In this issue of the EJNMMI, a study evaluating a promising new PET tracer for imaging prostate cancer (PCa), 68Ga-PSMA (PSMA), is published [1]. The paper deals with a retrospective analysis in 37 patients with biochemical recurrence of PCa who underwent both 18F-fluoromethylcholine (CHO) and PSMA PET/CT for the purpose of restaging. The aim of the study was to compare the diagnostic performance of the novel tracer with that of CHO. On a patient basis, the detection rates were 70.3 % and 86.5 % for CHO and PSMA, respectively. PSMA also showed a better performance at low PSA values. The authors conclude that PSMA PET/CT can detect PCa relapse and metastasis with significantly improved contrast when compared with CHO PET/CT. This advantage is related to higher tracer uptake by PCa lesions and low background signal, which allow the detection of small lymph node, bone and liver metastases. Although innovative and interesting, this study had some limitations: it was retrospective; few lesions were confirmed by histology; several criteria were used to validate the uptake areas; the results were compared against CHO PET/CT, which was considered as a standard even though this modality is not established; and, finally, the impact on patient management is not discussed.

The limitations of the study preclude recommendation of this new radiopharmaceutical in oncological guidelines, a fate that has befallen many other promising PET tracers. This is a critical issue for the future of nuclear medicine and deserves a commentary. The way we introduce new radiopharmaceuticals into the clinical arena is highly influenced by published data and we believe that efforts are needed to enhance the impact of our reports in the world of oncology.

Although many readers are well aware of PCa imaging, let us briefly summarize the state of the art of PET radiopharmaceuticals that are used to study PCa.

PCa is a considerable health issue and has displayed an increasing incidence worldwide during the last decade [2] – a trend that justifies the burgeoning medical interest in this disease. We can say that PCa is studied as much in men as breast cancer is studied in, mostly, women. As always in oncology, accurate detection of disease spread is crucial for treatment decisions and imaging could play a major role in identifying tumour extension, provided that sufficient diagnostic accuracy is properly demonstrated. Of course, PET imaging is competing on many levels with other diagnostic modalities, but we will not cover these aspects. As a consequence of the great clinical interest in PCa, several PET radiopharmaceuticals have been investigated for their ability to detect the disease [3]. The role of the “pan-tumour” radiopharmaceutical FDG is limited, mainly due to its low sensitivity, particularly in patients with a low Gleason score, low serum PSA values and localized disease [4]. However, it has been suggested that FDG may be useful when applied in specific clinical settings, such as in staging high-risk patients, in the assessment of the spread of hormone-resistant disease, in the evaluation of treatment response, and in the assessment of prognosis in patients with metastatic PCa [5]. Nevertheless, the limited availability of evidence means that it is difficult to go beyond the hypothesis of a possible utility of FDG in aggressive disease.

Among the different PET tracers proposed for PCa, choline derivatives are the most commonly used, especially in patients with biochemical failure [6]. Acetate and CHO have comparable accuracy in visualizing PCa, but most of the recent studies have focused on CHO [7]. Practical issues led to the development of fluorinated compounds, such as 18F-fluoroethylcholine and CHO, to overcome the limitations of 11C-labelled radiopharmaceuticals. These various CHO-based tracers have shown some differences in their biokinetics [8], and it is not yet established which radiopharmaceutical is more accurate because there have been no direct comparative studies on individual compounds. However, no significant differences have been observed in the clinical setting, and running such a comparative study would not be of any clinical value. Although not validated by prospective, randomized clinical trials, CHO is increasingly being used for imaging of primary and recurrent PCa [9]. For the diagnosis of primary PCa, CHO has shown controversial results due to the heterogeneity of the patient groups included in the published experience [10]. Choline has demonstrated low sensitivity for the detection of primary prostate lesions and tumour configuration has been shown to affect diagnostic accuracy. Additionally, specificity is limited because PCa cannot be distinguished from prostatic hyperplasia, prostatitis and high-grade intraepithelial neoplasia [11]. For these reasons, the use of CHO cannot be recommended as a procedure for PCa detection. At least, however, we know when not to use this technique.

In staging, it is usually stated that the use of CHO should be restricted to patients with high-risk disease. Clinical factors, high Gleason scores (>8) and high PSA levels ( >20 ng/ml), should be considered in patient selection in order to ensure that there is some benefit from the use of this procedure [12]. Nevertheless, data are also lacking on this possible application.

In cases of biochemical failure after primary treatment, the accuracy of CHO in localizing recurrent disease has been extensively reported, but mainly in retrospective studies [6, 13]. The higher the PSA level and the shorter the PSA doubling time, the better will be the predictive value of this imaging modality [14]. Indeed, when the PSA level is >1 ng/ml, CHO has better accuracy than standard diagnostic procedures [15], and a PSA doubling time of less than 3 months to a maximum of 6 months has been suggested to be predictive of CHO positivity [16–18]. It is well known that the choice of treatment depends on disease spread [19], and CHO has been used to guide surgical and radiation treatments [20–22]. In particular, it has been postulated that CHO is useful in salvage radiation treatment planning as it allows accurate definition of radiation field extension and guidance of dose escalation to positive areas [23]. When validated, this may offer a brilliant example of personalized medicine, not to mention the use of negative CHO results to propose a “watchful waiting” approach to patients.

Based on the available knowledge, there is insufficient evidence to support the withdrawal of antiandrogen therapy on a regular basis before CHO imaging in patients with hormone-resistant PCa [24, 25]. Again, we at least know what we should not do.

An apparently promising radiopharmaceutical is 18F-labelled anti-1-amino-3-18F-fluorocyclobutane-1-carboxylic acid (FACBC), a new synthetic amino acid that was developed for assessment of the anabolic component of tumour metabolism. FACBC shows high uptake in PCa deposits and appears to be safe for use in humans [26]. Preliminary data have shown that it might be superior to 11C-CHO in the detection of recurrent PCa, but to date the only available report is that of a study in a small series of patients in which the results did not reach statistical significance and two non-established diagnostic modalities were compared, with no histological confirmation of the uptake areas [27].

The second “new” radiopharmaceutical, reported in this issue of the EJNMMI, is PSMA. The molecule is a type II transmembrane protein with glutamate-carboxypeptidase activity that is overexpressed in PCa [28, 29]. PSMA expression levels seem to be directly related to androgen independence, metastasis and progression of the tumour [30], and this is why it has been widely investigated for the development of several radiopharmaceuticals, for both PET and SPECT [31–36]. The 68Ga-labelled PSMA ligand has been previously investigated, with studies suggesting a better sensitivity compared with CHO [37, 38].

A major problem commonly encountered in clinical studies employing new diagnostic modalities, such as PET/CT radiopharmaceuticals, lies in the difficulty of assessing the accuracy of the technique. Histological verification of findings is often not possible for practical and ethical reasons. One of the best, or worst, examples of this critical situation is the use of choline in PCa. The clinical setting in which this radiopharmaceutical could be more useful is precisely that in which the validation process for any finding is particularly difficult. This is why, although many papers have been published reporting clinical studies using CHO and many centres throughout Europe are using this technique in clinical routine, CHO is still considered experimental and is not recommended in guidelines [39–41]. In fact, comparative studies of CHO with standard approach methodologies, which are recommended in clinical guidelines, are scarce, and randomized controlled trials evaluating the impact of CHO on patient management and outcome have still not been published. This situation is in a way a paradox. In fact, since imaging is almost neglected in all clinical guidelines on PCa, it would be easy to demonstrate with a properly designed and conducted clinical trial the superiority of CHO, or any other radiopharmaceutical, over CT or bone scan, for both staging and biochemical relapse.

We are of the opinion that a strong behavioural change is needed in the attitude of the nuclear medicine community towards imaging validation.

The good old days when everyone was allowed to use every tracer (of course having demonstrated fulfilment of the radiation protection requirements) are definitively over. Now we have to compete with several other diagnostic modalities, most of which do not use ionizing radiation and are cheaper and less complicated. So, we have to demonstrate that PET radiopharmaceuticals are beneficial for our patients and that the community has sufficient reason to pay for their use. The level of evidence that is needed to accept a diagnostic modality is a matter of debate, and here we have to enter the scientific and political arena and fight to obtain a different, and easier, way to gain acceptance of a diagnostic modality in clinical guidelines. We should immediately start this effort in simple ways, keeping in mind that readily available, easy-to-use and approved radiopharmaceuticals are the only way to ensure that nuclear medicine “stands clear” in the clinical world. Let us just start by designing proper studies aimed at demonstrating, not at postulating. Let us gather data from different centres, with significant numbers of patients, and let us delay publication until we have data on outcome or on surrogate markers of outcome.

We have to publish our results in order that the full value of radiopharmaceuticals is recognized within the oncology community, with the proviso that we are able to demonstrate this value. In other words, we should aim to make full use of PET’s potential in clinical practice, but we should also write papers and publish results in the proper way. That is to say: writing PET into existence.