The current diagnostic tools of serum prostate-specific antigen (PSA), digital rectal examination (DRE) and transrectal ultrasound scan (TRUS)-guided biopsies to provide pre-surgical tumor grading of prostate cancer are only accurate in around 69% of patients [5].

Most prostate malignancies show increased uptake of choline-PET tracers. However, uptake in benign prostate hypertrophy has been shown, and some report an inability of these tracers to differentiate between benign and malignant prostate tissue [6]. A sensitivity of up to 90% and specificity of 86% have been reported for the detection of localized prostate malignancy [7] with choline PET/CT, but the accuracy is lower for smaller tumors.

There is not enough evidence currently to support the use of choline-PET/CT, or other tracers, for screening patients for malignancy. There may be a role for guiding biopsies in patients who have repeated negative prostate biopsies despite a high clinical suspicion [8] but this remains an area of research interest at present.

Given the uncertainty of the accuracy of choline-PET in differentiating benign from malignant tissue, the value of this technique for T-staging prostate tumors is limited. The spatial resolution of clinical PET/CT scanners in widespread use is insufficient to accurately assess the prostate capsule for evidence of involvement or breach. The development of PET/MRI may show T-staging benefits, as suggested in a 15-patient feasibility study using ¹⁸F-choline PET/MRI [9] (Figs. 5.3 and 5.8).

Identifying involved lymph nodes at diagnosis (N-staging) has significant clinical significance, but has been difficult to achieve accurately with all imaging modalities, including MRI. Contractor et al. showed that ¹¹C-choline PET/CT was more sensitive than MRI for nodal staging (p = 0.007), detecting more sub-centimeter involved nodes [10]. Whilst choline-PET/CT has demonstrated good specificity, the sensitivity is relatively low and is dependent on the size of the involved lymph node and the PSA levels. De Jong et al. reported sensitivity/specificity values of 80%/96%, respectively, but the mean PSA for the 67 patients studied was over 100 ng/ml. In contrast, Beheshti et al. reviewed 130 patients with a mean PSA of 27 ng/ml (suggesting earlier disease and/or less disease burden) and reported a sensitivity of only 45% for nodal analysis, but a specificity of 96% [8]; the sensitivity increased to 66% if only nodes larger than 5 mm were considered. Other studies have demonstrated similar values of sensitivities and specificities [10–12].

The diagnosis of distant metastatic disease has important treatment implications; metastatic prostate cancer is incurable, and therefore invasive and morbid treatment to the primary disease is unlikely to be appropriate. Prostate cancer most frequently spreads to the bone causing typically sclerotic deposits. The current most common imaging method for screening for metastatic bone disease is with standard scintigraphy using technetium-labeled diphosphonates which are incorporated into the bone matrix of metastatic deposits secondary to the excess osteoblastic activity. ¹⁸F-fluoride, as a PET tracer, has a similar mechanism of uptake, but offers potential benefits from the resolution of PET/CT imaging, offering quantification potential and providing tomographic information as routine (Fig. 5.12). Faster clearance also allows imaging as early as 1 h post-injection. Choline PET/CT has been compared with standard bone scintigraphy in prostate cancer patients; Picchio et al. reported a sensitivity for identifying bone metastases of 89% for ¹¹C-choline PET/CT and 100% for bone scintigraphy, but the specificity was much greater for ¹¹C-choline PET/CT at 98 vs. 75% for bone scintigraphy [13]. Similar results have been reported by other groups [14]. This advantage of choline as a tracer is likely because there is little increased uptake in chronic degenerative lesions, unlike with standard ^99mTc bone scintigraphy. Beheshti et al. reported that in one study, ¹⁸F-choline PET/CT identified early bone marrow involvement that was not visible on CT alone [8]. No evidence currently demonstrates the superiority of choline PET/CT compared with standard staging techniques for the identification of bone metastases from prostate cancer, but it may have value in certain individual cases for problem solving (Fig. 5.4).

Imaging needs to identify sites of disease relapse, in particular whether this relapse is local to the prostate, within local or distant lymph nodes or distant metastatic spread. This has important treatment implications; a confined local recurrence might still be cured with salvage treatment. It is not uncommon for prostate cancer patients to have disease recurrence suspected by serial serum PSA rises. TRUS-guided biopsy only detects local recurrence in about 25–54% of these patients and is particularly poor when PSA values are low [15, 16]. CT has only a low diagnostic accuracy for localizing recurrent disease [17].