Cyclooxygenase-2 (COX-2) is the key enzyme involved in prostaglandin synthesis. It appears to support the growth of a number of solid tumours including colon, breast, ovary, lung and uterine cervix and may be an important therapeutic target in at least some of these tumours. COX-2 expression has recently been evaluated (by immunohistochemistry using polyclonal anti-COX-2 antibodies) in multiple myeloma (MM) where expression was documented in 33–57% of patients. COX-2 expression in these studies was strongly associated with an adverse outcome. In addition there is some emerging data to suggest that the use of aspirin in MM may improve survival rates. In order to further evaluate this we have used a monoclonal antibody (Clone SP21, Labvision, Fremont, Ca) to assess COX-2 expression in both normal and neoplastic plasma cells. 52 specimens were assessed using standard streptavidin-biotin immunoperoxidase techniques using a known COX-2+ colon cancer as a positive control. Strong uniform COX-2 expression was seen in 32/33 (97%) of myeloma patients assessed and was also documented in all patients with MGUS (n=10). COX-2 expression was also documented in reactive plasmacytic lesions (oral mucosa, skin and lymph node, n=6) as well as normal bone marrow plasma cells (n=6). Megakaryocytes stained positively in all bone marrow biopsies examined and provided a useful positive internal control while erythroid, myeloid and lymphoid cells were consistently negative. We would conclude that COX-2 is strongly expressed by both normal and neoplastic plasma cells suggesting that COX-2 is a potential therapeutic target in MM. The apparent increase in the proportion of myeloma patients expressing COX-2 in the present study reflects the use of a monoclonal antibody in our immunohistology studies. The fact that polyclonal antibodies identify a lower proportion of patients who appear to have an inferior outcome suggests that the level of expression is of prognostic significance rather than its presence or absence. This is worthy of further study using more appropriate techniques such as RQ-PCR.

Disclosure: No relevant conflicts of interest to declare.

Author notes


Corresponding author