Over the years, a number of so-called tumor markers have been identified for various cancers. Tumor markers are proteins which circulate in the blood and are elevated in the presence of certain cancers. These markers tend to be identified with specific cancer types—prostate specific antigen (PSA) for prostate cancer, carcinoembryonic antigen (CEA) for colorectal cancer, alpha-fetoprotein (AFP) for hepatocellular carcinoma or testicular cancer, CA19-9 for cancers of the pancreaticobiliary tree.
These markers are elevated to varying degrees in the presence of tumor and can be used to assist in diagnosing the cancer at presentation. For instance, a patient who is seen with back pain and bone metastases who has an elevated PSA is very likely to have prostate cancer. Their other major use is for following the clinical course of a patient under treatment—either to determine if a person who is supposed to be disease-free has had a recurrence or to see if a person who is on treatment for advanced disease is responding to the chemotherapy he/she is receiving (in which case the marker should decrease).
Many efforts have been made to utilize these markers for the purposes of early detection or screening, i.e., to perform these blood tests on healthy asymptomatic average-risk individuals in order to detect and treat cancer early and thereby reduce mortality. For the most part, with the possible exception of PSA and prostate cancer, these attempts have failed. The main reason seems to be that these tests are not sensitive or specific enough to be utilized for the purposes of screening. They lead to too many inaccuracies when used for screening and overwhelm the screening programs with false positives and false negatives. Many of these markers have had randomized trials to evaluate their potential utility for screening and have not proved to be effective in reducing cancer-related mortality to a sufficient degree.
A good example is CA125, a tumor marker discovered in 1981, that is elevated in patients with ovarian cancer. There are over 20,000 new ovarian cancer cases annually in the U.S., of whom tragically, over half eventually die from their disease. As with many other cancers, those with earlier stage disease do better, and multiple efforts have gone into detecting ovarian cancer earlier, including simply doing frequent palpation of the ovaries on pelvic exam. This has proven futile.
Serum CA125 levels are increased in those with more advanced ovarian cancer and more advanced stages. Consequently, serum CA125 has been tested a number of times as a screening tool for those at average risk for this disease. The results have generally been equivocal and it has also led to increased numbers of false positives, which has diminished its potential benefits. To enhance its potential utility as a screening tool, it has been paired with transvaginal ultrasound, another potential tool for detecting ovarian cancer early.
The ultimate evaluation of these tests for screening purposes came as part of a large screening trial known as the PLCO study (Prostate Lung Colorectal Ovarian), a randomized trial of about 160,000 people (approximately half men and half women) to test four different screening modalities at once in a cost-efficient fashion. We described one of these studies in last week’s article, chest X-rays, the Lung in PLCO. Obviously, this week we refer to the Ovarian part of the PLCO trial, the attempt to test screening for ovarian cancer.
The trial, published in 2011 in the Journal of the American Medical Association, recruited women between 1993 and 2001. There were ultimately 39,106 in the experimental arm and 39,111 in the control arm. The experimental group were offered serum CA125 screening annually for six years and transvaginal ultrasound for four years. The median follow-up time was 12.4 years. By the end of that time, there were 212 ovarian cancers in the experimental group diagnosed and 176 in the control group. Importantly, there was no stage shift in the experimental group, i.e., the stage distribution in both groups was similar. Most importantly, there were 118 ovarian cancer deaths in the experimental group and 100 in the control group. (Why there were more deaths in the screened group as compared to the unscreened group was noted but unexplained by the investigators.) Screening naturally led to more false positives in the experimental group and therefore there were 1,771 oophorectomies in the experimental group versus 1,304 in the control group.
In summary, there was no benefit in terms of ovarian cancer mortality from screening and, in addition, the screening led to significantly more unnecessary surgery to the screenees because of the necessity to work up and evaluate women who had abnormal blood or ultrasound tests. This is a common, but little appreciated, downside of most screening tests. The end result is that screening is not recommended for women at average risk for ovarian cancer, though recommendations may be different for those at increased risk for this disease, such as those with BRCA mutations.
Alfred I. Neugut, MD, PhD, is a medical oncologist and cancer epidemiologist at Columbia University Irving Medical Center/New York Presbyterian and Mailman School of Public Health in New York.
This article is for educational purposes only and is not intended to be a substitute for professional medical advice, diagnosis, or treatment, and does not constitute medical or other professional advice. Always seek the advice of your qualified health provider with any questions you may have regarding a medical condition or treatment.
