Prospective study on the effect of short-term androgen deprivation therapy on PSMA uptake evaluated with ⁶⁸Ga-PSMA-11 PET/MRI in men with treatment-naïve prostate cancer

Currently, small-molecule imaging with gallium- or fluoride-labelled prostate-specific membrane antigen (⁶⁸Ga/¹⁸F-PSMA) has been rapidly taken into clinical use in many European countries [1‐3], although its utility in primary staging of prostate cancer still needs further validation [4‐6]. At present, according to the majority of published data, the main indication of ⁶⁸Ga-PSMA PET imaging is re-staging in presence of biochemical recurrence, especially at low prostate-specific antigen (PSA) values [7]. However, in recent years, there has been also a growing evidence on the promising role of ⁶⁸Ga-PSMA PET imaging in nodal and distant staging in patients with high-risk disease [8].

Based on in vitro studies and animal models, it is known that administration of androgen deprivation therapy (ADT) increases PSMA expression [9, 10]. Although this notion was primarily published more than 20 years ago [9], first clinical case report was published only 2 years ago. Hope et al. demonstrated a 7-fold increase in maximum standardised uptake value (SUVmax) of PSMA uptake after the initiation of ADT [10]. However, in two recent series, the effect of ADT on PSMA-PET findings was deemed as heterogeneous [11, 12]. However, in both of these series, all the metastases were analysed as whole and no region-based analysis was performed. In addition, no data is present if increase is also seen in physiologically avid organs. According to previous published data, we hypothesised that ADT might improve the performance of PSMA-PET imaging in primary staging of prostate cancer. In addition, the hypothesis was that different regions of prostate cancer (primary tumour in prostate and metastases in lymph nodes, bone, viscera) and physiologically avid organs as well (salivary glands, kidneys, liver, and spleen) respond differently to ADT.

The purpose of the current study is to demonstrate in patient and region-based analysis of the time course effect of ADT on PSMA uptake observed in repeated ⁶⁸Ga-PSMA-11 PET/MRI scans in men with newly diagnosed, treatment-naïve prostate cancer patients.

Nine patients were included in the study. Patients’ characteristics are shown in Table 1. At baseline, five had metastatic prostate cancer (other than regional lymph nodes and/or bone), while four patients presented with local or locally advanced disease. Visceral metastases were not present in any of the patients. All patients reached castration levels (serum testosterone < 1.7 nmol/L) within a period of 10 days after the initiation of ADT. Eight patients completed all the four PET/MRI scans, while one patient performed only three scans due to study withdrawal.

Patient-based observations are depicted in Fig. 1. In two patients (patients no. 2 and no. 7), no increase in SUVmax was observed, whereas in 7 (78%) patients, a heterogeneous change in PSMA uptake occurred. In patient no. 8, who had two bone metastases already at baseline, three new bone metastases were observed post-ADT.

A region-based analysis of the primary tumour in prostate, metastatic lesions, and the uptake pattern in physiologically avid organs is depicted in Figs. 2 and 3, and in Table 2. Also, the specific locations of the metastases are presented in Table 3. All the lesions in baseline and in follow-up, except for a discordance in two parailiac lymph node metastases, were detected by the two readers, Cohen’s Kappa 0.89 (95% CI, 0.79–0.99). After a consensus reading, in total, 16 prostate, 16 lymph node, and 23 bone lesions were detected and analysed.

There was a marked increase in SUVmax (maximum increase in SUVmax of 76%) in more than half of the bone metastasis and a less pronounced and not so frequent increase in lesions in prostate (29%), and lymph nodes (19%). The increase was observed within the first 3 to 4 weeks post-ADT. In all lesions, which were considered “decrease”, the maximum SUVmax decrease was 50% or less. Despite the decrease, none of the lesions disappeared during the follow-up.

The most pronounced increase in physiologically avid organs was observed in parotid glands (23%), and submandibular glands (22%). In other physiologically avid organs, the increase was less than 20%. The decrease was seen very seldom or not at all, and the mean decrease was less than 20% (Fig. 4).

In this prospective registered study, a heterogeneous increase of PSMA uptake was observed after ADT in treatment-naïve prostate cancer patients in repeated ⁶⁸Ga-PSMA PET/MRI scans. The most evident increase in SUVmax was observed 3 to 4 weeks post-ADT especially in bone metastasis. In one patient, with bone metastases already at baseline, three new bone metastases, which were already visible anatomically on MRI images, were observed. Of those lesions with decreasing SUVmax, none disappeared.

It has already been demonstrated that in both androgen-sensitive and androgen-resistant human prostate adenocarcinoma cells (LNCaP), the expression of PSMA is upregulated post-ADT and downregulated in the presence of testosterone or DHT [14‐16]. Although a case report published by Hope et al. corroborated these preclinical studies, demonstrating a 7-fold increase in PSMA SUVmax values after the initiation of ADT, our results are in line with the two very recently published studies [10‐12]. Aggarwal et al. performed a study in eight patients, scanned with ⁶⁸Ga-PSMA-11 PET before and after the initiation of ADT within a variable period of 2–4 weeks. The study demonstrated a heterogeneous increase in SUVmax, in 68% and in 41% of lesions in castration-sensitive men (n = 4) and in castration-resistant men (n = 4), respectively. The other study, conducted by Emmet et al., studied also both castration-sensitive (n = 8) and castration-resistant men (n = 7). Although they showed that the increase was more evidently seen in castration-resistant men, in the light of our data and the study by Aggarwal, it seems that the increase occurs also in castration-sensitive men. In fact, in the study by Emmet et al., of those eight men with castration-sensitive disease, half of them had Gleason score 7 prostate cancer. It might well be that the increase is more evidently seen in poorly differentiated tumours. Also, our study clearly demonstrates the biological difference between testosterone flare and PSMA flare. Since degarelix directly inhibits the action of GnRH, no PSA flare occurs, and serum testosterone decreases rapidly. Despite the rapid decrease in testosterone levels, the increase in PSMA uptake was still observed.

ADT is the gold standard treatment in patients with metastatic prostate cancer. The effect of continuous long-term ADT on reducing the visibility of castration-sensitive prostate cancer lesions on PSMA-PET has already been investigated; however, it is still uncertain if initiation of ADT could interfere with the staging results [17]. According to our preliminary findings, as none of the lesions disappeared during the observational period, ongoing short-term ADT does not represent a contraindication on performing a staging PSMA-PET. Although ADT did not significantly increase PSMA-PET staging performance as only in one patient new metastases were found, PSMA activity in the majority metastatic lesion reached the highest uptake 3 to 4 weeks post-ADT. Therefore, this time window might improve the detection rate, providing better lesion to background ratio. This aspect might be interesting especially in an oligometastatic disease, where bone metastases detected could then be selectively treated with radiotherapy. However, further studies are warranted to investigate the possible clinical impact of the phenomenon.

Since this is the first study where changes in PSMA SUVmax values were analysed on a region-based analysis, the fact that we observed the increase most evidently in bone metastasis is also interesting. This notion is supported also by two case reports demonstrating a similar increase of PSMA uptake in bone metastasis [18, 19]. Although the biology of the phenomenon is not known, authors hypothesised that increase of PSMA uptake in bone might be caused by osteoblastic turn over or bone reparation processes. However, most probably, this is not the case, since we observed the increase also in lymph nodes, in primary tumours, and in some of the physiologically avid organs suggesting a more general rather than organ-specific mechanism. This is, in fact, the first study to report that changes in PSMA-PET findings post-ADT are not restricted only to tumour tissue.

One might also question whether this flare phenomenon is dependent on the different mechanisms of action of ADT treatments. However, our results with GnRH antagonist therapy are similar with previous studies, in which LhRH agonist, antiandrogens, or new androgen signalling pathway modulators were administered [10‐12, 16]. Given these facts, it seems rather evident that PSMA is connected to the androgen pathway [10‐12, 16]. More recently, also mTOR pathway with mTOR inhibitor, rapamycin has shown to be linked to increase in PSMA uptake [16, 20]. Taken this all together, one might hypothesise that the observed effects of ADT on PSMA uptake in lesion level depict the androgen sensitivity of the specific lesions and also potentially the heterogeneity in aggressiveness of the lesions. Therefore, could it be possible to select those lesions that are prone to progress or are insensitive to ADT and selectively treat only those lesions? To understand this phenomenon, further studies with larger number of patients and longer follow-up are warranted.

Moreover, understanding what lies behind this phenomenon might raise a great interest from a theragnostic perspective. A preclinical study on a mouse model of castration-resistant prostate cancer (CRPC) demonstrated that pre-treatment with enzalutamide for 21 days followed by 177Lu-PSMA radioligand therapy (RLT) resulted in a significantly enhanced RLT-induced DNA-damage. However, pre-treatment with androgen receptor blockade did not show any additive effect on tumour growth reduction, suggesting that ADT might not necessarily guarantee an increased efficacy of RLT [21]. Nevertheless, considering the heterogeneity of PSMA expression and the lack of clinical studies about this specific theragnostic aspect, the possible synergistic effect of ADT + RLT still needs to be defined.

Moreover, although ¹⁷⁷Lu-PSMA radionuclide therapy is at the present moment used in castration-resistant prostate cancer patients, there are ongoing clinical trials to study its use also in castration-sensitive patients [22]. If proven to be effective in castration-sensitive men, according to our data, it would be reasonable to hypothesise that the timing between administration of ADT and of ¹⁷⁷Lu-PSMA radionuclide therapy is crucial. ADT could increase the tumour targeting and therefore increase the efficacy of ¹⁷⁷Lu-PSMA treatment during the time window of maximum PSMA uptake.

Also, our finding of increased PSMA uptake in normal salivary glands post-ADT needs to be taken into account as a possible increased therapy-related risk factor for significant xerostomia in patient candidates for ¹⁷⁷Lu-PSMA therapy.

The study has some limitations. First, the cohort is small and underpowered to infer firm conclusions. In addition, there is variability in time intervals of the scans between the different patients. Therefore, the results should be considered preliminary and the study as a proof of concept. However, all the patients were thoroughly examined, and eight of the nine patients completed the study by undergoing four sequential PSMA-PET/MRI scans. In addition, the strength of the study is its truly prospective and registered nature. Second, the cohort is heterogenous. However, this can also be seen as an advantage since with this cohort, we are able to demonstrate that, although the increase is minor, PSMA flare is not merely seen in metastatic patients.

Both in patient and region level, a heterogeneous increase of PSMA uptake was observed post-ADT, observed most evidently in bone metastases. The highest response on PSMA uptake was observed 3 to 4 weeks after ADT. Although the impact of ADT on ⁶⁸Ga-PSMA PET staging performance was minor in this small patient cohort, more research is needed to investigate whether ADT could significantly improve detection rate and have clinical impact in patients with oligometastatic disease. Moreover, results were encouraging that short-term usage of ADT does not seem to represent a contraindication to perform 68Ga-PSMA PET for staging purpose, since none of the lesions disappeared.