Suspicious index in Lyme carditis: Systematic review and proposed new risk score

Abstract

Background

Lyme carditis (LC), an early manifestation of Lyme disease that most commonly presents as high‐degree atrioventricular block (AVB), usually resolves with antibiotic treatment. When LC is not identified as the cause of AVB, a permanent pacemaker may be inappropriately implanted in a reversible cardiac conduction disorder.

Hypothesis

The likelihood that a patient's high‐degree AVB is caused by LC can be evaluated by clinical characteristics incorporated into a risk stratification tool.

Methods

A systematic review of all published cases of LC with high‐degree AVB, and five cases from the authors' experience, was conducted. The results informed the development of a new risk stratification tool, the Suspicious Index in LC (SILC) score. The SILC score was then applied to each case included in the review.

Results

Of the 88 cases included, 51 (58%) were high‐risk, 31 (35.2%) intermediate‐risk, and 6 (6.8%) low‐risk for LC according to the SILC score (sensitivity 93.2%). For the subset of 32 cases that reported on all SILC variables, 24 (75%) cases were classified as high‐risk, 8 (25%) intermediate‐risk, and 0 low‐risk (sensitivity 100%). Specificity could not be assessed (no control group). Notably, 6 of the 11 patients who received permanent pacemakers had reversal of AVB with antibiotic treatment.

Conclusion

The SILC risk score and COSTAR mnemonic (constitutional symptoms; outdoor activity; sex = male; tick bite; age < 50; rash = erythema migrans) may help to identify LC in patients presenting with high‐degree AVB, and ultimately, minimize the implantation of unnecessary permanent pacemakers.

1. INTRODUCTION

Lyme disease (LD) is a tick‐borne infection primarily caused by Borrelia burgdorferi bacteria.1 LD is a clinical diagnosis, which can be confirmed by serological testing.2 Cardiac complications occur in early disseminated LD, weeks to months after bacterial transmission.3 Up to 10% of LD involves the heart, with 90% of Lyme carditis (LC) causing atrioventricular block (AVB).4 Symptoms of LC include palpitations, dyspnea, dizziness, pre‐syncope, and syncope. Fatal cases of LC have been reported. The standard treatment for high‐degree AVB is pacing; however, the AVB in LC usually resolves with antibiotics, and unstable bradycardia can be supported by temporary pacing.5 Permanent pacemaker implantation is rarely needed.6

This manuscript proposes a risk stratification tool: Suspicious Index in LCs (SILC). SILC is a new (non‐validated) scoring system that evaluates the likelihood that a patient's high‐degree AVB is caused by LC.

2. METHODS

A database search was performed in Embase and Medline on October 3, 2017 (search terms: (“Lyme” OR “Lyme disease”) AND (“carditis,” “myocarditis,” “heart block,” “heart muscle conduction disturbance,” “heart conduction system,” “sick sinus syndrome,” “heart arrest,” “conduction,” “AV block,” “atrioventricular block,” “asystole,” “sinus pause,” OR “bundle branch block”)). Duplicate articles and reports on animal cases were removed. The article titles were screened for cases of LC or conduction disorders. The remaining articles were assessed for eligibility based on: (i) English language in the abstract and/or full text; (ii) case report or series; and (iii) high‐degree AVB reported. Articles that satisfied the aforementioned criteria were included in the study.

The following data on patient demographics, clinical presentation, disease progression, and treatment response were collected when available: (a) age, (b) gender, (c) history and timing of tick bite, (d) outdoor activities (previous 30 days), (e) visit to an endemic area (previous 30‐45 days), (f) presence of erythema migrans, (g) serological confirmation, (h) time since onset of LD symptoms, (i) presenting symptoms, (j) type of AVB, (k) high‐degree AVB resolution and timing, (l) antibiotic treatment and timing of response, (m) need for temporary/permanent pacemaker, (n) any other treatment, (o) administration of a stress test or electrophysiological study,and (p) follow up timing.

The SILC score is presented in Table 1.6 The rationales for the selected variables and point allocations in the SILC score are detailed in Section 3 and are based on the prevalence of indices from the systematic review of all published cases. The total summed score indicates low (0‐2), intermediate, (3–6) or high (7–12) suspicion of LC.

Table 1.

The suspicious index in Lyme Carditis score evaluates the likelihood that a patient's high‐degree atrioventricular block is caused by Lyme carditis

Variable	Value
Constitutional symptomsa	2
Outdoor activity/endemic area	1
Sex = male	1
Tick bite	3
Age < 50	1
Rash = erythema migrans	4

The total summed score indicates low (0‐2), intermediate (3‐6), or high (7‐12) suspicion of Lyme carditis.

^aFever, malaise, arthralgia, and dyspnea.

SILC was then retrospectively applied to evaluate whether the variables included in the model help to determine if a high‐degree AVB was caused by LC. SILC variables not reported in the cases were conservatively deemed to have a value of zero. Descriptive statistical analysis was first done for the whole cohort, and then repeated exclusively for cases that reported on all the SILC variables.

3. RESULTS

After duplicates were removed from the 1433 citations identified in our search, 922 studies remained. A further 814 articles were filtered down to 108 clinically relevant cases (as described in the methods). Another 34 articles were removed for: not in English,12 not a case report/series,4 or not reporting high‐degree AVB.18 The final 74 articles reported on 88 cases relevant to this study.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79(Figure 1).

Figure 1.

Flow of studies through the review process

3.1. Rationale and construction of the SILC risk stratification tool

1. Mean age was 35.8 ± 13.5 years. As LC is an under‐recognized cause of conduction abnormalities in this generally young, otherwise healthy population, one point was allocated in the SILC score for age < 50 years.

2. Seventy‐five percent of the patients were male, which is consistent with previous literature noting a strong 3:1 male predominance of LC, in the context of approximately equal prevalence of LD by sex. Therefore, one point was allocated for male sex.

3. Over one‐third (39.8%) of patients reported spending time in an endemic area, and 38.6% reported participating in outdoor activities within 30 to 45 days before presentation. Given pathophysiology of LD transmission from rodent host reservoirs to humans through the Ixodes tick, one point was given for exposure to endemic region and/or outdoor activity.

4. Constitutional symptoms, such as fever (28.4%) and fatigue or malaise (39.8%) were commonly reported. To differentiate LC from other etiologies of high‐degree AVB, two points were given for the presence of any characteristic symptoms of LD: fever, malaise, arthralgia, or dyspnea.

5. A tick bite was recognized in 30.7% of cases. As ticks often land in intertriginous areas, most patients with LC have not identified a tick bite. Since a recalled history of a tick bite is logically highly sensitive for LC, three points were allotted for this variable.

6. Although the pathognomonic erythema migrans rash was only reported in 50% of cases, its presence almost certainly indicates LD. Therefore, four points were given for erythema migrans.

3.2. Clinical course of LC cases

Patients presented to the hospital at a median of 3.0 (1.6‐4.1) weeks after the onset of symptoms attributable to LD. If recalled, the tick bite occurred a median of 4.5 (3.0‐6.5) weeks before presentation. Conduction disturbances reported were: third‐degree AVB (77.3%), second‐degree AVB (33.0%), and sinus pauses or asystole (12.5%). LD infection was confirmed serologically in 96.6% of cases.

Antibiotics were used in 93.2% of cases (61% ceftriaxone, 24.4% doxycycline, and 31.7% other). A temporary pacemaker was inserted in 36.4% of cases. A permanent pacemaker was implanted in 12.5% of cases. Notably, more than half of the 11 patients who received permanent pacemakers also had reversal of AVB with antibiotic treatment. High‐degree AVB resolved in 94.3% of cases, with a median time to resolution of 5 (3‐9) days.

Tests to evaluate AV conduction prior to discharge were used in 20.5% of cases (66.7% exercise stress test, 38.9% electrophysiological study). Follow‐up was reported in 61.4% of cases (7.8 [4.3‐24.7] weeks after discharge).

3.3. SILC score efficacy

SILC scores were calculated for each case. Of the 88 cases, 58% had a high SILC score (scores 7‐12), 35.2% had an intermediate SILC score (scores 3‐6), and 6.8% had a low SILC score (scores 0‐2). The sensitivity of the SILC risk stratification tool was 93.2%, considering that the population was previously identified as “having” the disease.

When the analysis was repeated on the 32 cases that reported on all the SILC variables, the sensitivity increased to 100%: 75% had a high SILC score, 25% had an intermediate SILC score, and none had a low SILC score (sensitivity 100%).

Specificity could not be assessed (no control group).

4. DISCUSSION

4.1. Management of Lyme carditis

In the absence of formal guidelines for the management of LC, we propose the following algorithm for the diagnosis and management of high‐degree AVB because of suspected LC.

Patients presenting with high‐degree AVB should have their SILC score evaluated. Patients with low (0‐2) SILC scores should proceed to standard treatment for high‐degree AVB (consideration of pacemaker implantation). Serological confirmation of Lyme infection (usually using enzyme‐linked immunosorbent assay and Western blot techniques) can confirm the diagnosis of LC.2 We suggest that a patient who presents with AVB and has an intermediate3, 4, 5, 6 to high7, 8, 9, 10, 11, 12 SILC score should be tested for serological evidence of LD and treated with antibiotics (±temporary‐permanent pacing for symptomatic bradycardia or strict cardiac monitoring for asymptomatic bradycardia).

Patients with serologically‐confirmed LD should continue the course of intravenous antibiotics for 10 to 14 days, followed by oral antibiotics for 4 to 6 weeks. If 1:1 conduction is restored, temporary pacing can be discontinued. Prior to discharge, and at a minimum of 10 days from admission, evaluation of AV conduction with a treadmill test is recommended for the assessment of conduction stability prior to discharge. If 1:1 conduction fails at <90 beats per minute (bpm), then a permanent pacemaker may be recommended. If 1:1 conduction is maintained >120 bpm, then the patient can be followed with an outpatient electrocardiogram in 4 to 6 weeks to confirm lack of rhythm or conduction abnormalities. If 1:1 conduction fails between 90 and 120 bpm, then a repeat stress test in 4 to 6 weeks may be considered before deciding on permanent pacing.80

4.2. Utility of the SILC score

Although LC may resolve spontaneously, prompt antibiotics can reduce complications.3 A recent study by Wan et al. reported on five cases of confirmed LC in which early treatment with antibiotics reversed high‐degree AVB in all cases.6 Only two patients needed temporary pacing, which was removed by day 5. Notably, several visits to the emergency department occurred before the correct diagnosis was established. All patients were asymptomatic and had normal electrocardiograms at 4 to 6 weeks follow‐up.6

Permanent pacemaker implantation is rarely necessary in LC management. However, in this review, 11 cases (12.5%) were implanted with permanent pacemakers in addition to antibiotics. In six of these permanent pacemaker cases, the underlying AVB resolved after a course of antibiotics, and the pacemaker was removed7, 8, 9, 10 or no longer required.11, 12 In these cases, the pacemakers were implanted before the LC diagnosis was serologically confirmed.7, 8, 9, 10, 11, 12 In the five remaining cases requiring permanent pacemaker implantation, the underlying heart block did not resolve despite antibiotic treatment.8, 13, 14, 15 In 4 of these 5 cases, the pacemaker was placed after a course of antibiotics had not resolved the heart block.10, 13, 14, 15 In one case, the pacemaker was placed before LC was diagnosed.8 Permanent pacemakers are rarely required, and should be limited to when AVB does not resolve after antibiotic treatment.

The SILC score and its mnemonic COSTAR may help physicians consider LC on the differential diagnosis when treating patients with high‐degree AVB, particularly for patients living in endemic regions. Unsurprisingly, SILC successfully identified 93.2% of the total previously published cases as intermediate or high risk of LC. Of all 88 cases, six scored low on the SILC score.14, 15, 16, 17, 18, 19 All six low‐scoring SILC cases lacked data on at least one of the SILC variables, and five of these lacked data on 3 or more SILC variables.14, 15, 16, 17, 19 If all relevant data were available, the low‐scoring SILC cases may have been categorized as higher‐risk on the SILC scale. This proposed score has not been prospectively validated yet; therefore, before concluding on its value, it would be reasonable to apply SILC to a prospective series.

4.3. Limitations

The retrospective nature of the study limited the amount of data available for collection. Many of the cases included in the study did not report on all the relevant categories for evaluation. Some of the included studies had only the abstracts in English, which may have led to the exclusion of more details from the full text not available in the abstract. Only approximately one‐third of the cases reported on all the SILC variables. Furthermore, because of the nature of a retrospective systematic review, the study only included serologically‐confirmed LC cases. Determining sensitivity in a sub‐selected diseased population may be questionable validity; but given the exploratory nature of the study, the authors have decided to include the results to confirm that the variables and allocation points used in the model help in capturing the vast majority of the confirmed cases. We were not able to assess cases of AVB that present with similar symptoms but have an etiology other than LC. Without a control group, the specificity of the SILC risk score could not be determined.

Due to inherent limitations in the retrospective application of a proposed risk score, future directions include a multi‐center prospective study to validate the SILC score and ultimately, test the hypothesis that an index to detect LC as a cause of AVB prevents unnecessary pacemaker implantations.

5. CONCLUSIONS

LC most often manifests as cardiac conduction disease and high‐degree AVB. The SILC risk score (and COSTAR mnemonic) may help to identify LC in patients presenting with high‐degree AVB. This condition should be treated with antibiotics, thereby preventing the unnecessary implantation of permanent pacemakers.

CONFLICTS OF INTEREST

The authors declare no potential conflict of interests.

Besant G, Wan D, Yeung C, et al. Suspicious index in Lyme carditis: Systematic review and proposed new risk score. Clin Cardiol. 2018;41:1611–1616. 10.1002/clc.23102