Combination regimen with statins and NSAIDs: A promising strategy for cancer chemoprevention

Abstract

Statins and nonsteroidal antiinflammatory drugs (NSAIDs) are commonly prescribed for lowering cholesterol and antiinflammation, respectively. Recently, their potential roles as cancer chemopreventive agents have been subject to intensive studies. Human trials have not provided conclusive results on the protective effects of statins against different cancers, while more convincing results have been observed for cancer preventive effects of NSAIDs, especially on colorectal cancer. A promising strategy to enhance the cancer preventive efficacy of statins and NSAIDs is to use them in combination, which may produce synergy and lower the dose required for each agent. This strategy is of particular interest for potential use of low doses of statins and NSAIDs on a long-term basis for cancer chemoprevention; increased risks for gastrointestinal and cardiovascular side effects associated with the use of NSAIDs have been observed in colorectal cancer chemopreventive trials. This article reviews the evidence for the cancer preventive actions of statins and NSAIDs, as well as their possible synergistic action and the mechanisms involved. © 2008 Wiley-Liss, Inc.

While the mortality from major chronic diseases, such as cardiovascular and cerebrovascular diseases, has decreased substantially during the past half century in the United States, the mortality from cancer has just begun to show a slight decrease. Cancer has become the number one killer of both men and women under age 85 years in the United States.1 A promising approach to controlling cancer is to prevent it before malignant events have occurred.2 Cancer chemoprevention refers to the use of natural or synthetic substances to inhibit, retard or reverse the carcinogenesis.3 A wealth of evidence from preclinical studies have convincingly demonstrated the cancer preventive efficacy of various agents in different animal models. However, the data from observational, case–control, cohort studies, and randomized trials in humans have overall demonstrated mixed results. Statins and nonsteroidal antiinflammatory drugs (NSAIDs) have been reported to be potential cancer chemopreventive agents. This review focuses on an attractive approach of chemoprevention using the combination of statins and NSAIDs.

Statin and cancer chemoprevention

Statins are small-molecule inhibitors of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, which are widely used as cholesterol-lowering drugs. Besides their use in the treatment of lipid disorders, statins have showed anticarcinogenic activities in many in vitro and in vivo preclinical studies, and attracted much attention in exploring their roles in cancer chemoprevention. Several observational human studies reported potential protective effects of statin use against overall risk of cancer,4–6 whereas other studies did not7, 8 (Table I). In a case–control study in Quebec, Canada, statin use was found to be associated with a 28% reduction in the risk of developing cancer (RR = 0.72, 95% CI = 0.57–0.92) in comparison with the use of bile acid-binding resins.4 In another case–control study where 3,129 cancer patients were identified and matched to 16,976 controls, statin use (80% as simvastatin) was associated with a 20% reduction in the overall risk of cancer (OR = 0.80, 95% CI = 0.66–0.96).5 The results also suggested the association between the reduced cancer risk with statin use for a longer period and in high dosages. Moreover, the protective effect of statin use was only present in the current users and the past users who had their therapies stopped for less than 6 months before the index date. In a Danish study, comparisons were made between overall and site-specific cancer incidence among 12,251 statin users and cancer incidence among nonusers and users of other lipid-lowering drugs (n = 1,257), and a slight reduction in the overall cancer risk was found in the statin users.6 A small population-based cohort study did not show the link between reduced overall cancer risk and lipid-lowering drug use.7 With limited statistical precision, the results from this study were compromised by the fact that the statin use was not separated from other lipid-lowering drugs. Using information from the General Practice Research Database, a study involving 3,244 patients and 14,844 controls found no substantial difference in the risk of overall cancer in the comparison of current statin users with subjects without hyperlipidemia and no recorded use of lipid-lowering drugs. However, untreated hyperlipidemia was associated with slightly increased risks of colon cancer (RR = 1.8, 95% CI = 1.2–2.8), prostate cancer (RR = 1.5, 95% CI = 1.1–2.0) and bladder cancer (RR = 1.9, 95% CI = 1.2–3.1).8

Table I. Studies on Statin Use and Cancer

Source	Location	Site of cancer	Medication	No. of participants1	Outcomes
L, Lovastatin; P, pravastatin; S, simvastatin; A, atorvastatin; C, cerivastatin; F, fluvastatin. 1 Expressed as either “cancer cases/control” for case-controlled studies, or “study population” for cohort studies and meta-analysis. – 2 Meta-analysis. 3 These statins were used in randomized controlled trials included in the meta-analysis.
Blais et al.4	Canada	Overall	L, P and S	542/5,420	Protective
Graaf et al.5	The Netherlands	Overall	S (80%), A, C, F and P	3,129/16,976	Protective
Friis et al.6	Denmark	Overall	A, C, F, L, P and S	334,754	Protective
Olsen et al.7	Denmark	Overall	Lipid-lowering drugs (52% statins)	1,882	Not protective
Kaye et al.8	USA	Overall	Statins	3,244/14,844	Not protective
Poynter et al.9	Israel	Colorectal	P and S	1,953/2,015	Protective
Jacobs et al.10	USA	Colorectal	Cholesterol-lowering drugs (62–86% as statins)	132,136	Not protective
Dale et al.11	Multiple2	Overall	A, C, F, L, P and S3	86,936	Not protective
Bonovas et al.12	Multiple2	Colorectal	L, P and S3	1,500,000	Not protective
Khurana et al.,13	USA	Lung	Statins	7,280/476,453	Protective
Khurana et al.14	USA	Kidney	Statins	1,446/482,287	Protective
Platz et al.15	USA	Prostate	Cholesterol-lowering drugs (mainly statins)	34,989	Protective
Murtola et al.16	Finland	Prostate	A, F, L, P and S	24,723/24,723	Protective
Farwell et al.17	USA	Overall	Mainly S and L	62,842	Protective
Setoguchi et al.18	USA	Lung, breastand colorectal	A, C, F, L, P and S	31,723	Not protective
Vinogradova et al.19	UK	Colorectal	A, C, F, P and S	5,686/24,982	Not protective

Recently, the Molecular Epidemiology of Colorectal Cancer (MECC) study has attracted much attention. This study involved 1,953 colorectal cancer patients in northern Israel, and it was found that statin use for 5 years or longer was associated with significantly reduced risk of colorectal cancer (OR = 0.50, 95% CI = 0.40–0.63).9 This association remained significant after adjustment for various confounding factors (OR = 0.57, 95% CI = 0.44–0.73). Based on a stratified analysis, the statin use provided significant protective effect among patients with a family history of colorectal cancer (OR = 0.41, 95% CI = 0.19–0.93). A protective effect of statins was also observed in the subgroup with inflammatory bowel disease (OR = 0.06, 95% CI = 0.006–0.55). In contrast to the results from the MECC study, a large cohort study involving 132,136 subjects demonstrated that there was no association between colorectal cancer risk and use of cholesterol-lowering drugs, mostly statins.10 The results from 2 meta-analyses also did not support the link between a reduced cancer risk and statin use. In a large meta-analysis of 27 studies (n = 86,936 participants), the relationship between statin use and the incidence and mortality of different types of cancers was studied. It was found that statins have a neutral effect on cancer and cancer death risk in randomized controlled trials.11 According to another meta-analysis of 18 published studies, there was no association between statin use and risk of colorectal cancer found in either randomized controlled studies (n = 6) or prospective cohort studies (n = 3).12 However, the meta-analysis of retrospective case–control studies (n = 9) indicated a statistically significant association between statin use and a modest reduction in the risk of colorectal cancer.

Several newly published human studies demonstrated the possible protective effects of statins against the risk of overall cancer and several specific cancers, in particular lung, prostate, renal and colorectal cancers (Table I). In a large case–control study involving 483,733 veterans from south central US, statin use for 6 months or longer was associated with a significant risk reduction of lung cancer by 55% (OR = 0.45, 95% CI = 0.42–0.48).13 The results indicated an increasing protective effect against lung cancer from the longer duration of statin use. A risk reduction of lung cancer by 77% was found among those who received statin therapy for 4 and more years (OR = 0.23, 95% CI = 0.20–0.26). It is noteworthy that statin use for 6 months or longer was even associated with a risk reduction of 53% against lung cancer in smokers (OR = 0.47, 95% CI = 0.43–0.51). A subsequent study in the same population showed a significant association between statin use and risk reduction of renal cell carcinoma (OR = 0.52, 95% CI = 0.45–0.60).14 Two trials reported consistent results that statin use was not associated with risk reduction in overall prostate cancer, but was associated with a reduced risk of the advanced prostate cancer.15, 16 In a large retrospective cohort study, the relationship between statin use and cancer incidence was investigated among the veterans using statins (n = 37,248) in reference to the veterans using antihypertensive medications but no statins (n = 25,594).17 Statin (mainly simvastatin and lovastatin) use was associated with a 26% risk reduction in the incidence of all cancers after adjusting for age and other potential confounders (HR 0.74, 95% CI = 0.70–0.87). Statin use was also associated with the reduced risk in the incidence of the top 3 cancers in the cohort, namely, prostate cancer (HR = 0.85, 95% CI = 0.78–0.93), lung cancer (HR = 0.75, 95% CI = 0.66–0.86) and colorectal cancer (HR = 0.62, 95% CI = 0.54–0.73). Compared to nonstatin users, 3 categories of equivalent simvastatin doses (≤10 mg, 11–39 mg, and ≥40 mg) were significantly associated with a reduced risk of all cancers after adjusting for multiple confounders (p_trend < 0.001). For individual cancers, simvastatin use of 40 mg or higher was associated with decreased incidences of lung cancer (HR = 0.73, 95% CI = 0.54–0.97), colorectal cancer (HR = 0.59, 95% CI = 0.41–0.85) and melanoma (HR = 0.64, 95% CI = 0.44–0.94). Moreover, the dose–response trend of the protective effects was statistically significant for lung cancer (p_trend < 0.002), colorectal cancer (p_trend < 0.001) and melanoma (p_trend < 0.004). The results from these studies support the hypothesis that statins may reduce the risk of cancer. However, overall results from observational studies still remain inconclusive. For example, in contrast to the aforementioned studies where statin use was associated with lower risk of cancer, Setoguchi et al.18 and Vinogradova et al.19 did not find protective effects of statin use against lung, breast or colorectal cancer in 2 studies with large population.

NSAIDs and cancer chemoprevention

Many observational epidemiologic studies have reported protective effects of NSAIDs against different cancers in the general population. The most consistent finding is that the regular use of aspirin and other NSAIDs was associated with lower incidence of adenomatous polyps and colorectal cancer, and lower mortality from colorectal cancer.20 The use of NSAIDs was also linked to the protective effects against cancers in other sites such as breast, esophagus, stomach, lung, prostate and ovary.21–29 However, the results for these cancer types are less consistent compared to those for colorectal cancer.

Randomized clinical studies have demonstrated chemopreventive effects of NSAIDs against colorectal cancer (Table II). Three short-term trials showed that sulindac treatments (4–9 months) decreased the number and size of adenomatous polyps, and induced regression of existing polyps in familial adenomatous polyposis patients.30–32 Similar protective effects were also observed in a long-term study, with the sulindac treatment for a mean period of 63.4 months.33 In contrast, sulindac treatment for 48 months did not prevent the development of adenomas in another long-term study, which involved 41 subjects who were genotypically affected by FAP but were phenotypically unaffected initially.34

Table II. Randomized Controlled Trials on the NSAID Use and Cancer

Source	Agent	No. of participants	Duration (months)	Dose (mg/day)	Outcomes
Giardiello et al.30	Sulindac	22	9	2 × 150	Protective
Labayle et al.31	Sulindac	10	2 × 4	300	Protective
Cruz-Correa et al.33	Sulindac	12	14–89	158	Protective
Giardiello et al.34	Sulindac	41	48	2 × 75 or 2 × 150	Not protective
Sandlar et al.35	Aspirin	635	Median, 30.9	325	Protective
Baron et al.36	Aspirin	1,121	34–40	81 or 325	Protective
Benamouzig et al.37	Aspirin	272	48	160 or 300	Protective
Steinbach et al.39	Celecoxib	77	6	2 × 100 or 2× 400	Protective
Bertagnolli et al.40	Celecoxib	2035	36	2 × 200 or 2 × 400	Protective
Arber et al.41	Celecoxib	1561	36	400	Protective

Aspirin has been studied in 3 prospective trials for prevention of sporadic colorectal adenomas. In a randomized double-blinded trial with 635 patients who had a previous history of colorectal cancer, the daily aspirin treatment of 325 mg significantly decreased the average number of adenomas.35 Compared with the placebo group, aspirin-treated group also had a lower percentage of patients who had at least 1 adenoma. Moreover, aspirin treatment reduced the relative risk of a new adenoma (RR = 0.65, 95% CI = 0.46–0.91). In another trial involving more subjects (n = 1,121), aspirin was administered at daily doses of 81 or 325 mg to patients with a recent history of colorectal adenoma.36 A moderate reduction in the recurrence of adenoma was found in the group given 81 mg of aspirin (RR = 0.81, 95% CI = 0.69–0.96), but no significant effect was observed in the group treated with a higher dose (325 mg) of aspirin. Similarly, a 40% reduction in the risk of advanced lesion was found in the low-dose (81 mg) aspirin-treated group, while no statistical significant effect was seen in the high-dose (325 mg) aspirin-treated group. In the APACC trial, 272 patients with a history of colorectal adenoma were randomly assigned to treatments of soluble aspirin (160 or 300 mg/day) or placebo.37 After 1 year of treatment, the percentages of patients who had at least 1 adenoma of more than 5 mm in diameter were 10% in aspirin-treated group and 23% in placebo group, suggesting a 56% risk reduction (RR = 0.44, 95% CI = 0.24–0.82). Aspirin treatment also decreased the mean burden of recurrent adenomas polyps (1.55 ± 0.53 mm in aspirin group vs. 4.03 ± 1.46 mm in placebo group, p = 0.001). In addition, a trend of increased protective effects from higher dose of aspirin was observed, but the results were not statistically significant.

Nonselective NSAIDs, such as sulindac and aspirin, inhibit both cyclooxygenase (COX)-1 and COX-2 activity. The long-term use of these nonselective NSAIDs has been linked with serious gastrointestinal side effects, including ulceration, bleeding and perforation, due to inhibition of the gastroprotective function of COX-1.38 To avoid the side effects associated with COX-1 inhibition, a class of selective COX-2 inhibiting NSAIDs, such as celecoxib and rofecoxib, have been developed. The first trial of a selective NSAID (celecoxib) was conducted among 77 patients with familial adenomatous polyposis.39 The treatment with 400 mg celecoxib twice a day for 6 months caused a 28% reduction in the mean number of colorectal polyps (p = 0.003), and a 30.7% reduction in the polyp burden defined as the sum of the polyp diameters (p = 0.001). Two large randomized trials of celecoxib have been conducted for prevention of sporadic colorectal adenomas. In the Adenoma Prevention with Celecoxib (APC) trial,40 patients who had adenomas removed were randomly assigned to receive placebo (n = 679), 200 mg celecoxib (n = 685) or 400 mg (n = 671) of celecoxib twice a day. The cumulative incidence of adenoma detected through year 3 was 60.7, 43.2, and 37.5% in the placebo, 200 mg celecoxib and 400 mg celecoxib groups, respectively, which corresponded to a risk ratio of 0.67 (95% CI = 0.59–0.77) in the 200-mg group and 0.55 (95% CI = 0.48–0.64) in the 400-mg group. Celecoxib treatments decreased the adenoma burden in both 200- (1.0 ± 0.1 cm) and 400-mg (0.9 ± 0.1 cm) groups in comparison to the placebo group (1.3 ± 0.1 cm). Moreover, celecoxib treatments reduced the cumulative incidence of advanced adenoma through year 3 with a risk ratio of 0.43 (95% CI = 0.31–0.61) in the 200-mg group and 0.34 (95% CI = 0.24–0.50) in the 400-mg group. In the Prevention of Colorectal Sporadic Adenomatous Polyps trial,41 subjects who had adenomas removed were randomly assigned to placebo (n = 628) and 400 mg celecoxib twice daily (n = 933) groups. The 3-year cumulative incidence of adenomas detected was 49.6% in the placebo group and 33.6% in the 400 mg of celecoxib group with a relative risk reduction of 36% (RR = 0.64, 95% CI = 0.56–0.75). The treatment with 400 mg celecoxib twice daily reduced the cumulative incidence of advanced adenoma through year 3, with a risk ratio of 0.49 (95% CI = 0.33–0.73).

However, long-term use of selective NSAIDs was also associated with an increase in cardiovascular side effects. In 2004, Merck withdrew rofecoxib from the worldwide market because of increased cardiovascular toxicity observed in a trial designed to test the efficacy of rofecoxib to prevent recurrence of colonic polyps.42 Consequently, recruitment for another trial with rofecoxib was stopped before the final efficacy results were obtained. A subsequent analysis on the rates of cardiovascular thrombotic events and death suggested that rofecoxib therapy was associated with increased frequency of adverse cardiovascular events.43 The APC trial with celecoxib was also discontinued early, because celecoxib use was found to be associated with a dose-related increase in the composite end point of death from cardiovascular causes, myocardial infarction, stroke or heart failure.44

Cancer chemoprevention by the combinations of statins and NSAIDs

Accumulating evidence has convincingly demonstrated the cancer preventive effects of NSAIDs, especially in colorectal cancer. However, the relative high dose required for the observed chemopreventive effect in human trials may discourage the singular use of NSAIDs on a long-term basis for cancer prevention because of possibly increased risk for serious gastrointestinal and cardiovascular side effects.38, 44, 45 Moreover, NSAID use at high doses is not favorable for their application as cancer chemopreventive agents, based on their unsatisfactory cost-effectiveness, especially in average-risk population.46, 47 An alternative approach is to combine NSAIDs with other chemopreventive agents. There is a growing body of evidence suggesting that the combination of cancer chemopreventive agents with different modes of action may produce synergistic type of interactions, which can result in considerably stronger protective effects against carcinogenesis than each chemopreventive agent could individually. The enhanced efficacy by the combination can also lower the dose required for each agent in the combination, and in turn reduce unwanted side effects possibly caused by high-dose single-agent administration. In numerous in vitro studies, NSAIDs have shown synergistic effects in combination with other therapeutics, such as statins, PPARγ ligands, inhibitors of the EGFR family and TRAIL receptor ligands.48 Some of these observations have been confirmed in animal models.48

The combination of NSAIDs and statins is especially of interest for cancer prevention. As mentioned earlier, statins are potential cancer chemopreventive agents. In addition, the noted efficacy in cardiovascular disease prevention and the relative safety of the statins have resulted in statins' widespread use and its recent conversion from prescription to over-the-counter drug in the United Kingdom.49 For instance, atorvastatin was the most commonly prescribed drug in the United States in 2006 (www.pharmacytimes.com). Several human studies have suggested potential protective effects against cancer by the combination regimen with statins and NSAIDs. In a large trial to study the outcomes in patients with coronary artery disease, pravastatin use was associated with 43% reduction in the number of newly diagnosed cases of colon cancer.50 It was noteworthy that 83% of patients in both pravastatin and placebo groups received a daily dose of aspirin, suggesting a possible enhanced protective effect by the interaction between pravastatin and aspirin.51 In a population-based case–control study, the effects of statins and aspirin were evaluated on the risk of colorectal cancer.52 Overall, 537 patients with histologically confirmed incident colorectal cancer and 612 control individuals were included in the study. Regular use of low-dose aspirin (94% of doses used were 100 mg) was associated with modest risk reduction on colorectal cancer (adjusted OR = 0.77, 95% CI = 0.55–1.07), while a stronger risk reduction was found associated with regular use of statins, mostly atorvastatin and simvastatin (OR = 0.65, 95% CI = 0.43–0.99). Most interestingly, combined use of both statins and low-dose aspirin for 5 or more years was associated with a much stronger risk reduction of 62% on colorectal cancer (OR = 0.38, 95% CI = 0.15–0.97). In the California Men's Health Study, the association of statin use and the risk of prostate cancer was examined among 69,047 eligible participants.53 The long-term use of statin for 5 years or longer was associated with a 28% risk reduction (RR = 0.72, 95% CI = 0.53–0.99). Moreover, stratified analysis indicated that regular NSAID users who also took statins for 5 years or longer had an even lower risk of prostate cancer (RR = 0.64, 95% CI = 0.44–0.93) in comparison with the whole cohort, while no protective effect of long-term statin use was observed among never or episodic NSAID users (RR = 1.05, 95% CI = 0.55–1.98). Although these studies had limitations, the results provided evidence that combinational use of statins and NSAIDs may produce stronger protective effects against colorectal and prostate cancer development in comparison with the singular use of either of the agents.

Several in vitro studies have demonstrated synergistic effects between statins and NSAIDs in inhibition of cell growth and induction of apoptosis in cancer cells.51, 54–56 Recently, the in vivo efficacy of this type of combination has been investigated in animal models for prostate and intestinal cancers. In a xenograft tumor model, daily introperitoneal injections of high dose of atorvastatin (10 μg/g body weight) or celecoxib (10 μg/g body weight) showed little or no effect on the formation of androgen-independent prostate tumors that occurred after subcutaneous injection of PC-3 cells in male severe combined immunodeficient mice.57 In contrast, the administration with a low-dose combination of atorvastatin (5 μg/g body weight) plus celecoxib (5 μg/g body weight) significantly delayed the formation of the tumors, and some of mice the remained tumor-free throughout the whole experiment period. The low-dose combination of the 2 agents also had a stronger inhibitory effect on the tumor burden (53.3% reduction in the tumor size) than the higher dose of individual agent (20% reduction for atorvastatin, and 24.4% reduction for celecoxib). In an azoxymethane (AOM)-induced colon carcinogenesis model, combination of lovastatin and sulindac produced stronger inhibition on colonic aberrant crypt foci than did either agent alone.54 Using the same AOM-rat model, Reddy et al. studied the efficacy of atorvastatin, aspirin and celecoxib, given singularly at high dose levels and in combination at lower doses against colon carcinogenesis.58 The combination of low dose of atorvastatin (100 ppm) and celecoxib (300 ppm) decreased the incidence and multiplicity of adenocarcinomas by 71 and 90%, respectively, and this low-dose combination was more effective than the individual high doses of atorvastatin (150 ppm, 34 and 37% reduction for incidence and multiplicity, respectively) or celecoxib (600 ppm, 61 and 76% reduction for incidence and multiplicity, respectively). Similarly, the low-dose combination of atorvastatin and aspirin produced the stronger inhibitory effects on the incidence and multiplicity of adenocarcinoma in comparison to the higher dose of individual agents alone. In another study using the APC^Min/+ mouse model, Swamy et al. assessed the effects of atorvastatin and celecoxib on the development of intestinal adenomatous polyps. It was observed that the combination of atorvastatin (100 ppm) and celecoxib (300 ppm) suppressed colonic adenomatous polyps completely and small intestinal adenomatous polyps by 86%, and these inhibitory effects were stronger than those produced by the treatment with atorvastatin (100 ppm) or celecoxib (300 ppm) alone.59 Taken together, these results convincingly demonstrated that combination regiments with statins and NSAIDs significantly enhanced the cancer preventive efficacy of either type of agents administered alone, which provides a strong rationale to use the statins/NSAIDs combination as a promising approach for cancer chemoprevention.

Mechanisms of actions

The mechanisms by which statins and NSAIDs inhibit cancer cell growth, induce apoptosis and suppress other procarcinogenic processes are not yet fully understood. In Figure 1, atorvastatin and celecoxib were used as examples to summarize the possible mechanisms of statins and NSAIDs as cancer chemopreventive agents. By inhibition of HMG-CoA reductase, the rate-limiting enzyme of mevalonate pathway, statins inhibit synthesis of isoprenoids, such as geranylgeranylpyrophosphate (GGPP) and farnesylpyrophosphate (FPP). These isoprenoids are essential for isoprenylation, membrane localization and subsequent activation of series of signaling proteins, such as Ras, Rho and Rac. Add-back experiments demonstrated that GGPP can prevent statin-induced apoptosis in cancer cells, while FPP showed no or limited preventive effects.54, 60, 61 These results suggested a more important role of GGPP in statin-induced effects than that of FPP. Consistent with these results, studies demonstrated the role of geranylgeranylated Rho proteins in statin-induced effects, while the results on farnesylated Ras have been controversial.62 Another possible mechanism for the cancer preventive effects of statins is due to the inhibition of cholesterol synthesis. Cholesterol is a major component of lipid rafts that regulates many important signaling pathways. Elevated circulation cholesterol increased the extent of protein tyrosine phosphorylation in xenograft tumors and promoted the tumor growth in mice.63 Simvastatin lowered cholesterol content in lipid rafts, inhibited Akt pathway signaling and induced apoptosis in prostate cancer cells. Supplementation of cholesterol reversed these inhibitory and apoptotic effects.63 However, other studies failed to prevent statin-induced growth arrest and apoptosis in different cancer cells by cholesterol supplementation.60, 64, 65

Figure 1. Molecular mechanisms of atorvastatin and celecoxib as cancer chemopreventive agents.

The anticarcinogenetic effects of NSAIDs have been previously attributed to their inhibitory effects on cyclooxygenases. Aberrant functioning of cyclooxygenase-2 (COX-2) has been associated with carcinogenesis by promoting cell survival, angiogenesis and metastasis.66–68 Recent studies, however, have also suggested COX-2-independent mechanisms for cancer chemoprevention by NSAIDs (see Ref. 69 for review). Celecoxib treatments induced cell cycle arrest at G1 phase in multiple tumor cells, in which expression levels of cyclin-dependent kinase (CDK) inhibitors were increased and expression levels of cyclins were decreased.70, 71 The growth-arresting effect of celecoxib might be mediated by its inhibitory action on Akt that can phosphorylate and inactivate p21^Cip1/Waf1 and p27^Kip1.72–74 It has been reported that celecoxib can activate both intrinsic and extrinsic pathways of apoptosis in various cancer cells.75–77 The possible mechanisms for the proapoptotic effect include inhibition of Akt,73 inhibition of Ca²⁺ ATPase activity,78 increase of ceramide level79 and inhibition of transcription activity of NFκB.76 Aspirin was reported to induce cell cycle arrest and apoptosis in multiple colon cancer cells through increasing the levels of mismatch repair proteins,80 activating p53 and p21^Waf1/Cip1 in an ataxia-telangiectasia-mutated kinase-dependent way,81 or activating NFκB pathway.82 APC/β-catenin pathway plays a pivotal role in the development of various cancers, especially colon cancer. Treatment with celecoxib rapidly decreased the DNA-binding activity of β-catenin, which was followed by extensive degradation of β-catenin through a caspase and proteasome-dependent pathway in colon cancer cells.83 Inhibitory effects on the APC/β-catenin signaling were also observed in other colon cancer cells after treatments with aspirin and indomethacin.84 By COX-independent mechanisms, NSAIDs showed inhibitory effects on angiogenesis that is critical to the growth, invasion and metastasis of tumors. Celecoxib suppressed the angiogenesis of pancreatic tumors in an orthotopic mouse model.85 In this model, celecoxib dose-dependently reduced VEGF expression at both the mRNA and protein levels. This was due to the inhibition by celecoxib on both DNA binding activity and transactivating activity of a transcription factor Sp1 that is critical for the expression of VEGF. Other studies demonstrated that celecoxib and sulindac exerted antiangiogenic effects by inhibiting proliferation of human umbilical vein endothelial cells,86 and by inhibiting activities of other proangiogenic factors, such as matrix metalloprotease 2 and 9, and the early growth response factor Egr-1.87–89 Aspirin, at therapeutic concentrations, was found to inhibit tube formation in a 3-dimensional collagen angiogenesis model, via a COX-independent mechanism, which may contribute to its cancer chemoprotective effects.90

The question of exactly how statins and NSAIDs work in a synergistic fashion to produce enhanced anticarcinogenic effects is largely unresolved. We have studied the mode of interactions between atorvastatin and celecoxib in colon cancer HCT116 and HT29 cells and demonstrated a significant synergistic action.56 The atorvastatin/celecoxib combination treatment (for 24 hr) caused cell cycle arrest at the G0/G1 phase, and this effect was much stronger than those caused by atorvastatin or celecoxib alone. Moreover, treatment with the combination (for 48 hr) induced significant apoptosis that could not be seen with atorvastatin or celecoxib treatment individually.56 These results are consistent with the findings from animal studies that combination treatments with atorvastatin and celecoxib significantly decreased proliferative index and increased apoptotic index in tumor tissues.57–59 Other studies in cancer cells also showed enhanced apoptosis by cotreatments with statins and NSAIDs.51, 54, 55, 57 Our studies showed that atorvastatin decreased the level of membrane bound RhoA, presumably through inhibition of isoprenylation, and its combination with celecoxib significantly enhanced the effect.56 This may suppress oncogenic activities of RhoA whose dysregulated functions are associated with cell cycle progression and increased invasiveness and metastasis of tumor cells.91 One possible mechanism for atorvastatin/celecoxib combination to induce cell cycle arrest is through inhibition of membrane association of RhoA, and this in turn may disrupt RhoA's negative regulation on both p21^Cip1/Waf1 and p27^Kip1 and increase the levels of these 2 CDK inhibitors.56, 62, 92 Interestingly, by an unknown mechanism, atorvastatin/celecoxib combination increased the level of membrane bound RhoB in contrast to RhoA. The increased membrane association of RhoB may contribute to the inhibitory effects of atorvastatin/celecoxib combination on cancer cell growth because of the possible tumor-suppressing role of RhoB.93 An important mechanism for celecoxib to induce apoptosis is its inhibition of the Akt pathway.73 It was found that celecoxib, at low levels where no inhibition on Akt was observed by itself, significantly synergized atorvastatin to abolish phosphorylation of Akt in colon cancer cells.56 Consistently, the same combination treatments decreased phosphorylation levels of PDK1 and PI3K, the upstream kinases of Akt. Moreover, the combination treatments potentially increased PTEN activity by inhibiting phosphorylation of PTEN at Ser380. All of these effects were either not seen or of a much lesser extent in the colon cancer cells treated with atorvastatin or celecoxib alone. The inhibition on the Akt pathway may play a pivotal role in the apoptosis induced by the atorvastatin/celecoxib combination.56 It is important to point out that the 2 human colon cancer cell lines that we used had impaired COX-2 expression.56 HCT116 cells lack COX-2 protein,94 whereas HT29 cells express enzymatically inactive COX-2 protein.95 As a consequence, the effects of celecoxib and its combination with atorvastatin observed in this study were independent of COX-2 activity. Our unpublished results on combination of licofelone, a dual inhibitor of COX-1 and 2 and 5-lipoxygenase, and atorvastatin did not find the significant synergy in the inhibition of HCT116 cell growth. Similarly, in H1299 human lung cancer cells, which express COX-2 and produce prostaglandin E2 (presumably through COX-2 activity),96, 97 we found strong synergy by the atorvastatin/celecoxib combination, but not by the atorvastatin/licofelone combination. These findings suggest that inhibition of COX-2 may not be a key component in the synergy between atorvastatin and celecoxib, at least in the colon and lung cancer cells tested. Further investigation is needed to confirm the role of COX-2 in the statins/NSAIDs combination.

Concluding remarks and future directions

Based on the earlier reviewed results from laboratory and epidemiological studies, we propose the combination of statins and NSAIDs as a promising regimen for cancer chemoprevention. Although the existing data are mostly on colon cancer, and some on prostate cancer, the combination of statins and NSAIDs may also be useful for the prevention of other cancers. In order to bring this concept to fruition, we suggest more research be carried out in the following areas:

• 1
  Laboratory studies: More studies in cell lines, xenograft models and animal carcinogenesis models are needed to explore the synergistic inhibitory effects of combinations of statins and NSAIDs against different cancers. The mechanisms of the synergistic action should be an exciting area for more research. These studies may reveal new targets for cancer prevention and new biomarkers for chemoprevention trials.
• 2
  Epidemiological studies: The exciting data on the synergistic anticancer actions of statins and NSAIDs should form the rationale for additional retrospective and prospective studies in humans on this subject.
• 3
  Chemoprevention trials: Randomized, double-blind, placebo-controlled trials are needed to clearly demonstrate the efficacy of the combinations of low doses of statins and NSAIDs. There are already enough data to provide a strong rationale for a trial on colon cancer. The advantage of using statins and NSAIDs is that their safety for human use and even their cancer preventive activities have been well studied. Therefore, it would be prudent to start trials with low doses of statins (e.g., 10 mg atorvastatin daily) in combination with NSAIDs (e.g., 100 mg celecoxib twice a day or 81 mg aspirin twice a day); for example, in a 2 × 2 factorial design.

For statins and NSAIDs that are mainly metabolized by cytochrome P450 3A4, possible drug–drug interactions between the 2 drugs should be studied.