Tina Kopsaftis

Comparative analysis of three risk assessment tools in Australian patients with prostate cancer

Whats known on the subject? and What does the study add?

Prostate-specific antigen (PSA) rate of decline post external beam radiotherapy predicts prostate cancer death

BACKGROUND AND PURPOSE To assess the association between PSA velocity (PSAV) in the first 24 months after external beam radiotherapy (EBRT) and prostate cancer-specific mortality (PCSM) and all cause mortality. MATERIALS AND METHODS All eligible patients in the South Australian (SA) Prostate Cancer Clinical Outcomes registry were followed. 848 Patients treated by definitive EBRT with more than one PSA recorded in the two year post-treatment were included. We calculated PSAV by linear regression. RESULTS The mean number of PSA measurements in the 2year period was 4.4 (SD1.9). The median PSAVs across quartiles (Q1-Q4) were -4.17, -1.29, -0.38 and 0.20ng/ml/yr. In multivariable analysis, a U-shaped relationship was seen between PSAV and PCSM with Q1-Q4 hazard ratios (HR) being 3.82 (1.46-10.00), 3.07 (1.10-8.58), 1, 5.15 (1.99-13.30) respectively. HR for all cause mortality in a similar model were 1.79 (1.07-2.98), 1.55 (0.93-2.59), 1.00 and 1.74 (1.04-2.90) for Q1 to Q4 respectively. A rapid PSA decline in the first year was a strong predictor of PCSM. However, in the second year PSA increase was positively associated with PCSM. CONCLUSION A rapid decline in PSA in the first year following EBRT is positively associated with PCSM. This may be a useful early indicator of the need for additional therapies.

Clinical and socio‐demographic profile of an Australian multi‐institutional prostate cancer cohort

Aims:  To describe the clinical and socio‐demographic data from a South Australian prostate cancer cohort (PCCOD).

Localised prostate cancer in elderly men aged 80-89 years, findings from a population-based registry

To investigate the rate of prostate cancer‐specific mortality (PCSM) and disease characteristics in patients diagnosed with localised prostate cancer at age 80–89 years in comparison with men diagnosed at age 70–79 years.

Prostate cancer outcomes and delays in care.

ObjectivesTo examine the survival effect of treatment delays from the time of confirmed diagnosis of prostate cancer to first treatment in an Australian population.MethodsThree thousand one hundred and forty patients were identified from the South Australian Prostate Cancer Clinical Outcomes Collaborative database for analysis. Selected patients had dates recorded for both diagnosis and treatment. We examined the effect of treatment delay (the time from diagnosis to date of first treatment) on survival using Cox and competing risks regression and compared quartiles of delay across the cohort. Adjustment was made for age, PSA levels, treatment modality and Gleason score. Outcomes included overall survival (OS) and prostate cancer-specific mortality (PCSM).ResultsQuartiles of delay were as follows (days)—Q1: 35, Q2: 86, Q3: 138.0, Q4: 264. Shorter delays were associated with hormonal treatment, high Gleason score and high PSA values. Measuring PCSM with Q2 as reference, age-adjusted associations were—Q1: sHR 4.37 (2.75–6.94), Q3: sHR 1.29 (0.73–2.28), Q4: sHR 1.55 (0.91–2.63). After additional adjustment for treatment type, Gleason score and PSA, Q1 remained at increased risk [sHR 2.46 (1.10–5.54)]. A similar trend was observed for OS. In analysis stratified by Gleason score, delays were not significantly associated with OS.ConclusionsFactors associated with shorter delay in treatment include high Gleason score, high PSA and hormonal treatment. After adjustment for these variables, increased delays were not associated with OS or PCSM in this cohort. The nonlinear association of delay with risk may explain conflicting reports in the literature.

Prostate cancer mortality is high in the elderly and can be reduced by selective individualized curative treatment

We read with interest, Bandini et al’s paper which, using a SEER-based cohort, reports radical prostatectomy (RP) or radiotherapy (RT) reduces prostate cancer (PCa) mortality in the elderly, and concludes curative therapy for localized prostate cancer should not be denied on the basis of chronological age alone [1]. We inform on our real-world data which supports these results [2]. The South Australian Prostate Cancer Clinical Outcomes Collaborative (SAPCCOC) is the longest running prostate cancer registry in the Southern Hemisphere. It commenced in 1998 and is a prospective, third-party collected-, longitudinal, observational registry which follows participants until death or withdrawal. We examined men from our registry between 2005 and 2014, aged 70–89 without evidence of metastatic disease at presentation, identifying 1888 participants and dividing them into two cohorts based on age at diagnosis. Cohort 1(“Co1”)—1428(76%) were aged 70–79, and cohort 2(“Co2”)—460(24%) aged 80–89. Differences associated with age were examined by 5-year groupings. We compared patient and disease demographics, treatments and survival outcomes. Curative intent included RP, external beam radiotherapy and brachytherapy. Non-curative therapies included hormonal treatments and observation. Pair-wise interaction with age, Charlson co-morbidity scores and year of diagnosis were included as covariates. Additionally, a propensity score-matched conditional binomial logistic regression was employed. Median follow-up was similar at 7.3 and 7.0 years. Causes of death were obtained based on mandatory reporting to the state Births, Deaths and Marriages registry. Co2 had higher Gleason scores, and pre-treatment PSA values (p < 0.001), but not T-stage (p = 0.17). The odds of receiving curative treatment in Co2 was lower than in Co1 (OR = 0.12; 95% CI = 0.09, 0.16; p < 0.001). Curative intent treatments (RP and RT) were used more frequently in Co1 (62%) compared with the older group (18%). There was a higher rate of RP in the younger group (18 vs 0%). Similarly, the rate of radiotherapy was higher in Co1 (44 vs 18%). In contrast, primary hormonal therapy (21 vs 4%) and watchful waiting (22 vs 16%) were more frequent in Co2. At the time of analysis, there were 310 deaths (22%) in Co1 of which 97 (31%) were due to PCa. In Co2 there were 263 deaths (57%) of which 73 (28%), were attributable to PCa. The younger group (Co1) had had better overall survival, with 87% 5-yearand 68% 10-year survival, vs 55% 5-year and 26% 10-year survival in older, which is not unexpected in view of the differential Charlson scores (Co1-mean score 1.3; Co2-mean score 1.9, p < 0001), however, the proportion of deaths due to prostate cancer was the same in each group (31 vs 28%). Older patients were at * Kim Moretti kim@theurologist.net.au

Delays in radical prostatectomy for prostate cancer and survival outcomes.

A recent publication, involving 619 Canadian patients undergoing radical prostatectomy (RP), suggests that increased delays to time of surgery may be associated with higher rates of biochemical recurrence (BCR) [1] in men with high-risk disease. However, BCR per se is not strongly correlated to the “hard” outcomes of developing metastases or increase cancer mortality that are more heavily influenced by Gleason score (GS) and pathological stage. The manuscript concludes “...further studies are warranted to assess the impact of SWT (surgical wait time) on cancer-specific survival and overall survival (OS)” and additional retrospective studies with larger cohorts were requested. In the state of South Australia, we have the longest running prostate cancer registry in the Southern Hemisphere, and have performed such an analysis, in a larger cohort, with a greater number of events, in respect of prostate cancer outcomes and delays in care [2], and draw your readers’ attention to these findings. Zanty and colleagues found that, on multivariate analysis, SWR did not affect BCR, except in high-risk disease, where an SWT cut point of 90 days was recommended. We identified 3140 eligible patients between 1998 and 2013, included all treatment types (RP, radiotherapy [RT], androgen deprivation [ADT]), local and advanced disease, and examined the effect of treatment delay (time from diagnosis to date of first treatment) on prostate cancer-specific mortality (PCSM) and OS, using Cox Proportional Hazards modelling and competing risks regression, comparing quartiles of delay across the cohort. Using quartiles avoids the introduction of arbitrary cut-off values. Adjustment was made for age, PSA levels, treatment modality, and GS. Quartiles of delay were as follows (days)—Q1:35, Q2:86, Q3:138.0, and Q4:264. Quartile two was used as the reference point and extensive subgroup analyses conducted. Shorter delays were associated with hormonal treatment, high GS, and high PSA values. PCSM, in an age-adjusted model, was highest in Q1 (HR-4.37) compared to Q3 (HR1.29) and Q4 (HR-1.55). After additional adjustment for GS, PSA value, and treatment type, Q1 still remained at highest risk (HR-2.46). The fact that shorter delays were associated with poorer outcomes is possibly explained, because patients with high PSA levels and high GS at biopsy are more likely to receive treatment sooner, given they represented more aggressive disease. When examining only patients treated with curative intent (RP or RT), delays in Q3 were associated with increased risk of PCSM (HR-2.07), though on sub analysis of RP alone and RT alone, no quartile showed increased risk. On examining the effect of NCCN risk profile, only those in “high-risk” group appeared to be at increased risk of poor outcomes, particularly in the context of a short time to treatment. However, this most likely reflects the underlying aggressive biology of the disease which led to these patients being treated more quickly in the first place. A similar trend was observed for OS; however, after stratification by GS, no association between delay and OS could be demonstrated, and within each GS stratum, delays in care were not associated with variation in overall survival. This comment refers to the article available at https ://doi. org/10.1007/s0034 5-017-2105-6.

Erratum: Stromal androgen receptor regulates the composition of the microenvironment to influence prostate cancer outcome [Oncotarget. 2015; 6:16135-16150]

Erratum: Stromal androgen receptor regulates the composition of the microenvironment to influence prostate cancer outcome Damien A. Leach, Eleanor F. Need, Roxanne Toivanen, Andrew P. Trotta, Helen M. Palethorpe, David J. Tamblyn, Tina Kopsaftis, Georgina M. England, Eric Smith, Paul A. Drew, Carole B. Pinnock, Peng Lee, Jeff Holst, Gail P. Risbridger, Samarth Chopra, Donald B. DeFranco, Renea A. Taylor and Grant Buchanan Oncotarget. 2015; 6:16135-16150 PMID: 25965833

793PPROSTATE CANCER (PCA) IN ELDERLY MEN – FINDINGS FROM SOUTH AUSTRALIAN PROSTATE CANCER CLINICAL OUTCOME COLLABORATIVE (SA-PCCOC)

ABSTRACT Aim: PCa is among the most common cancers. It is the cause of death in half of cases diagnosed. Data on elderly men, especially ≥80, is limited. We describe the patient and disease characteristics and the primary cause of death (PCa versus non-PCa) in patients (pts) diagnosed when aged ≥ 80 and compare them with pts diagnosed at the age of 70-79 using data from a large clinical registry. Methods: Data for elderly men (≥ 70 yrs) with PCa were obtained from SA-PCCOC (a longitudinal, observational registry of biopsy-proven PCa cases, throughout the Australian state of South Australia since 1998) including cause of death. Proportions were compared using a Chi squared test. A multivariable model used logistic regression to assess the variables: age at diagnosis, clinical stage and Gleason score at biopsy. Results: There were 601 pts in the ≥ 80 group (median age at diagnosis 84, range 80-101) and 1232 pts aged 70-79 (median age at diagnosis 74, range 70-79). At data cut-off, 431(72%) had died in the ≥ 80 group; 187 (43%; 95% CI 39-48%) caused by PCa. In the 70-79 group, 290 (24%), had died with PCa as the primary cause in 142 (49%; 95% CI 43-55%). Pts were more likely to die if diagnosed with PCa when over 80 compared to those aged 70-79 at diagnosis [OR 8.24; 95% CI 6.56-10.33]. We did not observe a statistically significant difference in the distribution of cause of death between the two groups (PCa versus non-PCa) [OR 0.80; 95% CI 0.59-1.09]. In pts aged ≥80, a Gleason score >7 at diagnosis was associated with death caused by PCa (p = 0.01); however this was not significant in multivariable analysis [OR 1.2; 95% CI 0.62-2.33]. In pts aged 70-79, death due to PCa was associated with clinical stage 3 [OR 3.56; 95% CI 1.19-11.44, p = 0.03], and Gleason score >7 [OR 3.01; 95% CI 1.22-7.66, p = 0.02] in multivariable analysis. Conclusions: A large proportion of elderly men with PCa are likely to die from the disease. The proportion of elderly men (≥80 years) dying from PCa is similar to those aged 70-79. To our knowledge this is the first study focusing on the population of pts older than 80 with PCa. Disclosure: All authors have declared no conflicts of interest.