Screening & Diagnosis of Prostate Cancer

You must undestand the concepts of SCREENING and DIAGNOSTIC TESTING before you can understand prostate cancer screening.

 

Screening has two primary properties:

  1. Screening is a DIAGNOSTIC PROCESS which uses a series of TESTS. Like any other clinical decision making process, it is based on PROBABILITIES.
    1. When the probability of disease being present is close to 100%, the presence of disease is confirmed.
    2. When the probability of disesae being present is close to 0% the disease is ruled out.
  2. What defines screening as a unique diagnostic process is that it attempts to identify disease often when there are NO SYMPTOMS. The ultimate goal is detection of disease at an EARLIER STAGE then would occur through routine stage with the hope of BETTER OUTCOMES.

 

Many patients who have screen detected cancers ask why they don't have any symptoms. The reason is that by the time most cancers cause symptoms the chances of cure is very low. By it's nature, screening is meant to detect cancer at an earlier stage before symptoms develop when the cancer has a better chance of cure. If one were to wait until symptoms are present, the outcome would probably not be as good.

 

Diagnosis & Testing

 

It is critical to understand the DIAGNOSTIC PROCESS is one of determining PROBABILITIES. While the goal is to be as certain as possible in a diagnosis, sometimes compromise is necessary.

An example of when compromise is necessary would be when the test which could confirm the diagnosis has a high likelihood of resulting in the death of the patient. In these cases, treatment might be started without a definitive diagnosis - so long as the risks of the treatment were acceptable!

 

Fortunately in much of medicine the diagnostic process is fairly straighforward and results in a conclusive diagnosis - confirming (or excluding) the presence of a disease. Stated another way, the result is often 'black and white'. However, in some cases the degree of certainty is more 'grey' - that is, after a series of tests there remains uncertainty in the diganosis. In some cases it is impossible to entirely exclude the presence of a disease. In these situations there are significant consequences - this will become apparant when we look at prostate cancer screening.

 

TESTS are the tools that allow a clinician to determine the probability of a disease and ultimately reach a diagnosis. Testing takes many forms:

  • history (symptoms)
  • physical examination
  • blood testing
  • imaging (e.g. prostate MRI)
  • genetic testing
  • biopsy

An ability to select the best test from the toolbox (or series of tests) for any particular patient is what physicians are experts in. Selection of the test is important because for most tests there are a number of compromises in respect to diagnostic performance, cost and potential harm to the patient. For example, biopsies are usually highly accurate at establishing a diagnosis but tend to have higher risks for patients and are more expensive.

 

To say a test is 'inaccurate' is unhelpful since every test has at least some degree of 'inaccuracy'. Medicine is based in large part on  quantifying the strengths and weakeses of every action that is taken. Medical studies and clinical trials are the powerful tools used to provide the evidence for the actions we take. Finally, a much neglected but important determinant of test performance is understanding how the test performs when it is used in various specific scenarios - this is often determines the ultimate value of a test.

 

The DIAGNOSTIC PERFORMANCE of a test describes how well the test does in assigning a probability that a disease is present. That is, how well it performs its primary function. ROC curves and likelihood ratios are 2 of the best measures. Measures such as accuracy, precision and false positives and negatives are more susceptible to the circumstances around testing and therefore are usually an oversimplification. There are multiple ways in which the performance of a test can be measured. 

  1. Receiver operator characteristic (ROC) curves
  2. Likelihood ratios (also see here)
  3. Accuracy and precision
  4. False positive and false negative

 

Diagnostic performance is very different then the CLINICAL UTILITY of a test. Diagnostic performance doesn't take into account the costs or harms of a test. Furthermore, just because a test might show a statistical improvement in the ability to make a diagnosis does not mean that it will affect clinical decision making. Sometimes, increased diagnostic peformance does not translate into better results or change the recommendations for further testing or treatment. It is often said in medicine that 'statistically significant does not always mean clinically significant'.


Interpretation and application of test results, therefore, is highly CONTEXTUAL. For example, the interpretation of PSA depends on the circumstances - screening, after surgery, after radation, during chemotherapy, etc.

 

Diagnostic Testing for Screening in Practice

 

As discussed, screening is a diagnostic process that uses tests to reach a diagnosis in patients who do not have symptoms. There may be multiple steps in this process. Each step of the screening process can end at one of 2 places:

1. An actual diagnosis - either establishing the presence of a disease or ruling out the presence of a disease. Screening colonoscopy is a good example of this - if a tumor is seen, a biopsy can be taken immediately and a diagnosis can be made.

2. Identify patients who require additional or repeated testing. PSA for prostate cancer screening is an example of this. PSA rarely provides enough certainty to confirm a diagnosis. It can only help in the estimation of risk to determine people in whom prostate biopsy might be worthwhile.

 

The trade-offs in the 2 types of tests used in this example (colonoscopy vs. PSA) can be seen: colonoscopy is expensive, invasive with a small but definite risk of a serious complications but high accuracy; PSA is an inexpensive, easy blood test which has low accuracy.

 

Note that the language of risk and probabilities of a test is quantitative (based on numbers) but that decisions about whether it is worthwhile to proceed is ultimately a value based one. For this reason, the recommended approach to making that decision is called 'Informed and Shared Decision Making'. It is important that you understand the benefits and harms of screening tests and make an informed choice about which screening tests are right for you. 

 

Just as any diagnostic process has limitations, so does screening. There are are, however, some very specific limitations to screening: 

  • false-positive and false-negative tests results are possible
  • screening can identify cancers which are not destined to cause any problems
  • screening may lead to treatment which may not improve the person's health or help them live longer - in fact, the treatments may worsen their health or cause them to die prematurely
  • screening tests can be harmful - while the needle poke needed to draw a PSA has virtually no risk, a perforated colon during screening colonoscopy can be life-threatening.
  • screening can be expensive and divert limited resources from other worthwhile diseseases

 

As you can see, screening is not always a 'slam dunk'. Early detection can most definitely result in better outcomes for some patients but can also result in overdiagnosis and overtreatment in others. When the stakes are higher the decision process gets alot more difficult. This is the area where prostate cancer screening lives. 

 

Prostate Cancer Screening

 

The rationale for prostate cancer screening comes down to 2 characteristics of the disease:

  • Prostate cancer is a serious disease (in some men). It accounts for 5% of all cancer deaths. 11% of men will receive a diagnosis of cancer in their lifetime. 
  • Advanced prostate cancer is not curable. It has to be caught early

 

Decision making in prostate cancer screening is complicated by a number of factors - it's a high stakes game in with the potential for overdiagnosis/overtreatment AND underdiagnosis/undertreatment all at the same time. 

  • Prostate biopsy is required to establish a diagnosis: it is currently the ONLY test that can do this. While biopsy is well tolerated in the majority of patients, unfortunately it can result in serious infection in about 1 in 25 patients. An even bigger issue is that screen-detected cancers are more likely to be clinically insignificant. 
  • There is NO currently available test that can exclude the presence of prostate cancer with anywhere close to 100% certainty.
  • Disease spectrum is high: prostate cancer has a split personality - some prostate cancers are lethal within a few years but others do well without any treatment decades after diagnosis - then there is everything in between.
  • The majority of patients die with prostate cancer, not because of it.
    • Of all new diagnoses, 98.2% will be alive at 5 years regardless of whether they receive any treatment
    • 100% of patient with localized or regional disease are alive at 5 years BUT 
    • only 30% of patients with distant disease are alive at 5 years despite aggressive treatment
  • Disease course occurs over a very long time in most patients: the vast majority of cancers diagnosed with screening are localized and all of these have a very long clinical course. The large majority of men will die from another cause 10-20 years after a diagnosis of prostate cancer is made. For those men who ultimately succumb to prostate cancer, this usually occurs 10-20 years after diagnosis and very rarely before that time.
    • The decision to proceed further along the diagnostic and therapeutic pathway is therefore made well before the disease itself is capable of causing any symptoms.
  • Treatments can cause serious harm. Surgery and radiation have the potential to cause significant urinary, sexual and bowel dysfunction. This can become worse over time.

 

All of these things are meant to be considered when one starts with the simple PSA test. It's alot to think about.

 

THE CHALLENGE WITH PROSTATE CANCER SCREENING IS STRIKING THE RIGHT BALANCE FOR EACH PATIENT

 

 

Should I have a prostate biopsy?

 

The PSA test is a pretty simple test - it only requires a blood sample. What happens next is more complicated.

The decision to have a biopsy is based on the comparitive risks and benefits of 2 courses of action - those of not doing a biopsy (i.e. continued observation) vs. those associated with doing a biopsy.

 

 

The decision to have a prostate biopsy is ultimately a personal decision since only you can determine what level of risk is acceptable and what your values are. The basic tenant of making such a decision is to make an informed decision based on some facts in consultation with an expert.

 

Facts regarding prostate cancer to keep in mind when deciding if a biopsy is right for you.

  1. There is no currently available test which can exclude the presence of prostate cancer with 100% certainty . Therefore every man has some risk of having prostate cancer no matter what his family history, age, race, PSA, or prostate exam findings. Having favorable features can significantly reduce the risk that you have prostate cancer, but the risk is never 0%.
  2. Not all prostate cancers require treatment because not all men who have prostate cancer either need or will benefit from treatment. This is discussed in the 'Natural History of Prostate Cancer' page.
  3. Some prostate cancers definitely benefit from treatment - in terms of increasing both the length and quality of life.

 

Questions which you should be asking before making a decision on having a prostate biopsy.

  1. What is the risk that I have cancer? More specifically, what is the risk that I have a high-grade cancer?
  2. If I have prostate cancer, what are the risks and benefits of receiving treatment vs. not receiving treatment (observation)?
  3. What are the risks of the biopsy itself and how can those risks be reduced?
  4. What are the limitations of biopsy?

 

What is the risk that I have cancer?

 

Before assessing what an individual's risk of having prostate cancer is, it is important to understand the baseline risk of prostate cancer - the prevalence of prostate cancer. It is not feasible or reasonable to biopsy every male, and therefore, it is helpful to know how one's risk compares to that of one's peers. There are a few ways in which prostate cancer can be diagnosed:

  1. At the time of autopsy: this gives the true prevalence of prostate cancer and is assessed after a man has passed away. Allows for complete evaluation of the entire prostate gland.
  2. Prostate biopsy in men 'for cause': in men who are felt to be at 'increased risk' of having cancer.

 

AUTOPSY PREVALENCE VS. BIOPSY PREVALENCE. A frequently quoted saying is that 'all men would eventually develop prostate cancer if they lived long enough'. Without question, the prevalence of prostate cancer increases in any age group over time. When based on autopsy data the prevalence of prostate cancer in 80-year-old men is between 50% and 80%. While it is true that most elderly men harbour prostate cancer, only about 20% of men receive a diagnosis of prostate cancer during their lifetime and furthermore fewer than one and 5 of these men (3-4% of all man) specifically die from prostate cancer. Therefore, there are striking differences between the prevalence of diagnosis when made at the time of autopsy compared to biopsy while a man is alive. Chosing to compare onself to an 80 year old man diagnosed at the time of autopsy is not appropriate.

 

The group that is the most important for comparison is that of peers in your age group who have a similar risk of harbouring prostate cancer as you. This sort of data has only been available since 2003 and was generated by doing biopsies in a large number of men as part of the Prostate Cancer Prevention Trial (PCPT) - including many who would have been previously been considered at low risk and never subjected to a prostate biopsy. Prior to this, it was felt that men with low PSA's (less than about 2.5 mcg/L) and normal prostate exams had a negligible risk of prostate cancer. Note that prostate biopsy underestimates the prevalence of prostate cancer compared to autopsy diagnosis because only part of the prostate is examined.

 

Estimating an individual's risk of harboring prostate cancer

 

Our understanding of an individual's risk of harbouring prostate cancer has changed significantly since the PCPT trial in 2003. Prior to this trial, only men with abnormalities on prostate exam or who were felt to have increased PSA levels were sent for biopsy. The PSA level at which a man was felt to be at risk was usually above 4 mcg/L, though other 'cut points' were used by various authorities. Some physicians tried to better select men by using different PSA thresholds based on age or race; others attempted to calculate the rate of change of PSA over time (PSA velocity); still others tried to evaluate the contribution to the overall PSA level by benign enlargement of the prostate (PSA density). Then the PCPT trial came along and demonstrated that:

 

THERE NO 'NORMAL' LEVEL OF PSA AND THAT THE USE OF SPECIFIC PSA VALUES ('CUT POINTS') TO DECIDE IF A MAN SHOULD HAVE A BIOPSY IS INAPPROPRIATE.

 

Despite this, as of 2011 laboratory requisitions are still printed with 'Reference' or 'Normal' ranges and decisions to perform prostate biopsies are often made on the basis of a PSA level alone. The risk of harboring prostate cancer in men who were previously considered to have neglible risk of having prostate cancer if a biopsy is performed*:

 

 

Age Range Overall Risk of Any Cancer
Risk of High Grade Cancer
50-60  14%  1.0%
60-70  14%  1.4%
70-80  14%  1.9%
80-90  14%  2.6%

* Men with no family history of prostate cancer, PSA less than 1 mcg/L, and no abnormality on prostate exam.

 

Therefore, virtually all men have at least a 1 in 10 risk of harboring prostate cancer. This can be quite alarming, until one recognizes that the majority of prostate cancers do not represent lethal tumors and are unlikely to affect the quality or quantity of life in most men when left untreated. Fortunately, the most important number is the risk of high-grade tumor since these tumors are much more likely to have an impact on a man's quality and quantity of life. The chances of a high-risk cancer are much lower than the risk of harboring prostate cancer overall.

 

Estimating Your Risk of Having Prostate Cancer

The following factors have been demonstrated to predict a man's risk of harbouring biopsy-detectable prostate cancer:

  1. PSA level
  2. Prostate exam abnormalities
  3. Age
  4. Race
  5. Family history of prostate cancer
  6. Prior biopsy not showing cancer (negative biopsy)

Note that PSA velocity (PSA change over time) has NO bearing on the risk of harboring prostate cancer once these other factors are considered. You can estimate your own risk of having a biopsy-detectable cancer by clicking the link below. There are some very important caveats when using the risk calculator:

  1. The accuracy of the estimate is only as good as the quality of information put into the calculator. The calculator does NOT take into account important information such as:
    1. The severity of the prostate exam abnormality. Some findings are very subtle where as others are quite striking.
    2. The strength of the family history. For example, prostate cancer which developed in a brother at a young age is of great importance where as cancer in a relative such as a cousin or maternal uncle has very little importance.
    3. Normal, non-cancer related fluctuations in PSA. For example, sudden large rises in PSA are usually due to non-cancerous causes such as infection or prostate infarction.
  2. The estimate of risk is a 'point estimate' based on the data available at the time of the calculation. Changes in the clincal picture can affect the calculated risk.
  3. The estimate is APPROXIMATE. There is inherent inprecision in any sort of calculation.
  4. Use of the calculator may not be appropriate in your particular situation. The PCPT data was generated from men who started with PSA's less than 3 mcg/L and normal prostate exams. Extrapolation to men outside of these circumstances may not be appropriate.

 

THE PRIMARY VALUE OF THE CALCULATOR IS TO RECOGNIZE THAT FACTORS OTHER THAN PSA PLAY AN IMPORTANT ROLE IN ASSESSING RISK AND TO PROVIDE A BALLPARK ESTIMATE OF THAT RISK

 

 

 

References:

Lange Textbook of Epidemiology

Cancer Screening Overview (PDQ) - Patient Version was originally publshed by the National Cancer Institute April 2017

 

On The Web

Genetic testing for BRCA1, BRCA2: saliva test. Myriad Genetic Laboratories, Ambry Genetics and Gene Dx (2-4 weeks)

Deciding on When a Prostate Biopsy Should be Done

Estimating Your Risk of Harboring Prostate Cancer

Prostate Biopsy Risk Calculator

About Prostate Biopsy: Interpretation & Limitations

Screening for Prostate Cancer

Prostate Cancer on the Web

 

Prostate Cancer Screening Is Controversial - There is the potential for benefit and harm. Read these links for more detail.


American Urological Association Guidelines on Prostate Cancer Screening

Canadian Urological Association Guidelines on Prostate Cancer Screening & Response to CTFPHC Recommendations

Canadian Taskforce on Preventative Health Care

National Institutes for Health/NCI PSA

New York Times Editorial on Prostate Cancer Screening 2016


General Prostate Cancer Web-Resources

Memorial Sloan-Kettering Cancer Center in New York is an excellent resource for information on prostate cancer. Balanced, unbiased discussions of the disease, including discussion regarding some of the controversies in prostate cancer.

General Information on Cancer

UNDERSTANDING CANCER - Metrovan Urology info on the principles of diagnosis, staging, prognosis and more.

American Cancer Society

BC Cancer Agency: Good general website from the British Columbia Cancer Agency. Has contact information on locations.

National Cancer Institute: Excellent source of understandable and mainly unbiased information. Several very good brochures on every stage of prostate cancer.

National Comprehensive Cancer Network: peer-reviewed expert content/prostate cancer guidance on evidence-based cancer diagnosis and management. Best for Prostate and Kidney Cancer. The most in-depth information is located in the physician section and requires registration.