by Ashley Hay
* * * * *
People like me who are against PSA screening are sometimes accused of being callous and indifferent to the suffering of those who have prostate cancer. In 2003, a very public stoush between Alan Coates, who was president of Cancer Council Australia, and Wayne Swan, who was Labor’s shadow community services minister, played out on national television. Because his father had died of prostate cancer at the age of 67, Swan underwent screening when he was 47. He was diagnosed with cancer and elected to undergo a radical prostatectomy. During an interview on ABC TV’s 7.30 Report, he said:
I believe that screening did save my life. My father had died an extremely painful death from prostate cancer. I think it is absolutely critical that men who are particularly at risk, if there’s a history in their family, that they should be tested. I guess the point is that early detection is the best protection.
Coates disagreed:
I decided personally that I will not be tested and I’ve held that decision for over ten years … If there was a really good effective screening test, then we mightn’t be having this discussion because it would be like it is for women with breast cancer or for people with bowel cancer. We know those screening modalities work. We know they save lives. We just don’t know that for screening for prostate cancer.
A Sydney urologist, Paul Cozzi, weighed in with his reservations about Coates’s position:
I don’t believe it’s appropriate for Professor Coates to give a personal opinion in this media under the auspices of the Cancer Council of Australia. I believe it’s up to the individual patient to discuss with the GP the pros and cons of PSA testing prior to proceeding with testing … I personally have had several patients that have been very confused by the recent information in the media as to whether they should proceed, even with the prostate biopsy. There is more confusion of course with regard to treatment but certainly these recent events in the media have not helped the general practitioners, the urologists or the patients …
Swan was even stronger in his condemnation:
I’m absolutely appalled by what he has said. This man is the head of the peak cancer organisation in this country. People would regard him as being highly educated and highly knowledgeable. I believe that the statements that he has made run against the weight of medical opinion …
The debate went on. Cozzi mentioned a major European study on prostate cancer screening that was due to release its results later that year. ‘I think when the data from this latest trial becomes available,’ he wrote, ‘we will hopefully have a lot more scientific evidence as to the benefit of screening ...’ Coates also appealed to the evidence: ‘If putting a fair presentation of the pros and cons causes confusion, then I’m sorry, but the evidence isn’t sufficient to remove that confusion, and I don’t think anybody is well-served by pretending that it’s better than it is.’
The personal experiences of the three men obviously coloured their views. Swan had had cancer, was operated on, and was alive. The urologist had operated on many men and they were still alive. Coates had decided not to be screened and he, in his 60s at the time, was still very much alive, too.
While the three men were debating the issue, two large randomised studies were under way – the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial, or PLCO, in the United States, and the European Randomized Study of Screening for Prostate Cancer, or ERSPC. The PLCO enrolled 77 000 men over an 11-year period and the ERSPC included nearly 182 000 men; in each case, half the men were allocated to prostate cancer screening and the other half to ‘usual’ care. Both studies published their results in the world’s premier medical weekly, the New England Journal of Medicine, in 2009. The ERSPC trialists concluded that screening saves lives; the PLCO group found no benefit in screening. Because PSA has been used much more widely in the United States than in Europe, those in favour of screening believe that the negative findings of the PLCO were invalid. A further two years of follow-up has not changed the conclusions of either study.
What is the layperson to make of two mega-studies that produce different results? And how different are they anyway?
The men in the screening group of the ERSPC trial were 21 per cent less likely to die of prostate cancer than were the men in the non-screening arm. Sounds like an important result, doesn’t it? I teach my first-year medical students to ask a simple question whenever they see a ‘relative’ risk reported like this: relative to what? To answer, you have to look at the absolute risk of cancer death in both groups. When you frame it this way, the significance of the findings dramatically changes. Why? Because the absolute risk of prostate cancer death in both groups was actually quite small – about 0.4 per cent in the screened group and about 0.5 per cent in the control group. Translated into absolute numbers, rather than ratios, this means that 1055 men have to be screened and 37 cases of prostate cancer have to be treated to prevent one prostate cancer-related death.
If we apply the ERSPC figures to the Australian setting, about 5965 of the 6130 men (or 97 per cent) who underwent a radical prostatectomy in 2012 derived no survival benefit from the operation. And if that is not enough, the ERSPC and the PLCO both found no difference in the all-cause death rate between the men who were screened and those who weren’t. In other words, screening might prevent a small number of cancer-related deaths (not a trivial result for the men concerned, of course) but it doesn’t save lives overall. This is the wide-angle, public health view. While it is usually a good thing for a doctor to look at you closer up, if that focus is applied to broader population-based programs it can cause more harm than good.
The Medicare data suggest that the publication of the results of the PLCO and ERSPC trials may have changed the practice of Australian medical practitioners to a small degree. The number of radical prostatectomies fell by a few hundred, from 6470 in 2009 to 6130 in 2012. (It is very difficult to get accurate numbers of patients treated with radiotherapy, so the fall in surgical procedures may have been replaced by an increase in radiotherapy.) But TRUS biopsies were also down by 15 per cent in the same period, from 34 280 to 29 029. General practitioners are the main drivers of PSA testing so it is plausible that the evidence has altered the practice of some of them. But it is hard to change the minds of those with personal experience of the disease.
* * * * *
The US Preventive Services Task Force is the peak American independent expert panel dealing with a variety of public health issues. For ten years, it had recommended against prostate cancer screening on the grounds that there was no high-level evidence in support of it. When the negative results of the American PLCO study and the absence of an all-cause mortality benefit in the European ERSPC trial were published, the issue was settled as far as the panel was concerned. The US Centers for Disease Control and other federal agencies followed suit, and now recommend against prostate cancer screening in any form.
The American Cancer Society, the American College of Physicians and the American Urology Association still believe that there is a place for prostate cancer screening, however. In an important, if partial, concession, the American Urology Association changed their guidelines in 2013 to recommend against screening for men under the age of 40 in any circumstances, and it doesn’t recommend screening for men 40–54 years of age at average risk (without a family history, that is). But it still recommends ‘shared decision-making’ for men aged 55–69 years, ‘based on a man’s values and preferences’. This nuance is missing from the website of the Prostate Cancer Foundation of Australia. The foundation remains passionately pro-screening and ‘disputes absolutely’ that ‘it would be better for men not to know whether they have the disease and therefore they should not be tested or treated’.
Each of the groups who advocate for some form of prostate cancer screening recommend that men should ‘talk to their doctor’. This is, in my opinion, an example of medical weasel words. I’ve used the prostate debate as a case study in the course on evidence-based medicine I’ve taught for a
decade to medical students at the Australian National University. The complexity of the arguments and the degree of epidemiological understanding required makes it hard for most of my students to form a balanced view, and most remain uncertain about the best course of action even after hours of lectures and tutorials. To say that laypeople, irrespective of their education and intelligence, can make an informed choice based on a short discussion with their GP is disingenuous. I suspect that the proponents of screening are aware that this nod to patient autonomy will result in patients agreeing to be screened, such will be their fear of having cancer.
* * * * *
Despite the evidence, though, personal experience will still tend to trump science. So, in the interests of balance, let me share the personal experience of one of my patients, who I’ll call ‘Dan’.
Ever since his 50th birthday, Dan’s wife had been pestering him to see their GP, but it was the inflight magazine article about the Qantas CEO’s diagnosis of prostate cancer that finally convinced him to act. He went to the pathology lab during his lunch break to have the tests his GP had ordered. He slept poorly that night and, for the first time ever, shouted at his daughter over breakfast the next morning. Less than a week later, the GP rang him. She explained that his PSA was just a bit over the normal range for a man of his age. She had written a referral to a urologist, and there was nothing to worry about. He rang the urologist’s rooms and made an appointment, then he cancelled the family trip to Noosa.
The urologist might have been taciturn, but he was neat and tidy and his rooms spoke of material success. After a few questions about Dan’s history and the stock market, he asked him to take off his trousers and underwear and to ‘pop up on the couch and face the wall’. Dan’s mouth was dry now. ‘This will be a bit uncomfortable,’ the urologist said, his right hand now gloved and lubricated. Dan couldn’t recall if the urologist told him that his prostate felt enlarged, but he did remember that he was booked in for a prostate biopsy a week later.
He was given an antibiotic on the morning of the procedure, and he sat in the waiting room overcome by a mounting sense of dread. This was not like him. After the biopsy, his wife picked him up and drove him home. They didn’t talk much in the car but he did ask her to slow down over the speed bumps outside the school. His wife made a few laconic (and not terribly funny) comments about his tentative cowboy swagger when he walked down the hall. He noticed that his urine was a little red – more rosé than pinot, but disturbing nevertheless. He went back to work the next morning and, apart from having to find a reason to spend most of the day on his feet, he felt quite normal.
While he was brushing his teeth that night, though, he suddenly felt a chill. Within half an hour he was shaking uncontrollably. He took a couple of paracetamol tablets and assumed he was coming down with the flu. He went to bed but spent a restless night, intermittently feeling hot then cold, and sleeping for only a few hours at a time. The following morning he felt sicker than he had ever experienced before and he began to feel a little panicky. His wife took him to their GP, who sent them straight to the emergency department.
Dan sat in the waiting room. While his wife spoke to the triage nurse he slid off his chair onto the floor. He was taken by wheelchair to a resuscitation cubicle, where his blood pressure was found to be low and his pulse was racing. His temperature was 39.8° Celsius. He noticed that he was starting to detach himself from the noise and action around him. He stopped worrying about things for the first time in 24 hours.
I first saw him a few hours later. The sequence of events that had led to his admission had been distilled into a single request from the emergency department consultant: ‘Would you accept the care of a 50-year-old man with septic shock?’ Septic shock is one of the most serious conditions in medicine. It has a mortality rate ranging from 10–60 per cent, depending on the type of germ involved. By the time I saw Dan he had been transferred to the intensive care unit and he had a tube down his throat and a machine was helping him to breathe. The emergency doctors suspected that the shock was caused by a bowel germ that had entered his bloodstream at the time of his prostate biopsy. The most likely culprit would be a Gram-negative bacterium known as E. coli.
I went with the ICU doctor in charge of the shift to talk to Dan’s wife. She asked if he was going to be all right. The ICU doctor managed to gently convey the gravity of the situation while leaving her with a sense of hope and qualified reassurance. Of all the germs to have circulating in your blood, E.coli is among the least worst. The mortality rate is around 10 per cent – so the odds were in his favour. He remained critically ill over the next 24 hours, but when I saw him the next morning he was sitting up in bed and eating lunch. The tube down his throat had been removed and his blood pressure and pulse were now normal. His wife was by his side. He couldn’t remember much of what happened leading up to his hospital admission and it appeared that he hadn’t connected the ICU admission with the prostate biopsy.
‘I’ve had some excellent news this morning, doctor,’ he said between mouthfuls of corned beef and mashed potato. I assumed he meant that he would be getting out of ICU later that day. ‘I asked one of the docs here to look up the result of my prostate biopsy,’ he said, smiling. He grabbed his wife’s hand. And then, with tears welling in his eyes, he told me that the biopsy was clear – which meant that he didn’t have cancer. ‘I’m a very lucky man,’ he said.
It was neither the time nor the place to do anything but agree.
* * * * *
Infection is one of the most important complications of prostate biopsy. (Others include blood in the urine and semen, rectal bleeding and difficulty passing urine, and these are usually referred to as ‘frequent’ and ‘minor’ in the urological literature.) Since the biopsy needle is passed through the wall of the rectum it is almost inevitable that some of the bacteria that live in the bowel will be carried with it.
The prostate gland is known as a ‘sterile site’. Under ordinary conditions it doesn’t have any bacteria in it, unlike the mouth, nose, stomach and bowel, where the ‘good’ germs and the ‘bad’ germs live side by side. The plumbing of the male urinary tract is interconnected, so an infection in the prostate can spread to the bladder and the kidneys. Bladder infections are usually relatively benign, but kidney infections are more serious and almost always require hospitalisation. Sometimes, as in Dan’s case, the infection moves directly from the prostate to the bloodstream. We see about 300 bloodstream infections each year in our hospital in Canberra, which serves a population of about 500 000. They are more common than a heart attack, and more lethal by a factor of at least two and as much as 12. As is the case with a heart attack, mortality rises if you don’t institute the correct treatment as a matter of urgency.
In the early days of TRUS biopsies as many as 70 per cent of men developed a urinary tract infection after the procedure. But it soon became clear that antibiotics taken before the procedure (known as ‘prophylaxis’) would reduce the risk of infection. The choices of effective prophylaxis are limited, and until recently a drug known as ciprofloxacin has been the most popular in a very narrow field. With prophylaxis, 1–5 per cent of men undergoing TRUS biopsy develop a bladder or kidney infection and 0.5–2 per cent develop a bloodstream infection. Most urologists consider something that occurs at a rate of between 0.5 to 2 per cent to be rare.
Let’s do some more maths, using 0.5 per cent, the low end of the estimated rate of bloodstream infection following TRUS biopsy. There were just over 29 000 prostate biopsies performed in Australia in 2012, giving a conservative calculated bacteraemia rate – the rate of bacterial infection – of 145 cases. Almost all of these patients will need hospital admission and many will end up in ICU. Most TRUS biopsies are performed in otherwise well men, so they would be expected to have a lower death rate from bacteraemia than patients with concomitant illnesses. There is currently no national surveillance of infections related to prostate biopsy so we don’t know what that rate is. Let’s ass
ume it’s a very small number. The deaths would be distributed across Australia and it is unlikely that any single urologist would see more than one or two in a professional career. Similarly, no single ICU will have a concentrated experience of TRUS-related septic shock. A distributed problem usually remains a hidden problem.
But this is all changing very rapidly. A Canadian study has shown a 400 per cent increase in the rate of serious infections following TRUS biopsy in the ten years up to 2005, and it is assumed that most of this rise is due to emerging antibiotic resistance. People who travel to India, China or South East Asia have roughly a 50 per cent chance of picking up a ciprofloxacin-resistant organism in their bowel that will stay there for up to six months after they return home. As we have seen, germs can live happily in your bowel and cause you no harm. Unless, that is, we stick a needle through your rectum into your prostate.
The problem is that there is no other antibiotic that can be taken by mouth that works as well as ciprofloxacin. The alternatives need to be given by an intravenous infusion, adding to the cost and time that prophylaxis takes. And the more we use these previously reserved antibiotics, the faster we drive resistance to them.
It’s important for infectious diseases doctors to give our urology colleagues antibiotic advice, but focusing on the microbiological detail misses the real point: the best way to prevent an infection associated with a TRUS biopsy is to avoid doing the biopsy in the first place. We don’t know it for a fact yet, but it is plausible that the reason that there is no difference in the all-cause mortality between those who are screened for prostate cancer and those who aren’t is that the reduction in cancer deaths is replaced by a corresponding rise in infection-related deaths. The emergence of more resistant bacteria will increase the risk of infection associated with every TRUS biopsy. Inevitably, the current drawn contest will turn into a one-sided bout, with infection winning hands down.