Methods - We conducted a population-based nationwide cohort study of 17,168 patients newly diagnosed with PCa between 1996 and 2013 using the National Health Insurance Research Database (NHIRD) of Taiwan. Cox proportional hazards models with 1:1 propensity score-matched analysis were used to investigate the association between ADT use and the risk of autoimmune diseases. The autoimmune diseases included Graves’ disease, Crohn’s disease, psoriasis, systemic lupus erythematosus, rheumatoid arthritis, ankylosing spondylitis, Guillain-Barre syndrome, Sjogren’s syndrome, myasthenia gravis, pernicious anemia, hereditary hemolytic anemia, polyarteritis nodosa, Celiac disease, uveitis, polymyalgia rheumatica, dermatomyositis, Hashimoto’s thyroiditis, hypersensitivity vasculitis, Behcet’s disease, polymyositis, alopecia areata, Wegener’s granulomatosis, ulcerative colitis, autoimmune hemolytic anemia, pemphigus, multiple sclerosis, systemic sclerosis, Goodpasture syndrome, giant cell arteritis, thromboangiitis obliterans, arteritis obliterans, and Kawasaki disease. The duration of ADT use as a time-dependent variable was also examined for its association with autoimmune diseases. We also performed six secondary analyses.
Results - Of the 17,168 selected PCa patients, 14,444 patients met all the inclusion and exclusion criteria. After propensity score matching, 5590 ADT users and 5590 non-ADT users were included in the study cohort. A propensity score-matched analysis (adjusted hazard ratio (aHR), 0.619, 95% confidence interval (CI), 0.51–0.75, P < 0.001) demonstrated a significantly decreased risk of autoimmune diseases in ADT users. A significant decrease in the risk of autoimmune diseases with increasing ADT duration was also demonstrated (P < 0.001).
Conclusions - We observed that ADT use in patients with PCa was associated with a decreased risk of autoimmune diseases. These novel findings provide a potential role for androgen deprivation therapy in the modification of inflammation and autoimmunity in Asian patients with prostate cancer.
Introduction
Prostate cancer (PCa) is the most common cancer in the United States1. Androgen deprivation therapy (ADT) has been a mainstay of treatment for advanced PCa since the pivotal study of Huggins and Hodges in 19412. A marked increase in the use of ADT has been observed in recent years. For example, greater than half of newly diagnosed PCa patients in Taiwan undergo ADT3. In addition, approximately 500,000 PCa patients receive ADT in the United States4.
Androgens play an important role in PCa development and may also regulate a certain part of human immunity5. It is interesting to investigate the immune effects of the castration achieved by ADT given that this information remains limited6.
Increasing evidence has further indicated that ADT is associated with several diseases, including osteoporosis, cardiovascular disease, cerebrovascular disease, and Alzheimer’s disease7,8. Moreover, an association between ADT and a decreased risk of ulcerative colitis has also been reported9. ADT suppresses the progression of PCa and may also alter gut autoimmunity9. In addition, ADT reduces several immune-related cytokines in PCa patients10,11. Given similar mechanisms, it has been theorized that ADT may exhibit some association with autoimmune diseases.
To date, however, there have been only limited investigations regarding the possible association between ADT and autoimmune diseases. We supposed that an association between ADT and autoimmune diseases may be of clinical importance and should be discussed. In this study, we used a large-scale nationwide database of Taiwan to determine whether the use of ADT is associated with the subsequent development of autoimmune diseases in PCa patients.
Methods
Data source and collection
We conducted a nationwide cohort study using data from Taiwan’s National Health Insurance Research Database (NHIRD). The NHIRD is an extensive database that contains data from the National Health Insurance (NHI) program. The NHI program is the unique medical insurance system of Taiwan that covers 99.5% of Taiwan’s 23 million residents12. For this study, we used the Registry for Catastrophic Illness Patient Database (RCIPD), a subdatabase of the NHIRD. In Taiwan, patients diagnosed with cancer or other major diseases are entitled to a waiver for medical payment after receiving a catastrophic illness certification. To that end, clinical information relevant to the diagnosis of any malignancy, including pathological reports and cytology reports, is sent to the NHI administration for review by experts to confirm the given diagnosis, before which the relevant waiver or waivers is/are approved. The RCIPD contains all the clinical information on patients with a confirmed malignancy13. Clinical diagnoses for the patients in the database were determined according to the International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM). This study was approved by the Institutional Review Board of the Tri-Service General Hospital (approval number: TSGHIRB NO B-104-21).
Study Population
The study subjects were selected by using RCIPD data for the period from January 1996 to December 2013. The accuracy of the PCa diagnoses of the selected subjects was confirmed by both ICD-9-CM codes and their inclusion in the RCIPD. All patients with PCa who had follow-up data for at least 180 days after the initial PCa diagnosis were enrolled in the study population (Supplementary Fig. 1). In general, patients who were newly diagnosed with PCa (ICD-9-CM: 185) between 1996 and 2013 were considered for inclusion. The exclusion criteria were as follows: patients who were diagnosed before 1 January 1997; patients who were younger than 40 years old at the time of diagnosis; patients who had a previous history of autoimmune diseases; and patients with less than 180 days follow-up after the PCa diagnosis.
Supplementary Figure 1. Study flowchart of cohort selection. ADT, androgen deprivation therapy
Those patients who were enrolled in the study population were then divided into two groups, ADT users and nonADT users, according to whether they underwent ADT.
Study outcomes and covariates
Those PCa patients who underwent ADT were clearly identified in the RCIPD. The use of ADT includes the use of GnRH agonists (leuprolide, goserelin, triptorelin, and buserelin), oral antiandrogens (cyproterone acetate, bicalutimide, flutamide, and nilutamide), and estrogens (diethylstilbestrol and estramustine)14.
In Taiwan, patients diagnosed with autoimmune diseases are also registered in the RCIPD, and the medical records of these patients are subjected to a detailed evaluation to ensure that the diagnostic criteria are met and to ensure further confirmation by experts assigned by the NHI administration. A total of 33 new onset autoimmune diseases were identified in the RCIPD through the use of ICD9-CM codes (Supplementary Table 1)15.
Supplementary Table 1. ICD-9 diagnostic codes of 33 autoimmune diseases
The exact incidence of autoimmune diseases among ADT users was determined by only including those who received an autoimmune disease diagnosis after the initiation of ADT at least 180 days after the PCa diagnosis. In addition, the incidence of autoimmune diseases among nonADT users was determined by exclusively including those who received an autoimmune disease diagnosis at least 180 days after the PCa diagnosis and after the median time to ADT use in this study. Censoring was defined as death, the dates of outcome incidence (i.e., autoimmune diseases), or until the end of the follow-up period on 31 December 2013, whichever came first.
PCa patients were classified into the following five age groups: <50 years, 50–60 years, 60–70 years, 70–80 years, and >80 years. The comorbidity covariates reported to be related to PCa in the past literature16 were studied, including diabetes mellitus (ICD-9-CM: 250), hypertension (ICD-9-CM: 401–405), hyperlipidemia (ICD-9-CM: 272), coronary heart disease (ICD-9-CM: 410–414), chronic kidney disease (ICD-9-CM: 585, 586, and 588), chronic liver disease (ICD-9-CM: 456, 571, and 572), and cerebral vascular accident (ICD-9-CM: 430–438).
Statistical analysis
The baseline characteristics of the patients were first analyzed using descriptive statistics. The two groups (ADT users and non-ADT users) were compared using the χ2 test for categorical variables. We calculated the autoimmune disease incidence rates (per 100,000 person-year), and incidence rate ratios were then estimated using Poisson regressions. The Kaplan−Meier (KM) curve was used to estimate the cumulative incidences of autoimmune diseases for the two groups, and the difference between two groups was estimated with log-rank test. Hazard ratios (HRs) were calculated using propensity score-matched, and multivariable-adjusted Cox proportional hazards models were used to test the association between ADT and autoimmune diseases.
Secondary analyses
We used six secondary analyses to minimize the influence of potential bias in this study. First, to ensure the accuracy of event definitions, we defined autoimmune diseases as only those cases from the RCIPD with catastrophic illnesses with a detailed review by experts. Second, we used a 1:1 nearest-neighbor propensity-score-matched analysis to reduce the influence of age and underlying comorbidities (diabetes mellitus, hypertension, hyperlipidemia, coronary heart disease, chronic kidney disease, chronic liver disease, and cerebral vascular accident). The propensity scores were estimated by a logistic regression model and an 8-to-1-digit method with a digit-based greedy matching algorithm. Third, we used competing risk regression models proposed by Fine and Gray to adjust for death from any cause17. Fourth, the use of ADT has further adjusted as a time dependent covariate during the follow-up period to avoid immortal time bias. Fifth, according to the duration of the use of ADT, ADT usage was categorized into three levels (no ADT, <12 months, and ≥12 months) to evaluate whether there was any “dose-response” effect in the relationship between ADT and autoimmune diseases. According to the different types of ADT, ADT use was categorized into four types (GnRH agonists alone, GnRH agonists + oral antiandrogens, oral antiandrogens alone, and estrogens only) to evaluate whether there were any relationships between types of ADT and autoimmune diseases (Supplementary Table 2). Finally, we used a falsification analysis to validate the associations among other unmeasured characteristics of patients9,18. We selected five diseases, including anaphylaxis, pneumonia, tuberculosis, appendicitis, and abdominal aortic aneurysm, without known or hypothesized associations in the literature as outcomes to evaluate their associations, if any, with ADT. Moreover, anemia, a well documented adverse event of ADT, was used to validate the methods via adjustment for anemia in the study cohorts.
Supplementary Table 2. Propensity Score-Matched Cox Regression Analysis for the Association between Types of ADT and Autoimmune Diseases
The proportional hazard assumption was tested using Schoenfeld residuals and a log-minus-log graph. SPSS version 22.0 for Windows (IBM, Armonk, NY, USA), and SAS version 9.2 (SAS Institute, Cary, NC, USA) computer software programs were used to perform all statistical analyses. STATA version 11.2 (StataCorp, College Station, TX, USA) software program was used to produce KM curve plots based on the number at risk. Comparison results with a P < 0.05 were considered statistically significant.
Results
Demographics
A total of 14,444 patients newly diagnosed with PCa were enrolled in the study after meeting all the inclusion criteria. Of those patients, 6210 underwent ADT, and 8234 were non-ADT users. After propensity score matching, 5590 ADT users and 5590 non-ADT users were included in the study cohort, with a median time from PCa diagnosis to ADT use of 14.6 days. Among those patients, 457 were newly diagnosed with autoimmune diseases during a median follow-up period of 3.82 years (interquartile range, 2.18–6.46 years). Of these patients, 155 (2.8%) patients were ADT users, and 302 (5.4%) were non-ADT users (Supplementary Figure 1).
The demographic characteristics of ADT in PCa patients are provided in Table 1. For the full cohort, patients who underwent ADT were significantly older and had significantly more comorbidities. After the propensity score matching, however, no statistically significant differences in ages and comorbidities were noted between ADT users and non-ADT users. The incidence of 33 autoimmune diseases among the ADT users and non-users are listed in Supplementary Table 3.
Table 1. Demographic characteristics of patients with prostate cancer according to the use of androgen deprivation therapy
Results - Of the 17,168 selected PCa patients, 14,444 patients met all the inclusion and exclusion criteria. After propensity score matching, 5590 ADT users and 5590 non-ADT users were included in the study cohort. A propensity score-matched analysis (adjusted hazard ratio (aHR), 0.619, 95% confidence interval (CI), 0.51–0.75, P < 0.001) demonstrated a significantly decreased risk of autoimmune diseases in ADT users. A significant decrease in the risk of autoimmune diseases with increasing ADT duration was also demonstrated (P < 0.001).
Conclusions - We observed that ADT use in patients with PCa was associated with a decreased risk of autoimmune diseases. These novel findings provide a potential role for androgen deprivation therapy in the modification of inflammation and autoimmunity in Asian patients with prostate cancer.
Introduction
Prostate cancer (PCa) is the most common cancer in the United States1. Androgen deprivation therapy (ADT) has been a mainstay of treatment for advanced PCa since the pivotal study of Huggins and Hodges in 19412. A marked increase in the use of ADT has been observed in recent years. For example, greater than half of newly diagnosed PCa patients in Taiwan undergo ADT3. In addition, approximately 500,000 PCa patients receive ADT in the United States4.
Androgens play an important role in PCa development and may also regulate a certain part of human immunity5. It is interesting to investigate the immune effects of the castration achieved by ADT given that this information remains limited6.
Increasing evidence has further indicated that ADT is associated with several diseases, including osteoporosis, cardiovascular disease, cerebrovascular disease, and Alzheimer’s disease7,8. Moreover, an association between ADT and a decreased risk of ulcerative colitis has also been reported9. ADT suppresses the progression of PCa and may also alter gut autoimmunity9. In addition, ADT reduces several immune-related cytokines in PCa patients10,11. Given similar mechanisms, it has been theorized that ADT may exhibit some association with autoimmune diseases.
To date, however, there have been only limited investigations regarding the possible association between ADT and autoimmune diseases. We supposed that an association between ADT and autoimmune diseases may be of clinical importance and should be discussed. In this study, we used a large-scale nationwide database of Taiwan to determine whether the use of ADT is associated with the subsequent development of autoimmune diseases in PCa patients.
Methods
Data source and collection
We conducted a nationwide cohort study using data from Taiwan’s National Health Insurance Research Database (NHIRD). The NHIRD is an extensive database that contains data from the National Health Insurance (NHI) program. The NHI program is the unique medical insurance system of Taiwan that covers 99.5% of Taiwan’s 23 million residents12. For this study, we used the Registry for Catastrophic Illness Patient Database (RCIPD), a subdatabase of the NHIRD. In Taiwan, patients diagnosed with cancer or other major diseases are entitled to a waiver for medical payment after receiving a catastrophic illness certification. To that end, clinical information relevant to the diagnosis of any malignancy, including pathological reports and cytology reports, is sent to the NHI administration for review by experts to confirm the given diagnosis, before which the relevant waiver or waivers is/are approved. The RCIPD contains all the clinical information on patients with a confirmed malignancy13. Clinical diagnoses for the patients in the database were determined according to the International Classification of Diseases, 9th revision, Clinical Modification (ICD-9-CM). This study was approved by the Institutional Review Board of the Tri-Service General Hospital (approval number: TSGHIRB NO B-104-21).
Study Population
The study subjects were selected by using RCIPD data for the period from January 1996 to December 2013. The accuracy of the PCa diagnoses of the selected subjects was confirmed by both ICD-9-CM codes and their inclusion in the RCIPD. All patients with PCa who had follow-up data for at least 180 days after the initial PCa diagnosis were enrolled in the study population (Supplementary Fig. 1). In general, patients who were newly diagnosed with PCa (ICD-9-CM: 185) between 1996 and 2013 were considered for inclusion. The exclusion criteria were as follows: patients who were diagnosed before 1 January 1997; patients who were younger than 40 years old at the time of diagnosis; patients who had a previous history of autoimmune diseases; and patients with less than 180 days follow-up after the PCa diagnosis.
Supplementary Figure 1. Study flowchart of cohort selection. ADT, androgen deprivation therapy
Those patients who were enrolled in the study population were then divided into two groups, ADT users and nonADT users, according to whether they underwent ADT.
Study outcomes and covariates
Those PCa patients who underwent ADT were clearly identified in the RCIPD. The use of ADT includes the use of GnRH agonists (leuprolide, goserelin, triptorelin, and buserelin), oral antiandrogens (cyproterone acetate, bicalutimide, flutamide, and nilutamide), and estrogens (diethylstilbestrol and estramustine)14.
In Taiwan, patients diagnosed with autoimmune diseases are also registered in the RCIPD, and the medical records of these patients are subjected to a detailed evaluation to ensure that the diagnostic criteria are met and to ensure further confirmation by experts assigned by the NHI administration. A total of 33 new onset autoimmune diseases were identified in the RCIPD through the use of ICD9-CM codes (Supplementary Table 1)15.
Supplementary Table 1. ICD-9 diagnostic codes of 33 autoimmune diseases
The exact incidence of autoimmune diseases among ADT users was determined by only including those who received an autoimmune disease diagnosis after the initiation of ADT at least 180 days after the PCa diagnosis. In addition, the incidence of autoimmune diseases among nonADT users was determined by exclusively including those who received an autoimmune disease diagnosis at least 180 days after the PCa diagnosis and after the median time to ADT use in this study. Censoring was defined as death, the dates of outcome incidence (i.e., autoimmune diseases), or until the end of the follow-up period on 31 December 2013, whichever came first.
PCa patients were classified into the following five age groups: <50 years, 50–60 years, 60–70 years, 70–80 years, and >80 years. The comorbidity covariates reported to be related to PCa in the past literature16 were studied, including diabetes mellitus (ICD-9-CM: 250), hypertension (ICD-9-CM: 401–405), hyperlipidemia (ICD-9-CM: 272), coronary heart disease (ICD-9-CM: 410–414), chronic kidney disease (ICD-9-CM: 585, 586, and 588), chronic liver disease (ICD-9-CM: 456, 571, and 572), and cerebral vascular accident (ICD-9-CM: 430–438).
Statistical analysis
The baseline characteristics of the patients were first analyzed using descriptive statistics. The two groups (ADT users and non-ADT users) were compared using the χ2 test for categorical variables. We calculated the autoimmune disease incidence rates (per 100,000 person-year), and incidence rate ratios were then estimated using Poisson regressions. The Kaplan−Meier (KM) curve was used to estimate the cumulative incidences of autoimmune diseases for the two groups, and the difference between two groups was estimated with log-rank test. Hazard ratios (HRs) were calculated using propensity score-matched, and multivariable-adjusted Cox proportional hazards models were used to test the association between ADT and autoimmune diseases.
Secondary analyses
We used six secondary analyses to minimize the influence of potential bias in this study. First, to ensure the accuracy of event definitions, we defined autoimmune diseases as only those cases from the RCIPD with catastrophic illnesses with a detailed review by experts. Second, we used a 1:1 nearest-neighbor propensity-score-matched analysis to reduce the influence of age and underlying comorbidities (diabetes mellitus, hypertension, hyperlipidemia, coronary heart disease, chronic kidney disease, chronic liver disease, and cerebral vascular accident). The propensity scores were estimated by a logistic regression model and an 8-to-1-digit method with a digit-based greedy matching algorithm. Third, we used competing risk regression models proposed by Fine and Gray to adjust for death from any cause17. Fourth, the use of ADT has further adjusted as a time dependent covariate during the follow-up period to avoid immortal time bias. Fifth, according to the duration of the use of ADT, ADT usage was categorized into three levels (no ADT, <12 months, and ≥12 months) to evaluate whether there was any “dose-response” effect in the relationship between ADT and autoimmune diseases. According to the different types of ADT, ADT use was categorized into four types (GnRH agonists alone, GnRH agonists + oral antiandrogens, oral antiandrogens alone, and estrogens only) to evaluate whether there were any relationships between types of ADT and autoimmune diseases (Supplementary Table 2). Finally, we used a falsification analysis to validate the associations among other unmeasured characteristics of patients9,18. We selected five diseases, including anaphylaxis, pneumonia, tuberculosis, appendicitis, and abdominal aortic aneurysm, without known or hypothesized associations in the literature as outcomes to evaluate their associations, if any, with ADT. Moreover, anemia, a well documented adverse event of ADT, was used to validate the methods via adjustment for anemia in the study cohorts.
Supplementary Table 2. Propensity Score-Matched Cox Regression Analysis for the Association between Types of ADT and Autoimmune Diseases
The proportional hazard assumption was tested using Schoenfeld residuals and a log-minus-log graph. SPSS version 22.0 for Windows (IBM, Armonk, NY, USA), and SAS version 9.2 (SAS Institute, Cary, NC, USA) computer software programs were used to perform all statistical analyses. STATA version 11.2 (StataCorp, College Station, TX, USA) software program was used to produce KM curve plots based on the number at risk. Comparison results with a P < 0.05 were considered statistically significant.
Results
Demographics
A total of 14,444 patients newly diagnosed with PCa were enrolled in the study after meeting all the inclusion criteria. Of those patients, 6210 underwent ADT, and 8234 were non-ADT users. After propensity score matching, 5590 ADT users and 5590 non-ADT users were included in the study cohort, with a median time from PCa diagnosis to ADT use of 14.6 days. Among those patients, 457 were newly diagnosed with autoimmune diseases during a median follow-up period of 3.82 years (interquartile range, 2.18–6.46 years). Of these patients, 155 (2.8%) patients were ADT users, and 302 (5.4%) were non-ADT users (Supplementary Figure 1).
The demographic characteristics of ADT in PCa patients are provided in Table 1. For the full cohort, patients who underwent ADT were significantly older and had significantly more comorbidities. After the propensity score matching, however, no statistically significant differences in ages and comorbidities were noted between ADT users and non-ADT users. The incidence of 33 autoimmune diseases among the ADT users and non-users are listed in Supplementary Table 3.
Table 1. Demographic characteristics of patients with prostate cancer according to the use of androgen deprivation therapy
The incidence of nine autoimmune diseases for which the total number of patients was ≥5 in the ADT user group is provided in Table 2.
Validating risk factors
Cox regression analysis results for the independent risk factors associated with autoimmune diseases in the propensity score-matched cohort are provided in Supplementary Table 4. The results of validation of the study methods by Cox regression analysis for the association of ADT and selected diseases are provided in Supplementary Table 5.
Supplementary Table 4. Independent Risk Factors of Autoimmune Diseases among Androgen Deprivation Therapy in Propensity Score-Matched Cohort Identity by Cox Regression Analysis
Table 2. Incidence (per 100,000 person-years) of autoimmune diseases (in which total N ≧ 5 in the ADT group) among ADT users and non-ADT users
Validating risk factors
Cox regression analysis results for the independent risk factors associated with autoimmune diseases in the propensity score-matched cohort are provided in Supplementary Table 4. The results of validation of the study methods by Cox regression analysis for the association of ADT and selected diseases are provided in Supplementary Table 5.
Supplementary Table 4. Independent Risk Factors of Autoimmune Diseases among Androgen Deprivation Therapy in Propensity Score-Matched Cohort Identity by Cox Regression Analysis
Supplementary Table 5. Multivariable Cox Regression Analysis for the Association of ADT for Validation
Association between the use of ADT and autoimmune diseases
A statistically significant negative association was between ADT use and the risk of autoimmune diseases according to the propensity score-matched analysis (adjusted hazard ratio (aHR), 0.619; 95% CI, 0.510–0.752; P < .001) (Table 3 and Supplementary Table 6).
Supplementary Table 6. Propensity Score–Matched Cox Regression Analysis and Multivariable-adjusted Cox Regression Analysis for the Association between ADT Use and Autoimmune Diseases
The KM disease-free curve for autoimmune diseases in relation to ADT use is presented in Fig. 1. The cumulative probability of being autoimmune disease-free was significantly increased among ADT users in the propensity score-matched cohort (P < .001). ADT users exhibited significantly increased autoimmune disease-free rates (98% over 3 years, 97% over 5 years, and 91% over 8 years) compared with nonusers (96% over 3 years, 94% over 5 years, and 86% over 8 years).
For the nine autoimmune diseases for which the number of patients was ≥5 in the ADT user group, we analyzed the association between autoimmune diseases and ADT use (Table 3). Compared with non-ADT users, ADT users exhibited significantly reduced risks of autoimmune diseases, such as Graves’ disease (aHR, 0.45; 95% CI, 0.23–0.90), psoriasis (aHR, 0.52; 95% CI, 0.28–0.95), and uveitis (aHR, 0.50; 95% CI, 0.26–0.96). Further, Cox regression analysis of the duration of ADT use revealed that patients with ≥12 months of ADT use exhibited the least risk of subsequent autoimmune diseases (aHR, 0.61; 95% CI, 0.49–0.77; P < .001) (Table 4). A significantly decreased risk of autoimmune disease with increasing ADT duration was also demonstrated (P for trend < .001).
Discussion
To the best of our knowledge, this is the first study to investigate the association between ADT use and the risk of subsequent autoimmune diseases in an Asian population. We enrolled 14,444 patients with PCa in a retrospective nationwide cohort. In this large-scale study, we demonstrated a decreased risk of autoimmune diseases in ADT users using both propensity score-matched and multivariable regression models. We also demonstrated an association between greater durations of ADT use and decreased risks of autoimmune diseases.
Several possible mechanisms can explain the effects of ADT use on the etiology of autoimmune diseases. Chronic inflammation has previously been found to be associated with both PCa and autoimmune diseases19. Proliferative inflammatory atrophy, which is associated with prostate inflammation, has the potential to further progress to PCa20. Relatedly, several inflammatory cytokines, such as interleukin (IL)-1, IL-6, and IL-17, have been reported to be induced by PCa19.
Androgens enhance a T helper (Th) cell type 1 response21. In addition, ADT suppresses androgen levels and reduces Th1 response while also reducing the levels of inflammatory cytokines. Morse and McNeel revealed that Th1 and Th17 cells decrease in the periphery after 24 months of ADT use in PCa patients22. Recent studies have also found that several inflammatory cytokines are reduced after ADT use in PCa patients, including IL-1β, IL2, tumor necrosis factor (TNF)-α, and interferon (IFN)-γ10,11. IL-1β and TNF-α play a crucial role in numerous autoimmune diseases. Thus, a reduction in the levels of IL1β and TNF-α after ADT use may further decrease the risk of autoimmune diseases. Saylor et al. demonstrated that patients with PCa who received 12 months of ADT exhibit reduced IL-6 levels compared with non-ADT controls23. IL-6, which plays a crucial role in PCa progression, is also an activator of STAT3 and activated nuclear factor-κB (NFκB) complexes24. NF-κB activation has been implicated in PCa development and progression25. In addition, Guzmán-Soto et al. revealed that leuprolide acetate promotes a significant reduction in NF-κB activation and reduced IL-1β, IL-17A, and TNF-α mRNA expression levels in experimental autoimmune encephalomyelitis rats26.
Significantly decreased risks of Graves’ disease, psoriasis, and uveitis were also observed among ADT users in our study. The typical characteristic of psoriasis is localized hyper-proliferation of keratinocytes. T-cell-mediated hyperproliferation (Th-1, Th-17, and Th-22 cells) and overexpression of proinflammatory cytokines play important roles in the pathophysiology of psoriasis27. In patients with psoriasis, inflammatory cytokines are markedly elevated, and cytokine interactions have been reported to activate STAT1, STAT3, and NF-κB28. IL-6 and IL-17 regulate keratinocyte proliferation in psoriatic lesions via STAT and NF-κB pathways28. ADT use reduces the number of Th1 and Th 17 cells and decreases IL-6 levels22. Therefore, the use of ADT in PCa patients may reduce the incidence of psoriasis.
Uveitis is also believed to be T-cell dependent, and the roles of Th1 and Th17 cells have become major topics of interest in past decades29. Relatedly, the use of ADT has previously been reported to reduce Th1 and Th17 cells22. Thus, the incidence of uveitis may be reduced due to ADT usage.
Graves’ disease is caused by thyroid-stimulating autoantibodies. In thyroid tissue, recruited Th1 lymphocytes may be responsible for enhanced IFN-γ and TNF-α production that sustain the intrathyroidal autoimmune process30. Thus, a reduction in Th1 cells caused by ADT may subsequently reduce the incidence of Graves’ disease. In addition, proinflammatory regulators, including chemokines and their receptor networks, appear to contribute to the pathogenesis of both autoimmune diseases and PCa30,31. CXCL10 and its receptor CXCR3 contribute to the pathogenesis of Graves’ disease31. Chemokines and their receptors, including CXCL8 and CXCR1/CXCR2, CXCL12 and CXCR4, also influence the development of PCa and metastases32,33. In addition, CXCL10 and CXCR3 are responsible for the regulation of prostate tumor growth31. Shen and Lentsch demonstrated that CXL10 levels are significantly reduced in PCa in vitro and the transgenic adenocarcinoma of the mouse prostate (TRAMP) mouse model34. Therefore, the incidence of Graves’ disease may be reduced among PCa patients. Further studies are warranted to investigate the relationships among ADT use, CXCL10/CXCR3 levels, and autoimmune diseases.
In contrast, Morse and McNeel conducted a study with a rodent animal model that revealed increased levels of Th1 cells in prostate T cells within 30 days after castration followed by a predominance of Th17 cells, which persisted up to 90 days post castration35. Koh et al. demonstrated that castration alone induced IFN-γsecretion in mouse splenocytes at 3 weeks post-castration36, whereas Khosla et al. demonstrated that short-term GnRH agonist use over 4 weeks resulted in increasing levels of circulating proinflammatory cytokines in healthy elderly men37. These three studies only investigated short-term use of ADT. A significantly decreased risk of autoimmune diseases with long-term ADT (≥12 months) use was observed in our study. We hypothesize that long-term use of ADT may cause adaptation of the immune response. However, further studies are warranted to more fully investigate the relationship between long-term ADT use and immune responses. In the secondary analysis by types of ADT, significantly decreased risks of incident autoimmune diseases were observed in the use of GnRH agonists alone, GnRH agonist + oral antiandrogens, and oral antiandrogens alone compared to non-ADT use.
Yang et al. reported a 23% increased risk of rheumatoid arthritis (RA) in 44,785 patients who received ADT for PCa in North America38. They claimed androgen-mediated thymic regeneration may increase the risk of RA. However, Kili-Drori et al. revealed no association between the use of ADT and the incidence of psoriasis and RA in a conference paper39. In our study, a decreased trend of RA without statistical significance was observed. We hypothesized that reduction of IL-1 beta and TNF-alpha after ADT use may contribute to this phenomenon10,11. In addition, there is considerable disparity in the risk of RA between the different populations. The prevalence of RA is highest in native American and north America populations but very low in Asia40.
We also investigated the incidence rates of the 33 autoimmune diseases in this study (Supplementary Table 3). A comparison between the incidence rates of the three significantly decreased autoimmune diseases among ADT users and the incidence of these diseases reported in worldwide populations is provided in Supplementary Table 7. The incidence rates of these three autoimmune diseases among ADT users were reduced compared with those reported in previous studies41-46.
Supplementary Table 7. Comparison of the Incidence (per 100,000 person-years) of Autoimmune Diseases among ADT Users and that in Worldwide General Populations
Supplementary Table 3. Incidence (per 100,000) of 33 Autoimmune Diseases among ADT users and non-ADT users
One strength of this study investigating the relationship between ADT use and autoimmune diseases is that it is a large-scale nationwide population-based study. The NHI system of Taiwan covers greater than 99% of the total population of Taiwan. In addition, Taiwan’s NHIRD is among the few nationwide databases maintained by an Asian country.
However, several limitations in this study should be noted. First, prostate-specific antigen levels, clinical stages of PCa, and Gleason scores of the patients were not available from the NHIRD database. Thus, it was difficult to further investigate the severity of PCa. Second, all the diagnoses were defined by using the ICD-9 coding system, such that misinterpretations of some data or the misclassification of some diagnoses may have occurred. Third, laboratory data are not available in the NHIRD. Thus, inflammatory markers could not be assessed. Laboratory tests may provide more information for discovering the mechanism between autoimmune diseases and ADT use. Family history of autoimmune diseases for PCa patients was also not available. Finally, this population-based study was a retrospective study; so further prospective studies are needed to fully evaluate the relationship between ADT use and autoimmune diseases.
In conclusion, we demonstrated that ADT use in patients with PCa reduced the risk of autoimmune diseases compared with a general population cohort of 14,444 PCa patients. Further studies are warranted to obtain a better understanding of the relationship between ADT and autoimmune diseases.
Compliance with ethical standards: Conflict of interest - The authors declare that they have no conflict of interest.
1. Division of Urology, Department of Surgery, Taoyuan General Hospital, Ministry of Health and Welfare, Taoyuan, Taiwan
2. Faculty of Medicine, National Yang-Ming University, Taipei, Taiwan
3. Graduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan
4. Department of Pathology and Graduate Institute of Pathology and Parasitology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan
5. Department of Urology, Chang Gung Memorial Hospital, Keelung, Taiwan
6. Cancer Medicine Center of Hualien Tzu Chi Hospital, Tzu Chi University, Hualien, Taiwan
References:
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin. 2016;66:7–30.
2. Huggins C, Hodges CV. Studies on prostatic cancer. I. The effect of castration, of estrogen and androgen injection on serum phosphatases in metastatic carcinoma of the prostate. CA Cancer J Clin. 1972;22:232–40.
3. Cancer Registry Annual Report, 2000–2011. Taiwan: Department of Health, the Executive Yuan, Republic of China; 2003–13.
4. Shahani S, Braga-Basaria M, Basaria S. Androgen deprivation therapy in prostate cancer and metabolic risk for atherosclerosis. J Clin Endocrinol Metab. 2008;93:2042–9.
5. Fijak M, Meinhardt A. The testis in immune privilege. Immunol Rev. 2006;213:66–81.
6. Aragon-Ching JB, Williams KM, Gulley JL. Impact of androgen deprivation therapy on the immune system: implications for combination therapy of prostate cancer. Front Biosci. 2007;12:4957–71.
7. Keating NL, O’Malley AJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol. 2006;24:4448–56.
8. Nead KT, Gaskin G, Chester C, et al. Androgen deprivation therapy and future Alzheimer’s disease risk. J Clin Oncol. 2016;34:566–71.
9. Klil-Drori AJ, Tascilar K, Yin H, Aprikian A, Bitton A, Azoulay L. Androgen deprivation therapy and the incidence of inflammatory bowel disease in patients with prostate cancer. Am J Epidemiol. 2016;184:15–22.
10. Kaczmarek P, Pokoca L, Niemirowicz J, Majewska E, Baj Z. Effect of luteinizing hormone-releasing hormone (LHRH) analogue treatment on a cytokine profile in prostate cancer patients. Pharmacol Rep. 2008;60:399–403.
11. Salman H, Bergman M, Blumberger N, Djaldetti M, Bessler H. Do androgen deprivation drugs affect the immune cross-talk between mononuclear and prostate cancer cells? Biomed Pharmacother. 2014;68:21–24.
12. Bureau of National Health Insurance DoH, Executive Yuan. The National Health Insurance Statistics. 2013. http://www.nhi.gov.tw/English/webdata/webdata.aspx?menu=11&menu_id=296&webdata_id=1942&WD_ID=296. Accessed 18 Sep 2014.
13. Wu CY, Chen YJ, Ho HJ, Hsu YC, Kuo KN, Wu MS, et al. Association between nucleoside analogues and risk of hepatitis B virus-related hepatocellular carcinoma recurrence following liver resection. JAMA. 2012;308:1906–14.
14. Liu JM, Chen TH, Chuang HC, Wu CT, Hsu RJ. Statin reduces the risk of dementia in diabetic patients receiving androgen deprivation therapy for prostate cancer. Prostate Cancer Prostatic Dis. 2018. https://doi.org/10.1038/s41391-018-0091-4. [Epubahead of print]
15. Chen SJ, Chao YL, Chen CY, Chang CM, Wu EC, Wu CS, et al. Prevalence of autoimmune diseases in in-patients with schizophrenia: nationwide population-based study. Br J Psychiatry. 2012;200:374–80.
16. Liu JM, Lin PH, Hsu RJ, Chang YH, Cheng KC, Pang ST, et al. Complementary traditional Chinese medicine therapy improves survival in patients with metastatic prostate cancer. Medicine (Baltimore). 2016;95:e4475.
17. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509.
18. Frakt AB. An observational study goes where randomized clinical trials have not. JAMA. 2015;313:1091–2.
19. De Nunzio C, Kramer G, Marberger M, Montironi R, Nelson W, Schröder F, et al. The controversial relationship between benign prostatic hyperplasia and prostate cancer: the role of inflammation. Eur Urol. 2011;60:106–17.
20. Nelson WG, De Marzo AM, Isaacs WB. Prostate cancer. N Engl J Med. 2003;349:366–81.
21. Zandman-Goddard G, Peeva E. Gender and autoimmunity. Autoimmun Rev. 2007;6:366–72.
22. Morse MD, McNeel DG. Prostate cancer patients on androgen deprivation therapy develop persistent changes in adaptive immune responses. Hum Immunol. 2012;71:496–504.
23. Saylor PJ, Kozak KR, Smith MR, Ancukiewicz MA, Efstathiou JA, Zietman AL, et al. Changes in biomarkers of inflammation and angiogenesis during androgen deprivation therapy for prostate cancer. Oncologist. 2012;17:212–9.
24. Nadiminty N, Lou W, Lee SO, Lin X, Trump DL, Gao AC. Stat3 activation of NF-{kappa}B p100 processing involves CBP/ p300-mediated acetylation. Proc Natl Acad Sci USA. 2006;103:7264–9.
25. Huang S, Pettaway CA, Uehara H, Bucana CD, Fidler IJ. Blockade of NF-kappaB activity in human prostate cancer cells is associated with suppression of angiogenesis, invasion, and metastasis. Oncogene. 2001;20:4188–97.
26. Guzmán-Soto I, Salinas E, Quintanar JL. Leuprolide acetate inhibits spinal cord inflammatory response in experimental autoimmune encephalomyelitis by suppressing NF-κB activation. Neuroimmunomodulation. 2016;23:33–40.
27. Lowes MA, Suarez-Farinas M, Krueger JG. Immunology of psoriasis. Annu Rev Immunol. 2014;32:227–55.
28. Lowes MA, Bowcock AM, Krueger JG. Pathogenesis and therapy of psoriasis. Nature. 2007;445:866–73.
29. Horai R, Caspi RR. Cytokines in autoimmune uveitis. J Interferon Cytokine Res. 2011;31:733–44.
30. Antonelli A, Ferrari SM, Corrado A, Di Domenicantonio A, Fallahi P. Autoimmune thyroid disorders. Autoimmun Rev. 2015;14:174–80.
31. Engl T, Relja B, Blumenberg C, Müller I, Ringel EM, Beecken WD, et al. Prostate tumor CXC-chemokine profile correlates with cell adhesion to endothelium and extracellular matrix. Life Sci. 2006;78:1784–93.
32. Murphy C, McGurk M, Pettigrew J, Santinelli A, Mazzucchelli R, Johnston PG, et al. Nonapical and cytoplasmic expression of interleukin-8, CXCR1, and CXCR2 correlates with cell proliferation and microvessel density in prostate cancer. Clin Cancer Res. 2005;11:4117–27.
33. Taichman RS, Cooper C, Keller ET, Pienta KJ, Taichman NS, McCauley LK. Use of the stromal cell-derived factor-1/CXCR4 pathway in prostate cancer metastasis to bone. Cancer Res. 2002;62:1832–7.
34. Shen H, Lentsch AB. Progressive dysregulation of transcription factors NF-kappa B and STAT1 in prostate cancer cells causes
proangiogenic production of CXC chemokines. Am J Physiol Cell Physiol. 2004;286:C840–847.
35. Morse MD, McNeel DG. T cells localized to the androgen-deprived prostate are TH1 and TH17 biased. Prostate. 2012;72:1239–47.
36. Koh YT, Gray A, Higgins SA, Hubby B, Kast WM. Androgen ablation augments prostate cancer vaccine immunogenicity only when applied after immunization. Prostate. 2009;69:571–84.
37. Khosla S, Atkinson EJ, Dunstan CR, O’Fallon WM. Effect of estrogen versus testosterone on circulating osteoprotegerin and other cytokine levels in normal elderly men. J Clin Endocrinol Metab. 2002;87:1550–4.
38. Yang DD, Krasnova A, Nead KT, Choueiri TK, Hu JC, Hoffman KE, et al. Androgen deprivation therapy and risk of rheumatoid arthritis in patients with localized prostate cancer. Ann Oncol. 2018;29:386–91.
39. Kili-Drori A, Tascilar K, Yin H, Aprikian AG, Azoulay L. Androgen deprivation therapy and the incidence of autoimmune diseases. J Clin Oncol. 2015;33:e16010.
40. Silman AJ, Pearson JE. Epidemiology and genetics of rheumatoid arthritis. Arthritis Res. 2002;4:S265–272.
41. Carlé A, Pedersen IB, Knudsen N, Perrild H, Ovesen L, Rasmussen LB, et al. Epidemiology of subtypes of hyperthyroidism in Denmark: a population-based study. Eur J Endocrinol. 2011;164:801–9.
42. Teng W, Shan Z, Teng X, Guan H, Li Y, Teng D, et al. Effect of iodine intake on thyroid diseases in China. N Engl J Med. 2006;354:2783–93.
43. Bell LM, Sedlack R, Beard CM, Perry CO, Michet CJ, Kurland LT. Incidence of psoriasis in Rochester, Minn, 1980−1983. Arch Dermatol. 1991;127:1184–7.
44. Donker GA, Foets M, Spreeuwenberg P, van der Werf GT. Management of psoriasis in general practice now more in agreement with the guidelines of the Dutch College of General Practitioners (NHG). Ned Tijdschr Geneeskd. 1998;142:1379–83.
45. Gritz DC, Wong IG. Incidence and prevalence of uveitis in Northern California; the Northern California Epidemiology of Uveitis Study. Ophthalmology. 2004;111:491–500.
46. Hwang DK, Chou YJ, Pu CY, Chou P. Epidemiology of uveitis among the Chinese population in Taiwan: a population-based study. Ophthalmology. 2012;119:2371–6.
Read a Commentary by Dr. Andrew J. Armstrong: Androgen Deprivation Therapy for Prostate Cancer and the Risk of Autoimmune Diseases - Commentary
Association between the use of ADT and autoimmune diseases
A statistically significant negative association was between ADT use and the risk of autoimmune diseases according to the propensity score-matched analysis (adjusted hazard ratio (aHR), 0.619; 95% CI, 0.510–0.752; P < .001) (Table 3 and Supplementary Table 6).
Supplementary Table 6. Propensity Score–Matched Cox Regression Analysis and Multivariable-adjusted Cox Regression Analysis for the Association between ADT Use and Autoimmune Diseases
Table 3. The association between androgen deprivation therapy and autoimmune diseases (in which total n ≧ 5 in the ADT users group) by propensity score-matched Cox regression analysis
The KM disease-free curve for autoimmune diseases in relation to ADT use is presented in Fig. 1. The cumulative probability of being autoimmune disease-free was significantly increased among ADT users in the propensity score-matched cohort (P < .001). ADT users exhibited significantly increased autoimmune disease-free rates (98% over 3 years, 97% over 5 years, and 91% over 8 years) compared with nonusers (96% over 3 years, 94% over 5 years, and 86% over 8 years).
Fig. 1 Kaplan−Meier curves according to androgen deprivation therapy (ADT) use for the cumulative probability of remaining autoimmune diseases-free in the propensity score-matched cohort. ADT androgen deprivation therapy
For the nine autoimmune diseases for which the number of patients was ≥5 in the ADT user group, we analyzed the association between autoimmune diseases and ADT use (Table 3). Compared with non-ADT users, ADT users exhibited significantly reduced risks of autoimmune diseases, such as Graves’ disease (aHR, 0.45; 95% CI, 0.23–0.90), psoriasis (aHR, 0.52; 95% CI, 0.28–0.95), and uveitis (aHR, 0.50; 95% CI, 0.26–0.96). Further, Cox regression analysis of the duration of ADT use revealed that patients with ≥12 months of ADT use exhibited the least risk of subsequent autoimmune diseases (aHR, 0.61; 95% CI, 0.49–0.77; P < .001) (Table 4). A significantly decreased risk of autoimmune disease with increasing ADT duration was also demonstrated (P for trend < .001).
Table 4. Propensity score-matched Cox regression analysis for the association between the duration of ADT use and autoimmune diseases
Discussion
To the best of our knowledge, this is the first study to investigate the association between ADT use and the risk of subsequent autoimmune diseases in an Asian population. We enrolled 14,444 patients with PCa in a retrospective nationwide cohort. In this large-scale study, we demonstrated a decreased risk of autoimmune diseases in ADT users using both propensity score-matched and multivariable regression models. We also demonstrated an association between greater durations of ADT use and decreased risks of autoimmune diseases.
Several possible mechanisms can explain the effects of ADT use on the etiology of autoimmune diseases. Chronic inflammation has previously been found to be associated with both PCa and autoimmune diseases19. Proliferative inflammatory atrophy, which is associated with prostate inflammation, has the potential to further progress to PCa20. Relatedly, several inflammatory cytokines, such as interleukin (IL)-1, IL-6, and IL-17, have been reported to be induced by PCa19.
Androgens enhance a T helper (Th) cell type 1 response21. In addition, ADT suppresses androgen levels and reduces Th1 response while also reducing the levels of inflammatory cytokines. Morse and McNeel revealed that Th1 and Th17 cells decrease in the periphery after 24 months of ADT use in PCa patients22. Recent studies have also found that several inflammatory cytokines are reduced after ADT use in PCa patients, including IL-1β, IL2, tumor necrosis factor (TNF)-α, and interferon (IFN)-γ10,11. IL-1β and TNF-α play a crucial role in numerous autoimmune diseases. Thus, a reduction in the levels of IL1β and TNF-α after ADT use may further decrease the risk of autoimmune diseases. Saylor et al. demonstrated that patients with PCa who received 12 months of ADT exhibit reduced IL-6 levels compared with non-ADT controls23. IL-6, which plays a crucial role in PCa progression, is also an activator of STAT3 and activated nuclear factor-κB (NFκB) complexes24. NF-κB activation has been implicated in PCa development and progression25. In addition, Guzmán-Soto et al. revealed that leuprolide acetate promotes a significant reduction in NF-κB activation and reduced IL-1β, IL-17A, and TNF-α mRNA expression levels in experimental autoimmune encephalomyelitis rats26.
Significantly decreased risks of Graves’ disease, psoriasis, and uveitis were also observed among ADT users in our study. The typical characteristic of psoriasis is localized hyper-proliferation of keratinocytes. T-cell-mediated hyperproliferation (Th-1, Th-17, and Th-22 cells) and overexpression of proinflammatory cytokines play important roles in the pathophysiology of psoriasis27. In patients with psoriasis, inflammatory cytokines are markedly elevated, and cytokine interactions have been reported to activate STAT1, STAT3, and NF-κB28. IL-6 and IL-17 regulate keratinocyte proliferation in psoriatic lesions via STAT and NF-κB pathways28. ADT use reduces the number of Th1 and Th 17 cells and decreases IL-6 levels22. Therefore, the use of ADT in PCa patients may reduce the incidence of psoriasis.
Uveitis is also believed to be T-cell dependent, and the roles of Th1 and Th17 cells have become major topics of interest in past decades29. Relatedly, the use of ADT has previously been reported to reduce Th1 and Th17 cells22. Thus, the incidence of uveitis may be reduced due to ADT usage.
Graves’ disease is caused by thyroid-stimulating autoantibodies. In thyroid tissue, recruited Th1 lymphocytes may be responsible for enhanced IFN-γ and TNF-α production that sustain the intrathyroidal autoimmune process30. Thus, a reduction in Th1 cells caused by ADT may subsequently reduce the incidence of Graves’ disease. In addition, proinflammatory regulators, including chemokines and their receptor networks, appear to contribute to the pathogenesis of both autoimmune diseases and PCa30,31. CXCL10 and its receptor CXCR3 contribute to the pathogenesis of Graves’ disease31. Chemokines and their receptors, including CXCL8 and CXCR1/CXCR2, CXCL12 and CXCR4, also influence the development of PCa and metastases32,33. In addition, CXCL10 and CXCR3 are responsible for the regulation of prostate tumor growth31. Shen and Lentsch demonstrated that CXL10 levels are significantly reduced in PCa in vitro and the transgenic adenocarcinoma of the mouse prostate (TRAMP) mouse model34. Therefore, the incidence of Graves’ disease may be reduced among PCa patients. Further studies are warranted to investigate the relationships among ADT use, CXCL10/CXCR3 levels, and autoimmune diseases.
In contrast, Morse and McNeel conducted a study with a rodent animal model that revealed increased levels of Th1 cells in prostate T cells within 30 days after castration followed by a predominance of Th17 cells, which persisted up to 90 days post castration35. Koh et al. demonstrated that castration alone induced IFN-γsecretion in mouse splenocytes at 3 weeks post-castration36, whereas Khosla et al. demonstrated that short-term GnRH agonist use over 4 weeks resulted in increasing levels of circulating proinflammatory cytokines in healthy elderly men37. These three studies only investigated short-term use of ADT. A significantly decreased risk of autoimmune diseases with long-term ADT (≥12 months) use was observed in our study. We hypothesize that long-term use of ADT may cause adaptation of the immune response. However, further studies are warranted to more fully investigate the relationship between long-term ADT use and immune responses. In the secondary analysis by types of ADT, significantly decreased risks of incident autoimmune diseases were observed in the use of GnRH agonists alone, GnRH agonist + oral antiandrogens, and oral antiandrogens alone compared to non-ADT use.
Yang et al. reported a 23% increased risk of rheumatoid arthritis (RA) in 44,785 patients who received ADT for PCa in North America38. They claimed androgen-mediated thymic regeneration may increase the risk of RA. However, Kili-Drori et al. revealed no association between the use of ADT and the incidence of psoriasis and RA in a conference paper39. In our study, a decreased trend of RA without statistical significance was observed. We hypothesized that reduction of IL-1 beta and TNF-alpha after ADT use may contribute to this phenomenon10,11. In addition, there is considerable disparity in the risk of RA between the different populations. The prevalence of RA is highest in native American and north America populations but very low in Asia40.
We also investigated the incidence rates of the 33 autoimmune diseases in this study (Supplementary Table 3). A comparison between the incidence rates of the three significantly decreased autoimmune diseases among ADT users and the incidence of these diseases reported in worldwide populations is provided in Supplementary Table 7. The incidence rates of these three autoimmune diseases among ADT users were reduced compared with those reported in previous studies41-46.
Supplementary Table 7. Comparison of the Incidence (per 100,000 person-years) of Autoimmune Diseases among ADT Users and that in Worldwide General Populations
Supplementary Table 3. Incidence (per 100,000) of 33 Autoimmune Diseases among ADT users and non-ADT users
One strength of this study investigating the relationship between ADT use and autoimmune diseases is that it is a large-scale nationwide population-based study. The NHI system of Taiwan covers greater than 99% of the total population of Taiwan. In addition, Taiwan’s NHIRD is among the few nationwide databases maintained by an Asian country.
However, several limitations in this study should be noted. First, prostate-specific antigen levels, clinical stages of PCa, and Gleason scores of the patients were not available from the NHIRD database. Thus, it was difficult to further investigate the severity of PCa. Second, all the diagnoses were defined by using the ICD-9 coding system, such that misinterpretations of some data or the misclassification of some diagnoses may have occurred. Third, laboratory data are not available in the NHIRD. Thus, inflammatory markers could not be assessed. Laboratory tests may provide more information for discovering the mechanism between autoimmune diseases and ADT use. Family history of autoimmune diseases for PCa patients was also not available. Finally, this population-based study was a retrospective study; so further prospective studies are needed to fully evaluate the relationship between ADT use and autoimmune diseases.
In conclusion, we demonstrated that ADT use in patients with PCa reduced the risk of autoimmune diseases compared with a general population cohort of 14,444 PCa patients. Further studies are warranted to obtain a better understanding of the relationship between ADT and autoimmune diseases.
Compliance with ethical standards: Conflict of interest - The authors declare that they have no conflict of interest.
Supplementary information The online version of this article (https://doi.org/10.1038/s41391-019-0130-9) contains supplementary material, which is available to authorized users.
Authors: Jui-Ming Liu1,2,3, Cheng-Ping Yu4, Heng-Chang Chuang1, Chun-Te Wu5, Ren-Jun Hsu3,4,61. Division of Urology, Department of Surgery, Taoyuan General Hospital, Ministry of Health and Welfare, Taoyuan, Taiwan
2. Faculty of Medicine, National Yang-Ming University, Taipei, Taiwan
3. Graduate Institute of Life Sciences, National Defense Medical Center, Taipei, Taiwan
4. Department of Pathology and Graduate Institute of Pathology and Parasitology, Tri-Service General Hospital, National Defense Medical Center, Taipei, Taiwan
5. Department of Urology, Chang Gung Memorial Hospital, Keelung, Taiwan
6. Cancer Medicine Center of Hualien Tzu Chi Hospital, Tzu Chi University, Hualien, Taiwan
References:
1. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin. 2016;66:7–30.
2. Huggins C, Hodges CV. Studies on prostatic cancer. I. The effect of castration, of estrogen and androgen injection on serum phosphatases in metastatic carcinoma of the prostate. CA Cancer J Clin. 1972;22:232–40.
3. Cancer Registry Annual Report, 2000–2011. Taiwan: Department of Health, the Executive Yuan, Republic of China; 2003–13.
4. Shahani S, Braga-Basaria M, Basaria S. Androgen deprivation therapy in prostate cancer and metabolic risk for atherosclerosis. J Clin Endocrinol Metab. 2008;93:2042–9.
5. Fijak M, Meinhardt A. The testis in immune privilege. Immunol Rev. 2006;213:66–81.
6. Aragon-Ching JB, Williams KM, Gulley JL. Impact of androgen deprivation therapy on the immune system: implications for combination therapy of prostate cancer. Front Biosci. 2007;12:4957–71.
7. Keating NL, O’Malley AJ, Smith MR. Diabetes and cardiovascular disease during androgen deprivation therapy for prostate cancer. J Clin Oncol. 2006;24:4448–56.
8. Nead KT, Gaskin G, Chester C, et al. Androgen deprivation therapy and future Alzheimer’s disease risk. J Clin Oncol. 2016;34:566–71.
9. Klil-Drori AJ, Tascilar K, Yin H, Aprikian A, Bitton A, Azoulay L. Androgen deprivation therapy and the incidence of inflammatory bowel disease in patients with prostate cancer. Am J Epidemiol. 2016;184:15–22.
10. Kaczmarek P, Pokoca L, Niemirowicz J, Majewska E, Baj Z. Effect of luteinizing hormone-releasing hormone (LHRH) analogue treatment on a cytokine profile in prostate cancer patients. Pharmacol Rep. 2008;60:399–403.
11. Salman H, Bergman M, Blumberger N, Djaldetti M, Bessler H. Do androgen deprivation drugs affect the immune cross-talk between mononuclear and prostate cancer cells? Biomed Pharmacother. 2014;68:21–24.
12. Bureau of National Health Insurance DoH, Executive Yuan. The National Health Insurance Statistics. 2013. http://www.nhi.gov.tw/English/webdata/webdata.aspx?menu=11&menu_id=296&webdata_id=1942&WD_ID=296. Accessed 18 Sep 2014.
13. Wu CY, Chen YJ, Ho HJ, Hsu YC, Kuo KN, Wu MS, et al. Association between nucleoside analogues and risk of hepatitis B virus-related hepatocellular carcinoma recurrence following liver resection. JAMA. 2012;308:1906–14.
14. Liu JM, Chen TH, Chuang HC, Wu CT, Hsu RJ. Statin reduces the risk of dementia in diabetic patients receiving androgen deprivation therapy for prostate cancer. Prostate Cancer Prostatic Dis. 2018. https://doi.org/10.1038/s41391-018-0091-4. [Epubahead of print]
15. Chen SJ, Chao YL, Chen CY, Chang CM, Wu EC, Wu CS, et al. Prevalence of autoimmune diseases in in-patients with schizophrenia: nationwide population-based study. Br J Psychiatry. 2012;200:374–80.
16. Liu JM, Lin PH, Hsu RJ, Chang YH, Cheng KC, Pang ST, et al. Complementary traditional Chinese medicine therapy improves survival in patients with metastatic prostate cancer. Medicine (Baltimore). 2016;95:e4475.
17. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94:496–509.
18. Frakt AB. An observational study goes where randomized clinical trials have not. JAMA. 2015;313:1091–2.
19. De Nunzio C, Kramer G, Marberger M, Montironi R, Nelson W, Schröder F, et al. The controversial relationship between benign prostatic hyperplasia and prostate cancer: the role of inflammation. Eur Urol. 2011;60:106–17.
20. Nelson WG, De Marzo AM, Isaacs WB. Prostate cancer. N Engl J Med. 2003;349:366–81.
21. Zandman-Goddard G, Peeva E. Gender and autoimmunity. Autoimmun Rev. 2007;6:366–72.
22. Morse MD, McNeel DG. Prostate cancer patients on androgen deprivation therapy develop persistent changes in adaptive immune responses. Hum Immunol. 2012;71:496–504.
23. Saylor PJ, Kozak KR, Smith MR, Ancukiewicz MA, Efstathiou JA, Zietman AL, et al. Changes in biomarkers of inflammation and angiogenesis during androgen deprivation therapy for prostate cancer. Oncologist. 2012;17:212–9.
24. Nadiminty N, Lou W, Lee SO, Lin X, Trump DL, Gao AC. Stat3 activation of NF-{kappa}B p100 processing involves CBP/ p300-mediated acetylation. Proc Natl Acad Sci USA. 2006;103:7264–9.
25. Huang S, Pettaway CA, Uehara H, Bucana CD, Fidler IJ. Blockade of NF-kappaB activity in human prostate cancer cells is associated with suppression of angiogenesis, invasion, and metastasis. Oncogene. 2001;20:4188–97.
26. Guzmán-Soto I, Salinas E, Quintanar JL. Leuprolide acetate inhibits spinal cord inflammatory response in experimental autoimmune encephalomyelitis by suppressing NF-κB activation. Neuroimmunomodulation. 2016;23:33–40.
27. Lowes MA, Suarez-Farinas M, Krueger JG. Immunology of psoriasis. Annu Rev Immunol. 2014;32:227–55.
28. Lowes MA, Bowcock AM, Krueger JG. Pathogenesis and therapy of psoriasis. Nature. 2007;445:866–73.
29. Horai R, Caspi RR. Cytokines in autoimmune uveitis. J Interferon Cytokine Res. 2011;31:733–44.
30. Antonelli A, Ferrari SM, Corrado A, Di Domenicantonio A, Fallahi P. Autoimmune thyroid disorders. Autoimmun Rev. 2015;14:174–80.
31. Engl T, Relja B, Blumenberg C, Müller I, Ringel EM, Beecken WD, et al. Prostate tumor CXC-chemokine profile correlates with cell adhesion to endothelium and extracellular matrix. Life Sci. 2006;78:1784–93.
32. Murphy C, McGurk M, Pettigrew J, Santinelli A, Mazzucchelli R, Johnston PG, et al. Nonapical and cytoplasmic expression of interleukin-8, CXCR1, and CXCR2 correlates with cell proliferation and microvessel density in prostate cancer. Clin Cancer Res. 2005;11:4117–27.
33. Taichman RS, Cooper C, Keller ET, Pienta KJ, Taichman NS, McCauley LK. Use of the stromal cell-derived factor-1/CXCR4 pathway in prostate cancer metastasis to bone. Cancer Res. 2002;62:1832–7.
34. Shen H, Lentsch AB. Progressive dysregulation of transcription factors NF-kappa B and STAT1 in prostate cancer cells causes
proangiogenic production of CXC chemokines. Am J Physiol Cell Physiol. 2004;286:C840–847.
35. Morse MD, McNeel DG. T cells localized to the androgen-deprived prostate are TH1 and TH17 biased. Prostate. 2012;72:1239–47.
36. Koh YT, Gray A, Higgins SA, Hubby B, Kast WM. Androgen ablation augments prostate cancer vaccine immunogenicity only when applied after immunization. Prostate. 2009;69:571–84.
37. Khosla S, Atkinson EJ, Dunstan CR, O’Fallon WM. Effect of estrogen versus testosterone on circulating osteoprotegerin and other cytokine levels in normal elderly men. J Clin Endocrinol Metab. 2002;87:1550–4.
38. Yang DD, Krasnova A, Nead KT, Choueiri TK, Hu JC, Hoffman KE, et al. Androgen deprivation therapy and risk of rheumatoid arthritis in patients with localized prostate cancer. Ann Oncol. 2018;29:386–91.
39. Kili-Drori A, Tascilar K, Yin H, Aprikian AG, Azoulay L. Androgen deprivation therapy and the incidence of autoimmune diseases. J Clin Oncol. 2015;33:e16010.
40. Silman AJ, Pearson JE. Epidemiology and genetics of rheumatoid arthritis. Arthritis Res. 2002;4:S265–272.
41. Carlé A, Pedersen IB, Knudsen N, Perrild H, Ovesen L, Rasmussen LB, et al. Epidemiology of subtypes of hyperthyroidism in Denmark: a population-based study. Eur J Endocrinol. 2011;164:801–9.
42. Teng W, Shan Z, Teng X, Guan H, Li Y, Teng D, et al. Effect of iodine intake on thyroid diseases in China. N Engl J Med. 2006;354:2783–93.
43. Bell LM, Sedlack R, Beard CM, Perry CO, Michet CJ, Kurland LT. Incidence of psoriasis in Rochester, Minn, 1980−1983. Arch Dermatol. 1991;127:1184–7.
44. Donker GA, Foets M, Spreeuwenberg P, van der Werf GT. Management of psoriasis in general practice now more in agreement with the guidelines of the Dutch College of General Practitioners (NHG). Ned Tijdschr Geneeskd. 1998;142:1379–83.
45. Gritz DC, Wong IG. Incidence and prevalence of uveitis in Northern California; the Northern California Epidemiology of Uveitis Study. Ophthalmology. 2004;111:491–500.
46. Hwang DK, Chou YJ, Pu CY, Chou P. Epidemiology of uveitis among the Chinese population in Taiwan: a population-based study. Ophthalmology. 2012;119:2371–6.
Read a Commentary by Dr. Andrew J. Armstrong: Androgen Deprivation Therapy for Prostate Cancer and the Risk of Autoimmune Diseases - Commentary