Quantitative Na-18F PET/CT Methodologies for Assessing Osteoblastic Tumor Burden

Objectives Standard Uptake Value (SUV) is a metric that has been clinically used for describing tumor burden in PET imaging. Although routinely used for 18F-FDG PET, maximum lesion SUVs for Na18F-avid lesions demonstrate very high values even in benign areas of increased bone turnover. Total lesion fluoride (TLF) and fluoride tumor volume (FTV) are recently described quantitative Na18F measures that may be better descriptors and may enable better prognostic information for malignant/metastatic osteoblastic lesions. The methodologies for determining such metrics can be rapidly established using whole-body threshold-based isocontouring but can be confounded by the inadvertent inclusion of benign regions of increased bone turnover. The objective of this study was to assess clinical methodologies for assessing and quantifying osteoblastic tumor burden in a variety of oncologic patients referred for diagnostic Na18F PET/CT.

Methods This retrospective analysis was performed in 36 oncology patients who underwent baseline Na-18F PET/CT imaging studies. The oncologic tumor types included prostate, breast, thyroid, and head/neck cancers. Following intravenous administration of 372 MBq of Na-18F, whole body PET/CT imaging was performed around 95 min following administration. Visual assessment of 18F-avid osteoblastic lesions was performed by a reader panel. Quantitative isocontour assessment of 18F-avid osteoblastic tumor burden was completed and the following parameters were assessed: highest SUVmax for all osteoblastic lesion (hSUVmax), average of SUVmax values for all 18F-avid osteoblastic lesions, total lesion fluoride uptake (TLF), and fluoride tumor volume (FTV). Isocontours were obtained by either automated threshold-based whole body contouring or manual contouring of individual osteoblastic metastatic lesions.

Results Qualitative assessment of the patient population yielded a spectrum of 18F-avid osteoblastic tumor burden ranging from a few isolated lesions to extensive multifocal osteoblastic disease. This patient population also demonstrated a wide spectrum of osteoblastic degenerative disease burden. Quantitative whole-body isocontour assessment of 18F-avid osteoblastic activity using a minimum threshold SUV of 10 demonstrated sufficient isocontouring for capturing osteoblastic metastatic lesions but it also captures the vast majority of active osteoblastic degenerative disease which confounds quantification efforts. Although more time consuming, quantitative lesion-by-lesion isocontour assessment of 18F-avid osteoblastic metastatic lesions minimizes any false-positive osteoblastic activity from underlying degenerative disease and provides a better assessment of tumor-specific osteoblastic disease burden.

Conclusions In general, the prevalence of underlying degenerative disease in our patient population limits the ability to rely on semi-automated whole body threshold-based isocontouring for the assessment and quantification of tumor-specific osteoblastic disease burden. Although such approaches are faster and require minimal training, the presence of benign osteoblastic degenerative disease may limit its clinical utilization due to high false positive rates or overestimation of tumor-specific disease burden. For nuclear medicine physicians and radiologists, the alternative approach of using lesion-by-lesion quantitative assessment may be more robust and more accurate for assessing tumor-specific osteoblastic disease burden.