Actinomycosis is a chronic suppurative condition caused by anaerobic or microaerophilic gram-positive bacteria of the genus Actinomyces. Actinomyces israelii is most commonly responsible for this condition but A. naeslundii, A. odontolyticus, A. meyeri, A. viscosus and A. gerencseriae may also cause disease in humans . These organisms are part of the usual flora of the mouth, gastrointestinal tract and female genital tract. We report a case of vertebral actinomycosis with associated spinal cord compression.
A 63-year-old Chinese man presented with a 2-month history of upper thoracic central and radicular back pain and difficulty walking. He had been born in China but lived most of his life in Taiwan. Recently, he had come to live with his daughter in New Zealand. During these 2 months he had developed increasing leg weakness and difficulty walking so that he required assistance to move about. He had neither fever nor weight loss.
Six months before admission he had a left hemispheric cerebrovascular accident with speech disturbance and right leg weakness that fully resolved within 2 months. He was found to be hypertensive then and began aspirin and a calcium channel blocker. There was no other past medical history and he was on no other medications. He was a nonsmoker and did not drink alcohol.
On examination he was afebrile and did not look unwell. He had bad dental caries. The only abnormalities on examination were upper motor neurone signs in his lower limbs. There was a sensory level at the T5 dermatome and his gait was ataxic. There was mild to moderate tenderness to percussion in the region of the third and fourth thoracic vertebrae.
Initial investigations revealed a haemoglobin of 155 g/L, a white cell count of 9.5 ×109/L with a normal differential and an ESR of 74 mm/hour. His renal function and liver function tests were normal apart from a mildly elevated GGT of 61 U/L. A chest radiograph demonstrated apparent widening of the mediastinum at the level of the aortic arch and a wedge compression deformity of the T4 vertebral body. There were no lung parenchymal abnormalities.
The initial impression was of tuberculosis or malignancy.
Magnetic resonance imaging of his spine showed moderate compression of the body of T4 with reduced signal in the marrow of T4, and to a lesser extent T3 and T5 (Fig. 1). There was involvement of both of the T4 pedicles. Abnormal soft tissue thickening was present in the prevertebral and paravertebral mid-thoracic region. There was soft tissue encroachment on the extradural space resulting in moderate deformity and compression of the spinal cord at the T4 level. There was no involvement of the intervertebral discs. Malignancy was thought to be more likely than infection.
A computed tomography (CT) guided vertebral biopsy was nondiagnostic so the patient proceeded to an open biopsy. Pus was seen as the muscle was stripped off the posterior elements of T4. A vertebrectomy and stabilization was required. The pus had not traversed the intervertebral discs. Gram stain of this material revealed gram-positive cocci and gram-negative bacilli. Histology demonstrated acutely and chronically inflamed granulation tissue with marked osteoblastic rimming, osteoclastic activity, reactive new bone formation and bone remodelling. No granulomas or acid-fast bacilli were seen. The appearances suggested chronic osteomyelitis and there was no evidence of malignancy.
Initial treatment was with intravenous (IV) flucloxacillin, gentamicin, and metronidazole. Culture of the surgical specimens grew Actinomyces israelii, Eikenella corrodens, Bacteroides ureolyticus and a microaerophilic gram-positive coccus. With the diagnosis now of vertebral actinomycosis and because all the organisms demonstrated susceptibility to penicillin, the patient was changed to benzylpenicillin 2MU (1.4g) IV every 4 hours. He received this for 6 weeks and was then given oral phenoxymethylpenicillin 1.0g qid and probenicid 0.5g bd for a total antibiotic duration of 12 months.
A CT scan of his chest performed after surgery found no significant hilar or mediastinal lymphadenopathy, and no focal pulmonary consolidation.
The neurological symptoms steadily improved with complete resolution by the end of treatment. There was no evidence of relapse 6 months after treatment was stopped. He was reviewed by the dental service and 12 teeth were extracted.
Actinomycosis involving the vertebral column is uncommon, according to a review in 1978 documenting a total of 118 cases . While extensive vertebral involvement is common, associated spinal cord compression has been reported in the English medical literature in only 13 cases, including this one [3–11]. The details of these cases are shown in Tables 1 and 2. The rate of spinal cord compression in those with vertebral column actinomycosis has never been precisely established, but the 118 cases of vertebral actinomycosis in the above review  contained four of the patients we have recorded in Table 1, giving a rate of approximately 3%. This compares to 6 to 17% in pyogenic bacterial osteomyelitis [12,13] and 39 to 50% in tuberculous osteomyelitis [14,15]
The vertebral column is usually involved by contiguous spread from another focus in the head and neck, or in the chest or abdomen [5,16,17]. Although hematogenous dissemination has been reported, it is considered rare [17–19]. Because there was no focus of infection found in our patient’s chest we suggest that hematogenous spread occurred from his dental caries.
The 13 patients with spinal cord compression had a median age of 41 (range 22–63) years. Eleven of these patients were male and one was female. Actinomycosis generally is more common in males, with reported ratios of males to females of between three and five to one [18,19].
Initial symptoms preceded admission by a median of 2 months (range 2 weeks to 2 years). Back pain was reported in all but one case and fever and weight loss were relatively uncommon. Draining sinuses and subcutaneous abscesses are a common and characteristic feature of actinomycosis [16–19] and were present in half of these patients. Three of them had had recent dental surgery and another had severe dental caries.
There are a number of characteristic radiological features of vertebral actinomycosis, [5,16,17,19] some of which were seen in the patients with spinal cord compression (Table 1). Several adjacent vertebrae are commonly affected and vertebral collapse is uncommon. The intervertebral discs are rarely affected, and the posterior elements of the vertebrae (namely the pedicles, lamina, and the spinous and transverse processes, as well as the heads of corresponding ribs) are often involved. The affected bone has been described as having a moth-eaten or honeycomb appearance [5,19]. Periosteal new bone formation is common .
Radiological appearances may be useful in distinguishing actinomycosis from other disorders affecting the vertebrae. Tuberculosis often involves more than two vertebrae and adjacent discs; collapse is common and the posterior elements are usually spared . Pyogenic bacterial osteomyelitis usually results in narrowing of the intervertebral disc and erosion of the adjacent vertebral endplates and the posterior elements are uncommonly involved . Metastatic malignancy can be confused with actinomycosis (as in our patient) because the posterior elements, especially the pedicles, are involved in most cases . Metastases may involve single or multiple vertebrae and vertebral body collapse is much more frequent than in vertebral actinomycosis.
In six cases there was culture and definitive identification of the responsible pathogens. Five were due to A.israelii and one to A.meyeri. In four cases, additional organisms were cultured.
Given the rarity and unusual radiological appearance of vertebral actinomycosis when compared to other causes of osteomyelitis, this diagnosis is often not considered until diagnostic specimens sent to microbiology or pathology departments provide the unexpected answer. When an initial impression was stated, six patients were thought to have either malignancy or tuberculosis.
The first three reported cases did not receive antibiotic therapy and all died (Table 2). There is no information regarding treatment or outcome for the fourth patient. The nine most recent cases from 1951 onwards were all treated with antibiotics. The total median duration of therapy was 7.5 months, ranging from 6 to 14 months. Of these nine patients, all had a significant improvement in their neurological signs, and five made a full recovery. There were no reported relapses or deaths. These outcomes compare favorably with spinal cord compression with an associated neurological deficit due to pyogenic or tuberculous vertebral osteomyelitis. Previous series of pyogenic vertebral osteomyelitis resulting in a neurological deficit report no improvement or deterioration in 17 to 25% of patients. There was a mortality rate of 15 to 17% [23,24]. A large study in 1978 of tuberculous vertebral osteomyelitis reported persisting deficit in 19% and a mortality rate of 14% in the subset of 42 patients with a neurological deficit . A more recent, smaller study in 1995 that included 10 patients with a neurological deficit due to tuberculous vertebral osteomyelitis reported neurological improvement in all patients except for two that died of unrelated causes .
We have reported a case of thoracic vertebral actinomycosis, probably due to bacteremic spread from dental caries, resulting in spinal cord compression. There were no obvious clinical symptoms or signs that led us to consider this diagnosis. Radiological clues that were in keeping with the diagnosis included the involvement of three vertebral bodies together with the pedicles of T4 and sparing of the intervertebral discs. The identification of A. israelii from the operative specimens was unexpected. Our patient made a complete recovery following 12 months of treatment with benzylpenicillin and then phenoxymethylpenicillin.
Vertebral actinomycosis with associated spinal cord compression is rare, with the majority of reported cases affecting the thoracic vertebrae. Alternative diagnoses, such as malignancy or tuberculosis, were often considered before the microbiology or pathology departments provided the correct diagnosis. Pointers to the diagnosis include the presence of draining sinuses or subcutaneous abscesses, the ability of this infection to cross anatomic boundaries, and the distinctive radiological appearances listed above. Prolonged treatment with antibiotics, most often benzyl penicillin followed by phenoxymethylpenicillin, and in some instances surgery, resulted in significant, if not complete, improvement in the neurological deficit of all these patients.
1. Smego RA, Foglia G. Actinomycosis. Clin Infect Dis 1998; 26:1255–63.
2. Lewis RP, Sutter VL, Finegold SM. Bone infections involving anaerobic bacteria. Medicine 1978; 57:279–305.
3. Parker CA. Actinomycosis and blastomycosis of the spine. J Bone Joint Surg (Am) 1923; 21:759–77.
4. Krumdiek N, Stevenson L. Spinal epidural abscess associated with actinomycosis. Arch Pathol 1940; 30:1223–6.
5. Cope VZ. Actinomycosis of bone with special reference to infection of the vertebral column. J Bone Joint Surg (Br) 1951; 33B:205–14.
6. Lane T, Goings S, Fraser DW, et al. Disseminated actinomycosis with spinal cord compression: report of two cases. Neurology 1979; 29:890–3.
7. Kannangara DW, Tanaka T, Thadepalli H. Spinal epidural abscess due to Actinomyces israelii.
Neurology 1981; 31:202–4.
8. Muller PG. Actinomycosis as a cause of spinal cord compression: a case report and review. Paraplegia 1989; 27:390–3.
9. Oruckaptan HH, Senmevsim O, Soylemezoglu F, et al. Cervical actinomycosis causing spinal cord compression and multisegment root failure: case report and review of the literature. Neurosurgery 1998; 43:937–40.
10. Yung BCK, Cheng JCK, Chan TTF, et al. Aggressive thoracic actinomycosis complicated by vertebral osteomyelitis and epidural abscess leading to spinal cord compression. Spine 2000; 6:745–8.
11. Eftekhar B, Ketabchi E, Ghodsi M, et al. Cervical epidural actinomycosis. J Neurosurg 2001; 95:132–4.
12. Mader JT, Calhoun J. Osteomyelitis. In: Mandell GL, Bennet JE, Dolin R, eds. Principles and practice of infectious diseases. 5th
ed. Philadelphia: Churchill Livingstone; 2000: 1182–96.
13. Sapico FL, Montgomerie JZ. Pyogenic vertebral osteomyelitis: Report of nine cases and review of the literature. Rev Infect Dis 1979; 1:754–76.
14. Medical Research Council Working Party on Tuberculosis of the Spine, Seventh Report: A controlled trial of anterior spinal fusion and debridement in the surgical management of tuberculosis of the spine in patients on standard chemotherapy: A study in two centres in South Africa. Tubercle 1978; 59:79–105.
15. Rezai AR, Lee M, Cooper PR, et al. Modern management of spinal tuberculosis. Neurosurgery 1995; 36:87–98.
16. Meyer M, Gall MB. Mycosis of the vertebral column: a review of the literature. J Bone Joint Surg (Br) 1935; 17:857–66.
17. Lubert M. Actinomycosis of the vertebrae. Am J Roentgenol 1944; 51:669–76.
18. Bates M, Cruickshank G. Thoracic actinomycosis. Thorax 1957; 12:99–124.
19. Young WB. Actinomycosis with involvement of the vertebral column. Clin Radiol 1960; 11: 175–82.
20. Chapman M, Murray RO, Stoker DJ. Tuberculosis of the bones and joints. Semin Roentgenol 1979; 14:266–82.
21. Babinchak TJ, Riley DK, Rotheram EB. Pyogenic vertebral osteomyelitis of the posterior elements. Clin Infect Dis 1997; 25:221–4.
22. Laredo J-DL, el Quessar AE, Bossard PB, et al. Vertebral tumours and pseudotumours. Radiol Clin North Am 2001; 39:137–63.
23. Eismont FJ, Bohlman HH, Soni PL, et al. Pyogenic and fungal vertebral osteomyelitis with paralysis. J Bone Joint Surg 1983; 65A:19–29.
24. Abramovitz JN, Batson RA, Yablon JS. Vertebral osteomyelitis: the surgical management of neurological complications. Spine 1986; 11: 418–20.