Abstract
An enormous body of data confirms the efficacy of aspirin, non-aspirin nonsteroidal antiinflammatories (NA-NSAIDs), and COX-2 selective inhibitors (COXIBs) against neoplastic development. These protective data are consistently reported in animal models of intestine, esophagus, lung, bladder, skin, and mammary cancers, and less consistently - but nevertheless positively - reported in other model systems as well, including prostate and oropharynx. Complementing these data, robust epidemiologic investigations have now confirmed preventive activity of these agents against colorectal, esophageal, stomach, lung, breast, and prostate cancer, and suggested some level of efficacy in others, including bladder and skin cancer. The efficacy of these agents in treating or suppressing neoplasia typically has been most profound when given early in the course of disease. In the colorectum, where there is the most data, preventive effects have been described across a spectrum of progression (e.g., ACF, adenomas, cancer, and cancer-associated mortality) following early treatment.
COX-2, in particular, is commonly over-expressed early in human carcinogenesis across a wide spectrum of tumors and many studies have correlated its over-expression with poorer outcomes among cancer patients. Although incompletely understood, the best-described chemopreventive activity of these agents on the molecular level relates to the inhibition of one or both cyclooxygenase enzymes (i.e., COX-1 and COX-2), although many other non-COX related mechanisms, such as stimulation of peroxisome proliferator-activated receptors and inhibition of Akt phosphorylation, have been reliably described. At the cellular or tissue level, the apparent mechanisms of action are numerous as well, including inhibiting proliferation, neoangiogenesis and invasion, and stimulating apoptosis and immunosurveillance.
Strategies targeting COX enzymatic activity have been translated into the clinic with great success in terms of efficacy, but with concerning toxicities following chronic administration. In clinical trials, aspirin reduces the risk of recurrent colorectal adenomas, cancer, and cancer-associated mortality, but increases the risk of serious upper gastrointestinal bleeding. Some observational studies even suggest that aspirin use may be associated with reductions in overall mortality. In several clinical trials, COXIBs have been shown to reduce recurrent colorectal adenomas, with even greater effects against more advanced lesions. However, these same trials have also demonstrated that chronic administration of COXIBs at high doses can be associated with a significant 2–3fold increased risk of serious cardiovascular events. Indeed, these concerning side effects have been described with most NA-NSAIDs and COXIBs, with the clear exception of aspirin (and possibly naproxen, in some studies).
Following the initial demonstration of COXIBs' efficacy against colorectal neoplasia, but with enthusiasm blunted by serious cardiovascular toxicities, a number of smaller studies recently have shown celecoxib's potential preventive effects against neoplasia more broadly, involving a number of additional tissues. These phase II trials of COXIBs have demonstrated reductions in: bronchial proliferation in former smokers, breast cell proliferation in women with breast cancer, invasive transitional cell bladder cancers prior to definitive cystectomy, and squamous and basal cell skin cancer incidence.
Therefore, one of the greatest challenges in developing these agents for prevention lies in trying to tip the therapeutic index to improve benefits and/or reduce risks. At least four strategies have evolved toward this goal: 1) modifying the recommended regimen to try to separate efficacy from safety concerns, 2) providing concomitant agents for risk reduction such as proton pump inhibitors to reduce GI toxicities, 3) identifying combinations of agents promising synergy, and 4) identifying patients most likely to benefit and/or least likely to be harmed by aspirin, NA-NSAIDs or COXIBs. Of these, combinatorial strategies have been extremely promising preclinically, and one human trial of sulindac + eflornithine has shown profound efficacy against recurrent colorectal adenomas in a placebo-controlled trial. The idea of using pharmacogenomics to identify susceptible individuals for treatment is also promising in early studies.
Alternatively, investigators are also looking to the biology of COX-2 expression and signaling to identify other therapeutic options that might offer an improved risk:benefit profile. The process of COX-2 expression is highly regulated at both the transcriptional and post-transcriptional levels, offering opportunities to reduce expression or inhibit post-translational stability of transcripts, and thereby suppress COX-2 signaling. A variety of dietary supplements and traditional medicines have recently been identified to influence these processes. Another approach is to reduce COX-2 signaling by inhibiting derived prostanoids' (e.g., PGE2) interactions with downstream EP receptors. Several pharmaceutical companies are exploring the feasibility, efficacy, and safety of EP receptor antagonists, and preliminary preclinical data appear promising. Other investigations are exploring the potential of a novel product associating aspirin or a NA-NSAID with phosphatidylcholine which reduces upper GI toxicities by 70% in short-term clinical studies. Alternative administrative strategies, such as intermittent dosing, deserve further exploration as well. It would be similarly important to understand the mechanisms underlying the serious cardiovascular events associated with these agents, so they might be mitigated. Hypotheses advanced thus far have focused on COXIBs' induction of an imbalance between prostacyclin and thromboxanes, or COXIBs' tendency to induce fluid retention and hypertension as preludes to cardiovascular events.
Given the toll of cancer and the tremendous opportunity for aspirin, NA-NSAIDs, and COXIBs' contributions to reducing cancer risks at a number of sites - and potentially, across several other common diseases of aging more broadly - there is an urgent need to refine and re-double our developmental efforts, rather than abandon one of the most consistent and powerful classes of chemopreventive agents known.
Citation Information: Cancer Prev Res 2011;4(10 Suppl):ED04-02.
RSS Feeds