Abstract

Objectives: Lyme borreliosis (LB) is the most common human tick-borne infection in Europe and the USA. In this study we set out to analyse the outcome of patients treated for Lyme disease via outpatient parenteral antibiotic therapy (OPAT) and the appropriateness of this treatment using current guidelines.

Methods: This was a retrospective review of all patients with suspected LB managed via OPAT in Glasgow in 2000–11.

Results: Of 72 patients treated for suspected LB, 35 patients (49%) were treated in accordance with guidelines and 36 (50%) were treated with no specific guidelines. A definite improvement was seen in 20 patients (28%). Adverse reactions were documented in 29 (40%) patients with neutropenia, and mild liver function derangement was most commonly observed.

Conclusions: These results show the complexity of translating well-substantiated regimens from clinical trials to actual clinical practice. OPAT was an effective way of administering parenteral therapy for Lyme disease but should not be undertaken lightly due to the rate of adverse events and low rates of success in certain patient groups seen in this study. In view of this, stricter criteria for inclusion to OPAT in line with published guidance should be applied to minimize patient harm and optimize success.

Introduction

Lyme borreliosis (LB) is the most common human tick-borne infection in both Europe and the USA.1,2 Antibiotic treatment is effective therapy, and for many of the manifestations of Lyme disease such as erythema chronicum migrans or Lyme arthritis, oral regimens such as doxycycline are highly effective. Indeed there is recent evidence from northern Europe that oral antibiotic therapy is as effective as parenteral treatment for neurological manifestations of LB if they are confined to the peripheral nervous system and/or meninges.3 However, for the central nervous system complications of Lyme such as encephalitis, myelitis and vasculitis, parenteral therapy remains the recommended option.

Within existing North American and European guidance, there is a slight variation in indications for parenteral therapy in early and late Lyme, possibly reflecting differing Borrelia burgdorferi sensu lato species predominating in these different countries.1 In contrast, there is consensus and indeed excellent evidence that antibiotics are not effective in patients with a variety of persisting symptoms following Lyme disease, often classified as post-Lyme syndrome.4 Although serological diagnosis does not have 100% sensitivity, particularly in early disease, virtually all patients with late stage disease have positive serology. There have been no substantiated reports of late stage neuroborreliosis that remain seronegative, and only single case reports of apparent seronegative late Lyme disease with other manifestations.5,6

Notwithstanding this accumulated evidence, many charitable and other organizations challenge this conventional view of the diagnosis and management of Lyme disease, often advocating prolonged courses of parenteral antibiotics even in the absence of specific serological evidence of Lyme infection.7

In Scotland, reports of LB continue to increase,8 and as a consequence, increasing referrals to the Glasgow infectious diseases unit have been observed. For those requiring intravenous (IV) therapy, the preferred parenteral agent is ceftriaxone for a duration of 14–21 days.1,3,9 Such a regimen requires either prolonged hospital admission or selected patients may be facilitated by outpatient parenteral antibiotic therapy (OPAT) programmes. To date, there are no studies that have described the outcome of this means of giving antibiotics for Lyme disease, its relation to initial clinical presentation, adverse reactions and clinical improvement. Therefore, we analyse retrospectively the outcome of patients treated with parenteral antibiotics for Lyme disease in an outpatient setting. We also reviewed the appropriateness of this treatment using current guidelines.

Materials and methods

A retrospective review was performed to identify all patients with suspected LB managed via OPAT between January 2000 and June 2011. In brief, OPAT management of suspected LB consisted of a pre-treatment consultation by an infectious diseases physician to determine applicability of therapy, to counsel on likely success of therapy and potential adverse consequences of treatment. Typically patients would be trained to self-administer IV ceftriaxone through a peripherally inserted mid-line or central venous catheter. As a minimum, patients would be reviewed by a clinical nurse specialist after the first and second week of IV therapy and on completion. Success or failure would be assigned following review by the infectious diseases physician at clinic after cessation of therapy.

Patients were identified using a prospectively maintained OPAT database.10 Clinical, demographic, diagnostic and management data were collected by review of patient records. Serological testing was performed on all patients including an ELISA and western blot for LB. Diagnostic and treatment characteristics were compared with published guidance at the time of patient treatment. The guidelines used were the Infectious Diseases Society of America (IDSA) guidelines published in 2000, which was then updated in 2006, the European Federation of Neurological Societies guidelines published in 2010 and the British Infection Association guidelines published in 2011.1,3,9,11 Outcome of therapy was retrospectively assigned to either definite improvement (where both patient and attending physician identified objective improvement in symptoms or signs) or other (which included modest/mild/transient or no or unknown improvement). Physician-anticipated outcome before therapy was recorded following review of the pre-treatment consultation and was compared with the actual outcome after therapy.

Results

Seventy-six patients were identified as having being treated for suspected LB via OPAT between January 2000 and June 2011 (78 treatment episodes). Clinical records were available for 72 patients, and 57 (79%) patients were treated from 2007 onwards. Median age was 42 years (range 15–89 years), and 58% were female. Referrals were received from 7 of 14 health boards within Scotland with the majority (68%) arising from Greater Glasgow and Clyde. Referrals were most frequently received from general practice (39%) and neurology (38%). Sixty percent of referrals were for suspected neuroborreliosis (Table 1). The majority of patients had positive or equivocal Lyme serology; however, 13 patients (18%) had negative serology.

Table 1

Clinical and diagnostic characteristics

 Patients, N (%) 

 
Suspected diagnosis at referral 
    Neuroborreliosis 43 (60) 
    Chronic LB 12 (17) 
    Lyme arthritis 11 (15) 
    Early LB 3 (4) 
    Cardiac LB 2 (3) 
    Uveitis secondary to LB 1 (1) 
General systemic symptoms 
    Fatigue 32 (44) 
    Flu-like symptoms 4 (6) 
    Myalgia 6 (8) 
    Night sweats 1 (1) 
    Galactorrhoea 1 (1) 
Neurological 
    Radicular pain or parasthesia 34 (47) 
    Headache 11 (15) 
    Muscle weakness 9 (13) 
    Seventh nerve palsy 7 (10) 
    Impaired mentation 7 (10) 
    Hearing loss or tinnitus 5 (7) 
    Seizures 1 (1) 
    Guillain–Barre syndrome 1 (1) 
    Atypical motor neuron disease 1 (1) 
    Mononeuritis multiplex 1 (1) 
    Ataxia 1 (1) 
Rheumatological 
    Arthralgia without evidence of synovitis 19 (26) 
    Inflammatory arthritis 5 (7) 
Cardiological 
    Complete heart block 1 (1) 
    Pericarditis 1 (1) 
    Heart failure 1 (1) 
Dermatological 
    Rash, unspecified 3 (4) 
    Erythema chronicum migrans rash 3 (4) 
    Cellulitis 2 (3) 
    Possible acrodermatitis chronica atrophicans 1 (1) 
Ophthalmological 
    Uveitis 1 (1) 
    Diploplia 1 (1) 
    Blurred vision and visual oscillations 1 (1) 
Serology 
    Positive/equivocal 57 (79) 
    Negative 13 (18) 
    Unknown 2 (3) 
Infectious diseases physician’s anticipated outcome of treatment 
    Possible or probable benefit likely 36 (50) 
    No benefit likely 36 (50) 
 Patients, N (%) 

 
Suspected diagnosis at referral 
    Neuroborreliosis 43 (60) 
    Chronic LB 12 (17) 
    Lyme arthritis 11 (15) 
    Early LB 3 (4) 
    Cardiac LB 2 (3) 
    Uveitis secondary to LB 1 (1) 
General systemic symptoms 
    Fatigue 32 (44) 
    Flu-like symptoms 4 (6) 
    Myalgia 6 (8) 
    Night sweats 1 (1) 
    Galactorrhoea 1 (1) 
Neurological 
    Radicular pain or parasthesia 34 (47) 
    Headache 11 (15) 
    Muscle weakness 9 (13) 
    Seventh nerve palsy 7 (10) 
    Impaired mentation 7 (10) 
    Hearing loss or tinnitus 5 (7) 
    Seizures 1 (1) 
    Guillain–Barre syndrome 1 (1) 
    Atypical motor neuron disease 1 (1) 
    Mononeuritis multiplex 1 (1) 
    Ataxia 1 (1) 
Rheumatological 
    Arthralgia without evidence of synovitis 19 (26) 
    Inflammatory arthritis 5 (7) 
Cardiological 
    Complete heart block 1 (1) 
    Pericarditis 1 (1) 
    Heart failure 1 (1) 
Dermatological 
    Rash, unspecified 3 (4) 
    Erythema chronicum migrans rash 3 (4) 
    Cellulitis 2 (3) 
    Possible acrodermatitis chronica atrophicans 1 (1) 
Ophthalmological 
    Uveitis 1 (1) 
    Diploplia 1 (1) 
    Blurred vision and visual oscillations 1 (1) 
Serology 
    Positive/equivocal 57 (79) 
    Negative 13 (18) 
    Unknown 2 (3) 
Infectious diseases physician’s anticipated outcome of treatment 
    Possible or probable benefit likely 36 (50) 
    No benefit likely 36 (50) 

Fifty-five patients (76%) had potential tick exposure and risk factors for LB, and 28 (39%) had a recollection of one or more tick bites. Fourteen (19%) patients had no recollection of a tick bite and no identifiable risk factor for LB. There was a wide range of symptoms reported (Table 1), and multiple symptoms were common. Most frequent symptoms were radicular pain or paraesthesia in 34 patients (47%), fatigue in 32 (44%) and arthralgia without evidence of synovitis in 19 (26%).

Thirty-five patients (49%) were treated in accordance with guidelines,1,3,9,11 and 36 (50%) were treated with no specific guidelines. An additional patient was given parenteral ceftriaxone for uveitis associated with positive Lyme serology, and there are no guidelines for the management of this rare manifestation of LB. Treatment regimens given that were non-compliant with the guidelines included: parenteral therapy for a clinical syndrome where parenteral therapy was not indicated (N = 16), prolonged duration of parenteral therapy (N = 6), oral therapy following parenteral therapy (N = 23) and repeated courses of parenteral therapy (N = 2).

Median duration of ceftriaxone was 21 days (interquartile range: 21–23 days; range: 1–43 days). The majority (95.8%) received 2 g daily. Twenty-three patients also received a subsequent oral antibiotic for a median of 3 weeks (range 1–8 weeks). A total of 91.7% of patients administered therapy themselves via a short-term vascular device after training.

Median follow-up was 4 weeks (range: 0–204 weeks) after cessation of ceftriaxone. A definite improvement was seen in 20 patients (28%), with either no improvement or only a modest, slight or transient improvement in 50 patients (69%) (Table 2). No outcome data were available in one patient, and one patient only received a single dose of ceftriaxone and developed anaphylaxis, subsequently receiving an oral regimen instead. Patients’ treatment response according to serological status and adherence to published guidance showed conflicting results. Poor outcome was observed in those with negative serology (23% improvement), those with positive serology in whom IV therapy was not recommended (9% improvement) and those who received IV therapy as per guidance (25% improvement). In contrast, improvement was observed in 63% of those with positive serology who received prolonged therapy beyond that recommended in published guidance (Figure 1).

Figure 1

Treatment response according to patients’ serology and guideline treatment recommendation. 1One patient presented with uveitis—no guidelines for this rare manifestation of LB. 2Two patients have no outcome data.

Figure 1

Treatment response according to patients’ serology and guideline treatment recommendation. 1One patient presented with uveitis—no guidelines for this rare manifestation of LB. 2Two patients have no outcome data.

Table 2

Outcome and adverse events

 Patients
 

 
 Na 
Symptom improvementb 
    Definite 20 28 
    Modest/slight/transient 24 33 
    Nil 26 36 
Correlation between anticipated and actual symptomatic outcomeb 
    Definite improvement and  probable/possible benefit anticipated 18 53 
    Modest/slight/transient or no  improvement and no anticipated  benefit 34 94 
    Definite improvement but no  anticipated benefit – 
    Modest/slight/transient or no  improvement but possible benefit  anticipated 16 – 
Adverse reactions 
    Neutropeniac 13 (4) 18 
    Line infection 3 (1) 
    Rash 3 (3) 
    Allergic reaction 2 (2) 
    Diarrhoead 
    Liver function deranged 11 
    Headache 
    Nausea 
    Fatigue 
    Oral thrush 
    Shingles 
    Dyspepsia 
 Patients
 

 
 Na 
Symptom improvementb 
    Definite 20 28 
    Modest/slight/transient 24 33 
    Nil 26 36 
Correlation between anticipated and actual symptomatic outcomeb 
    Definite improvement and  probable/possible benefit anticipated 18 53 
    Modest/slight/transient or no  improvement and no anticipated  benefit 34 94 
    Definite improvement but no  anticipated benefit – 
    Modest/slight/transient or no  improvement but possible benefit  anticipated 16 – 
Adverse reactions 
    Neutropeniac 13 (4) 18 
    Line infection 3 (1) 
    Rash 3 (3) 
    Allergic reaction 2 (2) 
    Diarrhoead 
    Liver function deranged 11 
    Headache 
    Nausea 
    Fatigue 
    Oral thrush 
    Shingles 
    Dyspepsia 

aNumber in brackets denotes number of patients who stopped parenteral ceftriaxone due to the adverse event noted.

bTwo patients have no outcome data: one failed to attend follow-up and one only received one dose of ceftriaxone due to anaphylaxis.

cOne patient developed agranulocytosis that resolved after cessation of treatment.

dOne Clostridium difficile-associated diarrhoea.

Thirty-six patients (50%) were assessed pre-treatment to have a possible or probable improvement with therapy and 36 were felt unlikely to benefit. The correlation between physician-anticipated outcome for the infectious diseases and the patients’ subjective response to treatment had a positive predictive value of 53%, with 18 of 34 patients felt likely to have a symptomatic improvement doing (no outcome data in two patients). However, 34 of the 36 patients not deemed likely to improve had either no improvement or only a modest, slight or transient improvement in symptoms at follow-up, giving the physician’s anticipated outcome a negative predictive value of 94%.

Adverse reactions were documented in 29 (40%) with a total of 39 drug reactions in these patients. Neutropenia and mild liver function derangement were most commonly observed (Table 2). Treatment was discontinued in 10 (14%) due to adverse events; significant neutropenia (<1 × 10*9/l) in 4 (with agranulocytosis in 1 of these), significant drug rash in 3 and anaphylaxis in 1 after the first dose of ceftriaxone. One patient had a line infection necessitating hospital admission with severe sepsis.

Discussion

This study demonstrates the ‘real-world experience’ of an infectious diseases tertiary referral centre in an area of high LB prevalence. It presents for the first time the outcome of patients with proven or suspected Lyme disease treated by OPAT. Previous studies have documented the efficacy of various antibiotic regimens, including parenteral therapy, in the treatment of various Lyme disease syndromes. The results of this study show the complexity of translating these well-substantiated regimens from clinical trials to actual clinical practice.

Individual patient management decisions—including management of patient expectations—can be difficult. Patients may have strong preconceived ideas about their putative diagnosis, the benefits of serological tests and what the ‘best’ treatment(s) are before consultation.12 Equally infectious diseases physicians may be inclined to give a trial of treatment in patients with otherwise undifferentiated or incurable neurological symptoms/signs in the expectation that little harm can be expected and that any potential benefit of therapy outweighs any small perceived risk of harm or lack of benefit.

Broadly, OPAT was an effective way of administering parenteral therapy for Lyme disease, but our results offer some useful insights into the benefits and risks of treatment for this group of patients and helps inform future practice.

One of the most striking results of this audit was the high rate of adverse events noted, occurring in 40% of patients, with events leading to cessation of parenteral treatment in 14%. In previous OPAT trials, ∼25% of patients have been found to experience adverse reactions.13 Of particular note were the 13 patients (18%) who developed neutropenia, with significant neutropenia in 4. In clinical trials for ceftriaxone, neutropenia was seen in <1% of patients, with agranulocytosis in <0.1%.14 The reason for the high proportion of patients developing neutropenia in this cohort is unclear. However, these results underscore the importance of discussing with patients the potential for significant side effects before embarking on treatment.

Concordance between the treatment received and that which was advised in the guidelines was just under 50%. This may have been partly due to a lack of European guidelines published in English until 2010 and concern that the IDSA guidelines were not fully applicable in Europe due to the different Borrelia species dominant in the UK compared with the USA, both in terms of the possibility of false-negative serological tests and that treatment requirements might be different. However, it is also possible that although some of these patients could have been on effective oral regimens, patient pressure may have contributed to the decision to try parenteral therapy.

The limitations of this study were as follows: it was retrospective in nature, the duration of follow-up was wide ranging and the number of patients in each treatment group was small limiting the conclusions that can be drawn regarding the efficacy of OPAT in these patients. Furthermore, without a pre-determined definition of success/failure, it is difficult to interpret outcomes in this cohort. Simply using the reported outcomes from clinic consultations may overestimate the success of therapy. We have attempted to address this by retrospectively assigned mild, moderate and transient improvement as failure for the purposes of this analysis. It is clear there was a range of responses with a low overall reporting of clear cut clinical improvement. Low reported success was evident except in those with serological evidence of infection and a guidance approved indication for IV therapy who received prolonged therapy. A placebo effect in this group in whom extra medical care was dedicated cannot be excluded. It is also notable that if the infectious diseases physician’s pre-treatment consultation prediction was that parenteral therapy was unlikely to be effective, this had a negative predictive value of 94%. As this prediction was discussed openly with patients before therapy, it is impossible to know whether the suggestion itself had any impact on perceived outcome by patients.

In conclusion, the dilemma of ‘withholding’ potential therapy from patients with debilitating symptoms who have often been tried and failed on other therapies is impossible to reconcile within evidence-based treatment guidelines and is likely to remain an individual physician patient decision based on a frank pre-treatment discussion of risk and benefit. OPAT remains a useful mode of therapy for this ambulant patient group but should not be undertaken lightly due to the rate of adverse events and low rates of success in certain patient groups seen in this study. In view of this, stricter criteria for inclusion to OPAT in line with published guidance should be applied to minimize patient harm and optimize success.

Acknowledgements

The authors are grateful to Lindsay Semple for assistance with data collection.

Conflict of interest: None declared.

References

1
British Infection Association
The epidemiology, prevention, investigation and treatment of Lyme borreliosis in United Kingdom patients: a position statement by the British Infection Association
J Inf
 , 
2011
, vol. 
62
 (pg. 
329
-
38
)
2
Duncan
CJA
Carle
G
Seaton
RA
Tick bite and early Lyme borreliosis
BMJ
 , 
2012
, vol. 
344
 pg. 
e3124
 
3
Mygland
A
Ljostad
U
Fingerle
V
Rupprecht
T
Schmutzhard
E
Steiner
I
EFNS guidelines of the diagnosis and management of European Lyme neuroborreliosis
Eur J Neurol
 , 
2010
, vol. 
17
 (pg. 
8
-
16
)
4
Klempner
MS
Hu
LT
Evans
J
Schmid
CH
Johnson
GM
Trevino
RP
, et al.  . 
Two controlled trials of antibiotic treatment in patients with persistent symptoms and a history of Lyme disease
N Engl J Med
 , 
2001
, vol. 
345
 (pg. 
85
-
92
)
5
Pikelj
F
Strle
F
Mozina
M
Seronegative Lyme disease and transitory atrioventricular block
Ann Intern Med
 , 
1989
, vol. 
111
 pg. 
90
 
6
Dejmkova
H
Hulinska
D
Tegzova
D
Pavelka
K
Gatterova
J
Vavrik
P
Seronegative Lyme arthritis caused by Borrelia garinii
Clin Rheumatol
 , 
2002
, vol. 
21
 (pg. 
330
-
4
)
7
Cottle
LE
Mekonnen
E
Beadsworth
MBJ
Miller
ARO
Beeching
NJ
Lyme disease in a British referral clinic
Q J Med
 , 
2012
, vol. 
105
 (pg. 
537
-
43
)
8
Slack
CS
Mavin
S
Yirrel
D
Ho-Yen
DO
Is Tayside becoming a Scottish hotspot for Lyme borreliosis?
J R Coll Physicians Edinb
 , 
2011
, vol. 
41
 (pg. 
5
-
8
)
9
Wormser
GP
Dattwyler
RJ
Shapiro
ED
Halperin
JJ
Steere
AC
Klempner
MS
, et al.  . 
The clinical assessment, treatment, and prevention of Lyme diseases, human granulocytic anaplasmosis, and babesiosis: clinical practice guidelines by the infections diseases society of America
Clin Inf Dis
 , 
2006
, vol. 
43
 (pg. 
1089
-
134
)
10
Seaton
RA
Sharp
E
Bezlyak
V
Weir
CJ
Factors associated with outcome and duration of therapy in outpatient parenteral antibiotic therapy (OPAT) patients with skin and soft-tissue infections
Int J Antimicrob Agents
 , 
2011
, vol. 
38
 (pg. 
243
-
8
)
11
Wormser
GP
Nadelman
RB
Dattwyler
RJ
Dennis
DT
Shapiro
ED
Steere
AC
, et al.  . 
Practice Guidelines for the Treatment of Lyme Disease
Clin Inf Dis
 , 
2000
, vol. 
31
 
Suppl. 1
(pg. 
S1
-
S14
)
12
Barbour
AG
Expert advice and patient expectations: laboratory testing and antibiotics for Lyme disease
JAMA
 , 
1998
, vol. 
279
 (pg. 
239
-
40
)
13
Chapman
ALN
Seaton
RA
Cooper
MA
Hedderwick
S
Goodall
V
Reed
C
, et al.  . 
Good practice recommendations for outpatient parenteral antimicrobial therapy (OPAT) in adults in the UK: a consensus statement
J Antimicrob Chemother
 , 
2012
, vol. 
67
 (pg. 
1053
-
62
)
14
Drug information online