Seasonality of Serum Prostate-Specific Antigen Levels and Its Impact on Prostate Cancer Screening in Hong Kong

Abstract

Serum prostate specific antigen (PSA) test is commonly used as a tool for screening and early detection of prostate cancer. Many factors can affect its sensitivity and specificity as a screening test of prostate cancer. Its efficacy has been much debated ever since its introduction for this purpose. There is a need to better understand the confounding factors, among them seasonal variation in serum PSA levels, in order to make the best use of this test. By understanding the normal seasonal variation in serum PSA in a specific population, the cutoff point for normality may need to be adjusted according to the time of the year in order to maintain the efficacy of PSA test as a tool for screening early prostate cancer.

The aims of this study were to investigate the occurrence of seasonal variations of serum PSA levels and the effects they might exert on the usage of PSA for prostate cancer screening in Hong Kong. Total PSA (tPSA) and free PSA (fPSA) data were collected from two major laboratories in Hong Kong. In order to safeguard patient confidentiality, no personal identifying information was collected. Data collected were date of test sample collection, client age, tPSA and fPSA levels, PSA measurement method and clinical history of prostate cancer, if available. Three meteorological parameters, namely daily mean air temperature, total bright sunshine duration, and global solar radiation, covering the test sample collection period, were obtained from the Hong Kong Observatory official web site.

A total of 15,863 tPSA and 491 fPSA test results from 31st Oct 2005 to 31st Dec 2007 were collected. The age of clients has a mean of 58.4 year and ranges from 13 to 102 year. The tPSA and fPSA values were correlated with moving averages of the meteorological parameters. Although the results varied somewhat between the two laboratories, overall the tPSA results were significantly correlated with the three meteorological parameters. The strongest correlation was with daily mean air temperature. No correlation was shown between the meteorological parameters with either fPSA or percent fPSA (%fPSA). tPSA tests performed during summer periods were more likely to have a value of >4 ng/mL (chi-square test, p < 0.001), which is the range of test results for which prostate biopsies would be recommended. It was estimated that there would be at least a 16% increase in abnormal tPSA values for tests performed between summers should these tests be ordered during the summer months instead. The study found no variations of fPSA and %fPSA with seasonal change.

In conclusion, due to seasonal variations and other factors that may affect serum PSA levels, a single abnormal tPSA result should be interpreted with caution. A marginally elevated tPSA value, especially when obtained during summer time, should be repeated and confirmed before biopsies are recommended.