Spondylodiscitis mainly concerns the lumbar region, being CSD rare and accounting for about 10% [8] of all spondylodiscitis with an estimated incidence of CSD of 6%[12]. In the current series, it was 8.9% (70/786 cases). On the contrary, Lu et al[18] recorded a higher incidence of 19% of cases that might be due to a large number of patients with intravenous drug abuse; the latter is a well-known risk factor for CSD[10, 21]. In our series, intravenous drug addiction was recorded in 17/70 (24.28%) cases.
CSD is usually caused by the spread of bacteria or other microorganisms from other body infection foci. The most common cause of CSD is bacterial, typically resulting from one of the following mechanisms:
1. Hematogenous spread represents the most common infection way in which bacteria come from a remote infection, such as the skin, urinary tract, heart, or respiratory system. From this distant site of infection, bacteria can disseminate by vascular pathways (spinal posterior arteries, segmental arteries, venous plexus of Batson…)[17, 20].
2. Contiguous spread from neighboring structures – such as the ear-nose-throat area and soft tissues of the neck – going directly to the cervical spine. This mechanism occurs often secondary to surgeries of the ENT area, such as pharyngolaryngeal operations.
3. Following surgery or invasive procedures concerning the cervical spine or other intradiscal invasive procedures such as discographies[16, 34]. Infections can be introduced during the procedure or may develop later due to impaired wound healing. This mechanism will lead to a microbiological profile similar to that of cervical prosthesis infections and is particularly relevant at the cervical level where a great number of cervical disc surgeries, as well as neighboring structures surgeries, are performed.
The most common bacterial pathogens responsible for cervical spondylodiscitis, notably resulting from hematogenous dissemination, include Staphylococcus aureus and Streptococcus species. However, other bacteria including fungal infections can also lead to this condition. Though rare in developed countries, Mycobacterium tuberculosis is involved in as high as 20% of cases, being on the frontline of osteoarticular infections worldwide[5]. Brucella is nowadays rarely involved in developed countries such as France as well[9]. In cases secondary to surgical procedures, germs involved are mainly cutaneous bacteria (coagulase-negatives Staphylococcus and Corynebacterium notably). Gram-negative bacilli, including Pseudomonas aeruginosa, are more rarely in cause. A large variety of other germs have been the object of case reports[1, 22, 23]. An underlying pathology such as one causing immunosuppression (HIV, diabetes, chemotherapy, drug abuse) can favor the occurrence of CSD and modify the microbiology of this affection.
Different studies have suggested that CSD is likely to have another synchronous non-contiguous infection in another distant region of the spine. This risk is not negligible, as up to 47% of patients in the Shousha et al[14] series suffered from this phenomenon. This might be in favor of the systemic nature of the disease in immunocompromised patients. In our current series, endocarditis was identified in 12 cases and prostatitis in 4. The latter suggests that any patient presenting with a spinal infection should not only undergo an MRI scan of the entire spine but also other investigations including cardiac ultrasound and/or total body (TB) CT scan and in selected cases either TB scintigraphy or PET scan depending on the associated presenting symptoms.
Surgical management of cervical spondylodiscitis is widely based on 2 main principles:
- firstly, identification of the microorganism involved and/or to decompress the spinal cord;
- Secondly, reconstruction of the resulting defect and achievement of a stable construct with or without instrumentation.
Going through the literature many studies are found implementing a complex and aggressive surgery involving cervical corpectomies with anterior fusion, and posterior arthrodesis with or without bone graft[11, 12, 14, 18, 19, 25, 26, 29]. On the other hand, such aggressive management can often result in an increased morbidity and mortality rate. Shousha et al[27] reported in their study various complications and specifically 3 patients died postoperatively due to septicemia, 1 experienced a construct metal failure, and 1 the esophagus perforation. An esophagus perforation was also reported by Schimmer et al5 and similar results have been reported by Stone et al[29].
The mortality rate after CSD is not well documented in the literature. Karadimas et al[14] reported a 12% mortality rate in their series, and Heyde et al[12] reported a mortality rate of 5%. On the contrary, in our series, the mortality rate was zero percent which is better than previously reported in pertinent literature. Similar results were also reported by Muzii et al[21] in 2006. The authors demonstrated in the management of 8 patients that a limited debridement associated with drainage, washing, and targeted antibiotic therapy without instrumentation could have very good clinical evolution and radiological fusion. They used a small catheter to wash out epidural collections by a minimal opening of the posterior longitudinal ligament. Finally, they did not record any cervical secondary deformity.
On the opposite, Včelák et al[32] in their series of 21 patients, in the acute phase, preferred anterior debridement with epidural empyema evacuation without anterior column reconstruction and posterior instrumented fusion in case of secondary kyphosis, and in the chronic phase, radical anterior debridement with titanium implant reconstruction instead improved the kyphosis from 15.9° before surgery to 6.1° at 1 year FU. No patient reported neurological deterioration postoperatively.
Instability due to infection may be defined as a 50% reduction in vertebral body height and/or vertebral angulation greater than 20 ̊ and it might be iatrogenic, occurring after large decompression and debridement[4]. The presence of an infection in the past was considered a contraindication to bone grafts and instrumentation. Many authors have instead demonstrated and recommended primary reconstruction, as the risk of secondary implant infection is negligible if debridement of infected bone and especially targeted antibiotic therapy is systematically performed[3, 7, 19, 24, 31].
The MISA of CSD used in our series to identify the microorganism involved and/or to decompress the spinal cord assuring as much as possible mechanical stability resulted in excellent clinical results. We performed the cervical micro-discectomy and debridement to obtain specimens for bacteriological and histopathological analysis followed by targeted antibiotic therapy. In selected case, we performed posterior longitudinal ligament opening to drain epidural PUS collections and to decompress the spinal cord without applying either local antibiotic or osteosynthesis/arthrodesis. Neither spine instability nor kyphotic deformation was recorded in all patients and bony fusion occurred in all and was also confirmed radiologically by CT scan.
The neurological symptoms in our series were characterized by progressive tetra paresis in 11(15.71%) patients and a shock syndrome concerned 5 (7.14%) patients. In the literature, such figures are higher reaching 40% [12, 27]. However, clinical improvement with good recovery occurs in more than 50% of patients [25, 27] and some even reported a 100% improvement[3, 4, 6, 13, 28]. In our study, 100% of patients with neurological disorders improved after surgery being in line and even better compared to the results previously reported in the literature[25].
To the best of our knowledge, this represents the largest series, ever reported in the literature. Our cooperative and multicentric experience resulted in excellent patient outcomes in all the 5 participating centers.
Although the literature shows that the use of cervical hardware does not increase the risk of infection recurrence, we do believe that a MISA guarantees excellent clinical and radiological results avoiding at the same time all the complications related to any more extensive/aggressive approach as well as to any instrumentation failure.