Still, Drs. Golde and Koo realized that, regardless of how NSAIDs might be working to decrease the risk of developing Alzheimer’s, conducting clinical trials of NSAIDs in populations at risk for Alzheimer’s or in those with the disease would be difficult. Long-term use of high-dose NSAIDs can cause stomach ulcers, kidney damage and gastrointestinal bleeding in anyone, and those side effects would be even more prevalent in the elderly. Moreover, if NSAIDs were working by lowering AB42, Dr. Golde knew very high doses of the NSAIDs would be needed to make a difference in Alzheimer’s risk.
This meant that a compound that could successfully and significantly lower AB42 must be one without such severe side effects. So, the first NSAIDs that Drs. Golde and Koo screened were known as COX2 inhibitors because they were believed to be safer. (NSAIDs reduce inflammation because they target enzymes that are known as cyclooxygenases or COX, and classic NSAIDs, such as ibuprofen and indomethacin, nonselectively inhibit the two types of COX enzymes, COX1 and COX2.)
But, again to their surprise, Drs. Golde and Koo found that many COX2 inhibitors actually had the opposite effect on AB42 -- rather then decreasing it, they increased it. The investigators then expanded their search to look more closely at compounds related to NSAIDs that might lower AB42 but result in greatly reduced COX activity.
So they tested a compound called r-flurbiprofen.
R-flurbiprofen is the mirror image of the COX-inhibiting drug s-flurbiprofen, but because it is structurally distinct (much as a person’s right and left hands have the same overall structure