Background
Ovarian cancer (OC) is the fifth leading cause of cancer death among women after lung, breast, colorectal, and pancreatic cancer [
1]. OC is highly treatable when diagnosed in the early stages; however, the majority of cases are diagnosed in late stage and the 5-year survival is approximately 27% [
1]. In order to detect early stages of OC, there is a need for a better understanding of the etiology of this disease as well as improved screening options. Currently, there are three screening procedures in use for OC detection: bimanual pelvic examination, serum CA125, and transvaginal ultrasonography (TVS) [
2]. Pelvic examinations are not effective in distinguishing a premalignant lesion from a normal ovary [
3]. Serum CA125 is elevated in 47% of women with early stage OC and elevated in 80-90% of advanced stage OCs [
4]. However, CA125 can also be elevated in healthy women and in patients with benign ovarian disease [
5]; as well as, in other cancers such as endometrial, pancreatic, lung, breast, colorectal, and certain gastrointestinal tumors [
6]. The pairing of TVS with serum CA125 improves OC diagnosis; however, sonography can result in false positive results and unnecessary surgery [
7]. Unfortunately, utilizing these methods either alone or in combination does not produce the desired results for early disease detection. Therefore, examining the molecular changes that occur at early disease onset may provide new approaches or biomarkers for OC detection.
Deregulation of inflammation due to overexpression of proinflammatory proteins contributes to the malignant phenotype by supporting cancer cell growth and tumor invasion. Since the proinflammatory cytokine, interleukin-1β (IL-1β), is constitutively expressed in OC [
8] and elevated in serum of OC patients [
9], in this pilot study, we sought to assess whether elevated urinary levels of IL-1β are associated with OC and related clinical parameters.
Discussion
While inflammation is an essential biological process for normal development and tissue homeostasis, it is also involved in a number of pathologic conditions including tissue injury, chronic inflammation, immunological diseases, and cancer [
11]. Epidemiological studies have shown a link between chronic inflammation and risk for cancer as evidenced by prolonged infection with
Helicobacter pylori and gastric cancer, inflammatory bowel disease and colon cancer, and prostatitis and prostate cancer [
12]. In the ovary, chronic inflammation resulting from repeated ovulatory wounding and repair promotes oxidative stress which enhances DNA replication errors and ultimately, oncogenesis [
13]. Inflammation is regulated by several factors that can either promote or inhibit inflammation and since epithelial OCs are highly inflammatory, this pilot study evaluated urinary levels of the proinflammatory mediator, IL-1β against clinical parameters in order to gain a better understanding of this disease. This study was able to evaluate four clinical parameters in relation to urinary IL-1β levels: (1) patient age, (2) diagnosis, (3) family history of cancer, and (4) BMI.
When we considered patient age at the time of sample collection, the highest levels of urinary IL-1β levels were found in the 20–29 and 30–39 years age groups followed by declining IL-1β levels as age increased. In keeping with our findings and according to the American Society for Reproductive Medicine, a woman's potential reproductive capacity begins to gradually decline at approximately 32 years of age and more rapidly decreases after 37 years of age [
14]. IL-1β has been suggested to play a role in female reproduction; specifically in ovulation and oocyte maturation, and inflammatory-linked mechanisms, such as production and activation of proteolytic enzymes, prostaglandin production, nitric oxide production, cellular metabolism, and steroidogenesis [
15],[
16]. Therefore, it seems likely that after 40 years of age, there would be a decrease in urinary IL-1β levels as oocyte maturation and ovulation decrease in preparation for menopause. In contrast, Vural, et al. found higher plasma levels of IL-1β in postmenopausal (≥48.6 years) women than in premenopausal (30.5 ± 2.5 years) women with levels of IL-1β decreasing below the premenopausal levels only after hormone replacement therapy [
17]. Therefore, it is possible that the sharp decrease in urinary IL-1β levels seen in our study in women >60 years may be due, in part, to the usage of hormone replacement therapy; however, this clinical information was not available for confirmation.
Urinary IL-1β levels alone had limited success in differentiating disease status. IL-1β is present in the serum and ascites of OC patients [
9] and has been shown to be involved with cancer tumorigenesis, angiogenesis, and metastasis [
18]. The inability of urinary IL-1β to differentiate between benign and malignancy may be confounded by the inflammatory nature of so many benign and cancer conditions. That is, many benign ovarian conditions develop in an inflammatory microenvironment. For instance, endometrioma, a form of endometriosis in the ovary, is a highly inflammatory condition [
19] which would expectedly result in high levels of urinary IL-1β [
20]. Proinflammatory markers including serum C-reactive protein, IL-6, and IL-8 have all been used in clinical studies in an attempt to differentiate between normal, benign tumor, and OC [
21]-[
23]. Immunohistochemical analyses have shown differential expression of IL-18 and its receptors in benign ovarian tumors, borderline ovarian tumors, and ovarian carcinomas [
24]. Therefore, developing and employing a panel of inflammatory mediators, including urinary IL-1β, may eventually benefit differential diagnostic and prognostic outcomes of OC.
Patient samples with confirmed family history of cancer were limited in this small pilot study. Nonetheless, urinary IL-1β levels tended to be highest in patients with benign ovarian disease and a (first degree) family history of breast and/or ovarian cancer compared to patients with benign ovarian disease, but without a family history of breast and/or ovarian cancer. Likewise, elevated levels of urinary IL-1β were found among OC patients with a family history of breast and/or ovarian cancer compared to OC patients from families without a family history of disease. This supports the recent dualistic model that epithelial OCs arise as either two types: Type 1 and Type 2 [
25]. Type 1 tumors may arise in a step-wise progression from a benign precursor lesion such as, the highly inflammatory condition, endometriosis. However, it is important to remember that not all individuals at risk for OC develop the disease so that secondary events, perhaps beyond a family history, may be necessary to promote disease.
Obesity as a risk factor of OC remains controversial. Recently, a meta-analysis of 47 epidemiological studies found increased OC risk with high BMI [
26]. The Ovarian Cancer Association Consortium investigated 15 case–control studies and found overweight and obese women were associated with increased risk of OC [
27]. The National Institutes of Health also found that BMI was significantly associated with increased OC risk in women who never used hormone therapy [
28]. Canchola, et al. found a positive association between OC risk and adult weight gain, waist circumference, and waist-to-hip ratio, but no association to overall obesity as classified by BMI [
29]. In agreement, Delort, et. al. also noted high waist-to-hip ratio associated with increased risk of OC though they found no association with BMI [
30]. In contrast, Schouten, et al. reported no overall association between BMI and risk of OC; however, they did report a positive association with high BMI and increased OC risk among premenopausal women [
31]. More recent prospective studies reported no significant relationship between BMI and OC risk, irrespective of menopausal status [
32],[
33]. Likewise, evidence reported by McGee, et. al. does not support a risk for OC with weight or weight gain among
BRCA1 or
BRCA2 mutation carriers [
34]. Interestingly, Engleland et al. also reported that the risk for OC was not associated with adult BMI, but suggested a possible increased risk in women who were obese in young adulthood [
35].
In this pilot study, one of the most apparent clinical features related to elevated urinary IL-1β was BMI. We found increased urinary IL-1β levels associated with higher BMI. Overweight and obese patients were most likely to be diagnosed with OC and ovarian benign disorders, respectively, while concomitantly demonstrating the highest average urinary IL-1β levels. In contrast, healthy controls with normal BMI exhibited the lowest average urinary IL-1β. Among our data was a single patient case classified as underweight, but with elevated urinary IL-1β levels and a diagnosis of OC. It is tempting to speculate that this individual may have had cachexia at the time of sample collection where weight loss due to the increased glucose, lipid, and protein requirements of the tumor [
36] could manifest as low BMI compounded with elevated urinary IL-1β as a result of advanced disease.
Obesity and elevated IL-1β levels in OC patients may contribute to OC mortality. While some studies have shown no association between BMI five to ten years prior to OC diagnosis and OC mortality [
37],[
38], they do suggest that obesity is associated with poor outcome [
39]. Consequently, obesity itself may not be the leading factor for increased OC mortality, but may act as a comorbidity factor. For instance, difficulty of proper chemotherapy dosages for obese patients may contribute to poor disease outcome. A study of dosing practices of clinicians found that a significant proportion of OC patients with advanced disease were overweight or obese, as seen in the current study, and under-dosing in obese populations was common [
40]. The variability in dosing when prescribing chemotherapy is largely due to concern for potential over-dosing and chemotherapy associated toxicities [
41]. High mobility group A2 (HMGA2) is a protein that can regulate transcription by altering chromatin architecture and facilitate the assembly of multiprotein complexes of transcriptional factors [
42]. It is overexpressed in serous OC tumors, but not in normal ovarian epithelial cells [
43]. OCs with high expression of HMGA2 and high BMI negatively affected overall survival [
44].
Lastly, obesity may increase tumor aggressiveness. Increased metabolic activity and glucose concentrations driven by the Warburg effect [
45] are associated with highly aggressive OC cell lines [
46]. In cancer, the Warburg effect is regarded as a characteristic metabolic process that may contribute to cell survival in a stressful environment, such as the stress of chronic inflammation [
47]. The Warburg effect suggests that cancer cells produce energy predominately by glycolysis and lactic acid production over oxidative phosphorylation [
45]. An in vivo obese mouse model demonstrated increased tumor size and tumors in these obese mice had a unique molecular makeup noted by upregulation of inflammation genes [
48]. Obesity in OC patients may further exacerbate disease by contributing to an inflammatory environment. Obesity-related type 2 diabetes is associated with chronic inflammation [
49]-[
51] and IL-1β levels have been shown to be correlated with obesity and obesity related disorders. Individuals with combined elevated plasma levels of IL-1β and IL-6 are at increased risk for developing type 2 diabetes [
52], but even mild weight loss in obese patients resulted in a 45% decrease in serum IL-1β levels over a three-year study period [
53]. Leptin, an adipocytokine involved in the pathogenesis of insulin resistance necessary for developing type 2 diabetes, induces β-cell apoptosis and impaired β-cell function by promoting IL-1β production in human pancreatic islets [
54]. Expression of leptin is also positively correlated with BMI [
55]. In the current study, obese patients tended to have the highest urinary IL-1β levels and increased urinary IL-1β may be indicative of advanced disease. Consequently, urinary IL-1β levels/BMI may prove to be useful prognostic indicator of gynecologic disease.
The greatest limitation of urinary IL-1β as a biomarker for OC is kidney function. Inflammatory mediators, including IL-1β, are typically found elevated in the urine and serum of patients with impaired kidney function [
56]-[
58]. Furthermore, elevated levels of IL-1β have been reported in vaginal secretions associated with gynecologic infections; however, Basso, et al. were unable to detect IL-1β in patient urine or serum [
59]. Interestingly, while the pH of our urine samples were within neutral range suggesting the absence of urinary tract infections and normal renal function, two samples in our study, 3119–2 and 3119–13, displayed unusually elevated urinary IL-1β levels. Unfortunately, clinical information pertaining to kidney injury/dysfunction or gynecological infection was unavailable as it may have contributed to excessive urinary levels of IL-1β. Clearly, the potential for confounding clinical parameters to influence the impact of urinary IL-1β levels for gynecologic disease warrants further investigation.
The data in this study are derived from a small pilot study, but is representative of our institutional clinical practice in regard to OC histology and stage distribution as well as published pilot studies examining IL-1β in urine, serum, and plasma [
17],[
56],[
58]-[
61] and other clinically relevant pilot studies [
62]-[
66]. A normal baseline value for urinary IL-1β in women has not yet been established in the literature [
67]. Therefore, further study with increased sample size may assist in the development of statistically significant baseline and threshold values that could be used to differentiate between healthy, benign disorders, and OC, as well as other clinical parameters such as metabolic disruption. Further, urinary levels of IL-1β levels/BMI as a prognostic indicator of gynecologic disease could impact clinical practice.
Competing interests
The authors declare that they have no competing interests.