengorged size. Because the Ixodes tick is so small, the bite is usually painless and goes unnoticed. After feeding, the tick drops off in 2–4 days. If a tick is found, it should be removed immediately. The best way to accomplish this is to use fine-tipped tweezers to pull firmly and repeatedly on the tick’s mouth part—not the tick’s body—until the tick releases its hold. Saving the tick in a bottle of alcohol for future identification may be useful, especially if symptoms develop.

The three stages of Lyme disease are classified based on early or late manifestations of the disease and whether it is localized or disseminated.

A. Symptoms and Signs

1. Stage 1, early localized infection

Stage 1 infection is characterized by erythema migrans (Figure 6–26). About 1 week after the tick bite (range, 3–30 days; median 7–10 days), a flat or slightly raised red lesion appears at the site, which is commonly seen in areas of tight clothing such as the groin, thigh, or axilla. This lesion expands over several days. Although originally described as a lesion that progresses with central clearing (“bulls-eye” lesion), often there is a more homogeneous appearance or even central intensification. About 10–20% of patients either do not have typical skin lesions or the lesions go unnoticed. Most patients with erythema migrans will have a concomitant viral-like illness (the “summer flu”) characterized by myalgias, arthralgias, headache, and fatigue. Fever may or may not be present. Even without treatment, the symptoms and signs of erythema migrans resolve in 3–4 weeks. Although the classic lesion of erythema migrans is not difficult to recognize, atypical forms can occur that may lead to misdiagnosis. Vesicular, urticarial, and evanescent erythema migrans have all been reported. Chemical reactions to tick and spider bites (these usually recede in 24–48 hours, whereas erythema migrans increases in size in this time period), drug eruptions, urticaria, and staphylococcal and streptococcal cellulitis have been mistaken for erythema migrans. Southern tick–associated rash illness (STARI) has a similar appearance, but it occurs in geographically distinct areas of the United States.

Completely asymptomatic disease, without erythema migrans or flu-like symptoms, can occur but is very uncommon in the United States. Serosurveys done as part of vaccination trials found asymptomatic seroconversion in 7% or less of the population.

2. Stage 2, early disseminated infection

Up to 50–60% of patients with erythema migrans are bacteremic, (especially if multiple lesions are present) leading to dissemination of the spirochete resulting in a wide variety of symptoms and signs and within days to weeks of the original infection, secondary skin lesions develop in about 50% of patients. These lesions are similar in appearance to the primary lesion but are usually smaller. Malaise, fatigue, fever, headache (sometimes severe), neck pain, and generalized achiness are common with the skin lesions. Most symptoms are transient. After hematogenous spread, some patients experience cardiac (4–10% of patients) or neurologic (10–15% of patients) manifestations, including myopericarditis, with atrial or ventricular arrhythmias and heart block. Neurologic manifestations include both the central and peripheral nervous systems. The most common CNS manifestation is aseptic meningitis with mild headache and neck stiffness. The most common peripheral manifestation is a cranial nerve VII neuropathy, ie, facial palsy (usually unilateral but can be bilateral, see Figure 24–1). A sensory or motor radiculopathy and mononeuritis multiplex occur less frequently. Conjunctivitis, keratitis, and, rarely, panophthalmitis can also occur. Rarely, skin involvement can be manifested as a cutaneous hypopigmented lesion called a borrelial lymphocytoma.

3. Stage 3, late persistent infection

Stage 3 infection occurs months to years after the initial infection and again primarily manifests itself as musculoskeletal, neurologic, and skin disease. In early reports, musculoskeletal complaints developed in up to 60% of patients, but with early recognition and treatment of disease, this has decreased to less than 10%. The classic manifestation of late disease is a monoarticular or oligoarticular arthritis most commonly affecting the knee or other large weight bearing joints. While these joints may be quite swollen, these patients generally report less pain compared to patients with bacterial septic arthritis. Even if untreated, the arthritis is self-limited, resolving in a few weeks to months. Multiple recurrences are common but are usually less severe than the original disease. Joint fluid reflects an inflammatory arthritis with a mean white blood cell count of 25,000/mcL (0.025/L) with a predominance of neutrophils. Chronic arthritis develops in about 10% of patients. The pathogenesis of chronic Lyme arthritis may be an immunologic phenomenon rather than persistence of infection. The observations that persons with chronic arthritis have an increased frequency of HLA-DR4 gene expression, antibodies to OspA and OspB protein in joint fluid (major outer surface proteins of B burgdorferi), lack B burgdorferi DNA in synovial fluid as detected by PCR, and often do not respond to antibiotics—all support the inference of an immunologic mechanism.

Rarely, the nervous system (both central and peripheral) can be involved in late Lyme disease. In the United States, a subacute encephalopathy, characterized by memory loss, mood changes, and sleep disturbance, is seen. In Europe, a more severe encephalomyelitis caused by B garinii is seen and presents with cognitive dysfunction, spastic paraparesis, ataxia, and bladder dysfunction. Peripheral nervous system involvement includes intermittent paresthesias, often in a stocking glove distribution, or radicular pain.

The cutaneous manifestation of late infection, which can occur up to 10 years after infection, is acrodermatitis chronicum atrophicans. It has been described mainly in Europe after infection with B afzelii, a genospecies that commonly causes disease in Europe but not the United States. There is usually bluish-red discoloration of a distal extremity with associated swelling. These lesions become atrophic and sclerotic with time and eventually resemble localized scleroderma. Cases of diffuse fasciitis with eosinophilia, an entity that resembles scleroderma, have been rarely associated with infection with B burgdorferi.

B. Laboratory Findings

The diagnosis of Lyme disease is based on both clinical manifestations and laboratory findings. The US Surveillance Case Definition specifies a person with exposure to a potential tick habitat (within the 30 days just prior to developing erythema migrans) with (1) erythema migrans diagnosed by a clinician or (2) at least one late manifestation of the disease and (3) laboratory confirmation as fulfilling the criteria for Lyme disease.

Nonspecific laboratory abnormalities can be seen, particularly in early disease. The most common are an elevated sedimentation rate of more than 20 mm/h seen in 50% of cases, and mildly abnormal liver biochemical tests are present in 30%. The abnormal liver biochemical tests are transient and return to normal within a few weeks of treatment. A mild anemia, leukocytosis (11,000–18,000/mcL) (0.011–0.018/L), and microscopic hematuria have been reported in 10% or less of patients.

Laboratory confirmation requires serologic tests to detect specific antibodies to B burgdorferi in serum, preferably by ELISA and not by indirect immunofluorescence assay (IFA), which is less sensitive and specific and can cause misdiagnosis. A two-test approach is recommended for the diagnosis of active Lyme disease, with all specimens positive or equivocal by ELISA then confirmed with a Western immunoblot assay that can detect both IgM and IgG antibodies. IgM antibody appears first 2–4 weeks after onset of erythema migrans, peaks at 6–8 weeks, and then declines to low levels after 4–6 months of illness. The presence of IgM antibody in patients with prolonged symptoms persisting for several months is likely to be a false-positive result. IgG occurs later (6–8 weeks after onset of disease), peaks at 4–6 months, and may remain elevated at low levels indefinitely despite appropriate therapy and resolution of symptoms. When a Western immunoblot is done during the first 4 weeks of illness, both IgM and IgG should be tested. If the initial ELISA is negative, the result is considered final; Western immunoblot should not be performed. A negative Western immunoblot following an initial positive ELISA indicates as false-positive ELISA; the patient does not have evidence of exposure to B burgdorferi. A positive immunoblot requires that antibodies are detected against two (for IgM) or five (for IgG) specific protein antigens from B burgdorferi. Patients who may have acquired infection outside of the United States may require additional tests that include peptide antigens (C6, PepC10), which are conserved across species and strains.

If a patient with suspected early Lyme disease has negative serologic studies, acute and convalescent titers should be obtained since up to 50% of patients with early disease can be antibody negative in the first several weeks of illness. A fourfold rise in antibody titer would be diagnostic of recent infection. In patients with later stages of disease, almost all are antibody positive. False-positive reactions in the ELISA and IFA have been reported in juvenile rheumatoid arthritis, rheumatoid arthritis, systemic lupus erythematosus, infectious mononucleosis, subacute infective endocarditis, syphilis, relapsing fever, leptospirosis, enteroviral and other viral illnesses, and patients with gingival disease (presumably because of cross-reactivity with oral treponemes). False-negative serologic reactions occur early in illness, and antibiotic therapy early in disease can abort subsequent seroconversion.

The diagnosis of late nervous system Lyme disease is often difficult since clinical manifestations, such as subtle memory impairment, may be difficult to document. Most patients have a history of previous erythema migrans or monoarticular or polyarticular arthritis, and the vast majority have antibody present in serum. When CSF is sampled from patients with encephalopathy, there may be evidence of inflammation (pleocytosis or elevated protein, or both), and localized antibody production, ie, a ratio of CSF to serum antibody of greater than 1.0. PCR has low sensitivity and is not recommended for routine diagnosis. Elevated CSF levels of the chemokine CXCL13 have been associated with CNS Lyme disease, but they can also occur in other infections such as neurosyphilis. Patients with late disease and peripheral neuropathy almost always have positive serum antibody tests, usually have abnormal electrophysiology tests, and may have abnormal nerve biopsies showing perivascular collections of lymphocytes; however, the CSF is usually normal and does not demonstrate local antibody production.

Caution should be exercised in interpreting serologic tests because they are not subject to national standards, and inter-laboratory variation is a major problem. In addition, some laboratories perform tests that are entirely unreliable and should never be used to support the diagnosis of Lyme disease (eg, the Lyme urinary antigen test, immunofluorescent staining for cell wall–deficient forms of B burgdorferi, lymphocyte transformation tests, using PCR on inappropriate specimens such as blood or urine). Finally, testing is often done in patients with nonspecific symptoms such as headache, arthralgia, myalgia, fatigue, and palpitations. Even in endemic areas, the pretest probability of having Lyme disease is low in these patients, and the probability of a false-positive test result is greater than that of a true-positive result. For these reasons, the CDC has established guidelines for laboratory evaluation of patients with suspected Lyme disease:

1. The diagnosis of early Lyme disease is clinical (ie, exposure in an endemic area, with clinician-documented erythema migrans), and does not require laboratory confirmation. (Tests are often negative at this stage.)

2. Late disease requires objective evidence of clinical manifestations (recurrent brief attacks of monoarticular or oligoarticular arthritis of the large joints; lymphocytic meningitis, cranial neuritis [facial palsy], peripheral neuropathy or, rarely, encephalomyelitis—but not headache, fatigue, paresthesias, or stiff neck alone; atrioventricular conduction defects with or without myocarditis) and laboratory evidence of disease (two-stage testing with ELISA or IFA followed by Western blot, as described above).

3. Patients with nonspecific symptoms without objective signs of Lyme disease should not have serologic testing done. It is in this setting that false-positive tests occur more commonly than true-positive tests.

4. The role of serologic testing in nervous system Lyme disease is unclear, as sensitivity and specificity of CSF serologic tests have not been determined. However, it is rare for a patient to have positive serologic tests on CSF without positive tests on serum (see below).

5. Other tests such as the T cell proliferative assay and urinary antigen detection have not yet been studied well enough to be routinely used.

Cultures for B burgdorferi can be performed but are not routine and are usually reserved for clinical studies. Aspiration of erythema migrans lesions has yielded positive cultures in up to 30% of cases, whereas culture of a 2-mm punch biopsy is positive in 50–70%. PCR of a skin biopsy is even more sensitive, with positivity rates of 80%. In early disease, blood cultures are positive in up to 50% if large volumes (9 mL) are used, but CSF is rarely culture positive. The ability to culture organisms from skin lesions is greatly influenced by antibiotic therapy. Even a brief course of several days will result in negative cultures. Special silver staining of chronically inflamed synovial tissue demonstrates spirochetes in one-third of patients.

PCR is very specific for detecting the presence of Borrelia DNA, but sensitivity is variable and depends on which body fluid is tested, the stage of the disease, and collection and testing technique. In general, PCR is more sensitive than culture, especially in chronic disease but is not available in many clinical laboratories. Testing should not be done on blood or urine but has been successfully performed on synovial fluid and CSF. Up to 85% of synovial fluid samples are positive in active arthritis. In contrast, 38% of CSF samples in acute CNS Lyme disease are PCR positive compared with only 25% in chronic CNS disease. However, whether a positive PCR indicates persistence of viable organisms that will respond to further treatment or is a marker for residual DNA (not active infection) has not been clarified. A negative PCR result does not rule out disease.

Several observational studies, including serosurveys of over 2000 pregnant women in endemic areas, and case control studies of nearly 800 infants with congenital heart disease have not found any association between seropositivity or tick exposure in prospective mothers and the prevalence of congenital malformations, fetal death, and prematurity. Based on several observational studies, B burgdorferi infection in pregnant women has not been associated with congenital syndromes, unlike other spirochetal illnesses such as syphilis.

Some patients and advocacy groups have claimed either a post–Lyme syndrome (in the presence of positive laboratory tests and after appropriate treatment) or “chronic Lyme disease” in which tests may all be negative. Both entities include nonspecific symptoms such as fatigue, myalgias, and cognitive difficulties (see Prognosis below). Expert groups are in agreement that there are no data to support that ongoing infection is the cause of either syndrome.

There is no human vaccine currently available. Simple preventive measures such as avoiding tick-infested areas, covering exposed skin with long-sleeved shirts and wearing long trousers tucked into socks, wearing light-colored clothing, using repellents, and inspecting for ticks after exposure will greatly reduce the number of tick bites. Environmental controls directed at limiting ticks on residential property would be helpful, but trying to limit the deer, tick, or white-footed mouse populations over large areas is not feasible.

Prophylactic antibiotics following tick bites is recommended in certain high-risk situations if all of the following criteria are met: (1) a tick identified as an adult or nymphal I scapularis has been attached for at least 36 hours; (2) prophylaxis can be started within 72 hours of the time the tick was removed; (3) more than 20% of ticks in the area are known to be infected with B burgdorferi; and (4) there is no contraindication to the use of doxycycline (not pregnant, age greater than 8 years, not allergic). As noted above, I scapularis ticks in the northeastern, mid-Atlantic, and upper midwestern states are infected with B burgdorferi at a high rate (greater than 20%) whereas I pacificus ticks on the West Coast are not (less than 5%). The medication of choice for prophylaxis is a single 200-mg dose of doxycycline. If doxycycline is contraindicated, no prophylaxis should be given and the patient should be closely monitored for early disease, since short course prophylactic therapy with other agents has not been studied, and if early disease does develop, appropriate therapy is very effective in preventing long-term sequelae. Individuals who have removed ticks (including those who have had prophylaxis) should be monitored carefully for 30 days for possible coinfections.

Lyme disease, babesiosis (see Chapter 35), and human granulocytic anaplasmosis (formerly human granulocytic ehrlichiosis) (see Chapter 32) are endemic in similar areas of the country and are transmitted by the same tick, I scapularis. Coinfection with two or even all three of these organisms can occur, causing a clinical picture that is not “classic” for any of these diseases. The presence of erythema migrans is highly suggestive of Lyme disease, whereas flu-like symptoms without rash are more suggestive of babesiosis or anaplasmosis. The complete blood count is usually normal in Lyme disease, but in patients with Lyme disease and babesiosis, anemia and thrombocytopenia are more common. Patients with Lyme disease and anaplasmosis are more likely to have leukopenia. Coinfection should be considered and excluded in patients who have persistent high fevers 48 hours after starting appropriate therapy for Lyme disease; in patients with persistent symptoms despite resolution of rash; and in those with anemia, leukopenia, or thrombocytopenia.

Antibiotic sensitivity of B burgdorferi has been established in vitro. Tetracycline is effective against the spirochete, but penicillin is only moderately so. Erythromycin is effective in vitro but has been disappointing in clinical trials. Ampicillin, ceftriaxone, azithromycin, cefuroxime, and imipenem are also effective in vitro, but aminoglycosides, ciprofloxacin, and rifampin are not.

Present recommendations for therapy are outlined in Table 34–4. For erythema migrans, antibiotic therapy shortens the duration of rash and prevents late sequelae. Doxycycline is most commonly used and has the advantage of being active against anaplasmosis, a common coinfection (Anaplasma phagocytophilum [formerly Ehrlichia]). It has proven effective in shorter courses of 10–14 days compared to other regimens. Amoxicillin is also effective and is recommended for pregnant or lactating women and for those who cannot tolerate doxycycline. Cefuroxime axetil is as effective as doxycycline, but because of its cost, it should be considered an alternative choice for those who cannot tolerate doxycycline or amoxicillin or for those in whom the medications are contraindicated. Erythromycin and azithromycin are less effective, associated with higher rates of relapse, and are not recommended as first-line therapy.

Isolated facial palsy (without meningitis or peripheral neuropathy) can be treated with doxycycline, amoxicillin, or cefuroxime axetil for 2–3 weeks. Although therapy does not affect the rate of resolution of the cranial neuropathy, it does prevent development of late manifestations of disease.

The need for a lumbar puncture in patients with seventh nerve palsy is controversial. Some clinicians perform lumbar puncture on all patients with facial palsy and others only if there are symptoms or signs of meningitis. If meningitis is present, therapy with a parenteral antibiotic is indicated. Ceftriaxone is most commonly used, but penicillin is equally efficacious. In European countries, doxycycline 400 mg/d orally for 14 days is frequently used and is comparable in efficacy to ceftriaxone.

Patients with atrioventricular block or myopericarditis (or both) can be treated with either oral or parenteral agents for 2–3 weeks. Hospitalization and observation is indicated for symptomatic patients, those with second- or third-degree block, and those with first-degree block with a PR interval of 300 milliseconds or more. Once stabilized, hospitalized patients can be transitioned to one of the oral regimens to complete therapy.

Therapy of arthritis is difficult because some patients do not respond to any therapy, and those who do respond may do so slowly. Oral agents (doxycycline, amoxicillin, or cefuroxime axetil) are as effective as intravenous regimens (ceftriaxone, cefotaxime, or penicillin). A reasonable approach to the patient with Lyme arthritis is to start with oral therapy for 28 days. If there is partial resolution, re-treat with an additional 28 days of the same oral regimen. However, if there has been no response or worsening with initial oral therapy, switch to intravenous ceftriaxone for 14–28 days. If arthritis persists after re-treatment, symptomatic therapy with nonsteroidal anti-inflammatory drugs is recommended. For severe refractory pain, synovectomy may be required.

Based on the limited published data, therapy of Lyme disease in pregnancy should be the same as therapy in other patients with the exception that doxycycline should not be used.

Clinicians may encounter patients with nonspecific symptoms (such as fatigue and myalgias) and positive serologic tests for Lyme disease who request (or demand) therapy for their illness. It is important in managing these patients to remember (1) that nonspecific symptoms alone are not diagnostic; (2) that serologic tests are fraught with difficulty (as noted above), and in areas where disease prevalence is low, false-positive serologic tests are much more common than true-positive tests; and (3) that parenteral therapy with ceftriaxone for 2–4 weeks is costly and has been associated with significant adverse effects, including cholelithiasis and Clostridioides difficile colitis. Parenteral therapy should be reserved for those most likely to benefit, ie, those with cutaneous, neurologic, cardiac, or rheumatic manifestations that are characteristic of Lyme disease.

Most patients respond to appropriate therapy with prompt resolution of symptoms within 4 weeks. True treatment failures are thus uncommon, and in most cases re-treatment or prolonged treatment of Lyme disease is instituted because of misdiagnosis or misinterpretation of serologic results (both IgG and IgM antibodies can persist for prolonged periods despite adequate therapy) rather than inadequate therapy or response. Prolonged courses of antibiotic therapy for nonspecific symptoms that persist after completion of appropriate assessment (and treatment, if necessary) for Lyme disease is not recommended. The terms "chronic Lyme disease" or "post-Lyme disease syndrome" have been applied to a heterogeneous group of patients with documented Lyme disease who have been adequately treated but have persistent poorly defined, nonspecific symptoms such as fatigue, arthralgias, myalgias, and neurocognitive complaints. Although some providers treat these patients with multiple or prolonged courses of oral or intravenous antibiotics, there are no data suggesting that this is efficacious, and the prolonged use of antibiotics in this setting is emphatically discouraged.

The long-term outcome of adult patients with Lyme disease is generally favorable, but some patients have chronic complaints. Joint pain, memory impairment, and poor functional status secondary to pain are common subjective complaints in patients with Lyme disease, but physical examination and neurocognitive testing fail to document the presence of these symptoms as objective sequelae. Similarly, in highly endemic areas, some patients with a diagnosis of Lyme disease note pain, fatigue, and an inability to perform certain physical activities when followed for several years. However, these complaints occur just as commonly in age-matched controls without a history of Lyme disease. Attempts to document chronic cardiac disease in patients treated for Lyme disease also have been unsuccessful. The long-term outcome of treated nervous system Lyme disease is favorable, with complete recovery in 75% of patients. Of the remaining individuals, only 12% had sequelae that affected their daily activities.

Immunity is not complete after Lyme disease. Reinfection, although uncommon, is predominantly seen in patients successfully treated for early disease (erythema migrans) in whom antibody titers do not develop. Clinical manifestations and serologic response are similar to an initial infection.

Consultation with an infectious diseases specialist with experience in diagnosing and treating Lyme disease can be helpful in atypical or prolonged cases.

Admission for parenteral antibiotics is indicated for any patient with symptomatic CNS or cardiac disease as well as those with second- or third-degree atrioventricular block, or first-degree block with a PR interval of 300 milliseconds or more.