Identifying high risk patients

Dr Elizabeth Rapley updates us on the latest research into prostate cancer from the Institute of Cancer Research

A search is underway to find biological markers that can predict prostate cancer aggressiveness. The Institute of Cancer Research (ICR) is investigating a number of promising biomarkers and have identified that a combination of three genetic abnormalities significantly impacts on how long a prostate cancer patient is likely to survive with the disease¹.

Prostate cancer affects 35,000 men per year in the UK. It is the most common cancer diagnosed in men and the second leading cause of cancer death (after lung cancer) in British men, with more than 10,000 deaths per year².

Clinically the behaviour of prostate cancer is highly variable and there is a real urgency to identify biomarkers that can accurately and reliably classify patients into clinically relevant subtypes. Critically there is a need to differentiate between those patients who have the aggressive disease and require intensive treatment and those who require less treatment or surveillance alone. Recurrent gene alterations within tumours can often provide key biomarkers that can be exploited for diagnosis, disease sub-classification, prognosis and therapy.

Deletion of all or part of the PTEN tumour suppressor gene on chromosome 10p23 is a frequent event that occurs in around 70% to 80% of prostate cancers³. More recently, fusions between the androgen-regulated transmembrane protease serine 2 gene, TMPRSS2, and E twenty-six (ETS) transcription factors were discovered in prostate cancer^3-4. Loss of the PTEN gene and rearrangements of the ETS gene family are considered to be important and common molecular events in prostate cancer. Furthermore, mouse models have suggested that these genes cooperate to promote carcinogenesis^5-6. However the combined impact of these abnormalities on survival in a large group of patients has not previously been examined.

In this study, tissue microarrays were constructed from unselected transurethral resection of prostate specimens. The patients received no initial treatment and were conservatively managed. Fluorescent in situ hybridisation (FISH) was used to detect rearrangements at the ERG and ETV1 loci and PTEN loss.

FISH results for both PTEN loss and ERG and ETV1 gene status was available for analysis from 308 patients. The primary endpoint of the study was death from prostate cancer or death from any other cause. Median follow up was 100 months (range 3-197 months). After 11 years of follow up, 25% of men had died from prostate cancer, 34% from other causes and only 22% were alive without disease progression.

Patients were stratified to those that had an ERG/ETV1 rearrangement (122 patients, 40%) and those that did not (186 patients, 60%). The patients were then further stratified into whether they had a PTEN loss or not. The data showed that PTEN loss alone and ERG/ETV rearrangement alone was not a significant predictor of clinical outcome; only when combined was the effect significant.

Nineteen patients (6.2%) had PTEN loss and no ERG/ETV rearrangement and defined a poor prognosis group of prostate cancers. This group had a significantly worse cause specific and overall survival at 13.7%. Those patients with ERG/ETV rearrangement and loss of PTEN (37 patients) or no loss of PTEN (85 patients) identified an intermediate prognosis group with 11 year survival at 41.0% and 59.8% respectively. The largest group of patients (n = 167, 54%) who had neither PTEN loss nor ERG/ETV1 gene rearrangements, were shown to have a good prognosis with a cause specific survival of 85.5% at 11 years.

Importantly, the study also highlighted that traditional indicators – such as the grade of the tumour – did not accurately predict cancer survival. Gleason grading of tumours is used to indicate which prostate cancers may need further treatment and a Gleeson score of > 7 in combination with other clinicopathological indicators usually prompts more aggressive treatment for a patient. However, in the poor prognosis group as defined by the molecular markers, 21% of patients had a Gleeson grade of seven or less. Moreover in the good prognosis group, prostate specific deaths occurred across all the Gleeson grades. The study underlined the importance of identifying the high risk group, based on the underlying genetic abnormalities, in a cohort of patients that would otherwise receive conservative management when, they should be receiving more intensive therapy for their prostate cancer.  

The data suggests that molecular characterisation of PTEN, ERG and ETV1 gene status might be used to determine risk of prostate cancer death, potentially helping to decide which patients should be managed conservatively and which should be treated aggressively.  

The research also has implications for ongoing and future clinical trials where patients are stratified by clinicopathological features alone. These results suggest that an imbalance of patients with different PTEN and ERG/ETV1 rearrangements could influence the trial outcome and should be taken into account.

The study further suggests that there may be other important molecular abnormalities to be identified in prostate cancer and these may account for the observation of poorer survival in patients with non-ERG/ETV1 gene rearranged/PTEN loss tumours.

The ICR is continuing its investigations into a number of related genetic alternations thought to play a role in prostate cancer behaviour. This includes clinical trials testing the use of 2+Edel – a duplication of the fusion of the TMPRSS2 and ERG genes – as a marker. Prostate cancers commonly contain one copy of this fusion gene, but a duplication is found in just 6.6% of prostate cancers (equivalent to about 1,800 UK cases each year). Patients with 2+Edel have only a 25% survival rate after eight years, compared to 90% for patients with no alterations in this region of DNA⁷. Furthermore, the use of ERG as a marker to predict patient response to the new prostate cancer drug abiraterone – discovered at the ICR and now licensed to Cougar Biotechnology Inc (now owned by Johnson & Johnson) – is also being examined. A Phase I/II study found the majority of patients whose tumours shrank significantly on this drug had an abnormality in the ERG gene⁸.

The ICR scientists would now like take to the ERG, ETV1 and PTEN markers forward into clinical trial to confirm whether their use in treatment selection improves patient outcomes.