6. Utility of next generation sequencing and isothermal microcalorimetry for identification of bacteria in presumed aseptic thoracolumbar revision spine surgery

BACKGROUND CONTEXT Recent studies have been able to detect the subclinical presence of bacteria in 10%-40% of patients undergoing revision spinal instrumentation procedures, suggesting that clinically silent bacteria may be responsible for presumed aseptic revisions. PURPOSE The purpose of this study is to determine the rate of bacterial presence in revision thoracolumbar spinal instrumentation surgery due to implant failure. In patients without clinical signs or symptoms of infection, we will utilize next generation sequencing (NGS) technology as well as cultures and isothermal microcalorimetry (IMC) to determine the rate of subclinical infection. STUDY DESIGN/SETTING A prospective pilot case-series at a single academic institution between September 2020 and September 2021. PATIENT SAMPLE Patients undergoing consecutive thoracolumbar revision spine surgery resulting from implant failure. OUTCOME MEASURES The main outcome measure was bacterial presence. Routine patient and clinical information including age, sex, body mass index, preoperative indication, and additional clinical/operative information were also identified. Lab values included ESR, CRP and procalcitonin. COVID tests results were documented. METHODS Patients undergoing thoracolumbar revision spine surgery due to implant failure by a single surgeon at a single institution were prospectively identified. Tissue and blood samples were taken intraoperatively and transported to a lab for analysis using IMC and NGS. RESULTS In the pilot study, 22 consecutive patients were included based on prespecified criteria. The average (SD) age of the patients was 65 (14) with 25% female sex. There were 82 total tissue samples obtained (mean 3.7 samples per patient), with a concordance of 93% between cultures and IMC, 85% between cultures and NGS, and 88% between IMC and NGS. The prevalence of occult surgical site infection was detected in 4/22 (18%) cases by NGS, 2/22 (9%) cases by IMC, and 0/22 (0%) by standard cultures. Two out of two (100%) infections detected by IMC and 2/4 (50%) of the NGS infections had grossly loose/broken pedicle screws without clear etiology for breaking, suggesting these are potentially real subclinical infection cases. There was one additional loosening of screws case without pathogen identification by any platform. However, no patients have undergone subsequent surgery for concern for infection to date. For that reason, considering loose/broken screw cases as true subclinical infections IMC had 67% sensitivity, 100% specificity, 100% PPV, and 95% NPV, while NGS was 67%, 90%, 50%, 94% respectively, and cultures were 0%, 100%, 0%, 87% respectively. Pathogens identified include Fusarium equiseti (fungi), Acinetobacter spp., Brevundimonas spp., Lecythophora spp. (fungi), Enterobacter cloacae, Escherichia coli, Enterococcus faecalis, and Staphylococcus haemolyticus. CONCLUSIONS The detection of infection in spinal implants is challenging. The use of IMC and NGS identified two occult cases of infection, while NGS also falsely identified two additional cases of infection in our pilot case series of thoracolumbar revision spine surgery, while one possible subclinical infection had no pathogens identified on any platform. Research using IMC and NGS to detect subclinical infection in revision spine surgery requires further investigation but indicates subclinical infection may be an explanation for some instances of revision surgery, particularly those involving loose or broken pedicle screws. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs. Recent studies have been able to detect the subclinical presence of bacteria in 10%-40% of patients undergoing revision spinal instrumentation procedures, suggesting that clinically silent bacteria may be responsible for presumed aseptic revisions. The purpose of this study is to determine the rate of bacterial presence in revision thoracolumbar spinal instrumentation surgery due to implant failure. In patients without clinical signs or symptoms of infection, we will utilize next generation sequencing (NGS) technology as well as cultures and isothermal microcalorimetry (IMC) to determine the rate of subclinical infection. A prospective pilot case-series at a single academic institution between September 2020 and September 2021. Patients undergoing consecutive thoracolumbar revision spine surgery resulting from implant failure. The main outcome measure was bacterial presence. Routine patient and clinical information including age, sex, body mass index, preoperative indication, and additional clinical/operative information were also identified. Lab values included ESR, CRP and procalcitonin. COVID tests results were documented. Patients undergoing thoracolumbar revision spine surgery due to implant failure by a single surgeon at a single institution were prospectively identified. Tissue and blood samples were taken intraoperatively and transported to a lab for analysis using IMC and NGS. In the pilot study, 22 consecutive patients were included based on prespecified criteria. The average (SD) age of the patients was 65 (14) with 25% female sex. There were 82 total tissue samples obtained (mean 3.7 samples per patient), with a concordance of 93% between cultures and IMC, 85% between cultures and NGS, and 88% between IMC and NGS. The prevalence of occult surgical site infection was detected in 4/22 (18%) cases by NGS, 2/22 (9%) cases by IMC, and 0/22 (0%) by standard cultures. Two out of two (100%) infections detected by IMC and 2/4 (50%) of the NGS infections had grossly loose/broken pedicle screws without clear etiology for breaking, suggesting these are potentially real subclinical infection cases. There was one additional loosening of screws case without pathogen identification by any platform. However, no patients have undergone subsequent surgery for concern for infection to date. For that reason, considering loose/broken screw cases as true subclinical infections IMC had 67% sensitivity, 100% specificity, 100% PPV, and 95% NPV, while NGS was 67%, 90%, 50%, 94% respectively, and cultures were 0%, 100%, 0%, 87% respectively. Pathogens identified include Fusarium equiseti (fungi), Acinetobacter spp., Brevundimonas spp., Lecythophora spp. (fungi), Enterobacter cloacae, Escherichia coli, Enterococcus faecalis, and Staphylococcus haemolyticus. The detection of infection in spinal implants is challenging. The use of IMC and NGS identified two occult cases of infection, while NGS also falsely identified two additional cases of infection in our pilot case series of thoracolumbar revision spine surgery, while one possible subclinical infection had no pathogens identified on any platform. Research using IMC and NGS to detect subclinical infection in revision spine surgery requires further investigation but indicates subclinical infection may be an explanation for some instances of revision surgery, particularly those involving loose or broken pedicle screws.