The value of 18F-FDG PET/CT in detection of osteomyelitis in patients with stage IV pressure ulcers

HATEM NASR; Mohammed Sambas; Hussein Farghaly; ALQARNI Abdullah; Sami Almutairy

Abstract

2512

Introduction: The management of advanced pressure ulcers is crucially dependent on the extent of sorrounding tissue involvement, specifically the presence of underlying bone involvement. This study aims to evaluate the value of F-18 Fluorodeoxyglucose (FDG) Positron Emission Tomography (PET) / computed tomography (CT) in accurately identifying underlying osteomyelitis (OM) in patients with stage IV pressure ulcers.

Methods: We retrospectively reviewed all patients with stage IV pressure ulcers who had F-18 FDG PET/CT imaging. The study included 35 patients (mean age of 69.11±19.78 years; 22 males). Data collected included patients’ demographic, clinical findings, laboratory results as leukocytic count, CRP and ESR as well as biopsies or tissue cultures and other imaging modalities as standard reference. Images interpretation was performed to assess the extent of osseous involvement corresponding to sites of grade IV pressure ulcers based on FDG uptake regardless of presence or absence of underlying morphological changes on the low dose CT portion of the study. SUVmax value as a quantitative parameter was used to assess the intensity of radiotracer uptake. Lesions were interpreted as positive (increased osseous uptake related to the site of the soft tissue ulcer), negative (no or minimal uptake or uptake not related to ulcer site). Both patient based and lesion based analysis were performed. For patients who had more than 1 grade IV ulcer, the higher SUVmax value was used. We used unpaired student T-test to compare means of continuous values between osteomyelitis positive and negative groups. ROC analysis was performed to assess the best cut off values for SUVmax and CRP to predict underlying osteomyelitis. Chi square test was used to assess difference in frequencies between categorical variables.

Results: Within our study population 22/35 (62.9%) were diabetic and 28/35 hypertensive (80%). The main diagnosis was CVA in 14/35 (40%), spinal cord injury (SCI) in 7/35 (20%) of whom 2 are traumatic, sepsis in 3/35 (8.6%), cardiac in 2/35 (5.7%), lung fibrosis in 2/35 (5.7%), MS in 2/35 (5.7%) while the remaining 5 patients had meningoencephalitis, anoxic brain injury following cardiac arrest, femur fracture, hip replacement and dementia. Ten patients had more than 1 grade IV pressure ulcer (4 patients had 2 ulcers, 4 patients had 3 ulcers and 2 patients had 5 ulcers). Lower grades pressure ulcers including impending ulcers were detected as well in multiple patients though were excluded from our analysis. We studied 50 grade IV pressure ulcer sites (30 sacrococygeal, 16 femoral trochanteric, 2 ischial, 1 calcaneal and 1 over dorsal spine). Tissue cultures were available for 46/50 (92%) sites and all were positive except 3. Biopsies were available for only 5 lesions in 3 (8.6%) patients with chronic osteomyelitis in 2 site and mixed acute and chronic in 3 sites. Nine patients (25.7%) had WBC scan, 12 patients (34.5%) had MRI, 12 patients (34.5%) had bone scan and 9 patients (25.7%) had diagnostic CT scan. The mean SUVmax was higher in patients with OM compared to those with no OM (5.81±0.84 vs 3.83±1.61; p<0.0001) and in lesions with OM compared to lesions with no OM (5.64±1.69 vs 3.35±1.21; p<0.0001). The mean serum CRP (mg/L) was higher in patients with OM compared to those with no OM (120.1±89.58 vs 66.86±42.45; p=0.047). The CRP level showed weal but significant positive correlation with number of pressure ulcers (r=0.378; p=0.025). Based on ROC analysis cutoff value for SUVmax >4.8 was found to best predict osteomyelitis (AUC=0.866; p=0.0001) with sensitivity, specificity, PPV, NPV and accuracy of 67.86%, 90.91%, 90.5%, 69.0% and 78% respectively (p<0.0001) while the best cutoff for CRP was >62 mg/L (AUC=0.663; p=0.0329) with sensitivity, specificity, PPV, NPV and accuracy of 88.89%, 39.13%, 63.2%, 75.0% and 68.0% respectively (p<0.013).

Conclusions: F-18 PET/CT could potentially represent a valuable and promising tool in detection of active OM associated with stage IV pressure ulcers. The higher SUVmax showed relatively high specificity though with suboptimal sensitivity while serum CRP level had fairly high sensitivity but with poor sensitivity.

Although other imaging modalities as MRI, radiolabeled WBC scan or bone scan could be helpful in detection of osteomyelitis / osseous involvement in similar clinical scenarios, head to head comparison with PET/CT requires further research.