Abstract
In the post-Women's Health Initiative era, many women are reluctant to use estrogen, while others are unable to take estrogen and approach their pharmacist for alternative treatments for hot flashes. A number of therapies have been proposed in the literature, often leaving us to wonder whether we should be recommending these options to our patients.
Some help in answering this question comes from a recent systematic review and meta-analysis of nonhormonal therapies for menopausal hot flashes reported in a recent issue of the Journal of the American Medical Association. 1 Nelson et al. considered all randomized controlled trials published in the English language and, based on inclusion and exclusion criteria, reviewed 43 trials, of which 24 were included in the meta-analysis (6 for SSRIs or SNRIs, 4 for clonidine, 2 for gabapentin, 6 for red clover isoflavone extracts and 6 for soy isoflavone extracts).
Highlights of the systematic review
Unique information provided by the systematic review relates to dose response and adverse effects. The trials of paroxetine revealed no significant difference in efficacy with the higher doses (20 mg/day and 25 mg CR/day), however, there was an increase in adverse effects. A dose-dependent response was observed in the trials with venlafaxine, where effects were greater with the 75 mg and 150 mg doses. A similar difference in response was found in the gabapentin trials, where 300 mg/day showed no difference, while 900 mg/day was associated with an improvement in symptoms. The trials with clonidine indicated an increase in adverse effects over placebo, while the adverse effects from isoflavone extracts did not differ from placebo.
What can we learn from the meta-analysis?
The meta-analysis allows for the combination of the results of the various trials to increase the strength of the evidence on efficacy of the various treatments.
The 6 SSRI or SNRI trials had a significant combined weighted mean difference in the number of daily hot flashes of −1.13 (CI: −1.70 to −0.57) compared to placebo. The sensitivity analysis of these trials demonstrated that in women with breast cancer and/or taking selective estrogen receptor modulators (SERMs), the reduction in symptoms remained significant whereas it was not significant in those trials where women were not taking SERMs (a difference of −0.17 [CI: −1.41 to 1.07], which is not significant, as the CI crosses 1).
The 4 trials of clonidine that were included had a weighted mean difference of −0.95 (CI: −1.44 to −0.47) after 4 weeks and this difference remained significant in trials that were 8 weeks in length. In the sensitivity analysis, the 2 trials that enrolled women with breast cancer and/or taking SERMs showed a significant reduction in hot flashes, whereas those trials that did not include this group did not demonstrate a significant difference.
The 2 trials of gabapentin used doses of 900 mg/day and had a combined weighted mean difference of −2.05 (CI: −2.80 to −1.30). Six trials with red clover isoflavones were analyzed and achieved a combined weighted mean difference of −0.44 (CI: −1.47 to 0.58), which was not significant. For the soy isoflavone trials, this weighted mean difference was not significant at 4–6 weeks, but was significant for the trials that continued to 12–16 weeks and 6 months (−0.97 and −1.22).
What do we do with this evidence?
The results of this meta-analysis show that there is evidence from fair-good quality trials of SSRIs or SNRIs and gabapentin and from poor-fair trials for clonidine, supporting their use. The evidence for red clover isoflavones is poor, while the results with soy isoflavones are mixed. The authors attempt to put these results in perspective by comparing the results of the SSRIs and SNRIs to results from similar work done with estrogen. They indicate that the combined estimate for SSRIs and SNRIs and clonidine is a reduction in 1 hot flash per day compared to a reduction in 2.5 to 3 hot flashes per day with estrogen. 2 The estimated reduction for gabapentin is 2 hot flashes per day, while no reduction in hot flashes was demonstrated by the evidence for the isoflavones.
Despite being a well-conducted meta-analysis, the study is still flawed by the methodological issues plaguing all trials studying hot flashes (placebo response, large dropouts) and the limitation to English-language studies. However, it does provide us with an interpretation of the evidence that does exist and helps us to provide an evidence-based answer to the question — “What can I do about these hot flashes?”
Patients (n = 1161) with mild to moderate hypertension were randomized to losartan, candesartan, or losartan plus hydrochlorothiazide for 12 weeks of treatment.9 Serum uric acid levels decreased with losartan alone, remained the same with the combination, and increased with candesartan monotherapy. A small open-label crossover study enrolled 10 men with a history of recurrent acute or chronic tophaceous gout to receive three weeks of micronized fenofibrate 200 mg once a day in addition to their usual dose of allopurinol therapy.10 Serum urate levels were reduced by 19% with the addition of fenofibrate and no significant adverse effects were observed.
Gout has been linked to many medications (see Table 3). By identifying patients who are at higher risk for hyperuricemia, pharmacists can monitor patients when suspected agents are added to their regimen.