Abstract
Cancer of the prostate is the most common cancer and the immediate cause of death in 5–6% of men in the Nordic countries. An additional group of men of similar proportion die with prostate cancer but from another cause. This obviously is a public health issue.
Prostate-specific antigen (PSA) is a protein produced by the cells of the prostate gland. The level of PSA in blood indicates the state of the gland. It was soon after its discovery that physicians realized that an increased level of this protein in the blood could be used as a marker for prostate cancer otherwise not noticeable. It seemed that PSA could serve as an ideal screening tool for early detection of prostate cancer.
However, the problem at this time in the mid-1980s was the lack of treatment with reasonable outcome and limited side effects. The balance between good and harm was not favorable enough to warrant population screening at this time. In addition, it soon turned out that the specificity and sensitivity of the PSA test were poorly defined with a great overlap.
Today – 30 years later – there are still no recommendations for mass screening from the health authorities on both sides of the Atlantic. On the contrary, hesitation and downright negative attitudes are the rule due to the weaknesses of the PSA test, the side effects of treatment and the highly variable natural history of the cancer [1,2].
Before discussing screening further, we need to draw attention to the difference between mass screening based on active invitation from a health authority, as opposed to the demand from persons in the population without signs and symptoms to have a screening test. In the first situation, the WHO rules need to be in place with a safe and validated screening test without side effects, a treatment that leads to improved outcome compared to treatment at the time of symptoms and the willingness of the society to finance the program. When screening is performed at the demand from individuals, there is a possibility to discuss all these factors individually, which is not the case in population screening programs.
Prior to the widespread use of PSA, most men presented with advanced disease. They sought medical care because of their symptoms due to the advanced cancer. Essentially, they could not be ‘cured’. Palliative care with radiation, hormone ablation, or both, was the usual option. However, since the introduction of PSA, only few men present this way and these men with advanced disease have not been in any sort of voluntary or population-based screening programs.
The discussion of sensitivity and specificity of PSA is not always as systematic as it should be. The fact is that a single figure indicating a very low level is usually compatible with a low risk of developing cancer. However, one single PSA that is ‘normal’ or elevated above the very low level must be evaluated in its context: is this an initial PSA or one in a longitudinal series (PSA velocity)? What is the PSA relative to the size of the prostate (PSA density)? How old is the patient (PSA goes up with age)? What is the ratio of free to total PSA? Is there a family history of prostate cancer and, if so, what were the specifics about that family member’s cancer?
It is obvious that the PSA test has a potential; why is it not possible to leverage what we know about PSA in large screening studies that are well designed and make sense? There are some possible explanations. The overall survival of a group in the population can be assessed in prospective, long-term studies, but may not be the most wanted aim. Some of these patients are elderly and enjoy a good quality of life. Living with a cancer that is untreated yet under control may be equally important. In the European study the rate of deaths due to cancer was significantly reduced, but all-cause mortality was not improved [2]. In the American PLOC study, no such benefit was found in the group overall; however, when stratified for age, men under 70 did show a survival benefit from screening [1]. An important warning is that – as expected – more cancers were found in the screened group with a corresponding increase in complications related to treatment.
Another point to discuss is the age in the target group. In the large studies, men were invited and followed in the ages from 55 to 74 years. It is possible that much more could be gained by inviting men in the ages below 55 or even younger. Like with breast cancer screening, the target group will be diluted with fewer cases per group invited, but the gain in life years could be larger.
Urologists who have to care for the cancer patients have generally over the years had more positive attitudes to the possible value of PSA than epidemiologists. It remains to be seen whether the use of PSA can be refined or if any other marker for this the most common cancer will be found.