Longer treatment: better outcome

Management of suspected Lyme borreliosis: experience from an outpatient antibiotic therapy service.’ White B, Seaton RA, Evans TJ. QJM 2013; 106:133–38.

Lyme disease in a Glasgow ID Clinic

This study looks at the ‘real world experience’ of an infectious diseases clinic in Glasgow using Outpatient Parenteral Antibiotic Treatment (OPAT), in the form of intravenous ceftriaxone, for treatment of suspected or proven Lyme disease. The conclusion given in the abstract is that “OPAT was an effective way of administering IV treatment for LB but should not be undertaken lightly due to the high rate of adverse events and low rates of success in certain patients groups seen in this study.”

However, there are other elements of the study that LDA thought worth drawing to clinicians’ attention. We present an outline of the study and LDA’s response below.

The paper is a retrospective review of all patients with suspected Lyme disease treated with outpatient intravenous ceftriaxone in Glasgow from January 2000 to June2011. It looks at outcomes, appropriateness of current guidelines when applied to the ‘real world’ of clinical practice and records any adverse effects of treatment. Most patients (91.7%) self-administered the treatment through a peripheral or central venous cannula following training.


76 patients were identified and records were available for 72, of which 79% were treated from 2007.

The clinic deals with referrals for 7 of the 14 Scottish Health Boards in a ‘highly endemic region.’

  • 68% of the referrals were from Greater Glasgow & Clyde Health Board.
  • Median age: 42 years (range 15 – 89 years). 56% were female.
  • 39% came from General practice and 38% from Neurology.
  • The majority of patients had positive/equivocal serology; 18% of patients were seronegative.
  • 76% had ‘potential tick exposure or risk factors for LB’ and 39% had recollection of one or more tick bites. 19% had ‘no recollection of a tick bite and no identifiable risk factor for LB.’

Suspected diagnoses at referral:

Neuroborreliosis 60%, Chronic Lyme borreliosis 17%, Lyme Arthritis 15%, Early Lyme borreliosis 4%, Cardiac Lyme borreliosis 3%, Uveitis 1%.

Common Symptoms included:

Nerve root pain and peripheral sensory problems 47%, Fatigue 44%, Joint pain 26%, Headache 15%, Muscle weakness 13%, Seventh Nerve/Facial palsy 10%, Cognitive problems 10%, Muscle pain 8%, Hearing Loss/ Tinnitus 7%, Inflammatory arthritis 7%.

In general: ‘A wide range of symptoms were reported and multiple symptoms common.’

Erythema migrans rash was present in only 4% of patients, with unspecified rash in 4%.

49% of patients were treated according to guidelines and 50% outside guidelines. Patients treated outside guidelines included those with negative serology, those who received longer or repeated courses of intra-venous treatment and 32% who received oral antibiotics following on from intravenous: median 3 weeks (range 1-8 weeks).

The median duration of ceftriaxone was 21 days (range 1-43 days).

Median follow-up was 4 weeks after stopping treatment (range 0-204 weeks).

Definite improvement was seen in 28% patients; modest/slight or transient improvement in 33% and no improvement in 36%. The authors combine the latter groups in their analysis.

The authors commented that ‘Patients’ treatment response according to serological status and adherence to published guidelines showed conflicting results.’

61% of those with positive serology who received prolonged therapy showed definite improvement as well as 23% of those with negative serology.

Poor outcome was seen in those with positive serology where guidelines recommended oral treatment (9% definite improvement), and those who received intravenous therapy as per guidelines (23% definite improvement).

Adverse reactions included a surprisingly high rate (18%) of neutropenia* (reduced number of neutrophils which are a type of white blood cell), abnormal liver function tests (11%), diarrhoea (6%) of which only one case was due to C. difficile, line infection (4%) with one patient stopping treatment as a result of severe sepsis. Less common were rash (4%), allergic reaction** (3%), headache, fatigue, nausea, oral thrush, shingles and dyspepsia (all 1%).

* One patient stopped treatment due to agranulocytosis (severe neutropenia) which then resolved. **One patient had anaphylaxis after the first dose of ceftriaxone.

The authors comment that their most striking finding was 18% rate of neutropenia, compared to less than 1% in other clinical trials of this drug, with agranulocytosis in less than 0.1%. Limitations of the study are that it is retrospective; it follows a small number of patients with a wide range of symptoms and has no pre-determined robust measure of success or failure of treatment.

The authors conclude that treatment ‘is likely to remain an individual physician patient decision based on a frank pre-treatment discussion of risk and benefit’.


In LDA’s letter of response we emphasised the need for more research as indicated by the James Lind Alliance process, which at the time was nearing its conclusion. We emphasised the high number of seropositive individuals who showed definite improvement with prolonged treatment (61%) as opposed to those treated within guidelines (20%). We argued that this was only ‘inconsistent’ because it appeared to cast doubt on guidelines when applied to the real world. We emphasised that this is the second UK paper to show improvement following treatment of seronegative individuals and call for improvements in study design.

We question whether the high rate of neutropenia could have been related to the group of patients, specifically querying whether other tick-borne infections such as Anaplasma, a known cause of neutropenia had been considered.

  • The authors have not replied to our letter – we wonder if they agree with us?
  • QJM is a high impact international journal.


Stricker, De Long and Johnson also responded, using a statistical method to show that in their opinion “the study demonstrated a significant benefit of prolonged vs. short-course antibiotic therapy in carefully chosen patients with ‘debilitating symptoms’ of Lyme Borreliosis.”

The authors, White et al, replied stating that application of Stricker et al’s statistical method was not valid and “subgroup analysis carried out in this fashion is fraught with errors in interpretation.” They call for the clinicians “to develop a larger, well-designed randomized control trial with robust outcome measures to provide clarity on this issue.”


A summary of all this?

  • The evidence on the ground indicates that guidelines for Lyme disease are not being adhered to because they are inadequate
  • They agree with this in the USA and in the Netherlands (the latter joined us for the JLA workshop)
  • LDA’s JLA project has confirmed that there are uncertainties in both diagnosis and treatment.

What are we waiting for?