Background
As a malignancy with particularly poor prognosis, novel therapeutic options are urgently required for the treatment of ovarian cancer[
1,
2] In 2009, approximately 25,000 women will be diagnosed in North America and most will die of their disease, making it the fifth leading cause of cancer mortality in women[
3] The majority of ovarian cancer cases present as advanced stage III or IV disease and treatment usually involves surgical cytoreduction, followed by adjuvant platinum/taxane chemotherapy, with about 70-80% response rates. While patients typically undergo a period of remission of 1-2 years, more than half eventually relapse. Some patients with recurrent disease become refractory to platinum treatment. They are generally next treated with regimens of gemcitabine, topotecan, and/or liposomal doxorubicin, but with very limited success[
4,
5] The reduced rate of response in these patients is typically due to the development of drug resistance[
6] Taken together, to directly increase the quality and longevity of life, new and immediate therapeutic approaches are urgently required to combat ovarian cancer.
We and others have shown that the statin family of drugs exhibit antiproliferative activity against cancer cells without causing collateral damage to normal cells[
7]. Statins inhibit the rate-limiting enzyme of the mevalonate (MVA) pathway, HMG-CoA reductase (HMGCR), and have been used for decades as safe and effective agents in the control of hypercholesterolemia[
7,
8] In addition to cholesterol, the MVA pathway gives rise to a number of crucial biochemical end-products, including ubiquinone, dolichol, isopentenyladenine, and isoprenoid precursors. Statins can trigger tumor cells to undergo a classic caspase-dependent, apoptotic response that is reversible by exogenous addition of MVA or the isoprenoid precursors, geranylgeranyl pyrophosphate (GGPP) and farnesyl pyrophosphate (FPP)[
7]. Thus, the statin family of drugs are immediately available for use as part of the arsenal of molecular targeted therapeutics to combat cancer.
Like most anti-cancer agents, statins demonstrate robust efficacy on some but not all tumor-types, emphasizing the importance of matching the agent with the sensitive, responsive cancer. Statins have been extensively shown to trigger apoptosis of cell lines derived from haematological malignancies, including acute myelogenous leukemia and multiple myeloma[
7,
9]. This preclinical data has been recently translated to Phase I/II clinical trials that have shown promising results when statins have been used in combination with standard chemotherapy[
10,
11] Similarly, median survival was doubled with the addition of statins to 5-fluorouracil in advanced hepatocellular carcinomas[
12] As was recently reviewed, solid tumor derived cell lines that have recently been shown to be statin sensitive include breast, colorectal, lung, prostate, and pancreatic, [
13]. however, preclinical work focusing on ovarian cancer is required to determine whether statins have the potential to be used to combat this tumor type as well. Very recently, preliminary reports have indicated that ovarian carcinoma is sensitive to statin-induced apoptosis, providing a unique alternative to treating this deadly disease[
14,
15].
To advance these findings, we demonstrate that lovastatin induces apoptosis of ovarian cancer cells in a p53-independent manner and synergizes with doxorubicin, a chemotherapeutic agent used to treat recurrent ovarian cancer. Lovastatin triggers ovarian tumor cells to undergo apoptosis by two mechanisms: first, by blocking HMGCR activity; and second, by increasing the level of doxorubicin within drug-resistant cells. Together, these data support further pre-clinical and clinical evaluations of statins as a new strategy to combat ovarian cancer and overcome drug resistance.
Discussion
Our work provides important evidence to support further pre-clinical and clinical evaluation of the statin family of drugs as anticancer agents against ovarian cancer. We show that a panel of ovarian cancer derived cell lines is sensitive to lovastatin-induced apoptosis, consistent with recent reports[
14,
15]. Mechanistically this apoptotic pathway is functionally blocked by exogenous MVA or the isoprenoid precursors GGPP and FPP. Moreover, we show that statin killing occurs irrespective of the mutational status of the tumor suppressor p53. Our results using a dominant negative p53 clearly indicate that lovastatin-induced apoptosis is substantially p53-independent and this is also supported by the observation that p53-null SKOV3 cells are able to undergo lovastatin-induced apoptosis. These observations are particularly important for ovarian cancer in which p53 mutation rates have been estimated between 23 and 79%[
32]. We also show that lovastatin can synergize with doxorubicin and potentiate apoptosis. Synergy is achieved by lovastatin blocking drug efflux through a MVA-independent mechanism that enables the intracellular retention and genotoxic action of doxorubicin. To the best of our knowledge, these latter features of statin-induced apoptosis have not yet been reported for ovarian cancer. Exploiting the unique ability of statins to drive apoptosis by the mevalonate-dependent mechanism alone warrants further evaluation of these agents in the treatment of ovarian cancer (Figure
6D, left side). In addition, using statins, like lovastatin, to synergize with chemotherapeutics that are P-gp substrates (Figure
6D, right side) may be a feature of lovastatin action that further maximizes ovarian cancer cell death and improves patient survival.
It is interesting to note that while several reports have shown that P-gp expressing cells were more sensitive statin-induced apoptosis, [
33‐
36] our results show the opposite trend (Figure
1A). Indeed, the MTT
50 results for lovastatin in A2780ADR and A2780CIS cells are roughly 5-fold higher than in the parental A2780 cells. The reason for this difference is unknown, but it is possible that the drug resistant cells have exploited additional mechanisms of resistance, such as increasing the expression of anti-apoptotic proteins.
As agents approved for use in humans, the MVA-dependent antiproliferative activity of statins has prompted several Phase I clinical trials of statins on a wide variety of late-stage cancers, and although statins were well tolerated, only limited responses were evident. More recently statins have been evaluated on cohorts of patients harboring a tumor-type that has been shown to be sensitive to statin-induced apoptosis in tissue culture studies. In these focused, tumor-specific, hypothesis-driven trials, statins have demonstrated some efficacy as a single agent[
29,
30,
37] but more wide-reaching effects were evident when statins were combined with chemotherapeutics [
10,
11,
38,
39]. Thus, our data identifying ovarian carcinoma as a statin-sensitive tumor type strongly supports the evaluation of statins in strategies to combat this disease.
A recent, retrospective epidemiological study showed that statin use in patients diagnosed with epithelial ovarian cancer is associated with improved survival[
40]. Although only a relatively small number of patients met the criteria for the study, multivariable analysis identified statin use as an independent positive prognostic factor after controlling for age, stage, grade, and suboptimal cytoreduction, providing clinical support for the use statin-based combinations in cancer treatment. Similar recent analyses of breast cancer also provided additional insights. For example, it appears that lipophilic statin use after breast cancer diagnosis has been associated with decreased risk of recurrence[
25,
26]. Overall, these recent studies provide supporting rationale for the use of statins as anticancer agents and suggest that lipophilic statins (lovastatin, simvastatin, atorvastatin, fluvastatin, and pitavastatin) may be more effective, perhaps because they are better able to penetrate solid tumors compared to hydrophilic statins. From a pharmacological perspective, the lipophilic statins that demonstrate higher plasma concentrations with longer retention times in the circulation include atorvastatin and fluvastatin. This suggests these lipophilic agents may best target the tumor and show higher anti-cancer efficacy
in vivo, consistent with a previous study comparing lipophilic and hydrophilic statins in ovarian cancer[
15].
Recent evidence suggests that there may be a connection between drug resistance and regulation of the MVA pathway. In MDR AML cells, HMGCR mRNA levels were not significantly elevated upon statin exposure in cells that showed preferential sensitivity to lovastatin[
36]. More recently, it was suggested that high levels of HMGCR mRNA correlates with sensitivity to statin-induced apoptosis[
15]. It will be interesting in the future to determine how HMGCR expression impacts statin sensitivity and whether it can be exploited as a biomarker.
Mechanistically, it is clear that statins target HMG-CoA reductase and similarly trigger tumor cells to undergo apoptosis[
7]. (Figure
6D, left side), however, several practical questions remain unresolved regarding statins as potential P-gp inhibitors (Figure
6D, right side). This new role of statins would be particularly important to consider in the management of ovarian cancer as survival and disease recurrence after taxane/carboplatin treatment has recently been associated with specific P-gp polymorphisms[
41]. Several classes of specific P-gp inhibitors have been developed but have unfortunately shown general cytotoxicity in clinical trials[
42]. This is thought to be due to targeting P-gp not only on tumor cells, but also on several normal vital organs that constitutively express P-gp. It would be easy to assume that statins blocking P-gp will similarly cause general cytotoxicity, however, it is not known whether statins and classic P-gp inhibitors are mechanistically or functionally similar. Lovastatin has been reported to inhibit P-gp in a limited number of biochemical studies with two very distinct caveats: none have used human cells overexpressing drug-selected human P-gp and the concentrations of drug used have been well beyond the physiologically achievable range [
43‐
46]. Moreover, the results of these studies have been in conflict when using either the acid or lactone form of the statin[
45,
46]. Importantly, we conducted our work with physiologically attainable concentrations of both doxorubicin and lovastatin in human cell systems selected to overexpress human P-gp.
It is also worth noting that Bcl-2 was unable to inhibit cell death induced by the combination of lovastatin and doxorubicin (Additional file
2: Supplemental Figure S2). While the reasons for this result are unclear, it is possible that the cells have become drug-resistant through means other than the MDR machinery, such as upregulation of one or more anti-apoptotic proteins, and thereby rendered forced expression of Bcl-2 incapable of rescuing cells further. Further study will be required to better understand the interplay of all mechanisms of drug resistance.
Statins ultimately need to advance to clinical trials where their inhibition of drug efflux can be monitored on both tumor and normal cells. Interestingly, other groups have reported that lovastatin protects normal cells from doxorubicin-induced cytotoxicity [
47‐
49] which, when combined with our data, suggests that statins may affect P-gp differently in normal cells compared to tumor cells. It is entirely possible that lovastatin functionally blocks P-gp in a manner that is distinct from classic P-gp inhibition. Evidence that statins can be successfully combined with various P-gp substrates is also established from their safe and effective combination in the polypharmacy of cardiac patients with hypercholesterolemia[
50]. Taken together, our results suggest the ability of statins to trigger apoptosis of ovarian cancer cells may be exploited in the treatment of this disease, and that the potential P-gp inhibitory properties of certain statins, like lovastatin, warrant further investigation. It is also of interest to note that at MTT
50 concentrations, but not higher, lovastatin had a slightly antagonistic relationship with cisplatin, a non-P-gp substrate (Figure
4A, Additional file
1: Supplemental Figure S1). This observation suggests that it could potentially be deleterious to combine lovastatin with cisplatin in the treatment of some patients. Furthermore, lovastatin and doxorubicin were also able to synergize in A2780 parental and A2780CIS cells. While this suggests that elements other than P-gp are involved in the interaction between these two drugs, the degree of synergy observed in A2780ADR cells is higher, indicating that inhibition of P-gp is likely an important mechanism of how lovastatin synergizes with doxorubicin. These results require further investigation to truly understand the manner by which lovastatin functionally interacts with other chemotherapeutics.
Determining which statin will maximally target different tumors, including ovarian, under different conditions will also be vital to advancing patient care. In the 14 completed and 20 or more ongoing clinical trials evaluating statins in the prevention or treatment of cancer, [
10‐
12,
30,
37‐
39,
51‐
57] the rationale for choosing a particular statin is not presented and appears random. Indeed, the ideal choice of statin as an anti-cancer agent remains unclear, however, evidence suggests lipophilic agents with pharmacologic properties that favor access to solid tumors is of high priority. Further work is required to better understand the activity of these statins as potential inhibitors of P-gp and to determine if this inhibition is specific to tumor cells
in vivo.
Acknowledgements
We would like to express our profound appreciation to Dr. Theodore J. Brown for his enthusiasm in the initiation of this project. Sincere thanks to members of the Penn lab for helpful discussions and critical review of the manuscript as well as to the following for providing necessary reagents: Dr. Theodore J. Brown (The Samuel Lunenfeld Research Institute, Toronto, ON, Canada; Skov3, Ovca429, Hey, and HOC-7 cells), Dr. Richard B. Pearson (Peter MacCallum Cancer Centre, Melbourne, Australia; DOV13, Ovca432, and OVHS-1 cells), Dr. Jeremy Squire (University of Toronto, Toronto, ON, Canada; A2780, A2780ADR, and A2780CIS cells), Dr. Aaron Schimmer (University of Toronto, Toronto, ON, Canada; CEM and CEMVBL cells), Dr. David Andrews (McMaster University, Hamilton, ON, Canada; Bcl-2 antibody), and Apoptex Corporation (lovastatin). This study was conducted with the support of the Ontario Institute for Cancer Research Network through funding provided by the Province of Ontario (LZP), a fellowship from the Leukemia and Lymphoma Society of Canada (AM), Excellence In Radiation Research for the 21st Century Strategic Training Initiative In Health Research award from the Canadian Institutes for Health Research (JWC), and Ontario Graduate Scholarships (JWC and CAG).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
AM and JWC carried out experiments, participated in the design of the study, and contributed to drafting the manuscript. CAG carried out experiments. LZP conceived of the study, participated in its design and coordination, and helped to draft the manuscript. All authors read and approved the final manuscript.