One popular way to rearrange statistics in cancer studies is to consider only survivors as those who have “completed treatment” and to not count the fatalities who, for obvious reasons, have failed to complete the trial.

After correcting for these inaccuracies, a large population-based study comparing overall and prostate-cancer survival in almost 60,000 men with prostate cancer treated by various methods concluded that previous studies “have generally overestimated the benefits of radical prostatectomy” (Lancet, 1997; 349: 906-10).

No controlled trials have successfully demonstrated that active intervention increases survival in men with prostate cancer. A 10-year randomised trial of men with early disease that compared radical prostatectomy with watchful waiting (monitoring disease progression without undergoing aggressive treatment) found that, although fewer men who had surgery died of prostate cancer, prostatectomy did not improve overall survival (N Engl J Med, 2002; 347: 781-9).

After more than 20 years, men who have had radical prostatectomy have survival rates that, on average, are the same as those with low-grade, low-stage, untreated prostate cancer (Scand J Urol Nephrol Suppl, 1995; 172: 65-72). In addition, if you are over 65 with low-grade prostate cancer and don’t opt for surgery, you have the same life expectancy as anyone else (JAMA, 1995; 274: 626-31).

Worryingly, surgery may not catch the cancer or even spread it (J Urol, 1996; 155: 238-42). Among Medicare patients in the US who underwent prostatectomy, 28 per cent reported needing follow-up treatment for cancer - such as radiation or androgen-deprivation therapy - four years after the operation (Urology, 1993; 42: 622-9).

Diagnosis and disease progression Each year, around 22,000 cases of prostate cancer are diagnosed in the UK (and 220,000 in the US), making prostate cancer the most common form of male cancer in the UK, and the second-most common (after skin cancer) in the US. But these statistics may simply be reflecting the fact that this disease is detected through a highly inaccurate screening blood test (see box).

Men with elevated prostate-specific antigen (PSA) levels are referred for biopsy - which can draw attention to very early prostate cancer in asymptomatic men - in the belief that early aggressive treatment is more beneficial than the evidence actually suggests (Am J Med, 1998; 104: 526-32). The majority of these men end up undergoing radical prostatectomy, although early disease does not necessarily warrant it (Prostate Cancer Prostatic Dis, 2005; e-pub ahead of print).

And even if you didn’t have advanced cancer in the first place, studies have found that taking a biopsy sample from an otherwise intact tumour can cause dissemination, or ‘seeding’, of cancer cells (Urologe A, 2005; 44: 64-7).

In most cases, prostate cancer is slow to grow, taking several years to become a detectable problem (and even longer to metastasise beyond the prostate). However, a small percentage of patients experience more rapidly growing, aggressive forms of this cancer. Biopsy can determine cancer aggressiveness through the so-called Gleason score. This is a number from one to 10 used to describe the more the tissue varies from normal. Nevertheless, there is no evidence to link the Gleason score to an increased chance of survival.

Types of surgery Radical prostatectomy can be carried out using various techniques such as the radical perineal approach (cutting through the perineum, the area lying between the testicles and anus), the radical retropubic approach (through the lower abdomen) or by laparoscopy (keyhole surgery).

The retropubic approach * is associated with greater loss of blood requiring transfusion (in one study all retropubic patients required transfusion compared with a little more than half in the perineal prostatectomy group (Br J Urol, 1994; 74: 626-9)

* has higher rates of urinary incontinence and urethral narrowing (J Urol, 1998; 160: 454-8; Br J Urol, 1994; 74: 626-9).

The perineal approach * results in higher rates of faecal incontinence (J Urol, 1998; 160: 454-8)

* is associated with fewer overall complications than the retropubic approach (J Urol, 2005; 173: 1863-70)

* is usually recommended for men with pubic arches that allow for easy removal of the prostate gland.

Nerve-sparing surgery * is an option if the cancer has not spread beyond the prostate and only a small amount of the surrounding healthy tissue is to be removed

* supposedly reduces the likelihood of erectile dysfunction (impotence) and urinary incontinence, but may be of limited benefit (JAMA, 2005; 293: 2648-53; J Natl Cancer Inst, 1997; 89: 1117-23).

Laparoscopic prostatectomy * employs a thin, tubelike instrument - or even a surgeon-controlled robotic arm - that allows the procedure to be done via tiny, minimally invasive, ‘keyhole’ incisions

* requires fewer days of bladder catheterisation

* results in less blood loss during surgery and possibly an earlier recovery of nocturnal continence (Urology, 2003; 62: 292-7)

* requires a skilled and experienced hand, as even small differences in technique can affect the outcome (J Urol, 2005; 173: 2099-103)

* has no proven advantage over open prostatectomy.

What doctors don’t tell you A large body of research shows that this get-in-there-early surgery can cause:

* erectile dysfunction in 60-89 per cent of men following radical prostatectomy (Urology, 1993; 42: 622-9). With nerve-sparing surgery, 60-85 per cent of men recover erectile function, but it could be up to two years later (JAMA, 2005; 293: 2648-53)

* persistent urinary incontinence. In one study five years after surgery, 30 per cent still had to wear pads or clamps, and more than 40 per cent reported occasional incontinence (Urology, 1993; 42: 622-9).

* deep vein thrombosis or fatal pulmonary embolism weeks after surgery (J Urol, 1997; 158: 6)

8 further surgery. In the US Medicare study, 6 per cent of radical prostatectomy patients needed additional surgery for incontinence, 20 per cent needed postoperative treatment to combat urethral narrowing and 15 per cent sought help for sexual dysfunction (Urology, 1993; 42: 622-9). With laparoscopy, nearly 4 per cent had to correct postsurgical complications (J Urol, 2002; 167: 51-6).

Kim Wallace