PROstate Cancer Imaging, Treatment and Toxicity (PROCITT)

1. ADT induced Osteoporosis Prostate cancer is a common malignancy in Australian men. In men with localized disease at the time of diagnosis, baseline PSA level, tumour stage and Gleason grade can be used to help stratify into risk categories. Men with high risk disease are defined by an absence of metastatic disease using conventional imaging, and any one of the following: a presenting PSA of >20, Gleason grade 8-10 disease on histology, or stage T3-4 disease.[1] Such men are often treated with a combination of radiotherapy to the prostate and pelvic lymph nodes, in conjunction with a course of adjuvant androgen deprivation therapy (ADT) of between 18-36 months.[2] Recent literature suggests that the greatest benefit from adjuvant ADT comes from the first 4-6 months of treatment, and although there is measurable benefit from prolonging the course of ADT, it follows the law of diminishing returns with progressively smaller benefit per unit of increased treatment time.[3] This is important, in that if cumulative toxicities are being inflicted by prolonging the treatment, there is likely to be a duration where the harm of further treatment will start to outweigh the shrinking disease control benefits. With greater clinical experience of the use of adjuvant ADT, there has become a better awareness of the toxicities associated with this treatment. Accelerated loss of bone mineral density has long been recognized as a complication of being hypogonadal. There is now good evidence that this leads to an approximately 7% higher risk of fractures for men with prostate cancer managed with ADT.[4] Osteoporotic fractures are associated with increased morbidity and mortality, and a high proportion of patients who suffer them never fully regain their pre-fracture level of functioning. There are Australian guidelines for the management of osteopaenia / osteoporosis for men managed with ADT.[5, 6] They recommend monitoring of bone mineral density (BMD) using annual DEXA scanning, supplementary Vitamin D with Calcium, and the use of bisphosphonate therapy for men with prevalent minimal trauma fracture or baseline BMD T-Score <-2.0. One point high-lighted is that there is a wide spectrum in the rate of bone mineral loss between patients and techniques of measurement, with figures as high as 8% per year reported. This is far in excess of a normal rate of bone loss amongst males of 0.5% per year.[7] Although validated nomograms exist for the general population combining DEXA findings with clinical parameters to predict long term fracture risks, no such tool exists for men rendered hypogonadal with the use of ADT.[8] Guidelines for men on ADT are empirical, and largely copy risk factors from the general population.[9] Beyond baseline BMD, the only clinical factor shown to have any accuracy in predicting bone loss for men on ADT is the change in serum P1NP (N-Terminal Pro-peptide of Type 1 Procollagen, a marker of bone formation).[10] One study showed that men in the highest tertile for P1NP after 6 months of ADT, had the greatest loss in BMD at 12 months. This finding has not been verified, and there remains a need to investigate the utility of other clinical parameters either at baseline or early in ADT therapy to find accurate predictors of which patients are at highest risk for accelerated BMD loss. Osteoporosis Imaging Currently, the only method to reliably determine which men are more rapid bone losers is to perform serial DEXA imaging. Thus, by the time that rapid bone loss occurs, it is too late to take measures to prevent it by interventions such as curtailing the duration of adjuvant ADT. Furthermore, we have level 2 evidence from a randomized clinical trial, that intervention with a bisphosphonate needs to be instigated at the commencement of ADT and continued throughout the duration of ADT to maximize bone density.[10] This study will aim to define a predictive tool combining baseline imaging and clinical characteristics to help determine which patients are at higher risk of accelerated bone loss prior to the initiation of ADT. Osteoporosis is a complex condition characterized by loss of both cortical and trabecular bone.[11] The structural basis of bone loss is poorly quantified by DEXA scanning which combines cortical and trabecular bone density in its measurement.[12] However, they can be separately and non-invasively quantified with the use of ultrasound (US), computerized tomography (CT), peripheral high resolution quantitative CT (pHR-QCT) or magnetic resonance imaging (MRI).[13] The last of these methods has the advantages of not being operator dependent, not requiring exposure to ionizing radiation and wide availability. A disadvantage is the relatively poor characterization of cortical and trabecular bone at a field strength of 1.5 T. There has been some work using CT imaging to separately quantify both cortical and trabecular BMD, as well as other parameters of trabecular bone quality. Much of this work has used pHR-QCT, which has revealed detailed changes in the porosity of cortical bone for men on ADT which is likely to weaken the bone, and as been termed 'trabecularization'.[14] Recent studies have compared this technique which has relatively limited accessibility, with more widely available technologies such as Quantitative CT (QCT) and Multidetector CT (MDCT).[15, 16] A very accurate correlation for Trabecular BMD was found between all 3 CT modalities. This raises the possibility that BMD can be estimated from the staging MDCT performed on all prostate cancer patients, without needing to expose them to the extra radiation dose required to perform a QCT. An advantage of MRI is that it also allows the collection of additional information regarding bone marrow (BM) including fat fraction and perfusions. These measures have previously shown some correlation with BMD measured by DEXA imaging, however the correlation is relatively poor, with a wide degree of unexplained variation.[17-19] BM has intimate proximity with trabecular bone, and paracrine factors such as the RANK-Ligand secreted from the BM plays a key role in recruiting bone resorbing osteoclasts.[20] It might therefore be that some of the variation in BMD measured with DEXA is due to baseline variation in BM quantity. There are also possible correlations between BM fat (BMF), and subcutaneous adipose tissue (SAT), visceral adipose tissue (VAT) and hepatic adipose tissue (HAT), all of which can be separately quantified by MRI.[21] This, is turn, may be linked with the deranged insulin levels and response linked with ADT administration, and posited as a cause of increased cardiovascular morbidity.[22] Other evidence shows that ADT induces a drop of haemoglobin from an average baseline value of 151 g/L down to 135 within 18 months of starting treatment.[23] No haemolytic process is evident, and the circumstantial evidence points to bone marrow suppression as being the mechanism for this. Such mild anaemia may also contribute to the insidious fatigue often seen in men treated with ADT. There is also some evidence from reanalysis of randomized trial data, that men who have the greatest drop in haemoglobin in the 3 months following initiation of ADT have a poorer overall survival in the setting of metastatic disease.[24] As such, measuring BM at baseline may help in predicting which patients are at risk of both losing bone faster, becoming anaemic, and suffering fatigue. It is therefore plausible that measurement of BM will add an important dimension to our knowledge of the bone as a functional unit as well as better explaining some of the toxicities associated with ADT.
2. Circulating Tumour Cells For a cancer to metastasize from the primary site of origin to other parts of the body, malignant cells must under a series of changes. One crucial step involves being able to use blood vessels to transport tumour cells around the body. Assays are now commercially available to measure these Circulating Tumour Cells (CTCs), including one which has FDA approval with the brand-name 'CellSearch'.[25, 26] This has superceded older approaches using reverse transcriptase polymerase chain reaction to detect CTC in men with prostate cancer.[27] Work over the last decade in patients with metastatic cancer has shown that the presence of CTCs in men with PC are a bad prognostic factor, with higher levels of CTCs correlating with poorer overall survival.[28] On the other side of the spectrum of tumour burden, work looking at patients undergoing a radical prostatectomy has shown only a very low incidence of CTCs (~20%) prior to surgery, which was no different to that measured in a cohort of healthy controls.[29] One issue with this study is that <5% of the patients involved would be predicted to eventually suffer metastatic failure, hence the chance of finding CTCs was likely to be very low based on the mainly low to intermediate risk patient cohort examined. Men with high risk PC have a much higher chance of eventual metastatic failure, of the order of 20-30%, or higher depending on their initial risk factors (PSA level, tumour stage and Gleason grade). At the time of diagnosis these men may therefore exhibit CTC levels intermediate between the metastatic and surgical cohorts previously considered. It may be that high risk PC patients with CTCs detected represent a very high risk group, and apart from providing important prognostic information for men, it could therefore warrant treatment intensification with increased duration of adjuvant ADT, or entry into clinical trials.
3. Prostate Radiotherapy Hypofractionation Radiotherapy (RT) has been shown to independently improve overall survival for men with high risk PC managed with ADT.[30] As such, standard of care for these men remains bimodality treatment with both RT and ADT.[1] RT has traditionally been given at doses of 1.8-2 Gy per day due to concerns about the potential for larger fraction sizes to cause late toxicity. Over the last 10 years multiple randomized controlled trials (RCTs) have shown that higher doses of RT (of the order of 74-80 Gy) lead to better rates of no biochemical evidence of disease (bNED).[31, 32] Due to the long natural history of PC, bNED is a validated surrogate endpoint looking at PSA control,[33] however the trial with the longest follow-up is now also beginning to show an improvement in Prostate Cancer Specific Survival (PCSS).[31] The use of such regimens leads to treatment durations of 8-10 weeks, which can be inconvenient for patients, consume a large proportion of the capacity of a RT department, and consequently be a significant factor in the existence of waiting lists for radiotherapy. There is strong data for PC suggesting that hypofractionation (that is, daily fraction sizes of >2 Gy) is particularly effective at maximizing tumour effect. Newer technologies such as image guided RT (IGRT) which ensures more accurate delivery of the RT, and intensity modulated RT (IMRT) which reduces unwanted radiation dose to adjacent normal structures are now in clinical use in Australia. They both have been used in phase 2 trials of Hypofractionated RT (HypoRT), with results for efficacy and late toxicity comparable to those reported in the literature for conventionally fractionated cohorts.[34, 35] There have been two small RCTs recently reported comparing HypoRT and conventionally fractionated populations, both showing no increased toxicity with the HypoRT, and better bNED.[36, 37] One of these focused mainly on high-risk men and included ADT, similar to the patient population eligible for PROCITT.[36]
4. Radiotherapy Volume When defining the RT treatment volume for a man with PC, traditional thinking has been to treat the prostate alone. However, for a local treatment modality such as RT or surgery, it is important to appreciate the natural patterns of spread of the disease. For instance, there are good consensus guidelines for patients with head and neck cancer to help radiation oncologists to know who are most likely to benefit from elective treatment of their cervical neck lymph nodes. This is because, despite the neck being negative at the time of diagnosis, surgical neck dissection series have helped to inform decision aids regarding the chance of a clinically normal neck harbouring sub-clinical disease. Nomograms have been constructed from large surgical PC cohorts to help define the risk of extracapsular extension, seminal vesicle involvement and lymph node involvement based on initial clinical parameters. Trying to treat all patients with the progressively larger treatment volumes required to include these areas would potentially increase toxicity without a high chance of improving efficacy. However, if a threshold risk level of 15-25% were required prior to including each elective target volume, we would aim to apply such treatments to patients most likely to benefit. Such concepts are already beginning to enter into consensus guidelines,[1, 38] and clearly represent a promising avenue of investigation. Of all of these expanded treatment volumes, only Whole Pelvic Radiotherapy (WPRT) has been investigated in men with PC in RCT.[39] Neither RCT found a significant benefit for the use of WPRT. However, many practice changing RCTs have used WPRT on all patients.[2, 40-42] One of the reasons for this discrepancy is likely to be that entry criteria for the largest WPRT RCT estimated a 15% risk of pelvic lymph node involvement.[39] Later work has shown that this only corresponded to a 2% pathological risk of nodal involvement. This emphasizes the need to use validated decision tools to select appropriate treatments.
5. Duration of neoadjuvant ADT

Often adjuvant ADT is given prior to commencing RT. This is known as neoadjuvant hormonal therapy (NHT). There is no clear guidance on what duration to give this for, although 3-6 months is a common approach. Results from an Australian randomized trial have shown 6 months of NAT to result in superior survival than 3 months.[43] Intuitively, it would seem that some patients would benefit from a shorter duration of NHT than others depending on their tumour response. There has been some preliminary work looking at an adaptive approach for this, where RT is started once a maximal PSA response has been achieved.[44] This has been shown to be feasible and effective in the phase 2 setting.