“The news is [not] all good”: misrepresentations and inaccuracies in Australian news media reports on prostate cancer screening

In 2003, 4733 male deaths were attributable to lung cancer and 2761 deaths to prostate cancer.10 In the context of an item on prostate cancer, the statement that every man will at some stage have a prostate “problem” may have reasonably been taken to mean that every man will have a prostate cancer problem at some stage, rather than (for example) benign prostatic hyperplasia.

The Urological Society of Australia and New Zealand’s policy states that “Population screening of asymptomatic men is not recommended”.11 The second media comment above implies a conflation of the importance of early detection via screening of breast, cervical and colon cancer with prostate cancer screening. This does not reflect Australian Government policy. Not once throughout the study period did any report mention that nearly all national and international cancer control and expert health technology assessment bodies do not support prostate population screening.

Claims that the PSA test is more accurate for prostate cancer than mammography is for breast cancer are inconsistent with the evidence. For example, in the Prostate Cancer Prevention Trial, the sensitivity (capacity to detect cancer if present) of PSA at a cut-off level of 4.1 ng/mL was 40% for cancer with Gleason grade ≥ 7. In other words, the test missed 60% of moderate- to high-grade cancers.12 At the same threshold (4.1 ng/mL), the false positive rate was about 10%, meaning that 10% of men with no prostate cancer received an abnormal PSA result. This compares with a sensitivity of 75%–95% for screening mammography,13 and a false positive rate of 2%–7%.13,14 Because the sensitivity of the PSA test is poor, clinicians may use an even lower threshold such as 2.5 ng/mL in calling a test result abnormal. In the Prostate Cancer Prevention Trial, using 2.6 ng/mL as the threshold for an abnormal result increased the sensitivity to 67%, but the false positive rate increased to 23%.

The assertions that PSA testing detects significant disease and that most men with PSA-detected prostate cancer will die of their cancer are totally indefensible on the evidence. Among men with clinically detected, well differentiated prostate cancer, long-term survival is excellent, with little or no reduction in life expectancy compared with men without prostate cancer.15 As screen-detected cancers are, on average, less aggressive than clinically detected cancers,16,17 it is misleading and alarmist to imply that most PSA screen-detected cancer will be life threatening. Indeed the reverse is true: most PSA screen-detected cancer is not life threatening. Data from the European Randomized Study of Screening for Prostate Cancer suggest that, with yearly or even 4-yearly PSA screening, 48%–56% of the cancers found would never have been diagnosed in the absence of screening.18 The over-detection of prostate cancer (ie, screening detection of low-grade cancers not destined to affect health or life expectancy) is one reason why prostate cancer rates increase so steeply after the introduction of PSA screening.19-22

Over-diagnosis tends to lead to over-treatment, and the harms caused by treatments are important (see below).

Claims attributing observed declines in prostate cancer mortality to PSA screening are based on observational studies, which provide unreliable evidence for addressing this question. Declines in prostate cancer mortality in the United States have been attributed to PSA screening, but also to improved treatments available for advanced prostate cancers.23 Similar falls have been observed in England and Wales, where PSA testing is less common and “has been discouraged”.24 In Canada, one study demonstrated that the decline in mortality is unlikely to be due to PSA testing, despite claims to the contrary.25 The efficacy of PSA screening will not be resolved until the PSA trials currently in progress report their results.

There are some studies (randomised trials and observational studies) comparing different management options for prostate cancer (eg, comparing surgery [radical prostatectomy] and watchful waiting26). However, the applicability of this study to the question of the efficacy of PSA screening is limited, as it recruited subjects who predominantly had clinically detected, rather than screen-detected, cancers.26 As noted, many screen-detected cancers will not threaten survival in the same way as clinically detected cancers, and therefore data from such studies may not be applicable to screen-detected cancers. Treatment trials in men with screen-detected cancers are needed.

Considerable evidence contrasts markedly with these optimistic claims made by screening advocates. A follow-up study of men undergoing radical prostatectomy reported that, after 5 years, 14%–16% of men had urinary incontinence, and 79% had erections insufficient for intercourse.27 A review undertaken by the US Preventive Services Taskforce (USPST) found 15%–50% of men experienced long-lasting urinary problems after prostatectomy, compared with 2%–16% after radiation therapy.15 The same review found that 20%–70% and 20%–45% of men experience reduced erectile function after prostatectomy and radiation therapy, respectively, and 6%–25% experience bowel problems after radiation therapy.15 Adjuvant therapies such as androgen deprivation therapy also have adverse effects: 5%–25% experience breast swelling and 50%–60% experience hot flushes. All these measures were recorded at least 12 months after treatment.15

Significantly, the USPST review noted that current evidence on whether newer nerve-sparing surgical procedures reduce rates of adverse effects was mixed “outside of excellent academic centers”,15 a point particularly relevant to the debate in Australia, where the purported benefits of robotically assisted prostate surgery using the da Vinci surgical system have featured in four television items. On 60 Minutes, an extended segment was introduced with:

However, only three Australian facilities have the equipment, and only 11 surgeons are trained in its use.28 Although advantages reportedly exist in terms of decreased patient pain and recovery times, long-term oncological benefits are unclear, nor is there clear evidence of appreciable reductions in the incidence of incontinence or erectile dysfunction.29,30

Given that the da Vinci system has only been in use since the late 1990s, the absence of long-term results is predictable, but this “lack of current clear evidence that there is true benefit associated with this technology”31 has not stopped Australian proponents from highlighting its efficacy. “[A]ggressive marketing and commercial interests”31 are key factors in promoting this costly technology.

Men considering these claims should be aware that optimistic claims about adverse effects after prostate surgery are commonly made by urologists. However, when ratings of urinary symptoms made by patients and by their urologists are compared, significantly more adverse ratings are made by patients.32

Finally, the adverse psychological effects of receiving a cancer diagnosis, and then having to make difficult choices between treatments, all of which have substantial risks of adverse effects, should not be overlooked. Men who choose watchful waiting over any of these treatment options have to live with their cancer diagnosis. Data from the trial of prostatectomy versus watchful waiting suggest that the quality-of-life effect of a cancer diagnosis followed by watchful waiting or surgery is substantial.33