The most award winning
healthcare information source.
TRUSTED FOR FOUR DECADES.
Abstract & commentary
Synopsis: Soslow and colleagues studied the expression of COX-2 in tumor specimens. Whereas COX-1 was constitutively expressed at low levels, COX-2 expression was limited to malignant, premalignant and adjacent nonneoplastic epithelium. The potential for inhibition of COX-2 in oncology is discussed.
Source: Soslow RA, et al. Cancer 2000;89:2637-2645.
Twenty cases each of lung, colon, and breast cancer specimens were randomly obtained for a retrospective analysis of cyclooxygenase (COX) expression as determined by immunohistochemistry. The results were reported as an immunohistochemical score (IHS) ranging from 0-12. The IHS was determined by the product of two numbers representing 1) the percentage of immunoreactive cells (scale 0-4) and 2) the staining intensity (scale 0-3). The IHS was considered strong if this product was 9-12, moderate if 5-8, weak if 1-4, and negative if 0. COX-1 expression was found to be ubiquitous with no variation between benign or neoplastic tissue whereas COX-2 expression varied depending on location.
The 20 lung cancer cases included 16 with adenocarcinoma or large cell carcinoma, the remainder being either squamous cell carcinoma (3) or carcinoid (1). Overall, 90% of the lung cancers expressed significant levels of COX-2. Benign tissue over 2 cm from the malignant tumor showed no COX-2 expression, although 4 out of 14 samples from tissue adjacent to malignant areas did show COX-2 expression.
The 20 cases of colon cancer were all adenocarcinomas with the exception of one neuroendocrine carcinoma. Again, benign tissue over 2 cm from the malignant tumor showed no COX-2 expression. On the other hand, 71% of the adenocarcinomas contained moderate to strong COX-2 expression. The likelihood of COX-2 expression varied with the differentiation of the colon cancer. This ranged from no expression in poorly differentiated tumors to frequent expression in adenomas (86%).
The 20 breast cancer cases contained infiltrating carcinoma (17) or DCIS alone (3). As was seen with the other tumor types, no COX-2 expression was found in distant benign tissue. Whereas only 41% of the infiltrating carcinomas expressed COX-2, expression was seen in 80% of DCIS lesions.
COMMENT by Kenneth W. Kotz, MD
The formation of prostaglandins from arachidonic acid is catalyzed by the COX enzyme. The well-known anti-inflammatory effects of NSAIDs are mediated by inhibition of COX. The discovery of two COX isoforms led to the ability to specifically inhibit COX-2, allowing for more specific control of inflammation with less side effects. COX-1 is generally thought to be constitutively expressed in many cells and function as a "housekeeping" enzyme, examples including protecting the gastric mucosa and maintaining renal blood flow. On the other hand, COX-2 expression seems to be induced in response to injury or inflammation. Inhibiting COX-2 will result in decreased prostaglandin formation thereby mediating the inflammatory response. The lack of COX-2 expression in locations such as platelets allows for an increased therapeutic ratio. The categorization of the relative roles of COX-1 and COX-2 as "housekeeping" and "inflammatory response" enzymes, respectively, may be oversimplified.1
Two COX-2 inhibitors are available in the United States. Rofecoxib (Vioxx) is approved for use in osteoarthritis, acute pain in adults, and primary dysmenorrhea. Celecoxib (Celebrex) is approved for use in osteoarthritis, rheumatoid arthritis, and of particular interest to oncologists, familial adenomatous polyposis (FAP). In this latter disorder, the drug was shown to reduce the number of adenomatous colorectal polyps, but with unproven effects on cancer reduction, morbidity, or mortality. Whereas the dose used in arthritis is 200-400 mg per day, the dose recommended for FAP is 800 mg total per day, increasing the risk of side effects. In a study of 83 patients, the 800 mg total dose reduced polyp formation by 28%, compared with 12% for 200 mg total per day, and 5% for placebo.
Studies of COX-2 inhibition will interest oncologists because of the potential role of COX-2 in neoplastic physiology. First, COX-2 overexpression may prevent normal pathways of apoptosis. Second, COX-2 may promote angiogenesis. Third, COX-2, which can be expressed in response to tumor promoters, may impair immune surveillance possibly through the effects of prostaglandin E2.1 Inhibition of these processes may halt the transformation of premalignant lesions and possibly even reverse established malignant behavior. The potential role of COX-2 in the development of malignant clones is suggested by the results of Soslow et al who observed moderate immunoreactivity for COX-2 not only in adjacent nonneoplastic tissue, but also in premalignant lesions (atypical adenomatous hyperplasia of the lung, colonic adenomas, and DCIS of the breast).
There are numerous preclinical studies, some referenced by Soslow et al, that support testing the potential of inhibiting COX-2 in oncologic diseases. In fact, there are several open clinical trials looking at the role of COX-2 inhibition in oncology. Examples include: 1) the use of celecoxib with trastuzumab in women with HER2/neu overexpressing metastatic breast cancer that is refractory to trastuzumab; 2) the use of celecoxib in preventing disease recurrence in patients with bladder cancer; 3) the use of celecoxib to prevent cancer in patients with Barrett’s esophagus; and 4) the use of celecoxib to prevent polyp formation in patients treated for sporadic adenomatous polyps. Certainly, it is too early to use COX-2 inhibitors for prevention or treatment of malignant or premalignant conditions outside the context of a clinical trial, but it is certainly exciting to see if the ability to selectively inhibit COX-2 will be useful to the practice of oncology.
1. Buttar N, et al. Mayo Clin Proc 2000;75:1027-1038.