Aggregatibacter aphrophilus spinal epidural abscess

Abstract

A 76-year-old woman with a rare case of spinal epidural abscess (SEA) that had no risk factors for such type of infection, presented symptoms of back pain, progressive neurological deficit of the lower limb and loss of sphincter control. A gadolinium-enhanced MRI confirmed the diagnosis of an SEA. The patient underwent laminectomy with surgical drainage, where cultures showed the presence of Aggregatibacter aphrophilus, a bacterium of the HACEK group (Haemophilus species, Aggregatibacter species, Cardiobacterium hominis, Eikenella corrodens, and Kingella species), rarely involved in SEA. Following surgery, the patient was treated with intravenous ceftriaxone for 6 weeks, and this gave excellent results.

Background

Epidural abscesses with Aggregatibacter aphrophilus are very rare. There are no specific guidelines for the antibiotic treatment of HACEK-caused spinal epidural abscess (SEA). The application of third generation cephalosporins can be considered a good choice, however, given their known activity on this type of bacteria. In the case of our patient, excellent results were obtained following the treatment of laminectomy with surgical drainage, antibiotic therapy with ceftriaxone for 6 weeks and all of this combined with daily physiotherapy.

Case presentation

We present the case of a 76-year-old woman who had progressive back pain below the left scapula. After 8 days of symptomatic treatment (painkillers and rest), she displayed balance disorders, due to weakness of the left lower limb and sphincter disorder, alternating incontinence and urinary retention. Her medical history included several uncomplicated surgeries (hysterectomy, appendicectomy, partial thyroidectomy, bilateral total hip prosthesis) and gastric ulcer 10 years prior. The patient was in retirement and lived alone. She had used neither tobacco, alcohol nor drugs and had not had any recent dental care. The only medication being taken at that time was vitamin D and thyroid hormones.

On admission to the emergency department, the patient presented urinary incontinence along with persistent back pain but without any associated fever. Clinical examination showed slurred gait and hyper-reflexia of the right lower extremity (knee and ankle). There was nothing particular about palpation of the spine nor was there anything of note about the rest of the clinical examination.

Investigations

Blood tests on admission showed an increased C reactive protein (CRP) of 322 mg/L, white cell count at 16,7 ×10⁹/L with a neutrophilic formula. Chest X-ray showed bilateral pleural effusions of a small volume. Blood cultures were negative. Thoracolumbar CT scan showed no evidence of spondylodiscitis, osteolytic lesions, dural sac or intervertebral disc abnormalities.

A full spine MRI was performed with contrast. It showed a clear thickening of the meningeal membranes, along the entire length of the spine but predominantly at T4, T5 and T6 with epidural abscess at those levels, as well as narrowing of the spinal cord, confirming the diagnosis of SEA (figures 1 and 2). There was no associated myelitis.

Figure 1.

T1-weighted, gadolinium-enhanced sagittal MRI showing the spinal epidural abscess (red arrows).

Figure 2.

T1-Weighted, gadolinium-enhanced MRI axial at the T5 level showing the spinal epidural abscess (red arrows).

Treatment

A urinary catheter was inserted on admission. The patient was treated empirically with antibiotics (ceftazidime and vancomycin) from the day of admission. She underwent laminectomy with drainage of the abscess on day 3 after admission. Two bacteriological samples were collected, both showed the presence of A. aphrophilus after 24 hours of incubation. The bacteria were identified by Matrix Assisted Laser Desorption Ionisation-Time of Flight, Brucker with a very high confidence index (log score 2.47). No other bacterium was found in the culture. Sensitivity was tested using the E-test technique and showed a minimum inhibitory concentration of 0.19 mg/L for amoxicillin and 0.016 mg/L for cefotaxime. In view of these results, vancomycin was discontinued and ceftazidime was replaced by ceftriaxone (2 g twice daily) from day 7 and continued for a total duration of 6 weeks. The patient benefited from daily physiotherapy sessions for up to 3 months after surgery.

Outcome and follow-up

The urinary catheter was removed after 3 weeks with excellent sphincter control recovery. The inflammatory syndrome regressed progressively with a normalised CRP 3 weeks after admission. Three months after surgery, MRI control was reassuring and there was no recurrence of collection or signs of epiduritis. Neurologically, the evolution was favourable with progressive recovery. Three months after the antibiotic therapy, the neurological clinical examination was completely normalised. There was no pain or sensorimotor deficits and the patient was considered to have been cured.

Discussion

SEA is a rare condition with an incidence ranging from 3 to 12 per 10 000 admissions. Its incidence has been increasing over the past two decades, likely due to the ageing population and the availability of MRI for early diagnosis.^{1 2} There are a significant number of risk factors associated with the development of SEA. These are mainly comorbidities (diabetes, alcoholism, haemodialysis, cancer, …), factors promoting bacteraemia (intravenous drug use, skin lesion, concomitant infection, …) or local bacterial spread (epidural infiltration, acupuncture, tattooing, …).³ A recent review has counted nearly fifty different risk factors associated with SEA listed in the literature. Interestingly, the authors question their usefulness in clinical practices. All the more so since 20%–50% of patients with a diagnosis of SEA had none of these risk factors.¹ In this case, the patient actually had none of the risk factors traditionally described.

From a clinical point of view, the most frequent symptom is back pain, found in 70%–100% of cases, lasting from several days to several weeks. Fever is found in one out of two cases. The main neurological manifestations are muscular weakness, radiculopathies and sphincter control disorders. At least one of these neurological symptoms is present in a third of the cases, but the classic triad of back pain, fever and neurological deficit is found in only a minority of cases.^{1 3}

For the diagnosis, the biological assessment almost always shows an inflammatory syndrome with leucocytosis. Blood cultures have to be performed, but only in half of the cases a causative agent is found.^{1 3} The reference imaging for the diagnosis of SEA is gadolinium-enhanced MRI with sensitivity and specificity higher than 90%. Contrast CT has a lower sensitivity, especially in early stages, but its use should be considered if MRI cannot be performed.¹

Treatment is based on surgery and antibiotic therapy. Surgical management is necessary both to perform laminectomy at the sites of abscess and to take microbiological samples to guide the antibiotic treatment.¹ Antibiotic treatment is sometimes offered alone in the absence of neurological signs, but recent data show a higher mortality rate and persistent neurological deficit without surgery.²

The pathogens involved in SEA (excluding mycobacteria) are mainly Staphylococcus aureus (60%–90%), among which the proportion of methicillin-resistant S. aureus varies greatly depending on the local ecology (15%–40%). At a lower frequency (5%–10% each) are Streptococci, coagulase-negative Staphylococcus and gram-negative bacilli (among which Escherichia coli and Pseudomonas aeruginosa are the most prevalent). In 5%–10% of cases, the causative agent remains unknown. Finally, 1% of cases involve rare micro-organisms such as certain parasites and fungi.^{1 3 4} The route of spread is haematogenic in half of the cases and by direct contiguity in a third of cases, the remaining part stays without an identified source.¹

Aggregatibacter aphrophilus is a fastidious gram-negative bacilli belonging to the HACEK organisms which are a part of the normal oropharyngeal microbiota. This family includes bacteria involved in rare and serious infections, particularly endocarditis (1.4%–3% of cases) and osteoarticular infections.⁵ The name A. aphrophilus was proposed in 2006 to group together two species with similar molecular characteristics to replace their former names: Haemophilus aphrophilus and Haemophilus paraphrophilus.⁶ A. aphrophilus is a bacterium frequently found in human dental plaque. Pathologies involving this micro-organism are mainly infectious endocarditis, bone infections and brain abscesses.^{5 7}

A. aphrophilus is a pathogen very rarely involved in epidural abscesses. In 2009, a review counted 17 cases of spinal infections involving this bacterium. These were mainly vertebral osteitis (11 cases out of 17) and spondylodiscitis (4 out of 17). In this series, only one case was an epidural abscess with no other associated osteoarticular infection.⁸ In this case, the patient was successfully treated with drainage surgery and antibiotics for 5 weeks (intravenous amoxicillin and ceftriaxone for 3 weeks and ciprofloxacin orally for 2 weeks).⁹ Two other similar cases have been reported. One was successfully treated with antibiotics alone: cefotaxime for 2 weeks then ciprofloxacin for 6 weeks.¹⁰ The other patient had been successfully treated with drainage surgery and intravenous ceftriaxone for 6 weeks.¹¹

Learning points.

• Aggregatibacter aphrophilus is a bacterium of dental plaque, exceptionally involved in epidural abscesses.

• Back pain is the far most frequent symptom in spinal epidural abscesses. Fever and neurological deficits are less frequently observed.

• Gadolinium-enhanced MRI is the gold standard in the diagnosis of spinal epidural abscess.

• There are no specific guidelines for the treatment of HACEK spinal epidural abscesses which are extremely rare. A treatment combining surgery with drainage and a third generation cephalosporin antibiotic treatment for 6 weeks gave excellent results in our case.