Session Information
Session Type: ACR Poster Session C
Session Time: 9:00AM-11:00AM
Background/Purpose: Few factors have been consistently linked to antibiotic-refractory Lyme arthritis (ARLA) other than premature intra-articular glucocorticoid injections. We sought to identify clinical and treatment factors associated with ARLA in children.
Methods: We identified children age 18 years and younger with a diagnosis of Lyme disease seen in pediatric rheumatology clinics across three hospital systems who met clinical and Western blot laboratory criteria for Lyme arthritis. ARLA was defined as having persistent, documented arthritis at least 2 months after adequate antibiotic treatment (≥56 days of oral or ≥14 days of parenteral antibiotics) per IDSA/Red Book guidelines. We compared prespecified demographic, disease, and early treatment characteristics of primary interest between children with ARLA and children whose arthritis resolved within 3 months using descriptive statistics and multivariable logistic regression.
Results: There were 46 children with ARLA and 119 children whose arthritis resolved within 3 months. Multiple clinical factors were independently associated with ARLA in univariable analysis (Table 1). These factors included older age; presentation with unilateral knee synovitis or continuous synovitis ≥6 weeks; and marked clinical worsening (e.g., joint capsule rupture, joint recruitment) after antibiotic initiation. Presentations with severe joint pain or with non-knee joints were associated with rapid resolution. There were few instances of premature glucocorticoid joint injections, limiting statistical power to detect a difference between groups. In multivariable analysis, ARLA was associated with age ≥10, clinical presentations with prolonged arthritis, knee-only synovitis, lack of severe features (fever, severe pain, high inflammatory markers), and worsening arthritis while on treatment (Table 2). Results were consistent after adjustment for clinical center and when using an alternate definition of persistent arthritis ≥6 months after antibiotic initiation. In exploratory analyses of variables that were not of primary interest, antibiotic regimens below the recommended dose were associated with ARLA (P = 0.03), but low-frequency regimens (e.g., twice daily amoxicillin) were not (P = 0.48).
Conclusion: Pediatric antibiotic-refractory Lyme arthritis is associated with multiple clinical factors, including older age, prolonged knee synovitis at diagnosis, and clinical worsening on antibiotics. In contrast, younger children presenting with non-knee synovitis, severe pain, and a robust inflammatory response often respond quickly to antibiotics.
| Table 1. Clinical and treatment characteristics of primary interest | |||
| Characteristic |
Arthritis resolved within 3 months of antibiotic initiation (N = 114) |
Antibiotic-refractory Lyme arthritis (N = 46) |
P-value |
| Demographics |
|
|
|
| Age in years, median (IQR) |
9.6 (7.0, 11.9) |
12.0 (9.1, 14.7) |
<0.01 |
| Male sex, N (%) |
82 (69) |
28 (61) |
0.33 |
| Clinical presentation |
|
|
|
| Duration of current joint symptoms in days, median (IQR) |
5 (2, 14) |
22 (7, 60) |
<0.01 |
| Prior self-resolving episodes of joint pain/swelling, N (%) |
41 (34) |
11 (24) |
0.19 |
| Acute migratory arthritis, N (%) |
4 (3) |
0 |
0.21 |
| ≥6 weeks of continuous joint symptoms, N (%) |
3 (3) |
9 (20) |
<0.01 |
| Fever within 2 weeks of diagnosis not from another cause, N (%) |
38 (30) |
8 (22) |
0.34 |
| Severe pain preventing mobility or requiring hospitalization, N (%) |
28 (24) |
0 |
<0.01 |
| More than 1 joint involved,1 N (%) |
29 (24) |
2 (4) |
<0.01 |
| Non-knee joint involved, N (%) |
23 (19) |
1 (2) |
<0.01 |
| Laboratory values |
|
|
|
| Number of Western blot IgG bands (N), median (IQR) |
10 (8, 10) |
9 (9, 10) |
0.21 |
| Maximum erythrocyte sedimentation rate (mm/hr), median (IQR) |
35 (19, 49) |
17 (8, 34) |
<0.01 |
| Treatment and clinical course |
|
|
|
| Glucocorticoid joint injection before first antibiotic course, N (%) |
2 (2) |
2 (4) |
0.32 |
| Marked worsening within 6 weeks of antibiotic initiation,2 N (%) |
6 (5) |
10 (22) |
<0.01 |
| Characteristics of spondyloarthritis within 6 weeks of antibiotic initiation,3 N (%) |
2 (2) |
1 (2) |
0.83 |
| Exam or imaging with chronic changes within 6 weeks of antibiotic initiation,4 N (%) |
18 (15) |
4 (9) |
0.28 |
| IQR, interquartile range. 1Two knees would count as two joints 2New massive effusion, joint capsule rupture, or joint recruitment 3Personal history of psoriasis, inflammatory bowel disease, acute anterior uveitis, or inflammatory back pain; presence of enthesitis, tendonitis, or dactylitis on exam 4 Flexion contracture, muscle atrophy, condylar hypertrophy, erosions on imaging | |||
| Table 2. Multivariable model of clinical factors associated with antibiotic-refractory Lyme arthritis or prolonged arthritis | ||||
|
Antibiotic-refractory Lyme arthritis |
Arthritis lasting at least 6 months |
|||
| Characteristic |
Adjusted odds ratio (95% CI) |
P-value |
Adjusted odds ratio (95% CI) |
P-value |
| Age ≥10 years |
2.7 (1.1, 6.5) |
0.03 |
2.2 (1.01, 4.9) |
0.05 |
| Continuous joint symptoms for ≥6 weeks at presentation |
19.0 (3.6, 99.4) |
<0.01 |
10.2 (2.3, 45.1) |
<0.01 |
| Presenting in knee(s) only |
10.1 (1.9, 54.0) |
<0.01 |
4.3 (1.3, 14.9) |
0.02 |
| Severity features present1 |
0.3 (0.1, 0.8) |
0.02 |
0.5 (0.2, 1.1) |
0.07 |
| Worsening within 6 weeks of antibiotic initiation2 |
5.3 (1.5, 19.2) |
0.01 |
3.5 (1.03, 12.1) |
0.04 |
| CI, confidence interval. 1Recent, otherwise unexplained fever, severe pain limiting mobility, hospitalization for pain, or erythrocyte sedimentation rate ≥40 mm/hour 2 New massive effusion, joint capsule rupture, or joint recruitment within 6 weeks of antibiotic initiation | ||||
To cite this abstract in AMA style:
Horton DB, Taxter AJ, Groh B, Sherry DD, Rosé CD. Clinical and Treatment Factors Associated with Antibiotic-Refractory Lyme Arthritis in Children [abstract]. Arthritis Rheumatol. 2016; 68 (suppl 10). https://acrabstracts.org/abstract/clinical-and-treatment-factors-associated-with-antibiotic-refractory-lyme-arthritis-in-children/. Accessed .« Back to 2016 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/clinical-and-treatment-factors-associated-with-antibiotic-refractory-lyme-arthritis-in-children/