Ticking the right boxes: classification of patients suspected of Lyme borreliosis at an academic referral center in the Netherlands

Accepted Manuscript Ticking the right boxes; classification of patients suspected of Lyme borreliosis at an academic referral center in the Netherlands Jeroen Coumou , Eduard A. Herkes , Matthijs C. Brouwer , Diederik van de Beek , Sander W. Tas , Gerty Casteelen , Michèle van Vugt , Markus V. Starink , Henry J.C. de Vries , Bob de Wever , Lodewijk Spanjaard , Joppe W.R. Hovius , Dr. PII:

S1198-743X(14)00100-1

DOI:

10.1016/j.cmi.2014.11.014

Reference:

CMI 99

To appear in:

Clinical Microbiology and Infection

Received Date: 17 June 2014 Revised Date:

5 November 2014

Accepted Date: 14 November 2014

Please cite this article as: Coumou J, Herkes EA, Brouwer MC, van de Beek D, Tas SW, Casteelen G, van Vugt M, Starink MV, de Vries HJC, de Wever B, Spanjaard L, Hovius JWR, Ticking the right boxes; classification of patients suspected of Lyme borreliosis at an academic referral center in the Netherlands, Clinical Microbiology and Infection (2014), doi: 10.1016/j.cmi.2014.11.014. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT 1

Original article

2 Ticking the right boxes; classification of patients suspected of

4

Lyme borreliosis at an academic referral center in the

5

Netherlands.

6 7

Running title: Patients suspected of Lyme borreliosis

8

RI PT

3

Jeroen Coumou1,#, Eduard A. Herkes1,#, Matthijs C. Brouwer2,3,

10

Diederik van de Beek2,3, Sander W. Tas2,4, Gerty Casteelen2,5,

11

Michèle van Vugt2,6 Markus V. Starink2,7, Henry J.C. de Vries2,7,8,

12

Bob de Wever2,9, Lodewijk Spanjaard2,9 & Joppe W.R. Hovius1,2,5,*

13

M AN U

SC

9

14

1

15

Multidisciplinary

16

Neurology;

17

7

Dermatology; 8Center for Infection and Immunity Amsterdam and

18

9

Medical Microbiology, Academic Medical Center; University of

19

Amsterdam; The Netherlands

borreliosis

TE D

Lyme

Rheumatology;

5

Psychiatry;

6

3

Departments

of

Infectious Diseases;

EP

4

Center,

AC C

20

Center for Experimental and Molecular Medicine; 2Amsterdam

21

# These authors contributed equally to the described work.

22 23

* corresponding author

24

Dr. Joppe W.R. Hovius, Department of Internal Medicine, Division

25

of Infectious Diseases, Academic Medical Center, University of

26

Amsterdam, Meibergdreef 9, Amsterdam 1105 AZ, The Netherlands

27

Email address: [email protected] Tel.: +31 20 566 59 10

28

Fax : +31 20 697 71 92

ACCEPTED MANUSCRIPT Abstract

30

To provide better care for patients suspected of Lyme borreliosis

31

(LB) we founded the Amsterdam Multidisciplinary Lyme borreliosis

32

Center (AMLC). The AMLC reflects a collaborative effort of the

33

departments of internal medicine/infectious diseases, rheumatology,

34

neurology, dermatology, medical microbiology and psychiatry. In a

35

retrospective case series characteristics of 200 adult patients referred

36

to the AMLC were recorded and patients were classified as having

37

LB, post treatment Lyme borreliosis syndrome (PTLBS), persistent

38

B. burgdorferi s.l. infection despite antibiotic treatment, or as having

39

no LB. In addition, LB, PTLBS and persistent B. burgdorferi s.l.

40

infection cases were classified as ‘definite’, ‘probable’ or

41

‘questionable’. Of the 200 patients, 120 (60%) did not have LB and

42

31 (16%) had a form of localized or disseminated LB, of which 12

43

were classified as definite, 6 as probable and 13 as questionable. In

44

addition, 34 patients (17%) were diagnosed with PTLBS, of which

45

22 (11%) as probable and 12 (6%) as questionable. A total of 15

46

patients (8%) were diagnosed with persistent B. burgdorferi s.l.

47

infection, of which none were classified as definite, 3 as probable and

48

12 as questionable. In conclusion, in line with previous studies, the

AC C

EP

TE D

M AN U

SC

RI PT

29

49

number of definite and probable (persisting) LB cases was low. The

50

overall high number of questionable cases illustrates the fact that it

51

can sometimes be challenging to either rule out or demonstrate an

52

association with a B. burgdorferi s.l. infection, even in an academic

53

setting. Finally, we were able to establish alternative diagnoses in a

54

large proportion of patients.

ACCEPTED MANUSCRIPT 55

Keywords

56

Lyme Disease, Tertiary Care Centers, Chronic Disease, Retrospective

57

Studies, Multidisciplinary Communications

AC C

EP

TE D

M AN U

SC

RI PT

58

ACCEPTED MANUSCRIPT Introduction

60

Lyme borreliosis (LB) is the most common tick-borne disease in the

61

North-Eastern part of the United States of America and in Europe in

62

zones of temperate climate [1]. LB is caused by spirochetes of the

63

Borrelia burgdorferi sensu lato (s.l.) group [2]. In the Netherlands,

64

the number of LB cases appears to be on the rise, from 100 per

65

100.000 inhabitants in 2005 to 134 per 100.000 inhabitants in 2009

66

[3]. Similarly, the number of visits to Dutch general practitioners

67

(GPs) for tick bites rose from 371 per 100.000 in 2001 to 446 and

68

564 in 2005 and 2009 respectively. Recently, the Dutch ministry of

69

Health has asked for concerted action on ticks and LB and asked for

70

the development of a nation-wide collaborative effort between

71

medical and scientific institutes focusing on LB to improve LB care

72

and research in the Netherlands.

M AN U

SC

RI PT

59

TE D

73

Diagnosis and treatment of early localized LB in the Netherlands is

75

mostly done by GPs, but in case of atypical localized or disseminated

76

disease patients are often referred to medical specialists. According

77

to international guidelines and the recently updated Dutch national

78

guideline, objective clinical findings of early localized LB include

AC C

EP

74

79

erythema migrans (EM) and objective clinical findings of

80

disseminated LB include Borrelia lymphocytoma, multiple EM,

81

Lyme arthritis, Lyme carditis, Lyme neuroborreliosis (LNB) amongst

82

other more rare manifestations [4-7]. Acrodermatitis chronica

83

atrophicans (ACA) is usually referred to as late LB. In case of an

84

EM, which is pathognomonic for LB, no further testing is

85

recommended since EM can precede the antibody response [8, 9]. In

86

contrast, serological testing of antibodies against B. burgdorferi s.l.

ACCEPTED MANUSCRIPT in serum is required to confirm the diagnosis of disseminated LB.

88

When appropriate, the diagnosis of disseminated LB can be further

89

supported by evidence from additional diagnostics, including culture

90

and PCR of B. burgdorferi s.l. on skin, synovial fluid or

91

cerebrospinal fluid (CSF) or suggestive histopathological findings.

92

Although the prognosis of LB after recommended antibiotic

93

treatment is good and microbiological failure appears to be an

94

infrequent event as discussed elsewhere [4], patients may experience

95

long

96

recommended antibiotic treatment. This condition has been referred

97

to as post-treatment Lyme borreliosis syndrome or post Lyme disease

98

syndrome (PTLBS or PLDS) [4, 7], and randomized controlled trials

99

did now show substantial or long-lasting beneficial effects of

debilitating

subjective

symptoms

despite

SC

and

M AN U

100

lasting

RI PT

87

additional antibiotic treatment compared to placebo [10-13].

TE D

101

With the available serological tests it is difficult to differentiate

103

between active and past B. burgdorferi s.l. infection and 4-8 % of the

104

Dutch population has detectable antibodies against B. burgdorferi s.l.

105

[14]. Therefore, it is not recommended to test patients with subjective

106

symptoms without objective clinical findings compatible with LB.

AC C

EP

102

107

Nonetheless, approximately 70 % of the serological tests ordered by

108

GPs are from such patients [15]. To establish a diagnosis in a patient

109

presenting with subjective symptoms, with a history of tick bites

110

and/or antibodies against B. burgdorferi s.l. - previously treated or

111

not treated for LB - can be a challenge for physicians. On one hand,

112

misdiagnosis of LB can lead to (multiple) antibiotic courses, without

113

effect, but with (serious) side effects or a delay in identification and

114

management of the actual underlying cause of the complaints [16].

ACCEPTED MANUSCRIPT 115

On the other hand, when the clinical presentation is less clear or

116

when diagnostic tests are not performed as, or when, they should be,

117

a missed diagnosis could result in prolonged or progressive illness.

118 Therefore, in an attempt to offer better care for patients suspected of

120

LB, we have initiated the Amsterdam Multidisciplinary Lyme

121

borreliosis Center (AMLC). At the AMLC, various medical

122

specialists, including infectious diseases specialists, neurologists,

123

dermatologists and rheumatologists, collaborate to establish a

124

diagnosis in the referred patient – either LB or an alternative

125

diagnosis – and treat accordingly. In this report, we describe the

126

characteristics of the first 200 adult patients who were referred to the

127

AMLC.

AC C

EP

TE D

M AN U

SC

RI PT

119

ACCEPTED MANUSCRIPT 128

Materials and Methods

129 The AMLC

131

The AMLC is located at the outpatient clinic of the Academic

132

Medical Center (AMC) of the University of Amsterdam in the

133

Netherlands. The AMLC is open to referral of patients by GPs from

134

the Amsterdam region and medical specialist from all over the

135

Netherlands. Referrals were accepted - after being centrally judged

136

by an infectious disease specialist (JH or MvV) - when there was a

137

suspicion of LB, either based on the described symptoms or signs or

138

the results of previous diagnostic tests, or when the referring

139

physician specifically requested referral. Based on the provided

140

clinical information patients were invited to the appropriate

141

outpatient clinic, i.e. the outpatient clinic of Infectious Diseases,

142

Neurology, Rheumatology or Dermatology. In addition, the

143

department of medical microbiology was frequently consulted.

144

Within each department there were one or two dedicated specialists

145

who were responsible for patients suspected of LB. At all AMLC

146

outpatient clinics information on tick bites, symptoms compatible

147

with LB, previous serological testing and antibiotic treatment was

AC C

EP

TE D

M AN U

SC

RI PT

130

148

obtained and a physical examination in search of objective clinical

149

findings compatible with LB was conducted. In the majority of cases

150

a B. burgdorferi s.l. C6-EIA (IgM/IgG, Immunetics) and upon

151

indication an immunoblot (either IgM and/or IgG) (Mikrogen) was

152

performed by the department of medical microbiology. Tests were

153

considered positive based on the manufacturer’s cut-off or

154

interpretation criteria. Patients suspected of LNB were seen by

155

neurologists. At the department of neurology, lumbar punctures were

ACCEPTED MANUSCRIPT performed when patients were suspected for LNB (for LNB criteria

157

see below). Patients suspected of Lyme arthritis were seen by

158

rheumatologists. A synovial fluid aspiration for a B. burgdorferi s.l.

159

PCR (see Supplementary Information) was performed at the

160

discretion of the treating physician. Patients with skin lesions were

161

seen at the dermatology outpatient clinic, where skin biopsies for B.

162

burgdorferi s.l. culture (see Supplementary Information), PCR or

163

histology were performed upon indication. After this initial

164

(multidisciplinary) evaluation it was determined whether the patient

165

required (additional) antibiotic treatment for LB. Antibiotic treatment

166

regimes, dosages and duration were in concordance with the recent

167

national CBO guideline [6]. Additional testing - such as blood tests

168

and imaging, to rule out or establish an alternative diagnosis – and

169

therapeutic interventions were left to the discretion of the treating

170

physicians. No systematic follow-up of patients was present at the

171

AMLC. However, follow-up was collected from the documented

172

clinical impression of the treating physician or from the patient’s

173

experience if documented. Further follow-up was usually performed

174

through the GP, who was given written advice for further

175

management.

AC C

EP

TE D

M AN U

SC

RI PT

156

176 177

Consecutive retrospective case series and classifications

178

Case record forms from patients, which were referred between

179

January 2011 and April 2013, were retrospectively reviewed (see

180

below) using standardized forms. Information on (pre-visit)

181

diagnostic test results, medical history, objective clinical findings,

182

subjective symptoms and previous treatment was recorded and

183

analyzed using the SPSS (version 21) software. Based on this

ACCEPTED MANUSCRIPT information, patients were classified into different categories: early

185

localized LB or disseminated LB if patients were not previously

186

treated (Table 1A), and persistent B. burgdorferi s.l. infection or

187

PTLBS if patients were previously treated (Table 1B). To address the

188

likelihood of a causal relationship between complaints and an active

189

or past B. burgdorferi s.l. infection, these four categories were further

190

classified as ‘definite’, ‘probable’ and ‘questionable’. Definite cases

191

have a low risk, probable cases a low to intermediate risk and

192

questionable cases a high risk of being misclassified. Of note,

193

alternative diagnoses were not found or were considered unlikely in

194

all of the case definitions mentioned above. Finally, patients not

195

fulfilling criteria for any of these categories were classified as not

196

having LB (No LB), in some of which an alternative diagnosis could

197

be found or considered. All cases were reviewed by two reviewers

198

(EH and JC). If there was disagreement between the two reviewers

199

about the classification or both could not classify the patient into a

200

distinct category, cases were classified by JH. Our retrospective

201

analysis is in accordance with the AMC research code, which is

202

based on the Helsinki Declaration of 1975. For a more detailed

203

description of our classifications see below and Table 1A and Table

AC C

EP

TE D

M AN U

SC

RI PT

184

204

1B.

205 206

Definite early localized LB and disseminated LB Cases with definite

207

LB included: I) patients presenting with objective clinical findings

208

compatible with LB as described in (inter)national guidelines and

209

supportive evidence from laboratory tests, such as B. burgdorferi s.l.

210

serological tests, culture, PCR or suggestive histopathological

211

findings [4-6]; II) cases with neurological findings compatible with

ACCEPTED MANUSCRIPT LNB, as described by the European Federation of Neurological

213

Societies (EFNS) guideline, with a pleocytosis in CSF and positive

214

intrathecal anti-B. burgdorferi s.l. IgG antibody index [17]. LNB

215

cases with a duration of symptoms longer than six months were

216

considered as late disseminated LB; III) patients presenting with

217

objective clinical findings reminiscent of LB, e.g. atypical skin

218

lesions or a polyarthritis without involvement of large joints, which

219

were supported by positive B. burgdorferi s.l. culture, PCR and/or

220

suggestive histopathological findings.

SC

221

RI PT

212

Probable early localized LB and disseminated LB Cases were

223

classified as probable LB when I) objective clinical findings

224

reminiscent of LB were present in combination with B. burgdorferi

225

s.l. antibodies; II) neurological findings compatible with LNB, with

226

only a pleocytosis in CSF or a positive intrathecal anti-B. burgdorferi

227

s.l. IgG antibody index, in combination with positive serological tests

228

for B. burgdorferi s.l. antibodies in serum.

TE D

EP

229

M AN U

222

Questionable disseminated LB Patients, in which no objective

231

clinical findings compatible with, or reminiscent of, LB were present

AC C

230

232

in combination with B. burgdorferi s.l. antibodies in serum, as

233

determined by serological tests. In addition there was either a relation

234

between the onset of symptoms with a tick bite or a non-documented

235

EM or LB manifestation in the past.

236 237

PTLBS The classification of probable PTLBS was in line with the

238

PLDS criteria from the Infectious Diseases Society of America

239

(IDSA) guideline [4]. We did not designate this as definite PTLBS,

ACCEPTED MANUSCRIPT since this diagnosis cannot be definite in our opinion. Cases

241

classified as questionable PTLBS were identical to probable PTLBS

242

cases, except for the fact that the preceding LB episode was

243

questionable (see above for the definitions) and they were required to

244

have positive serological tests for antibodies against B. burgdorferi

245

s.l. We designated these patients as questionable PTLBS rather than

246

medically unexplained symptoms (MUS), since a relation with a

247

previous B. burgdorferi s.l. infection could not be fully excluded.

RI PT

240

SC

248

Persistent B. burgdorferi s.l. infection Patients presenting with

250

objective clinical findings compatible with LB that were previously

251

treated with antibiotics were classified as definite (positive B.

252

burgdorferi s.l. culture) or probable (positive B. burgdorferi s.l. PCR

253

and/or

254

burgdorferi s.l. infection. Patients without objective clinical findings,

255

but with subjective symptoms progressive over time despite previous

256

recommended or inappropriate antibiotic treatment for a documented

257

LB episode, were classified as questionable. In addition, questionable

258

persistent B. burgdorferi s.l. infection included patients presenting

259

with progressive subjective symptoms despite previous inappropriate

histopathological

findings)

persistent

B.

AC C

EP

TE D

suggestive

M AN U

249

260

antibiotic treatment for a questionable LB episode (see above for the

261

definitions) in the past. We considered antibiotic treatment

262

inappropriate when not meeting guideline recommendations [4-6],

263

i.e. too short of duration, insufficient dosage, insufficient frequency,

264

use of a non-recommended ineffective antibiotic or the simultaneous

265

use of supplements, such as calcium tablets, together with

266

tetracyclines.

ACCEPTED MANUSCRIPT Results

268

Of the patients referred to the AMLC, most were referred by GPs

269

(n=162, 81%) (Table 2). Fatigue was the most reported complaint,

270

i.e. by 141 (71 %) patients. Other common reported symptoms

271

included arthralgia, myalgia, paresthesia and headache (Table 2).

272

Skin lesions were the most reported objective clinical finding, i.e. in

273

31 (16 %) patients. More than half of the patients (n=108, 54 %) had

274

symptoms that were present for more than one year at the time of

275

presentation at the AMLC, of which only three had objective clinical

276

findings that were progressive over time. Prior to referral to the

277

AMLC, for the majority of patients serological testing was performed

278

and approximately half of the patients had received antibiotic

279

treatment based on a suspicion of LB (Table 2).

M AN U

SC

RI PT

267

280

A B. burgdorferi s.l. C6-EIA on serum as part of the AMLC’s

282

diagnostic work-up was done in 168 (84 %) patients and was

283

considered positive in 66 tested sera (40 % of tested sera)

284

(Supplementary Table 1). In the remaining 32 patients, the treating

285

specialist at the AMLC deemed additional testing unnecessary, based

286

on a low a priori chance of having LB or previous serological test

AC C

EP

TE D

281

287

results (Table 2). For many patients immunoblots had been

288

performed prior to referral to the AMLC. Therefore, an immunoblot

289

was done only in 74 (37 %) patients, of which 28 (38 % of tested

290

sera) were positive or indeterminate (Supplementary Table 1). A total

291

of 20 PCRs on skin biopsies, synovial fluid and CSF were done to

292

strengthen or confirm the diagnosis of EM, ACA, Lyme arthritis or

293

LNB. In addition, 29 lumbar punctures to detect specific intrathecal

294

antibody production - by C6 EIA - and pleocytosis in CSF were

ACCEPTED MANUSCRIPT performed to confirm or rule out LNB (Supplementary Table 1). We

296

also tested blood samples of 29 patients - either on the patients’

297

explicit request or because patients had a reported positive PCR

298

blood test from a commercial laboratory prior to referral - using our

299

clinically validated PCR - and found no positives (data not shown).

RI PT

295

300

A graphical summary of the referral process and the analysis at the

302

AMLC is shown in Figure 1A (Figure 1A). Based on the criteria

303

shown in Table 1 we concluded that 120 (60%) patients did not have

304

LB (Table 3). In 43 of these patients an alternative diagnosis was

305

established (Supplementary Table 2), among which for example

306

seven patients with osteoarthritis. Patients were also diagnosed with

307

HIV-infection, polymyalgia rheumatica or multiple sclerosis,

308

amongst other diagnoses.

TE D

309

M AN U

SC

301

An active form of LB not previously treated with antibiotics was

311

diagnosed in 31 (16 %) patients, of which only 12 (6 %) were

312

classified as definite LB, including 5 EM, 2 multiple EM, 1 Lyme

313

arthritis, 1 LNB and 3 ACA (Figure 1B). In addition, we classified 6

314

patients with probable LB, including 3 cases with skin lesion(s) - 2

AC C

EP

310

315

atypical EM and 1 atypical multiple EM - and 3 LNB cases

316

supported by pleocytosis in CSF and B. burgdorferi s.l. antibodies in

317

serum (not in CSF). The remaining 13 LB patients were classified as

318

questionable LB. The most reported symptoms in patients with

319

questionable LB were fatigue, arthralgia, paresthesia, myalgia and

320

headache. In 10 patients with questionable LB a tick bite related to

321

the onset of the symptoms was reported and in the other three the

322

patients reported a non-documented and untreated EM in the past.

ACCEPTED MANUSCRIPT 323 From the 200 referred patients, 104 received prior antibiotic

325

treatment. Of these patients, 34 (17 %) were diagnosed with PTLBS.

326

We classified 22 patients (11 %) as probable PTLBS, meeting the

327

criteria of the published case definition [4], and 12 patients (6 %) as

328

having questionable PTLBS (Table 3). Finally, 15 patients (8 %)

329

were classified as persistent B. burgdorferi s.l. infection, of which

330

none were classified as definite, three as probable and the majority

331

(n=12) as questionable. The three patients with probable persistent B.

332

burgdorferi s.l. infection included one patient that was diagnosed

333

with a persisting EM - based on ongoing inflammation observed by

334

histopathological examination of a skin section obtained by skin

335

biopsy - after antibiotic treatment for an EM that had lasted for two

336

months. However, B. burgdorferi s.l. culture and PCR on skin

337

samples were negative. The second patient was diagnosed with

338

persisting Lyme arthritis after previous antibiotic treatment for Lyme

339

arthritis, supported by a B. burgdorferi s.l. PCR on synovial fluid.

340

The third patient presented with recurrent arthritis of the left ankle

341

and IgG antibodies against B. burgdorferi s.l. in serum. Prior to the

342

onset of these symptoms, the patient had been treated for an EM with

AC C

EP

TE D

M AN U

SC

RI PT

324

343

doxycycline for 10 days, which was followed by a peripheral facial

344

nerve paresis that had resolved over time. These three patients

345

received antibiotic retreatment at the AMLC and clinically improved.

346

In the twelve patients with questionable persistent B. burgdorferi s.l.

347

infection the most reported symptoms were fatigue (n=8, 67 %),

348

arthralgia (n=7, 58 %), paresthesia (n=6, 50 %), headache (n=6, 50

349

%) and myalgia (n=4, 33 %). In eight of the patients with

350

questionable persistent B. burgdorferi s.l. infection, previous

ACCEPTED MANUSCRIPT antibiotic treatment was regarded as inappropriate, because treatment

352

was either too short or because patients had taken calcium or other

353

supplements, which could have lowered absorption of tetracyclines

354

from the intestine. The remaining four patients in the questionable

355

persistent B. burgdorferi s.l. infection category reported progressive

356

subjective symptoms after recommended treatment for a documented

357

LB episode and no alternative explanation was evident.

RI PT

351

358

Antibiotic treatment was given to 50 patients (25 %) by physicians at

360

the AMLC, which included 27 of the 31 patients with objective

361

findings compatible with LB. The remaining four patients had

362

already started with antibiotic therapy, initiated by the referring

363

physician. All patients with probable and questionable persistent B.

364

burgdorferi s.l. infection (n=15) were treated with antibiotics.

365

Finally, eight patients retrospectively classified as having no LB or

366

PTLBS received antibiotic treatment at the AMLC.

M AN U

TE D

367

SC

359

Limited information on follow-up - several weeks to months - was

369

available for only 98 (49 %) patients, making it insufficient for a

370

thorough analysis on follow-up. Nonetheless, we compared the

AC C

EP

368

371

follow-up data from patients with objective clinical findings

372

compatible with or reminiscent of LB – i.e. patients with definite or

373

probable LB and probable persisting B. burgdorferi s.l. infection – to

374

that of patients with merely subjective symptoms – i.e. questionable

375

LB and questionable persisting B. burgdorferi s.l. infection. In

376

addition, we analyzed the follow-up data of both probable and

377

questionable PTLBS patients. From 17 out of 21 patients with

378

objective clinical findings compatible with or reminiscent of LB

ACCEPTED MANUSCRIPT follow-up data was available. All of these 17 patients improved. In

380

contrast, from 17 out of 25 questionable cases follow-up data was

381

available and only 8 (47 %) reported improvement. In addition, of 17

382

out of 34 cases with PTLBS follow-up was data available - since

383

these were usually referred back to the GP - and 15 out of these 17

384

patients (88 %) reported improvement.

AC C

EP

TE D

M AN U

SC

RI PT

379

ACCEPTED MANUSCRIPT Discussion

386

In this retrospective case series, we classified 200 patients that were

387

referred to our multidisciplinary LB referral clinic. The relatively low

388

number of patients with LB in our study may reflect the societal

389

concerns on LB diagnostics and treatment, the difficulty of excluding

390

LB from the differential diagnosis, lack of awareness of the current

391

national guidelines by the referring physicians or the lack of power to

392

discriminate between a past and active infection with current

393

serological tests. In addition, the low number of active B. burgdorferi

394

s.l. infections among patients referred to the AMLC could be caused

395

by previous referral to (multiple) other medical specialists, extensive

396

testing on LB and antibiotic treatment prior to consultation (Table 2).

397

Notably, in 43 of the 120 (36 %) AMLC patients that did not have

398

LB an alternative diagnosis was established (Supplementary Table

399

2). This illustrates there is a serious risk of improper treatment and

400

misdiagnosis in case of (self) referral to ‘LB literate’ doctors, who

401

often diagnose these patients with ‘chronic Lyme disease’ and

402

prescribe prolonged non-recommended antibiotic treatment [18, 19].

SC

M AN U

TE D

A total of 31 (16 %) patients were classified as having early localized

AC C

404

EP

403

RI PT

385

405

or disseminated LB. These observations are similar to both reports

406

from the US [20, 21], as well as a recent study from a British LB

407

referral clinic on 115 patients [22], in which 23 % of the patients

408

suspected of LB were thought to have been infected with B.

409

burgdorferi s.l. Another study on LNB in Germany reported that of

410

the 113 patients suspected of chronic LNB, one patient (<1 %) had

411

acute LNB, eight patients (7 %) had an acute LB - without LNB - and

412

six patients (5 %) had residual symptoms after previously proven and

ACCEPTED MANUSCRIPT 413

treated LNB or LB [23]. Collectively, this illustrates the low number

414

of patients with strong evidence of an active B. burgdorferi s.l.

415

infection in LB referral clinics. Indeed, the majority of LB patients in

416

our study (n=13) were classified as questionable LB.

RI PT

417 Patients classified as questionable LB presented with subjective

419

symptoms only. In general, these subjective symptoms have no

420

predictive value for LB. However, patients with questionable LB in

421

our study had positive B. burgdorferi s.l. serological tests in

422

combination with either a relation between the onset of symptoms

423

with a tick bite or a non-documented EM or LB manifestation in the

424

past. In addition, after careful examination by multiple specialists at

425

the AMLC, another explanation could not be demonstrated. It could

426

be debated whether instead of questionable LB we could have

427

designated these patients as possible, improbable or dubious LB.

428

Regardless, based on the CBO guideline, which recommends treating

429

cases with a low a priori chance for LB and positive B. burgdorferi

430

s.l. serological tests, we chose to treat these patients with antibiotics.

431

We do not recommend that GPs and physicians outside LB referral

432

clinics follow our approach and would like to emphasize that careful

AC C

EP

TE D

M AN U

SC

418

433

exclusion of other causes and consultation of other specialists is of

434

paramount importance for this patient category. We discussed with

435

these patients that if the recommended antibiotic therapy had no

436

long-lasting effects, LB was unlikely to be the cause of their

437

symptoms. The benefit of our approach might be that both physician

438

and patient can focus on additional investigations in search of the

439

etiology or - perhaps more often – adequate management of

440

medically unexplained symptoms when antibiotics did not sort any

ACCEPTED MANUSCRIPT effects. On the contrary, when symptoms did resolve, a positive

442

response to antibiotics does not necessarily mean the patient was

443

infected with B. burgdorferi s.l., since a placebo effect, an

444

immunomodulatory effect of the antibiotic or the mere effect of time

445

could have been alternative explanations.

RI PT

441

446

Another 34 (17 %) of the referred patients were diagnosed with

448

PTLBS, of which 12 were classified as questionable PTLBS, since a

449

relation with preceding B. burgdorferi s.l. infection could not be

450

ignored, because of B. burgdorferi s.l. serology and no evident

451

alternative cause. However, an active infection with B. burgdorferi

452

s.l. was considered highly unlikely. The symptoms of the PTLBS

453

patients are non-specific and share similarities with those of chronic

454

fatigue syndrome, fibromyalgia or MUS. The treatment for both

455

patients diagnosed with PTLBS and MUS is similar and does not

456

include antibiotic treatment, but an individual approach to achieve

457

acceptance and improvement of quality of life in which the GPs plays

458

a central role. Specialized MUS centers can provide multidisciplinary

459

education and advice or even cognitive therapy. The AMLC indeed

460

internally referred four patients to a psychiatrist with an expertise in

AC C

EP

TE D

M AN U

SC

447

461

the management of MUS. Recently, after the completion of this

462

study, the AMLC begun a collaboration with a MUS center at the VU

463

University Medical Center Amsterdam.

464 465

Finally, of the 15 patients diagnosed with persistent B. burgdorferi

466

s.l. infection, none were classified as definite and three were

467

classified as probable. The absence of definite persistent B.

468

burgdorferi s.l. infection cases and the low number of probable

ACCEPTED MANUSCRIPT persistent B. burgdorferi s.l. infection cases in our study is not

470

unexpected, since a persistent B. burgdorferi s.l. infection after

471

recommended antibiotic treatment appears to be a rare event [4, 24].

472

It is also possible that complaints, if caused by an active B.

473

burgdorferi s.l. infection are the result of a reinfection, rather than a

474

persistent infection [25]. The 12 patients diagnosed with questionable

475

persistent B. burgdorferi s.l. infection had subjective symptoms only,

476

similar to patients diagnosed with questionable LB, with the

477

difference that they had received (inappropriate) antibiotic treatment

478

for a prior (questionable) LB episode and that their symptoms were

479

progressive over time (Table 1B). Although partially against

480

published trials [10-13], and not recommended by the IDSA

481

guideline [4], retreatment of questionable persistent B. burgdorferi

482

s.l. infection cases, especially those which had received prior

483

inappropriate treatment, is in accordance with recommendations from

484

the recent Dutch national guideline [6], and was the result of a

485

compromise between physician and patient. As we did with

486

questionable LB patients, we discussed the pros and cons of

487

antibiotic treatment with questionable persistent B. burgdorferi s.l.

488

infection patients. Specifically, we discussed the fact that if they did

AC C

EP

TE D

M AN U

SC

RI PT

469

489

not respond to antibiotic therapy a persisting B. burgdorferi s.l.

490

infection was unlikely, and we discussed the option of treatment for

491

MUS. Our classification ‘questionable persistent B. burgdorferi s.l.

492

infection’ might be useful to describe the patient population in a

493

(tertiary) Lyme clinic. However, since the risk of misclassifying

494

these patients is high, such cases should only be used with caution for

495

future clinical or research purposes. Furthermore, this classification

496

should not be confused with ‘chronic Lyme disease’, which is a

ACCEPTED MANUSCRIPT 497

misnomer describing patients with chronic subjective symptoms

498

which are attributed to LB, but that is in fact a heterogeneous group

499

as previously described [17, 26].

500 Although incomplete and limited, our follow-up analysis showed that

502

antibiotic treatment resulted more often in improvement in patients

503

with objective clinical findings compatible with or reminiscent of LB

504

compared to patients in which only subjective symptoms were

505

present. In future studies we will strive for more accurate and

506

complete follow-up over a longer period of time, which will be

507

facilitated in the near future by a multi-center prospective study

508

assessing the risk of, and the risk factors for, developing persisting

509

symptoms after treated LB. In addition, once the number of well-

510

defined (definite and probable) LB cases increases, we will perform

511

multiple logistic regression analysis to identify negative and/or

512

positive predictors for LB.

SC

M AN U

TE D

513

RI PT

501

To conclude, LB is an infectious disease to which specific objective

515

clinical findings have been attributed. However, by many LB is

516

invariably linked to a wide range of subjective symptoms, limited

AC C

EP

514

517

diagnostic test options, as well as poor treatment options and

518

outcomes. This affects the use of diagnostic tests for, and treatment

519

of, LB by physicians. In the current study we have used established,

520

and have proposed new, criteria to categorize patient populations at

521

LB referral centers. Using these criteria, we show that we were able

522

to exclude LB in many cases, to establish alternative diagnoses for a

523

significant group of patients and to categorize most of the patients

524

into distinct classifications. Using the currently available diagnostic

ACCEPTED MANUSCRIPT tests, for some patients - especially questionable LB and questionable

526

persisting B. burgdorferi s.l. infection cases - it is difficult to

527

determine whether these patients indeed had a symptomatic B.

528

burgdorferi s.l. infection. Future tests might be able to better

529

distinguish between past and active B. burgdorferi s.l. infections and

530

could thus partially resolve these issues and guide antibiotic

531

treatment. Until these tests are developed, validated and widely

532

available, physicians with both experience with, and affinity for, LB

533

should determine the likelihood of an active infection with B.

534

burgdorferi s.l. in each individual patient. The benefits of a tertiary

535

referral center for LB - such as the AMLC - are that this evaluation is

536

done in a multidisciplinary and systematic manner by experienced

537

specialists, it can initiate and engage in basic and clinical research on

538

LB and it will uncover alternative diagnoses. Thus, tertiary LB

539

referral centers are in direct interest of LB (suspected) patients.

SC

M AN U

TE D

AC C

EP

540

RI PT

525

ACCEPTED MANUSCRIPT Acknowledgements

542

We are grateful to prof. dr. Peter Speelman, chair of the 2004 CBO

543

Lyme borreliosis guideline, and to prof. dr. Peterhans van den Broek,

544

chair of the 2013 CBO Lyme borreliosis guideline, for critically

545

proofreading the manuscript. We are also grateful to Anneke Oei for

546

excellent work for the culturing of B. burgdorferi s.l.

AC C

EP

TE D

M AN U

SC

547

RI PT

541

ACCEPTED MANUSCRIPT 548

References

549

1.

550

disease. J Clin Invest 2004; 113: 1093-1101.

551

2.

552

Grunwaldt E, Davis JP. Lyme disease-a tick-borne spirochetosis?

553

Science 1982; 216: 1317-1319.

554

3.

555

Notermans DW, van Pelt W. Ziekte van Lyme in Nederland 1994–

556

2009: Aantal huisartsconsulten blijft toenemen. Is voorlichting en

557

curatief beleid genoeg? Infectieziekten Bulletin 2010; 21: 84–87.

558

4.

559

assessment, treatment, and prevention of Lyme disease, human

560

granulocytic

561

guidelines by the Infectious Diseases Society of America. Clin Infect

562

Dis 2006; 43: 1089-1134.

563

5.

564

clinical case definitions for diagnosis and management in Europe.

565

Clin Microbiol Infect 2011; 17: 69-79.

566

6.

567

http://www.diliguide.nl/document/1314

Steere AC, Coburn J, Glickstein L. The emergence of Lyme

RI PT

Burgdorfer W, Barbour AG, Hayes SF, Benach JL,

SC

Hofhuis A, Harms MG, van der Giessen JWB, Sprong H,

M AN U

Wormser GP, Dattwyler RJ, Shapiro ED, et al. The clinical

and

babesiosis:

clinical

practice

TE D

anaplasmosis,

EP

Stanek G, Fingerle V, Hunfeld KP, et al. Lyme borreliosis:

AC C

CBO Guideline Lymeziekte 2013

568

7.

569

Lyme borreliosis. Lyme borreliosis in the Netherlands. Neth J Med

570

2011; 69: 101-111.

571

8.

572

study of serologic tests for lyme disease. Clin Infect Dis 2008; 47:

573

188-195.

574

9.

575

Wormser GP, Steere AC. 2-tiered antibody testing for early and late

Coumou J, van der Poll T, Speelman P, Hovius JW. Tired of

Steere AC, McHugh G, Damle N, Sikand VK. Prospective

Branda JA, Aguero-Rosenfeld ME, Ferraro MJ, Johnson BJ,

ACCEPTED MANUSCRIPT 576

Lyme disease using only an immunoglobulin G blot with the addition

577

of a VlsE band as the second-tier test. Clin Infect Dis 2010; 50: 20-

578

26.

579

10.

580

function in post-treatment Lyme disease: do additional antibiotics

581

help? Neurology 2003; 60: 1916-1922.

582

11.

583

antibiotic treatment in patients with persistent symptoms and a

584

history of Lyme disease. N Engl J Med 2001; 345: 85-92.

585

12.

586

treatment of post Lyme disease (STOP-LD): a randomized double

587

masked clinical trial. Neurology 2003; 60: 1923-1930.

588

13.

589

placebo-controlled trial of repeated IV antibiotic therapy for Lyme

590

encephalopathy. Neurology 2008; 70: 992-1003.

591

14.

592

Boven CP. Prevalence of Lyme borreliosis in The Netherlands. Ned

593

Tijdschr Geneeskd 1991; 135: 2288-2292.

594

15.

595

sensu lato serology in the Netherlands: guidelines versus daily

RI PT

Kaplan RF, Trevino RP, Johnson GM, et al. Cognitive

SC

Klempner MS, Hu LT, Evans J, et al. Two controlled trials of

M AN U

Krupp LB, Hyman LG, Grimson R, et al. Study and

TE D

Fallon BA, Keilp JG, Corbera KM, et al. A randomized,

EP

Nohlmans MK, van den Bogaard AE, Blaauw AA, van

AC C

Coumou J, Hovius JW, van Dam AP. Borrelia burgdorferi

596

practice. Eur J Clin Microbiol Infect Dis 2014; 33: 1803-1808.

597

16.

598

“chronic Lyme disease”. JAMA Internal Medicine 2014.

599

17.

600

Schmutzhard E, Steiner I. EFNS guidelines on the diagnosis and

601

management of European Lyme neuroborreliosis. Eur J Neurol 2010;

602

17: 8-16.

Nelson C, Elmendorf S, Mead P. Neoplasms misdiagnosed as

Mygland A, Ljostad U, Fingerle V, Rupprecht T,

ACCEPTED MANUSCRIPT 603

18.

604

appraisal of "chronic Lyme disease". N Engl J Med 2007; 357: 1422-

605

1430.

606

19.

607

and ethical concerns associated with advocacy of Lyme disease. The

608

Lancet infectious diseases 2011; 11: 713-719.

609

20.

610

The consequences of overdiagnosis and overtreatment of Lyme

611

disease: an observational study. Ann Intern Med 1998; 128: 354-362.

612

21.

613

Psychiatric comorbidity and other psychological factors in patients

614

with "chronic Lyme disease". Am J Med 2009; 122: 843-850.

615

22.

616

Beeching NJ. Lyme disease in a British referral clinic. QJM :

617

monthly journal of the Association of Physicians 2012; 105: 537-543.

618

23.

619

Eiffert H, Schmidt H. The diagnostic spectrum in patients with

620

suspected chronic Lyme neuroborreliosis--the experience from one

621

year of a university hospital's Lyme neuroborreliosis outpatients

622

clinic. Eur J Neurol 2011; 18: 547-555.

Feder HM, Johnson BJ, O'Connell S, et al. A critical

RI PT

Auwaerter PG, Bakken JS, Dattwyler RJ, et al. Antiscience

SC

Reid MC, Schoen RT, Evans J, Rosenberg JC, Horwitz RI.

M AN U

Hassett AL, Radvanski DC, Buyske S, Savage SV, Sigal LH.

TE D

Cottle LE, Mekonnen E, Beadsworth MB, Miller AR,

AC C

EP

Djukic M, Schmidt-Samoa C, Nau R, von Steinbuchel N,

623

24.

624

Subjective symptoms after treatment of early Lyme disease. Am J

625

Med 2010; 123: 79-86.

626

25.

627

Differentiation of reinfection from relapse in recurrent Lyme disease.

628

N Engl J Med 2012; 367: 1883-1890

629

26.

630

confusing entity. Tijdschrift voor Infectieziekten; 7: 20-29.

Cerar D, Cerar T, Ruzic-Sabljic E, Wormser GP, Strle F.

Nadelman RB, Hanincova K, Mukherjee P, et al.

Hovius JW, Speelman P (2012) Chronic Lyme disease: a

Table 1. Schematic overview of the classification used for diagnosis and probability of Lyme borreliosis (LB)

ACCEPTED MANUSCRIPT

Patients suspected of LB Not previously treated with antibiotics

A

Early localized Lyme borreliosis

RI PT

- Objective clinical findings compatible with disseminated LB4 - Antibodies against B. burgdorferi s.l.1 and/or supportive laboratory evidence5

- Typical EM

- Atypical macular skin lesion - Antibodies against B. burgdorferi s.l.1 and positive PCR or culture from skin biopsy

- Objective clinical findings reminiscent of LB6 - Positive supportive laboratory evidence7

- Atypical macular skin lesion - History of tick bites - Antibodies against B. burgdorferi s.l.1

- Neurological findings suggestive of LNB7 - Antibodies against B. burgdorferi s.l.1 - Either a pleocytosis in CSF or positive intrathecal IgG AI

Probable

SC

Definite

Disseminated Lyme borreliosis3

Questionable

M AN U

- Objective clinical findings reminiscent of LB6 - Antibodies against B. burgdorferi s.l.1 - Subjective symptoms only8 - Antibodies against B. burgdorferi s.l.1 - Non-documented LB episode in the past9 or a relation between the onset of symptoms and a tick bite

N.A.2

Patients suspected of LB Previously treated with antibiotics

TE D

B

Post treatment Lyme borreliosis syndrome (PTLBS)

Probable

Questionable

EP

N.A.10

- No resolution of previous documented LB episode despite prior antibiotic therapy14 - Evidence of persistent infection (positive culture)

In line with Wormser et al.11 - Previous documented LB episode12 - Current presentation with only subjective symptoms8, despite prior recommended antibiotic therapy13 - Resolution of symptoms over time

- No resolution of previous documented LB episode despite prior antibiotic therapy14 - Supportive evidence of persistent infection, i.e. B. burgdorferi s.l. PCR and/or suggestive histopathological findings

- Questionable LB episode in the past9 - Antibodies against B. burgdorferi s.l.1 - Current presentation only with subjective symptoms8 despite prior recommended antibiotic therapy13 - Resolution of symptoms over time

- Previous documented LB episode12 - Persisting subjective symptoms despite antibiotic therapy14 - No resolution or worsening of symptoms over time - Antibodies against B. burgdorferi s.l.1

AC C

Definite

Persistent B. burgdorferi s.l. infection

- Questionable LB episode in the past9 - Prior inappropriate antibiotic therapy13 - No resolution or worsening of symptoms over time - Antibodies against B. burgdorferi s.l.1

ACCEPTED MANUSCRIPT Patients not fulfilling these criteria were classified as No LB. Other evident explanations were excluded in patients fulfilling one of these criteria. Definite cases have a low risk, probable cases a low to intermediate risk and questionable case a high risk of being misclassified. For a more detailed description and explanation, see the Material and Method section. 1 As determined in serum by a B.

(Mikrogen).

2

RI PT

burgdorferi s.l. C6-EIA (IgM/IgG, Immunetics) and/or immunoblot (either IgM and/or IgG) With clinical judgment, (repeated) serology and/or skin biopsies for PCR or culture it

should be possible to distinguish between a probable early localized LB from a non-LB related skin manifestations.

3

If duration of symptoms was less than 6 months, cases were classified as early

SC

disseminated LB. If duration of symptoms was more than 6 months, or it was defined as ACA, which is usually classified as late LB, cases were classified as late disseminated LB. 4 Based on Stanek et al.

M AN U

[5], www.eucalb.com, which is in line with the Dutch CBO guideline [6] and include Borrelia lymphocytoma, multiple EM, Lyme arthritis, Lyme carditis and LNB and ACA. 5 B. burgdorferi s.l. 6

culture or PCR and/or suggestive histopathological findings.

These include atypical skin lesions,

polyarthritis without involvement of large joints, conduction disorders of the heart other than AV-

TE D

nodal conduction disorders or neurological symptoms which could be attributed to LB other than a meningoradiculitis, meningoencephalitis or polyradiculitis.

7

Based on the EFNS guideline by

Mygland et al. [17]. 8 These include non-specific symptoms, such as widespread musculoskeletal pain

EP

(arthralgia or myalgia), paresthesia or complaints of cognitive impairment with or without fatigue. LB episode in the past reported by patient, not witnessed by a physician. 11

For more details, see Wormser et al. [4].

AC C

cannot be definite.

12

10

9

In our opinion PTLBS

Previous objective clinical findings

compatible with LB4, which were witnessed by a physician and were diagnosed as LB.

13

For a

description of recommended and inappropriate treatment see material and methods section. 14 Either recommended or inappropriate treatment. ACA: acrodermatitis chronica atrophicans. AI: antibody index. CSF: cerebrospinal fluid. EM: Erythema migrans. LB: Lyme borreliosis. LNB: Lyme neuroborreliosis. N.A.: Not applicable.

ACCEPTED MANUSCRIPT

Table 2. Characteristics, presenting symptoms and pre-visit LB-related diagnostic work-up of 200 patients referred to the Amsterdam Multidisciplinary Lyme borreliosis Center. n=200

Male

82

Female

118

Age in years (median, range)

46, 18-80

Tick bites (time since last tick bite)

n=200

41 %

No tick bite

96

48 %

59 %

0-6 months

59

30 %

5

3%

39

20 %

6-12 months More than a year

SC

Referred by

RI PT

Gender

162

81 %

Time unknown

1

1%

Specialist

38

19 %

Previous B. burgdorferi s.l. serology2

170/200

85 %

of which positive

127

75 %

Previous B. burgdorferi s.l. PCR

5/200

3%

of which positive

1

20 %

32/200

16 %

27

84 %

M AN U

General Practitioner Previous referrals to other specialists for current complaints

59 %

Fatigue

141

71 %

Arthralgia

98

49 %

Paresthesia

68

34 %

Headache

54 45

Duration of symptoms

27 % 23 %

<6 weeks

16

6 weeks-3 months

20

8%

3-6 months

23

12 %

6-12 months

32

16 %

More than one year

108

54 %

No symptoms

1

1%

AC C

10 %

3

of which positive

EP

Myalgia

Other non-recommended test

TE D

Symptoms (top 5)

118 1

104/200

52 %

4

78

75 %

4

14

13 %

Other antibiotic treatment < 1 month

6

6%

Other antibiotic treatment5 > 1 month

6

6%

Antibiotic treatment Doxycycline 100 mg bid <1 month Doxycycline 100 mg bid >1 month 5

ACCEPTED MANUSCRIPT

The sum of the percentages per group can exceed 100% due to rounding. 1 Patients could have reported multiple symptoms. 2 EIA/ELISA or EIA/ELISA

RI PT

and immunoblot. 3 Other non-recommended B. burgdorferi s.l. tests include PCR on blood, Dark Field Microscopy Live Blood Analysis, lymphocyte transformation test (LTT), and reduced expression of CD57 on mononuclear cells. Tests were considered positive by the (commercial) laboratory that had performed the test. 4 May be combined with other antibiotic treatment. 5 May include: amoxicillin, atovaquone, azithromycin, ceftriaxone, ciprofloxacin,

AC C

EP

TE D

M AN U

SC

clarithromycin, metronidazole. PCR: Polymerase chain reaction. bid: Twice a day.

ACCEPTED MANUSCRIPT

Table 3. Classification of LB in the 200 patients referred to the Amsterdam Multidisciplinary Lyme borreliosis Center. Post Treatment Lyme borreliosis Syndrome (PTLBS)

Persistent B. burgdorferi s.l. infection

n=34

n=15

n=120

n=7

n=7

n=17

120 (100 %)

5 (71 %)

4 (57 %)1

3 (18 %)3

N.A.

0 (0 %)

2

4

22 (65 %)

3 (20 %)5

12 (70 %)

12 (35 %)

12 (80 %)

N.A.

2 (29 %)

2 (29 %)

Questionable

N.A.

N.A.

1 (14 %)

2 (12 %)

M AN U

Probable

SC

Definite

Disseminated Lyme borreliosis <6 months (early) >6 months (late)

RI PT

No Lyme Early localized borreliosis Lyme borreliosis

Classification is based on the criteria and definitions as shown in Table 1A and 1B. For LNB criteria, see Material and Methods. 1 One Lyme neuroborreliosis (LNB), one Lyme arthritis and two multiple erythema migrans (MEM). 2 One LNB, with pleocytosis but without B. burgdorferi s.l. antibody production in

EP AC C

text. N.A.: Not applicable, see Table 1.

TE D

CSF, and one MEM. 3 Three ACA. 4 Two late LNB, with pleocytosis but without B. burgdorferi s.l. antibody production in CSF. 5 Cases are described in the

ACCEPTED MANUSCRIPT A

Persisting symptoms after antibiotic treatment

Clinical and serological inconclusive

Second opinion requested by patient

M AN U

Suspicion of disseminated LB (or atypical EM)

Medical specialist

Amsterdam Multidisciplinary Lyme Center* Dermatology

Infectious diseases

Neurology

LB excluded 120

B

EP

No alternative diagnosis 77

AC C

Alternative diagnosis 43

Rheumatology

Psychiatry

TE D

Medical microbiology

Local LB 7

Diss. LB 24

LB excluded or succesful treatment for LB

SC

GP

RI PT

Patient suspected of LB

PTLBS 34

Persistent LB 15

ACCEPTED MANUSCRIPT Borrelia burgdorferi s.l. qPCR A qPCR targeting the B. burgdorferi s.l. ospA gene was performed using the LightCycler TaqMan Master (Roche, Penzberg, Germany) in a LightCycler 480 RealTime PCR System (F. Hoffmann-La Roche, Basel, Switzerland). With this qPCR we

RI PT

were able to detect 10 Fg/uL of B. burgdorferi genomic DNA. In addition, this qPCR was able to detect B. burgdorferi, B. garinii, B. afzelii, B. japonica, B. valaisiana and B. andersoni, and was therefore designated as a B. burgdorferi s.l. catch-all qPCR.

SC

There was no crossreaction with genomic DNA from B. hermsii, B. crocidurae and B.

M AN U

anserina, as well as a whole range of other gram negative bacteria.

Optimal reaction conditions in a final volume of 20ul were LightCycler FastStart TaqMan,

LightCycler®

FastStart

Enzyme,

AAAAATATTTATTGGGAATAGGTCT-3’)

primers,

0,9

0,25

µM µM

µM

forward

(5’-

reverse

(5'-

probe

(6FAM-

TE D

CACCRGGYAMRTCTACTGAA-3’)

and

0,9

5’AGCATGTAAGCAAAATG-3’-MGBNFQ), Uracil-DNA Glycosylase (UNG) (0,2 U/ml) and 5 µl of template DNA. Cycling conditions included an initial activation of

EP

UNG at 50°C for 2 min and a denaturing phase at 95°C for 10 min, followed by 50

AC C

cycles of a 15 s denaturation at 95°C followed by a 60 s annealing-extension step at 60°C (Ramp rate 20°C/s and a single point measurement at 60 °C) and a cooling-cycle of 40°C for 1 min. Analysis was performed using the second derivative calculations for crossing point values. Curves were also assessed visually.

ACCEPTED MANUSCRIPT Borrelia burgdorferi s.l. culture

Skin biopsies, cerebrospinal fluid obtained from lumbar punctures or synovial biopsies were cultured in Modified Kelly’s Medium (MKM) in 10 mL glass tubes at

RI PT

33ºC. Prior to incubation, skin biopsies are disinfected with chlorhexidine in ethanol before added to MKM with 50 µg/mL rifampicin and 100 µg/mL fosfomycin. CSF samples were centrifuged (6000 g for 20 minutes) after which the pellet was added to

SC

MKM. Biopsies from synovial tissue or fluid were added directly to MKM. Cultures

AC C

EP

TE D

M AN U

were checked weekly under dark field microscopy for up to eight weeks.

ACCEPTED MANUSCRIPT Supplementary table 1. Diagnostic work-up at the AMLC

168/200

84 %

66/168

39 %

74/200

37 %

of which positive

28/74

Lumbar punctures

29/522

Borrelia burgdorferi s.l. C6-EIA (IgM/IgG) of which positive

Pleocytosis and positive Borrelia IgG index3

1/29

Only pleocytosis

5/29

RI PT

Immunoblot IgM and/or IgG1

14/314

45 %

4/14

29 %

2/125

17 %

1/2

50 %

4/296

14 %

0/4

0%

12/314

39 %

M AN U

of which positive

Borrelia burgdorferi s.l. PCR Synovial fluid/Synovium of which positive Borrelia burgdorferi s.l. PCR CSF of which positive

TE D

Histopathological examination of skin section

of which supportive of active B. burgdorferi s.l. infection 4/12

EP

Borrelia burgdorferi s.l. culture

3%

17 %

33 %

17/2007

9%

0/17

0%

Mikrogen, interpretation criteria according to manufacturer’s recommendation. For this analysis

AC C

1

56 %

SC

Borrelia burgdorferi s.l. PCR Skin

of which positive

38 %

indeterminate immunoblots (n=12) were considered positive. 2 Number of patients seen at department of Neurology. 3 B. burgdorferi s.l. IgG index = Lyme index CSF / Lyme index serum. 4 Number of patients presenting with skin lesions. Number of lumbar punctures.

7

5

Number of patients seen at department of Rheumatology.

6

Of which 13 from skin biopsies, 2 from CSF and 2 from synovial

tissue. CSF: cerebrospinal fluid. EIA: Enzyme Immuno Assay, PCR: Polymerase Chain Reaction.

ACCEPTED MANUSCRIPT

Supplementary table 2. List of other diagnosis established per specialism Dermatology (n = 8) Allergy/Eczema (n=4)

Subcorneal pustular dermatosis (Sneddon Wilkinson) General Internal Medicine (n = 14) Hypovitaminosis D (n=7) Polymyalgia rheumatica Monoclonal gammopathy of undetermined significance Chronic pain syndrome Malnutrition Hypothyroidism Irritable bowel syndrome Infectious diseases (n = 3) Upper respiratory tract infection (n=2)

TE D

HIV-infection Neurology (n = 8)

Benign paroxysmal positional vertigo Complex regional pain syndrome-I

Parkinson(-like)

AC C

Neuropathy

EP

Stroke Multiple sclerosis

Medication overuse headache Polymyositis

Reumatology (n = 8) Osteoarthritis (n=7) Fibromyalgia

Psychiatry (n = 2) Delusional parasitosis Depression

M AN U

Paraneoplastic syndrome

SC

Pityriasis lichenoides chronica

RI PT

Morphea (n=2)