Inflammation and Cancer: Is the Link as Simple as We Think?

Abstract

For investigative cutaneous biologists a link between inflammation and skin cancer is a topic frequently discussed when multi-stage chemical carcinogenesis and photocarcinogenesis models are presented. Indeed, the inflammatory process includes the release of cytokines, growth factors, chemotactic polypeptides, and prostaglandins, which have become favorite targets for prevention of chemical induced as well as UV-light-induced skin cancer (Marks and Furstenberger, 2000; Tripp et al, 2003; Wilgus et al, 2003; Bowden, 2004). Dissecting the mediators of inflammation in cutaneous carcinogenic pathways has revealed key roles for prostaglandins, cyclooxygenase-2, tumor necrosis factor-a (TNF-a), AP-1, nuclear factor-kB (NF-kB), signal transducer and activator of transcription (STAT)3, and others (Buckman et al, 1998; Moore et al, 1999; Suganuma et al, 1999; Young et al, 1999; Chan et al, 2004; Lind et al, 2004). Besides these experimental results, several clinical conditions associated with inflammation appear to predispose the patient to increased susceptibility for skin cancer including discoid lupus erythematosus, dystrophic epidermolysis bullosa, and chronic wound sites. Despite this vast collection of data and clinical observations, however, there are several dermatological settings in which chronic, and even lifelong, skin diseases associated with inflammation do not predispose to conversion of lesions into malignancies such as psoriasis, atopic dermatitis, and Darier’s disease. The purpose of this commentary is to summarize a rapidly accumulating body of evidence providing new molecular insights linking inflammation with the promotion of tumorigenesis, and to remind the reader that such a link may not be as simple as currently portrayed because certain types of inflammatory processes in skin (and possibly other tissues as well) may also serve a tumor suppressor function. Here we focus on such a dualistic viewpoint, and offer a new perspective on potential molecular mechanisms that regulate the pro-oncogenic, as well as the tumor suppressor, effects of chronic inflammation, and tumor development using psoriasis as our primary disease model. Over the past few months, several publications in leading biomedical journals grappled with an important issue in oncology, namely, defining potential links between chronic tissue damage, inflammation, and the development of cancer (Beachy et al, 2004; Bhowmick et al, 2004; Greten et al, 2004; Pikarsky et al, 2004). In an essay concept entitled ‘‘Mending and Malignancy,’’ Beachy et al (2004) point out the clinical link between persistent heartburn and adenocarcinoma of the esophagus. They focus on injured epithelium and highlight molecular pathways contributing to an expanded stem cell pool that, together with mutagenic agents, contribute to carcinogenesis. They conclude that understanding the response to epithelial injury may lead to useful strategies for cancer prevention and therapy. In addition, Balkwill and Coussens (2004) reviewed the role of the NF-kB signal transduction pathway that can regulate inflammation and also promote malignancy. Their review summarized the latest findings revealed in a letter to Nature by Pikarsky et al (2004). Using Mdr2 knockout mice in which hepatitis is followed by hepatocellular carcinoma (HCC), Pikarsky et al implicated TNF-a upregulation in tumor promotion of HCC, and suggest that TNF-a and NF-kB are potential targets for cancer prevention in the context of chronic inflammation. A similar conclusion was reached with respect to NF-kB by an independent group of investigators using a model of experimental dextran sodium sulfate-induced colitis, in which inactivation of the IkB kinase resulted in reduced colorectal tumors (Greten et al, 2004). Way before the molecular clues for cancer had been investigated, Hawkins (1835) made the clinical observation that squamous cell carcinoma can be a long-term sequela of chronic osteomyelitis in the overlying skin. Furthermore, it has been argued that the immune response contributes significantly to fostering a tumorigenic process through CD4þ T cells, or other inflammatory links (Daniel et al, 2003; Vakkila and Lotze, 2004). Although there are many other clinical conditions supporting the concept that inflammation is a critical component of tumor progression (e.g., reflux esophagitis/ esophageal cancer; inflammatory bowel disease/colorectal cancer) (Coussens and Werb, 2002), there is at least one notable example that does not fit this paradigm. As described below, psoriasis is a chronic cutaneous inflammatory disease, which is seldom if ever accompanied by cancer, suggesting the relationship between tissue repair, inflammation, and development of cancer may not be as simple as portrayed by the aforementioned reviews and experimental results. Besides psoriasis, other noteworthy observations pointing to more complexity include the observation that in the Mdr2 knockout mice, we rarely detect bile duct tumors despite extensive inflammation, NF-kB activation, and abundant proliferation of bile ducts in portal spaces (Pikarsky et al, 2004). Moreover, in a skin cancer mouse model, NF-kB was shown to inhibit tumor formation (Dajee et al, 2003). Thus, the composition of inflammatory Abbreviations: NF-kB, nuclear factor-kB; TGFb, transforming growth factor b; TNF-a, tumor necrosis factor-a