Is C-Reactive Protein an Inflammation Opsonin That Signals Colon Cancer Risk?

Colorectal cancer is a leading cause of cancer-related morbidity and mortality in the Western world. A personal history of inflammatory bowel disease (IBD) is a well-established risk factor for colorectal cancer. Recent data suggest that the risk of colorectal cancer for individuals with IBD increases by 0.5% to 1.0% yearly, 8 to 10 years after diagnosis.¹ The magnitude of colorectal cancer risk increases with early age at IBD diagnosis, longer duration of symptoms, and extent of the disease.¹ In addition, a wealth of experimental, epidemiological, and randomized trial evidence suggests that long-term use of aspirin and other nonsteroidal anti-inflammatory drugs (NSAIDs) can reduce the risk of colorectal adenoma and cancer.^{2 - 4} In both instances, the duration of either the inflammatory process or the use of anti-inflammatory agents is emerging as a key predictor of increased and decreased colorectal cancer risk, respectively.

In this issue of THE JOURNAL, Erlinger and colleagues⁵ report that C-reactive protein (CRP) is predictive of an increased incidence of colon cancer but not rectal cancer in the CLUE II cohort, a prospective study of 22 887 residents of Washington County, Maryland, who provided a blood sample in 1989. Their finding is additional evidence that persistent inflammation is a risk factor for the development of colon cancer.

C-reactive protein has an interesting history and was one of the key molecules ushering the entrance of molecular medicine in the last century. Discovered in 1930 at Rockefeller Hospital in New York City,⁶ CRP was uncovered as a serum fraction denoted "C" present in the serum of febrile patients admitted with pneumonia. This fraction contained a protein that precipitated polysaccharide components of the pneumonia bacteria. The authors of that report, Tillett and Francis,⁶ were studying the immune response in patients with pneumonia to the specific capsular polysaccharide and to pneumococcal nuclear protein fractions. In the late 1920s, pneumonia was a major cause of death in New York City. A number of investigators were drawn to study this mysterious and dangerous disease and confirmed that the precipitant had the pneumococcal carbohydrate components. It quickly became apparent that CRP as well as other proteins rapidly increased in plasma in response to infection.⁷ Today, it is believed that CRP binds to damaged cellular tissues and components, serving as an opsonin that prepares this debris for phagocytosis and clearance by macrophages and leukocytes. Thus, CRP binds to blood cells including platelets and leukocytes.⁸

It is thought that a certain degree of local inflammation and tissue injury is needed to signal the liver, the site of CRP biosynthesis, to initiate synthesis and release of the protein into the blood⁹ ; therefore, a signal from inflammatory foci may correlate with colon cancer development.

It is possible that the persistent production of CRP reflects the signaling of the innate immune system. Taking the view that a tumor is a parasite to the host,¹⁰ the proliferation of the neoplastic cell type and disruption of healthy tissue architecture and function increase at a rate that exceeds the ability of the cytotoxic lymphocytes and macrophages to destroy and clear these aberrant cell types. Signaling the liver to produce the opsonin CRP should enhance and facilitate the clearance of the tumor debris. This may be the first recognition of a communication link between the colon and liver, indicating that a local inflammatory response is ongoing. Thus, the initial proliferation in a neoplasm may be recognized by the system as a local inflammation. In many tumors subject to early histological examination, there is evidence of an early leukocytic infiltrate or leukocytic rim that likely arises from debris.

A weakness of the study by Erlinger et al⁵ is that CRP was measured only once. Given the protracted preclinical period for colorectal cancer, it is possible that increased CRP level is a consequence of the disease, rather than a risk factor. The authors addressed this issue by showing the absence of an association between CRP level and follow-up time among colon cancer cases. In future studies, it would be important to better characterize the temporal relationship between CRP level and appearance of colon cancer through serial measurements. Another shortcoming of the study is that use of aspirin and other NSAIDs was assessed only by self-report within the last 48 hours before samples were taken for CRP level determination. Also, use of aspirin and use of other NSAIDs were combined in the analyses. It is now clear that aspirin acts differently with cyclooxygenase (COX)-1 and COX-2 to inhibit prostaglandins and has a unique action on vascular COX-2¹¹ that triggers production of novel endogenous anti-inflammatory mediators.¹² These properties of aspirin are not shared by COX-2 inhibitors.¹³ Consequently, the potential impact of aspirin and other NSAIDs was not optimally assessed. Last but not least, the small number of cases and controls with a family history of colorectal cancer in this study precluded analysis of a possible association between CRP level and familial colon cancer.

Several studies have shown that this might indeed be the case. The observational data regarding NSAIDs and colorectal neoplasia are quite strong, as the finding in observational studies of a protective effect of NSAIDs is reinforced by data from recent clinical trials showing that aspirin prevents sporadic adenomas.^{3 - 4} Furthermore, use of the NSAIDs sulindac and celecoxib has led to regression of existing colorectal polyps in patients with familial polyposis.^{14 - 15} Aspirin is well appreciated for its ability to block key components in eicosanoid biosynthesis. One is important in thrombogenesis, inhibiting platelet thromboxane production via inhibition of COX-1 and the other via inhibition of the prostaglandin cascade by acting on both COX-1 and COX-2.^{16 - 17} A recent study showed that selective COX-2 inhibition does not modify CRP levels,¹⁸ suggesting that COX-2 inhibitors may not have a direct impact on CRP and could influence colon cancer by other, yet unknown mechanisms.¹⁹

Early studies with CRP showed that administration of prostaglandin E₂ to rabbits and humans led to a rapid increase in CRP level. However, prostaglandin E₂ did not stimulate CRP production in isolated hepatocytes indicating that extrahepatic influences are needed.²⁰ Conversely, CRP can affect eicosanoids that play key roles in inflammation. Also, CRP decreases production of prostacyclin.²¹ Thus, CRP can have a diverse impact on the generation of chemical mediators, such as eicosanoids and particularly prostaglandins.

In addition, it would be interesting to determine if use of aspirin results in more protection against colon cancer and possibly adenoma development among individuals with persistently high CRP levels. This would contribute experimental evidence that inflammation is causally related to colorectal neoplasia.

The future of CRP as a marker of colon cancer risk remains to be further defined. The provocative findings of Erlinger et al⁵ are likely to stimulate basic scientists and epidemiologists alike to further explore the relationship between inflammation and colon cancer. More than 70 years after being discovered in patients with pneumonia, CRP is now linked with one of the most common forms of cancer.