AVOCADO/SOYBEAN UNSAPONIFIABLES (ASU) IN OSTEOARTHRITIS (OA) TREATMENT: A META-ANALYSIS WITH FOCUS ON EFFICACY

Background: Complementary and alternative medicine (CAM) is widely used as self-medication by patients with rheumatologic diseases. One of the therapies, which has shown some promise in the treatment of osteoarthritis (OA), is based on avocado/soybean unsaponifiables (ASU).Objectives: To evaluate the efficacy of the existing ASU preparations in the treatment of OA by carrying out a meta-analysis on published randomised controlled trials (RCTs) involving ASU treatment of knee and/or hip OA.Methods: Systematic searches were performed of the bibliographic databases Medline, EMBASE, Chemical Abstracts, Cinahl, Scopus, Web of Science, as well as The Cochrane Controlled Trials Register. Also, the reference lists from the retrieved trials were searched. Selected criteria: RCTs evaluating the effectiveness and safety of ASU in OA; Both single blinded and double blinded studies were eligible. There were no restrictions on language. The OMERACT III outcome variables[1] was considered for analysis. Two investigators independently extracted data for the OMERACT III outcome variables published in more than one study (pain and the Lequesne index); the Jadad quality score was applied to assess the quality of the included published trials. This score included randomisation, double blinding, and explicit description of dropouts or an intention-to-treat methodology applied. We calculated the effect size (ES) as the standardised mean difference, with 95% confidence intervals (CI), using a random effects model based on the published mean (SD) values at the end of trial as outcome. Data were analysed using the Review Manager (RevMan) computer program, version 4.2.Results: Four RCTs and 662 patients were included in the review; hip OA: 275 and knee OA: 387, respectively. The studies had two outcomes in common: pain (VAS: 0-100 mm) and the Lequesne index. All 4 studies used 300 mg ASU per day for a weighted (arithmetric) mean duration of 6 months (range: 3 months to 1 year). The weighted mean age was 64 years at baseline in an average population with 64% females. The weighted Jadad quality score was 4.5 (range: 3 to 5), indicating a low risk of bias. Overall, there was a significant relative risk (RR) of being a responder to ASU compared to placebo treatment of 1.38 (95% CI: 1.03 to 1.85), with a corresponding (Peto Odds Ratio, placebo-risk) adjusted Number Needed to Treat (NNT) of: 7 (5 to 14). Overall, ASU was superior to placebo in pain reduction (P=0.01): ES=-0.42 (-0.75 to –0.09); and in decreasing physical disability measured by the Lequesne index (P<0.0001): ES=-0.48 (-0.70 to –0.26). The number and severity of adverse events in the ASU-treated patients and in the placebo group were similarly small.Conclusion: According to our analyses, there is a strong ('platin') evidence that patients suffering from both knee and hip OA, taking 300 mg ASU on a daily basis, experience an increased "risk" of improvement compared to patients taking placebo, with a NNT of 7. In general, the size of improvement indicates a small-to-moderate clinical effect of ASU considering the pooled efficacy in both hip and knee OA patients. There is also evidence that knee OA patients will benefit more from ASU-treatment (moderate-to-large clinical effect) than hip OA patients. There might therefore be therapeutic properties to be explored in the avocado/soybean-unsaponifiables. It is, though, worth noticing that all published RCTs are based on the ASU preparation from Pharmascience, France, and further studies, using other types of ASU, are needed to support the present data. This evidence is based on studies of rather short duration and long-term effects of ASU therapy remain to be clarified.References: 1. Bellamy et al. J Rheumatol. 1997;24:799-802.Citation: Ann Rheum Dis, volume 65, supplement II, year 2006, page 227Session: Osteoarthritis – clinical aspects and treatment