Quite a few epidemiological studies including meta-analyses indicate that prostate inflammation is associated with increased risk of prostate cancer. The cause of inflammation in the prostate is speculated to be several microorganisms that cause prostatitis or sexually transmitted infections. Other specific microorganisms, such as xenotropic murine leukemia virus-related virus, are also reported to relate to the development of prostate cancer; however, the contribution of this microorganism to prostate cancer development needs to be carefully interpreted. Environmental factors, especially dietary factors, might also be associated with prostate cancer development. Among related dietary factors, charred meat carcinogen 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine might be a link between environmental factors and inflammation, because 2-amino-1-methyl-6-phenylimidazo[4,5-b]pyridine has the potential to accelerate prostate inflammation through its estrogenic effect. In light of these findings, preventing or reducing prostate inflammation might be one strategy for chemoprevention of prostate cancer.
The incidence of and mortality rates for prostate cancer are extremely different among regions and races worldwide, with black men in the USA having high rates and Asian men having low rates. Differences in genetic background might partially account for this phenomenon; however, a study of immigrants from Asian countries to Western countries suggests that environmental factors might influence the development of prostate cancer.1,2 In the present review, we focus on prostate inflammation as the environmental contributor to prostate carcinogenesis, and discuss the role of prostate inflammation in prostate carcinogenesis.
Evidence linking inflammation and prostate cancer
Chronic inflammation has long been linked to cancers such as those of the liver, esophagus, stomach, large intestine and urinary bladder.3 Inflammation might influence the pathogenesis of cancers by inflicting cell and genome damage, triggering restorative cell proliferation to replace damaged cells, and elaborating a portfolio of cytokines that promote cell replication, angiogenesis and tissue repair.3
In terms of prostate cancer, several case–control studies and a meta-analysis have shown that there is a significant increase in the relative risk of prostate cancer in men with prostatitis (Table 1).4–19 Daniels et al. carried out a cross-sectional analysis from a prospective cohort study of 5821 men older than 65 years using a self-reported history of prostatitis and prostate cancer, and found a positive association between a history of prostatitis and a history of prostate cancer (OR 5.4, 95% CI 4.4–6.6).6 Dennis et al. showed by a meta-analysis of 11 case–control studies that there is a statistically significant summary OR for prostate cancer of 1.57 for ever having had prostatitis.5 These epidemiological studies inevitably suffer from biases, the two most notable of which are that patients with prostatitis are more likely to be followed by a urologist and thus might be more likely to be evaluated for prostate cancer (detection bias), and patients with prostate cancer might be more likely to remember or be willing to report a previous episode of prostatitis than men without prostate cancer (recall bias).20 These studies are also influenced by the variable quality of prostatitis confirmation and an inability to classify the type of prostatitis or to detect asymptomatic prostatitis.21 To clarify the association between prostatitis and the risk of prostate cancer, additional well-designed epidemiological studies are required.
Table 1. Studies analyzing the association between prostatitis/STI and prostate cancer
|1980||Japan||Niijima et al.||Case–control||187||Records||1.0||NA||NA||NA|| 7 |
|1985||Japan||Mishina et al.||Case–control||100||Questionnaire||1.64||0.69–3.92||NA||NA|| 8 |
|1988||USA||Honda et al.||Case–control||216||Interview||2.2||1.2–4.3||NA||NA|| 9 |
|1993||Japan||Nakata et al.||Case–control||294|| ||4.46||2.71–7.36||NA||NA|| 10 |
|1994||USA||Hiatt et al.||Case–control||238||Records||1.1||0.5–2.3||NA||NA|| 11 |
|1996||USA||Zhu et al.||Case–control||175||Questionnaire||1.1||0.6–2.2||NA||NA|| 12 |
|2000||USA||Hayes et al.||Case–control||981||Interview, serology||NA||NA||Gonorrhea or syphilis 1.6||1.2–2.1|| 13 |
|2004||USA||Roberts et al.||Case–control||409||Records||1.7||1.1–2.6||NA||NA|| 14 |
|2005||USA||Patel et al.||Case–control||700||Interview||1.8||1.1–2.9||1.0||0.7–1.5|| 15 |
|2005||Cuba||Fernandez et al.||Case–control||273||Interview||NA||NA||1.7||1.1–2.5|| 16 |
|2006||USA||Sutcliffe et al.||Cohort (n = 36 033)||2263||Questionnaire||1.08||0.96–1.20||Gonorrhea 1.04, syphilis 1.06||Gonorrhea 0.79–1.36, syphilis 0.44–2.59|| 17 |
|2008||USA||Huang et al.||Case–control||868||Serology, self report||NA||NA||Any STI 1.3||1.0–1.6|| 18 |
|2010||USA||Cheng et al.||Cohort (n = 68 675)||1658||Questionnaire||1.30||1.10–1.54||1.02||0.91–1.15|| 19 |
Pathological evidence: Proliferative inflammatory atrophy as a lesion that connects inflammation and prostate cancer
Epithelial atrophy is divided into two categories: diffuse and focal.22 Diffuse atrophy is a lesion that develops uniformly in the prostate after androgen deprivation. This atrophic lesion consists of a prominent basal cell layer that underlies cuboidal luminal cells. In contrast to diffuse atrophy, focal atrophy develops in the presence of androgen, is seen heterogeneously in the prostate and contains a non-prominent basal cell layer. Focal atrophy was suggested to relate to prostate cancer as early as 1954.22,24 Many of these lesions are associated with acute and chronic inflammation, and occur diffusely in the peripheral zone where prostate cancer predominantly develops.25 Although these lesions show atrophic morphology, epithelial cells of these lesions tend to be proliferative, and De Marzo et al. named this lesion PIA.26 There are morphological transitions between PIA and high-grade PIN lesions, where 40% of high-grade PIN lesions merge directly with PIA,27 which might suggest that PIA is a lesion that links inflammation and prostate cancer development.
GST are inducible enzymes that catalyze the detoxification of reactive electrophiles and oxidants produced from inflammation, and they protect against neoplastic transformation. Luminal columnar cells in areas of PIA show high levels of GSTP1 and GSTA1 expression, suggesting that this lesion is exposed to cellular stress.28,29 Interestingly, the expression of GSTP1, which is a major class of GST, is downregulated by GSTP1 promoter CpG island hypermethylation in prostate cancer.29 Because GSTP1 is a major enzyme that protects cells from oxidative genome damage, one might hypothesize the following: if a specific precursor lesion of prostate cancer exists, such a lesion should lack the ability to induce GSTP1 by promoter hypermethylation, and this lesion would be vulnerable to oxidants and electrophiles, would accumulate genome damage, and eventually would transform into PIN and prostate cancer cells. What is the GSTP1 promoter methylation status in PIA lesions? Nakayama et al. microdissected PIA lesions and normal areas from human prostate cancer specimens, and found that 6.3% of PIA lesions showed methylation in the GSTP1 CpG island, whereas no normal epithelium showed methylation in the GSTP1 CpG island.30 In addition, they analyzed high-grade PIN lesions and adenocarcinoma lesions from non-microdissected specimens, and found that 68.8% (22/32) of high-grade PIN lesions and 90.9% (30/33) of adenocarcinoma lesions showed methylation in the GSTP1 CpG island. This finding might also support the hypothesis that PIA is a lesion that connects inflammation and prostate cancer. Although the prevalence of hypermethylation of the GSTP1 CpG island in PIA appears to be low, PIA lesions usually spread widely in the peripheral zone in elderly men, and the absolute area of PIA that contains hypermethylated GSTP1 CpG islands is not small and might have a role in tumorigenesis. PIA, PIN and prostate cancer share other genetic characteristics as well. We found that in PAH lesions, a variant of PIA, mutations in the p53 gene occurred at a similar frequency (5%) in both PAH and high-grade PIN, but did not occur in normal epithelium.31
As the importance of focal atrophy lesions is recognized, it will be necessary to define them clearly. De Marzo et al. proposed criteria for subclassification of focal atrophy.22 In this report, they classified focal atrophy into four patterns: (i) simple atrophy; (ii) PAH; (iii) simple atrophy with cyst formation; and (iv) partial atrophy. In this classification system, simple atrophy and PAH are considered to be PIA. This classification was validated by 34 pathologists with the median kappa value being 0.80. A future study examining the role of atrophic lesions in prostate carcinogenesis should use this classification to avoid confusion in the definition and terminology of focal atrophy.
Types of prostatitis and prostate cancer
Prostatitis is an extremely common condition in adult men. There are several types of symptomatic and asymptomatic prostate inflammations, and the definition of prostate inflammation needs to be defined clearly to examine the association of prostate inflammation in prostate carcinogenesis. The National Institutes of Health issued a consensus statement on the classification of prostatitis in 1999 to formalize the diagnostic criteria for prostatitis.32 According to these criteria, prostatitis is divided into four categories: (i) acute bacterial prostatitis (category I); (ii) chronic bacterial prostatitis (category II); (iii) CP/CPPS (category III); and (iv) asymptomatic inflammatory prostatitis (category IV). CP/CPPS (category III) is the most common prostatitis syndrome, comprising 90–95% of prostatitis cases.33 Its clinical presentation includes symptoms of pain or discomfort in the pelvic region and possibly urinary or ejaculatory symptoms. This condition is extremely common in adult men: 2–10% of adult men suffer from symptoms compatible with CP at any time, and approximately 15% of men suffer from symptoms of prostatitis at some point in their lives.34 There is no apparent racial difference in the prevalence of CP.34 There are two subgroups in this category. Patients with white blood cells in their prostatic fluid, post-prostate massage urine or seminal fluid are classified in the inflammatory category (category IIIA), and patients without white blood cells in any of these samples are classified in the non-inflammatory category (category IIIB). Even though microorganisms are not detected after bacterial culture, probably because of inhibitory substances in semen, prostate fluid and prostate tissue, there is empirical support for a potential role of genitourinary tract colonization and/or infection in CP/CPPS; for example, Chlamydia trachomatis, Ureaplasma urealyticum, protozoan pathogen Trichomonas vaginalis, Neisseria gonorrhea, herpes simplex types I and II, and cytomegalovirus.35
Asymptomatic inflammatory prostatitis (category IV) is diagnosed in men undergoing evaluation for other genitourinary tract problems, such as infertility and elevated PSA. The prevalence of this category is unknown, but the prevalence of category IV prostatitis was estimated to be as high as 11.2–32.3% in a population of men with elevated PSA.36,37 The association between the prostatitis subdivided into these categories and prostate cancer is not well analyzed, as aforementioned. Roberts et al. reported that acute chronic bacterial prostatitis (category I or category II) was associated with prostate cancer, whereas CPPS (category III) was not associated with prostate cancer at all.4 In terms of category IV prostatitis, there is no report relating this specific type of prostatitis with prostate cancer because of the difficulty in defining the incidence of this type of prostatitis in a non-selected population-based study.38
Because prostatitis (symptomatic or asymptomatic) increases serum PSA level,39,40 elevated PSA in men with negative prostate biopsy indicates that these men had prostatitis at some point. Recently, it was reported that men with an elevated baseline PSA level at a young age have a greater risk of prostate cancer diagnosis during the next several decades.41 One possible interpretation of this phenomenon might be that prostatitis, which causes an elevated serum PSA level, leads to future prostate carcinogenesis. From this interpretation, it might be possible that treatment of prostatitis with antibiotics could help to reduce the development of prostate cancer.
Causes of prostatic inflammation
Sexually transmitted infections and prostate cancer
The microorganisms related to CP/CPPS are also known to cause STI, which have also been hypothesized to increase the risk of prostate cancer. Several case–control studies investigating the association of STI with prostate cancer are summarized in Table 1. There are two articles on meta-analyses that analyzed the association between the risk for prostate cancer and history of STI. Dennis et al. analyzed 23 case–control studies and calculated a RR of prostate cancer of 1.44 (95% CI 1.24–1.66) for a history of any STI, 2.30 (1.34–3.94) for a history of syphilis and 1.36 (1.15–1.61) for a history of gonorrhea.42 Taylor et al. analyzed 29 case–control studies and calculated a RR of prostate cancer of 1.48 (95% CI 1.26–1.73) for a history of any STI, 1.35 (1.05–1.83) for a history of gonorrhea and 1.39 (1.12–2.06) for a history of HPV infection.43 Studies included in these meta-analyses assessed STI exposure by self-report of STI history, and the majority featured case–control designs with retrospective assessment of STI exposure, thus allowing for the potential of recall bias and detection bias. Furthermore, none of these studies accounted for exposure to multiple STI or to repeated infections with the same organism. The possibility also exists for the subjects to have had an asymptomatic infection, especially with Chlamydia or HPV.43
More recently, epidemiological studies have begun to investigate associations between individual STI and prostate cancer by serology, but there is no conclusive consensus on an association between specific microorganisms and prostate cancer. Some articles reported positive association between prostate cancer risk and circulating antibodies against syphilis,44C. trachomatis,45 HSV-2,46 and HPV serotypes 1647,48 and 18,34 which are two of the high-risk types for cervical cancer, whereas the other articles reported no detected associations for HPV serotypes 11, 16 or 23.18,44,47,49,50 Studies were null for antibodies against Chlamydia species45 and herpesvirus 851 and 2,52,53 whereas Hoffman et al. observed significantly higher odds of HHV-8 seropositivity in prostate cancer patients than in controls.54 The inconsistency of these results might be caused by many confounding factors related to STI. Sutcliffe et al. found several correlates of STI history, such as African–American race, foreign birth, southern residence, cigarette smoking, greater alcohol consumption, higher ejaculation frequency, vasectomy and high cholesterol.55 In investigation of STI and later chronic disease, such as benign prostatic hyperplasia and prostate cancer, these confounders should be taken in account.
Specific microorganisms in prostate cancer
Are there any other specific pathogens that cause prostate inflammation and eventually result in prostate cancer? One candidate is Propionibacterium acnes, a bacterium that lives on the skin and thrives in blocked follicles and which is associated with acne. Cohen et al. detected P. acnes from 35% of prostate cancer tissue obtained from prostatectomy samples, and a significantly higher degree of prostatic inflammation was observed in cases culture positive for P. acnes.56 By using prostate samples from benign prostatic hyperplasia patients, Alexeyev et al. found that P. acnes is the most prevalent bacterium in this disease, and that its prevalence was higher in samples from patients subsequently diagnosed with prostate cancer.57 When P. acnes was cocultured with the prostate epithelial cell line RWPE1, cytokines and chemokines, such as IL-6 and IL-8, respectively, were secreted from infected cells, and long-term exposure to P. acnes altered cell proliferation and enabled anchorage-independent growth of infected epithelial cells, thus initiating cellular transformation.58 Sutcliffe et al. showed that severe acne, as measured by tetracycline use for four or more years, is associated with prostate cancer risk.59 Although there are other epidemiological studies that show either a negative or no correlation between acne and the development of prostate cancer,60–62 and the role of P. acnes in prostate carcinogenesis remains to be elucidated, P. acnes might have at least some role in the induction of prostatitis.
Individuals with a single mutated copy of the RNASEL gene (bearing mutation R462Q) are reported to be associated with increased risk of prostate cancer.63 The RNASEL gene encodes RNase L, which is an endoribonuclease that functions in the molecular pathways of interferon action against viral infections.64 The R462Q substitution in the RNASEL gene produces a threefold decrease in the catalytic activity of the parental wild-type enzyme,63 which leads to decreased natural defense against viruses. Is there a virus associated with prostate cancer just as the HPV is associated with cervical cancer? In 2006, it was reported that XMRV was identified in prostate tumors predominantly of patients homozygous for the R462Q RNASEL variant.65 After this finding was published, there were several reports that confirmed the presence of XMRV in prostate cancer.65,67 However, there are also several reports in which XMRV could not be or was rarely detected in prostate specimens68–70 or in the blood from prostate cancer patients.71 Aloia et al. tested nearly 800 prostate cancer samples using a combination of real-time PCR and immunohistochemistry, and there was no evidence for XMRV in the prostate cancer samples they obtained.72 Recently, it was reported that XMRV was generated from recombination of two endogenous murine retroviruses during the passage of the prostate cancer xenograft in mice, and there was a possibility that positive results of XMRV presence, which was detected by PCR, might result from contamination of mouse DNA.73,74 Furthermore, Sato et al. reported that some commercially available PCR-kits are contaminated with an endogenous murine leukemia viral genome.75 At present, it is recognized that the causal role of XMRV in prostate tumorigenesis is small.
Environmental factors and prostate inflammation
In Japan, the incidence of and mortality rate for prostate cancer are still low compared with those in Western countries, but they are increasing dramatically; in the past 30 years, the incidence and mortality rate have increased by 5.8-fold and 2.0-fold, respectively (Center for Cancer Control and Information Services, National Cancer Center, Japan). Factors that account for this increase are speculated to be several environmental factors, which are strongly suggested from a study of men immigrating from Japan to Hawaii or California.1,2 In a review by De Marzo et al., several contributors to prostate inflammation were enumerated including infectious agents, hormonal changes, physical trauma, urine reflux and dietary habits.76 Among these environmental factors, diet is the factor that has changed dramatically in Japan in these past 50 years. A traditional Japanese diet used to consist of rice served with meals high in soybean products and fish, and low in red meat. Within 50 years, the Japanese diet has completely changed. There has been a 4.5-fold increase in the intake of animal fat and a 4.0-fold increase in the intake of animal protein, whereas there was only a slight (1.1-fold) increase in total energy. If there is a factor in the diet that relates to the increase in prostate cancer in Japan, it might be something included in the animal products. There are several case–control and prospective studies that suggest a relationship between red meat consumption and prostate cancer risk. Wright et al. carried out a population-based prostate cancer case–control study comprising 1309 cases and 1267 controls, and found that red meat consumption was positively associated with prostate cancer risk (OR 1.43, 95% CI 1.11–1.84).77 A prospective study consisting of 65 548 men with a follow-up period of 11 years resulted in 5113 prostate cancer cases, and indicated that red meat consumption was significantly associated with a higher risk of prostate cancer among black men (RR 2.0, 95% CI 1.0–4.2), but not among white men.78 There is yet another prospective study comprising 175 343 USA men with a follow up of 9 years. There were 10 313 prostate cancer cases, and the authors noted that consumption of red meat was associated with elevated prostate cancer risk (hazard ratio 1.07, 95% CI 1.00–1.14). Furthermore, they found that intake of barbecued/grilled meat was also associated with elevated prostate cancer risk (hazard ratio 1.09, 95% CI 1.02–1.17).79 It is known that high-temperature cooking methods, such as grilling, barbecuing or pan-frying, produce multiple mutagens and carcinogens, such as HCA, polycyclic aromatic hydrocarbons and N-nitroso compounds, and thus, methods of cooking meat might relate to prostate cancer risk. From a population-based case–control study, John et al. reported that grilled and well-done meat is associated with increased risk of prostate cancer.80 There are three prospective studies indicating that meat cooked at high temperature is associated with increased risk of prostate cancer.81–83 Cross et al.81 and Koutros et al.82 further analyzed estimated intake of HCA and polycyclic aromatic hydrocarbons by using the National Cancer Institute's CHARRED database (http://charred.cancer.gov) and found that intake of HCA, such as PhIP, MeIQx and DiMeIQx is associated with increased risk of prostate cancer. Although there is one prospective study that could not find an association between specific HCA intake and prostate cancer,84 some HCA, such as PhIP and MeIQx, have been shown to cause prostate cancer in a rodent model,85–87 and they might have some role in the development of human prostate cancer.
HCA: Possible link between diet, inflammation and prostate cancer
HCA are formed by pyrolysis of amino acids in the presence of creatine and carbohydrate during the cooking of a variety of muscle meats. HCA are reported to be present in the parts per billion (ng/g) range in meats cooked by ordinary household methods,86,88 and the most abundant HCA is PhIP followed by MeIQx and DiMeiQx.89 The formation of HCA in meats depends on cooking temperature and duration. After absorption of HCA by the intestine, they are metabolically activated by hepatic cytochrome P450 1A2-mediated N-hydroxylation followed by O-esterification of the N-hydroxylamines, which are catalyzed by N-acetyltransferases and/or sulfotransferases. These reactive ester derivatives covalently bind to guanine residues of genomic DNA to form DNA adducts, and eventually cause mutation.90
Shirai et al. fed male F344 rats with PhIP at a dose of 400 ppm for 52 weeks and found that 66.7% of these rats developed carcinoma in the ventral prostate.85 Interestingly, the carcinoma induced by PhIP is lobe specific, just as is human prostate cancer, which preferentially develops in the peripheral zone.12 We thought that by pursuing the mechanism of lobe-specific development of prostate cancer, we could find a clue to understanding the mechanism of prostate cancer development.91 We took advantage of Big Blue transgenic F344 rats, which have the λLIZ shuttle vector that includes the cII gene, which is the target for the mutagenesis studies. By collecting the genomic DNA from each tissue of interest, we could quantify MF of the cII gene with the λSelect-cII mutation detection system (Stratagene, La Jolla, CA, USA) and could compare MF in each tissue of interest. We separately collected genomic DNA from the ventral, dorsolateral and anterior lobes of the prostate, and compared MF. Surprisingly, MF of all prostate lobes were increased after treatment with PhIP, suggesting that all prostate lobes were targets of PhIP-induced mutation. Considering that the ventral lobe of the prostate is the only lobe that develops carcinoma from PhIP, PhIP-induced mutation is a necessary, but not sufficient, condition.
What other factors contribute to prostate carcinogenesis? Because we have been interested in inflammation in prostate carcinogenesis, we analyzed infiltration of inflammatory cells after 8 weeks of PhIP treatment. Although there was no difference in the number of inflammatory cells related to acquired immunity, such as lymphocytes, there was a significant increase in inflammatory cells related to innate immunity, such as mast cells, specifically in the ventral prostate. Because mast cells are an important cellular source of tumor necrosis factor-α in peripheral tissue,92 we postulated that an increase in mast cells in response to PhIP treatment might make the ventral prostate prone to acute/chronic prostatitis. To confirm this hypothesis, we treated F344 male rats with PhIP and induced prostatitis by transurethral injection of lipopolysaccharide, after which half of the rats died from severe prostatitis in the PhIP-treated group, and no rats died in the saline-treated group (unpubl. data). These results suggest that PhIP promotes prostate inflammation by increasing mast cell number in the prostate lobe that develops into prostate carcinoma.
Estrogenic hormones and prostate inflammation
The promotion of prostate inflammation by PhIP might be explained by its estrogenic effect. By using COS-1 cells transiently transfected with an estrogen-responsive reporter gene, Lauber et al. found that PhIP mediated transcription through estrogen receptor α, but not β, and this transcription was inhibited by the pure estrogen receptor antagonist ICI 182 780.93 It is known that estrogen induces non-bacterial prostatitis in rats.94,95 Although the exact mechanism by which this occurs has not been clearly elucidated, Harris et al. showed that treatment of male Wistar rats with estradiol-17β upregulated mRNA of IL-1β, IL-6, macrophage inflammatory protein-2, and inducible nitric oxide synthase in the prostate, which might account for the promotion of inflammation in the prostate by estrogen.96 Recently, it was reported that endogenous estrogen leads to sequential induction of prostatic chronic inflammation and prostatic premalignancy.97 In that report, the authors developed mice with endogenous elevated estrogen levels as a result of aromatase overexpression and found that the prostates of these mice had chronic inflammation by 40 weeks-of-age. Interestingly, an increase in mast cells in the prostate preceded chronic inflammation in this model, suggesting the importance of mast cells in the development of estrogen-induced prostatitis. After prostatitis, premalignant prostatic intra-epithelial neoplasia lesions emerged by 52 weeks-of-age, which appears to link estrogens to prostatitis and premalignancy in the prostate.
From these findings, mast cells appear to have a role that connects diet, inflammation and prostate cancer, and such a role of mast cells in human prostate cancer was recently reported. After evaluating prostate needle biopsy specimens from 104 prostate cancer patients, Nonomura et al. reported that an increased number of mast cells is associated with poor prognosis.98 Johansson et al. analyzed peritumoral and intratumoral mast cells in the prostate specimens of prostate cancer patients who underwent watchful waiting and found that peritumoral, not intratumoral, mast cells stimulate the expansion of human prostate tumors.99 From these results, mast cells might have some role in the biology of human prostate cancer.
Inflammation, possible target for prostate cancer prevention
From the laboratory and epidemiological findings aforementioned, it is assumed that by quenching inflammation in the prostate, the development of prostate cancer might be prevented. Cyclooxygenase enzymes (COX 1 and COX 2) are responsible for formation of key mediators of inflammation, such as prostaglandin, prostacyclin and thromboxane. By inhibiting COX enzymes with NSAIDs, such as aspirin and ibuprofen, inflammation is attenuated.
Epidemiological studies that examined the association between use of aspirin and other NSAIDs, and the risk of prostate cancer are summarized in Table 2.100–124 Most of these reports suggest an inverse relationship between use of NSAIDs and the risk of prostate cancer. Mahmud et al. carried out a systematic meta-analysis of epidemiological studies where the outcome was prostate cancer incidence or mortality and the exposure was use of NSAIDs. They found that aspirin use decreased the total risk of prostate cancer by 17% (95% CI 0.77–0.89) and decreased the risk of advanced prostate cancer by 19% (95% CI 0.72–0.92). In terms of non-aspirin NSAIDs, although it was not significant, there was a 19% decrease in risk for total prostate cancer (95% CI 0.78–1.02).125 As shown in Table 2, the results of reports investigating the association of NSAIDs with the risk of prostate cancer seem to be inconsistent. The reason for this inconsistency might be that doses and treatment durations of the drugs differ between the studies, and screening bias in which NSAID users are more likely to be screened for prostate cancer, possibly because of more frequent contact with healthcare providers, might have some influence on the results. Another reason might be that with one exception, duration of exposure to NSAIDs is too short. It is widely accepted that any potential effects of NSAID use on cancer incidence are likely to involve a considerable induction period.126 Well-designed observational studies with adequate exposure measurements, sufficient duration of exposure and careful adjustment for screening are required.
Table 2. Studies analyzing the association between use of NSAIDs and prostate cancer
|1998||Europe||Case–control||317||Questionnaire||NSAIDs and aspirin||0.88||0.64–1.20|| 100 |
|1998||USA||Case–control||319||Records||Aspirin||1.60||0.82–3.11|| 101 |
|2000||USA||Case–control||417||Interview||NSAIDs||0.34||0.23–0.58|| 102 |
|2000||UK||Case–control||570||Database||NSAIDs||1.34||1.09–1.64|| 103 |
|2003||Canada||Case–control||2221||Database||Aspirin||0.82||0.71–0.95|| 104 |
|2004||Europe||Case–control||296||Database||Ibuprofen||2.18||1.94–2.45|| 105 |
|2004||UK||Case–control||2183||Database||Aspirin||0.70||0.61–0.79|| 106 |
|2006||USA||Case–control||1029||Questionnaire||Aspirin||1.05||0.89–1.25|| 107 |
|2006||Europe||Case–control||1261||Interview||Aspirin||1.10||0.81–1.50|| 108 |
|2006||USA||Case–control||506||Interview||NSAID and aspirin||0.67||0.52–0.87|| 109 |
|2006||Canada||Case–control||494||Interview||Aspirin||0.58||0.36–0.91|| 110 |
|2006||Canada||Case–control||2025||Database||NSAIDs and COX2 inhibitor||0.71||0.58–0.86|| 111 |
|2010||USA||Case–control||1001||Interview||Aspirin||0.79||0.65–0.96|| 112 |
|2011||Canada||Case–control||9007||Database||Propionates||0.90||0.84–0.95|| 113 |
|1989||USA||Cohort (n = 22 781)||149||Questionnaire||Aspirin||0.90||0.63–1.30|| 114 |
|1994||USA||Cohort (n = 12 668)||123||Interview||Aspirin||0.95||0.66–13.5|| 115 |
|2002||USA||Cohort (n = 47 882)||2479||Questionnaire||Aspirin||1.05||0.96–1.14|| 116 |
|2002||USA||Cohort (n = 90 100)||2574||Questionnaire||Aspirin||0.76||0.60–0.98|| 117 |
|2002||USA||Cohort (n = 1 362)||91||Interview||NSAIDs||0.45||0.28–0.73|| 118 |
|2003||Europe||Cohort (n = 172 057)||324||Database||NSAIDs||1.30||1.16–1.45|| 119 |
|2004||USA||Cohort (n = 14 470)||121||Interview||Aspirin||1.11||0.60–2.05|| 120 |
|2005||USA||Cohort (n = 70 144)||4853||Questionnaire||Aspirin and NSAIDs||0.95||0.86–1.05|| 121 |
|2005||USA||Cohort (n = 1 244)||141||Database||Aspirin and NSAIDs||0.71||0.49–1.02|| 122 |
|2011||USA||Cohort (n = 78 485)||8092||Questionnaire||Acetaminophen||0.62||0.44–0.87|| 123 |
|2011||USA||Cohort (n = 51 529)||4858||Questionnaire||Aspirin||0.90||0.80–1.02|| 124 |
Quite a few epidemiological studies suggest a link between prostate inflammation and prostate cancer. Although it is unclear whether the dramatic increase in prostate cancer incidence and mortality in Japan is related to increase of inflammation in the prostate, dramatic changes in dietary habit, such as increased consumption of cooked meats, might influence the development of prostate cancer in the Japanese population through its effect in promoting prostate inflammation. Targeting prostate inflammation might be an attractive way to reduce or prevent prostate cancer. NSAIDs and COX-2 inhibitors are reported to be candidates for prostate cancer prevention, but the results are inconsistent. At present, new strategies other than chemoprevention are required. Targeting of dietary habits might be another way to prevent prostate cancer by attenuating prostate inflammation.