Low back pain after a dental procedure: a case of Streptococcus viridans vertebral osteomyelitis

Abstract

Vertebral osteomyelitis due to Streptococcus viridans following a dental procedure is a rarely reported phenomenon. We discuss the case of a 67-year-old immunocompetent woman who presented with low back pain of 3 weeks duration associated with subjective fever and chills. On admission, the MRI of the lumbar spine showed L5-S1 vertebral osteomyelitis with associated paravertebral and epidural abscesses. Subsequently, detailed history was retaken and the patient reported having had a maxillary tooth extraction followed by a dental implant 2 months prior to the onset of her symptoms. Blood and abscess fluid cultures grew S. viridans. Transthoracic echocardiogram showed no evidence of endocarditis. The patient was started on intravenous ceftriaxone but her treatment course was complicated by agranulocytosis requiring a switch to vancomycin. She required a total of 9 weeks of intravenous antibiotics for complete clinical cure.

Background

Low back pain is the fifth most common reason for office visits in the USA.¹ The majority of patients seen in the primary care setting have non-specific low back pain. However, occasionally, serious causes such as vertebral osteomyelitis (VO) might be overlooked. Common risk factors for VO include diabetes mellitus, injection drug use, degenerative spine disease, alcohol abuse, malignancy and any other immunocompromised state.² Compared to Staphylococcus aureus, Streptococcus viridans is a rare cause of VO. It often has a subacute presentation with no fever and no neurological symptoms.³ This leads to a delay in diagnosis, with the potential for increased morbidity and worse neurological outcomes. Cases of infective endocarditis associated with VO due to S. viridans following dental work have been reported in the literature.^{4 5} To the best of our knowledge, this is the first reported case of isolated VO with epidural abscess due to S. viridans following a dental procedure. This case emphasises the importance of obtaining a dental history and considering VO in the differential diagnosis of new onset back pain.

Case presentation

A 67-year-old woman with no significant medical history presented to the emergency department of our hospital, with the symptom of low back pain of 3 weeks duration. Associated symptoms included occasional subjective fevers, chills and night sweats for 3–4 days. She reported no bowel/bladder dysfunction, no lower extremity weakness, no numbness and no saddle anaesthesia. On presentation, she had normal vital signs and was afebrile. On physical examination, she had localised tenderness to palpation in the lumbosacral region while the rest of the examination was unremarkable, including a negative straight leg raise test. The patient denied any history of intravenous drug use. She reported a history of penicillin allergy characterised by generalised hives, with no angio-oedema and no airway compromise.

Prior to this presentation, she had visited her family physician's office twice in the recent past, for low back pain. At that time she had denied any history of fever or chills and her examination was reported as normal. Her pain was attributed to a musculoskeletal origin and she was prescribed cyclobenzaprine on the first visit and acetaminophen/hydrocodone after the second visit. A plain radiograph of the lumbosacral spine was taken on her second visit, which was reported as normal.

Investigations

On initial investigations, significant laboratory abnormalities included C reactive protein (CRP) 41.34 mg/dL (normal : <1.00 mg/dL), erythrocyte sedimentation rate 109 mm/hour (normal 0–20 mm/hour) and white cell count of 13 500/µL (normal 4800–10 800/µL) with an absolute neutrophil count of 10 300/µL. The rest of the routine blood work was unremarkable. Owing to concern for VO, MRI of the lumbar spine with contrast was performed, which showed discitis and osteomyelitis at the L5-S1 level, with associated paraspinal as well as epidural abscess at a similar level, as shown in figure 1. Following this, detailed history was retaken and the patient reported having had a maxillary tooth extraction followed by a dental implant placement 2 months prior to developing these symptoms. She underwent computerised tomography-guided paraspinal abscess drainage and placement of a 12 French drainage catheter in the paraspinal abscess cavity on day 2 of hospitalisation. The initial two sets of blood cultures and abscess fluid cultures returned positive for viridans group streptococci, which were sensitive to penicillin and ceftriaxone. Transthoracic echocardiogram (TTE) performed on day 3 of hospitalisation showed no vegetation and no valvular abnormality.

Figure 1.

MRI of the lumbar spine with contrast showing discitis and osteomyelitis at the L5-S1 level. Arrows pointing to paraspinal abscess at the L5-S1 level and rim-enhancing epidural collections at L5, S1 and S2 levels, suggesting probable epidural abscesses with significant narrowing of the spinal cord at these levels.

Treatment

The patient spiked a fever of 38.3°C on day 1 of hospitalisation and was started on intravenous vancomycin because of the history of penicillin allergy. She was switched to intravenous ceftriaxone on day 6 of hospitalisation after discussing the benefits of once daily dosing and the risks associated with possible cross-reactivity due to history of penicillin allergy. On day 10 of hospitalisation, she was discharged to a nursing home, with the drainage catheter in place due to persistent purulent drainage and presence of residual cavity. Ceftriaxone 2 g intravenous once daily was continued for a planned duration of 6 weeks. Follow-up with interventional radiology and infectious disease was arranged with instructions for weekly blood work to trend her blood counts and inflammatory markers.

Outcome and follow-up

Two weeks after discharge, the drainage catheter was removed by the interventional radiologist after fluoroscopy showed no residual cavity. The patient developed neutropenia (absolute neutrophil count (ANC) of 1300/µL) on day 28 of intravenous ceftriaxone. The rest of the cell lineages were not affected. Ceftriaxone was discontinued at that time and vancomycin was initiated. Two days later, she developed agranulocytosis with ANC of 0/µL and the staff at the nursing home reported low-grade fevers. The patient was readmitted to the hospital, with concern for worsening infection versus drug fever. She was continued on intravenous vancomycin and repeat MRI of her lumbar spine fortunately showed improvement of the underlying osteomyelitis and the associated abscesses. Her agranulocytosis was deemed to be due to ceftriaxone as ANC improved to 0.3/µL 5 days after discontinuation. She was given filgrastim (granulocyte colony-stimulating factor) with normalisation of her neutrophil count. She was discharged on intravenous vancomycin and required a total of 9 weeks of treatment for complete clinical recovery.

Discussion

VO accounts for 1% of all skeletal infections and is most commonly caused by S. aureus. The common site of involvement is lumbar vertebrae followed by the thoracic region. The clinical presentation is often non-specific, with back pain being the most common symptom.² Importantly, fever can be absent at presentation, thus distracting the physician from the possibility of an infectious aetiology, leading to a delay in diagnosis, as was seen in this patient. S. viridans is an unusual cause of VO. A study by Murillo et al³ reported that, compared to S. aureus, the S. viridans patient group had a subacute presentation, with fewer patients having fever and neurological symptoms, lesser values of CRP elevation and significantly longer diagnostic delays with mean duration of symptoms until diagnosis of 44 days. In the same study, the mean age of patients with VO due to S. viridans was 68 years.

Since the clinical presentation of VO is often non-specific, imaging is an essential part of the diagnostic work up. Plain radiographic changes take weeks to months to appear and may not be helpful.⁶ Our patient had a plain radiograph of the spine 2 weeks prior to presentation, which showed no radiographic signs of osteomyelitis. MRI is the most sensitive radiographic technique for diagnosis of both VO and epidural abscess.⁷ It can detect early infection, evaluate the extent of the disease affecting the spine and delineate soft tissues as well as neural structures. The 2015 Infectious Disease Society of America (IDSA) guidelines on the diagnosis and treatment of native VO⁸ recommends against the need for image-guided biopsy in patients with blood cultures positive for S. aureus and clinical evidence of VO. However, in patients with bacteraemia due to Streptococcus species and clinical evidence of VO, the need for tissue diagnosis is left to the treating physician's discretion. We felt that our patient had clinical, laboratory and radiological findings consistent with S. viridans VO and hence we decided not to pursue image-guided biopsy.

Cases of VO and concomitant infective endocarditis due to S. viridans following dental procedures have been reported.^{4 5} In the two reported cases, patients had TTE evidence of mitral valve vegetation with associated mitral regurgitation. The most likely mechanism of infection in such cases is haematogenous spread from significant bacteraemia related to dental interventions.⁹ Our patient had no clinical evidence of infective endocarditis and the TTE was normal. Hence we believe that she had isolated VO that resulted from the bacteraemia related to the dental procedure.

The IDSA guidelines recommend penicillin G or ceftriaxone as the treatment of choice for VO due to streptococcal species.⁸ Vancomycin use is recommended as an alternative in patients who are allergic to the first-line medications. Our patient was allergic to penicillin and was hence started on ceftriaxone because of the ease of once daily dosing. Once the patient developed agranulocytosis she was switched to vancomycin as an alternative treatment. Although rare, ceftriaxone induced agranulocytosis has been reported in the literature.¹⁰ A single randomised clinical trial showed that 6 weeks of intravenous antibiotic treatment is non-inferior to 12 weeks of treatment for pyogenic VO.¹¹ We proceeded with prolonged treatment because of the severity of initial infection with associated paravertebral and epidural abscesses, and a complicated treatment course with agranulocytosis and fever during the fifth week of treatment.

Learning points.

• Although rare, vertebral osteomyelitis should be considered in patients with new onset low back pain and a history of recent dental work.

• Streptococcus viridans is an uncommon cause of vertebral osteomyelitis, and often results in a subacute presentation.

• Fever can be absent in cases of vertebral osteomyelitis. Hence diagnosis of this entity requires a high index of suspicion.