European Journal of Neurology 2014
doi:10.1111/ene.12420
Oral doxycycline for Lyme neuroborreliosis with symptoms of encephalitis, myelitis, vasculitis or intracranial hypertension D. Bremella and L. Dotevalla,b a
Department of Infectious Diseases, Institute of Biomedicine, Sahlgrenska Academy at the University of Gothenburg, G€ oteborg; and Department of Communicable Disease Control, V€ astra G€ otaland Region, G€ oteborg, Sweden
b
Keywords:
central nervous system Lyme disease, doxycycline, Lyme borreliosis, Lyme neuroborreliosis Received 14 January 2014 Accepted 24 February 2014
Background and purpose: The treatment recommendation for Lyme neuroborreliosis with central nervous system (CNS) symptoms is intravenous ceftriaxone, according to current American and European guidelines. For Lyme neuroborreliosis with peripheral nervous system (PNS) symptoms, treatment with intravenous ceftriaxone and oral doxycycline is considered equally effective. The purpose of this study was to evaluate the efficacy of oral doxycycline in the treatment of Lyme neuroborreliosis with CNS symptoms. Methods: Patients with Lyme neuroborreliosis who had undergone cerebrospinal fluid (CSF) sampling before and after treatment at the Department of Infectious Diseases, Sahlgrenska University Hospital, during the period 1990–2012, were included in this retrospective study. The CSF mononuclear cell count was used as a surrogate marker of treatment outcome. Comparisons of CSF mononuclear cell counts were made between patients with CNS symptoms and patients with PNS symptoms before and after treatment with oral doxycycline. Results: Twenty-six patients classified as having CNS symptoms and 115 patients classified as having PNS symptoms were included. The decline in CSF mononuclear cell counts did not differ significantly between the two groups of patients. All patients with CNS disease showed a marked clinical improvement after treatment, even though 62% had remaining symptoms at the end of follow-up. Conclusion: Treatment with oral doxycycline resulted in a similar decrease in CSF mononuclear cell counts in patients with Lyme neuroborreliosis with CNS symptoms compared with patients with Lyme neuroborreliosis with PNS symptoms. The results indicate that oral doxycycline is an effective treatment for Lyme neuroborreliosis irrespective of the severity of symptoms.
Introduction The most common symptoms of European Lyme neuroborreliosis (henceforth referred to as Lyme neuroborreliosis – LNB) are painful meningoradiculitis (Garin Boujadoux Bannwarth syndrome) and facial nerve palsy. These symptoms are considered to be mainly of peripheral nervous system (PNS) origin, although the exact location of the pathological changes responsible remains to be conclusively established [1,2]. Symptoms originating from the central nervous system (CNS) are less common and the exact incidence is not known. Borrelia burgdorferi infection of the CNS is known to cause encephalitis, segmental Correspondence: D. Bremell, Department of Infectious Diseases, Institute of Biomedicine, Sahlgrenska Academy at the University of Gothenburg, SE-416 85 G€ oteborg, Sweden (tel.: +46 31 343 41 69; fax: +46 31 19 61 37; e-mail: [email protected]).
© 2014 The Author(s) European Journal of Neurology © 2014 EFNS
myelitis, vasculitis and intracranial hypertension. The clinical symptoms of these rare pathological conditions may include ataxia, paraparesis, sphincter dysfunction, Parkinson-like symptoms, confusion and severe cognitive impairment [3,4]. The outcome of LNB with severe symptoms and late presentation is often less favorable, although there are no controlled studies of different treatment options or sequelae after LNB with CNS symptoms [5]. The diagnosis of LNB is based on anamnestic and clinical data, as well as cerebrospinal fluid (CSF) analyses. Of the laboratory analyses, the presence of CSF mononuclear pleocytosis is considered mandatory [6]. Levels of CSF mononuclear cells decrease after treatment initiation, with a steep decline in the first weeks, after which the decline levels off, so that the curve in a graph plotting the cell count against time after treatment initiation takes on a logarithmic
1
2
D. Bremell and L. Dotevall
or exponential form [7,8]. The decline in CSF mononuclear cell count has been used as a surrogate marker of treatment effect [9,10]. According to European guidelines, LNB with only PNS symptoms should be treated with either oral doxycycline or intravenous ceftriaxone for 14 days. For LNB with CNS manifestations, intravenous ceftriaxone is the preferred treatment [1]. However, at the Department of Infectious Diseases, Sahlgrenska University Hospital, Gothenburg, oral doxycycline has been the standard-of-care treatment for LNB since 1987, irrespective of the severity of symptoms. The purpose of this study was to evaluate the efficacy of oral doxycycline for LNB with symptoms indicating CNS involvement. Since detailed clinical outcome is difficult to evaluate retrospectively, the main objective was to analyze the decline in mononuclear CSF cell counts after antibiotic treatment as a surrogate marker of treatment effect.
nial hypertension. Encephalitis was defined as focal or disseminated symptoms due to inflammation of the brain parenchyma according to the clinical assessment [11]. Diagnoses and findings classified as being indicative of PNS involvement were radicular pain, cranial nerve paresis and/or sensory disturbances. Laboratory tests
Two different tests were used for the analysis of antibodies to B. burgdorferi in CSF during the study period. Until 26 June 2006 a commercial enzyme-linked immunosorbent assay for immunoglobulin G (IgG) and IgM antibodies to purified native B. burgdorferi flagellum was used (Dako, Glostrup, Denmark). Starting on 26 June 2006, CSF samples were analyzed with a commercial sandwich chemiluminescence immunoassay, using recombinant B. burgdorferi antigens to detect IgG and IgM antibodies (DiaSorin, Saluggia, Italy).
Materials and methods Statistics Patients
The study was conducted at the Department of Infectious Diseases, Sahlgrenska University Hospital, Gothenburg, as a retrospective study. The inclusion criteria were (i) a first visit to the clinic during the period 1990–2012, (ii) treatment with oral doxycycline and (iii) CSF sampling before and within 14 weeks after treatment. In the period 1990–2000, all patients with LNB were asked to participate in a structured follow-up program that included CSF sampling 4– 8 weeks after treatment and follow-up visits for at least 6 months after treatment. In the period 2000– 2012, patients with LNB were not followed according to a standardized protocol but follow-up CSF samplings were performed liberally, especially in patients with more severe symptoms. Patients’ hospital records were reviewed and data on predetermined parameters were collected. The study was approved by the regional ethical review board at the University of Gothenburg. Case definitions
The diagnostic criteria for LNB were (i) symptoms consistent with LNB and other explanations excluded, (ii) CSF mononuclear pleocytosis (>5 cells/ll) and one of the following: (a) preceding erythema migrans within 3 months or (b) the detection of B. burgdorferi-specific antibodies in CSF. Diagnoses and findings classified as being indicative of CNS involvement were encephalitis, myelitis, CNS vasculitis and/or intracra-
Quantitative variables are presented as the median (range). The Spearman rank correlation was used for correlation analysis. The Mann–Whitney U test was used for comparisons between groups. For the analysis of paired data, the Wilcoxon matched pairs test was used. For the analysis of slope coefficients, ANCOVA was used. All calculations were performed using Prism 5.0 (GraphPad Software, LaJolla, CA, USA). P values
doi:10.1111/ene.12420
Oral doxycycline for Lyme neuroborreliosis with symptoms of encephalitis, myelitis, vasculitis or intracranial hypertension D. Bremella and L. Dotevalla,b a
Department of Infectious Diseases, Institute of Biomedicine, Sahlgrenska Academy at the University of Gothenburg, G€ oteborg; and Department of Communicable Disease Control, V€ astra G€ otaland Region, G€ oteborg, Sweden
b
Keywords:
central nervous system Lyme disease, doxycycline, Lyme borreliosis, Lyme neuroborreliosis Received 14 January 2014 Accepted 24 February 2014
Background and purpose: The treatment recommendation for Lyme neuroborreliosis with central nervous system (CNS) symptoms is intravenous ceftriaxone, according to current American and European guidelines. For Lyme neuroborreliosis with peripheral nervous system (PNS) symptoms, treatment with intravenous ceftriaxone and oral doxycycline is considered equally effective. The purpose of this study was to evaluate the efficacy of oral doxycycline in the treatment of Lyme neuroborreliosis with CNS symptoms. Methods: Patients with Lyme neuroborreliosis who had undergone cerebrospinal fluid (CSF) sampling before and after treatment at the Department of Infectious Diseases, Sahlgrenska University Hospital, during the period 1990–2012, were included in this retrospective study. The CSF mononuclear cell count was used as a surrogate marker of treatment outcome. Comparisons of CSF mononuclear cell counts were made between patients with CNS symptoms and patients with PNS symptoms before and after treatment with oral doxycycline. Results: Twenty-six patients classified as having CNS symptoms and 115 patients classified as having PNS symptoms were included. The decline in CSF mononuclear cell counts did not differ significantly between the two groups of patients. All patients with CNS disease showed a marked clinical improvement after treatment, even though 62% had remaining symptoms at the end of follow-up. Conclusion: Treatment with oral doxycycline resulted in a similar decrease in CSF mononuclear cell counts in patients with Lyme neuroborreliosis with CNS symptoms compared with patients with Lyme neuroborreliosis with PNS symptoms. The results indicate that oral doxycycline is an effective treatment for Lyme neuroborreliosis irrespective of the severity of symptoms.
Introduction The most common symptoms of European Lyme neuroborreliosis (henceforth referred to as Lyme neuroborreliosis – LNB) are painful meningoradiculitis (Garin Boujadoux Bannwarth syndrome) and facial nerve palsy. These symptoms are considered to be mainly of peripheral nervous system (PNS) origin, although the exact location of the pathological changes responsible remains to be conclusively established [1,2]. Symptoms originating from the central nervous system (CNS) are less common and the exact incidence is not known. Borrelia burgdorferi infection of the CNS is known to cause encephalitis, segmental Correspondence: D. Bremell, Department of Infectious Diseases, Institute of Biomedicine, Sahlgrenska Academy at the University of Gothenburg, SE-416 85 G€ oteborg, Sweden (tel.: +46 31 343 41 69; fax: +46 31 19 61 37; e-mail: [email protected]).
© 2014 The Author(s) European Journal of Neurology © 2014 EFNS
myelitis, vasculitis and intracranial hypertension. The clinical symptoms of these rare pathological conditions may include ataxia, paraparesis, sphincter dysfunction, Parkinson-like symptoms, confusion and severe cognitive impairment [3,4]. The outcome of LNB with severe symptoms and late presentation is often less favorable, although there are no controlled studies of different treatment options or sequelae after LNB with CNS symptoms [5]. The diagnosis of LNB is based on anamnestic and clinical data, as well as cerebrospinal fluid (CSF) analyses. Of the laboratory analyses, the presence of CSF mononuclear pleocytosis is considered mandatory [6]. Levels of CSF mononuclear cells decrease after treatment initiation, with a steep decline in the first weeks, after which the decline levels off, so that the curve in a graph plotting the cell count against time after treatment initiation takes on a logarithmic
1
2
D. Bremell and L. Dotevall
or exponential form [7,8]. The decline in CSF mononuclear cell count has been used as a surrogate marker of treatment effect [9,10]. According to European guidelines, LNB with only PNS symptoms should be treated with either oral doxycycline or intravenous ceftriaxone for 14 days. For LNB with CNS manifestations, intravenous ceftriaxone is the preferred treatment [1]. However, at the Department of Infectious Diseases, Sahlgrenska University Hospital, Gothenburg, oral doxycycline has been the standard-of-care treatment for LNB since 1987, irrespective of the severity of symptoms. The purpose of this study was to evaluate the efficacy of oral doxycycline for LNB with symptoms indicating CNS involvement. Since detailed clinical outcome is difficult to evaluate retrospectively, the main objective was to analyze the decline in mononuclear CSF cell counts after antibiotic treatment as a surrogate marker of treatment effect.
nial hypertension. Encephalitis was defined as focal or disseminated symptoms due to inflammation of the brain parenchyma according to the clinical assessment [11]. Diagnoses and findings classified as being indicative of PNS involvement were radicular pain, cranial nerve paresis and/or sensory disturbances. Laboratory tests
Two different tests were used for the analysis of antibodies to B. burgdorferi in CSF during the study period. Until 26 June 2006 a commercial enzyme-linked immunosorbent assay for immunoglobulin G (IgG) and IgM antibodies to purified native B. burgdorferi flagellum was used (Dako, Glostrup, Denmark). Starting on 26 June 2006, CSF samples were analyzed with a commercial sandwich chemiluminescence immunoassay, using recombinant B. burgdorferi antigens to detect IgG and IgM antibodies (DiaSorin, Saluggia, Italy).
Materials and methods Statistics Patients
The study was conducted at the Department of Infectious Diseases, Sahlgrenska University Hospital, Gothenburg, as a retrospective study. The inclusion criteria were (i) a first visit to the clinic during the period 1990–2012, (ii) treatment with oral doxycycline and (iii) CSF sampling before and within 14 weeks after treatment. In the period 1990–2000, all patients with LNB were asked to participate in a structured follow-up program that included CSF sampling 4– 8 weeks after treatment and follow-up visits for at least 6 months after treatment. In the period 2000– 2012, patients with LNB were not followed according to a standardized protocol but follow-up CSF samplings were performed liberally, especially in patients with more severe symptoms. Patients’ hospital records were reviewed and data on predetermined parameters were collected. The study was approved by the regional ethical review board at the University of Gothenburg. Case definitions
The diagnostic criteria for LNB were (i) symptoms consistent with LNB and other explanations excluded, (ii) CSF mononuclear pleocytosis (>5 cells/ll) and one of the following: (a) preceding erythema migrans within 3 months or (b) the detection of B. burgdorferi-specific antibodies in CSF. Diagnoses and findings classified as being indicative of CNS involvement were encephalitis, myelitis, CNS vasculitis and/or intracra-
Quantitative variables are presented as the median (range). The Spearman rank correlation was used for correlation analysis. The Mann–Whitney U test was used for comparisons between groups. For the analysis of paired data, the Wilcoxon matched pairs test was used. For the analysis of slope coefficients, ANCOVA was used. All calculations were performed using Prism 5.0 (GraphPad Software, LaJolla, CA, USA). P values
