Acta Neurochir (2014) 156:1189–1196 DOI 10.1007/s00701-014-2058-0
CLINICAL ARTICLE - SPINE
A retrospective study of 113 consecutive cases of surgically treated spondylodiscitis patients. A single-center experience Ehab Shiban & Insa Janssen & Maria Wostrack & Sandro M Krieg & Florian Ringel & Bernhard Meyer & Michael Stoffel
Received: 25 October 2013 / Accepted: 7 March 2014 / Published online: 27 March 2014 # Springer-Verlag Wien 2014
Abstract Background Recommendations for the operative treatment of spondylodiscitis are still a controversial issue. Methods A retrospective review identified 113 consecutive patients who underwent surgical debridement and instrumentation for spondylodiscitis between 2006 and 2010 at our department. Results The mean age at presentation was 65 years; 78 patients were male (69 %). Distribution of the inflammation was lumbar in 68 (60 %), thoracic in 19 (17 %) and cervical in 20 (18 %) cases. Six patients (5 %) had two concomitant noncontiguous spondylodiscitis foci in different segments of the spine. Epidural abscess was found in 33 patients (29 %). One hundred four patients (92 %) had pain. Neurological deficit Presentation at a conference: Global Spine Congress, Hong Kong, China April 2013 E. Shiban (*) : I. Janssen : M. Wostrack : S. M. Krieg : F. Ringel : B. Meyer : M. Stoffel Department of Neurosurgery, Klinikum rechts der Isar, Technische Universität München, Ismaninger Str. 22, 81675 Munich, Germany e-mail: [email protected] I. Janssen e-mail: [email protected] M. Wostrack e-mail: [email protected] S. M. Krieg e-mail: [email protected]
was found in 40 patients (35 %). In the thoracic and lumbar cases, dorsal instrumentation alone was considered sufficient in 26 cases; additional interbody fusion from the posterior was performed in 44 cases. A 360° instrumentation was performed in 22 cases. In the cervical cases, only ventral spondylodesis and plating were performed in eight cases, only dorsal instrumentation in five and 360° instrumentation in seven. Postoperative intravenous antibiotics were administered for 14.4±9.3 (mean ± SD) days followed by 3.2±0.8 (mean ± SD) months of oral antibiosis. Complete healing of the inflammation was achieved in 111 (98 %) cases. Two patients died because of septic shock, both with fulminant endocarditis. Pain resolved in all cases. Neurological deficits were completely resolved in 20 patients, and 14 patients had a partial recovery. Conclusion The results of our retrospective study show that surgical treatment of spondylodiscitis with a staged surgical approach (if needed) and a short 1-2-week period of intravenous antibiotics followed by 3 months of oral antibiotics is appropriate for most patients in whom conservative treatment has failed or is not advisable. Furthermore, surgical treatment of newly diagnosed spondylodiscitis might be recommended as an initial treatment option in many cases. Thereby the choice of fusion material (autologous bone, titanium, PEEK) seems less important. Keywords Spondylodiscitis . Spinal instrumentation . Antibiotic regime
F. Ringel e-mail: [email protected] B. Meyer e-mail: [email protected]
Introduction
M. Stoffel e-mail: [email protected]
Spinal infections can be a source of significant morbidity and mortality from spinal instability, neurological deficits and ongoing sepsis. Non-specific pyogenic spondylodiscitis is a rare clinical condition that can present with uncharacteristic chronic
M. Stoffel Department of Neurosurgery, Helios Klinikum, Krefeld, Germany
1190
back pain to a life-threatening course with rapidly progressing neurological deficits. The incidence of spondylodiscitis is approximately 1:250,000 per year. Pyogenic spondylodiscitis represents only 2–4 % of all cases of bone infection [16]. The incidence has been on the rise during the last few decades. This has been attributed to the increase in the elderly population with comorbidities, mainly diabetes mellitus, intravenous drug abuse and impaired immunocompetence (immunosuppressive drugs, human immunodeficiency virus, infections, chronic alcoholism) [2, 16, 45]. While the traditional method of managing spondylodiscitis has been prolonged immobilization, this regime is associated with significant complications and treatment failure. With increasing expertise in spinal instrumentation and fusion techniques, the approach for managing spondylodiscitis needs to be revisited. Hence, this study aims to assess the effectiveness of surgical intervention as a primary modality in spondylodiscitis.
Material and methods We performed a retrospective review of patients who underwent surgical treatment for spondylodiscitis between 2006 and 2010 at our department. One hundred twenty-three consecutive patients were identified. Ten patients were treated conservatively because of (1) severe comorbidities that caused the patients to have a high surgical mortality rate and (2) minor pain symptoms and regressive inflammation signs after antibiotic treatment. One hundred thirteen patients underwent surgical debridement and instrumentation. The following refers to these patients. Data were gathered through review of patients’ case notes, relevant imaging and electronic records. Gd-enhanced MR imaging of the whole spine was obtained in all cases. CT scans of the affected region were performed in all cases in order to better evaluate the extent of bone destruction. Our department of neuroradiology and the senior author examined all images. Patients were considered to have fever if their temperature was ≥ 38.5 °C. C-reactive protein (CRP) and complete blood count tests were performed in all cases using routine laboratory techniques. Neurological deficit was classified as mild when limited to problems involving one or two nerve roots. Severe deficits were classified according to the ASIA impairment scale. All patients were advised about both treatment modalities as conservative treatment with immobilization followed by mobilization in a brace or surgical treatment. However, for patients with a neurological deficit, abscess or failed conservative treatment, surgery was strongly recommended (Fig. 1). The surgical approach and method of spinal instrumentation were determined for each patient individually according to the comorbidities, location of infection and extent of bony destruction. For the cervical spine, the surgical approach was influenced foremost by the localization of the inflammation in
Acta Neurochir (2014) 156:1189–1196
Fig. 1 Number of patients with a neurological deficit, epidural abscess and/or failed conservative treatment for whom surgical treatment was strongly recommended
relation to the cervical spinal cord. Accordingly, a ventral approach was chosen if the main focus of the infection was ventral to the spinal cord; otherwise, a dorsal approach was chosen. Additional posterior surgery was performed in order to either enhance stability or provide additional decompression from the posterior. Options used were an anterior cervical discectomy or corpectomy and fusion using PEEK, titanium or autologous bone from the iliac crest depending on the surgeon’s preference with or without dorsal instrumentation or dorsal instrumentation alone. For patients with inflammation in the thoracolumbar spine, options used were dorsal instrumentation with or without interbody fusion. In cases without any destruction of the endplates and disc height reduction, then dorsal instrumentation without interbody fusion was performed. Interbody fusion was performed if the height of the affected disc was normal or increased. If an interbody fusion from the posterior was technically not feasible, then interbody fusion was performed through a ventral or lateral approach. Antibiotic treatment included a short period of broadspectrum intravenous antibiotic therapy, a tailoring according to the antibiogram and switching over to oral antibiotics at the time of clear improvement of clinical and laboratory parameters for a period of 3 months. Cure of inflammation was considered complete if patients exhibited no signs or symptoms of active localized infection by clinical examination and inflammatory markers at 3 months postoperatively. Neurological outcomes mentioned are at a minimum of 3month postoperative follow-up. To our knowledge, our study is the single largest series of spondylodiscitis treatment in which all patients underwent surgical debridement of infected disc space followed by spinal instrumentation.
Results The median age at presentation was 67 years (range 40–94). Seventy-eight patients were male (69 %); 104 patients (92 %)
Acta Neurochir (2014) 156:1189–1196
had pain. Neurological deficit was found in 40 patients (35 %); 36 patients had a mild deficit, and 4 were suffered from paraparesis. CRP was elevated in all patients. Median CRP value upon admission was 7.7 mg/dl, and it declined significantly in all cases upon discharge (Fig. 2). White blood cell (WBC) count was elevated only in 52 patients (46 %). Median WBC value on admission was 9.3 g/l, and it was significantly reduced at discharge (Fig. 3). Fever was diagnosed on admission in 24 patients (21 %) and resolved in all patients prior to discharge. Spondylodiscitis was confined to one level in 84 (74 %), two contiguous levels in 24 (21 %) and three contiguous levels in 5 (4 %) cases. Epidural abscess was found in 33 patients (29 %). Distribution of the inflammation was lumbar in 68 (60 %), thoracic in 19 (17 %) and cervical in 20 (18 %) cases. Six patients (5 %) had two occurrences of concomitant noncontiguous spondylodiscitis in different segments of the spine. The suspected source of infection was identified in 103 (91 %) of the cases. The causes were as follows: prior surgery in 56 (49 %), repetitive facet joint infiltration in 14 (12 %), diverticulitis in 9 (8 %), endocarditis in 9 (8 %), renal insufficiency and infection in 6 (5 %), wound healing disorders secondary to diabetes mellitus in 6 (5 %), parodontitis in 2 and parotitis in 1 case. No source of infection could be suspected in ten (9 %) cases (Fig. 4). Successful isolation of the organism from the infected space was possible in 82 patients (73 %). In 31 patients (27 %), organisms could not be isolated. Eighteen (58 %) of these 31 patients had antibiotic therapy prior to surgery. Staphylococci were encountered in 58 (51 %) patients. In 39 cases Staphylococcus aureus was identified. Multiresistant bacteria were found in 15 cases (13 %), and in 9 (8 %) cases of sporadic bacteria were identified. As mentioned above, the surgical approach was tailored for each patient depending on comorbidities, localization of
Fig. 2 Diagram showing C-reactive protein (CRP) levels throughout the hospital stay. CRP was elevated in all cases. Median CRP value upon admission was 7.7 md/dl, and it declined significantly in all cases upon discharge
1191
Fig. 3 Diagram showing white cell count (WCC) levels throughout the hospital stay. WCC was elevated only in 52 patients (46 %). Median WCC value at admission was 9.3 g/l, and it was significantly reduced at discharge
infection and extent of bone destruction. In the thoracic and lumbar cases, dorsal instrumentation alone was considered sufficient in 26 cases; additional interbody fusion from the posterior was performed in 44 cases (38 autologous bone and 6 PEEK cages); 360° instrumentation was performed in 22 cases (5 PEEK, 14 titanium and 3 autologous bone) (Fig. 5). In cervical disease, ventral spondylodesis was performed in eight cases (3 PEEK cages and 5 autologous bone from the iliac crest), ventral plating, dorsal instrumentation alone in five cases and 360° instrumentation in seven cases (4 PEEK cages and 3 autologous bone from the iliac crest) (Fig. 5). Surgery-related complications were seen in seven patients (6 %). Two patients had a postoperative epidural bleed, screw correction was indicated in three cases, and two patients had developed a wound-healing disorder. Further complications were observed in 12 patients (11 %), 2 patients developed pulmonary emboli, 4 patients developed pneumonia and 6 had a urinary tract infection requiring escalation of the antibiotic treatment. A broad-spectrum antibiotic, mostly clindamycin, was administered initially till the results of the antibiogram were available. An organism-sensitive specific antibiotic was initiated based on the antibiogram report. Postoperative intravenous antibiotics were administered for 14.4±9.3 (mean ± SD) days followed by 3.2±0.8 (mean ± SD) months of oral antibiotics. Patients in whom the organism was not isolated received empirical broad-spectrum antibiotics. All patients continued with oral antibiotics for 3 months without major complications. Forty-seven patients (41 %) reported bowel discomfort but did not need to discontinue the medication. Five patients (4 %) developed diarrhea, and the antibiotic was paused for 1 week. Two patients developed an a facial exanthema, and the antibiotic was paused for 2 days. None of the patients developed any other serious side effects. All patients
1192
Acta Neurochir (2014) 156:1189–1196
Fig. 4 Summary of the suspected source of infections
had a strict follow-up regime including weekly physical examination and blood analysis with their primary care physician. Final follow-up at 3 months was performed at our institution in all surviving patients. Pain resolved in all cases. From the 36 patients with a mild neurological deficit, 20 had a full recovery, 12 had a partial recovery, and 4 were unchanged postoperatively. From the four patients with a severe neurological deficit, two had a partial recovery and two were unchanged postoperatively. Septicemia and multiorgan failure occurred in two patients, leading to death. Both patients had a fulminant endocarditis. Spinal surgery was a prerequisite for further urgent cardiac surgery. Eradication of infection was ultimately achieved in all patients as documented by normalization of CRP and clinical findings. One patient had a relapse of the inflammation following dorsal instrumentation alone. This was followed by debridement and anterior interbody fusion upon relapse.
Fig. 5 Summary of the surgical procedures
Discussion Obviously, this study suffers from some limitations due to its retrospective nature. Due to the nature of our department as a spine center, most of our patients (62 %) were referred to us with a clear indication for surgery (Fig. 1). For the remaining 38 % of patients, both surgical and conservative treatments were an option, and conservative treatment might have achieved good results as well. However, we tend to recommend surgery because it allows rapid mobilization and shortens hospital stay. This is on the one hand a major limitation because most of the patients were treated surgically. On the other hand, this uniquely large consecutive series of a surgically treated cohort with spondylodiscitis combined with detailed written and electronic notes allows some substantial conclusions.
Acta Neurochir (2014) 156:1189–1196
Similar to previous studies, the patient population was mostly elderly and male. In this study, 69 % of the patients were male and 57 % over the age of 60. The higher incidence in the male population remains to be explained. The association with age is believed to be due to the higher incidence of comorbidities in this higher age group [11, 21]. Also in concordance with the literature, the distribution of the inflammation was mostly seen in the lumbar spine followed by the thoracic and cervical spine [11, 21]. Among the inflammatory markers we found CRP to be most reliable in affirming the diagnosis of infection as well as for monitoring the response to treatment. CRP was elevated in all patients. Median CRP value upon admission was 7.7 mg/dl and significantly declined in all cases upon discharge. CRP had returned to normal in all survivors at 3-month follow-up. These findings are similar to previously published literature [11, 12, 54]. WBC count was not very reliable as it was elevated in only 52 patients (46 %). Immunocompromised patients and patients over the age of 60 years were more likely to have normal WBC levels despite infection [6]. Clinical findings of spondylodiscitis can be non-specific. However, back or neck pain is very common, but could be absent in up to 15 % of cases [49]. In our series 92 % of all patients presented with pain. Fever is not frequently experienced and occurs in only 20–50 % of cases [38]. Fever was diagnosed on admission in 24 patients (21 %) and resolved in all cases prior to discharge. Similar to other studies, a neurological deficit was found in 35 % of cases and was more likely to be present with epidural abscess or delayed diagnosis [44]. The surgical approach was tailored for each case individually and a staged surgical approach was implemented. All patients underwent surgical debridement and instrumentation, whereas staged surgery (a combination of both a ventral and a posterior approach) was needed in 25 % and 35 % in the thoracolumbar and cervical cases, respectively. Postoperative intravenous antibiotics were administered until CRP significantly declined and clinical symptoms started to resolve. Thereby, a very short duration of intravenous antibiotic administration of 14.4±9.3 (mean ± SD) was sufficient. This was followed by oral antibiotic administration of 3.2±0.8 (mean ± SD) months. Using such an algorithm of a staged surgical approach if needed and short-term postoperative intravenous antibiotics and subsequent oral antibiotics for 3 months, complete healing of the inflammation was achieved initially in 98 %, and ultimately in all cases, marked pain reduction was observed in all cases and neurological improvement was observed in 88 % of cases. The conservative method of managing spondylodiscitis includes external immobilization by orthesis or bed rest and prolonged duration of parenteral antibiotic administration and analgesics. This method aims at achieving spontaneous healing and fusion of the infected spinal segment [27, 29, 30, 41, 46, 51]. However, reported failure rates following conservative treatment range from 12 % to 18 % [3, 8, 24,
1193
40, 52]. Both mortality and morbidity rates are similar following conservative or surgical treatment in most patients with spondylodiscitis [52]; this is however not true for patients with spondylodiscitis caused by drug multiresistant bacteria. In these patients morbidity and mortality rates of up to 38 % have been reported, mostly following conservative treatment [1]. Fifteen patients (13 %) in our study were infected by multiresistant bacteria. Of those, 12 patients (92 %) had full recoveries, and 1 patient died 3 weeks postoperatively because of fulminant endocarditis. Spinal surgery was a prerequisite for further urgent cardiac surgery. Neither controlled comparative trials nor standard guidelines are present for the optimal antibiotic and duration of treatment. Duration of intravenous antibiotics administered varied from 10 days [35], 3 weeks [9], 30 days [17] to 6 weeks [43, 45] in different series. In this study 14.4±9.3 (mean ± SD) days of intravenous antibiotic administration was relatively short, but achieved significant improvement in clinical and laboratory findings in all surviving cases (98 %). The duration of oral antibiotics administered ranges from 6 weeks [9, 43], 59 days [17] to 3 months [35]. The administration of parenteral and oral antibiotics tends to be longer with a mean of 21 days and 45 days respectively if the infected disc tissue is not eradicated [9]. An observational retrospective study comparing duration of treatment of less than 6 weeks in one group or more than 6 weeks in the other group showed similar results of recovery rates, relapse and death [48]. Although high-level evidence is lacking, there is a general consensus on discontinuing long-term oral antibiotics only after normalization of ESR, CRP and leucocyte count [17, 19, 35, 43]. Surgical strategies in treating spondylodiscitis vary considerably. Some favor radical debridement [4, 13, 16, 20, 25, 26, 32, 34, 35, 41, 43, 45, 47, 50]. Others, however, have reported excellent results using minimally invasive dorsal percutaneous pedicle-screw rod spondylodesis, with the aim to achieve immediate pain relief and patient mobilization, without eradication of infected intervertebral disc space [9, 17, 22, 35, 39]. The specific surgical approach for debridement also varies. Some authors favor anterior approaches for debridement as it allows direct access to the infected intervertebral disc space, paraspinal abscess and stabilization by interbody graft [10, 14, 36, 37, 45]. Anterior lumbar interbody fusion followed by anterior stabilization also allows for good restoration of lordosis and anterior column support [13, 14, 42]. Supplementation by posterior instrumentation provides better deformity correction, correction of sagittal alignment and faster rates of fusion [10, 15, 18, 24, 28, 41]. Two-stage approaches of initial anterior surgical debridement and fusion followed by delayed posterior instrumentation after a period of intravenous antibiotics has been practiced in order to prevent increased risk of infection after spinal instrumentation [10]. Fukuta et al. reported excellent outcomes following placement of posterior instrumentation in the first stage
1194
followed by anterior debridement and fusion in the second stage [15]. Satisfactory results have also been reported following single-stage anterior or posterior interbody fusion and pedicle screw fixation [31, 33, 37]. The most effective surgical approach—anterior, posterior or a combination of both, single stage or two stages—needs to be tailored on an individual case basis [16, 20, 26, 27, 34, 35, 47]. However, most reports suggest that single-stage debridement, fusion and internal fixation do not lead to higher rates of wound infection or hardware-related complications [45]. In our series we found staged surgery, if needed, to be sufficient in most cases. Primarily in the thoracic and lumbar cases, dorsal instrumentation alone was considered sufficient in 26 cases. In 66 cases substantial destruction of the endplates was present, and therefore interbody fusion was deemed necessary. When possible, this was performed in the same procedure from the posterior with either autologous bone material (n= 38) or PEEK cages (n=6). If interbody fusion from the posterior was not mechanically feasible, then an anterior approach (n=22) was performed in a second surgical procedure, mostly 1 week thereafter, allowing the patient to recuperate from the first operation. As shown, this rationale of staged surgery (if needed) proved to be sufficient in most cases. One patient had a relapse 1 month postoperatively following dorsal instrumentation alone. This was followed by debridement and anterior interbody fusion upon relapse. Use of instrumentation in treating spinal infections has been a subject of intense debate. Several authors have expressed their concern about the use of alloplastic implants at the site of active infection [5, 13, 14, 36, 37]. Bacteria tend to colonize stainless steel and polymethyl-methacrylate heavily, thereby reducing antibiotic effectiveness. However, titanium is less prone to bacterial adherence [7]. Interbody fusion using autologous bone graft has long been considered good practice to treat spondylodiscitis [33, 47]. In a retrospective review of 60 patients diagnosed with pyogenic spondylodiscitis, Pee et al. [43] performed a single-stage anterior debridement and fusion using either an autologous iliac bone strut or cage (titanium, titanium mesh, PEEK) followed by posterior pedicle screw fixation. In this study no difference in clinical (pain, functional disability) and radiological outcome (correction of segmental lordosis and fusion rate) was noted between the strut and cage group. However, the subsidence rate was higher in the strut group. Duration until subsidence was also shorter in the strut group than in the cage group. Several studies have demonstrated favorable results using the titanium mesh cage compared to autologous bone grafts for anterior interbody fusion in infected spines [18, 23, 31]. Similar to other reports, the use of PEEK cages in our series (N=19) showed successful outcomes in terms of clinical and laboratory parameters and radiological findings without an increase in relapse rates [4, 37, 43, 53].
Acta Neurochir (2014) 156:1189–1196
Conclusion The results of our retrospective study show that surgical treatment of spondylodiscitis with a staged surgical approach (if needed) and a short period of intravenous antibiotics of 12 weeks followed by 3 months of oral antibiotics is appropriate in most patients in whom conservative treatment has failed or is not advisable because of the presence of a neurological deficit or an epidural abscess. Furthermore, with relatively low rates of surgical complication and eradication of the inflammation in all cases, surgical treatment of newly diagnosed spondylodiscitis might be recommended as an initial treatment option in many cases. Therefore, the choice of fusion material (autologous bone, titanium, PEEK) seems less important.
Conflicts of interest The authors declare that they have no conflict of interest affecting this study. The study was completely financed by the Department of Neurosurgery.
References 1. Al-Nammari SS, Lucas JD, Lam KS (2007) Hematogenous methicillin-resistant Staphylococcus aureus spondylodiscitis. Spine (Phila Pa 1976) 15:2480–2386 2. Belzunegui J, Intxausti JJ, De Dios JR, Del Val N, Rodríguez Valverde V, González C, Queiro R, Figueroa M (2000) Haematogenous vertebral osteomyelitis in the elderly. Clin Rheumatol 19:344–347 3. Bettini N, Girardo M, Dema E, Cervellati S (2009) Evaluation of conservative treatment of non specific spondylodiscitis. Eur Spine J 1:143–150 4. Brase A, Ringel F, Stüer C, Meyer B, Stoffel M (2010) Debridement and fusion with polyetheretherketone implants in purulent spondylodiscitis: a clinical experience with nine patients. Acta Neurochir 152:2001–2004 5. Carragee EJ (1997) Instrumentation of the infected and unstable spine: a review of 17 cases from the thoracic and lumbar spine with pyogenic infections. J Spinal Disord 10:317–324 6. Carragee EJ (1997) Pyogenic vertebral osteomyelitis. J Bone Joint Surg Am 79:874–880 7. Chang CC, Merritt K (1994) Infection at the site of implanted materials with and without preadhered bacteria. J Orthop Res 12: 526–531 8. Cottle L, Riodan T (2008) Infectious spondylodiscitis. J Infect 56: 401–412 9. Deininger MH, Unfried MI, Vougioukas VI, Hubbe U (2009) Minimally invasive dorsal percutaneous spondylodesis for the treatment of adult pyogenic spondylodiscitis. Acta Neurochir 151:1451– 1457 10. Dimar JR, Carreon LY, Glassman SD, Campbell MJ, Hartman MJ, Johnson JR (2004) Treatment of pyogenic vertebral osteomyelitis with anterior debridement and fusion followed by delayed posterior spinal fusion. Spine 29:326–332 11. Dufour V, Feydy A, Rillardon L, Redondo A, Le Page L, Bert F, Belmatoug N, Fantin B (2005) Comparative study of postoperative and spontaneous pyogenic spondylodiscitis. Semin Arthritis Rheum 34:766–771
Acta Neurochir (2014) 156:1189–1196 12. Euba G, Narváez JA, Nolla JM, Murillo O, Narváez J, GómezVaquero C, Ariza J (2008) Long-term clinical and radiological magnetic resonance imaging outcome of abscess-associated spontaneous pyogenic vertebral osteomyelitis under conservative management. Semin Arthritis Rheum 38:28–40 13. Eysel P, Hopf C, Vogel I, Rompe JD (1997) Primary stable anterior instrumentation or dorsoventral spondylodesis in spondylodiscitis? Results of a comparative study. Eur Spine J 6:152–157 14. Fang D, Cheung KM, Dos Remedios ID, Lee YK, Leong JC (1994) Pyogenic vertebral osteomyelitis: treatment by anterior spinal debridement and fusion. J Spinal Disord 7:173–180 15. Fukuta S, Miyamoto K, Masuda T, Hosoe H, Kodama H, Nishimoto H, Sakaeda H, Shimizu K (2003) Two-stage (posterior and anterior) surgical treatment using posterior spinal instrumentation for pyogenic and tuberculotic spondylitis. Spine 28:E 302–308 16. Hadjipavlou AG, Mader JT, Necessary JT, Muffoletto AJ (2000) Hematogenous pyogenic spinal infections and their surgical management. Spine 25:1668–1679 17. Hanaoka N, Kawasaki Y, Sakai T, Nakamura T, Nanamori K, Nakamura E, Uchida K, Yamada H (2006) Percutaneous drainage and continuous irrigation in patients with severe pyogenic spondylitis, abscess formation, and marked bone destruction. J Neurosurg Spine 4:374–379 18. Hee HT, Majd ME, Holt RT, Pienkowski D (2002) Better treatment of vertebral osteomyelitis using posterior stabilization and titanium mesh cages. J Spinal Disord Tech 15:149–156 19. Heyde CE, Boehm H, El Saghir H, Tschöke SK, Kayser R (2006) Surgical treatment of spondylodiscitis in the cervical spine: a minimum 2-year follow-up. Eur Spine J 15:1380–1387 20. Hopf C, Meurer A, Eysel P, Rompe JD (1998) Operative treatment of spondylodiscitis—what is the most effective approach? Neurosurg Rev 21:217–225 21. Hopkinson N, Stevenson J, Benjamin S (2001) A case ascertainment study of septic discitis: clinical, microbiological and radiological features. QJM 94:465–470 22. Ito M, Abumi K, Kotani Y, Kadoya K, Minami A (2007) Clinical outcome of posterolateral endoscopic surgery for pyogenic spondylodiscitis: results of 15 patients with serious comorbid conditions. Spine 32:200–206 23. Kuklo TR, Potter BK, Bell RS, Moquin RR, Rosner MK (2006) Single-stage treatment of pyogenic spinal infection with titanium mesh cages. J Spinal Disord Tech 19:376–382 24. Karadimas EJ, Bunger C, Lindblad BE, Hansen ES, Høy K, Helmig P, Kannerup AS, Niedermann B (2008) Spondylodiscitis. A retrospective study of 163 patients. Acta Orthop 79:650–659 25. Khan IA, Vaccaro AR, Zlotolow DA (1999) Management of vertebral discitis and osteomyelitis. Orthopedics 22:758–765 26. Klockner C, Valencia R (2003) Sagittal alignment after anterior debridement and fusion with or without additional posterior instrumentation in the treatment of pyogenic and tuberculous spondylodiscitis. Spine 28:1036–1042 27. Klockner C, Valencia R, Weber U (2001) Sagittal alignment following surgical therapy of destructive spondylodiscitis: ventral or ventrodorsal procedure—a comparison of results. Orthopade 30: 965–976 28. Krödel A, Krüger A, Lohscheidt K, Pfahler M, Refior HJ (1999) Anterior debridement, fusion, and extrafocal stabilization in the treatment of osteomyelitis of the spine. J Spinal Disord 12:17–26 29. Krodel A, Sturz H (1989) Differentiated surgical and conservative treatment of spondylitis and spondylodiscitis. Z Orthop Ihre Grenzgeb 127:589–596 30. Krodel A, Sturz H, Siebert CH (1991) Indications for and results of operative treatment of spondylitis and spondylodiscitis. Arch Orthop Trauma Surg 110:78–82 31. Korovessis P, Petsinis G, Koureas G, Iliopoulos P, Zacharatos S (2006) Anterior surgery with insertion of titanium mesh cage and
1195
32.
33.
34.
35.
36.
37. 38.
39.
40. 41. 42.
43.
44.
45.
46.
47.
48.
49.
50.
posterior instrumented fusion performed sequentially on the same day under one anesthesia for septic spondylitis of thoracolumbar spine: is the use of titanium mesh cages safe? Spine 31:1014–1019 Korovessis P, Repantis T, Iliopoulos P, Hadjipavlou A (2008) Beneficial influence of titanium mesh cage on infection healing and spinal reconstruction in hematogenous septic spondylitis: a retrospective analysis of surgical outcome of twenty-five consecutive cases and review of literature. Spine (Phila Pa 1976) 33:E759–67 Lee JS, Suh KT (2006) Posterior lumbar interbody fusion with an autogenous iliac crest bone graft in the treatment of pyogenic spondylodiscitis. J Bone Joint Surg Br 88:765–770 Linhardt O, Kruger A, Krodel A (2004) First results of anterior versus posterior instrumentation-fusion in the treatment of spondylodiscitis. Z Orthop Ihre Grenzgeb 142:73–78 Mann S, Schütze M, Sola S, Piek J (2004) Nonspecific pyogenic spondylodiscitis: clinical manifestations, surgical treatment, and outcome in 24 patients. Neurosurg Focus 17(6):E3 McGuire RA, Eismont FJ (1994) The fate of autogenous bone graft in surgically treated pyogenic vertebral osteomyelitis. J Spinal Disord 7: 206–215 Mondorf Y, Gaab MR, Oertel JM (2009) PEEK cage cervical ventral fusion in spondylodiscitis. Acta Neurochir 151:1537–1541 Mylona E, Samarkos M, Kakalou E, Fanourgiakis P, Skoutelis A (2009) Pyogenic vertebral osteomyelitis: a systematic review of clinical characteristics. Semin Arthritis Rheum 39:10–17 Nagata K, Ohashi T, Ariyoshi M, Sonoda K, Imoto H, Inoue A (1998) Percutaneous suction aspiration and drainage for pyogenic spondylitis. Spine 23:1600–1606 O’Daly BJ, Morris SF, O’Rouke SK (2008) Long-term function outcome in pyogenic spinal infection. Spine 33:E246–E253 Ogden AT, Kaiser MG (2004) Single stage debridement and instrumentation for pyogenic spinal infections. Neurosurg Focus 17:E5,1–5 Pavlov PW, Meijers H, van Limbeek J, Jacobs WC, Lemmens JA, Obradov-Rajic M, de Kleuver M (2004) Good outcome and restoration of lordosis after anterior lumbar interbody fusion with additional posterior fixation. Spine 29:1893–1900 Pee YH, Park JD, Choi YG, Lee SH (2008) Anterior debridement and fusion followed by posterior pedicle screw fixation in pyogenic spondylodiscitis: autologous iliac bone strut versus cage. J Neurosurg Spine 8:405–412 Pigrau C, Almirante B, Flores X, Falco V, Rodríguez D, Gasser I, Villanueva C, Pahissa A (2005) Spontaneous pyogenic vertebral osteomyelitis and endocarditis: incidence, risk factors, and outcome. Am J Med 118:1287 Przybylski GJ, Sharan AD (2001) Single-stage autogenous bone grafting and internal fixation in the surgical management of pyogenic discitis and vertebral osteomyelitis. J Neurosurg Spine 94:1–7 Quiñones-Hinojosa A, Jun P, Jacobs R, Rosenberg WS, Weinstein PR (2004) General principles in the medical and surgical management of spinal infections: a multidisciplinary approach. Neurosurg Focus 17:E1,1–15 Rath SA, Neff U, Schneider O, Richter HP (1996) Neurosurgical management of thoracic and lumbar vertebral osteomyelitis and discitis in adults: a review of 43 consecutive surgically treated patients. Neurosurg 38:926–933 Roblot F, Besnier JM, Juhel L, Vidal C, Ragot S, Bastides F (2007) Optimal duration of antibiotic therapy in vertebral oseomylitis. Semin Arthritis Rheum 36:269–277 Sakkas LI, Davas EM, Kapsalaki E, Boulbou M, Makaritsis K, Alexiou I, Tsikrikas T, Stathakis N (2009) Hematogenous spinal infection in central Greece. Spine (Phila Pa 1976) 34:E513–518 Schinkel C, Gottwald M, Andress HJ (2003) Surgical treatment of spondylodiscitis. Surg Infect 4:387–391
1196 51. Sindern E, Gläser E, Bötel U, Malin JP (1993) Spondylodiscitis with spinal and radicular deficits. Limits of conservative treatment. Nervenarzt 64:801–805 52. Valancius K, Hansen ES, Høy K, Helmig P, Niedermann B, Bünger C (2012) Failure modes in conservative and surgical management of infectious spondylodiscitis. Eur Spine J 22: 1837–1844
Acta Neurochir (2014) 156:1189–1196 53. Walter J, Kuhn SA, Reichart R, Kalff R, Ewald C (2010) PEEK cages as a potential alternative in the treatment of cervical spondylodiscitis: a preliminary report on a patient series. Eur Spine J 19:1004–1009 54. Zarrouk V, Feydy A, Sallès F, Dufour V, Guigui P, Redondo A (2007) Imaging does not predict the clinical outcome of bacterial vertebral osteomyelitis. Rheumatology (Oxford) 46:292–295
CLINICAL ARTICLE - SPINE
A retrospective study of 113 consecutive cases of surgically treated spondylodiscitis patients. A single-center experience Ehab Shiban & Insa Janssen & Maria Wostrack & Sandro M Krieg & Florian Ringel & Bernhard Meyer & Michael Stoffel
Received: 25 October 2013 / Accepted: 7 March 2014 / Published online: 27 March 2014 # Springer-Verlag Wien 2014
Abstract Background Recommendations for the operative treatment of spondylodiscitis are still a controversial issue. Methods A retrospective review identified 113 consecutive patients who underwent surgical debridement and instrumentation for spondylodiscitis between 2006 and 2010 at our department. Results The mean age at presentation was 65 years; 78 patients were male (69 %). Distribution of the inflammation was lumbar in 68 (60 %), thoracic in 19 (17 %) and cervical in 20 (18 %) cases. Six patients (5 %) had two concomitant noncontiguous spondylodiscitis foci in different segments of the spine. Epidural abscess was found in 33 patients (29 %). One hundred four patients (92 %) had pain. Neurological deficit Presentation at a conference: Global Spine Congress, Hong Kong, China April 2013 E. Shiban (*) : I. Janssen : M. Wostrack : S. M. Krieg : F. Ringel : B. Meyer : M. Stoffel Department of Neurosurgery, Klinikum rechts der Isar, Technische Universität München, Ismaninger Str. 22, 81675 Munich, Germany e-mail: [email protected] I. Janssen e-mail: [email protected] M. Wostrack e-mail: [email protected] S. M. Krieg e-mail: [email protected]
was found in 40 patients (35 %). In the thoracic and lumbar cases, dorsal instrumentation alone was considered sufficient in 26 cases; additional interbody fusion from the posterior was performed in 44 cases. A 360° instrumentation was performed in 22 cases. In the cervical cases, only ventral spondylodesis and plating were performed in eight cases, only dorsal instrumentation in five and 360° instrumentation in seven. Postoperative intravenous antibiotics were administered for 14.4±9.3 (mean ± SD) days followed by 3.2±0.8 (mean ± SD) months of oral antibiosis. Complete healing of the inflammation was achieved in 111 (98 %) cases. Two patients died because of septic shock, both with fulminant endocarditis. Pain resolved in all cases. Neurological deficits were completely resolved in 20 patients, and 14 patients had a partial recovery. Conclusion The results of our retrospective study show that surgical treatment of spondylodiscitis with a staged surgical approach (if needed) and a short 1-2-week period of intravenous antibiotics followed by 3 months of oral antibiotics is appropriate for most patients in whom conservative treatment has failed or is not advisable. Furthermore, surgical treatment of newly diagnosed spondylodiscitis might be recommended as an initial treatment option in many cases. Thereby the choice of fusion material (autologous bone, titanium, PEEK) seems less important. Keywords Spondylodiscitis . Spinal instrumentation . Antibiotic regime
F. Ringel e-mail: [email protected] B. Meyer e-mail: [email protected]
Introduction
M. Stoffel e-mail: [email protected]
Spinal infections can be a source of significant morbidity and mortality from spinal instability, neurological deficits and ongoing sepsis. Non-specific pyogenic spondylodiscitis is a rare clinical condition that can present with uncharacteristic chronic
M. Stoffel Department of Neurosurgery, Helios Klinikum, Krefeld, Germany
1190
back pain to a life-threatening course with rapidly progressing neurological deficits. The incidence of spondylodiscitis is approximately 1:250,000 per year. Pyogenic spondylodiscitis represents only 2–4 % of all cases of bone infection [16]. The incidence has been on the rise during the last few decades. This has been attributed to the increase in the elderly population with comorbidities, mainly diabetes mellitus, intravenous drug abuse and impaired immunocompetence (immunosuppressive drugs, human immunodeficiency virus, infections, chronic alcoholism) [2, 16, 45]. While the traditional method of managing spondylodiscitis has been prolonged immobilization, this regime is associated with significant complications and treatment failure. With increasing expertise in spinal instrumentation and fusion techniques, the approach for managing spondylodiscitis needs to be revisited. Hence, this study aims to assess the effectiveness of surgical intervention as a primary modality in spondylodiscitis.
Material and methods We performed a retrospective review of patients who underwent surgical treatment for spondylodiscitis between 2006 and 2010 at our department. One hundred twenty-three consecutive patients were identified. Ten patients were treated conservatively because of (1) severe comorbidities that caused the patients to have a high surgical mortality rate and (2) minor pain symptoms and regressive inflammation signs after antibiotic treatment. One hundred thirteen patients underwent surgical debridement and instrumentation. The following refers to these patients. Data were gathered through review of patients’ case notes, relevant imaging and electronic records. Gd-enhanced MR imaging of the whole spine was obtained in all cases. CT scans of the affected region were performed in all cases in order to better evaluate the extent of bone destruction. Our department of neuroradiology and the senior author examined all images. Patients were considered to have fever if their temperature was ≥ 38.5 °C. C-reactive protein (CRP) and complete blood count tests were performed in all cases using routine laboratory techniques. Neurological deficit was classified as mild when limited to problems involving one or two nerve roots. Severe deficits were classified according to the ASIA impairment scale. All patients were advised about both treatment modalities as conservative treatment with immobilization followed by mobilization in a brace or surgical treatment. However, for patients with a neurological deficit, abscess or failed conservative treatment, surgery was strongly recommended (Fig. 1). The surgical approach and method of spinal instrumentation were determined for each patient individually according to the comorbidities, location of infection and extent of bony destruction. For the cervical spine, the surgical approach was influenced foremost by the localization of the inflammation in
Acta Neurochir (2014) 156:1189–1196
Fig. 1 Number of patients with a neurological deficit, epidural abscess and/or failed conservative treatment for whom surgical treatment was strongly recommended
relation to the cervical spinal cord. Accordingly, a ventral approach was chosen if the main focus of the infection was ventral to the spinal cord; otherwise, a dorsal approach was chosen. Additional posterior surgery was performed in order to either enhance stability or provide additional decompression from the posterior. Options used were an anterior cervical discectomy or corpectomy and fusion using PEEK, titanium or autologous bone from the iliac crest depending on the surgeon’s preference with or without dorsal instrumentation or dorsal instrumentation alone. For patients with inflammation in the thoracolumbar spine, options used were dorsal instrumentation with or without interbody fusion. In cases without any destruction of the endplates and disc height reduction, then dorsal instrumentation without interbody fusion was performed. Interbody fusion was performed if the height of the affected disc was normal or increased. If an interbody fusion from the posterior was technically not feasible, then interbody fusion was performed through a ventral or lateral approach. Antibiotic treatment included a short period of broadspectrum intravenous antibiotic therapy, a tailoring according to the antibiogram and switching over to oral antibiotics at the time of clear improvement of clinical and laboratory parameters for a period of 3 months. Cure of inflammation was considered complete if patients exhibited no signs or symptoms of active localized infection by clinical examination and inflammatory markers at 3 months postoperatively. Neurological outcomes mentioned are at a minimum of 3month postoperative follow-up. To our knowledge, our study is the single largest series of spondylodiscitis treatment in which all patients underwent surgical debridement of infected disc space followed by spinal instrumentation.
Results The median age at presentation was 67 years (range 40–94). Seventy-eight patients were male (69 %); 104 patients (92 %)
Acta Neurochir (2014) 156:1189–1196
had pain. Neurological deficit was found in 40 patients (35 %); 36 patients had a mild deficit, and 4 were suffered from paraparesis. CRP was elevated in all patients. Median CRP value upon admission was 7.7 mg/dl, and it declined significantly in all cases upon discharge (Fig. 2). White blood cell (WBC) count was elevated only in 52 patients (46 %). Median WBC value on admission was 9.3 g/l, and it was significantly reduced at discharge (Fig. 3). Fever was diagnosed on admission in 24 patients (21 %) and resolved in all patients prior to discharge. Spondylodiscitis was confined to one level in 84 (74 %), two contiguous levels in 24 (21 %) and three contiguous levels in 5 (4 %) cases. Epidural abscess was found in 33 patients (29 %). Distribution of the inflammation was lumbar in 68 (60 %), thoracic in 19 (17 %) and cervical in 20 (18 %) cases. Six patients (5 %) had two occurrences of concomitant noncontiguous spondylodiscitis in different segments of the spine. The suspected source of infection was identified in 103 (91 %) of the cases. The causes were as follows: prior surgery in 56 (49 %), repetitive facet joint infiltration in 14 (12 %), diverticulitis in 9 (8 %), endocarditis in 9 (8 %), renal insufficiency and infection in 6 (5 %), wound healing disorders secondary to diabetes mellitus in 6 (5 %), parodontitis in 2 and parotitis in 1 case. No source of infection could be suspected in ten (9 %) cases (Fig. 4). Successful isolation of the organism from the infected space was possible in 82 patients (73 %). In 31 patients (27 %), organisms could not be isolated. Eighteen (58 %) of these 31 patients had antibiotic therapy prior to surgery. Staphylococci were encountered in 58 (51 %) patients. In 39 cases Staphylococcus aureus was identified. Multiresistant bacteria were found in 15 cases (13 %), and in 9 (8 %) cases of sporadic bacteria were identified. As mentioned above, the surgical approach was tailored for each patient depending on comorbidities, localization of
Fig. 2 Diagram showing C-reactive protein (CRP) levels throughout the hospital stay. CRP was elevated in all cases. Median CRP value upon admission was 7.7 md/dl, and it declined significantly in all cases upon discharge
1191
Fig. 3 Diagram showing white cell count (WCC) levels throughout the hospital stay. WCC was elevated only in 52 patients (46 %). Median WCC value at admission was 9.3 g/l, and it was significantly reduced at discharge
infection and extent of bone destruction. In the thoracic and lumbar cases, dorsal instrumentation alone was considered sufficient in 26 cases; additional interbody fusion from the posterior was performed in 44 cases (38 autologous bone and 6 PEEK cages); 360° instrumentation was performed in 22 cases (5 PEEK, 14 titanium and 3 autologous bone) (Fig. 5). In cervical disease, ventral spondylodesis was performed in eight cases (3 PEEK cages and 5 autologous bone from the iliac crest), ventral plating, dorsal instrumentation alone in five cases and 360° instrumentation in seven cases (4 PEEK cages and 3 autologous bone from the iliac crest) (Fig. 5). Surgery-related complications were seen in seven patients (6 %). Two patients had a postoperative epidural bleed, screw correction was indicated in three cases, and two patients had developed a wound-healing disorder. Further complications were observed in 12 patients (11 %), 2 patients developed pulmonary emboli, 4 patients developed pneumonia and 6 had a urinary tract infection requiring escalation of the antibiotic treatment. A broad-spectrum antibiotic, mostly clindamycin, was administered initially till the results of the antibiogram were available. An organism-sensitive specific antibiotic was initiated based on the antibiogram report. Postoperative intravenous antibiotics were administered for 14.4±9.3 (mean ± SD) days followed by 3.2±0.8 (mean ± SD) months of oral antibiotics. Patients in whom the organism was not isolated received empirical broad-spectrum antibiotics. All patients continued with oral antibiotics for 3 months without major complications. Forty-seven patients (41 %) reported bowel discomfort but did not need to discontinue the medication. Five patients (4 %) developed diarrhea, and the antibiotic was paused for 1 week. Two patients developed an a facial exanthema, and the antibiotic was paused for 2 days. None of the patients developed any other serious side effects. All patients
1192
Acta Neurochir (2014) 156:1189–1196
Fig. 4 Summary of the suspected source of infections
had a strict follow-up regime including weekly physical examination and blood analysis with their primary care physician. Final follow-up at 3 months was performed at our institution in all surviving patients. Pain resolved in all cases. From the 36 patients with a mild neurological deficit, 20 had a full recovery, 12 had a partial recovery, and 4 were unchanged postoperatively. From the four patients with a severe neurological deficit, two had a partial recovery and two were unchanged postoperatively. Septicemia and multiorgan failure occurred in two patients, leading to death. Both patients had a fulminant endocarditis. Spinal surgery was a prerequisite for further urgent cardiac surgery. Eradication of infection was ultimately achieved in all patients as documented by normalization of CRP and clinical findings. One patient had a relapse of the inflammation following dorsal instrumentation alone. This was followed by debridement and anterior interbody fusion upon relapse.
Fig. 5 Summary of the surgical procedures
Discussion Obviously, this study suffers from some limitations due to its retrospective nature. Due to the nature of our department as a spine center, most of our patients (62 %) were referred to us with a clear indication for surgery (Fig. 1). For the remaining 38 % of patients, both surgical and conservative treatments were an option, and conservative treatment might have achieved good results as well. However, we tend to recommend surgery because it allows rapid mobilization and shortens hospital stay. This is on the one hand a major limitation because most of the patients were treated surgically. On the other hand, this uniquely large consecutive series of a surgically treated cohort with spondylodiscitis combined with detailed written and electronic notes allows some substantial conclusions.
Acta Neurochir (2014) 156:1189–1196
Similar to previous studies, the patient population was mostly elderly and male. In this study, 69 % of the patients were male and 57 % over the age of 60. The higher incidence in the male population remains to be explained. The association with age is believed to be due to the higher incidence of comorbidities in this higher age group [11, 21]. Also in concordance with the literature, the distribution of the inflammation was mostly seen in the lumbar spine followed by the thoracic and cervical spine [11, 21]. Among the inflammatory markers we found CRP to be most reliable in affirming the diagnosis of infection as well as for monitoring the response to treatment. CRP was elevated in all patients. Median CRP value upon admission was 7.7 mg/dl and significantly declined in all cases upon discharge. CRP had returned to normal in all survivors at 3-month follow-up. These findings are similar to previously published literature [11, 12, 54]. WBC count was not very reliable as it was elevated in only 52 patients (46 %). Immunocompromised patients and patients over the age of 60 years were more likely to have normal WBC levels despite infection [6]. Clinical findings of spondylodiscitis can be non-specific. However, back or neck pain is very common, but could be absent in up to 15 % of cases [49]. In our series 92 % of all patients presented with pain. Fever is not frequently experienced and occurs in only 20–50 % of cases [38]. Fever was diagnosed on admission in 24 patients (21 %) and resolved in all cases prior to discharge. Similar to other studies, a neurological deficit was found in 35 % of cases and was more likely to be present with epidural abscess or delayed diagnosis [44]. The surgical approach was tailored for each case individually and a staged surgical approach was implemented. All patients underwent surgical debridement and instrumentation, whereas staged surgery (a combination of both a ventral and a posterior approach) was needed in 25 % and 35 % in the thoracolumbar and cervical cases, respectively. Postoperative intravenous antibiotics were administered until CRP significantly declined and clinical symptoms started to resolve. Thereby, a very short duration of intravenous antibiotic administration of 14.4±9.3 (mean ± SD) was sufficient. This was followed by oral antibiotic administration of 3.2±0.8 (mean ± SD) months. Using such an algorithm of a staged surgical approach if needed and short-term postoperative intravenous antibiotics and subsequent oral antibiotics for 3 months, complete healing of the inflammation was achieved initially in 98 %, and ultimately in all cases, marked pain reduction was observed in all cases and neurological improvement was observed in 88 % of cases. The conservative method of managing spondylodiscitis includes external immobilization by orthesis or bed rest and prolonged duration of parenteral antibiotic administration and analgesics. This method aims at achieving spontaneous healing and fusion of the infected spinal segment [27, 29, 30, 41, 46, 51]. However, reported failure rates following conservative treatment range from 12 % to 18 % [3, 8, 24,
1193
40, 52]. Both mortality and morbidity rates are similar following conservative or surgical treatment in most patients with spondylodiscitis [52]; this is however not true for patients with spondylodiscitis caused by drug multiresistant bacteria. In these patients morbidity and mortality rates of up to 38 % have been reported, mostly following conservative treatment [1]. Fifteen patients (13 %) in our study were infected by multiresistant bacteria. Of those, 12 patients (92 %) had full recoveries, and 1 patient died 3 weeks postoperatively because of fulminant endocarditis. Spinal surgery was a prerequisite for further urgent cardiac surgery. Neither controlled comparative trials nor standard guidelines are present for the optimal antibiotic and duration of treatment. Duration of intravenous antibiotics administered varied from 10 days [35], 3 weeks [9], 30 days [17] to 6 weeks [43, 45] in different series. In this study 14.4±9.3 (mean ± SD) days of intravenous antibiotic administration was relatively short, but achieved significant improvement in clinical and laboratory findings in all surviving cases (98 %). The duration of oral antibiotics administered ranges from 6 weeks [9, 43], 59 days [17] to 3 months [35]. The administration of parenteral and oral antibiotics tends to be longer with a mean of 21 days and 45 days respectively if the infected disc tissue is not eradicated [9]. An observational retrospective study comparing duration of treatment of less than 6 weeks in one group or more than 6 weeks in the other group showed similar results of recovery rates, relapse and death [48]. Although high-level evidence is lacking, there is a general consensus on discontinuing long-term oral antibiotics only after normalization of ESR, CRP and leucocyte count [17, 19, 35, 43]. Surgical strategies in treating spondylodiscitis vary considerably. Some favor radical debridement [4, 13, 16, 20, 25, 26, 32, 34, 35, 41, 43, 45, 47, 50]. Others, however, have reported excellent results using minimally invasive dorsal percutaneous pedicle-screw rod spondylodesis, with the aim to achieve immediate pain relief and patient mobilization, without eradication of infected intervertebral disc space [9, 17, 22, 35, 39]. The specific surgical approach for debridement also varies. Some authors favor anterior approaches for debridement as it allows direct access to the infected intervertebral disc space, paraspinal abscess and stabilization by interbody graft [10, 14, 36, 37, 45]. Anterior lumbar interbody fusion followed by anterior stabilization also allows for good restoration of lordosis and anterior column support [13, 14, 42]. Supplementation by posterior instrumentation provides better deformity correction, correction of sagittal alignment and faster rates of fusion [10, 15, 18, 24, 28, 41]. Two-stage approaches of initial anterior surgical debridement and fusion followed by delayed posterior instrumentation after a period of intravenous antibiotics has been practiced in order to prevent increased risk of infection after spinal instrumentation [10]. Fukuta et al. reported excellent outcomes following placement of posterior instrumentation in the first stage
1194
followed by anterior debridement and fusion in the second stage [15]. Satisfactory results have also been reported following single-stage anterior or posterior interbody fusion and pedicle screw fixation [31, 33, 37]. The most effective surgical approach—anterior, posterior or a combination of both, single stage or two stages—needs to be tailored on an individual case basis [16, 20, 26, 27, 34, 35, 47]. However, most reports suggest that single-stage debridement, fusion and internal fixation do not lead to higher rates of wound infection or hardware-related complications [45]. In our series we found staged surgery, if needed, to be sufficient in most cases. Primarily in the thoracic and lumbar cases, dorsal instrumentation alone was considered sufficient in 26 cases. In 66 cases substantial destruction of the endplates was present, and therefore interbody fusion was deemed necessary. When possible, this was performed in the same procedure from the posterior with either autologous bone material (n= 38) or PEEK cages (n=6). If interbody fusion from the posterior was not mechanically feasible, then an anterior approach (n=22) was performed in a second surgical procedure, mostly 1 week thereafter, allowing the patient to recuperate from the first operation. As shown, this rationale of staged surgery (if needed) proved to be sufficient in most cases. One patient had a relapse 1 month postoperatively following dorsal instrumentation alone. This was followed by debridement and anterior interbody fusion upon relapse. Use of instrumentation in treating spinal infections has been a subject of intense debate. Several authors have expressed their concern about the use of alloplastic implants at the site of active infection [5, 13, 14, 36, 37]. Bacteria tend to colonize stainless steel and polymethyl-methacrylate heavily, thereby reducing antibiotic effectiveness. However, titanium is less prone to bacterial adherence [7]. Interbody fusion using autologous bone graft has long been considered good practice to treat spondylodiscitis [33, 47]. In a retrospective review of 60 patients diagnosed with pyogenic spondylodiscitis, Pee et al. [43] performed a single-stage anterior debridement and fusion using either an autologous iliac bone strut or cage (titanium, titanium mesh, PEEK) followed by posterior pedicle screw fixation. In this study no difference in clinical (pain, functional disability) and radiological outcome (correction of segmental lordosis and fusion rate) was noted between the strut and cage group. However, the subsidence rate was higher in the strut group. Duration until subsidence was also shorter in the strut group than in the cage group. Several studies have demonstrated favorable results using the titanium mesh cage compared to autologous bone grafts for anterior interbody fusion in infected spines [18, 23, 31]. Similar to other reports, the use of PEEK cages in our series (N=19) showed successful outcomes in terms of clinical and laboratory parameters and radiological findings without an increase in relapse rates [4, 37, 43, 53].
Acta Neurochir (2014) 156:1189–1196
Conclusion The results of our retrospective study show that surgical treatment of spondylodiscitis with a staged surgical approach (if needed) and a short period of intravenous antibiotics of 12 weeks followed by 3 months of oral antibiotics is appropriate in most patients in whom conservative treatment has failed or is not advisable because of the presence of a neurological deficit or an epidural abscess. Furthermore, with relatively low rates of surgical complication and eradication of the inflammation in all cases, surgical treatment of newly diagnosed spondylodiscitis might be recommended as an initial treatment option in many cases. Therefore, the choice of fusion material (autologous bone, titanium, PEEK) seems less important.
Conflicts of interest The authors declare that they have no conflict of interest affecting this study. The study was completely financed by the Department of Neurosurgery.
References 1. Al-Nammari SS, Lucas JD, Lam KS (2007) Hematogenous methicillin-resistant Staphylococcus aureus spondylodiscitis. Spine (Phila Pa 1976) 15:2480–2386 2. Belzunegui J, Intxausti JJ, De Dios JR, Del Val N, Rodríguez Valverde V, González C, Queiro R, Figueroa M (2000) Haematogenous vertebral osteomyelitis in the elderly. Clin Rheumatol 19:344–347 3. Bettini N, Girardo M, Dema E, Cervellati S (2009) Evaluation of conservative treatment of non specific spondylodiscitis. Eur Spine J 1:143–150 4. Brase A, Ringel F, Stüer C, Meyer B, Stoffel M (2010) Debridement and fusion with polyetheretherketone implants in purulent spondylodiscitis: a clinical experience with nine patients. Acta Neurochir 152:2001–2004 5. Carragee EJ (1997) Instrumentation of the infected and unstable spine: a review of 17 cases from the thoracic and lumbar spine with pyogenic infections. J Spinal Disord 10:317–324 6. Carragee EJ (1997) Pyogenic vertebral osteomyelitis. J Bone Joint Surg Am 79:874–880 7. Chang CC, Merritt K (1994) Infection at the site of implanted materials with and without preadhered bacteria. J Orthop Res 12: 526–531 8. Cottle L, Riodan T (2008) Infectious spondylodiscitis. J Infect 56: 401–412 9. Deininger MH, Unfried MI, Vougioukas VI, Hubbe U (2009) Minimally invasive dorsal percutaneous spondylodesis for the treatment of adult pyogenic spondylodiscitis. Acta Neurochir 151:1451– 1457 10. Dimar JR, Carreon LY, Glassman SD, Campbell MJ, Hartman MJ, Johnson JR (2004) Treatment of pyogenic vertebral osteomyelitis with anterior debridement and fusion followed by delayed posterior spinal fusion. Spine 29:326–332 11. Dufour V, Feydy A, Rillardon L, Redondo A, Le Page L, Bert F, Belmatoug N, Fantin B (2005) Comparative study of postoperative and spontaneous pyogenic spondylodiscitis. Semin Arthritis Rheum 34:766–771
Acta Neurochir (2014) 156:1189–1196 12. Euba G, Narváez JA, Nolla JM, Murillo O, Narváez J, GómezVaquero C, Ariza J (2008) Long-term clinical and radiological magnetic resonance imaging outcome of abscess-associated spontaneous pyogenic vertebral osteomyelitis under conservative management. Semin Arthritis Rheum 38:28–40 13. Eysel P, Hopf C, Vogel I, Rompe JD (1997) Primary stable anterior instrumentation or dorsoventral spondylodesis in spondylodiscitis? Results of a comparative study. Eur Spine J 6:152–157 14. Fang D, Cheung KM, Dos Remedios ID, Lee YK, Leong JC (1994) Pyogenic vertebral osteomyelitis: treatment by anterior spinal debridement and fusion. J Spinal Disord 7:173–180 15. Fukuta S, Miyamoto K, Masuda T, Hosoe H, Kodama H, Nishimoto H, Sakaeda H, Shimizu K (2003) Two-stage (posterior and anterior) surgical treatment using posterior spinal instrumentation for pyogenic and tuberculotic spondylitis. Spine 28:E 302–308 16. Hadjipavlou AG, Mader JT, Necessary JT, Muffoletto AJ (2000) Hematogenous pyogenic spinal infections and their surgical management. Spine 25:1668–1679 17. Hanaoka N, Kawasaki Y, Sakai T, Nakamura T, Nanamori K, Nakamura E, Uchida K, Yamada H (2006) Percutaneous drainage and continuous irrigation in patients with severe pyogenic spondylitis, abscess formation, and marked bone destruction. J Neurosurg Spine 4:374–379 18. Hee HT, Majd ME, Holt RT, Pienkowski D (2002) Better treatment of vertebral osteomyelitis using posterior stabilization and titanium mesh cages. J Spinal Disord Tech 15:149–156 19. Heyde CE, Boehm H, El Saghir H, Tschöke SK, Kayser R (2006) Surgical treatment of spondylodiscitis in the cervical spine: a minimum 2-year follow-up. Eur Spine J 15:1380–1387 20. Hopf C, Meurer A, Eysel P, Rompe JD (1998) Operative treatment of spondylodiscitis—what is the most effective approach? Neurosurg Rev 21:217–225 21. Hopkinson N, Stevenson J, Benjamin S (2001) A case ascertainment study of septic discitis: clinical, microbiological and radiological features. QJM 94:465–470 22. Ito M, Abumi K, Kotani Y, Kadoya K, Minami A (2007) Clinical outcome of posterolateral endoscopic surgery for pyogenic spondylodiscitis: results of 15 patients with serious comorbid conditions. Spine 32:200–206 23. Kuklo TR, Potter BK, Bell RS, Moquin RR, Rosner MK (2006) Single-stage treatment of pyogenic spinal infection with titanium mesh cages. J Spinal Disord Tech 19:376–382 24. Karadimas EJ, Bunger C, Lindblad BE, Hansen ES, Høy K, Helmig P, Kannerup AS, Niedermann B (2008) Spondylodiscitis. A retrospective study of 163 patients. Acta Orthop 79:650–659 25. Khan IA, Vaccaro AR, Zlotolow DA (1999) Management of vertebral discitis and osteomyelitis. Orthopedics 22:758–765 26. Klockner C, Valencia R (2003) Sagittal alignment after anterior debridement and fusion with or without additional posterior instrumentation in the treatment of pyogenic and tuberculous spondylodiscitis. Spine 28:1036–1042 27. Klockner C, Valencia R, Weber U (2001) Sagittal alignment following surgical therapy of destructive spondylodiscitis: ventral or ventrodorsal procedure—a comparison of results. Orthopade 30: 965–976 28. Krödel A, Krüger A, Lohscheidt K, Pfahler M, Refior HJ (1999) Anterior debridement, fusion, and extrafocal stabilization in the treatment of osteomyelitis of the spine. J Spinal Disord 12:17–26 29. Krodel A, Sturz H (1989) Differentiated surgical and conservative treatment of spondylitis and spondylodiscitis. Z Orthop Ihre Grenzgeb 127:589–596 30. Krodel A, Sturz H, Siebert CH (1991) Indications for and results of operative treatment of spondylitis and spondylodiscitis. Arch Orthop Trauma Surg 110:78–82 31. Korovessis P, Petsinis G, Koureas G, Iliopoulos P, Zacharatos S (2006) Anterior surgery with insertion of titanium mesh cage and
1195
32.
33.
34.
35.
36.
37. 38.
39.
40. 41. 42.
43.
44.
45.
46.
47.
48.
49.
50.
posterior instrumented fusion performed sequentially on the same day under one anesthesia for septic spondylitis of thoracolumbar spine: is the use of titanium mesh cages safe? Spine 31:1014–1019 Korovessis P, Repantis T, Iliopoulos P, Hadjipavlou A (2008) Beneficial influence of titanium mesh cage on infection healing and spinal reconstruction in hematogenous septic spondylitis: a retrospective analysis of surgical outcome of twenty-five consecutive cases and review of literature. Spine (Phila Pa 1976) 33:E759–67 Lee JS, Suh KT (2006) Posterior lumbar interbody fusion with an autogenous iliac crest bone graft in the treatment of pyogenic spondylodiscitis. J Bone Joint Surg Br 88:765–770 Linhardt O, Kruger A, Krodel A (2004) First results of anterior versus posterior instrumentation-fusion in the treatment of spondylodiscitis. Z Orthop Ihre Grenzgeb 142:73–78 Mann S, Schütze M, Sola S, Piek J (2004) Nonspecific pyogenic spondylodiscitis: clinical manifestations, surgical treatment, and outcome in 24 patients. Neurosurg Focus 17(6):E3 McGuire RA, Eismont FJ (1994) The fate of autogenous bone graft in surgically treated pyogenic vertebral osteomyelitis. J Spinal Disord 7: 206–215 Mondorf Y, Gaab MR, Oertel JM (2009) PEEK cage cervical ventral fusion in spondylodiscitis. Acta Neurochir 151:1537–1541 Mylona E, Samarkos M, Kakalou E, Fanourgiakis P, Skoutelis A (2009) Pyogenic vertebral osteomyelitis: a systematic review of clinical characteristics. Semin Arthritis Rheum 39:10–17 Nagata K, Ohashi T, Ariyoshi M, Sonoda K, Imoto H, Inoue A (1998) Percutaneous suction aspiration and drainage for pyogenic spondylitis. Spine 23:1600–1606 O’Daly BJ, Morris SF, O’Rouke SK (2008) Long-term function outcome in pyogenic spinal infection. Spine 33:E246–E253 Ogden AT, Kaiser MG (2004) Single stage debridement and instrumentation for pyogenic spinal infections. Neurosurg Focus 17:E5,1–5 Pavlov PW, Meijers H, van Limbeek J, Jacobs WC, Lemmens JA, Obradov-Rajic M, de Kleuver M (2004) Good outcome and restoration of lordosis after anterior lumbar interbody fusion with additional posterior fixation. Spine 29:1893–1900 Pee YH, Park JD, Choi YG, Lee SH (2008) Anterior debridement and fusion followed by posterior pedicle screw fixation in pyogenic spondylodiscitis: autologous iliac bone strut versus cage. J Neurosurg Spine 8:405–412 Pigrau C, Almirante B, Flores X, Falco V, Rodríguez D, Gasser I, Villanueva C, Pahissa A (2005) Spontaneous pyogenic vertebral osteomyelitis and endocarditis: incidence, risk factors, and outcome. Am J Med 118:1287 Przybylski GJ, Sharan AD (2001) Single-stage autogenous bone grafting and internal fixation in the surgical management of pyogenic discitis and vertebral osteomyelitis. J Neurosurg Spine 94:1–7 Quiñones-Hinojosa A, Jun P, Jacobs R, Rosenberg WS, Weinstein PR (2004) General principles in the medical and surgical management of spinal infections: a multidisciplinary approach. Neurosurg Focus 17:E1,1–15 Rath SA, Neff U, Schneider O, Richter HP (1996) Neurosurgical management of thoracic and lumbar vertebral osteomyelitis and discitis in adults: a review of 43 consecutive surgically treated patients. Neurosurg 38:926–933 Roblot F, Besnier JM, Juhel L, Vidal C, Ragot S, Bastides F (2007) Optimal duration of antibiotic therapy in vertebral oseomylitis. Semin Arthritis Rheum 36:269–277 Sakkas LI, Davas EM, Kapsalaki E, Boulbou M, Makaritsis K, Alexiou I, Tsikrikas T, Stathakis N (2009) Hematogenous spinal infection in central Greece. Spine (Phila Pa 1976) 34:E513–518 Schinkel C, Gottwald M, Andress HJ (2003) Surgical treatment of spondylodiscitis. Surg Infect 4:387–391
1196 51. Sindern E, Gläser E, Bötel U, Malin JP (1993) Spondylodiscitis with spinal and radicular deficits. Limits of conservative treatment. Nervenarzt 64:801–805 52. Valancius K, Hansen ES, Høy K, Helmig P, Niedermann B, Bünger C (2012) Failure modes in conservative and surgical management of infectious spondylodiscitis. Eur Spine J 22: 1837–1844
Acta Neurochir (2014) 156:1189–1196 53. Walter J, Kuhn SA, Reichart R, Kalff R, Ewald C (2010) PEEK cages as a potential alternative in the treatment of cervical spondylodiscitis: a preliminary report on a patient series. Eur Spine J 19:1004–1009 54. Zarrouk V, Feydy A, Sallès F, Dufour V, Guigui P, Redondo A (2007) Imaging does not predict the clinical outcome of bacterial vertebral osteomyelitis. Rheumatology (Oxford) 46:292–295
