Factors associated with sequelae after treatment of hematogenous pyogenic vertebral osteomyelitis

Factors associated with the time required for CRP normalization in pyogenic spondylitis: A retrospective observational study

Background

Treatment for pyogenic spondylitis tends to be prolonged; however, few studies have examined the factors associated with the time required for infection control. Therefore, we analyzed a consecutive cohort of patients to identify factors associated with the time required to control infection in pyogenic spondylitis. This study aimed to clarify the factors linked to the duration necessary for achieving infection control in cases of pyogenic spondylitis, using C-reactive protein (CRP) normalization as an indicator.

Methods

In this retrospective observational study, we investigated 108 patients diagnosed with pyogenic spondylitis. We evaluated the number of days from the first visit to CRP normalization; for cases wherein CRP did not normalize, the number of days to the date of final blood sampling was evaluated. In the present study, infection control in pyogenic spondylitis was defined as a CRP falling within the normal range (≤0.14 mg/dL). We performed univariate and multivariate Cox regression analyses to identify various factors associated with the time required for CRP normalization in pyogenic spondylitis.

Results

The mean time required for CRP normalization was 148 days. Univariate Cox regression analysis showed that the serum creatinine level, estimated glomerular filtration rate (eGFR), lymphocyte percentage, neutrophil percentage, CRP level, CRP-albumin ratio, and neutrophil-to-lymphocyte ratio were significantly associated with the time required to control infection. Multivariate Cox regression analysis showed that a higher neutrophil percentage, diabetes mellitus, and a lower eGFR were the independent factors associated with a longer infection control time.

Conclusions

We found that a higher neutrophil percentage, diabetes mellitus, and a lower eGFR were significantly associated with a longer time for CRP normalization in pyogenic spondylitis. These findings may help identify patients with pyogenic spondylitis who are at a high risk for an extended infection control period.

Clinical phenotyping of spondylodiscitis and isolated spinal epidural empyema: a 20-year experience and cohort study

Background

The incidence of spondylodiscitis (SD) and isolated spinal epidural empyema (ISEE) has been increasing in the last decades, but the distinct differences between both entities are poorly understood. We aimed to evaluate the clinical phenotypes and long-term outcomes of SD and ISEE in depth.

Methods

We performed a chart review and analyzed data from our cohorts of consecutive SD and ISEE patients who were treated and assessed in detail for demographic, clinical, imaging, laboratory, and microbiologic characteristics at a university neurosurgical center in Germany from 2002 to 2021. Between-group comparisons were performed to identify meaningful differences in both entities.

Results

We included 208 patients (72 females: age 75 [75 32-90] y vs. 136 males: 65 [23-87] y, median [interquartile range], p < 0.001), of which 142 (68.3%) had SD and 66 (31.7%) had ISEE. Patients with SD were older than ISEE (ISEE: 62 y vs. SD: 70 y, p = 0.001). While SD was more common in males than females (males: n = 101, 71.1% vs. females: n = 41, 28.9%, p < 0.001), there was no sex-related difference in ISEE (males: n = 35, 53.0% vs. females: n = 31, 47.0%, p = 0.71). Obesity was more frequent in ISEE than in SD (ISEE: n = 29, 43.9% vs. SD: n = 37, 26.1%, p = 0.016). However, there were no between-group differences in rates of diabetes and immunodeficiency. In the entire study population, a causative pathogen was identified in 192 (92.3%) patients, with methicillin-susceptible staphylococcus aureus being most frequent (n = 100, 52.1%) and being more frequent in ISEE than SD (ISEE: n = 43, 65.2% vs. SD: n = 57, 40.1%, p = 0.003). SD and ISEE occurred most frequently in the lumbar spine, with no between-group differences (ISEE: n = 25, 37.9% vs. SD: n = 65, 45.8%, p = 0.297). Primary infectious sources were identified in 145 patients (69.7%) and among this skin infection was most common in both entities (ISEE: n = 14, 31.8% vs. SD: n = 25, 24.8%, p = 0.418). Furthermore, epidural administration was more frequent the primary cause of infection in ISEE than SD (ISEE: n = 12, 27.3% vs. SD: n = 5, 4.9%, p < 0.001). The most common surgical procedure in SD was instrumentation (n = 87, 61%) and in ISEE abscess evacuation (n = 63, 95%). Patients with ISEE displayed lower in-hospital complication rates compared to SD for sepsis (ISEE: n = 12, 18.2% vs. SD: n = 94, 66.2%, p < 0.001), septic embolism (ISEE: n = 4/48 cases, 8.3% vs. SD: n = 52/117 cases, 44.4%, p < 0.001), endocarditis (ISEE: n = 1/52 cases, 1.9% vs. SD: n = 23/125 cases, 18.4%, p = 0.003), relapse rate (ISEE: n = 4/46, 8.7% vs. SD: n = 27/92, 29.3%, p = 0.004), and disease-related mortality (ISEE: n = 1, 1.5% vs. SD: n = 11, 7.7%, p = 0.108). Patients with SD showed prolonged length of hospital stay (ISEE: 22 [15, 30] d vs. SD: 38 [29, 53] d, p < 0.001) and extended intensive care unit stay (ISEE: 0 [0, 4] d vs. SD: 3 [0, 12] d, p < 0.002).

Conclusions

Our 20-year experience and cohort analysis on the clinical management of SD and ISEE unveiled distinct clinical phenotypes and outcomes in both entities, with ISEE displaying a more favorable disease course with respect to complications and relapse rates as well as disease-related mortality.

Can we predict favourable quality of life after surgically treated vertebral osteomyelitis? Analysis of a prospective study

PURPOSE

Vertebral osteomyelitis (VO) is a severe clinical entity associated with significant morbidity and mortality. Several studies have showed that successful treatment of VO patients leads to significantly improved quality of life (QoL). Nevertheless, QoL levels of these patients remained below those of the general population. There are rarely studies focusing on predicting factors for favourable QoL after surgically treated VO. The aim of this study was to identify factors influencing positively the QoL of patients undergoing surgery for VO.

METHODS

We conducted a prospective monocentric study including surgically treated VO patients from 2008 to 2016. Data were collected before (T0) and 1 year (T1) after surgery. Primary outcome was favourable QoL defined as back pain with disability restricting normal life activity with a cutoff value ≥ 12 on Oswestry Disability Index (ODI).

ETHICS

Ethical approval was given by the Faculty of Medicine at the University of Cologne (09-182).

RESULTS

A total of 119 patients surviving 1 year after surgically treated VO were analysed. Favourable QoL was achieved in 35/119 patients. On multivariate analysis, younger age (hazard ratio = HR: 0.95; 95% CI 0.91-0.99; p = 0.022), lower albumin (HR: 0.9; 0.83-0.98; p = 0.019) an ASA score ≤ 2 (HR:4.24; 95%CI 1.42-12.68; p = 0.010), and a lower preoperative leg pain on the VAS (HR: 0.86; 95% CI 0.76-0.97; p = 0.018) were identified as independent risk factors for favourable QoL. Interestingly, the absence of neurological deficits was not predictive for a favourable outcome by means of QoL.

CONCLUSION

One-third of surgically treated VO patients (29%) in our cohort achieved favourable QoL by means of ODI. Our findings can facilitate an estimation of the prognosis when informing the patient before surgery, and underscore that spine disability questionnaires, such as ODI, measuring QoL, are mandatory to evaluate comprehensively the outcome of this entity.

The Burden of Vertebral Osteomyelitis—An Analysis of the Workforce before and after Treatment

Although vertebral osteomyelitis (VO) has a major impact on morbidity, functional status, and quality of life, data concerning the influence on the patient’s ability to work (ATW) are lacking. Therefore, the aim of this study was to analyze the work status after VO-treatment as well as risk factors associated with loss of the ATW. We conducted a post-hoc analysis of data from a prospective VO-registry (2008–2019) supplemented by workforce data. Primary endpoint was the work status after one year (T1). Univariate analysis comparing patients’ characteristics “at-work” versus “not-at-work” at T1 was performed. Of a total of 335 VO-patients, n = 52 (16%) were part of the workforce at time of diagnosis (T0), of which 22 (42%) failed to be part of the workforce at T1. A higher number of comorbidities and a body mass index (BMI) < 25 kg/m² were associated with a reduced ATW. VO in working age patients is a debilitating condition and associated with reduced patients’ ATW. Patients engaged in heavy physical work mostly had a BMI < 25 kg/m² and therefore were more severely affected and no longer able to keep their workforce. More support in retraining should be offered after successful treatment to maintain ATW and reduce the socio-economic burden.

Clinical characteristics of pyogenic vertebral osteomyelitis, and factors associated with inadequate treatment response

OBJECTIVES

Pyogenic vertebral osteomyelitis (PVO) is associated with a high burden of disease. Our study aimed to describe characteristics at presentation of PVO, the risk of inadequate treatment response (ITR), relapse, and death, and to determine risk factors for ITR.

METHODS

Patients with an ICD-10 discharge code for PVO and admission to a major Danish university hospital between November 2016 and April 2019 were included. ITR was defined as clinical, microbiological, and/or radiological progression during treatment. Data were collected through review of medical records, and logistic regression was used to determine adjusted odds ratios (aOR).

RESULTS

Of 106 patients included, 87% presented with pain in the spine, 97% elevated CRP, 14% severe sepsis, and 13% with a history of previous spinal surgery. 39% were infected with Staphylococcus aureus and 9% with Escherichia coli. 31% responded inadequately to treatment, and risk factors for ITR were previous spinal surgery (aOR 19.29; 95% confidence interval (CI) 2.20-169.08), severe sepsis (aOR 4.59; 95% CI 1.28-15.41), and infection with Escherichia coli (aOR 8.10; 95% CI 1.71-38.45). 13% experienced relapse within the first 2 years, while the 1-year crude mortality was 12%.

CONCLUSION

Staphylococcus aureus is still the main pathogen in PVO patients, and the risks of relapse and mortality remain high. Factors found to be associated with ITR were previous spinal surgery, severe sepsis, and infection with Escherichia coli.

Microbial Etiology of Pyogenic Vertebral Osteomyelitis According to Patient Characteristics

Background

It is difficult to select an appropriate empirical antibiotic treatment regimen for patients with culture-negative pyogenic vertebral osteomyelitis (PVO). Having knowledge of the distribution of microorganisms according to patient characteristics can help clinicians make informed choices regarding empirical antibiotics. The aim of this study was to determine the microbial distribution among individuals with PVO according to their demographic and clinical characteristics.

Methods

We reviewed the medical records of patients admitted to our hospital with culture-confirmed PVO between January 2005 and December 2017 and collected data on demographics, underlying diseases, and radiographic and microbiological results. Statistical analysis was performed to identify associations between specific bacteria and specific patient characteristics.

Results

A total of 586 patients were included in the study. The prevalence of Staphylococcus aureus infections was higher in young patients than in old patients, while gram-negative bacterial infections and Enterococcus were more prevalent in older patients. Gram-negative bacterial infections were more common in women than in men (32.1% vs 16.4%; P < .05), in patients with cirrhosis than in those without (32.7% vs 21.1%; P < .05), and in patients with a solid tumor than in those without (31.0% vs 20.7%; P < .05). Methicillin-resistant S. aureus infections were more prevalent in patients with chronic renal disease than in those without (34.4% vs 14.7%; P < .05).

Conclusions

The microbial etiology of PVO varies according to patient characteristics. Patient characteristics should thus be considered when choosing empirical antibiotics in patients with culture-negative PVO.