Anterior cervical discectomy and fusion (or disk arthroplasty) has a low complication rate, with a reported infection rate lower than 3%.1,2 Most surgical site infections (SSI) occur in the early postoperative period. Delayed SSI, defined as occurring after 3 months from surgery, is a very rare complication.3 We report a case of delayed prevertebral cervical abscess diagnosed 2 years after two-level cervical arthroplasty, discuss the diagnosis and etiology, and review the literature.

A 55 years-old male consulted for a right cervical mass that had been slowly growing for eight weeks. Past medical history included high blood pressure, hyperlipidemia, smoking and alcoholic hepatopathy. Two years before, the patient was treated of disk herniation in another center, with C4–C5–C6 discectomy and arthroplasty (M6-C, Spinal Kinetics, Sunnyvale, CA, USA), with an uneventful postoperative period.

Physical examination revealed a fluctuant and tight right laterocervical mass, without fever, dysphagia, dyspnea or neurological deficits. Estimated sedimentation rate (ESR) was 20mm/h, C-reactive protein was (CRP) <1mg/dL and white blood cell (WBC) count was 6500/mm3 with 60% of neutrophils. Human immunodeficiency virus test was negative. Flexible laryngeal and esophageal endoscopy revealed an extrinsic laryngeal constriction but no intraluminal pathology. CT scan showed a right cervical prevertebral hypodens mass with mild peripheral iodine contrast enhancement at C4–D2 level (Fig. 1). MR revealed a T1WI hypointense and T2WI and STIR hyperintense mass (Fig. 1). The barium swallow did not show pass of contrast into the cavity (Fig. 2). Our primary diagnosis was delayed SSI because of the patient's risk factors. However, due to the slow growing of the collection, we could not exclude tuberculosis or post-surgical pseudomeningocele because of the lack of systemic inflammatory reaction and MR findings.

Fig. 1.

CT and MR. (A) Iodine contrast-enhanced CT scan, showing a right cervical prevertebral hypodense mass with mild peripheral contrast enhancement. (B) Sagittal cervical spine MR, showing a T1 hypointense mass. (C) Sagital cervical spine MR, showing a T2 hyperintense mass, suggestive of either cerebrospinal fluid collection or abscess.

(0.13MB).

Fig. 2.

Barium swallow. Barium swallow, showing left displacement and constriction of the esophagus, with no pass of contrast into the cavity.

(0.11MB).

Finally, owing to slow but continuous mass enlargement, an ultrasound guided puncture was performed, obtaining pus, but it did not provide enough evacuation neither microbiological diagnosis. Consequently, we indicated surgical exploration.

Intraoperatively, we found a capsulated cavity containing pus that extended to the prevertebral space in contact with the prostheses, without any associated tracheal or esophageal lesion. After abscess debridement, prostheses were removed and substituted for iliac crest bone graft (Fig. 3), and rifampicin was instilated.

Fig. 3.

Pre- and intraoperative pictures of the abscess. (A) Lateral view. (B) Anterior view. (C) Aspiration of abscess content.

(0.11MB).

We actively looked for a cerebrospinal fluid leak, aided by Valsalva maneuver, but we could not demonstrate it. In order to rule out a pseudomeningocele, we performed beta-2-transferrine test. It yielded a positive result but, as later discussed, we assumed it as a false positive. The bacterial culture and polymerase chain reaction analysis (PCR) were positive for Propionibacterium acnes, so antibiotic was adjusted from vancomycin and meropenem to amoxicillin-clavulanate 1g every 8h for 6 weeks. The postoperative period was uneventful. At 6 months follow-up, the patient remained asymptomatic.

Anterior cervical discectomy is a common procedure with a low infection rate, being most of them during the early postoperative period.1,2 In our literature review we found 7 cases of delayed SSI, defined as infection occurring after 3 months from surgery.3

Bhise4 reported a patient with fever and pus-discharging sinus over the posterior triangle of the neck, without any cavitation around the fused segment, and suggested a low-grade infection at implant site as the probable etiology. Christiano5 reported a case that presented with dysphagia and fever secondary to a prevertebral abscess, and stated the cause of infection was elusive. Hur6 presented a case of neck pain and progressive spinal cord syndrome due to an epidural abscess and proposed micro injury of esophagus due to implants as a causal factor. Jin7 reported a patient who consulted for blood-tinged sputum and a laryngeal mass, CT and MR revealed a prevertebral abscess associated to osteomyelitis, and they reported the cause of infection as elusive. Kuriloff8 published a case of a patient who consulted for dysphagia, cervical inflammation and fever, whose CT revealed a prevertebral abscess, and that was secondary to an esophageal fistula. Tsitsopoulos9 presented a patient with dysphagia and pus draining from incision due to a fistula from prevertebral space, and suggested it may have originated from the newly formed bone at the fusion site. Violon10 reported a case of neck pain and dysphagia due to prevertebral abscess and stated an insidious esophageal wall perforation might be evoked as the possible source of the infection. All cases had a good outcome after surgical revision and antibiotherapy, except the patient reported by Hur,6 who died due to sepsis. See Table 1 for a detailed review of these reported cases.

In these reported cases, the predominant microbiological isolation was Streptococcus spp., associated to 1 case of Veilonella spp. and 2 of Peptostreptococcus spp., which are members of the salivary and gut microbiome.11

We could not find a common pattern between clinical manifestation, blood test results, type of implant and associated osteomyelitis in these patients with delayed SSI.

Different causes of delayed SSI are proposed among reported cases: low grade infection at implant site, microinjury of esophagus, definite esophageal fistula, fistulation to skin from the newly formed bone at fusion site, but infection was also reported as elusive in 2 cases.

Regarding treatment options, in all cases except in Tsitsopoulos’,9 surgical debridement and removal of implants was performed, followed by antibiotic adjusted by microorganism culture.

To our knowledge, this is the only report of a delayed cervical spine SSI manifesting as a slowly growing cervical mass as the only clinical manifestation without any associated inflammatory findings.

The microbiological isolation of P. acnes, confirmed by PCR assay, implies some discussion. First, it is a member of skin and gut floras,11 which could favor the hypotheses of either latent implant contamination or acquired and spontaneously healed esophageal microperforation. Second, it is known as a microfilm creator,12 which could support the latent implant contamination hypothesis, and favors surgical removal of implants. Third, the patient underwent arthroplasty instead of fusion, which raises the question if the different implant properties could be a risk factor for a delayed implant-related SSI by biofilm-producing bacteria. Finally, it has a slow growth,12 which would justify this unusual course for a cervical abscess.

The beta-2-transferrin test is an elementary tool in the diagnosis of cerebrospinal fluid leak.13 MR findings raised the possibility of a spontaneously closed and subsequently infected pseudomeningocele. Therefore, we performed beta-2-transferrin test in the sample of purulent material. We obtained a positive result, which lead to strict control of meningeal irritation signs or cerebrospinal fluid fistula. Nevertheless, the lack of meningeal symptomatology, the non-confirming imaging tests, the intraoperative findings and the neuraminidase secretion of most of P. acnes strains,14 a known cause of false-positive test due to transferrine enzymatic processing,13 lead us to assume it as a false positive.

Acquired esophageal/tracheal microperforation, latent clinically-silent implant infection and bacteremia15 have been described as possible causes of delayed SSI in the cervical spine. In this case, micro-trauma related and spontaneously healed esophageal micro-perforation could not be completely ruled out. Nonetheless, due to P. acnes slow growth and microfilm formation,12 a latent implant infection was the infection cause we assumed.

We report an illustrative case of a delayed cervical paravertebral abscess following uneventful cervical disk arthroplasty. Our patient presented with a slowly growing paravertebral abscess appearing 2 years after cervical surgery without any inflammatory parameters. Delayed SSI is a rare but potentially severe complication after anterior cervical discectomy. Revision surgery is almost always required followed by long-course antibiotic treatment.

The manuscript submitted does not contain information about medical device(s)/drug(s). No funds were received in support of this work. No benefits in any form have been or will be received related directly or indirectly to the subject of this manuscript.

The present research has not received any specific scholarship from public, commercial, or non-profit agencies.