Article Text

Cochrane systematic review
Clonidine, SSRIs, SNRIs and gabapentin reduce hot flushes in women with a history of breast cancer; relaxation therapy may have a mild effect in the short term
  1. Deborah Fenlon
  1. Macmillan Survivorship Research Group, University of Southampton, Faculty of Health Sciences Southampton, UK
  1. Correspondence to Deborah Fenlon
    Macmillan Survivorship Research Group, Faculty of Health Sciences, University of Southampton, Highfield Campus, Southampton SO17 1BJ, UK; dfenlon{at}

Statistics from

Commentary on:

As treatments and early detection of breast cancer improve, the number of women affected by the consequences of treatment increases. Menopausal issues, such as hot flushes, may affect the majority of these women. Hot flushes may also be more troublesome in the context of breast cancer. At the same time women who have had breast cancer are advised not to take hormone therapies, which are the most effective treatment of this problem, and many are reluctant to take medications following a cancer diagnosis. Nursing staff are frequently called on to advice women in this position as to how to manage their hot flushes and therefore need to keep updated as to the latest evidence around effective and safe interventions. However, there is a large amount of literature on a wide range of potential interventions, and this carefully considered Cochrane review of the available evidence for non-hormonal treatments is therefore welcome in order to inform practice.

This article is a review of the literature on non-hormonal therapies conducted according to strict Cochrane guidelines. Only randomised controlled trials (RCTs) have been included. The rigour of this method means that the findings are reliable; however, it may mean that some studies with useful data on promising interventions have been excluded if they do not meet the strict entry criteria of the review. Studies which met the inclusion criteria included studies on pharmacological and non-pharmacological interventions, including selective serotonin and serotonin–norepinephrine reuptake inhibitors (SSRIs and SNRIs antidepressants), clonidine, gabapentin, relaxation therapy, homeopathy, vitamin E, magnetic devices and acupuncture.

The authors found that the pharmacological measures of gabapentin, clonidine and the antidepressants all had some effect in reducing the incidence and severity of hot flushes. The only non-pharmacological measure found to reduce hot flushes was relaxation, although the effect of this was not maintained after 3 months.

No critique is presented in this article of the way that the studies were analysed and the rigour of the analysis. In some of the findings, the poorer quality studies were apparently given equal weight to the more rigorous studies. For example, the comparison of the two studies on relaxation stated that one had no effect and the other had a beneficial effect on hot flushes. Although it states that the one with no effect was smaller, it does not make it clear that this was purely a pilot study and was not powered to detect a statistically significant difference. Similarly the study on a magnetic device only included 11 women. It would be premature to reject an intervention on the basis of this study alone as any effect is likely to be missed with such a small sample.

It is surprising that the article on vitamin E is reported as having no benefit for hot flushes as the authors of this article report a statistically significant benefit. Barton and colleagues1 question whether it is a clinically significant reduction in flushes as it was only a reduction of one flush per day. However, there was a statistically significant reduction in incidence of flushes and so to state that there was no effect and that vitamin E should not be considered for practice or even further research is inaccurate.

The authors state that one of the limitations is the poor reporting of adverse effects, which was either absent or under-reported. The conclusions and implications for practice of this article must therefore be tempered by other research findings that give a more complete picture of adverse effects.

Inevitably, the disadvantage of systematic reviews is the time that it takes to gather and analyse the data, during which time further evidence is becoming available. The most recent publication contained within this review is 2007. Since this time there have been a number of useful RCTs providing further evidence, including for the use of acupuncture. The review may be in danger of discounting potentially useful interventions if it overplays the results of small negative studies.

The article concludes that there is only evidence to support the use of the three pharmacological interventions and relaxation to reduce the frequency and intensity of hot flushes in women with breast cancer. They suggest that because the safety profile of the antidepressants is good that the use of these drugs is promising. Overall, the absolute benefit of all the interventions is small and that each woman's preference should guide recommendations. Bearing this in mind, the recommendations should include suggestions on research methodology that is about patient preference as well as absolute benefit.

Although this article conducts a useful review of the available evidence, it is important to remember that absence of evidence does not mean absence of effect and that interventions where the research evidence is emerging should not be discarded too soon. Furthermore, adverse effects may be under-reported, and so patient preference should drive choice of appropriate intervention in clinical practice.


View Abstract


  • Competing interests DF has received an honorarium from Roche for a presentation at the Champions for Change Conference.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.