European Urology

Prostate Cancer

Which biopsy procedure is optimal and which men should undergo it are pivotal questions regarding prostate cancer diagnosis. Scattoni et al attempt to address the first question by presenting a flowchart that aims to utilize the lowest number of biopsy cores to detect 95% of cancers that would have been detected with 24 cores [1]. Their flowchart uses digital rectal examination findings, prostate volume, and patient age to direct patients to one of four possible 10–16-core biopsy schemes.

This study begs the larger question of what the ultimate biopsy decision support system might look like if expanded beyond cancer detection rates to the more important end point of clinical outcome. We believe this question to be more important: Which men should undergo any biopsy scheme, and how should it be performed, if at all?

Although prostate biopsy is widespread and regarded as a routine office procedure, some men are probably harmed, broadly speaking, by the act of biopsy, while other men are more likely to benefit. Several characteristics would seem likely to distinguish these two groups.

The man most likely to benefit from biopsy has a high risk of having “curable” prostate cancer. Although this term is traditionally applied simply to any situation in which it is possible to render the patient free of the disease, for prostate cancer we would define a curable cancer as one that both justifies treatment and can be effectively treated. The details of this assessment are not easy. The concept that a tumor needs treatment implies that harm will ensue without intervention, while effectively treated means that intervention has done more good than harm. Thus, treatment should ultimately lead to improved disease-related outcomes without creating morbidity that outweighs this benefit. The likelihood of risk of disease and ability to cure it, if present, is almost impossible to establish in an individual patient, especially prior to biopsy. Another characteristic of a man who is likely to benefit from biopsy is anxiety based on suspected prostate cancer; the man should feel relief if his biopsy is negative and can proceed to management if positive. A man who does not care about sexual dysfunction may also be more likely to benefit from biopsy because the possibility of sexual dysfunction caused by prostate cancer treatment would not create a negative impact for him. If all else is equal, a younger man in otherwise excellent health would tend to benefit from biopsy because he should have a life expectancy in excess of his expected time of prostate cancer progression, although it is not known how many years of potential life expectancy are required to receive this benefit.

A man less likely to benefit from biopsy has a high risk of having the extremes of indolent or incurable prostate cancer. Diagnosing a cancer that is likely to remain indolent and asymptomatic throughout life only carries the weight of psychological distress, reducing quality of life. Unnecessarily reducing a man's quality of life without prolonging his life expectancy is doing harm. At the opposite end of the spectrum, early warning of an inevitably incurable cancer might only do harm to many men by reducing their quality of life prematurely by the knowledge of inescapable cancer death, especially if ineffective treatments further reduce their quality of life. Finally, a man whose life expectancy is shorter than would be required for most cancers to create morbidity or untimely death is unlikely to benefit from biopsy.

Therefore, the ultimate biopsy decision support tool needs to enable decisions about who to biopsy. Given the above arguments, the tool would likely need to incorporate a man's preferences for possible future health states. Additionally, it will need to estimate life expectancy and, thus, incorporate comorbidity information. This will take some effort but is undeniably possible.

Assuming this future decision support system recommends biopsy, we next need to decide what the biopsy is supposed to achieve, First, it should rule in or out a curable prostate cancer. This is very different from detecting any cancer and thus expands the traditional metric for biopsy performance: simple cancer detection sensitivity. Prognostic information to determine whether treatment would benefit the patient would be vital because a cancer that is too indolent to cause harm or too impervious to treatment to justify even trying should probably remain unrecognized until clinical symptoms dictate. Second, a biopsy scheme is preferred if it provides incremental prognostic accuracy in those diagnosed with curable prostate cancers. A third benefit of one possible biopsy scheme over another occurs when a scheme provides improved treatment delivery. An extended biopsy scheme, for example, would be preferred if it guided the surgical approach, resulting in a lower positive margin rate, or enabled selective or “focal” treatment. Given the accomplishment of those three tasks, one would prefer the biopsy scheme with the least morbidity, which might limit both the number and location of cores to an optimal scheme.

Unfortunately, defining who has curable prostate cancer—that which needs treatment and for which treatment exists—is a big problem. For starters, we never know in an individual patient whether he benefited from his treatment because we cannot know what would have happened, favorably or unfavorably, had he forgone intervention. Thus, it is not so easy to learn from our mistakes. Probably the closest we can come is to use all available and emerging data to model a prospective patient's life expectancy and morbidity with and without treatment. This goal is the intent of a large body of research within the comparative effectiveness heading, providing prospective patients with the predictions of benefits and harms as a function of the available treatment options.

In summary, the type of biopsy or no biopsy at all should be selected based on that which maximizes that patient's quality-adjusted life expectancy. To accomplish this goal, we have tremendous work remaining toward the development of the ultimate biopsy decision support tool.