Page 18 - Towards personalized therapy for metastatic prostate cancer: technical validation of [18F]fluoromethylcholine

Chapter 1
and helps to increase the specificity of PET, by identifying benign causes for enhanced tracer uptake. Secondly, the (low-dose) CT data generate an attenuation map to correct PET data, thus improving the visual quality and quantitative accuracy of the correlated acquisitions. Nowadays, fully three-dimensional PET/CT systems are state of the art, with time-of-flight (TF) capability (i.e., high resolution allowing the exact calculation of the location of the photon’s events) which improves the signal to noise ratio [26].
PET is an extremely sensitive imaging modality for which only small amounts of radiolabelled molecules (tracers, ~nM) need to be injected. The measurement of the tracer distribution allows for quantitative assessment of tissue function without affecting the underlying physiology [27]. Appropriate quantification of a new radiopharmaceutical requires kinetic modeling. This implies dynamic PET scanning starting at injection of the tracer and arterial blood sampling (to acquire the arterial input function and to measure potential tracer metabolites). With current PET scanners, the field of view of such dynamic scans is limited to about 15-25 cm. However, hybrid imaging techniques (e.g., PET/CT, PET/MRI) using new specific tracers might enable a quantitative assessment of response in metastatic sites (e.g., in lymph nodes and bone), using a single, non-invasive scan procedure, in a whole body (WB) setting. Acquisitions in WB setting are essential in the context of malignancy, since these allow characterizing all possible metastatic lesions in the body.
The quantitative approach, although reliable, is technically demanding. For clinical practice it is important to develop simple and accurate quantitative methods, which are complementary to visual image interpretation, thus minimizing the interobserver variability [28]. PET data analysis methods can be divided in three major groups:
1. qualitative analysis or visual assessment,
2. semi-quantitative analysis, including standardized uptake value (SUV)
and lesion-to-background ratio, and
3. absolute quantitative analysis, including (a) nonlinear regression (NLR),
(b) Patlak graphical analysis and (c) simplified quantitative methods.
Visual, qualitative assessment of the attenuated and non-attenuated PET images is the basis of any PET study interpretation. A pathological [18F]fluoromethylcholine ([18F]FCH) avid lesion is defined as a focus with higher choline uptake than the surrounding background, incompatible with accumulation in a physiological site. The major concern regarding this assessment is the significant inter- and intra-observer variability of PET image interpretation [28].
16