Coxib Therapy and Colorectal Cancer

Evidence suggests that NSAIDs can prevent the development of CRC.(44) CRC is the second leading cause of cancer-related mortalities in the United States, approximately 57,000 in 1999.(4,45) In the United States, 93% of all CRC cases occur in patients over 50 years of age, and the 5-year survival rate for patients with CRC is approximately 60%.(46) Worldwide, CRC accounts for approximately 556,000 mortalities.

Familial adenomatous polyposis (FAP) is a condition considered to be a precursor to CRC.(47,48) It is a rare condition caused by a defect in the gene APC (adenomatous polyposis coli), normally a tumor suppressor, that predisposes one to develop hundreds of colonic polyps. If left untreated, polyps can lead to colon cancer.(49)

Familial adenomatous polyposis, colorectal cancer, and COX

COX-2 is believed to play a role in the development of FAP and CRC. While COX-1 is constitutively expressed in normal GI mucosa, the level of COX-2 is low or undetectable.(50-52) In animal models of FAP or CRC, however, increased expression of COX-2 has been demonstrated.

One study was conducted in multiple intestinal neoplasia (MIN) mice, a model for FAP in humans. In adenomas harvested from MIN mouse intestine, the levels of COX-2 mRNA and protein were approximately threefold higher than levels of COX- 2 in normal mucosa from the same mouse. These findings implicate COX-2 expression at an early, preinvasive stage of CRC.(50) A second study with rats found increased levels of COX-2, but not COX-1, mRNA and protein in colon tumors that developed following treatment with a colorectal carcinogen.(51)

The same differential expression of the COX isozymes has been detected in human colorectal neoplasia. For example, 86% of tumor samples harvested from patients with CRC contained greater levels of COX-2 mRNA relative to those in the same patient's noncancerous mucosa. In 43% of the colorectal adenomas examined, an increase in COX-2 gene expression was also detected, again showing upregulation at an early stage in colorectal carcinogenesis. However, the level of COX-1 mRNA in all carcinomas examined was equivalent to the level seen in normal mucosa.(52)

COX-2, NSAIDs, apoptosis, and tumorigenesis

Apoptosis, or programmed cell death, is an active process that removes mutated or damaged cells, thus contributing to the prevention of cancer development. Disruptions in apoptosis and COX-2-mediated processes may provide some explanation for the promotion of colorectal tumor formation by COX-2 upregulation.

Briefly, upregulation of COX-2 results in decreased levels of the COX substrate, arachidonic acid (AA), and simultaneously, increased production of COXmediated eicosanoids.(52,53) COX-2-mediated prostaglandins stimulate cell proliferation, and other COX-2-mediated factors regulate tumor angiogenesis (tumor growth beyond 2 to 3 mm in size is dependent on tumor angiogenesis).(52-54) Loss of constraint of tumor cell growth is thought to result from decreases in AA, which ultimately result in lower levels of ceramide, a potent inducer of apoptosis.(55) (AA stimulates sphingomyelinase activity to catalyze the conversion of sphingomyelin to ceramide.)

Recent in vitro studies have implicated a key role of COX-2 in mediating mitogenic growth factor signaling and in the downregulation of apoptosis in human colon cancer cell lines.(48) Notably, NSAIDs have been shown to reverse this COX-2 effect in human colon cancer cell lines, promoting apoptosis. In one study, cancer cells were treated with the nonselective NSAID sulindac or its active metabolite, sulindac sulfide. Only sulindac sulfide resulted in dose-dependent apoptosis, which was not reversed by exogenous prostaglandin E2 (PGE2), the major eicosanoid in colon tumors, or by other prostaglandins. Furthermore, exogenous AA, but not a control fatty acid, was a potent inducer of apoptosis, presumably due to increased levels of ceramide. In this experimental model, sulindac sulfide treatment elevated ceramide levels tenfold relative to untreated cells. A synergistic effect on apoptosis was seen when sulindac sulfide and AA were combined.(55)

Similar effects were seen with indomethacin, which also displays tumor-suppressive activity in intestinal epithelial cells. In indomethacin-treated cells, there was a three- to four-fold increase in AA and a six-fold increase in ceramide; 94% of the treated cells underwent apoptosis.(55)

An in vitro study with the coxib SC58125 found increased rates of apoptosis in a human colon-cancer cell line that maintains high constitutive COX-2 expression and prostaglandin production.(56)

Tumor-related angiogenesis mostly relies on tumor cell expression of angiogenic factors and endothelial tube formation. The role of COX inhibition on these processes was investigated in an in vitro model of tumor angiogenesis. Endothelial cells and colon carcinoma cells engineered to differentially express COX-1 and/or COX-2 were co-cultured and exposed to aspirin or to NS-398, a COX-2-selective inhibitor. Inhibition of COX-2 activity by either agent reduced tumor cell production of angiogenic factors. However, aspirin or a COX-1 antisense oligonucleotide, but not NS-398 or a COX-2 antisense oligonucleotide, inhibited endothelial tube formation. Furthermore, tumor cell expression of angiogenic factors resulted in upregulated endothelial cell expression of COX-1. These results suggest that NSAIDs may inhibit angiogenesis by two mechanisms: inhibition of COX-2 activity in colon carcinoma cells to downregulate production of angiogenic factors, and inhibition of COX-1 activity in endothelial cells to suppress endothelial tube formation.(53)

Another study examined the role of COX-1 and COX-2 in tumor growth and angiogenesis using isografts of Lewis lung carcinoma (LLC) cells in COXdeficient "knockout" mice (COX-1-/- or COX-2-/-) or coxib-treated (celecoxib or SC-58125) wild-type mice. Tumor growth was diminished both in size and speed in COX-2 null mice compared with untreated wild-type mice. However, no such difference in tumor growth was observed between COX- 1 null mice and control mice. Furthermore, prior treatment with a coxib inhibited tumor growth, but to a lesser degree than tumor growth in COX-2 null mice. Angiogenesis was also measured using this model, and results from these experiments suggested that COX-2 activity is essential for tumor angiogenesis, implying again that COX-2 activity promotes tumor growth.(57)

The chemopreventive effect of COX inhibition has been seen in various animal models of colon cancer. The tumor load in MIN mice was decreased significantly and in a dose-dependent manner by the nonselective NSAID piroxicam.(58) These results were confirmed in a study of MIN mice treated with sulindac.(59)

Celecoxib demonstrated a chemopreventive effect in male rats in all phases of colon carcinogenesis: initiation, promotion, and progression. The incidence of azoxymethane-induced colon tumors was inhibited in celecoxib-treated rats by 93%; the multiplicity of colon tumors was inhibited by 97%, and the overall colon tumor burden was suppressed by more than 87%.(60)

Rofecoxib resulted in a similar dose-dependent reduction in the number and size of intestinal and colon polyps in MIN (ApcD716) mice. Using a rofecoxib dose comparable in plasma concentration to that achieved in humans treated with rofecoxib 25 mg once daily, there was a 55% reduction in the number of all intestinal polyps and an 80% reduction in the number of polyps more than 1 mm in size.(49)

Based on these preclinical findings, large epidemiological studies were conducted to examine the impact of NSAID use on the development of colon cancer. Almost every study found a strong correlation between continuous NSAID use and decreased incidence of CRC in humans.(47)

The mounting evidence from preclinical and epidemiological studies was the basis for clinical trials of NSAID treatment for individuals with FAP. Results from three controlled clinical trials found that treatment with sulindac resulted in substantial regression of adenomatous polyps.(61-63) However, virtually all patients experienced regrowth of adenomatous polyps after sulindac therapy was discontinued.(7,54,64)

In a recent clinical trial, celecoxib 400 mg twice daily for 6 months in 30 patients with FAP resulted in a 28% reduction in the mean number of colorectal polyps (P = .003) and a 30.7% reduction in polyp size (P = .001).(48) Based on these findings, celecoxib received US Food and Drug Administration approval for the treatment of FAP.

Chart 1. Implication of COX-2 in the promotion of colon cancer
There is substantial evidence that the COX-2 isozyme plays a crucial role in the promotion of FAP and CRC.
— Significant upregulation of COX-2 but not COX-1 occurs in animal models and human samples of FAP polyps and colorectal tumors.(50-52)

—COX-2-generated prostaglandins produce angiogenic factors and promote tumor angiogenesis.(53)

—PGE2, produced by COX-2 in colon tumors, suppresses apoptosis in human CRC cell lines and colon tumors.(55)

—Both celecoxib and rofecoxib have a COX-2-specific chemopreventive effect in animal models of CRC when compared with nonselective NSAIDs.(49,60)

—Celecoxib is approved as an adjunct to standard care for the treatment of FAP, a premalignant condition that leads to colon cancer if not treated.

Cancer Pain

Cancer, the second leading cause of death in the United States, is often associated with uncontrolled pain.(8) In 1986, the World Health Organization (WHO) developed a three-step therapeutic guideline, called the WHO analgesic ladder, to improve the management of increasing levels of cancer pain.(65) NSAID therapy is recommended by the WHO for use at all three steps on the analgesic ladder, either alone or in combination with an opioid and adjuvant analgesic (other drugs that enhance analgesic effects).(8,66-68)

Inflammation and cancer pain

Cancer pain is often triggered by the release of inflammatory cytokines from active tumors.(8) NSAIDs produce analgesia in part by inhibiting the release of these inflammatory mediators, thus reducing nocioceptive transmission.(8,66,69)

The most common cause of cancer pain is tumor infiltration of bone.(8,68) Bone metastases occur as a consequence of breast cancer, prostate cancer, lung cancer, or multiple myeloma.(8) One likely mechanism of pain secondary to bone metastasis is the secretion of prostaglandins by carcinomas.(8,68) For this reason, NSAIDs should be included in any regimen to control pain associated with bone metastasis.(8,68-70)

Lema MJ. Emerging options with coxib therapy. Cleve Clin J Med 2002;69:SI76-84.