Cancer Screening

We believe that decisions regarding the benefit of cancer screening are best made by the individual patient and physician. We do not feel that such decisions should be left to organizations that set themselves up as "authorities" in such matters. Their pronouncements are generally couched in very general terms. Their conclusions may make good headlines, but are too often wrong when it comes to individual people.

To give you an idea why we say this, let's look at a table put together by one statistician¹ to illustrate the odds that screening for a particular cancer will prevent death from that cancer. (No doubt someone else may come up with slightly different numbers.)

Cancer site	Screening Method	Chances of Preventing Death from Cancer
Breast	Annual breast exam with mammography	1 in 120 to 430 persons
Colon	Annual stool blood test	1 in 340 persons
Colon	Annual stool blood test with flexible sigmoidoscopy every 5 years	1 in 90 persons
Cervix	Pap smear every 3 years	1 in 100 persons

For example, by this estimate, between 120 and 430 women must undergo annual mammography and exam to prevent one of those women from dying of breast cancer. As it happens, that number is very much under question. There is some evidence that mammograms, no matter how many you do, do not make a measurable difference in the death rate from breast cancer.

In fact, all these numbers are at best an estimate. Nonetheless, you need to understand the principle. Screening is based upon an estimate of odds--odds that you have the disease in question, odds that the screening test will pick it up in time to do something positive to help you.

Many of us may know someone who had a breast cancer found on routine screening, and who was treated early and successfully as a result. We may conclude that "mammography is a life-saving measure." But then we need to answer several questions.

• When should mammography begin? What age?
• How often should we do it?
• Is there an age when we should stop?
• How much are we willing to pay for our mammography program? (Maybe we are working in a less-developed country, and people have money for a school or a mammography center, but not both.)

To answer those questions, we have to work out a detailed analysis of the odds. How many cancers will we detect?

Early detection: helpful or not?

If we detect cancer early, how likely is that to make a difference in treatment? Sometimes if we find a cancer early, the person dies just when they would have died anyway. For example, pretend that you can look down on Earth and identify everyone who was going to die of pancreatic cancer on January 1st, 2000. You go back in time to 1990, and tell the doctors to screen half of that group for cancer every month, and to give the other half normal care.

What the doctors find is that, on average, the screened group show up positive on January 1st, 1998, while the unscreened group develop symptoms and are diagnosed at a later stage of their cancer, on July 1st, 1998. Once diagnosed, the doctors treat them all as best they can. But pancreatic cancer being what it is, and our skills being what they are, everyone still dies on January 1st, 2000. The statisticians go in and say "Look, those you screened lived six months longer than those you didn't screen!" You, however, know better and do not call the newspaper.

So, in evaluating a screening test, finding the disease early is of no use unless early treatment gives us an advantage. This is why we focus on certain diseases.

To help you with more practical details about this, here are some numbers about prostate cancer screening.

The PSA test for prostate cancer

Of a thousand men who have a PSA (prostate specific antigen) blood test to screen for prostate cancer, these will be the results.

• 900 men will have normal results on the test.
• 100 men will have abnormal results. These men will have an ultrasound test, and possibly a biopsy. The biopsy is done by placing a needle in the prostate to look for cancer.
• 70 of the 100 men with abnormal results will have no prostate cancer.
• 30 of the 100 men with abnormal results will have prostate cancer.
• Assuming the 30 men with cancer are treated with radiation and surgery, about 9 would become impotent due to the treatment. Another 2 would leak urine.

The 70 men who have a positive PSA test but no prostate cancer are said to be false positive on this test. Because of the large numbers of false positives on the PSA test, some authorities do not recommend routine screening using this test. In fact, we still don't know whether, even if the test were accurate, early treatment of prostate cancer prolongs men's lives.²

Many men with prostate cancer do well without surgery and radiation. We have some patients who have used no treatment whatsoever and whose cancer has grown so slowly that it has caused them less difficulty than other illnesses of aging that they have. However, not all prostate cancer behaves the same. Some cancers, especially in younger men, can be quite aggressive and lead to early death.

Prostate cancer in perspective

Although 10 men in 100 will eventually be diagnosed with prostate cancer at some point, only 2 or 3 men in 100 will actually die from that illness. The biggest risk for men over the age of 50 is heart disease, stroke and other kinds of cancer besides prostate cancer.

Who should get a PSA test

We are more prone to do this test in men who have a family history of prostate cancer, African-American men, or men with an abnormality on examination of the prostate gland.

There are advances in the PSA test. We are testing for "complexed PSA" and following rates of change of PSA over time. This improves the usefulness of the test.

Summary regarding prostate cancer

According to one expert writing in 1996³, there was no evidence that screening for prostate cancer improved health or longevity. [ Note added 2012- evidence continues to mount that the PSA does not effectively reduce death from prostate cancer. ]

Summary of recommendations regarding cancer screening

If you are over 50, we ask you to come in once a year for cancer screening. We check your skin head-to-toe to look for skin cancer. We may biopsy a suspicious area. We help you decide what colon cancer screening method is best for you. For women, we do a breast exam and may recommend, on an individual basis, mammography. For men, we do a prostate exam and may recommend, on an individual basis, a PSA test. We use your personal and family history, and the results of your physical exam, to help you avoid cancer or find it early.

In contradistinction to those who say "every woman over fifty should have a mammogram every year," we ask "How do you know that you are not better off just offering a mammogram to any woman who feels something in her breast and thinks that she needs a mammogram?" We also tell you what measures you can take to reduce your risk of these common cancers.

For most people, the individual approach to cancer surveillance is more suitable than the "across-the-board" recommendations of those who make the headlines.

1. Archives Internal Medicine (1996 May 27; 156 (12): 1069)
2. Archives Internal Medicine (1996 June 24;156 (12): 1333)
3. Archives Internal Medicine (1996 May 27; 156 (12): 1069)

7/16/07