Associate Professor Nick Brook, University of Adelaide
Has anyone noticed a storm of controversy surrounding the diagnosis and treatment of prostate cancer? This centres on concerns about the overtreatment of prostate cancer, and may have particular relevance in older men, who mostly have a shorter life expectancy than their younger counterparts.
But men are living longer and in better health, perioperative management has advanced, and minimally invasive surgical treatments have lessened the acute physiological impact of treatment. Dogma has been that there should be a cut-off at 70 years of age when considering curative treatment for prostate cancer, but the tired catchphrase – ‘physiological not chronological age’ – is actually a very useful one, and has relevance to this topic.
It’s worth examining why we worry about treating localised prostate cancer with curative intent in older men. The reasons are four-fold:
• These men may not benefit from treatment, as they may die of other causes before their prostate cancer becomes clinically relevant
• Treatment may not be tolerated, and may cause morbidity and, rarely, mortality
• Most men with curable disease who are left untreated do not die from prostate cancer within 10 years of diagnosis.
• For those who die within 10 years of diagnosis, the disease was probably incurable at diagnosis
These last two points are taken from a recent presentation by Patrick Walsh, and reflect an understanding of the natural history of prostate cancer, and its heterogeneity.
With changes in demographics and treatments, should we be pushing for surgical treatment in older men with localised prostate cancer, or is this overtreatment? Are these men better off on surveillance/watchful waiting? Before we can answer these questions, some basic points need reviewing:
Prostate cancer is not one disease
First, prostate cancer as a disease is heterogeneous in its classification and behaviour. Gleason grading is absolutely central to determining how the cancer is likely to behave. We know that this classification trumps other variables in predicting outcomes, whether these are positive margin rates, extra-capsular extension, seminal vesicle invasion, lymph node status, recurrence after treatment, or prostate cancer mortality.
We can use the Albertson tables to indicate likely mortality from prostate cancer and non-prostate cancer causes over a period of time for a given Gleason score and age at diagnosis. It is really quite simple; when we ask “does it matter if we treat or not?”, we get an indication that a man in his early 60s with Gleason 6 disease has a very different proportional chance of dying from prostate cancer in the next 15 years than a man in his early 70s.
Likewise, a man in his 70s with Gleason 7 (a weakness of the tables is that 3+4 and 4+3 are combined) is, of course, proportionally more likely to die of another cause than he is to die of prostate cancer, but this is not true for a man in his 60s at diagnosis. So, age and Gleason score are combined in these tables to give us a reasonably powerful tool when we grapple with the question of whether to treat or not. These Albertson tables have been around for a long time, have recently been updated, and are greatly underused.
Active surveillance or watchful waiting?
Second, the terms ‘surveillance’ and ‘watchful waiting’ are separate entities that are often confused.
Watchful waiting is based on the premise that some patients will not benefit from treatment of their primary cancer. The decision is made at the outset to forgo definitive treatment, and instead provide palliative intervention for local progression or metastasis if/when it occurs.
Active surveillance is very different, and is based on the understanding that some but not all patients may benefit from localised treatment. The idea is to monitor closely and
1) identify those men with localised cancers that are likely to progress, providing timely treatment for them
2) to avoid treatment and associated treatment-related complications in men with cancers that are unlikely to progress.
If we consider again the Albertson tables, we can see why active surveillance makes sense for those cancers that are less likely to cause trouble, but also makes sense for older patients, perhaps with intermediate risk cancer.
A number of different active surveillance protocols are in use. They vary slightly (some have stricter criteria), and include the Johns Hopkins, Toronto, Miami, and UCSF protocols. The one we are encouraged to use in Australia as part of an international protocol study is PRIAS (Prostate Cancer Research International Active Surveillance).
Results of active surveillance
What do we know about outcomes from active surveillance? Klotz’s group in Toronto reported on 450 men followed with active surveillance, about 50% of whom were over 70 years at diagnosis, most with 3+3=6 but 17% of men had 3+4=7. Importantly 10-year cancer specific survival was 97%.
There was no difference in prostate cancer mortality for those men on AS over or under 70 years of age at diagnosis. Obviously though, non-prostate cancer death in those over 70 at diagnosis and commencement of AS was much greater than those under the age of 70.
This provides further evidence that age does matter; we already know our older patients are more likely to die of other causes. It also substantiates the idea that for carefully selected patients, AS is a sensible option.
What factors should we consider when choosing radical prostatectomy or ‘surveillance’ in older men?
When we are considering RRP in older men, there are three key questions we should consider:
1. In this man’s lifetime, will cancer control be an issue, i.e. do we need to perform radical prostatectomy to control his cancer or will surveillance (or watchful waiting) suffice?
2. Is the perioperative risk higher than is acceptable, and is it higher than in younger men? Does his age/co-morbidity preclude safe surgery?
3. Is his risk of long-term side effects that affect quality of life (incontinence and erectile dysfunction) too high, and is this higher than in younger men?
Let’s look at some evidence for these three areas:
The two randomised studies that we have were of watchful waiting (rather than surveillance) versus radical prostatectomy: The Scandinavian Prostate Cancer Group 4 Study randomised 695 men, 75 years old or less, with localised prostate cancer to radical prostatectomy or watchful waiting (not surveillance). The intervention for progression in the WW group was hormone ablation. Median follow-up was 12.8 years.
Briefly, compared to watchful waiting, radical prostatectomy reduced prostate cancer deaths in men under 65 years (51% RR reduction, p=0.008), but not in those over 65 years (17% RR reduction, NS). Likewise, occurrence of metastasis (itself, an important endpoint) was significantly reduced in the radical prostatectomy group in under 65s, but not in the over 65s.
This randomised study suggests that age does matter when considering the effect of RRP on cancer control.
The PIVOT study (Prostate Intervention Versus Observation Trial) looked at a similar number of men, randomised to RRP or observation (essentially watchful waiting, with palliative therapy or chemotherapy on progression). At 12 year follow up, there was no benefit of radical prostatectomy over observation, and there was no age effect.
This study has been heavily criticised as it was underpowered (the study was initially powered for 2000 men but only 731 were randomised), and for the very small number of prostate cancer deaths in each arm. There were far fewer deaths overall in PIVOT, and the men as a cohort had more co-morbid conditions than SPCG4. Difference in outcomes from the two studies may also be because the SPCG4 men were mostly PSA naïve, whereas the PIVOT cohort came from the early PSA testing era.
Perioperative risk in older men
Can we safely take older men through an operation and the perioperative period? Do the (generally) age-associated co-morbidities impart too much risk?
An excellent retrospective study from Ontario of 11,000 men who underwent RRP helps address this question. Importantly, it showed the following:
• Increasing age is associated with increased medical/surgical complications
• There is a small but significant increase in 30-day mortality with age, even when adjusted for comorbidity
• The number of co-morbidities is more important than age in determining mortality risk.
My reading of this is that age does matter, but medical fitness is more important is in determining post radical prostatectomy complications and death. Older men who are fit (with minimal co-morbidities) are low risk and can be considered for surgery.
Long term side effects that affect quality of life – continence and erectile function in older men
Lets conclude by looking at the potential long-term side effects of radical prostatectomy (incontinence and erectile dysfunction) that can have a major impact on quality of life. Is there any evidence for an age effect?
Many of the papers reporting side effects from radical prostatectomy are set about with bias, uncertainties and confusing definitions of continence and erectile function. However, a stand out paper from Massachusetts General Hospital looked at 430 men treated for localised prostate cancer with different modalities, and reported pre- and 36-month post-treatment sexual and continence function. Importantly, the authors stratified post-op outcomes according to pre-treatment function. Looking at erectile function, for nerve sparing radical prostatectomy (the gold-standard for erectile preservation), we see a remarkable reduction in sexual function. For those men with normal erectile function pre-op, only 8% were normal 36 months post-operatively. 28% of men deteriorated from normal to intermediate function, and 64% went from normal to poor. If we look at those who were intermediate before surgery, the figures post-op figures are worse.
The paper did not examine an age effect, but we can extrapolate from population studies of non-prostate cancer men. We know that erectile function deteriorates with age, and it is therefore likely that a cohort of older men will have worse pre-op function, and are therefore more likely to have worse post-surgery erectile function.
Admittedly there are weaknesses in this argument; the extrapolation and presumption, and the possibility that men with poor erectile function pre-operatively may not be concerned about post-operative erectile function, i.e. that factor may not affect their quality of life.
The paper reports a much less severe reduction in continence after surgery. However, but the outcomes are poorer for men who had less-than normal continence pre-operatively. Some recent data is available on the effect age on continence after robotic radical prostatectomy; this indicates that although early continence 3-6 months after surgery is slower to recover in men >70 years, after that time the results are equal to men <70 years.
This is a complex dilemma and a common clinical issue we struggle with regularly. As life expectancy increases, the likelihood of prostate cancer clinical progression increases. Less invasive surgery and improved perioperative management have expanded the pool of men that can be safely treated. We need to carefully examine the rationale for radical prostatectomy, watchful waiting and surveillance in this dynamic demographic of ‘older’ men, but we are becoming more sophisticated in our decision making in this area.
For cancer control, one randomised trial suggests that men over the age of 65 do not benefit from RRP compared to WW. In carefully selected men, true active surveillance seems safe, and it is likely to be particularly safe for men over 70. We may even be able to extend surveillance criteria in these older men.
In terms of surgical safety, age does matter but it is not as important as co-morbidity.
For those long-term side effects of radical prostatectomy that impact on quality of life, it is likely that older men have worse potency outcomes than younger men. Low-level evidence indicates that continence outcomes may be equivalent in older and younger men.