Association Between Statin Exposure and Incidence and Prognosis of Prostate Cancer

Abstract

It is widely thought that statins have huge therapeutic potential against prostate cancer (PCA). This study aimed to investigate the effect of statin exposure on PCA incidence and prognosis. PubMed, Web of Science, Embase, and Cochrane databases were searched for observational studies on the association between statin exposure and PCA from inception until July 2022. The primary endpoints were the incidence of PCA and the survival rate. A total of 21 studies were included in this meta-analysis. The pooled estimates showed that exposure to hydrophilic statins was not associated with the incidence of PCA (odds ratio [OR]=0.94, 95% CI=0.88-1.01, P=0.075), while the incidence of PCA was significantly decreased in populations exposed to lipophilic statins compared with the nonexposed group (OR=0.94, 95% CI=0.90-0.98, P=0.001), mainly in Western countries (OR=0.94, 95% CI=0.91-0.98, P=0.006). Subgroup analysis showed that simvastatin (OR=0.83, 95% CI=0.71-0.97, P=0.016) effectively reduced the incidence of PCA. The prognosis of PCA in patients exposed to both hydrophilic (hazard ratio [HR]=0.57, 95% CI=0.49-0.66, P<0.001) and lipophilic (HR=0.65, 95% CI=0.58-0.73, P<0.001) statins were better than in the nonexposed group, and this improvement was more significant in the East than in Western countries. This study demonstrates that statins can reduce the incidence of PCA and improve prognosis, and are affected by population region and statin properties (hydrophilic and lipophilic).

Prostate cancer (PCA) is currently the second most common cancer in men worldwide and the sixth leading cause of cancer mortality in 2020.¹ The disease burden of PCA is projected to increase to 499,000 deaths and 1.7 million new cases by 2030.² The incidence and mortality of PCA vary greatly among different countries and regions. In Eastern countries, the incidence of PCA is lower, but the mortality is higher, while the incidence and mortality of PCA are opposite in Western countries.³ In addition to family history, genetic predisposition, and race, occupational and environmental risk factors are also considered.² Currently available therapeutic approaches against PCA include radical prostatectomy, pelvic lymph node dissection, androgen deprivation therapy, and radiation therapy.⁴ However, there is a certain risk of erectile dysfunction and urinary incontinence, resulting in decreased quality of life after surgery.⁵ Therefore, reducing the incidence and improving the prognosis of this patient population is very important.

Statins are drugs that are effective in improving cardiovascular outcomes.⁶ The past decade has witnessed unprecedented discoveries, including the potential of statins to fight against cancers, such as gastric cancer,⁷ liver cancer⁸ and so on, but the effects of statins in the prevention and treatment of PCA remain unclear. Although some meta-analyses^9–12 found no association between statin exposure and the incidence of PCA, others¹³ indicated huge potential in reducing the risk of PCA. Given that statins can be classified into different groups based on their hydrophilic and lipophilic properties, there may be a correlation between the anticancer effects of statins and their properties. For example, one study showed that lipophilic statins effectively reduced the incidence of gastric cancer, while exposure to hydrophilic statins did not.⁷ Only one meta-analysis examined the association between statins (hydrophilic or lipophilic) and the risk of PCA at present, and the results were not statistically significant.¹² In addition, no studies have hitherto reported the effect of different properties (hydrophilic or lipophilic) of statins on the prognosis of PCA patients. Furthermore, there are regional differences in the incidence and prognosis of PCA. Accordingly, we sought to conduct a meta-analysis based on previously published observational studies to investigate the relationship between statin exposure and PCA incidence and survival. Importantly, further analysis is performed according to the properties of statins (hydrophilic or lipophilic) in different regions.

METHODS

Search Strategy

This meta-analysis searched PubMed, Web of Science, Embase, and Cochrane databases for literature studies on the effects of the properties of statins (hydrophilic or lipophilic) exposure on the incidence or the survival rate of PCA as of July 2022, according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.¹⁴ We used a combination of subject words and free words to complete the formulation of search terms, which included the key words (“Hydroxymethylglutaryl CoA Reductase Inhibitors” [Mesh] OR “Inhibitors, Hydroxymethylglutaryl-CoA Reductase” OR “Reductase Inhibitors, Hydroxymethylglutaryl-CoA” OR “HMG-CoA Reductase Inhibitor” OR “Statin” OR “Statins, HMG-CoA” OR “HMG-CoA Statins” OR “Statins, HMG CoA” OR “Inhibitors, HMG CoA Reductase” OR “Reductase Inhibitors, HMG-CoA” OR “HMG CoA Reductase Inhibitors” OR “Inhibitors, Hydroxymethylglutaryl-Coenzyme A” OR “Hydroxymethylglutaryl-Coenzyme A Inhibitors” OR “Inhibitors, Hydroxymethylglutaryl Coenzyme A” OR “Inhibitors, Hydroxymethylglutaryl-CoA” OR “Hydroxymethylglutaryl-CoA Inhibitors” OR “Hydroxymethylglutaryl-CoA Reductase Inhibitor” OR “Reductase Inhibitor, Hydroxymethylglutaryl-CoA” OR “THRaST” OR “atorvastatin” OR “bervastatin” OR “cerivastatin” OR “crilvastatin” OR “compactin” OR “dalvastatin” OR “fluindostatin” OR “fluvastatin” OR “glenvastatin” OR “lovastatin” OR “mevastatin” OR “pitavastatin” OR “pravastatin” OR “rosuvastatin” OR “simvastatin” OR “tenivastatin” OR “astatin”) AND (“Prostatic Neoplasms” [Mesh] OR “Prostate Neoplasms” OR “Neoplasm, Prostate” OR “Neoplasms, Prostatic” OR “Prostate Cancers” OR “Cancers, Prostate” OR “Cancer of the Prostate” OR “Prostatic Cancers” OR “Cancers, Prostatic” OR “Cancer of Prostate” OR “prostatic carcinoma” OR “CRPC”). The retrieved articles were screened by two researchers independently. Besides, the references of included articles were also searched. The PROSPERO registration number for this meta-analysis is CRD42022360390.

Inclusion and Exclusion Criteria

This meta-analysis selected related studies exploring statin exposure with the incidence or/and the survival rate of PCA. When different papers based on the same work were available, we selected the most recent or more comprehensive one. The inclusion criteria involved: (1) observational studies (cohort or case-control studies); (2) comparative studies of the effects of statin exposure of different properties (hydrophilic or lipophilic) on the incidence or survival rate of PCA compared with people without statin exposure; (3) relevant data available.

The exclusion criteria included: (1) documents not written in English; (2) unavailable for full text; (3) the association of statin properties (hydrophilic or lipophilic) with PCA incidence or survival rate was not reported; (4) not containing interesting outcomes (eg, odds ratio [OR] for PCA incidence and hazard ratio [HR] for survival rate with hydrophilic or lipophilic statins vs. nonstatin use).

Data Extraction and Quality Assessment

Two researchers extracted the relevant data independently. A third researcher would extract the data under the premise of discrepancies in data extraction. The following information was involved: author, publication year, study period, country, administration time, adjustment factor, and inclusion and exclusion criteria. Data were subgroup analyzed by countries and type of statins. The observational studies were assessed by the Newcastle-Ottawa Quality Assessment Scale (NOS) checklist. Each study was rated on a scale of 0 to 9, with a score of >6 considered a high-quality article.

Statistical Analysis

This meta-analysis was conducted using STATA 12.0 software. I ² and P-value were used to evaluate the heterogeneity among the included articles. Based on the criteria I ² >50 and P-value ≤0.1, the study was considered to have high heterogeneity. Due to the differences in study design, baseline information of patients, and the type and exposure dose of statins, the random-effects model was adopted to improve the robustness of this study. In the meantime, OR and 95% CIs were used to analyze the effect of statin exposure with different properties (hydrophilic or lipophilic) on the incidence of PCA, while HR and CI were used to analyze the effect of statin exposure with different properties (hydrophilic or lipophilic) on the survival rate of PCA. A P-value <0.05 was statistically significant. Sensitivity analysis was used to test the results’ stability, and Begg test was used to estimate publication bias.

RESULTS

Study Selection

A total of related 1998 studies were initially retrieved from the four databases based on the formulated search terms. After excluding duplicates (n=422) and studies that did not meet the requirements after browsing the title and abstract (n=1534), the remaining 42 studies were read in full text, and 21 did not meet the requirements were excluded. Finally, only 21 studies were included in this meta-analysis.^15–35 Among these, 14^15–28 investigated the association between exposure to statins of different properties (hydrophilic or lipophilic) and the incidence of PCA, and 7^29–35 explored the association between exposure to statins with different properties (hydrophilic or lipophilic) and PCA survival. Details of the study screening process are provided in Figure 1.

FIGURE 1.

A schematic flowchart for the selection of articles included in this meta-analysis.

Study Characteristics and Quality Evaluation

A total of 14 studies reported the association between statin exposure and PCA incidence in the United States (n=7), United Kingdom (n=2), Japan (n=1), Denmark (n=1), Finland (n=1), Canada (n=1), and China (n=1). There were 6 cohort studies and 8 case-control studies covering the period from 1987 to 2016. The inclusion criteria were primarily newly histologically confirmed PCA treated with at least one statin prescription, while the main exclusion criteria were a history of previous PCA or other cancers. In most cases, the statin course specified in the studies was longer than 6 months. Most analyses included adjustments for age and race, and some included body mass index, smoking, aspirin, diabetes, and alcohol use. The remaining details and characteristics are given in Table 1 and Supplementary Table 1 (Supplemental Digital Content 1, http://links.lww.com/AJCO/A467).

TABLE 1.

Characteristics of Included Studies in the Meta-analysis of the Correlation Between Statins and the Incidence of PCA

References	Country	Period	Inclusion criteria	Exclusion criteria	Administration time/dose of statins	Adjustment factor	Research type
Agalliu et al15	USA	2002.1.1-2005.12.31	Histologically confirmed PCA	Nonparticipation or nonresponse of the interview	≥3 mo	Age, race, PCA screening	Case control study
Boudreau et al16	USA	1990.1.1-2005.8.31	Group Health’s integrated group practiced ≥2 y; aged 45-79; residing in 1 of 13 Washington counties; no prior PCA	—	≥365 d	Age, diabetes, NSAID hypercholesterolemia, other lipid-lowering drug	Cohort study
Coogan et al17	USA	1987-2001	Histologically confirmed PCA within the previous year; no prior cancer	—	≥365 d	Age, race, year of interview, study center, education, number of doctor visits 2 y before hospitalization, religion, alcohol, BMI	Case control study
Fowke et al18	USA	2002-2010	Age ≥40; histologically confirmed PCA; no prior PCA	Uncertain pathology diagnosis; exogenous testosterone or steroid reductase inhibitor; BMI ≤18.5; prostate volume or PSA data were missing	—	Age, race, family history, BMI, WHR, height, aspirin, treatment for CVD, diabetes, BPH, PSA (continuous), prostate volume (continuous), number of PSA tests (1, 2, >2)	Case control study
Fujimoto et al19	JAPAN	2005.1-2013.7	First prescription after July 2005; histologically confirmed PCA after July 2005	Diagnosed at the time of statin exposure	—	Temporal trends in statins and events	Case control study
Haukka et al20	Finnish	1996-2005	Purchased at least one prescription of any statin; no prior PCA	—	≥5 y	Sex, age, group (statin user on nonuser)	Cohort study
Hippisley-Cox et al21	UK	2002.1.1-2008.6.30	Using the computer-based EMIS for at least a year	Without a postcode-related Town-send score; prescribed statins before	—	—	Cohort study
Ho, Wei et al22	China	2001.1.1-2008.12.31	Aged 55-100; newly IHD diagnosis	Diagnosed with any cancers; IHD diagnosis within 1 y before cohort entry date; used statin 180 d before the cohort entry date	—	Comorbidities, comedication, number of outpatient visits, number of hospitalizations	Cohort study
Jespersen et al23	Denmark	1997-2010	Histologically confirmed PCA	—	—	Age, comorbidities, aspirin, nonaspirin NSAID, education	Case control study
Lustman et al24	USA	2001.1.1-2009.12.31	Enrolled with the Central Region of Clalit Health Services; aged 45-85	Histologically confirmed PCA; uncertain pathology diagnosis; incomplete follow-up	≥6 mo	Age, immigrant, income	Cohort study
Righolt et al25	Canada	2000-2014	Registered with MH during 2000-2014 with ≥5 y of insurance coverage; age ≥40	—	—	Age, regional health authority of residence, length of drug use coverage, income, number of doctor visits 5 y before the index date, screening indicator, chronic cardiovascular disease (excluding hypertension), diabetes, other lipid-lowering drug, metformin, other oral hypoglycemic drugs, insulin, aspirin, non-aspirin NSAID, other individual statins	Case control study
Shannon et al26	USA	1997.5-2004.8	Prostate biopsy in the PV AMC	Prior PCA; another cancer; cognitive impairment	≥2 mo	Age, race, BMI, NSAID, diabetes, caloric intake, other lipid-lowering drug	Case control study
Vinogradova et al27	UK	1998.1.1-2008.7.1	Any cancer	Secondary cancers; diagnosed with cancer before index date; record of mastectomy or prescriptions for tamoxifen; fewer than 6 y of medical records before index date for the main analysis and fewer than 10 y for the further analysis	—	Townsend quintile, BMI, smoking, MI, CHD, diabetes, hypertension, stroke, RA, NSAID, Cox2-inhibitors, aspirin	Case control study
Wang et al28	USA	1994.11-2016.1	At least one visit to the urologic clinic because of any prostatic conditions	Aged ≤18; follow‐up <12 mo	—	Race, family history, smoking, BMI, hypertension, diabetes, hyperlipidemia, aspirin, BPH, chronic kidney disease, ACEI, insulin, PSA, vitamin E/multivitamin, finasteride, metformin, a testosterone supplement, selenium, ASCVD, number of PSA tests	Cohort study

ACEI indicates angiotensin-converting enzyme inhibitors; ASCVD, atherosclerotic cardiovascular disease; BMI, body mass index; BPH, benign prostate hyperplasia; CHD, coronary heart disease; CVD, cerebrovascular disease; —, data unavailable; EMIS, Egton Medical Information System; IHD, ischemic heart disease; MH, Manitoba Health; MI, myocardial infarction; NSAID, nonsteroidal antiinflammatory drug; PCA, prostate cancer; PSA, prostate-specific antigen; PV AMC, Portland Veterans Affairs Medical Center; RA, rheumatoid arthritis; SEER, Washington Surveillance, Epidemiology, and End Results cancer registry; WHR, waist-hip ratio.

Studies that compared statin exposure with PCA prognosis (n=7) were from the United States (n=2), China (n=2), Denmark (n=1), Canada (n=1), and the United Kingdom (n=1). There were 6 cohort studies and 1 case-control study covering the period from 1994 to 2016. The inclusion criteria were primarily newly histologically confirmed PCA, and exclusion criteria included unclear diagnostic information, diagnosis of other cancers, etc. The relevant adjustment factors for the included studies were age, race, sex, body mass index, clinical stage, and other treatments were also in the range of adjustment. The remaining details and characteristics are provided in Table 2 and Supplementary Table 2 (Supplemental Digital Content 2, http://links.lww.com/AJCO/A468).

TABLE 2.

Characteristics of Included Studies in the Meta-analysis of the Correlation Between Statins and the Survival Rate of PCA

References	Country	Period	Inclusion criteria	Exclusion criteria	Administration time/dose of statins	Adjustment factor	Research type
Goldberg et al29	Canada	1994.1.1-2016.9.30	Age ≥66; a single negative TRUS-Bx; no prior PCA	—	—	Age, rurality index, index year, ADG comorbidity score, proton pump inhibitors, alpha-blockers, 5-alpha-reductase inhibitors, chloroquine, dipyridamole, glaucoma eye drops	Cohort study
Larsen et al30	Denmark	1998-2011	Denmark resident; aged 35-85; histologically confirmed PCA; no prior cancer	Nonmelanoma skin cancer	—	Age, calendar period, clinical stage, Gleason score, radical prostatectomy, aspirin, nonaspirin NSAID, antihypertensives, other cardiovascular drugs, diabetes, COPD, IHD, congestive heart disease, marital status, moderate-to-severe kidney or liver diseases, income, education	Cohort study
Marcella et al31	USA	1997-2000	Only White and Black men who died from PCA between 1997 and 2000; aged 55-79; be married	Incomplete information	—	Education, BMI, waist size, comorbidity number, race, age	Case control study
Sun et al32	China	1998.1.1-2010.12.31	Histologically confirmed PCA	Age ≤20	≥6 mo	Age, sex, hormone therapy, treatment and comorbidities of diabetes, hypertension, stroke, CAD, COPD	Cohort study
Tan et al33	USA	2008.1-2011.12	Histologically confirmed PCA; enrolled in Medicare Part A and B; enrolled in Medicare Part D beginning at least 3 mo before cancer diagnosis	Enrolled in health care maintenance organizations; diagnosed at autopsy; missing a diagnosis date; death date equal to or less than the diagnosis date	—	Cancer stage, ADT, radiation therapy, surgery, salvage radiation, secondary cancer therapy, propensity scores, imbalanced variables after propensity scores adjustment, metformin use	Cohort study
Wu et al34	China	2008-2014	Received only ADT in the first year after their cancer diagnosis	Missing age or a diagnosis date; missing information on the cancer stage; another cancer; died within 1 y after the cancer diagnosis; T1 or T2 disease; surgery, radiation, or local treatment	—	Cancer stage, cancer grade, year of the cancer diagnosis, metformin, NSAID, aspirin	Cohort study
Yu et al35	UK	1998.4.1-2012.10.1	Histologically confirmed PCA; at least 1 y of follow-up	Metastatic disease; <1 y of up-to-standard medical history in the CPRD before the PCA diagnosis	—	Age, year of diagnosis, ethnicity, alcohol, smoking, obesity, chronic kidney disease, MI, ischemic stroke, transient ischemic attack, peripheral artery disease, prior cancer, PSA, Gleason grade, metformin, sulfonylureas, thiazolidinediones, insulins, other oral antihypoglycemic agents, ACEI, ARBs, calcium channel blockers, β-blockers, diuretics, other antihypertensive drugs, aspirin, NSAID, 5α-reductase inhibitors, prediagnostic statin, PSA, prostatectomy, radiation therapy, chemotherapy, ADT	Cohort study

ACEI indicates angiotensin-converting enzyme inhibitors; ADG, Johns Hopkins’ Aggregated Diagnosis Groups; ADT, androgen deprivation therapy; ARBs, angiotensin receptor blockers; BMI, body mass index; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; CPRD,Clinical Practice Research Datalink; IHD, ischemic heart disease; MI, myocardial infarction; NSAID, nonsteroidal antiinflammatory drug; PCA, prostate cancer; TRUS-Bx, transrectal ultrasound-guided prostate biopsy; —, data unavailable.

All observational studies described above were assessed for quality using NOS. All results met the quality standards of this meta-analysis (Supplementary Table 3, Supplemental Digital Content 3, http://links.lww.com/AJCO/A469 and Supplementary Table 4, Supplemental Digital Content 4, http://links.lww.com/AJCO/A470).

Association Between Exposure to Statins With Different Properties (Hydrophilic or Lipophilic) and PCA Incidence

In this meta-analysis, statins were divided into hydrophilic (rosuvastatin, pravastatin) and lipophilic (atorvastatin, simvastatin, fluvastatin, lovastatin) groups according to their characteristics. There was no difference in the incidence of PCA between the exposed and nonexposed populations with hydrophilic statins (OR=0.94, 95% CI=0.88-1.01, P=0.075) (Fig. 2A). On the other hand, the incidence of PCA was significantly lower in populations exposed to lipophilic statins than in the nonexposed group (OR=0.94, 95% CI=0.90-0.98, P=0.001, Fig. 2B). Further, stratification according to the type of statins showed no difference in the incidence of PCA between the population exposed to rosuvastatin, pravastatin, atorvastatin, fluvastatin, and lovastatin and the nonexposed population. Notably, the incidence of PCA was significantly lower in the simvastatin-exposed group than in the nonexposed group (OR=0.83, 95% CI=0.71-0.97, P=0.016).

FIGURE 2.

Forest plot of the total risk of prostate cancer in the population exposed to hydrophilic statins and lipophilic statins (A, hydrophilic statins, P=0.075; B, lipophilic statins, P=0.001).

Subgroup analysis was further performed according to regions; the incidence of PCA in Eastern (OR=0.91, 95% CI=0.72-1.15, P=0.439) and Western (OR=0.96, 95% CI=0.90-1.03, P=0.241) populations exposed to hydrophilic statins did not differ from the nonexposed group. For lipophilic statins, there was no difference in the incidence of PCA between the exposed and nonexposed groups in Eastern countries (OR=0.87, 95% CI=0.71-1.08, P=0.201). However, in Western countries, the incidence of PCA was significantly lower in people exposed to lipophilic statins than in nonexposed groups (OR=0.94, 95% CI=0.91-0.98, P=0.006). Additional details are provided in Table 3.

TABLE 3.

Subgroup Analysis of Statins and the Incidence of PCA

	No. studies	OR	95% CI	P	Heterogeneity (I 2) (%)
Hydrophilic statins
Rosuvastatin	5	0.84	0.65-1.09	0.191	72.0
Pravastatin	8	0.98	0.90-1.07	0.619	48.7
Lipophilic statins
Atorvastatin	8	0.98	0.90-1.07	0.678	77.9
Simvastatin	10	0.83	0.71-0.97	0.016	95.5
Fluvastatin	6	1.00	0.89-1.13	0.983	25.8
Lovastatin	6	0.94	0.81-1.10	0.445	38.3
Hydrophilic in Eastern countries	2	0.91	0.72-1.15	0.439	60.8
Hydrophilic in Western countries	10	0.96	0.90-1.03	0.241	47.6
Lipophilic in Eastern countries	2	0.87	0.71-1.08	0.201	61.7
Lipophilic in Western countries	12	0.94	0.91-0.98	0.006	91.1

CI indicates confidence interval; OR, odds ratio; PCA, prostate cancer.

Association of Exposure to Statins of Different Properties (Hydrophilic or Lipophilic) With PCA Prognosis

Compared with the nonexposed group, the prognosis of PCA was improved in the population exposed to hydrophilic (HR=0.57, 95% CI=0.49-0.66, P<0.001) (Fig. 3A) and lipophilic (HR=0.65, 95% CI=0.58-0.73, P<0.001) (Fig. 3B) statins.

FIGURE 3.

Forest plot of prostate cancer survival in the population exposed to hydrophilic statins and lipophilic statins (A, hydrophilic statins, P<0.001; B, lipophilic statins, P<0.001).

Subgroup analysis of drugs showed that patients exposed to these statins (rosuvastatin, pravastatin, atorvastatin, simvastatin, fluvastatin, and lovastatin) had a better prognosis than the non-exposed population. Among them, improved outcomes were most pronounced in those exposed to rosuvastatin (HR=0.51, 95% CI=0.38-0.69, P<0.001).

During subgroup analysis, according to regions, in both Eastern (HR=0.50, 95% CI=0.42-0.60, P<0.001) and Western (HR=0.71, 95% CI=0.62-0.82, P<0.001) countries, the population exposed to hydrophilic statins had a better prognosis than the non-exposed group. Similarly, Eastern (HR=0.61, 95% CI=0.53-0.71, P<0.001) and Western (HR=0.79, 95% CI=0.72-0.88, P<0.001) populations exposed to lipophilic statins had better prognosis compared with nonexposed populations. The degree of improvement in prognosis was more significant in Eastern than in Western populations. The remaining relevant details are provided in Table 4.

TABLE 4.

Subgroup Analysis of Statins and the Survival rate of PCA

	No. studies	HR	95% CI	P	Heterogeneity (I 2) (%)
Hydrophilic statins
Rosuvastatin	3	0.51	0.38-0.69	<0.001	83.4
Pravastatin	3	0.55	0.46-0.67	<0.001	57.5
Lipophilic statins
Atorvastatin	3	0.59	0.43-0.81	0.001	94.2
Simvastatin	3	0.69	0.52-0.92	0.010	92.1
Fluvastatin	2	0.64	0.50-0.82	<0.001	76.5
Lovastatin	3	0.73	0.55-0.97	0.031	87.5
Hydrophilic in Eastern countries	2	0.50	0.42-0.60	<0.001	75.3
Hydrophilic in Western countries	5	0.71	0.62-0.82	<0.001	14.0
Lipophilic in Eastern countries	2	0.61	0.53-0.71	<0.001	87.9
Lipophilic in Western countries	5	0.79	0.72-0.88	<0.001	58.8

HR indicates hazard ratio; PCA, prostate cancer.

Publication Bias and Sensitivity Analysis

Begg test was used to assess publication bias. There was no significant publication bias (P=0.544) (Supplemental Fig. 1, Supplemental Digital Content 5, http://links.lww.com/AJCO/A471) when comparing hydrophilic statin exposure to the incidence of PCA. The sensitivity analysis results showed that the statistical results were stable after removing each study one by one (Supplemental Fig. 2, Supplemental Digital Content 6, http://links.lww.com/AJCO/A472). At the same time, no significant publication bias was found between lipophilic statin exposure and PCA incidence (P=0.141) (Supplemental Fig. 3, Supplemental Digital Content 7, http://links.lww.com/AJCO/A473), and sensitivity analysis results were stable (Supplemental Fig. 4, Supplemental Digital Content 8, http://links.lww.com/AJCO/A474). Similar findings were reported for publication bias (P=0.350) (Supplemental Fig. 5, Supplemental Digital Content 9, http://links.lww.com/AJCO/A475) and sensitivity analysis (Supplemental Fig. 6, Supplemental Digital Content 10, http://links.lww.com/AJCO/A476) between hydrophilic statin exposure and PCA survival. Moreover, publication bias was not significant (P=0.256) (Supplemental Fig. 7, Supplemental Digital Content 11, http://links.lww.com/AJCO/A477) and sensitivity analysis (Supplemental Figure 8, Supplemental Digital Content 12, http://links.lww.com/AJCO/A478) yielded stable results between lipophilic statin exposure and PCA survival.

DISCUSSION

Obesity is well-recognized as a risk factor for PCA, and it has been reported that the risk of PCA in obese men was 1.5 times higher than in nonobese men.³⁶ There is a growing consensus that men with higher levels of dyslipidemia are likely to have an increased risk of PCA.^37–39 Based on this, the present study investigated whether statins could reduce the incidence of PCA. The results of this study showed that hydrophilic statins (rosuvastatin, pravastatin) were not associated with the incidence of PCA, but lipophilic statins, especially simvastatin, could effectively reduce the incidence of PCA. Besides, unlike in Eastern populations, lipophilic statins reduced the risk of PCA in Western populations.

Many mechanisms have been reported to account for the ability of statins to reduce the incidence of PCA. First of all, statins are well-established as specific and potent inhibitors of 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGCR), which could inhibit the metabolism of farnesyl pyrophosphate and thus affect many aspects of tumorigenesis.⁴⁰ Besides, an increasing body of evidence suggests that cholesterol homeostasis disorders in the synthesis, storage, uptake, efflux, and metabolism stages were often related to the occurrence of tumors.^41–43 LNCaP human PCA cells contained cholesterol-rich lipid rafts that constitutive signaling. Statins exposure leads to the depletion of lipid raft cholesterol, followed by downregulation in the PI3K-AKT pathway and apoptosis, thereby inhibiting tumor growth.^44,45 Another mechanism worth elucidating is that statins induce cell apoptosis and reduce cell proliferation, probably mediated by a downregulation of FOXO1/AKT phosphorylation in PCA cells, which probably yields a potential benefit in PCA prevention.⁴⁶

Lipophilic statins have a greater intracellular pathway and cross biofilms more easily than hydrophilic statins.⁴⁷ Lipophilic statins inhibited the proliferation of kinds of PCA cell lines by inducing G1 cell cycle arrest. For example, this inhibition was achieved with lovastatin at a concentration of 0.5 mmol/L, while hydrophilic pravastatin must be present at a 200-fold concentration to achieve cell cycle arrest because it did not enter cells efficiently enough.⁴⁸ In addition, the bioavailability of HMG-CoA reductase inhibitors in human tissues is reportedly restricted by first-pass metabolism in the liver.⁴⁹ Current evidence suggests that the pharmacokinetic profiles of lipophilic statins are the opposite of hydrophilic statins and are distributed in many tissues.⁵⁰ Further stratification according to statin drug revealed that only simvastatin was effective, probably because simvastatin, a dual targeting agent of both steroid and cholesterol biosynthesis, yielded the strongest growth inhibitory effect, as previously reported by Neuwirt et al.⁵¹ In addition, the number of studies on simvastatin in this meta-analysis was 10, which was the largest, and it was more conducive to produce statistically significant differences. In addition, one study substantiated that atorvastatin activated the transcription of LC3 through JNK and Erk signaling pathways, which led to the increase of LC3-II as an important component of autophagosomes, thereby leading to autophagy of PC3 cells and inhibiting the occurrence of PCA.⁵² However, there was no difference in atorvastatin-related outcomes in our study.

PCA is a disease exhibiting significant geographic variations in terms of incidence. In this respect, the incidence of PCA in Western countries is significantly higher than in Eastern countries, accounting for the more significant protective effect. In addition, there were more studies on lipophilic statins in the West (n=12) than in the East (n=2). Accordingly, Western countries were more likely to have statistical results. Moreover, black race and family history are well-established risk factors for PCA. Meanwhile, genetic factors and their interactions with the internal microenvironment and environmental conditions influence the development of PCA.² The associated androgen-receptor (AR) gene duplication has been reported to raise AR activity, leading to an increased risk of PCA, mainly in African Americans.⁵³ Therefore, it is reasonable to believe that genetic or racial factors play an indispensable role in the effect of statins on the incidence of PCA. Accordingly, the protective effect in Western countries was more significant than in Eastern countries.

Interestingly, both hydrophilic and lipophilic statins improved PCA prognosis in the East and the West. In contrast, hydrophilic and lipophilic statins yielded significantly better effects in Eastern populations than in Western populations in improving prognosis. Moreover, it has been reported that statins impact the farnesylation of ras and related proteins, thus inhibiting the growth and progress of ras-dependent tumor cells.^54,55 It is widely acknowledged that statins also possess anti-inflammatory properties, and inflammation affects PCA growth. Thus, statins probably suppress the progression of PCA by inhibiting inflammation. For instance, in a study of men, preoperative statin administration was significantly associated with a decreased risk of inflammation in peripheral malignant glands in radical prostatectomy specimens.⁵⁶ In addition, statins could effectively reduce PCA cell viability by inducing apoptosis or arrest in the G1 phase of PCA cells.⁵⁷ Moreover, Rosuvastatin yielded the most effective results, probably because the epithelial-mesenchymal transition of cancer cells is an important process in the invasion and metastasis of cancer. Deezagi and Safari⁵⁸ reported that rosuvastatin could inhibit epithelial-mesenchymal transition markers effectively. Furthermore, rosuvastatin inhibited angiogenesis in vitro by targeting endothelial cell function and significantly inhibited PCA growth in the mouse xenograft tumor model.⁵⁹ Moreover, a study showed that rosuvastatin was the most effective compound (among pravastatin, atorvastatin, simvastatin, lovastatin, cerivastatin, rosuvastatin, and fluvastatin) in inhibiting tumors in animal models.⁶⁰

The prognosis of PCA varied greatly between the East and the West, probably because prostate-specific antigen screening has not been widely adopted in Eastern countries, and the prostate-specific antigen level in Eastern subjects was lower than that in Western ones, which was likely to affect the diagnosis of PCA, leading to the more advanced pathologic stage of PCA and poor prognosis.^61–63

To our knowledge, this is the first comprehensive and systematic meta-analysis to explore the relationship between the properties of statins (hydrophilic or lipophilic) and the incidence and prognosis of PCA. Besides, this study demonstrates that the effect of statins on the incidence and prognosis of PCA is influenced by a combination of statins’ properties and regional factors. However, there are some limitations. First of all, there is a lack of high-quality RCT studies. Furthermore, differences in race, age, duration, dose, and confounding adjustment factors of statin therapy can lead to significant heterogeneity. However, further subgroup analysis could not be performed due to the lack of relevant data. In addition, the adjustment factors of the included studies were not uniform, which affected the reliability of the results to a certain extent. The use of statins for PCA prevention and their potential to improve prognosis warrants further exploration.

CONCLUSION

Available evidence suggests that hydrophilic statin exposure is not associated with decreased incidence of PCA. In contrast, exposure to lipophilic statins, primarily simvastatin, significantly reduced PCA risk, especially in Western populations. Both hydrophilic and lipophilic statins effectively improved PCA prognosis, and the improvement was more significant in Eastern than Western countries.