In principle, a prostate biopsy is a straightforward process. An ultrasound probe inserted into the rectum helps the clinician to locate the gland, and hollow needles are used to penetrate and collect tissue from a dozen sites across the prostate. The procedure is done with a local anaesthetic, and takes less than half an hour. More than one million such biopsies are performed every year in Europe and the United States. But a growing number of clinicians think this is too many.
One of their main concerns is that it is left to chance whether tumours in the prostate are caught by a needle — unlike biopsies in other tissues, which target abnormalities already spotted on imaging. Although the random approach is a good way to find hidden tumours, it can also miss the clinically meaningful disease, providing a misleading view of actively growing tumours and thereby resulting in delayed diagnosis of aggressive cancer. Moreover, a sizable fraction of the tumours that are detected might best be left hidden. “One-third of men above the age of 50 have indolent disease,” says Hashim Ahmed, a surgeon at Imperial College London. “These biopsies are picking up that pool of indolent disease.”
Such slow-growing tumours generally pose little risk to life and can instead be monitored for any changes — a strategy known as active surveillance. However, Ahmed notes that many people still opt for surgery or radiotherapy, which are associated with a host of long-term urological, sexual and other side effects. There are also risks posed by the biopsy itself, which requires perforating the rectal wall to obtain the necessary tissue. A 2012 clinical trial in the United Kingdom found that one-third of men experienced moderate or severe consequences from their first biopsy1. “It was pretty grim reading,” Ahmed says. “A lot of men had a very bad experience with bleeding, pain, discomfort and infection.”
Alternative strategies for imaging and sampling the prostate, including the use of magnetic resonance imaging (MRI) and transperineal rather than transrectal biopsy, are starting to transform the diagnostic process. Many studies suggest that these approaches can help physicians to focus on serious disease and prevent unnecessary interventions. But the high cost and steep learning curve associated with these innovations might be delaying their uptake in some places, rendering the real-world impact of these strategies unclear.
Look before you leap
A fundamental limitation of a transrectal ultrasound scan (TRUS) is that it charts out only the contours of the prostate, not the composition of the tissue itself. “The ultrasound is just to make sure that the biopsies are performed in the prostate and evenly distributed within the gland,” says Olivier Rouvière, a radiologist at the Edouard Herriot Hospital in Lyon, France.
The emergence of sophisticated multiparametric MRI strategies in the mid-2000s represented a sea change. This took advantage of the increasing magnetic power available for MRI, and employed multiple imaging strategies in parallel to capture previously inaccessible details about prostate density, vascularization and other important tissue characteristics. Armed with this fresh view into the prostate, clinicians at last saw the opportunity to detect tumours without inserting a single needle.
A 2017 trial called PROMIS, headed by Ahmed and his colleague Mark Emberton at University College London (UCL), offered an important proving ground for this technology. A total of 576 men were subjected to both a multiparametric MRI and a TRUS-biopsy, followed by an even more comprehensive biopsy process to thoroughly map the prostate. The results showed that although MRI was more likely to produce false positives, it was also much less likely to miss real cancer, with a false-negative rate of just 7% versus 52% for TRUS-biopsy2.
A subsequent trial known as PRECISION went a step further, testing whether MRI could spare some people from having to undergo a biopsy at all. In the MRI arm of this trial, only people with a positive result from imaging were subjected to a biopsy; this was then targeted at the area marked on the MRI3. In total, almost one-third of men avoided biopsies entirely. “The MRI-targeted strategy detected more of the significant cancers and fewer of the insignificant cancers,” says Caroline Moore, a urologist at UCL who was one of the lead investigators.
The use of multiparametric MRI before a biopsy is now widespread in the United Kingdom, and is gaining momentum in the rest of Europe and in the United States. Ahmed estimates that among the hospitals in his subnetwork of the National Health Service (NHS), known as the West London NHS Trust, 40–45% of men are being spared biopsy thanks to a negative MRI.
But MRI is an added expense, and analysis is not always straightforward. Moore often reviews results from other centres. “We see lots of bad scans, and then we see some good scans that haven’t been read very well,” she says. Experience is an asset, and high-volume imaging centres tend to fare better than smaller clinics. Several groups have been working on artificial intelligence (AI) algorithms that can interpret MRI results, but Rouvière says that these can be confounded by factors such as the differences between images generated with different instruments or protocols. “We are not there yet,” he says.
Keeping it clean
When a biopsy is required, clinicians now have access to a more sanitary way to do it. One of the most serious issues associated with a TRUS-biopsy is the infiltration of pathogenic microbes from the gut into the bloodstream. A 2017 analysis by the American Urological Association reported that sepsis — a life-threatening immune response to an infection — occurred after routine biopsies in 0.3% to 3.1% of cases4.
Antibiotics can minimize, but not eliminate, this risk. Michael Gorin, a urologist at the Icahn School of Medicine at Mount Sinai in New York City, notes that this heavy reliance on antibiotics might be contributing to the emergence of drug-resistant bacteria. Urologists prescribe more antibiotics than any other type of non-primary care practitioner in the United States. Gorin points out that his profession routinely deals with a variety of urological infections, but that “there’s no doubt that the more than one million prostate biopsies performed each year in this country using the transrectal approach is driving up our specialty’s numbers”.
Since the early 1980s, clinicians have known that they can mitigate the risk of complications by performing transperineal biopsies — accessing the prostate through the region between the genitals and the anus. However, this typically required the use of general anaesthesia to eliminate the pain from repeated needle probing at this site, making transperineal biopsy impractical for routine use.
Over the past decade, however, a variety of techniques have emerged for performing transperineal biopsies under local anaesthesia, so that the procedure can be done in an outpatient visit. This type of biopsy mostly eliminates the risk of severe infectious consequences and reduces antibiotic use. For example, a 2022 systematic review by Gorin and colleagues found only a 0.05% incidence of sepsis in more than 37,000 people undergoing transperineal biopsy5.
Moore exclusively performs transperineal biopsies, and tends to give her patients just a single dose of an antibiotic immediately before their biopsy, “but none to go home with, and none beforehand”, she says. “There are quite big studies of people doing no antibiotics, which seems to be OK, but we haven’t quite bitten the bullet yet on that one.” Importantly, this technique does not compromise diagnostic performance — and some clinicians think it might even catch tumours in regions of the prostate that are difficult to sample transrectally.
Transperineal biopsy is becoming the standard for numerous urologists. Gorin routinely performs the procedure under local anaesthesia, and Ahmed oversaw a transition to the technique for hospitals across the West London NHS Trust in early 2017. But not everybody is adopting this strategy: Gorin estimates that only about 10% of US practitioners are using transperineal biopsy.
Rouvière is also not all-in on this approach. He is sceptical about its ability to access more of the prostate, and his team is still grappling with making the procedure more comfortable for patients. “We did not find yet a perfect way to perform transperineal biopsy under only local anaesthesia,” he says. Ahmed agrees that making people comfortable and relaxed in this regimen can take considerable effort, and that some people ultimately opt for sedation or general anaesthesia. The process also takes considerably longer and requires extra equipment, including specialized ultrasound probes. Gorin thinks that these costs could be part of the reason for the low uptake of transperineal biopsies in the United States. “For a procedure that you don’t get paid a lot of money for but you do a lot, it could potentially mean a big drain on the practice,” he says.
Even with these technological advances, the field of prostate cancer is still grappling with some fundamental questions. One of these is how best to perform and analyse biopsies to minimize overtreatment while preventing as many cancer deaths as possible.
On the one hand, it remains unclear whether people undergoing a biopsy after an MRI scan should have a full systematic biopsy, in which a dozen tissue cores are collected from distinct sites, or just targeted sampling of the spots flagged in the scan. Rouvière thinks systematic biopsies will eventually fall out of use, but he continues to perform them on people with positive MRIs or high levels of prostate-specific antigen — a widely used biomarker for prostate cancer — to ensure that nothing gets overlooked in the diagnostic process. “It’s a safety net,” he says. Capturing more tissue cores can also offer clearer information about how aggressive a tumour is, which will guide treatment decisions.
On the other hand, some researchers have raised questions about whether MRI’s deeper view of the prostate might also be creating opportunities for overtreatment by uncovering more intermediate-risk tumours that do not necessarily pose a clear and present danger. Rouvière points out that even though MRI can clearly improve prostate cancer detection and spare some unnecessary medical procedures, there is no clear evidence that it improves patient survival.
But in the end, overtreatment is mainly the result of the decision-making guidelines that clinicians follow. Fresh technological insights could lead to more evidence-based strategies for designating which tumours are clinically significant. For example, Ahmed notes that MRI scans routinely uncover tumours that have historically been considered just dangerous enough to merit medical intervention. “But there’s concern now that those are also indolent or slow-growing, and could therefore be on active surveillance,” he says. “So active surveillance is shifting more and more into the medium-risk groups, and I think that’s a good thing.”