Article Text

Download PDFPDF

Lyme disease: diagnosis and management
  1. Amy L Ross Russell1,
  2. Matthew S Dryden2,
  3. Ashwin A Pinto1,
  4. Joanna K Lovett1
  1. 1 Neurology Department, Wessex Neurosciences Centre, Southampton General Hospital, Tremona Road, Southampton, UK
  2. 2 Department of Microbiology and Infection, Hampshire Hospitals NHS Foundation Trust & Rare and Imported Pathogens Department, PHE, Porton, Salisbury, UK
  1. Correspondence to Dr Amy L Ross Russell, Neurology Department, Wessex Neurosciences Centre, Southampton General Hospital, Southampton SO16 6YD, UK; amy.rossrussell{at}


Lyme disease (borreliosis) is a tick-borne bacterial infection caused by the spirochaete Borrelia burgdoferi, transmitted by hard-backed Ixodes ticks. Actual numbers of cases are increasing and it appears that the distribution across the UK is widening; however, it occurs most frequently in area of woodland, with temperate climate. It typically presents in mid to late summer. Lyme disease is a multisystem disease. The nervous system is the second most commonly affected system after the skin. Other systemic manifestations, such as carditis, keratitis, uveitis and inflammatory arthritis, rarely occur in European Lyme disease. In 2018, the National Institute for Health and Care Excellence has updated its guidelines on the diagnosis and management of Lyme disease. Here, we highlight important aspects of this guidance and provide a more detailed review of the clinical spectrum of neuroborreliosis, illustrated by cases we have seen.

  • lyme disease
  • neuroborreliosis

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Lyme disease (borreliosis) is a tick-borne bacterial infection with which every UK neurologist should be familiar. The reported number of cases is increasing,1 and the distribution across the UK appears to be widening. It is uncertain whether this is from environmental and ecological changes, or through greater medical and public awareness of the infection. This year, the UK’s National Institute for Health and Care Excellence (NICE) produced updated guidelines on the diagnosis and management of this disease.2 In this review, we highlight important aspects of this guidance and provide a more detailed review of the clinical spectrum of neuroborreliosis, illustrated by cases we have seen.


Lyme disease is a consequence of infection with members of the Borrelia burgdorferi group of spirochaetes. It is a zoonosis—a disease that exists in animals but can be transmitted to humans, in this case by hard-backed Ixodes ticks. In Europe, this is the Ixodes ricinus subgroup, which feeds on mammalian hosts, especially deer, sheep and rodents. Humans are incidental hosts. Each stage of the tick life cycle has a single blood feed. The tick may acquire the infection during its larval or nymph stage. Humans are therefore usually infected via bites from nymphs or, less commonly, adults (figure 1).

Figure 1

Life cycle of Borrelia burgdorferi. 34

The life cycle of the spirochaete dictates the seasonality and geographical distribution of Lyme disease. Ticks mature into larvae and nymphs that feed over the warm summer months, and case incidence therefore mirrors the local climate. Most UK cases present in June to September, following tick bites in May to August (figure 2). Other northern hemisphere tick-borne infections, endemic in parts of Euroasia and North America, such as tick-borne encephalitis, rickettsia, Ehrlichia, Coxiella infections and human babesiosis, have not been detected in the UK. Anaplasma may be transmitted but appears to be of low clinical significance.

Figure 2

Month of tick bite, first symptom and confirmation of diagnosis.4

The incidence of Lyme disease in the UK is 1.7/100 000.3 Since October 2010, all cases identified by National Health Service laboratories are reported to Public Health England. As cases of erythema migrans are often treated empirically, without serological testing, this incidence is likely to be an underestimate, and local rates will vary depending on local health practitioner behaviours, as well as the distribution of tick populations.

Within the UK, Lyme disease is most common in the South West of England (particularly in and around the New Forest), Thetford Forest, the Lake District and in the highlands of Scotland. The incidence in the Winchester, Hampshire area is 9.8/100 000, over five times higher than the national incidence.4 These prevalent areas are wooded, with a temperate climate, and relatively high deer populations. However, it can be acquired in most regions of the UK. It has even been acquired in some city parks, for example Richmond Park in London.

Internationally, Lyme disease is seen in the environmentally similar areas of Scandinavia and Central Europe (mainly Austria, Slovenia and Germany), the Northeast and Upper Mid-west USA, and some parts of Asia.5 There are no accurate international data on incidence of neuroborreliosis, and obtaining such data is challenging due to variation in case definition.

There are several different Borrelia subspecies, which vary in geographical distribution and in their clinical manifestations of infection. In Europe, including the UK, the prevalent organisms are B. afezelii and B. garinii, although infections from Borrelia burgdorferi sensu stricto, B. spielmanii and B. bavariensis also occur. B. garinii is associated with neuroborreliosis more frequently than the other subspecies, and this may explain why UK case series report higher rates of neurological involvement (16% and 25%)4 6 7 than those from other countries.5 8

Lyme disease is a systemic condition and affects multiple organs, particularly the heart causing carditis and conduction abnormalities, the joints causing an inflammatory arthritis, the eyes causing uveitis and keratitis, and the skin causing lymphocytoma and acrodermatitis chronica atrophicans.2 Arthritis and carditis are rare complications in Europe. In the United States, where B. burgdorferi sensu stricto is the only pathogenic organism, Lyme arthritis is more common, and carditis, although still less than 1%, is reported more frequently.5

Asymptomatic seropositivity

A high proportion of people living in endemic areas will be seropositive without symptoms. In one UK study, 25% of New Forest workers were seropositive.9 In southern Germany, approximately 20% of the population develop seropositivity over their lifetime, with rates increasing with each decade.10 Incidental seropositivity suggests previous asymptomatic exposure, not active disease. In the absence of symptoms, this does not require treatment.


Checking for, and early removal of, ticks is preventative. Tick bites that go unnoticed are more likely to transmit the bacterium as these are usually attached for longer, which dramatically increases the risk of transmission. Removal of a tick within 24 hours usually prevents infection, whereas when an infected tick is attached for over 48 hours the risk of transmission is high.11 Antibiotics are not recommended for tick bites without skin changes. There have been attempts at vaccine development, but trials have proved disappointing and none is available for human use.

Clinical consequences of infection

Erythema migrans and other skin lesions

The typical rash associated with Lyme disease is called erythema migrans; it looks like a target, with localised erythema spreading out from the site of inoculation (figure 3). About half of patients with neuroborreliosis report no rash, possibly because the primary infection has been in an inaccessible site or under the hairline.

Figure 3

Erythema migrans: the typical rash associated with borreliosis.

Erythema migrans appears 7–10 days after inoculation and usually resolves spontaneously. The rash can be quite varied but is invariably spreading. Rashes that are hot, itchy and resolve within 48 hours are more likely to be caused by allergy and not by borreliosis.2 Identifying the skin lesion is important diagnostically, as other causes of painful radiculitis, such as Herpes zoster, may also present with different, characteristic rashes.

Erythema migrans should be treated empirically without waiting for serological confirmation. Antibodies to B. burgdorferi may take 4–8 weeks to be detected.

Early neuroborreliosis

Neurological involvement is the second most common clinical manifestation in the UK and Europe, after erythema migrans. Approximately 5% of individuals with an untreated erythema migrans will develop neuroborreliosis, usually 4–6 weeks after tick exposure.12 Over 95% patients present within 6 months of infection, and any neurological manifestations during this time are termed ‘early’ neuroborreliosis.

Only 25% of patients with neuroborreliosis recall a tick bite and about 50% report a localised skin reaction13 so the absence of a history of tick bite or rash does not exclude the diagnosis.

Case 1

A middle-aged man presented with a 3-week history of intermittent fever, malaise and lethargy, and a 1-day history of right facial droop. He reported a recent erythematous rash on his torso and showed photographs showing a rash consistent with erythema migrans. On examination, there was only a right-sided lower motor neuron facial weakness. Serum C reactive protein was mildly elevated at 30 mg/L and his erythrocyte sedimentation rate was 32 mm/hour. Infectious mononucleosis screen, syphilis, HIV, hepatitis B and C serology were negative. CSF contained 63 white cells (60 lymphocytes and three neutrophils). Protein concentration was 1.03 g/L and glucose 3.1 mmol/L (paired plasma glucose 5.8 mmol/L). His serum screening enzyme immunoassay was reactive to both IgG and IgM. Reference laboratory testing confirmed a reactive enzyme immunoassay to IgG, IgM and C6, with a positive immunoblot to IgG and IgM. He was treated for 21 days with intravenous ceftriaxone. At 1 month follow-up his facial nerve weakness had fully resolved.

Case 2

A middle-aged man presented 8 weeks after the onset of a target-shaped rash on his right leg. The rash had enlarged, and been associated with significant swelling of his calf and ankle. Two weeks later he began experiencing sharp pain radiating from his buttock, down the lateral aspect of the right thigh and into the lower leg, which increased when sitting, moving and at night. On examination, there was isolated weakness of knee flexion with reduced pinprick sensation over L5 and S1 dermatomes. His right ankle jerk was absent. Plantar responses were flexor. Serum C reactive protein and erythrocyte sedimentation rate were normal. Cerebrospinal fluid (CSF) contained 154 white cells (145 lymphocytes, 9 neutrophils) and 68 red cells. Protein concentration was 1.53 g/L, with 3.8 mmol/L glucose (paired plasma glucose 5.8 mmol/L). MR scan of lumbar spine showed no compressive lesion of L5/S1. His serum screening enzyme immunoassay was reactive to both IgG and IgM. He was treated with 21 days of intravenous ceftriaxone. At 3-month follow-up, his strength was recovered, his pain was well controlled, but there was residual impaired pinprick sensation over an L5 distribution.

Bannwarth’s syndrome: painful lymphocytic meningoradiculitis

The pathology in early neuroborreliosis is localised predominantly to the CSF and meninges. Therefore, most European patients with neuroborreliosis present with features of Bannwarth’s syndrome. This is typically described as a triad of painful radiculitis, peripheral motor deficits and lymphocytic CSF, but patients may have an incomplete syndrome.12 13 In two UK case series, 91% and 97.5% of neuroborreliosis cases had facial palsy, multiple cranial neuropathies, radiculopathy or clinically apparent meningoencephalitis.4 6 The pain is usually the first and most prominent symptom, and in about 25% the only symptom.13 It is radicular, lancinating and typically worse at night.

The facial nerve is preferentially affected, often bilaterally, although almost always with asymmetrical onset. A facial palsy is the most common motor deficit in children. Other cranial nerves may be involved, most commonly third and sixth cranial nerve palsies in about 5% of neuroborreliosis cases.12

As with other neurotropic infections, the location of the tick bite may be relevant to the clinical syndrome: patients with tick bites close to their faces probably more commonly present with a facial palsy, and those bitten on limbs more commonly present with painful radiculitis. However, this association is difficult to establish and somewhat inconsistent.

Patients may complain of migrating areas of hyperesthesia or dysaesthesia, or non-specific cognitive symptoms of fatigue, depression and poor sleep, suggesting a mild encephalopathy. However, there is usually no objective evidence of sensory or cognitive deficit and these symptoms do not usually represent parenchymal central nervous system (CNS) infection.

Other CNS involvement

CNS involvement with vasculitis, myelitis or encephalitis occurs in less than 5% of cases. Most patients present with symptoms and signs of either cerebral infarction or myelopathy. Although there may be mild cognitive change, florid encephalitis with reduced consciousness and seizures is extremely unusual.14 15 There are interesting similarities to the clinical spectrum of meningovascular syphilis (caused by another spirochaete), which also presents with multiple lacunar and posterior circulation strokes, chronic pachymeningitis and progressive myelitis.

Case 3

A middle-aged man presented with a 5-day history of pain in both arms, and progressive numbness of his palms and inside his arms. He felt uncharacteristically depressed and very tired, and was struggling with fine motor tasks. He had reduced strength of grasp and grip and reduced light touch sensation on the palms, but neurological and systemic examination was otherwise normal. Serum C reactive protein and erythrocyte sedimentation rate were normal. MR scan of cervical spine showed a high-density signal extending between C5/6 and T1/2, with cord expansion (figure 4). The appearances were initially thought most consistent with an infiltrating neoplasm. The patient only remembered a history of skin reaction on the affected shoulder when camping 6 weeks before—diagnosed as cellulitis by his general practitioner and treated with erythromycin—when asked about a rash in relation to the diagnosis of Lyme. His serum screening Borrelia enzyme immunoassay was reactive, and this was confirmed on IgM and IgG immunoblot. He was treated with 1 week of intravenous ceftriaxone and a further 8 weeks of oral doxycycline. The patient declined to undergo a lumbar puncture. At 6-week follow-up, his symptoms had settled, neurological examination was normal and MR scan findings had resolved.

Figure 4

MR scan of spinal cord showing central cord lesion with expansion and T2 signal change.15

Case 4

A middle-aged woman presented with sudden onset of slurred speech, left-sided weakness and vertigo. She reported 8 months of patchy altered sensation over her body and scalp. She had no relevant medical or family history and took no medications. She was well travelled, including visiting Northern America in the previous year.

On examination, she had broken smooth pursuit movements and dysarthria. She had left-sided weakness and incoordination. She was mildly confused and lethargic. The plantar responses were extensor. MR scan of brain showed extensive signal abnormality on T2W images within the pons and medulla and with restricted diffusion in the pons on the right side, with patchy enhancement at this site following contrast administration (figure 5). MR angiography of the intracranial and extracranial vessels was normal. CSF contained 218 white cells and 3 red cells. The CSF protein concentration was 3.14 g/L and the CSF glucose was 1.9 mmol/L (paired plasma glucose was not sent).

Figure 5

Axial T2 (above) and axial T1 following contrast, and diffusion-weighted (Apparent Diffusion Coefficient) (below) MRI , showing right-sided enhancement in the pons.

She was covered empirically for bacterial meningitis including quadruple antituberculous therapy (with corticosteroids) given her travel history and CSF findings, until serum screening Borrelia enzyme immunoassay result was found to be reactive, confirmed with positive C6 antibodies, and positive immunoblot to IgG and IgM. CSF also contained antibodies to Borrelia. She was treated with 3 weeks of intravenous ceftriaxone and made a good recovery.

This is a case of probable acute posterior circulation stroke in the setting of a lymphocytic meningoencephalitis secondary to Borrelia infection. There is restricted diffusion in the brainstem, respecting the midline, suggesting a vascular territory. The infarct presumably resulted from small vessel vasculitis though MR angiography showed no vessel changes; a neuroinflammatory lesion is an alternative explanation.

Imaging in neuroborreliosis

Cases 3 and 4 both show how MR imaging can show clinically relevant inflammatory or neurovascular changes in the brain or spinal cord. Although the MR angiogram in case 4 was normal, there can be changes consistent with medium or large vessel vasculitis. It is less clear whether the ‘non-specific’ white matter changes in case 4 were relevant. White matter lesions often occur, especially in late or chronic Lyme syndromes. However, these are also common incidental findings and one study showed no significantly increased incidence of such changes in a small cohort of 16 patients diagnosed with chronic neuroborreliosis when compared with matched controls.16

CNS vasculitis

Lyme disease may cause both large and small vessel vasculitis in the nervous system. The most common clinical manifestation of this is ischaemic stroke, most commonly in the posterior circulation. Multiple aneurysms, venous sinus thrombosis and subarachnoid haemorrhage have also been reported.17 18 Pathological specimens of patients with CNS vasculitis from neuroborreliosis have shown fibroblastic thickening of the intima of leptomeningeal arteries, similar to the changes seen in meningovascular syphilis, and in some cases, spirochaetes have been cultured from affected areas of the brain8 15

Case 5

A middle-aged man from the New Forest (and a keen golfer) presented with a 3-week history of lower back pain radiating to the right thigh, associated with right thigh numbness. This had been preceded by a rash over his right thigh. On examination, there was weakness of right finger and elbow extension, right hip flexion and extension and right foot dorsiflexion. The right knee jerk was absent. There was an area of hypersensitivity around the trunk at the level of T10. Spinal cord imaging showed an area of signal change at T10–12 extending into the conus and with enhancement of the cauda equina. Nerve conduction studies showed changes consistent with a mononeuritis multiplex affecting the right radial nerve and right saphenous nerve. Lyme serology, sent at first presentation, was negative but repeat testing, 1 week later, showed reactivity to IgG. CSF studies showed 760 white cells (90% lymphocytes), protein 3.3 g/L and glucose 3.5 mmol/L (plasma 5.4 mmol/L), positive intrathecal oligoclonal bands, and was positive for Borrelia IgM. He was diagnosed with Lyme disease causing myelitis, radiculitis and a mononeuritis multiplex. He was treated with intravenous ceftriaxone, which resulted in an almost complete recovery.

Peripheral neuropathy

A facial neuropathy and polyradiculopathy, as found in Bannwarth’s syndrome, are the most common forms of peripheral neuropathy in Lyme. However, Borreliosis can also, less commonly, cause other mononeuropathies, including other cranial neuropathies. There may be a mononeuritis multiplex, and this may occur in the setting of other manifestations of neuroborreliosis and CSF changes, as in our case 5, although purely peripheral nervous system involvement with normal CSF can also occur.

Sural biopsies from patients with neuroborreliosis have shown perivascular inflammation and axonal damage, with Wallerian degeneration and consequent demyelination.19 Although spirochaetes are easily isolated from skin lesions, they have not been demonstrated in the affected peripheral nerves.

Later, a peripheral, often asymmetric, sensory neuropathy can occur in association with acrodermatitis chronica atrophicans (see under late neuroborreliosis).

Laboratory diagnosis

Routine haematology and biochemistry profiles are usually normal in neuroborreliosis, with normal or mildly elevated serum C reactive protein and erythrocyte sedimentation rate.4

Direct visualisation of the spirochaete by culture is rarely possible and not performed routinely. PCR is likewise not sensitive or specific enough to offer a reliable diagnostic test.20 It can be performed on samples of skin from the leading edge of erythema migrans, biopsies of acrodermatitis chronica atrophicans or synovium, joint fluid and CSF.

Routine diagnosis is based on serology. Responsible medical authorities, including NICE, recommend a two-tiered serological method of enzyme immunoassay plus confirmatory test, usually immunoblots. Confirmation of suspected infection should start with ELISA/enzyme immunoassay. If this is positive, the sample is forwarded for immunoblot at a reference laboratory.

The enzyme immunoassay is highly sensitive in established Lyme disease, and immunocompetent individuals with disseminated neuroborreliosis almost invariably have a strong IgG response. However, enzyme immunoassay is not highly specific,21 and several conditions can give false-positive results (box 1). Positive results are therefore forwarded to reference laboratories for confirmation.

Box 1

Causes of false-positive Borrelia enzyme immunoassay results

Causes of false-positive enzyme immunoassay:

  • Alternative infections: UK

    1. Syphilis

    2. Leptospirosis

    3. Helicobacter pylori

    4. Epstein-Barr virus

    5. Infective endocarditis

  • Alternative infections: Non UK

    1. Tick-borne relapsing fever

  • Autoimmune disorders

    1. Systemic lupus erythematosus

  • Intravenous immunoglobulin administration

This combination offers extremely high specificity and sensitivity. Negative serology does not rule out early infection within 6 weeks, as early in the disease there is lower sensitivity and false negatives can occur, especially in facial palsy where a significant minority are initially seronegative. However, symptoms of neuroborreliosis usually begin several weeks after inoculation, and therefore circulating antibodies should be present. If the enzyme immunoassay is negative but neuroborreliosis remains a concern, or if the serology results are conflicting, convalescent serology after a period of 1 month is helpful and will confirm or exclude neuroborreliosis. A negative IgG enzyme immunoassay a month after symptom onset is reliable (even if the IgM immunoblot is positive).

If symptoms and clinical suspicion of neuroborreliosis remain beyond 12 weeks, and initial serology was negative, new NICE guidance recommend sending both enzyme immunoassay and immunoblot samples for testing. Discussion of test results with a microbiologist may help in this situation.

Other diagnostic tests

Private laboratories, including via online sites, offer a variety of other diagnostic tests. These are not recommended for Lyme diagnosis as they are rarely appropriately clinically validated and generally have poor specificity. Patients may present with the results of such tests, citing these as evidence of Lyme disease. The Centers for Disease Control and Prevention list the following tests as not being fit for purpose for Lyme diagnosis22:

  • Culture, immunofluorescence staining or cell sorting of cell wall-deficient or cystic forms of B. burgdorferi.

  • Lymphocyte transformation tests (ELISPOT).

  • Quantitative CD57 lymphocyte assays.

  • ‘Reverse Western blots’.

  • Capture assays for antigens in urine.

  • In-house criteria for interpretation of immunoblots.

  • Measurements of antibodies in joint fluid (synovial fluid).

  • IgM or IgG tests without a previous ELISA/enzyme immunoassay/immunofluorescence assay.

Previously exposed/infected individuals

In previously infected individuals, it is expected to find persistent IgG response (on both enzyme immunoassay and immunoblot). It is also common to find a persistent IgM response. This does not imply active infection in the absence of clinical signs and symptoms. Specific antibodies against Bb similarly persist in CSF and may be present for years. These are not suitable biomarkers of reinfection or antibiotic effect

Diagnosis of suspected reinfection or incomplete treatment depends on evidence of active CNS inflammation and serology results. High background seropositivity and testing in patients with a low pretest probability will reduce the diagnostic value and lead to overdiagnosis.

CSF examination

Not all patients need a lumbar puncture because the diagnosis can be made clinically and with serology. However, CSF should be tested if there is any diagnostic doubt and may help in the acute setting while waiting for serology. The European Fedaration of Neurological Sciences (EFNS) case definition requires demonstration of a CSF pleocytosis (Box 2).

Box 2.

European Federation of Neurological Societies case definition for neuroborreliosis.24

European Federation of Neurological Societies case definition:


1. Neurological symptoms suggestive of neuroborreliosis AND

2. CSF pleocytosis AND

3. Bb-specific antibodies produced intrathecally

PROBABLE: 2/3 above criteria, with convalescent serology confirming antibody production (if criteria 3 is not initially met).

CSF usually shows a leucocytosis (10–1000/mm3), predominantly or exclusively lymphocytic, but may be normal, especially early. CSF protein concentrations are usually higher than seen in other acute infections without the reduction in glucose concentration seen in tuberculous meningitis. There are often unmatched CSF oligoclonal IgG bands.

CSF should be tested for intrathecal Borrelia antibody production with a ratio of CSF to serum antibodies (IgG or IgM) reflecting intrathecal synthesis.23 These antibodies appear at approximately 2 weeks and in one neuroborreliosis case series were universally present by 2 months.12 A positive CSF IgG response with active inflammation has a diagnostic predictive value above 95%.13 A positive CSF response without positive blood serology can occur, but is rare.


Retrospective case series show that most cases of neuroborreliosis resolves without treatment within 3 to 6 months.24 Antibiotics help to accelerate bacterial clearance and prevent progression of the infection. This should mitigate the stroke risk associated with acute neuroborreliosis, reduce damage from inflammation and limit axonal degeneration. Most importantly, antibiotics rapidly improve a patient’s symptoms, particularly their pain. Secondary axonal damage may have occurred, and will recover slowly, and sometimes incompletely. If there is a strong clinical suspicion, therefore, clinicians should consider treatment with appropriate antibiotics immediately, while serology is awaited.

Penicillin G, doxycycline, ceftriaxone and cefotaxime are all effective in early neuroborreliosis.25 Neuroborreliosis with CNS involvement is most often treated with intravenous ceftriaxone, but several studies have shown that oral doxycycline is as safe and effective as ceftriaxone for all manifestations of neuroborreliosis.26–29 Doxycycline has the advantage of oral preparation, but is contraindicated in pregnancy and lactation.

Current European Federation of Neurological Societies and NICE recommendations are to use either doxycycline 200 mg daily or intravenous ceftriaxone 2 g daily (or intravenous penicillin or intravenous cefotaxime) for neuroborreliosis with peripheral nervous system or meningeal-limited disease (table 1). In cases with CNS parenchymal manifestations, such as myelitis, encephalitis or vasculitis, current guidance is to treat with intravenous ceftriaxone, 2 g daily for 21 days. CSF pleocytosis, in the presence of peripheral nervous system disease, does not require treatment for CNS infection. There is no evidence of Borrelia resistance to first-line treatment2 23

Table 1

Recommended treatment for neuroborreliosis2

There is no current recommendation for the use of corticosteroids in neuroborreliosis. About half of the CNS vasculitis cases in one series received corticosteroids in addition to antibiotics, with a few cases receiving other forms of immunosuppression. These reviews showed no consistently significant advantage from using corticosteroids.17 30 It is currently unclear whether there are groups of patients who would benefit from the addition of corticosteroid to antibiotic therapy.

Most patients fully recover with treatment, although this can take several months. Sometimes there is permanent damage, for example to cranial or peripheral nerves.

It is helpful to warn patients that there may be an exacerbation in their symptoms after starting antibiotic therapy due to a Jarisch-Herxheimer reaction. This usually starts within 12 hours of antibiotic treatment and is due to rapid destruction of multiple bacteria. It does not usually require treatment to be stopped, and the advice for patients would be to seek medical attention if unwell, but to continue to take their antibiotics.2

Late and chronic Lyme disease

Late neuroborreliosis is defined as clinical symptoms lasting over 6 months with evidence of active infection. This is rare and comprises less than 2% of cases of neuroborreliosis. It has been described predominantly in mainland Europe, but there have been few cases reported in the UK, although none in the Devon or Hampshire series.4 6

European clinical and pathological reviews suggest that the pathological process is an induced CNS vasculitis, with persistent inflammation of vessels induced by chronic infection.12 With few cases and overlapping cohorts, it remains difficult to define clearly and differentiate early CNS neuroborreliosis and late neuroborreliosis. As this appears to be the result of the same pathological process as that seen in CNS parenchymal disease, there is a question as to why we distinguish the two.

Unless treated, this disease may not be self-limiting, and antibiotic treatment is necessary for recovery.

We do not know why in some people the spirochaetes are cleared, whereas in other immunocompetent people they persist and induce a profound inflammatory response in surrounding vessels. This variability may result from host immune factors or variation in Borrelia subspecies; we need further research to look at individual host responses.

Late disease can also cause a chronic symmetrical generalised axonal sensory peripheral neuropathy. This almost invariably exists with the skin lesion acrodermatitis chronica atrophicans. This lesion is most frequently reported in Europe and in association with infection by B. afezelii. In cases of acrodermatitis chronica atrophicans–associated neuropathy, serology is usually strongly positive, but CSF is normal, suggesting the inflammation is limited to the peripheral nervous system. There were no cases of chronic peripheral neuropathy or acrodermatitis chronica atrophicans in two UK case series, suggesting that this is rare in UK-acquired Borreliosis.4 6

‘Chronic Lyme disease’ is different and is less consistently defined. People within this cohort broadly fit into three categories:

  1. Persistent symptoms following laboratory confirmed and treated neuroborreliosis.

  2. Atypical symptoms that occur in the setting of positive serology, which may be due to asymptomatic past infection

  3. Symptoms attributed to Borreliosis but with persistently negative, equivocal or inconsistent serology results.

Typically, these patients suffer chronic, often relapsing, non-focal or non-specific neurological symptoms, such as fatigue, cognitive dysfunction and paraesthesia. There have been no objective clinical or laboratory marker of active infection or immune activation demonstrated12 and studies have shown that these symptoms are no more common in patients with positive serology or a history of tick bites than in the general population in endemic areas.31 32 This complex cohort probably includes patients with alternative diagnoses including chronic fatigue and functional neurological syndromes. We need specific definitions of the subgroups within this cohort, as well as further research looking at biomarkers of disease in each patient group, in order to target appropriate therapy.

Once active infection is ruled out or treated according to guidelines, and other diagnoses are excluded, the management of these patients should be supportive. Both seropositive and seronegative patients with persisting symptoms such as those described above have been repeatedly shown not to respond to prolonged courses of antibiotics, when compared with placebo.33 NICE guidelines therefore state that these patients should not be offered multiple or prolonged courses of antibiotics without compelling evidence of active infection. They should be offered symptomatic management and appropriate support and referral to alternative services where appropriate.2

Differential diagnosis

Early neuroborreliosis should be considered in cases of Bell’s palsy, especially in tick-exposed individuals, or where the Bell’s palsy is bilateral or associated with mild meningitic symptoms. It is often mistaken for varicella zoster infection, which also causes a painful myeloradiculopathy, rash and lymphocytic meningitis. It should be considered when finding an unexpected lymphocytic CSF, for example when investigating cases of suspected Guillain-Barré syndrome, or myelitis. It is part of the differential diagnosis for CNS vasculitis, alongside other infections including HIV, tuberculosis and syphilis.

Late and chronic Lyme has been reported as mimicking several other chronic neurological conditions, including multiple sclerosis and amyotrophic lateral sclerosis. Some of these reports are the result of coincidental positive serology from a previous asymptomatic infection, and some arise because of a few comparable features (such as intrathecal oligoclonal bands). However, a neurologist can usually distinguish between the diagnoses on clinical grounds. Neuroborreliosis is a monophasic illness that can be differentiated from the typical relapsing–remitting course and common CNS features of multiple sclerosis. Likewise, a neuroborreliosis-associated polyradiculopathy might cause multifocal denervation, but it is painful and lacks other features (bulbar involvement, brisk reflexes) of amyotrophic lateral sclerosis. Some cases remain difficult, however, for example if the patient has presented early in their disease course, and it will only become more obvious after further investigation or follow-up. If there is any doubt and neuroborreliosis could be part of the differential diagnosis clinically, we recommend serology testing early. An incidental positive result can then be investigated, for example by CSF testing, and discussed and dismissed early, where appropriate.

Recurrent symptoms

In our experience, symptom recurrence or chronic infection is extremely unlikely with antibiotic therapy equivalent to the new NICE treatment guidelines. Reinfection from a new tick bite is possible, so if symptoms recur, investigations (which may include CSF examination) should be repeated and a second course of treatment considered. We have not encountered any cases where objective neurological manifestations of active neuroborreliosis have progressed despite appropriate antibiotic treatment.


Neuroborreliosis is a result of infection with Borrelia burgdorferi, and in the UK this is usually of the strain B. garinii. The vast majority of cases present with all or part of Bannwarth’s syndrome—cranial neuropathy (usually facial) and/or painful meningoradiculitis. Atypical forms of neuroborreliosis include myelitis, CNS vasculitis and mononeuritis multiplex.

There is a spectrum of disease, broadly labelled ‘chronic Lyme disease’ that represents a complex cohort of patients. There is no evidence that prolonged courses of antibiotics are effective if there is no evidence for active infection. Treatment should be patient centred and focused on symptom control.

Serum confirmation of infection requires a two-stage process, the second stage at a reference laboratory. Serology may be negative early in infection but will usually be positive by the onset of neurological symptoms. Late manifestations are almost invariably associated with positive serology.

CSF is abnormal, with an abundance of lymphocytes, a high protein concentration and positive oligoclonal bands, but usually a normal glucose concentration.

Treatment should be given for 3 weeks. There is good evidence that doxycycline, which can be taken orally, is as effective as intravenous ceftriaxone in treating early peripheral neuroborreliosis, but CNS parenchymal disease should ideally be treated with intravenous ceftriaxone initially, with consideration of subsequent step down to doxycycline when the patient is well.

Key points

  • A significant proportion of people living in endemic areas will have positive serology without symptoms suggesting prior, asymptomatic infection.

  • The vast majority of cases of neuroborreliosis will present with cranial neuropathy, meningitis or radiculitis. Pain is often the first and worst symptom and is classically worse at night.

  • Lyme disease is an uncommon cause of carditis, uveitis, keratitis and inflammatory arthritis; these clinical manifestations are much less common in UK and European-acquired Lyme.

  • Diagnosis is via two-tier serology testing, which must be confirmed at a reference laboratory. Alternative testing methods are not considered sufficiently sensitive or specific and are therefore not reliable.

  • Neuroborreliosis can cause an induced vasculitis in the central nervous system, similar to changes seen in other organs; this is rare and probably the explanation for chronic presentations with myelitis, recurrent ischaemic events and pachymeningitis.

  • Treatment guidelines are available from the National Institute for Health and Care Excellence and recommend 3 weeks of appropriate antibiotic therapy. Patients with chronic symptoms should be evaluated and investigated, but there is no evidence for repeated courses of antibiotics without ongoing evidence of infection or inflammation, and no evidence of bacterial resistance to first-line antibiotics.


We thank Dr Haider Katifi for reviewing the manuscript and Dr Matthew Harris for contributing a case.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.


  • Correction notice This article has been corrected since it was

    published Online First. The references and citations were not correctly matched.

  • Contributors ALRR prepared the manuscript with contributions from JL and MD, and with review and revisions by all authors.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Provenance and peer review Commissioned; externally peer reviewed by John Halperin, New Jersey, USA.

Other content recommended for you