Female sex is associated with a lower risk of bone metastases and favourable prognosis in non-sex-specific cancers

As previously reported, bone is the third most common metastatic site for malignant cancer [1, 2]. Bone metastases (BM) can cause a series of skeletal-related events (SREs), including pain, bone fractures, spinal cord compression and hypercalcemia [3, 4]. The incidence of SRE has been reported in several studies, the cumulative SRE incidences were respectively reported to be 47, 31.4, and 38.0% in breast cancer, prostate cancer and multiple myeloma patients [5]. Another study reported SRE in 26% patients with prostate cancer, 70% in renal cell carcinoma, and 58% in urothelial carcinoma [6]. Both BM and SRE were thought to negatively impact patient survival. The number of bone metastasis was significantly correlated with the mortality of patients [7, 8]. The latest study showed that the 1-year survival rate of cancer patients with BM ranged from 10% in lung cancer patients to 51% in breast cancer patients [2]. Bone-modifying agents were accepted to significantly retard the first occurrence of SRE [6, 9]. Bisphosphonate was reported to be correlated with the better outcome in patients with bone metastasis [10, 11]. Treatments that prevent SREs have been reported to significantly improve daily function [12].

Accurate estimation of BM is crucial for the prevention and treatment of BM. Previous studies have presented an inconsistent prevalence of BM, which is influenced by various factors [13‐17]. It was reported that male sex was one of the independent risk factors for the development of BM [18]. Several studies revealed fewer BM in female patients than in male patients [19‐21]. In addition, few studies have investigated the disparity in prognosis between the sexes among patients with BM [22, 23]. The absent consensus regarding the influence of sex on BM was thought to be caused by the small sample size and different characteristics of cancers [17]. Thus, a further study based on a large population is warranted to examine the correlation between sex and the occurrence and prognosis of BM in various cancers.

The National Cancer Institute’s Surveillance, Epidemiology, and End Results (SEER) program was established in 1973 and comprises approximately 30% of the total US population; it is an important data source for epidemiologic analyses. In this study, using the records in the SEER datasets, we aimed to investigate the disparity in the prevalence of BM between the sexes for different non-sex-specific cancers and to evaluate the effect of the female sex on the occurrence and prognosis of BM.

To our knowledge, the present study was the first to systematically examine the disparity of the occurrence and prognosis of BM among the sexes in millions of patients with non-sex-specific cancers. The prevalence of BM in different cancers ranged from 0.2% (brain) to 24.4% (miscellaneous), and the pooled prevalence of BM for all patients was 2.0% (95% CI: 1.4–2.9%). The current study showed that the trend in the prevalence of BM dramatically decreased for ages between 11 and 40 years old, plateaued for ages between 41 and 90 years old and increased for ages > 90 years old in both male and female patients.

On the other hand, the present study showed that female patients had a lower prevalence of BM than male patients, and the female-to-male pooled prevalence ratio was consistent across different age groups. Compared with male sex, female sex had a protective effect on the development of BM as suggested by meta-analysis. A series of clinical studies suggested that the occurrence of BM was more prevalent in male patients, which was in line with our results [18, 21]. One of the potential explanations for the results may be that sex hormones have different influences on the development of BM.

Sex hormones have been accepted as key factors in musculoskeletal health for both males and females [24]. Both androgens and oestrogens can affect the proliferation of osteoblasts and osteoclasts. In both males and females, oestrogens were reported to directly inhibit osteoclasts and exert an effect on the maintenance of bone mass [24, 25]. Moreover, androgens are thought to contribute directly to male periosteal bone expansion, mineralization, and trabecular bone maintenance, which are important in the pathogenesis of BM [26, 27]. The latest review suggested that hormonal status affected the occurrence of BM [28]. In our study, the relationship between the occurrence of BM and age was in contrast with the tendency of human sex hormone levels to change with age [29]. Thus, we proposed the hypothesis that female hormones play a significant protective role in regulating BM in non-sex-specific cancer patients. Oestrogens were reported to be able to reverse the inhibitory effects of osteoblasts on osteogenic differentiation through regulation of the RANKL–osteoprotegerin pathway [30, 31], which further supports our hypothesis.

In the present study, compared with male sex, female sex was a favourable prognostic factor for patients with respiratory system cancers, which was consistent with the results of previous studies [32, 33]. Positive expression of the hormonal receptor such as ER-α was reported to be one of the favourable prognostic factors for lung cancer patients. Furthermore, increased ER-α expression was reported in the lung tissue of female patients [23, 34], which may explain our results indicating that female sex is a favourable prognostic factor for respiratory system cancer patients. At the same time, a previous report suggested that the better prognosis of female lung cancer patients could also be attributed to the better response to EGFR inhibitor treatment for females [35, 36].

The aforementioned findings suggest that male cancer patients and those with male-leaning genetic variations or hormonal status have higher odds of developing BM. Second, cancer patients with high male sex hormone levels at initial diagnosis can potentially be selected as candidates for screening for BM. Third, hormone therapy could be a potential therapeutic strategy for non-sex-specific cancer patients with BM. More studies will be needed to investigate the mechanism of the protective effect of sex hormones on the development and prognosis of BM. The use of hormonotherapy as a treatment option for non-sex-specific cancer should also be confirmed in biological research and clinical trials. All the possible findings may improve the prevention, screening and treatment of BM in patients with cancer.

There are some limitations in the present study. First, the diagnostic approach for initial BM among cancer patients was not recorded by the SEER database, and the asymptomatic cases and the patients who developed BM later during the disease course were not recorded in the database. Accordingly, the prevalence of BM may have been underestimated to some extent, and more studies are needed to further confirm the results. Second, there were significant heterogeneities in the meta-analyses. Although the random effect model could partly solve the problem, the combined results were relatively conservative, and the results should be interpreted with caution. In addition, skeletal-related events were not recorded in the SEER database, resulting in difficulty in evaluating their influence on survival and quality of life. All these weaknesses should be improved upon in future studies.

In summary, the prevalence of BM in both sexes inversely fluctuated with changes in age. Among non-sex-specific cancer patients, compared with male sex, female sex was associated with less frequent development of BM, and the trend was consistent with changes in age. Female sex exclusively showed a protective effect on the prognosis of respiratory system cancers. All the findings in the present study will provide useful guidelines for screening for BM and prediction of survival. More studies should be conducted to investigate the underlying mechanisms of hormone disparities and to discover more effective treatment methods from the view of hormone disparities.