Lyme disease is the most common tick-borne illness in the United States. It is particularly endemic to the Northeast (Figure). The number of cases has been steadily rising since 1995.1 It is caused by the spirochete Borrelia burgdorferi and is transmitted by the deer tick, Ixodes scapularis. As many as 80% of patients with untreated Lyme disease will experience joint symptoms.2,3 Owing to their resemblance in clinical presentation, Lyme arthritis and septic arthritis are often difficult to differentiate. However, their treatment is different. Lyme arthritis is treated exclusively with antibiotics, while acute bacterial septic arthritis is treated with surgical intervention in addition to antibiotics. Ninety-two percent of patients who present with Lyme arthritis are younger than 20 years.4 The knee is the most common joint affected in Lyme arthritis, involving up to 90% of cases. The elbow is rarely the first joint to present with symptoms.5,6 Therefore, Lyme disease is not commonly included in the differential diagnosis for children presenting with isolated elbow pain. The authors report 4 cases of Lyme arthritis presenting in the elbow (Table).
A 14-year-old healthy boy presented with a 7-day history of right elbow pain. He noted swelling and stiffness. He did not recall any trauma, skin rashes, or tick bites. He was afebrile, and his vital signs were normal. On examination, he had an effusion and full range of motion with tenderness. Radiographs showed a joint effusion. Lyme titers yielded positive results after 4 days. The patient was successfully treated with doxycycline. No surgery was performed.
A 4-year-old healthy girl presented with a 1-day history of right elbow pain and swelling. The patient denied any history of trauma, rashes, or tick bites. She was afebrile, and her vital signs were normal. On examination, she had edema and erythema. Her elbow range of motion was 30° to 90°, and she had tenderness. Radiographs showed a joint effusion. The patient's white blood cell count was 15.12 cells/µL, C-reactive protein was 1.19 mg/L, and ESR was 32 mm/h. Aspiration revealed 91,121 cells/mm3 with 96% neutrophils and a negative result on Gram staining. Lyme titers yielded positive results 6 days later. The patient was successfully treated with amoxicillin. No surgery was performed.
A 5-year-old healthy boy presented with a 1-day history of right elbow pain. He noted swelling and stiffness. He did not recall any trauma, skin rashes, or tick bites. He was afebrile, and his vital signs were normal. On examination, he had an effusion, limited range of motion in all directions secondary to pain, and tenderness to palpation. Radiographs showed a joint effusion. The patient's white blood cell count was 5.87 cells/µL and ESR was 29 mm/h. Lyme titers yielded positive results after 4 days. The patient was successfully treated with amoxicillin. No surgery was performed.
A 6-year-old healthy boy presented with a 2-day history of left elbow pain and swelling. The patient denied any history of trauma, rashes, or tick bites. He was afebrile, and his vital signs were normal. On examination, he had an effusion. Elbow range of motion was 30° to 115°, and he had tenderness. Radiographs showed a joint effusion. The patient's white blood cell count was 9.7 cells/µL, C-reactive protein was 2.96 mg/L, and ESR was 47 mm/h. Aspiration revealed 83,748 cells/mm3 with 85% neutrophils and a negative result on Gram staining. Owing to the concern of a septic joint, irrigation and debridement were performed. Cultures yielded negative findings. Lyme titer yielded positive results 5 days later. The patient was successfully treated with amoxicillin.
Lyme disease occurs after an initial Ixodes tick bite. The spirochete, B burgdorferi, spreads to the synovial tissue, triggering an immune response and subsequent synovial hypertrophy. Of the patients left untreated, 50% to 60% of them will develop Lyme arthritis.3 Although joint symptoms may begin as early as a few days or weeks after the tick bite, Lyme arthritis usually appears in the late stage of Lyme disease a few months after the tick bite.2 Typically, episodes of Lyme arthritis involve a single large joint, with the knee being the most common. A 90-patient study by Gerber et al5 found that the knee (90%) was the most commonly affected joint and that Lyme disease affected the elbow in 7% of the cases. However, they did not differentiate between monoarticular and polyarticular presentations.5 Similarly, a case series by Milewski et al6 reviewed 123 patients with Lyme arthritis and found that 86% of them had knee involvement, while 3% had elbow involvement.
According to the Centers for Disease Control and Prevention, Lyme disease is the most common tick-borne illness in the United States and is endemic to the Northeast.7,8 Between 2001 and 2016, a total of 253,690 cases of Lyme disease were reported.9 However, most cases go unreported, and it is estimated that there are actually 296,000 to 376,000 cases annually in the United States.1 Lyme disease is most prevalent in boys aged 5 to 9 years.1 Arthritis is the second most common symptom and is present in 30% of patients.9 In the current study, 4 cases of Lyme arthritis presented isolated to the elbow.1
Physical examination, synovial fluid analysis, and blood work play a significant role in diagnosing septic arthritis. Kocher and Zurakowski10 reported 4 criteria that suggested an increased risk of a septic hip: temperature above 38.5°C, white blood cell count above 12,000 cells/µL, ESR greater than 40 mm/h, and an inability to ambulate. According to Kocher and Zurakowski,10 if 1 criterion is present, the risk of a septic hip is 3%; if 2 criteria are present, the risk is 40%; if 3 criteria are present, the risk is 93%; and if 4 criteria are present, the risk is 99.6%.10 A fifth criterion of C-reactive protein above 2.5 mg/L was later added by Caird et al.11 Although the study by Kocher and Zurakowski10 focused on septic arthritis of the hip, its results have since been applied to other joints as well.
Baldwin et al12 evaluated 189 pediatric patients in an attempt to find predictive factors for differentiating between septic arthritis and Lyme arthritis of the knee. Of the 189 patients reviewed, 46 patients had septic arthritis and 140 had Lyme disease. Using multivariate analysis, they found that pain with short arc motion, C-reactive protein greater than 4 mg/L, age younger than 2 years, and history of fever were predictive of septic arthritis. Furthermore, they showed that the risk of septic arthritis was 2% when none of the factors were present, 18% with 1 of the factors present, 45% with 2 of the factors present, 84% with 3 of the factors present, and 100% with all 4 present.12 Applying these criteria to the current cases reveals that the likelihood of them being septic arthritis is low.
Although important, history and inflammatory markers are not always reliable. Fever has been noted in 22% to 51% of children with Lyme arthritis.6,13,14 Culp et al15 evaluated 43 patients with Lyme arthritis and found that ESR is elevated in most patients with Lyme arthritis, with 30 of 36 patients with Lyme arthritis showing elevated ESR. Therefore, ESR alone is not helpful in differentiating Lyme arthritis from bacterial septic arthritis. Furthermore, less than 50% of patients with Lyme arthritis reported exposure to ticks, and less than 45% of patients described a rash.15 In addition, geography and season are not always reliable because individuals may travel from a Lyme endemic area to a non-Lyme endemic area and because it could take months for Lyme arthritis to present. Therefore, Lyme arthritis should always be part of the differential diagnosis regardless of season, location, and joint affected.
Septic arthritis and Lyme arthritis share many features.16–18 The 2 pathologies have significant overlap in their synovial fluid profiles. Nonnucleated cell counts of 25,000 to 100,000 cells/mm3 and elevated neutrophil content on the differential diagnosis are common in the 2 pathologies. Bachman and Srivastava19 reviewed 29 patients with Lyme disease. Of the patients, 15 children had joint pain as their presenting symptom. The average synovial fluid nucleated cell count was 94,023 cells/mm3 with 84% neutrophils. Seven children with Lyme disease were initially diagnosed with septic arthritis, and 6 underwent surgical intervention.19 Heyworth et al20 evaluated 46 pediatric patients who presented with a painful hip and had a nonnucleated cell count between 25,000 and 75,000 cells/mm3 on synovial fluid analysis. Of the 46 patients within this range, 33% had septic arthritis and 28% had Lyme arthritis. They further subdivided the patients into 2 subgroups consisting of patients with 25,000 to 50,000 cells/mm3 and those with 50,000 to 75,000 cells/mm3. They found a significant difference in favor of patients having septic arthritis in the higher cell group as compared with other pathologies. However, Heyworth et al20 did not find a significant difference between septic arthritis and Lyme arthritis specifically. They did not find any other significant differences.
The mainstay in diagnosing Lyme arthritis is serologic testing. It is recommended by the Centers for Disease Control and Prevention that a 2-test approach be performed. First, the samples are tested for antibodies to B burgdorferi by enzyme-linked immunosorbent assay. Those with equivocal or positive results are then tested by a Western blot smear.21 Lyme arthritis usually manifests later in Lyme disease; therefore, the patients have had time to develop immunoglobulin G antibodies. Once the diagnosis of Lyme disease is confirmed, it is treated with antibiotics.22 Currently, because of logistics and cost-effectiveness, the serology for Lyme disease is processed twice a week in the authors' hospital's laboratory. In the past, enzyme-linked immunosorbent assay and Western blot samples were sent to an outside laboratory. Both of these practices lead to a delay in diagnosis. In the current study, there was an average delay of 4.75 days between patient presentation and positive result. Willis et al18 evaluated 10 children who presented to the emergency department with an acutely swollen joint suspicious for bacterial septic arthritis. All were found to have Lyme disease. Of the 10 patients, 7 underwent operative intervention. Surgical intervention was planned for the remaining 3 patients, but the availability of a rapid test for Lyme disease prevented the surgery.18 In this case, the presence of prompt Lyme testing prevented unnecessary surgery. However, a delay in diagnosis may lead to unnecessary surgery in the case of Lyme arthritis or continued joint damage in the case of a septic joint. Therefore, it is important that hospitals in Lyme endemic areas have the ability to obtain Lyme serology immediately.
The Infectious Diseases Society of America recommends that patients with Lyme arthritis be initially treated with a 30-day course of oral doxycycline, 100 mg twice daily, or amoxicillin, 500 mg 3 times daily.23 Unless there are concomitant neurologic abnormalities, oral regimens are the initial treatment of choice. If there is persistent mild swelling after a 30-day regimen of oral antibiotics, then a 30-day oral regimen should be repeated. However, if there is persistent moderate-to-severe joint swelling after a 30-day regimen, then the treatment should be switched to intravenous ceftriaxone, 2 g/d for 2 to 4 weeks.22
This study had several limitations. First, treatment bias may have been introduced because the patients were managed by different attending physicians. Second, the retrospective nature of the case series makes this study vulnerable to incomplete documentation and missing information regarding the patients' presentations. Third, this case series was small, leaving it underpowered to potentially find significant differences between patients presenting with Lyme arthritis vs septic arthritis.