CANCER SCREENING: Flickering Lights on the Christmas Tree

Joan Pont, MD, FACP

The swing in cancer screening over the last 50 years is like putting hundreds of lights on a Christmas tree, then realizing that most of them don’t work and need to be removed. With current technology, screening is effective in very few diseases. Physicians want the repertoire extended, and we await new advances in accurate serology and imaging.

The structural concept of screening is that a disease begins. We use a modality to detect its presence, and we do something to avoid death from that particular disease in the future, i.e., find it at a curable stage. We also need to consider balancing factors, such as not causing an equal amount of unintended morbidity or mortality in treating patients with false positives (no cancer altogether) or indolent disease that would not become manifest if left untreated (lead time bias). Specifically, knowing a patient has a disease eight years before death versus two years before death might make you think that treatment or detection lengthened survival time, whereas the natural history might be identical, and the six-year apparent difference is from acknowledging the disease’s presence at different times.

The idea of cancer screening goes back to 1928, when Dr. George Papanicolaou first reported cervical cancer cell detection. Careful follow-up brought this concept into a functional and mature state, leading to the publication in 1943 of Papanicolaou’s landmark book, Diagnosis of Uterine Cancer by the Vaginal Smear. The book launched widespread dissemination of the Pap smear and led to a dramatic reduction in deaths from cervical cancer.

To physicians, the Pap smear must have been a jolt, like landing on the moon or curing pellagra with niacin. Yet these advances are diametrically opposed. Landing on the moon takes the efforts of many smart people and loads of money to create a machine safe and powerful enough to get there and back. Avoiding pellagra takes eating corn prepared with alkali, or fruits and vegetables, or fortified foods—all simple maneuvers with great clinical benefit.

Cancer screening involves a little of both. There are huge endeavors involving technology and people for detection, screening, treatment and follow-up; but there are also simple steps we take every day in the office. The evaluation of symptoms or signs suggestive of cancer is very different from cancer screening. Weight loss, fever, bleeding, new and increasing localized pain require a diagnostic workup independent of screening recommendations. When a man has a breast lump, he may be evaluated for breast cancer. However, it would be of very limited value to screen all men for breast cancer.

My recommendations for cancer screening are aligned with programs that have shown aggregate benefit to the defined population. They are summarized in the U.S. Preventive Services Task Force recommendations.[1]

Cervical cancer screening for women between 21 and 65 years old reduces cervical cancer mortality. Colorectal cancer screening is estimated to increase aggregate life of adults over 50 by eight months compared with an unscreened population. Some people may have life extended, but many get tested and derive no benefit. That is where sophisticated risk/benefit calculations based on multiple studies boil down to practical recommendations.

Breast cancer screening with mammography every 1-2 years for women 50-75 years old reduces breast cancer mortality. We generally extend screening to women 40-75 years of age, acknowledging a delayed benefit. Lung cancer screening in ever-smokers is being better defined as research continues.

Those are the blinking lights on the Christmas tree. Other screening modalities have been explored, but with the technology available, they remain unproven. Ovarian cancer screening with ultrasound and serology did not diminish mortality from that dreaded disease. Prostate cancer screening with PSA may well represent a near-neutral balance of risk and benefit. Skin, pancreatic and esophageal cancers have not yielded their biologic destiny to being plucked out just in time. Essentially, by the time we can see them, it is too late for meaningful intervention.

Much research is ongoing. We need version 2.0 of the PSA test to exclude indolent disease not requiring surgery or radiation therapy. Thus we could use those tools only on cancers that are destined to progress.

If used correctly, streamlined cancer screening recommendations free up many health care personnel. Cervical cancer screening is a beautiful example of a win-win situation. Dropping down the frequency of testing to every three years and stopping at 65 years of age means primary care physicians, gynecologists and cytologists can address other health issues.

 Cancer screening is an exciting subject. It is full of possibilities. Great technical challenges remain that we hope will lend themselves to efficacious solutions in the future.

 Reference

1. USPSTF, “Recommendations for Adults: Cancer,” www.uspreventiveservicestaskforce.org/adultrec.htm#cancer (2011).

---

Dr. Pont, an internist, is an assistant physician in chief at Kaiser San Rafael.

Email: joan.pont@kp.org