Aspergillus osteomyelitis of the lumbar spine complicated with orbital apex syndrome: A potential role of the Batson's plexus in disease propagation

Abstract

We report a rare case of orbital apex syndrome following epidural steroid injections of the lumbar spine in an immunocompetent individual with osteomyelitis and discitis caused by Aspergillus fumigatus. We suspect that the craniospinal venous system, also known as the Batson's plexus, was the main route for steroid-facilitated disease propagation from the spine to intracranial structures.

1. Introduction

Aspergillus species are ubiquitous opportunistic fungi found in the environment. Invasive aspergillosis generally occurs in immunocompromised hosts, and is most commonly caused by Aspergillus fumigatus [1]. Infection usually develops through inhalation of Aspergillus conidia into the lungs or sinuses [2]. Osteomyelitis is rare but may develop by direct extension from a visceral focus, systemic dissemination via bloodstream invasion, traumatic inoculation, or contamination at the time of surgery [3–5]. Lumbar vertebral bodies and intervertebral disks are the most common sites involved [3] and affected patients generally present with back pain [3,5]. Orbital apex syndromes (OAS) are characterized by visual loss and ophthalmoplegia as a result of a variety of inflammatory, infectious, neoplastic, traumatic, and vascular conditions that ultimately lead to cranial nerve damage [6]. A number of infectious diseases involving the central nervous system, paranasal sinuses, and periorbital structures have been associated with OAS [6].

Here we describe an unusual case of an immunocompetent individual who developed A. fumigatus osteomyelitis/discitis of the lumbar spine complicated with right OAS following epidural steroid injections. Although OAS due to aspergillosis has been previously described [7–14], it is usually associated with paranasal sinuses involvement. To the best of our knowledge, this is the first case report of OAS resulting from disease propagation from a distant infectious focus. We suspect that fungal spread via the Baston's plexus venous system was the mechanism of disease extension from the spine to the orbital apex.

2. Case report

A 71 year old man presented to an outside hospital in February 2010 with a six-week history of severe low back pain at rest associated with numbness of the lower extremities. The patient had received an epidural injection with 3 ml (15 mg) of dexamethasone mixed with 2 mL of bupivacaine 0.25% at the level of L5–S1 three days prior to presentation. On day two of admission a second epidural steroid injection was administered. Twenty four hours later the patient developed a fever of 38.3 °C and worsening pain. Magnetic resonance imaging (MRI) of the lumbosacral spine showed evidence of L2–3 and L3–4 discitis and adjacent L3 and L4 vertebral osteomyelitis (Fig. 1). Erythrocyte sedimentation rate (ESR) and C-reactive protein at that time were 117 mm/h and 17.2 mg/L, respectively. The patient was empirically treated with intravenous (IV) ceftriaxone 2 g and vancomycin 1 g every 12 h. On day six, a computer tomography (CT)-guided bone biopsy of L3 vertebral body was performed. Gram stain showed rare leukocytes and no organisms. Bacterial culture, acid fast bacilli (AFB) and KOH smears were negative. Tuberculin skin testing and interferon-gamma release assay (Quantiferon TB-Gold) were also negative.

Fig. 1.

Sagittal view of fat suppressed post-gadolinium T1 weighted MRI of lumbar spine before (left panel) and after anti-fungal therapy (right panel). Findings are compatible with osteomyelitis of the end plates of L2–L3 and L3–L4 (arrows) with extradural, paraspinal and muscle involvement (arrow heads). Significant improvement of the osteomyelitis and extradural changes is seen at week 20.

On day twelve, the patient developed new onset diplopia with right periorbital pain, ptosis and decreased vision. Magnetic resonance angiography of the neck and brain showed no evidence of cavernous sinus thrombosis or arteritis. The MRI of the brain showed subtle changes in the right sphenoid bone and periocular muscles.

Upon transfer to our institution, almost two weeks from his initial presentation, the patient had no fever and his vital signs were stable. On physical exam there was right upper eyelid ptosis, 4 mm fixed pupil, with impaired eye abduction and adduction. Vision on the right visual field was significantly reduced. Fundoscopic examination of the right eye revealed mild venous congestion without papilledema. Left eye examination was normal. The lumbar spine was tender to palpation diffusely. White blood count on admission was 13,500 cells/mm³ with 86% neutrophils. Repeat MRI of the brain and orbits revealed an inflammatory process involving the dura of the lesser wing of the right sphenoid bone extending into the right orbital apex (Fig. 2). CT of the chest showed a new small area of consolidation in the left lower lobe.

Fig. 2.

Coronal view of fat suppressed post-gadolinium T1 weighted MRI of the orbits before (left panel) and after (right panel) anti-fungal therapy. Imaging shows an inflammatory process involving the dura of the lesser wing of the right sphenoid bone (arrow); focal myositis changes in the right orbital apex and inferolateral border of the right optic nerve (arrow heads) with compression of third and sixth cranial nerves that pass through the upper and lower heads of the lateral rectus muscle. Significant improvement is seen at week 7.

We started dexamethasone (4 mg IV every 6 h), continued vancomycin and switched ceftriaxone to cefepime (2 g IV every 8 h) plus metronidazole (500 mg IV every 8 h). Oral voriconazole 300 mg twice daily was added empirically. Patient's vision improved markedly within 48–72 h of this regimen. Periorbital pain resolved and back pain improved. One week later, diplopia resolved and the patient recovered the abduction and adduction movements of the right eye. Fungal culture (performed on brain heart infusion, mycocell and sabouraud dextrose agars) from the bone biopsy done at outside hospital grew A. fumigatus after 10 days of incubation. At this time (day 16 after initial presentation), antibiotic therapy was stopped, dexamethasone was tapered down and the patient was discharged home on oral voriconazole 300 mg twice daily. Vision returned to normal and right ptosis resolved one month later. Nasal and sputum fungal cultures were negative. Mycobacterial bone culture remained negative. Repeat MRI of the lumbar spine and orbits seven weeks after initiation of voriconazole showed marked improvement (Fig. 2). The left lower lobe consolidation resolved. Follow-up CT of the sinuses at week 18 of therapy was normal. Similarly, MRI of the lumbosacral spine at week 20 of therapy showed continued improvement with development of discogenic sclerosis (Fig. 1).

Following course was complicated with adrenal cortical insufficiency requiring long-term low dose dexamethasone 0.75 mg daily. Voriconazole was discontinued at week 24 of therapy. Two weeks after discontinuation of voriconazole, the patient developed worsening back pain that required admission for pain control. CT of the spine was unchanged. Repeat biopsies from the lumbar spine (L2–L5 intervertebral disks) showed no evidence of osteomyelitis. Pathology and both fungal culture and PCR for A. fumigatus were negative on this repeat biopsy specimen confirming successful eradication of the infection with antifungal therapy. Currently, 14 months after the end of therapy the patient has a stable chronic back pain not interfering with his activities of daily living. His visual symptoms never recurred.

3. Discussion

Vertebral aspergillosis in immunocompetent individuals has been previously reported [5,15,16]. Although a considerable number of these patients had history of trauma or spinal surgery, specific risk factors remain unknown. Our patient had no prior neurosurgery and extensive laboratory work-up (including serum and urine protein electrophoresis, quantitative immunoglobulins, HIV antibodies, nitroblue tetrazolium test, and CD4 lymphocyte count) failed to reveal the presence of an underlying immunodeficiency. The presence of an occult malignancy was also considered; however, a recent colonoscopy and prostatic specific antigen tests were normal and upon follow-up the patient had gained weight. Thus, immunosenescence associated with old age might have been this patient's only risk factor for invasive aspergillosis. Aging has been associated with impaired innate [17,18] and adaptive immune responses [19,20]. Cell-mediated immunity seems to be mainly affected by impaired ability of elderly hosts to generate naïve T cells and to mount robust memory T cell responses [21,22].

Hematogenous route has been reported as the main mode of acquisition of vertebral aspergillosis [4,5]. We suspect that the lungs may have been the initial route of infection as suggested by the presence of a new consolidation on CT scan, followed by hematogenous dissemination of A. fumigatus to the spine. Direct inoculation from epidural injections has been reported [5], and it constitutes another possible route of infection in this patient as his symptoms clearly worsened after receiving the epidural injections. Contiguous spread from the lungs is less likely as it generally affects the thoracic rather than the lumbar spine [16].

Spread of an ongoing infection in the spine to the orbital apex was likely facilitated by the injected corticosteroids into the epidural space. In the absence of clinical or radiological evidence of sinusitis, we hypothesized that Aspergillus might have disseminated from the spine to the orbital apex via the craniospinal venous system or Batson's plexus, a valveless venous system that surrounds the vertebral column connecting the intracranial veins and sinuses with the vertebral and pelvic venous plexuses. Because of its distribution and valveless nature, the Batson's plexus provides a unique vascular route for the spread of tumors, infection or emboli [23]. For example, the Batson's plexus has been implicated in egg embolism and worm migration to the central nervous system in neuroschistosomiasis [24,25]. Similarly, the Batson's plexus has been proposed to be the main route of infection in cases of spinal tuberculosis in children [26] and, more recently, in a case of epidural abscess caused by group B Streptococcus in a patient with a rectal ulcer [27]. The suspicion of fungal dissemination to the orbital apex is also supported by the fact that our patient's orbital symptoms were preceded by back pain. Furthermore, his rapid clinical and radiographic improvement with antifungal therapy strengthen our hypothesis, making Tolosa–Hunt syndrome, an idiopathic inflammatory syndrome of the orbital apex [10,28], less likely.

Reversal of underlying predisposing conditions, medical therapy and surgery should be considered during treatment of invasive aspergillosis. Currently, voriconazole is considered the drug of choice for the treatment of invasive aspergillosis [29], and it has been successfully used either alone or in combination with surgical debridement [5,29]. Although most patients with vertebral aspergillosis required surgical debridement in a recent case series [5], this was not necessary in our patient because of the prompt and marked clinical recovery with medical therapy alone.

Adrenal cortical insufficiency has been reported with disseminated fungal infections [30–31]. However, imaging of the adrenal glands was repeatedly normal on CT scan of the abdomen. Voriconazole related adrenal insufficiency has been reported in less than 2% of patients [32] and may have been the etiology in this case in addition to receiving high dose corticosteroids.

In conclusion, OAS can occur as a result of fungal spread from a distant infectious focus, as described in this patient with Aspergillus osteomyelitis of the lumbar spine. Epidural steroid injections can favor spread of infection to the orbital apex via the Batson's plexus which constitutes an important route for intracranial disease propagation. If given early, prolonged therapy with voriconazole alone may obviate the need for surgical debridement.

Consent

Written informed consent was obtained from the patient prior to publication of this case report. A copy of the written consent is available for review by the Editors upon request.

Competing interests

None declared.

Authors' contribution

JFC, SV and MEK collected the clinical and laboratory data and participated in the drafting and edition of the manuscript. MT was responsible for acquisition and interpretation of radiological data. All the authors read and approved the final version of the manuscript.