van Leeuwen et al fail to explain why the prostate-specific antigen (PSA) by itself is misleading and how physicians can gather better data to achieve better outcomes with fewer side effects.1 A person may have a high prostate-specific antigen (PSA) level due to an enlarged prostate gland (which is quite common), irritation, inflammation, infection, or other changes of the prostate gland. Thus, a PSA level of 10 ng/mL may not necessarily require biopsies or aggressive therapy. I illustrate with 1 example so physicians will have a better view of likely noncancer variability (a subject that in my opinion is poorly understood and omitted in most studies). For over 5 years, patient PP had a PSA level between 9 and 11 ng/mL. He had a negative biopsy the first time his PSA level was found to be elevated. His prostate gland was substantially enlarged (but no exact measures were available), with the usual consequences. One day his PSA level was 21 ng/mL. The patient changed his diet and behavior drastically (losing weight, exercising more, eating healthier foods, and other behaviors). Within a few weeks, the PSA level declined to 15 and then 11 ng/mL. For the past 3 years, the patient's PSA level has remained between 9 and 11 ng/mL. When the PSA level was highly elevated, 1 of the top prostate cancer experts in the United States insisted that a biopsy be performed immediately, without waiting for repeated testing or the effects of the drastic diet and behavior changes proposed by the patient. The patient refused a biopsy. Physical examination of the prostate for the past 3 years has found no nodules. After having a PSA level of 10 ng/mL for >10 years, and 3 years after having a PSA level > 20 ng/mL, no cancer has been detected to date. PSA values of approximately 10 ng/mL are not necessarily indicative of a high probability of cancer, nor deserve biopsies. Although the relevant factors are not known, it is possible for a 10-point variation in PSA to be the result of factors other than cancer, which are correctable via diet, behavior modification, antibiotics (for infections), and other treatment models (different from cancer treatment). We need far better data regarding the probability of cancer given PSA level, prostate size, body size, prostate status (inflamed, infected, irritated, etc), types of diet, and other factors. We also need a probability distribution of the variability of PSA values over relatively short time periods (measured in months). In the interim, physicians should carefully evaluate the patient and consider alternative treatments (such as substantial weight loss and a healthier diet) before proceeding with a biopsy. Biopsies do have undesirable side effects, and could lead to unnecessary treatment, again with undesirable side effects. Eduardo Siguel MD, PhD*, * Gaithersburg, Maryland.
Read full abstract