Association between gout and subsequent breast cancer: a retrospective cohort study including 67,598 primary care patients in Germany

Breast Cancer (BC) is the most common cancer type in women worldwide, with an estimated 2.1 million newly diagnosed cases in 2018 [1]. In Germany, approximately 68,950 women were diagnosed with BC in 2016 and 18,570 women died of the disease [2]. The number of BC cases worldwide is expected to increase to 2.7 million annually by 2030 [3‐5]. Notably it has been estimated that about 20% of all BC-related deaths can be attributed to modifiable risk factors such as alcohol use, obesity (BMI ≥ 30), and reduced physical activity [6]. For example, obesity is not only linked with increased BC incidence, but is also associated with a poor outcome in terms of increased BC recurrence and reduced survival compared to normal-weight BC patients [7‐9]. Recent studies support the hypothesis that obesity-related inflammation is a major contributor to this association [10]. In particular, it has been well documented that systemic inflammation as a hallmark of cancer plays a crucial role in breast carcinogenesis and progression [11]. It is also hypothesized that physical activity lowers BC risk in postmenopausal women by lowering systemic inflammatory markers, including TNF-α, IL-6, and C-reactive protein (CRP) [12, 13].

In view of the above, the association between BC and gout as a common inflammatory disease has attracted much interest during the past few years. Briefly, gout is the most common form of inflammatory arthritis worldwide. Despite being one of just a few curable rheumatic diseases, both the prevalence and incidence of gout are increasing across the globe so that it is being hypothesized in the literature that a modern gout epidemic is currently developing that is on a similar scale to the obesity epidemic [14, 15]. This increasing disease burden of gout is a consequence of lifestyle changes, obesity and dietary factors. However, genetic polymorphisms are also responsible for the development of gout and hyperuricemia [16‐18]. Mechanistically, the pathogenesis of gout can be divided into different phases: The initial development of hyperuricemia (serum urate > 6.8 mg/dl), followed by the deposition of monosodium urate crystals (MSU) usually in tissues with low temperature or acidic pH, such as joints and periarticular structures [19]. Finally, MSU crystal deposition can lead to an acute gout flare, particularly at the first metatarsophalangeal joint, the midfoot, and the knee with the rapid onset of painful acute inflammatory arthritis characterized by a swollen, hot, and red joint [20, 21]. Notably, a gout flare is a clinical diagnosis, but the diagnosis is also generally considered to be accurate when made in a primary care setting using well-validated diagnostic criteria [22].

In the past, Strasak et al. showed in a prospective study including more than 28,000 older Austrian women (aged ≥ 50 years) that hyperuricemia (> 5.41 mg/dL) was independently associated with an increased risk of total cancer mortality (HR: 1.27, 95% CI 1.08–1.48). In particular, malignant neoplasms of the breast and female genital organs were positively associated with hyperuricemia [23]. In line with these results, a metanalysis of three different prospective cohort studies with a total of 50,358 patients revealed that gout patients were at an increased risk of cancer, particularly urological cancers, digestive system cancers, and lung cancer. However, due to several study limitations of these works (e.g., limited sample size), the association between gout and breast cancer remains unclear [24‐27].

Aiming to explore this topic in more detail, we conducted a retrospective cohort study including 67,598 primary care patients in Germany to investigate the cumulative incidence of breast cancer as a function of gout and to explore the association between gout and subsequent breast cancer using univariable Cox regression analysis.

Our study revealed differences in the cumulative breast cancer incidence between the gout and non-gout cohorts. The Cox regression analysis indicated a significant association between gout and subsequent breast cancer in the overall population. In line with our results, Strasak and colleagues reported in a prospective study that high serum uric acid (SUA) levels (> 5.41 mg/dL) were independently associated with an increased risk of total cancer mortality (p < 0.0001). In particular, the authors found an independent, positive association with deaths due to malignant neoplasms of the breast and female genital organs [23]. In addition, Levine and colleagues reported a positive association between SUA levels and total cancer mortality among females aged 55–64 years, even after adjustment for multiple risk factors (e.g., age, relative weight, smoking) [34]. Despite being the main characteristic of gout, hyperuricemia is only one factor of many that contribute to the transition from hyperuricemia to clinically evident gout [16]. However, it has been shown that patients with established gout also have an increased risk of cancer [25‐27, 35]. In particular, a national population study from Taiwan showed that the annual incidence of cancer in gout patients was more than twice as high as in the normal population (8.7 vs. 4.2 cases per 1000 patient-years, p < 0.001). Gout was most closely associated with prostate cancer, with an age- and sex-adjusted HR of 1.71 (1.45–2.02) [25]. Another retrospective cohort study from Taiwan revealed that gout patients are more likely to develop urological cancers such as prostate, bladder, and renal cancers in particular [26]. Finally, Boffetta and colleagues showed in a large prospective study of 16,857 gout patients admitted to hospitals in Sweden that the incidence of cancers of the oral cavity and pharynx, colon, liver and biliary tract, pancreas, lung, skin (melanoma and nonmelanoma), endometrium, and kidney was higher among gout patients [27]. In summary, gout is associated with a broad spectrum of different tumor entities, meaning that the prevalent type of cancer in gout patients could be dependent on culture, vary between countries, and may also be influenced by tumor site-specific risk factors [36].

In the past, the literature has indicated that SUA exhibits antioxidant properties by scavenging reactive oxygen species (ROS) and, therefore, protects against cancer [37]. However, today there is increasing evidence that SUA plays a crucial role as a pro-oxidant by promoting inflammatory reactions and oxidative stress, thereby contributing to cancer development. Focusing on the pro-tumorigenic role of SUA in breast cancer progression, it is hypothesized that SUA might be partially responsible for the low-grade inflammation in the breast tumor microenvironment that contributes to tumor cell proliferation and metastasis [38]. Contrary to this, a prospective population-based study revealed that SUA levels were significantly inversely associated with breast cancer risk and overall cancer mortality, but not with the risk of lung, prostate, and colorectal cancer. However, if the antioxidant potential of uric acid is an underlying cause of the observed associations between SUA and breast cancer and cancer mortality, it would be reasonable to expect that such a universal mechanism might also cause similar associations with other cancer types and not just with breast cancer [39].

Finally, the age-stratified analyses in our study revealed that gout was only strongly associated with subsequent breast cancer in the age group ≤ 50 (HR: 1.58; 95% CI: 1.10–2.27). Interestingly, a national population-based study from Korea found a higher risk of cancer, as well as both all-cause and cancer mortality, in middle-aged patients (41–55 years) with gout compared to the general population. The subgroup analysis of this study showed that the risk of stomach cancer, head, and neck cancer and hematologic or lymphoid cancers was higher in middle-aged gout patients than in controls, whereby the risk of BC was not calculated. Notably, these results are in line with our observation that middle-aged gout patients have the highest cancer risk compared to controls [36].

Based on the fact that the median age at menopause among women from industrialized countries ranges between 50 and 52 years [40], one can assume that the menopause status may explains the exclusive significant association of gout and BC in women aged < 50. Indeed, it is known that postmenopausal women experience a higher occurrence of gout, while estrogen's uricosuric effect shields premenopausal women [41]. Given the positive association between hyperuricemia and higher cancer incidence (incl. BC), these results suggest that BC should be more prevalent in gout patients aged > 50 [24]. However, we observed the opposed effect. Notably, conflicting results have been reported in other studies focusing on gout incidence and menopausal status. According to an analysis of National Health and Nutrition Examination Survey data from 1999 to 2010, the occurrence of hyperuricemia was not associated with menopause [42]. Finally, further research is required to examine how hyperuricemia and gout are linked to breast cancer, considering factors such as age, inflammation, and menopausal status.

Our retrospective cohort study has several strengths: The german disease analyzer (DA) is a large European outpatient database containing data from 2898 practices with about 7.8 million patients in Germany. The representativeness of the diagnoses it contains has already been validated [43, 44]. Furthermore, DA provides continuously updated data generated directly from practice computers based on patient data (diagnoses, demographic data, prescriptions, etc.) and has successfully been used for several studies in various disciplines [43, 45‐47]. Notably, we performed propensity score matching based on diabetes and obesity as it is well known that both are strongly related to breast cancer [48]. We also included the use of diuretics in the matching process based on their tendency to induce hyperuricemia which can in turn lead to the development of gout. Furthermore, patients diagnosed with cancer prior to the index date were excluded since hyperuricemia (due to increased urate overproduction) can occur in patients with disorders involving high cell turnover such as advanced tumor diseases or highly proliferative myeloproliferative disorders [49, 50]. However, the DA does not contain any information on external confounding factors (e.g., alcohol and socioeconomic status). Importantly, alcohol consumption and lifestyle factors are important risk factors for developing gout and no information on them was available for this study [15]. Also, SUA levels are not recorded in the DA database and the association between hyperuricemia and BC risk, therefore, could not be evaluated. Finally, there is also a lack of hospital data and information on mortality.

In conclusion, our study provides evidence for the association between gout and subsequent breast cancer diagnosis, particularly in the youngest age group.