Background

Naproxen is an over-the-counter nonsteroidal anti-inflammatory drug (NSAID). It inhibits COX-1 and COX-2, both isoforms of the cyclooxygenase enzyme, and reduces inflammation through reduced prostaglandin synthesis.

In the Alzheimer's field, interest in NSAIDs arose when epidemiological studies started reporting lower rates of Alzheimer's or cognitive decline among people who had been taking these drugs for chronic treatment of inflammatory conditions (e.g., Mar 1997 news; in't Veld et al., 1998; Nov 2001 news; Vlad et al., 2008; Obermann et al., 2013). Experimental studies supported the argument that inflammation plays a role in Alzheimer's disease, prompting a wave of clinical trials of various NSAIDs, e.g., ibuprofen, rofecoxib, celecoxib, and R-flurbiprofen.

Findings

Three clinical trials of naproxen in Alzheimer's disease have been conducted.

In 2000 and 2001, the Alzheimer's Disease Study Group conducted a 40-center trial that compared one year of treatment with 220 mg twice daily of naproxen or 25 mg once daily of rofecoxib to placebo in 351 people with mild to moderate Alzheimer's disease. The trial assessed whether these drugs would slow cognitive decline, but found that neither showed any consistent benefit over placebo (Jul 2002 conference news; Aisen et al., 2003).

The second trial was the Alzheimer's Disease Anti-Inflammatory Prevention Trial, aka ADAPT. Funded by the NIA, this primary prevention study started in 2001 to enroll 2,625 people 70 or older who had a parent or sibling with Alzheimer's or another dementing illness of aging. Conducted at six U.S. sites, this study aimed to determine whether naproxen or celecoxib can delay the onset of AD or age-related cognitive decline. ADAPT was to conduct annual cognitive assessments for five to seven years; however, the trial soon became embroiled in a public controversy (see Sep 2002 news). In September 2004, Merck withdrew the related NSAID celebrex. In December 2004, the NIA halted dosing in ADAPT, citing an increase in cardiovascular side effects in its naproxen arm (Sep 2004 news; Dec 2004 news). A separate report raising concerns about a slightly elevated risk of heart attack with long-term naproxen and other NSAID treatment soon followed (Graham et al., 2005). ADAPT safety data were subsequently published but failed to end ongoing controversy over the decision to halt the trial (Nov 2006 news).

Efficacy data from ADAPT showed that neither naproxen nor celecoxib delayed incident dementia or cognitive decline; naproxen was reported to have hastened cognitive decline slightly (ADAPT Research Group 2007; ADAPT Research Group 2008). ADAPT came to exemplify conflicting findings between observational epidemiology, which continued to report protective effects of NSAIDs on cognition, and RTCs, which were negative (May 2008 news story).

Analysis of ADAPT data continued for some years. Results from a further two years of follow-up until 2007 indicated a dichotomy whereby people who were symptomatic at baseline worsened on naproxen compared to placebo, but people who were cognitively normal at baseline were less likely to decline and had a healthier CSF biomarker signature at that point (Jul 2009 conference news; Breitner et al., 2011). A subsequent follow-up evaluation of some 1,500 participants in 2010 and 2011 reported that one to three years of preventive treatment with naproxen in people with a family history of AD conferred no protection against cognitive decline (Alzheimer's Disease Anti-inflammatory Prevention Trial Research Group 2013; ADAPT-FS Research Group 2014 ).

Additional analyses attempted to distinguish between slow and fast decliners and to identify a time window during the prodromal phase of AD in which naproxen might be beneficial. A possible benefit of naproxen only during the presymptomatic phase was seen as being consistent with both ADAPT and epidemiological data (e.g., Leoutsakos et al., 2011).

A third trial, started in 2012 in Montreal, enrolled 195 cognitively normal people 55 and older who have a parent or multiple siblings with AD; biomarker evidence of preclinical AD was not required for inclusion. Participants were randomized to 220 mg twice daily of naproxen or placebo for two years. Primary outcomes were the global score on the Repeatable Battery for Assessment of Neuropsychological Status, as well as a composite Alzheimer progression score (APS) derived from multiple cognitive and biomarker measures of preclinical Alzheimer's disease. Safety parameters, naproxen kinetics, CSF biomarkers of AD pathology, and plasma/CSF markers of inflammation were secondary outcome measures. Efficacy data indicated no benefit of naproxen over placebo on the trajectory of the APS, or on any of the underlying measures. The naproxen group had more adverse events, including hypertension, gastrointestinal, and vascular or cardiac problems. While the trial was underpowered to have detected a potential small naproxen effect, the authors conclude such an effect would be difficult to demonstrate in a randomized prevention trial, and that the drug’s proven adverse effects would raise ethical concerns (see Meyer et al., 2019, Apr 2019 news).