Clinical, Laboratory, Imaging and Electrocardiographic Differences between Patients with Lyme Disease and Patients with Lyme Disease and B. divergens Antibodies

1. Introduction

In recent years, tick-borne diseases (TBDs) seem to have increased substantially worldwide [1,2]. Several tick-borne bacterial, viral, and protozoan pathogens cause infection in humans and animals. Among these infections are Lyme disease or borreliosis, but also other less common diseases such as babesiosis, the prevention and treatment of which is hampered by suboptimal diagnostics [1,2,3].

Asturias (Northwestern Spain) has a mild humid weather and it is covered by 210 Kha of natural forest, extending over 61% of its land area. In these forested areas, ticks and wild roe and fallow deer abound, especially in its most mountainous east and southwest parts. In these areas Ixodes ricinus carriers of Borrelia burgdorferi sensu lato (s.l). have been frequently observed [4,5]. These circumstances explain the high seroprevalence of Lyme disease in Asturias, higher than in other regions of Spain, reaching to 5.1% in healthy blood donors of this region [6]. Human babesiosis has also been reported in Asturias [7,8,9]

Although it affects primarily cattle in Europe, Babesia divergens, a parasite also transmitted by I. ricinus, can infect humans as well causing babesiosis. B. divergens-like DNA sequences have also been reported in red deer (Cervus elaphus), roe deer (Capreolus capreolus), reindeer (Rangifer tarandus) and fallow deer [10,11] However, there is no reported evidence that Babesia isolates from deer are infective to humans or cattle so far [12].

Coinfection with B. burgdorferi and Babesia microti is relatively common in northwestern and upper midwestewern United States. Approximately 10% of patients with Lyme disease in southern New England are co-infected with babesiosis in locations where both diseases are zoonotic [13]. In Europe B. microti and B. divergens positive antibodies were observed in 16.3% of Swedish patients seropositive for B. burgdorferi [14].Very recently an Irish longitudinal TBDs study, reported 30% of patients with positive antibody responses to tick-borne pathogens (TBP) including B. burgdorferi species and B. microti. These patients were treated with combination antibiotics that effectively relieved TBI symptoms with good patient tolerance [15].

Since 2011 two cases of severe infection by B. divergens have been reported in Asturias, probably the iceberg “tip” of an undetected infected population [8,9]. Of note, we have recently detected a B. divergens seroprevalence rate of 39.2% in 120 patients with Lyme disease in a retrospective study in Asturias (2015-2017) confirming that persons with Babesia-positive antibodies exceed the number of clinically diagnosed human babesiosis cases, so far. This result also warned of possible consecutive or simultaneous undetected co-infections in B. burgdorferi s.l.-infected patients [16].

The number of symptoms and duration of illness in American patients with concurrent Lyme disease and babesiosis due to B. microti are higher than in patients with either infection alone [13,17,18,19].

Illness characteristics of patients infected with both B. burgdorferi s.l. and B. divergens pathogens have not been reported so far. Diagnostic tools on when to suspect and how to manage babesiosis in patients with Lyme disease are lacking as well as how to distinguish co-infections from mono-infections or uninfected patients which could affect the course and severity of the disease [15,20].

To shed more light on these questions, we designed a retrospective study to compare clinical, laboratory and other studies ̶ imaging and electrocardiographic (ECG) ̶ characteristics of the aforementioned Asturian adult patients with previously confirmed Lyme disease and 39.2 % of them with confirmed IgG antibodies against B. divergens [16].

2. Material and Methods

2.1. Patients and Study Design

This retrospective study included 120 patients ≥18 years old residing in Asturias (Northwestern Spain) with an IgG positive serology for Lyme disease from 2015 through 2017 assessed at the Microbiology Service of the Hospital Universitario Central de Asturias (HUCA), Oviedo, Spain.

Detection of B. burgdorferi s.l. antibodies was done using an automated qualitative test (Vidas, BioMerieux, Madrid, Spain) and by immunoblot (Borrelia IgG IgM EcoLine, Sekisui Diagnostics, Russelsheim, Germany) [6,16]. Patients’ serum samples positive for Lyme disease were stored at -80ºC at the Microbiology Service of the HUCA for further studies. These patients had been followed at the Infectious Diseases, Neurology, Rheumatology and/or Dermatology Services of the HUCA or other regional affiliated hospitals because B. burgdorferi serology assessment is centralized at the HUCA at regional level. Patients’ electronic medical records were searched and their demographic, epidemiological, including Lyme disease-predisposing factors, clinical, laboratory, imaging, ECG and other data were collected.

Early localized Lyme disease included patients with erythema migrans and positive B. burgdorferi s.l. IgG and/or IgM serology. This stage occurred within 1-28 days following the tick bite. Early disseminated Lyme disease was made of patients with multiple erythema migrans, early neuroborreliosis, and carditis. This stage developed 3-12 weeks after the initial infection. Later Lyme disease included patients with arthritis, acrodermatitis chronica atrophicans and late neuroborreliosis. This stage developed months or years after the initial infection. Those with confirmed Lyme arthritis had asymmetrical, monoarticular or oligoarticular arthritis, a synovial fluid with 10,000-25,000 cell/mm³ and a positive anti-Borrelia IgG antibody detection and positive Borrelia IgG and/or IgM serology. Other patients with rheumatological symptoms (arthralgias, myalgias) were also included as rheumatological Lyme if they had positive Borrelia IgG and/or IgM serology and no other cause to explain their symptoms. Patients with Lyme carditis had mostly syncope, chest pain or dysnea, positive Borrelia IgG and/or IgM serology and no other cause to explain their heart symptoms. Patients with confirmed neuroborreliosis had compatible clinical symptoms and/or signs (mostly limbs paresthesia/paresia, gait disturbance, cranial neuritis and headache), cerebrospinal fluid (CSF) pleocytosis and CSF positive Borrelia IgG serology. Late Lyme disease occurred months or years after the initial infection. The typical symptoms consisted of neurological and rheumatological involvement. Past Borrelia infection included asymptomatic patients with positive Borrelia IgG serology [21,22]. In a subsequent study, stored serum samples seropositive for B. burgdorferi s.l. were used again to detect IgG antibodies against B. divergens by an in-house indirect fluorescent assay (IFA) and by Western Blot at the Centro Nacional de Microbiología, Instituto de Salud Carlos III, Majadahonda, Madrid, Spain [16].

2.2. Statistical analysis

The continuous variables did not follow a Gaussian distribution, according to the Kolmogorov-Smirnov test and, therefore, non-parametric tests were used for analysis. The values are reported as median (IQ range) or as percentage for continuous or categorical variables, respectively.

The group of patients infected with B. burgdorferi s.l. who also had B. divergens IgG antibodies and the group of patients only infected with B. burgdorferi s.l. were compared with the Mann-Whitney U test for continuous and the chi-square test and the Fisher’s exact text, when appropriate, for categorical variables. A stepwise logistic regression analysis model was constructed to identify the factors independently associated with the diagnosis of B. burgdorferi s.l. plus positive B. divergens IgG antibodies. A P value <0.05 for a two-tailed test was considered statistically significant. Calculations were carried out with the statistical software SPSS v. 25 (IBM Corp., Armonk, New York, USA).

3. Results

A total of 120 patients, 73 monoinfected with B. burgdorferi s.l. and 47 infected with B. burgdorferi who also had B. divergens IgG antibodies were included in the study [16]. The median age of the series was 57.5 years (IQ range 43.8-72.5) and 80 patients (66.7%) were men.

Table 1 shows the demographic, epidemiologic and Lyme disease predisposing factors of B. burgdorferi s.l monoinfected patients and patients infected with B. burgdorferi s.l. who also had B. divergens IgG antibodies. It can be appreciated that 65.8% developed tick-bites risky outdoor activities, mostly as a hobby, 28.3% had non-professional animal contact and only 10.8% were farmers or cattle breeders. Interestingly only 38.3 % of the Lyme disease patients recalled a recent tick bite. There were no patients with asplenia and only 4 (0.03%) had a known immunodepression. No differences between both groups of infected patients regarding demographic, epidemiologic and Lyme disease predisposing factors were observed.

The clinical features and complications of the patients are reported in Table 2. Overall, 48.3% had neurological symptoms, raising to 53.2% in B. burgdorferi s.l. infected patients and with B. divergens IgG antibodies, 14/120 (11.7%) had unilateral or bilateral facial paralysis and 25.8% cutaneous symptoms with erythema migrans reported in 22/120 (18.3%). On the other hand 32.5% had osteomuscular symptoms, mostly arthralgias (27.5%), and 20.8% had constitutional symptoms, mostly fever (11.7%). There were no differences between both groups of infected patients regarding these clinical manifestations. On the other hand 13/120 (10.8%) had cardiorespiratory symptoms, with dyspnea as the most frequent (5/120, 4.12%). Of note, B. burgdorferi s.l. infected patients and with B. divergens IgG antibodies had more frequently cardiorespiratory symptoms compared to B. burgdorferi s.l. monoinfected patients (9/47 [19.1%] vs. 4/73 [5.65%], P=0.02) especially dyspnea (4/47 [8.5%] vs. 1/73 [1.4%], P=0.07). No differences in the physical exam between both groups of infected individuals were detected. No differences regarding type of Lyme disease, its complications or presence of an alternative diagnosis between both groups of infected patients were observed.

Table 3 depicts the laboratory, radiologic imaging and EGC studies of the patients. Anemia, or increased levels of lactate dehydrogenase (LDH) or bilirrubin, parameters associated with hemolytic anemia, were not observed neither in both group of patients. Normal levels of C-reactive protein (CRP) or erythrocyte sedimentation rate (ESR) were observed in both groups. Atrioventricular (AV) block was detected in 5/47 (15.6%) patients with Lyme disease and B. divergens IgG antibodies compared to 1/73 (2.7%) monoinfected individuals (P=0.09). No other imaging or laboratory studies differences were observed between both groups of patients.

The microbiological diagnostic procedures are described in Table 4. Positive B. burgdorferi IgG serology tests was observed in monoinfected and in those with B. divergens IgG by definition. In addition B. burgdorferi IgM serology was positive in > 55% of all the patients included in the study. Positive antibodies against B. divergens assessed by indirect immunofluorescence assay (IFI) were identified in 47 serum samples from patients, and 40 serum of the total were confirmed by Western-blot (40/47-85.1%) using B. divergens extract proteins. Moreover, IgG antibodies against the major surface antigen (Bd37) of B. divergens were found in 19 /47 (47.5%) of the samples [16,23,24].

Table 5 shows the treatment administered to the patients. The most commonly administered antibiotic was doxycycline and there were no significant differences between both groups of patients in any of the regimens administered.

A forward stepwise logistic regression analysis model, including all variables with a P value ≤0.1 in the univariate analyses, revealed that the only significantly independent predictive factor for B. burgdorferi s.l.- infection of patients with B. divergens IgG antibodies was the presence of cardiorespiratory symptoms (OR 4.184, 95% CI 1.205-14.493, P=0.024).

4. Discussion

Borrelia burgdorferi s.l. infected patients who also had B. divergens IgG antibodies had more frequently cardiorespiratory symptoms, mostly dyspnea, compared to B. burgdorferi s.l. monoinfected individuals. These cardiorespiratory symptoms were unrelated to anemia that both groups of patients did not have, and might be enhanced by ECG AV block. This cardiac arrhythmia could be caused by summative myocardial damage induced by both pathogens.

Babesiosis due to B. divergens is the most severe among the infections due to all Babesia species in humans with a quite high mortality, up to 40% in hospitalized patients, although nowadays it is decreasing [2]. Other Babesia species such as B. microti have a lower mortality rate reaching 10-20% in hospitalized individuals [25,26]. Mortality attributable to Babesia is not due exclusively to the pathogen or the host but also to the interaction between both of them. This interaction explains the severity of some cases, two of them reported from Asturias, one in a young immunocompetent patient, and the lower severity of others such as those reported in the present study [8,9].

Previous reports showed that experimental coinfection by B. burgdorferi s.l. and B. microti induced increased arthritis severity in mice which correlated with a significant reduction of expression of the IL-10 and IL-13 cytokines [27]. B. burgdorferi-B. microti coinfection in humans increased the intensity and duration of the illness and its symptoms in American patients [13,17]. Although the number and duration of babesial symptoms were similar in older and younger of 50 years infected by B. microti and babesial-Lyme disease coinfected subjects, more older adults were admitted to the hospital than younger in other American study cohort [19]. Fatigue, but also headache, nausea, sweating, anorexia, chills, emotional lability and splenomegaly were more frequently reported in B. burgdorferi-B. microti coinfected individuals with symptoms’ duration lasting up to three months compared to those individuals monoinfected with B. burgdorferi s.l. No cardiac abnormalities have been described in these coinfected American patients, though [13].

To our knowledge this is the first report studying clinical differences, laboratory , imaging and ECG of B. burgdorferi s.l.- B. divergens-infection in the literature and the second in which B. divergens antibodies are detected among B. burgdorferi-infected patients [14]. A recent French meta-analysis found eight reports of B. burgdorferi s.l-Babesia spp. coinfection in the literature but all of them were due to B. microti [28].

Overall, 53.2% of patients with B. burgdorferi s.l. and with IgG antibodies against B. divergens had minor neurologic symptoms, mostly facial paralysis and muscle weakness. These symptoms seem more related to B. burgdorferi s.l. than to B. divergens infection. Neurologic symptoms of babesiosis are headache, confusion/delirium, impaired consciousness, ataxia/gait disorder and vision impairment not described by patients with B. burgdorferi s.l. and with IgG antibodies against B. divergens in our study, except for headache that was reported in 13.3 % of the patients (Table 2) [30]. Only 22.2% of the patients reported constitutional complains (fever, asthenia, anorexia, weight loss). In addition their clinical course was mild and short unlike the complicated clinical course of those with B. burgdorferi s.l.-B. microti coinfections [17,19]. A possible explanation for this discordance might be differences in the susceptibility of individual patients to infection.

It is intriguing that patients infected with B. burgdorferi s.l. and with IgG antibodies against B. divergens had more frequently cardiorespiratory symptoms, mostly dyspnea compared to B. burgdorferi s.l. monoinfected patients. These were the only significant differential clinical symptoms between both groups of patients. One possible explanation for the cardiac dysfunction in the group with Lyme disease and B. divergens IgG antibodies might be a summative potential myocardial damage of both microorganisms manifested by ECG AV block present in five individuals. Both B. burgdorferi s.l. and B. divergens might affect AV conduction leading to AV block. B. burgdorferi AV block is a well-known complication of Lyme disease [31]. Babesia bigemina infection induced changes in cardiac function biomarkers and D-dimer in cattle [32]. Babesia canis infection induced cardiac disorders in dogs with 32% of AV block [33]. Two reports showed myocardial damage and serious arrhtymia associated with severe babesiosis in two American patients [34,35]. A recent report showed that 19.6% of hospitalized American patients with acute babesiosis due to B. microti developed cardiac complications, mostly atrial fibrillation, heart failure, QT interval prolongation and cardiac ischemia [36]. No AV block I in that study was observed, though. All the patients had high parasitemia and had received antimicrobials (macrolides, quinine) which might have enhanced their heart arrythmia. Interestingly cardiac complications were not worse in 12 of them with confirmed B. burgdorferi s.l.-B. microti coinfection.

We assume that B. divergens might induce a myocardial damage similar to B. microti. Unluckily, neither troponin, creatin kinase (CK), natriuretic peptides (B-type natriuretric peptide [BNT] or N-terminal pro-B-type natriuretic peptide ([NT-proBNT]), other cardiac function biomarkers or echocardiograms were done in our study being retrospective. In addition ECG was performed in only 58.4% of our patients.

Thus, the mean weakness of the study is derived of its retrospective design lacking universal ECG, cardiac biomarkers and echocardiograms testing that might provide deeper insight into the mechanism of cardiac complications in the patients infected with Lyme disease and B. divergens IgG antibodies. Other weakness of the study is the uncertainty of the timing, simultaneous or sequential, of both infections, by B. burgdoferi s.l. and B. divergens. Patients could be firstly infected by B. divergens to become infected by B. burgdoferi subsequently. If this be the case we might be comparing virtually two groups with a B. burgdoferi s.l. monoinfection. The fact that even so clinical differences between both groups of infected patients were observed suggests that disparity might be higher in cases of a confirmed simultaneous coinfection.

In summary patients with Lyme disease and B. divergens IgG antibodies had a mild clinical course, and did not develop anemia or other analytical abnormalities. Their only, but significant differential symptoms were cardiorespiratory manifested as dyspnea and ECG AV block.

5. Conclusions

Patients with suspected Lyme disease could be screened for babesiosis as well to obtain relevant information on epidemiology and prevalence of both concurrent zoonosis in endemic areas. Detection of sequential and simultaneous infections might also provide useful information to the physician to monitor and treat, with the most effective antimicrobial regimen, patients living in these areas. This concern is especially important in those immunocompromised in whom babesiosis has a more severe evolution and could lead even to a fatal outcome.