Abstract
Background/Aim: Inflammation plays an important role in prostate cancer (PCa). We evaluated proinflammatory cytokines regarding differential diagnosis of PCa in men with PSA levels between 2-10 ng/ml. Patients and Methods: Serum samples of 79 men (PSA 2-10 ng/ml) were analyzed for 10 proinflammatory cytokines (IL-6, IL-8, TNF-α, IFN-γ, IL-10, IL-1β, IL-2, IL-4, IL-12p70, IL-13) and results were evaluated regarding presence of PCa and disease severity. Results: Significant differences between PCa patients and controls were found for IL-6 (p=0.002), IL-8 (p=0.030), and TNF-α (p=0.009), although they were not predictors of PCa in a logistic regression analysis. In addition, IL-6 and TNF-α levels were significantly higher in patients with high-risk PCa (p<0.05). No significant differences were observed regarding the other cytokines. Conclusion: In patients with PSA levels between 2-10 ng/ml, IL-6, IL-8, and TNF-α are associated with PCa, and IL-6 and TNF-α are associated with high-risk PCa.
Prostate cancer (PCa) has been reported to be the second cancer, immediately after lung cancer, with the highest incidence in men worldwide, and indeed it is the cancer with highest incidence in Europe, North, Central and South America, in several African countries and, also, in Australia (1). The high incidence of PCa is probably due to the massive use of prostate-specific antigen (PSA), mostly for screening purposes. Consequently, a significant proportion of all diagnosed cancers are clinically insignificant, and most likely would be clinically silent during an individual’s lifetime. This overdiagnosis of PCa also leads to overtreatment of low-risk PCa, with a negative impact on patients’ quality of life (2, 3). The concerns about overdiagnosis and overtreatment have led to an extensive debate about the utility of PSA-based PCa screening, mostly after the results of two large prospective studies on this subject were published: the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial (4), and the European Randomized Study of Screening for Prostate Cancer (ERSPC), although this last study showed that after a 16-years follow-up period, PSA screening significantly reduced PCa mortality (5). The basis for the overdiagnosis issue is the limited specificity of PSA for cancer, as several non-cancerous events may increase PSA serum levels (6). Therefore, there is a need to find new biomarkers that can help in PCa diagnosis, mostly in the detection of more aggressive and clinically significant forms (6, 7).
Inflammation has been associated with several types of cancers. It may promote carcinogenesis through several mechanisms, such as the promotion of cell proliferation and angiogenesis, as well as inducing DNA damage (8, 9). Prostatitis is a common finding in the peripheral zone of the prostate, where PCa also tends to develop (10). Advanced age is a known risk factor for PCa development, and other emerging risk factors are being considered, such as diet (11), as suggested by studies that have documented an association between serum lipids levels and PCa (12). With the advancement of age, the prostate becomes increasingly exposed to tissue injury, resulting from diet factors, infection, or others, which can lead to an inflammatory response (8). Exposure of the prostate to these agents can result in histological lesions of the prostate epithelium, known as proliferative inflammatory atrophy (PIA). These inflammatory lesions can either progress directly to carcinoma, or indirectly by giving rise, in the first place, to a high-grade prostatic intraepithelial neoplasia (HGPIN) (13). Endocrine factors can also promote prostate inflammation, as is the case with an increase in estrogens in the presence of testosterone (14).
Studies of prostatic inflammation in rat models have shown an activation of inflammasome-mediated proinflammatory cytokines, leading to an inflammatory state of the prostate similar to benign prostatic hyperplasia in humans (15), and it is postulated that these lesions could evolve further to tumor development (8).
All the evidence lead to an important role of inflammation in PCa development. However, there are still only a few studies on the role of proinflammatory cytokines in PCa. Therefore, our objective was to evaluate whether the serum levels of proinflammatory cytokines are related to the presence of PCa and, in particular, to the most aggressive forms of the disease.
Patients and Methods
Study design. In this observational study, from a cohort of 235 patients recruited at the Urology Department of the Central Lisbon University Hospital Center, who underwent a prostate biopsy for PCa suspicion and with a PSA level between 2 and 10 ng/ml (Hybritech calibration), we selected a group of 79 patients for this study. All patients were prospectively followed up until staging was concluded. Staging was performed according to the American Joint Committee on Cancer (AJCC) TNM staging system (16). Patients were divided into four groups according to the severity of their disease, and selection was performed to obtain an even number of patients in each group, and with a similar age distribution. 19 patients without PCa were included in the control group. The remaining 60 patients were divided into three groups of 20 patients each, that we classified, according to the PCa stage, as: low-risk (stages I and IIA), high-risk (stages IIB and IIC) and very high-risk (stages IIIB, IIIC, IVA and IVB). All participants provided written informed consent. The study protocol was approved by the Institutional Review Board and complied with the Helsinki Declaration.
All men included had an initial or repetition prostate biopsy performed with at least 12 cores. Pathological assessment of specimens was performed by the same expert genitourinary pathologist. Biopsy results were reported according to the definitions of the 2014 consensus conference of the International Society of Urological Pathology (17). Men with atypical small acinar proliferation or HGPIN were considered to have no PCa. In the subset of patients who underwent radical prostatectomy, we considered for staging purposes only the total Gleason score (GS) of the prostatectomy, instead of the GS of the biopsy, whenever there was a difference between them. Blood collection for biomarkers assessment was done on the same day of the biopsy.
Patients. Seventy-nine patients from the Urology Department of the Central Lisbon University Hospital Center, with a prostate biopsy scheduled for suspected PCa, were selected. The inclusion criteria were: no previous history of PCa, PSA level between 2 and 10 ng/ml (Beckman Coulter Hybritech®), no previous transurethral resection of the prostate, no therapy with drugs that may affect PSA concentration (5-α-reductase inhibitors and androgens), no urinary infection contemporary to blood collection, or acute bacterial prostatitis in the three months prior to the biopsy, without hemophilia, history of multiple blood transfusions, or chronic renal failure. Heavily hemolyzed serum samples also implied the exclusion of patients. The 79 patients were selected based on their cancer status, in order to assemble four groups with similar subject numbers and age-matching: one control group (without PCa) and three other groups based on disease severity, as previously described.
Blood samples processing and laboratory assays. Blood was collected in tubes without any anticoagulant and containing a separator gel. Samples were centrifuged and refrigerated (2-8°C) within less than 3 h after the blood draw. Serum aliquots were separated from each sample and frozen at –80°C within 8 h after the blood draw. Samples were transferred on dry ice to the Institute of Laboratory Medicine of the German Heart Center Munich, Clinics at the Technical University Munich, and stored at –80°C until analysis of the biomarkers.
Samples were thawed only once to perform a panel of 10 proinflammatory cytokines: interferon gamma (IFN-γ), IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13, and tumor necrosis factor alpha (TNF-α). These biomarkers were assayed using an electrochemiluminescent immunoassay, using the Meso Scale Discovery® V-Plex® multi-spot assay system on the MESO QuickPlex SQ 120 instrument (Meso Scale Diagnostics, LLC, Rockville, MD, USA). Beckman Coulter Hybritech® PSA was performed on the Access 2 immunoassay analyzer (Beckman Coulter, Brea, CA, USA).
Statistical analysis. Continuous variables are described as median and interquartile range (IQR). Mann–Whitney U-test or Kruskal-Wallis test were used to compare them across cancer risk groups. Values below detection, which can still be ranked, were included in the non-parametric test for IL-6 and IL-10. Univariate logistic regression analysis was used to predict the outcomes. Odds ratios (ORs) with 95% confidence intervals were also calculated. Statistical significance was defined as a two-sided p-value <0.05. Statistical analysis was performed using IBM® SPSS® Statistics 26.0 (IBM Corp., Armonk, NY, USA, 2019).
Results
Age and biomarker values. The median age of the participants was 69 years (IQR=62-73), with a minimum of 51 years and a maximum age of 87 years. Patients with PCa had a median age of 70 years (IQR=62-73) and controls (patients without PCa) had a median age of 67 years (IQR=61-73). Men in all three risk groups had median ages between 69 and 70 years. Thus, the age distribution was similar between all groups of participants, including controls, and there were no significant age differences between them (p=0.166). Overall, PSA had a median value of 6.0 ng/ml (IQR=4.4-8.1), with a minimum of 2.4 ng/ml and a maximum of 9.9 ng/ml. The results of biomarkers measured in all groups of patients are shown in Table I.
Measured parameters in the different groups of patients.
IL-6 and IL-8. For IL-6, significant differences were observed between patients with and without PCa (p=0.002; Figure 1A) as well as between all groups (p=0.012). IL-6 levels were significantly higher in patients with high- or very high-risk disease than in those without cancer or with low-risk disease (p=0.029; Figure 1B). Similarly, for IL-8, there was a significant difference between patients with and those without PCa (p=0.030; Figure 1C). However, there were no significant differences between all groups (p=0.128) and between patients without PCa or low-risk cancer versus high- or very high-risk cancer (p=0.092; Figure 1D).
Distribution of IL-6 (A, B), IL-8 (C, D) and TNF-α (E, F) values between different groups of patients (for IL-6, an outlier value of 169.3 pg/ml in the high risk cancer group is not shown, although it was used to calculate the box-plot). *p<0.05; **p>0.05.
TNF-α, IFN-γ and IL-10. TNF-α levels were significantly different between patients with and without PCa (p=0.009; Figure 1E), and there was also a significant difference when comparing levels in patients without cancer or with low-risk disease, with patients with high-risk or very high-risk PCa (p=0.020; Figure 1F). When comparing all groups of patients, the differences were also significant (p=0.047). No significant differences were observed in IFN-γ and IL-10 levels.
IL-1β, IL-2, IL-4, IL-12p70 and IL-13. The other cytokines were not detectable in the majority of the 79 patients and could not be evaluated, e.g. IL-1β (n=78), IL-2 (n=76), IL-4 (n=78), IL-12p70 (n=78) and IL-13 (n=78). In contrast, IL-6 was not detectable in 26 (11/19 non-PCa patients and 15/60 PCa patients) and IL-10 in 43 cases (9/19 non-PCa patients and 34/60 PCa patients).
Overall PCa prediction. Logistic regression analysis was performed, along with PSA, only for the cytokines that showed significant differences between patients with and without PCa. On univariate logistic regression analysis, IL-6, IL-8, and TNF-α were not predictors of PCa (p=0.051, p=0.963, and p=0.275, respectively), and PSA was not predictive of PCa (p=0.138; Table II).
Univariate logistic regression analysis for the prediction of overall and of high or very high risk prostate cancer.
High- or very high-risk PCa prediction. When considering high- or very high-risk PCa as the outcome, only PSA showed predictive ability on univariate logistic regression analysis (p=0.001). Again, IL-6 (p=0.258), IL-8 (p=0.996), and TNF-α (p=0.165) were not predictors of high- or very high-risk PCa on logistic regression analysis (Table II).
Discussion
Of the 10 cytokines measured in this special cohort of men with PSA between 2 and 10 ng/ml, only IL-6, IL-8, and TNF-α showed significantly higher values in patients with PCa than in controls (p=0.009 for TNF-α versus p=0.030 for IL-8 and p=0.002 for IL-6). Moreover, only TNF-α (p=0.020) and IL-6 (p=0.029) levels were significantly higher in the most aggressive forms of the disease. Moreover, IL-6 levels were undetectable in 26 patients. On the other hand, using logistic regression analysis, both IL-6, IL-8, and TNF-α were not predictors of overall PCa and were unable to discriminate the most aggressive forms of PCa in this specific cohort.
Our results show that IL-6, IL-8, and TNF-α cannot be used as general diagnostic biomarkers for PCa. However, the differences between these three proinflammatory biomarkers in cancer and non-cancer groups suggest that inflammation plays a relevant role in the pathogenesis of PCa, which is also detectable in this very specific patient cohort. These findings are consistent with those of other studies. Several published results support the view that IL-6 is an oncogenic factor in PCa (18). For instance, it was shown that benign prostate cells can develop a malignant phenotype and a stem-like behavior upon treatment with exogenous IL-6 (19). IL-6 expression occurs in several PCa cell lines, mostly in those that do not express the androgen receptor and have a greater malignant capacity (20). The oncogenic effects of IL-6 in PCa cells have been reported to result from the activation of several signaling pathways, such as the signal transducer and activator of transcription (STAT) factor-3 (18).
IL-8 has been reported to contribute to cell proliferation under androgen depletion conditions, being associated with PCa aggressiveness and with the loss of androgen receptors that occur in metastatic disease (21, 22). According to Veltri et al., increased IL-8 concentrations were found with increasing PCa stages. These authors also verified that IL-8 performed better than total PSA, and similar to the free to total PSA ratio, in distinguishing PCa from benign prostatic hyperplasia. However, Veltri et al. performed this study over a broader range of PSA concentrations (23), not limited to the 2-10 ng/ml PSA range as we did. The expression of IL-8 in human PCa cells has also been associated with angiogenesis, tumorigenicity, and metastasis development (24).
TNF-α has also been implicated in PCa development. It has been associated with the induction of C-C chemokine receptor 7 expression, thereby increasing the metastatic potential of PCa cells in lymph node metastasis (25). Nakashima et al. found significantly higher serum levels of TNF-α in patients with advanced PCa and cachexia than in patients with less severe forms of PCa (26). Elevated serum levels of this biomarker have also been found in PCa patients with bone metastasis, although the difference was not statistically significant (27). A polymorphism of the TNF-α gene, 308/CT, has also been implicated in an increased susceptibility to PCa development (28).
Our study has certain limitations. First, the number of participants (n=79) was small, especially in the control group (n=19). When considering overall PCa as the outcome, PSA was not shown to be a predictor, most certainly because the control group was too small, and the group was restricted to patients in the PSA gray zone between 2 and 10 ng/ml. When considering high- or very high-risk PCa as the outcome, the control group became larger (n=29), and PSA could reveal itself to be a predictor (p=0.001) on univariate analysis, as expected. Another limitation of this study was the absence of data about patients’ comorbidities that could also affect the proinflammatory cytokines levels (e.g., inflammatory or infectious conditions, other cancers, tissue injury). In Figure 1, there are obvious outliers in the control group (patients without PCa), which may be accounted for by other non-documented disease states.
We believe that this study can contribute not only to highlight the importance of inflammation in PCa, but mainly to draw attention to the role of IL-6, IL8 and TNF-α in its pathogenesis. Even when PSA levels are not very high, i.e., below 10 ng/ml, our study shows that serum levels of these biomarkers are already significantly higher in patients with PCa, when compared to controls.
We would like to highlight that the patients used in this study were selected from a special cohort with PSA levels in the “gray zone”, where decision making about whether or not to perform a biopsy is usually difficult. It seems that subtle inflammatory processes occur, which can be inferred from the quantification of some, but not all, cytokines in blood circulation. These markers showed a weak association with the presence of cancer (IL-6, IL8, and TNF-α) and also with the presence of high-risk cancer (IL-6 and TNF-α). However, in this special cohort, they did not show any diagnostic value. Nevertheless, the inclusion of markers that play a more indirect role in tumor pathogenesis can also be considered when looking for meaningful marker patterns. In addition, their relevance to prognosis, if a cancer is present, should be investigated. Likewise, its role during the treatment of PCa should also be investigated. In this regard, Tanji et al. demonstrated that both hormone therapy and radiation therapy can significantly increase the serum levels of some proinflammatory cytokines, while decreasing the levels of others. For example, these authors found that IFN-γ serum levels increased significantly during radiotherapy (29).
In summary, this study suggests that, for men with PSA levels between 2 and 10 ng/ml, IL-6, IL-8, and TNF-α are associated with PCa, since their serum levels are significantly higher in cases than in controls. IL-6 and TNF-α tend to be associated with more severe forms of PCa, as their serum levels are significantly elevated in patients with more advanced stages of the disease. However, these three proinflammatory cytokines were not predictors of PCa. IL-1β, IL-2, IL-4, IL-12p70, and IL-13 could not be detected at quantifiable concentrations. Therefore, their role in PCa pathogenesis could not be assessed in the present study. IFN-γ and IL-10 did not show an association with PCa in this group of patients. Larger studies should be conducted to investigate the role of IL-6, IL-8, and TNF-α in PCa, taking into account other comorbidities, besides PCa, that could affect their serum levels.
Footnotes
Authors’ Contributions
The Authors confirm their contribution to the paper as follows: study conception and design: Manuel M. Garrido and Stefan Holdenrieder; data collection: Manuel M. Garrido and Kimberly Krüger; analysis and interpretation of results: Manuel M. Garrido and Ruy M. Ribeiro; draft manuscript preparation: Manuel M. Garrido; critical review of the manuscript: Stefan Holdenrieder, Ruy M. Ribeiro, Kimberly Krüger, João T. Guimarães and Luís C. Pinheiro. All Authors approved the final version of the manuscript.
Conflicts of Interest
The Authors declare that there are no conflicts of interest in relation to this study.
- Received April 13, 2021.
- Revision received May 2, 2021.
- Accepted May 18, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.