Introduction

Infection after spine surgery is a clinically important problem and may be associated with long-term compromise of outcome and pseudarthrosis. The incidence of postoperative infection after spinal surgery is reported to vary between 0.7% and 20%. ¹ – ³ Early infection commonly presents with overt symptoms within 2 to 3 weeks after surgery. Prompt identification and aggressive treatment usually results in resolution of the infection with retention of hardware, without significant long-term sequelae. ¹ , ³ – ⁸ Delayed infection, on the other hand, can present months to years after surgery with such constitutional symptoms as chronic pain, wound drainage, implant failure, and pseudarthrosis. ⁶ – ¹¹ The Centers for Disease Control and Prevention (CDC) proposed a duration of 1 year as the standard time period for a delayed deep infection after surgery if an implant is present. ¹² Reported rates of pseudarthrosis following surgical site infection (SSI) range from 5 to 20%. ⁹ , ¹³ Overall, the incidence of pseudarthrosis after lumbar fusion is estimated to be 5 to 35%, and in up to 50% of cases after cervical fusion. ¹⁴ , ¹⁵ Patients usually pre sent with recurrent pain or neurologic symptoms, and pseudarthrosis is responsible for up to half of the revision surgeries. Whereas 30% patients may be asymptomatic, the symptoms of pseudarthrosis are nonspecific and can be due to underlying infection, disease progression, implant failure, or adjacent segment disease. ¹⁴ , ¹⁵ Therefore, symptomatic pseudarthrosis should be investigated by a complete clinical, radiological, and laboratory workup to rule out underlying infection as a possible secondary cause. ³ , ¹⁴ – ¹⁶

A high index of suspicion for subclinical infection should be maintained in the absence of clear signs of infection, as a recent report found positive intraoperative cultures in up to 44% cases of revisions for pseudarthrosis, which were otherwise not suspected of harboring infection. ¹⁶ These findings indicate that the likelihood of infection in cases of pseudarthrosis of the spine is high, and infection should be considered in the differential diagnosis of acute and delayed pseudarthrosis and implant failure. The surgical treatment of infection in the presence of a solid fusion is generally agreed to consist of debridement and removal of instrumentation. ¹ , ³ , ⁵ – ¹¹ , ¹³ , ¹⁷ , ¹⁸ Infection in the setting of an incomplete fusion is an important and common clinical challenge, and strategies for management may include primary or delayed implant exchange. The purpose of this chapter is to describe the pathophysiology of postoperative infection and pseudarthrosis in spine surgery and to present an evidence-based approach to the management of postoperative spine infection and pseudarthrosis.

The Problem of Biofilm

Bacteria can adhere to the surface of foreign materials and form a matrix of extracellular polymeric substances, within which they embed. ³ , ¹⁰ These foreign materials include devascularized autograft, allograft, and hardware, which are commonly used in spinal fusion. ¹⁹ This biofilm offers protection against antibiotics, phagocytes, and other immune responses. These bacteria often demonstrate an altered phenotype with respect to their growth rate and gene expression. All these factors lead to increased difficulty in isolating and eradicating the microorganisms responsible for infection. Most of the common organisms that cause postoperative spinal infections, such as Staphylococcus and Propionibacterium, have the ability to produce biofilm. ³ The implant material has also been shown to affect the rate of bacterial adherence and subsequent biofilm production. Studies have found that pure titanium has a lower infection rate as compared with titanium alloys and stainless steel. ²⁰ , ²¹ Interbody cages made of polyetheretherketone (PEEK) are widely used for cervical and lumbar fusion, but show a relatively high risk of biofilm formation. ²² , ²³ Silicon nitride, a relatively recent entrant in the United States market, unlike titanium and PEEK, has a positively charged hydrophilic surface and has been shown to inhibit bacterial colonization and biofilm formation. ²² , ²³ Surgeon familiarity with materials and the propensity of bacterial adherence is important when considering the use of hardware in the presence of infection.

The age of the biofilm is also important as early infections have less tenacious and immature biofilm, leading to adequate removal by debridement and antibiotics. ³ Maturation of biofilm and persistence of infection without implant removal is an important consideration in the management of delayed infection, ¹ , ³ especially in the presence of unhealed fusion. Culturing microorganisms from mature biofilm colonies on implants is difficult and, even if grown, antibiotic susceptibility testing is often inaccurate. ³ , ¹¹ Propionibacterium is a facultative anaerobe implicated in a significant number of late infections, and requires prolonged culture for detection. ³ , ⁷ , ¹⁶ , ¹⁸ Sampredo et al ²⁴ described a technique of vortexing and sonification of spinal implants preceding culture, and was shown to be more sensitive than peri-implant cultures.

Diagnosis

Clinically the recurrence or persistence of pain after fusion should lead to a high suspicion for the presence of pseudarthrosis. Radiographically, evidence of subsidence, hardware failure, or implant loosening suggests pseudarthrosis. The time at which pseudarthrosis is considered established is not uniform, and is not described after spinal fusion. Based on the trauma literature for fractures, it is considered that failure to show any progressive radiographic evidence of fusion for at least 3 months after the time period in which normal fracture union (usually 3 to 6 months) would have occurred, is evidence of nonunion. ²⁵ If present, constitutional symptoms, pain at incision site, tenderness to palpation, or wound drainage indicates an underlying infection. ³

The initial assessment of pseudarthrosis should include plain radiographs with flexion-extension views and thin-cut computed tomography (CT). ¹⁴ , ¹⁵ In addition to these modalities, the workup for infection should include laboratory markers such as total leukocyte count (TLC), erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP). CRP has been shown to be more sensitive for the detection of infection as it is relatively stable for each individual, has a narrow normal range, and is minimally affected by medications and other pathologies (except liver failure). ² , ³ , ⁷

Given the low virulence of microorganisms causing delayed infection, clinical signs and elevated laboratory markers may be absent. Therefore, a degree of suspicion is warranted in all cases of pseudarthrosis. ¹⁶ , ¹⁸ Measurement of nutritional parameters, including serum albumin, total protein, calcium, and vitamin D is important to identify opportunity to optimize immune function and bone healing capacity.

Magnetic resonance imaging (MRI) with and without contrast is of great value for diagnosing diskitis, osteomyelitis, and epidural abscess, ³ , ⁸ although the presence of instrumentation requires special protocols for metal artifact suppression. ⁸ Additional modalities that may be considered include radionuclide bone scintigraphy using technetium-99m methylene diphosphonate (MDP), especially in cases of inconclusive MRI findings. ²⁶ Whereas technetium-99m MDP scan may be affected by underlying bone formation, gallium-67–and indium-111–labeled leukocyte scans are more specific for infection. ^{3 18}F-fluorodeoxyglucose (FDG)–positron emission tomography (PET) can be a useful modality for diagnosing delayed infection, but is expensive and requires equipment that is not universally available. ²⁷ Even in the presence of unequivocal clinical, laboratory, and imaging findings of infection, definitive diagnosis can only be reached by bacteriological examination of infected tissues. ²⁶ Biopsy is considered superior to blood cultures in pathogen detection and can be routinely performed in suspicious cases. Although user dependent, CT-guided biopsy is a useful technique to obtain microbiological diagnosis and antibiotic suseptibility. ²⁶ , ²⁸

Shifflett et al ¹⁶ evaluated 578 cases of “presumed aseptic” revision spine surgeries at a single institution between 2008 and 2013. In their analysis, they included patients who had a negative clinical, laboratory, and radiographic workup for infection and hence “presumed aseptic.” They had obtained an average of 5.5 intraoperative cultures (range, 1–14) in 19.4% (n = 112) of cases based on the surgeon’s intraoperative clinical judgement or as routine. Of all the cases that had positive intraoperative cultures (n = 45), pseudarthrosis was the primary revision diagnosis in 53.3% patients. Interestingly, 55.6% of all uninstrumented cases had a positive culture as opposed to only 32.9% of the instrumented cases. In patients revised for pseudarthrosis, Propionibacterium acnes was cultured in 54.2% cases, Staphylococcus species in 58.3%, polymicrobial infections in 25%, and gram-negative organisms in 12.5% cases. P. acnes is a common culprit in delayed infection and has been implicated in cases of pseudarthrosis. ³ , ⁷ – ⁹ , ¹⁶ , ¹⁸ Additionally, given the association between P. acnes and degenerative disk disease, the authors hypothesized that this exposure at the index surgery could very well play a role in the subsequent development of pseudarthrosis. The limitations of this retrospective study precluded the formulation of any specific recommendations, but the authors emphasized that applying clinical judgment and determining the presence of pseudarthrosis warrant a high degree of suspicion in the absence of clear indicators of infection.