Discussion
The focus of most previous studies was on evaluation of the correlation between breast cancer and melatonin levels while there is quite limited number of studies examining the correlation between melatonin and prostate cancer. A cross-sectional study carried out by Bartsch et al. revealed that there were lower melatonin levels among the men who had prostate cancer than that in men with BPH21. In a case-cohort study in the Icelandic population, it was found that subjects whose first morning urinary 6-sulfatoxymelatonin (aMT6s) levels were below the median level were exposed to a 4-fold higher risk than the subjects with aMT6s levels above the median level (hazard ratio = 4.04; 95% CI = 1.26-12.98)22. On the contrary, in the present study, no significant correlation was found between morning urinary aMT6 levels and total risk of prostate cancer. In fact, in our study, there was a statistically significant negative correlation between the melatonin/cortisol (M/C) ratio and first-morning urinary melatonin-sulfate levels and the risk of advanced (end-stage) prostate cancer and general prostate cancer. In our study, we strongly believe that our results will contribute to the literature on prostate cancer risk and the circadian hormones. Melatonin had the protective effect on risk of cancer due to inhibiting growth of cancer cell, which protects cells against damage of DNA, and promotes DNA repair after it occurs1,23,24. Blask et al. recently used both steroid receptor-negative and receptor-positive human breast cancer xenografts in rats to perform a series of experiments and to discover an inverse correlation between tumor activity and melatonin level25. Similar results with prostate cancer xenografts were reported in similar research group 26. In other studies, a decrease in the malignant prostate tumor cells growth has been reported with the administration of both pharmacological and physiological doses of melatonin27.
In our study, we further measured another important circadian hormone i.e. cortisol, secreted by the adrenal cortex. Cortisol has been found to regulate both inflammation and immunity. This hormone deficiency can disrupt the immune system, and immune reactions may be suppressed by excessive hormone 28. Furthermore, there has been association between the chronic irregularity of rhythm of the circadian cortisol and increasing inflammation which is highly effective in carcinogenesis28,29. Mirick et al. found the effect of circadian disruption on the pattern of cortisol secretion release, increasing the risk of cancer30. Even though the present study found no significant correlation between prostate cancer and spot morning urinary cortisol level, we found an inverse correlation between advanced (end-stage) prostate cancer and prostate cancer and the M/C ratio. In the previous studies, the M/C ratio was related to the depression severity and different types of depression31,32. However, as far as we know, our study is one of the few studies evaluating the correlation between prostate cancer and the M/C ratio.
Whether to undergo an invasive prostate biopsy for subjects with abnormal PSA levels can be clinically determined with difficulty. The results of our study which support the conclusion that combined low M/C ratios and PSA levels exceeding 10 ng/mL indicates that a person is exposed to advanced (end-stage) prostate cancer and prostate cancer, suggesting that the M/C ratio should be considered a biomarker. The urine M/C ratio will be additionally used to decide if a prostate biopsy is necessary with increasing PSA levels. More studies are needed in order to examine the clinical utility of evaluation by combining PSA and M/C ratio for detecting the prostate cancer and the disease stage.
In our study, the male control group also had a high PSA level exceeding 10 ng/mL. The serum of patients with prostate cancer and BPH was expected to have higher concentrations of PSA 33. An estimated 50% of 50-year-old men, 75% of 80-year-olds, and 90% of 85-year-old men have histological evidence of BPH34. Men aged above 50 had about 99% of prostate cancer35. This is particularly problematic in older men with more prevalent BPH due to increase of glandular volume with BPH so that the PSA level and the sampling errors number related to prostate biopsy is enhanced34. While prostate cancer detection rates in Caucasian men can be as high as 40% for men whose PSA is 4-10 ng/mL, prostate cancer is observed in only 20% of Chinese men with the same level of PSA36. In our control group, the inclusion criteria include either remarkable or unremarkable digital rectal exams but with BPH confirmed histologically. For this reason, some male patients with undiagnosed prostate cancer may have been included in our control group. However, the inclusion of such patients in the control group would only lead to an underestimation of the correlation that is happening.
One of the strengths of our study is that we evaluated the most important risk factors of prostate cancer in our analyses, including PSA level and clinical stage. Furthermore, two important circadian biomarkers were evaluated and then the M/C ratio was discovered to be more relevant for prostate cancer. There are some limitations in the present study. The first limitation is its cross-sectional and case-control nature. For this reason, a precise causal relationship cannot be inferred. If sleep interruption or circadian disruption after diagnosis of a cancer causes melatonin levels to drop, there is a possibility of inverse causation between prostate cancer and circadian hormones. A further limitation is our use of one spot morning urine biomarkers single measurement not representing long-term exposure levels. Even though we have extensive knowledge of various covariates and can easily control possible confounders, we still did not know about factors such as the presence of sleep disorders or drug use. As a consequence, the related exposures and risks were gathered with a questionnaire, leading to some recall bias.