A 25-year-old man with a history of epilepsy developed a gradually progressive swollen, red, and painful left first toe. There was no preceding trauma. There were no fevers or chills, and no other extremities were involved. He had never had these symptoms in the past. He was seen initially by his primary care provider, who diagnosed cellulitis and prescribed cephalexin 500 mg 3 times a day for 1 week, with no improvement. A trial of colchicine 0.6 mg twice a day was attempted for 1 week, but this also generated no improvement and a referral was made to rheumatology.
At the initial rheumatologic evaluation the patient reported that his symptoms had started in September, and had been present for close to 2 months. He described the left first toe as “swelling up like a globe.” The second toe developed pain and swelling within 2–3 weeks, although not as severe. By 1 month the pain had spread to his left ankle. Walking was becoming much more difficult due to the pain. He was taking ibuprofen, with minimal improvement. He denied systemic symptoms. There was no eye redness or pain, no burning with urination or penile discharge, no back pain, and no nausea or vomiting, and his bowels were normal. There was no history of psoriasis.
The medical history was pertinent for well-controlled idiopathic generalized epilepsy, depression, asthma, and environmental allergies. There was no history of surgeries.
His medications included valproic acid, ethosuximide, sertraline, cetirizine, and inhaled albuterol as needed.
He worked as an artist, primarily oil painting. He was living with his parents. He denied tobacco or drug use, with occasional alcohol use. There was no sexual activity in the last year.
There were 2 uncles with gout. Both parents were alive and healthy.
Vital signs were normal, and there were normal results on eye, mouth, neck, cardiac, pulmonary, and abdominal examinations. There was prominent folliculitis on the back. The extremity examination was unremarkable, with the exception of the left foot. The left first toe was diffusely swollen, red, and tender, suggestive of dactylitis (Figure 1). The second toe also had some swelling and redness, but to a lesser degree. There was no pain or swelling of the ankle. Tibial and dorsalis pedis pulses were normal.
Laboratory tests and imaging
Laboratory testing revealed a normal complete blood count, comprehensive metabolic panel, erythrocyte sedimentation rate, and C-reactive protein level. A serum uric acid level was normal at 5.8 mg/dl. Testing for HLA–B27 was negative. Serologic testing for hepatitis B and C was negative. Plain films of the left foot were normal, with no evidence of fracture or bony abnormality.
PATIENT'S INITIAL COURSE
It was thought that this represented dactylitis and the patient was started on diclofenac 50 mg twice a day, with instructions to increase to 3 times a day if needed. Over the next week, he had little relief from the diclofenac, and the pain was increasing in severity. He had difficulty sleeping due to the pain, and with the sleep deprivation he had a seizure. Magnetic resonance imaging (MRI) of the toe was ordered, and he was started on prednisone 20 mg daily. This treatment had minimal impact on his pain and swelling, so the dose was increased to 40 mg daily. Even at this dose, there was minimal improvement in his symptoms. MRI of the toes revealed extensive subcutaneous edema in the first, second, and to a lesser extent, the third toes. There was patchy fluid signal intensity within the bone marrow involving the metatarsal head, proximal phalanx, and terminal phalanx of the first toe, with lesser involvement of the second and third toes (Figure 2). There was no tenosynovial or synovial inflammation, nor was a joint effusion present.
The patient's first symptom of a painful swollen toe suggested a monarthritis. An acute monarthritis is most likely caused by infection, trauma, or crystal disease. The patient did not recall any trauma, but with a history of epilepsy it was reasonable to consider occult trauma during a seizure. Cellulitis was a reasonable initial consideration, but with prolonged symptoms and no response to antibiotics this possibility was unlikely. Infectious etiologies such as syphilis and gonococcus remain on the differential diagnosis. The examination, however, indicates that the primary involvement was with the periarticular soft tissues, and therefore additional diagnoses need to be considered.
The patient's swollen toe mimics dactylitis. Dactylitis refers to diffuse swelling of a toe, classically described as a “sausage digit.” This appearance typically results from tenosynovitis of the toe and is most commonly seen in the class of disorders termed seronegative spondylarthritis. Dactylitis most frequently accompanies psoriatic arthritis, but also occurs with reactive arthritis, ankylosing spondylitis, and arthritis associated with inflammatory bowel disease. Less common diseases that can manifest with dactylitis include sarcoidosis, syphilis, tuberculosis, gout, and sickle cell disease. This patient's MRI had no evidence of tenosynovitis or joint effusion, as would be expected in dactylitis. Further, the lack of response to nonsteroidal antiinflammatory drugs or prednisone is also unusual, leading to the conclusion that this is not a typical representation of dactylitis.
Given the presence of the folliculitis on the back, another entity to consider is acne-associated arthritis. This is an uncommon entity usually related to severe acne in the form of acne conglobata or hidradenitis suppuritiva. The patient had no cystic areas and no scarring, making this diagnosis less likely.
MRI indicates bone marrow edema as the primary pathology. This finding raises the concern about osteomyelitis, yet the lack of cortical irregularities on MRI plus the prolonged course make this questionable. Bone marrow edema is also seen in psoriatic arthritis (1). Osteonecrosis, osteoarthritis, tumor-related bone marrow edema, as well as trauma leading to a bone contusion or microfracture should also be considered with these MRI results. Reflex sympathetic dystrophy, also known as complex regional pain syndrome (CRPS), has also been associated with bone marrow edema on MRI (2). Usually CRPS arises in an extremity following trauma or surgery, is very painful, and is associated with color changes such as hyperemia, swelling, and altered sweating. The diagnosis is sometimes difficult to establish. A standard imaging modality to aid in making the diagnosis is the 3-phase bone scan. These scans usually show increased activity in the first few months, followed by patchy osteoporosis later in the course.
At a followup visit, the patient entered the room in a wheelchair due to the severity of the pain. The prednisone was discontinued, and a diagnosis of CRPS was considered. He was prescribed gabapentin 300 mg 3 times a day, a lidocaine patch, tramadol 50 mg 3 times a day, and physical therapy. A 3-phase bone scan revealed diffusely increased activity throughout the left foot, most severe at the periarticular regions and especially in the hallux (Figure 3). These findings were thought to be consistent with CRPS.
There was no clear etiology for the development of CRPS, so additional laboratory work was obtained to identify a potential cause of peripheral neuropathy that may lead to this condition. Results included normal serum protein electrophoresis, negative syphilis serology, negative rheumatoid factor, and a negative antinuclear antibody and extractable nuclear antigen panel. Enzyme-linked immunosorbent assay (ELISA) testing for combined IgM and IgG antibodies to Borrelia burgdorferi was positive with a value of 3.05 (negative cutoff 0.69) using the Immunetics C6 B burgdorferi ELISA kit. A confirmatory immunoblot demonstrated IgM antibodies to bands p41, p39, and p23, and IgG bands p93, p66, p58, p41, p39, p28, and p18 (Figure 4). Nerve conduction studies of the left lower extremity were performed, showing low-amplitude nerve compound motor and sensory action potentials in the tibial nerve. These findings were thought to be consistent with a diagnosis of neuritis occurring in the context of Lyme disease or Lyme neuritis. Upon further questioning, the patient said he had spent many hours outside painting in Southern Vermont, and although he had seen many ticks, he did not recall any engorged ticks or any rash. The presence of peripheral neuritis without arthritis or rash suggests that this was a manifestation of early disseminated Lyme disease.
The patient was treated with doxycycline 100 mg twice a day for 4 weeks. By the second week his pain had diminished by greater than 50%, and by the end of the 4 weeks of antibiotic therapy, his pain was minimal. Additional Lyme testing was not obtained, as it would not affect clinical decision making. He continues to have minimal numbness and loss of pin and vibration sense in the affected toes. He was tapered off of gabapentin, and the lidocaine patch was discontinued.
CRPS is also known as reflex sympathetic dystrophy, causalgia, algodystrophy, and Sudeck's atrophy. This entity usually occurs in a single extremity following tissue trauma of some kind. The diagnosis requires regional pain and sensory changes to an extremity. The pain is often greater than expected from the injury that produced it. Symptoms often include swelling, color changes, temperature change, and altered sweating, and often persist for weeks to months (3). Laboratory studies are generally not helpful. Bone scintigraphy is not very sensitive but has high specificity, displaying a characteristic pattern of increased uptake in the affected extremity as compared to the normal extremity (4, 5).
This syndrome is usually separated into 2 forms: CRPS-1, which occurs without evidence of nerve injury, and CRPS-2, which occurs in the setting of known nerve injury (6). Some authors suggest that the difference is really arbitrary, since the degree of nerve involvement cannot always be appreciated (7, 8). Classically, CRPS-2 would be defined by an injury such as a crush injury to a neurovascular bundle or to a transection of a nerve. However, it is becoming more evident that some degree of nerve inflammation or injury is present as an underlying insult leading to CRPS, even if the degree of nerve insult is subclinical (9).
The pathophysiology of CRPS is poorly understood, and is complicated by the heterogeneity seen among cases. It is likely that there is considerable variation in the inciting events as well as the propagating factors that combine to produce the manifestations of CRPS in individual patients. Despite this variability, 3 recent reviews have identified some common features that contribute to the development of this syndrome, beginning with some form of tissue damage, which may be from trauma or ischemia (7, 10, 11). Local inflammation ensues with an increase in cytokines, including interleukin-1β (IL-1β), IL-2, IL-6, tumor necrosis factor α, bradykinin, and substance P, leading to peripheral nerve sensitization (7). Vasoconstriction leads to increased hypoxia, therefore propagating the pathology. Central spinal sensitization occurs, further magnifying pain perception. Autonomic sympathetic nervous system dysregulation may play a role in some patients (10). Exactly how these factors combine to create the disease in some, but not all, patients experiencing nerve injuries is not clear.
The case described above suggests a correlation between Lyme neuritis and the development of CRPS. Such a correlation can be supported in light of the understanding of the pathophysiology of CRPS described above. As the inciting event for the development of CRPS appears to be some form of nerve injury followed by local inflammation, it is conceivable that an acute or chronic nerve infection could serve as that source of injury and inflammation, and this would be most consistent with the rapid improvement in his symptoms with antibiotic therapy directed against the causative organism. Borrelia affecting peripheral nerves is a well-described entity (12). The most common presentation involves cranial neuropathies; however, the European literature has a higher incidence of reported painful radiculitis (12). Logigian and Steere reported a series of patients with neuropathy due to Lyme disease, of which more than one-half of the patients had slowing of either peripheral motor or sensory nerve conduction, with 76% of patients improving with ceftriaxone therapy (13). Although this series reports on late disseminated Lyme disease, similar results on nerve conduction studies may be seen in early disseminated disease, as was seen in our patient (14). Since the literature supports the hypothesis that CRPS can develop in conjunction with a neurotropic infection such as herpes zoster (15–17), this raises the possibility that B burgdorferi, another neurotropic infectious agent, might cause the same problem in certain clinical situations.
There are a few prior reports of CRPS correlated with Lyme disease. Sibanc and Lesnicar (18) described a 46-year-old man who developed CRPS of the left leg, which was refractory to all treatments. The patient later recalled a tick bite on the left foot months earlier, followed by a circular erythematous rash. Serologies were positive for Lyme disease, and negative for other infections. He was treated with ceftriaxone and, like this patient, had a dramatic improvement in the CRPS. In a study by Kohler and Thoden (19), 42 patients with neuroborreliosis and painful meningopolyneuritis were evaluated, with CRPS being identified in 3 patients.
Other case reports of Lyme disease–associated CRPS show less direct associations with peripheral nerve dysfunction. Bruckbauer et al reported a case of a man who developed contralateral CRPS after being treated for neuroborreliosis that presented with radicular pain of the left leg, and was confirmed by finding the spirochete in a skin biopsy tissue culture (20). Neumann et al described 5 patients with a correlation between Lyme disease and development of CRPS (21). In 1 of these patients the spirochete was identified in histology, while in 2 others late Lyme disease was diagnosed by the presence of acrodermatitis chronic atrophicans and positive IgG serologies for Borrelia. The fourth patient with CRPS also had prior circumscribed scleroderma that had responded to penicillin.
The patient described in this article developed CRPS of the left foot, as confirmed by the presence of pain, swelling, and color changes; lack of response to prednisone; appropriate MRI findings; and characteristic findings on bone scintigraphy. Additionally, he had evidence of Lyme neuritis based on the diagnostic serologies and results of nerve conduction studies. The success of antibiotic therapy in improving all elements of CRPS further establishes the linkage with B burgdorferi as the causative organism.
This case provides many important points. It is supportive of the hypothesis that some form of nerve injury underlies the pathophysiology of CRPS and demonstrates the ability of CRPS to mimic dactylitis. It also illustrates the propensity for Lyme disease to affect peripheral nerves, perhaps through direct infection. Additionally, it supports the possibility of Borrelia leading to the development of CRPS. Awareness of these relationships may lead to improved diagnosis and appropriate therapy in some patients with idiopathic CRPS.
Lyme disease with peripheral neuritis and associated complex regional pain syndrome mimicking dactylitis.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Jones had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Jones, Albert, Thadani, Rigby.
Acquisition of data. Jones, Albert, Thadani, Rigby.
Analysis and interpretation of data. Jones, Albert, Thadani, Rigby.