Acupuncture as adjunctive therapy in assisted reproduction: remaining uncertainties

Two randomized trials provide suggestive evidence that acupuncture might be an effective adjunctive therapy in assisted reproductive technologies, leading to improved pregnancy rates. However, additional research addressing remaining scientific and methodologic issues is needed before routine use of acupuncture can be recommended. Two randomized trials provide suggestive evidence that acupuncture might be an effective adjunctive therapy in assisted reproductive technologies, leading to improved pregnancy rates. However, additional research addressing remaining scientific and methodologic issues is needed before routine use of acupuncture can be recommended. Two provocative studies in this issue of Fertility and Sterility suggest that the use of acupuncture increases clinical pregnancy rates when used as adjunctive therapy in couples undergoing IVF. Westergaard et al. (1Westergaard L.G. Mao Q. Krogslund M. Sandrini S. Lenz S. Grinsted J. Acupuncture on the day of embryo transfer significantly improves the reproductive outcome in infertile women a prospective, randomized trial.Fertil Steril. 2006; 85: 1341-1346Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar) randomized couples to three groups (no therapy, acupuncture on the day of ET, and acupuncture on the day of therapy and 2 days later) and found significantly increased clinical and ongoing pregnancy rates in women receiving acupuncture on the day of ET but not in women receiving it on both the transfer day and 2 days later. Dieterle et al. (2Dieterle S. Ying G. Hatzmann W. Neuer A. Effect of acupuncture on the outcome of in vitro fertilization and intracytoplasmic sperm injection a randomized, prospective, controlled clinical study.Fertil Steril. 2006; 85: 1347-1351Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar) randomized women to either acupuncture performed on acupuncture points thought to be associated with reproduction on transfer day and 3 days later (the “active” group), or acupuncture performed on points thought to be unrelated to reproduction on the same days; the investigators also found significantly higher clinical and ongoing pregnancy rates in the group receiving the “active” acupuncture.Both studies were prospective, randomized, controlled trials, with a priori sample size calculations and appropriate standardization of other factors that might influence the outcomes of interest. Given this rigor and the need to identify effective adjunctive methods for ART, should infertility clinics begin routinely offering acupuncture to couples on the basis of these results? On the basis of some of the questions raised by these studies, I agree with the authors of both articles that additional research is needed.Although both articles offer possible explanations for the mechanism of action of acupuncture in enhancing the results of IVF, many basic questions of clinical significance remain. Both studies used similar acupuncture points, but there was not complete concordance in the points used; what are the implications of this, and how should a reader, especially those of us unfamiliar with the conceptual framework behind acupuncture, interpret the similar results of the two studies given these differences? What is the significance of the use of topical application of additional traditional Chinese medicine in one study (2Dieterle S. Ying G. Hatzmann W. Neuer A. Effect of acupuncture on the outcome of in vitro fertilization and intracytoplasmic sperm injection a randomized, prospective, controlled clinical study.Fertil Steril. 2006; 85: 1347-1351Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar) but not the other (1Westergaard L.G. Mao Q. Krogslund M. Sandrini S. Lenz S. Grinsted J. Acupuncture on the day of embryo transfer significantly improves the reproductive outcome in infertile women a prospective, randomized trial.Fertil Steril. 2006; 85: 1341-1346Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar)?The optimal control for studies of acupuncture is not obvious, and the search has generated considerable discussion and ongoing research (3Dincer F. Linde K. Sham interventions in randomized clinical trials of acupuncture—a review.Complement Ther Med. 2003; 11: 235-242Abstract Full Text Full Text PDF PubMed Scopus (150) Google Scholar, 4Miller F.G. Kaptchuk T.J. Sham procedures and the ethics of clinical trials.J R Soc Med. 2004; 97: 576-578Crossref PubMed Scopus (49) Google Scholar, 5Bausell R.B. Lao L. Bergman S. Lee W.L. Berman B.M. Is acupuncture analgesia an expectancy effect? Preliminary evidence based on participants’ perceived assignments in two placebo-controlled trials.Eval Health Prof. 2005; 28: 9-26Crossref PubMed Scopus (147) Google Scholar, 6White P.J. Methodological concerns when designing trials for the efficacy of acupuncture for the treatment of pain.Adv Exp Med Biol. 2004; 546: 217-227Crossref PubMed Scopus (12) Google Scholar, 7Goddard G. Shen Y. Steele B. Springer N. A controlled trial of placebo versus real acupuncture.J Pain. 2005; 6: 237-242Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). Especially given the lack of strong evidence on the basic mechanisms by which acupuncture might enhance reproduction, the appropriate choice of control is critical. Although both studies made reasonable choices for controls, both also illustrate some of the difficulties with the available choices.The use of “routine care” without a control intervention, as in the study by Westergaard et al. (1Westergaard L.G. Mao Q. Krogslund M. Sandrini S. Lenz S. Grinsted J. Acupuncture on the day of embryo transfer significantly improves the reproductive outcome in infertile women a prospective, randomized trial.Fertil Steril. 2006; 85: 1341-1346Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar), raises the possibility of a placebo effect. The authors make two arguments against a placebo effect. First, they state that the lack of observed difference between women receiving acupuncture on the day of ET only vs. ET plus 2 days later is evidence against a placebo effect, implying that a placebo given on 2 days should have a greater impact then a placebo given on only 1 day. Second, they argue that the fact that the effectiveness of acupuncture seems to be limited to women aged <38 years argues against a placebo effect. In addition to the statistical issues involved with this study, discussed below, it is not at all clear how these two points argue against a placebo effect, especially for an active treatment that the authors themselves speculate might work through neuroendocrine changes. It is certainly possible that placebo effects cause neuroendocrine changes that result in a more favorable intrauterine environment, and it is certainly possible that the effect of these changes is dependent on their timing relevant to the events involved in ET and implantation. An age-dependent placebo effect, mediated through neuroendocrine changes, is also not biologically implausible; one can imagine a variety of age-dependent mechanisms, such as egg quality, that might affect outcomes in ways that any neuroendocrine changes caused by a placebo effect would not be able to overcome.The study by Liederle et al. (2Dieterle S. Ying G. Hatzmann W. Neuer A. Effect of acupuncture on the outcome of in vitro fertilization and intracytoplasmic sperm injection a randomized, prospective, controlled clinical study.Fertil Steril. 2006; 85: 1347-1351Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar) provides an example of the complexities introduced by another option for a control—performing active acupuncture on points thought to be unrelated to the outcome of interest. In this case, one can not rule out an effect of the active treatment. Given the lack of knowledge on the potential mechanism of action of acupuncture, a detrimental effect of this treatment on reproductive outcomes can not be ruled out, as the authors acknowledge.The call from both authors for additional research using alternative controls, such as “placebo needles” (8White P. Lewith G. Hopwood V. Prescott P. The placebo needle, is it a valid and convincing placebo for use in acupuncture trials? A randomised, single-blind, cross-over pilot trial.Pain. 2003; 106: 401-409Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar), is certainly justified.Both studies were powered on the basis of a predefined clinically relevant difference in clinical pregnancy rates. Although this outcome has the attractive qualities of requiring fewer subjects and shorter follow-up observation than live birth rate, using this outcome for primary sample size estimation means that studies will be underpowered to detect differences in live birth rate. Use of “easier” surrogate outcomes is common in studies of infertility treatments but ultimately limits the value of the data generated by such studies (9Legro R.S. Myers E.R. Surrogate outcomes in studies of polycystic ovary syndrome.Hum Reprod. 2004; 19: 1697-1704Crossref PubMed Scopus (37) Google Scholar).Another issue raised by the study of Westergaard et al. is the appropriate use and analysis of a three-armed trial. The authors state that the Fisher’s exact test was used to compare differences between groups. However, they do not state whether there was an underlying hypothesis of ordered response—whether they hypothesized that treatment on both the day of transfer and 2 days later would result in improvement over treatment on the day of transfer only, which in turn would be better than results from the no-treatment arm. If this was the case, then either the sample size should have been increased, or the P value considered significant decreased, to account for the multiple planned comparisons (10Curran D. Sylvester R.J. Hoctin Boes G. Sample size estimation in phase III cancer clinical trials.Eur J Surg Oncol. 1999; 25: 244-250Abstract Full Text PDF PubMed Scopus (16) Google Scholar). Alternatively, a “decision rule” analysis could have been stipulated, again driven by the underlying hypotheses (11Myers E.R. Silva S. Hafley G. Kunselman A. Nestler J.E. Legro R.S. et al.Sample size estimation for the Pregnancy in Polycystic Ovary Syndrome (PPCOS) Trial.Contemp Clin Trials. 2005; 26: 271-280Abstract Full Text Full Text PDF PubMed Scopus (18) Google Scholar). In any event, analyzing a three-armed trial using sample sizes based on a two-armed trial is not appropriate and increases the chances of a Type I error (falsely concluding a real difference in treatments).The omission of the explicit hypotheses behind the three-armed trial also creates difficulties in interpreting the results of no statistical difference between the group receiving treatment on two days and the control group. Table 4 in the Westergaard article suggests that clinical pregnancy rates were similarly higher than in controls in the two groups receiving acupuncture, but that the group receiving acupuncture on day 2 had higher rates of earlier pregnancy loss; the study was not powered to determine whether the observed differences were statistically significant. Data showing that acupuncture on the day of transfer increased pregnancy rates but that subsequent acupuncture on the day of implantation increased loss rates would clearly be of both scientific and clinical interest.If further studies prove that acupuncture is an effective adjunct therapy, at least in some patients, then additional data on how acupuncture might be incorporated into clinical practice will be needed. Who will provide the acupuncture—acupuncture specialists, or clinic staff trained in acupuncture methods? What are the costs associated with implementing an acupuncture component to the typical IVF practice? Is the incremental cost-effectiveness of acupuncture reasonable compared with other alternatives?Good clinical research, especially trials involving randomization, is always difficult and always involves a series of trade-offs between methodologic purity and practicality. The authors of these studies are to be commended for using the methods of the randomized trial to address an intriguing question that involves two areas of medicine in which randomized studies are all too infrequent (12Birch S. Hesselink J.K. Jonkman F.A. Hekker T.A. Bos A. Clinical research on acupuncture. Part 1. What have reviews of the efficacy and safety of acupuncture told us so far?.J Altern Complement Med. 2004; 10: 468-480Crossref PubMed Scopus (113) Google Scholar). As both sets of investigators acknowledge, further research is needed to address remaining questions. At a minimum, I would suggest that such research [1] use standardized acupuncture methods to enhance comparability between studies, [2] use alternative controls, such as “placebo” needles, [3] use live birth as the primary outcome, [4] specify a priori the hypotheses and justification for the number of comparison groups in the study, and [5] base sample size on the live birth outcome and the number of study arms. If such studies continue to demonstrate effectiveness, then demonstrating cost-effectiveness and feasibility in clinical practice should be the final step before recommending routine use of acupuncture in IVF. Two provocative studies in this issue of Fertility and Sterility suggest that the use of acupuncture increases clinical pregnancy rates when used as adjunctive therapy in couples undergoing IVF. Westergaard et al. (1Westergaard L.G. Mao Q. Krogslund M. Sandrini S. Lenz S. Grinsted J. Acupuncture on the day of embryo transfer significantly improves the reproductive outcome in infertile women a prospective, randomized trial.Fertil Steril. 2006; 85: 1341-1346Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar) randomized couples to three groups (no therapy, acupuncture on the day of ET, and acupuncture on the day of therapy and 2 days later) and found significantly increased clinical and ongoing pregnancy rates in women receiving acupuncture on the day of ET but not in women receiving it on both the transfer day and 2 days later. Dieterle et al. (2Dieterle S. Ying G. Hatzmann W. Neuer A. Effect of acupuncture on the outcome of in vitro fertilization and intracytoplasmic sperm injection a randomized, prospective, controlled clinical study.Fertil Steril. 2006; 85: 1347-1351Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar) randomized women to either acupuncture performed on acupuncture points thought to be associated with reproduction on transfer day and 3 days later (the “active” group), or acupuncture performed on points thought to be unrelated to reproduction on the same days; the investigators also found significantly higher clinical and ongoing pregnancy rates in the group receiving the “active” acupuncture. Both studies were prospective, randomized, controlled trials, with a priori sample size calculations and appropriate standardization of other factors that might influence the outcomes of interest. Given this rigor and the need to identify effective adjunctive methods for ART, should infertility clinics begin routinely offering acupuncture to couples on the basis of these results? On the basis of some of the questions raised by these studies, I agree with the authors of both articles that additional research is needed. Although both articles offer possible explanations for the mechanism of action of acupuncture in enhancing the results of IVF, many basic questions of clinical significance remain. Both studies used similar acupuncture points, but there was not complete concordance in the points used; what are the implications of this, and how should a reader, especially those of us unfamiliar with the conceptual framework behind acupuncture, interpret the similar results of the two studies given these differences? What is the significance of the use of topical application of additional traditional Chinese medicine in one study (2Dieterle S. Ying G. Hatzmann W. Neuer A. Effect of acupuncture on the outcome of in vitro fertilization and intracytoplasmic sperm injection a randomized, prospective, controlled clinical study.Fertil Steril. 2006; 85: 1347-1351Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar) but not the other (1Westergaard L.G. Mao Q. Krogslund M. Sandrini S. Lenz S. Grinsted J. Acupuncture on the day of embryo transfer significantly improves the reproductive outcome in infertile women a prospective, randomized trial.Fertil Steril. 2006; 85: 1341-1346Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar)? The optimal control for studies of acupuncture is not obvious, and the search has generated considerable discussion and ongoing research (3Dincer F. Linde K. Sham interventions in randomized clinical trials of acupuncture—a review.Complement Ther Med. 2003; 11: 235-242Abstract Full Text Full Text PDF PubMed Scopus (150) Google Scholar, 4Miller F.G. Kaptchuk T.J. Sham procedures and the ethics of clinical trials.J R Soc Med. 2004; 97: 576-578Crossref PubMed Scopus (49) Google Scholar, 5Bausell R.B. Lao L. Bergman S. Lee W.L. Berman B.M. Is acupuncture analgesia an expectancy effect? Preliminary evidence based on participants’ perceived assignments in two placebo-controlled trials.Eval Health Prof. 2005; 28: 9-26Crossref PubMed Scopus (147) Google Scholar, 6White P.J. Methodological concerns when designing trials for the efficacy of acupuncture for the treatment of pain.Adv Exp Med Biol. 2004; 546: 217-227Crossref PubMed Scopus (12) Google Scholar, 7Goddard G. Shen Y. Steele B. Springer N. A controlled trial of placebo versus real acupuncture.J Pain. 2005; 6: 237-242Abstract Full Text Full Text PDF PubMed Scopus (36) Google Scholar). Especially given the lack of strong evidence on the basic mechanisms by which acupuncture might enhance reproduction, the appropriate choice of control is critical. Although both studies made reasonable choices for controls, both also illustrate some of the difficulties with the available choices. The use of “routine care” without a control intervention, as in the study by Westergaard et al. (1Westergaard L.G. Mao Q. Krogslund M. Sandrini S. Lenz S. Grinsted J. Acupuncture on the day of embryo transfer significantly improves the reproductive outcome in infertile women a prospective, randomized trial.Fertil Steril. 2006; 85: 1341-1346Abstract Full Text Full Text PDF PubMed Scopus (130) Google Scholar), raises the possibility of a placebo effect. The authors make two arguments against a placebo effect. First, they state that the lack of observed difference between women receiving acupuncture on the day of ET only vs. ET plus 2 days later is evidence against a placebo effect, implying that a placebo given on 2 days should have a greater impact then a placebo given on only 1 day. Second, they argue that the fact that the effectiveness of acupuncture seems to be limited to women aged <38 years argues against a placebo effect. In addition to the statistical issues involved with this study, discussed below, it is not at all clear how these two points argue against a placebo effect, especially for an active treatment that the authors themselves speculate might work through neuroendocrine changes. It is certainly possible that placebo effects cause neuroendocrine changes that result in a more favorable intrauterine environment, and it is certainly possible that the effect of these changes is dependent on their timing relevant to the events involved in ET and implantation. An age-dependent placebo effect, mediated through neuroendocrine changes, is also not biologically implausible; one can imagine a variety of age-dependent mechanisms, such as egg quality, that might affect outcomes in ways that any neuroendocrine changes caused by a placebo effect would not be able to overcome. The study by Liederle et al. (2Dieterle S. Ying G. Hatzmann W. Neuer A. Effect of acupuncture on the outcome of in vitro fertilization and intracytoplasmic sperm injection a randomized, prospective, controlled clinical study.Fertil Steril. 2006; 85: 1347-1351Abstract Full Text Full Text PDF PubMed Scopus (124) Google Scholar) provides an example of the complexities introduced by another option for a control—performing active acupuncture on points thought to be unrelated to the outcome of interest. In this case, one can not rule out an effect of the active treatment. Given the lack of knowledge on the potential mechanism of action of acupuncture, a detrimental effect of this treatment on reproductive outcomes can not be ruled out, as the authors acknowledge. The call from both authors for additional research using alternative controls, such as “placebo needles” (8White P. Lewith G. Hopwood V. Prescott P. The placebo needle, is it a valid and convincing placebo for use in acupuncture trials? A randomised, single-blind, cross-over pilot trial.Pain. 2003; 106: 401-409Abstract Full Text Full Text PDF PubMed Scopus (190) Google Scholar), is certainly justified. Both studies were powered on the basis of a predefined clinically relevant difference in clinical pregnancy rates. Although this outcome has the attractive qualities of requiring fewer subjects and shorter follow-up observation than live birth rate, using this outcome for primary sample size estimation means that studies will be underpowered to detect differences in live birth rate. Use of “easier” surrogate outcomes is common in studies of infertility treatments but ultimately limits the value of the data generated by such studies (9Legro R.S. Myers E.R. Surrogate outcomes in studies of polycystic ovary syndrome.Hum Reprod. 2004; 19: 1697-1704Crossref PubMed Scopus (37) Google Scholar). Another issue raised by the study of Westergaard et al. is the appropriate use and analysis of a three-armed trial. The authors state that the Fisher’s exact test was used to compare differences between groups. However, they do not state whether there was an underlying hypothesis of ordered response—whether they hypothesized that treatment on both the day of transfer and 2 days later would result in improvement over treatment on the day of transfer only, which in turn would be better than results from the no-treatment arm. If this was the case, then either the sample size should have been increased, or the P value considered significant decreased, to account for the multiple planned comparisons (10Curran D. Sylvester R.J. Hoctin Boes G. Sample size estimation in phase III cancer clinical trials.Eur J Surg Oncol. 1999; 25: 244-250Abstract Full Text PDF PubMed Scopus (16) Google Scholar). Alternatively, a “decision rule” analysis could have been stipulated, again driven by the underlying hypotheses (11Myers E.R. Silva S. Hafley G. Kunselman A. Nestler J.E. Legro R.S. et al.Sample size estimation for the Pregnancy in Polycystic Ovary Syndrome (PPCOS) Trial.Contemp Clin Trials. 2005; 26: 271-280Abstract Full Text Full Text PDF PubMed Scopus (18) Google Scholar). In any event, analyzing a three-armed trial using sample sizes based on a two-armed trial is not appropriate and increases the chances of a Type I error (falsely concluding a real difference in treatments). The omission of the explicit hypotheses behind the three-armed trial also creates difficulties in interpreting the results of no statistical difference between the group receiving treatment on two days and the control group. Table 4 in the Westergaard article suggests that clinical pregnancy rates were similarly higher than in controls in the two groups receiving acupuncture, but that the group receiving acupuncture on day 2 had higher rates of earlier pregnancy loss; the study was not powered to determine whether the observed differences were statistically significant. Data showing that acupuncture on the day of transfer increased pregnancy rates but that subsequent acupuncture on the day of implantation increased loss rates would clearly be of both scientific and clinical interest. If further studies prove that acupuncture is an effective adjunct therapy, at least in some patients, then additional data on how acupuncture might be incorporated into clinical practice will be needed. Who will provide the acupuncture—acupuncture specialists, or clinic staff trained in acupuncture methods? What are the costs associated with implementing an acupuncture component to the typical IVF practice? Is the incremental cost-effectiveness of acupuncture reasonable compared with other alternatives? Good clinical research, especially trials involving randomization, is always difficult and always involves a series of trade-offs between methodologic purity and practicality. The authors of these studies are to be commended for using the methods of the randomized trial to address an intriguing question that involves two areas of medicine in which randomized studies are all too infrequent (12Birch S. Hesselink J.K. Jonkman F.A. Hekker T.A. Bos A. Clinical research on acupuncture. Part 1. What have reviews of the efficacy and safety of acupuncture told us so far?.J Altern Complement Med. 2004; 10: 468-480Crossref PubMed Scopus (113) Google Scholar). As both sets of investigators acknowledge, further research is needed to address remaining questions. At a minimum, I would suggest that such research [1] use standardized acupuncture methods to enhance comparability between studies, [2] use alternative controls, such as “placebo” needles, [3] use live birth as the primary outcome, [4] specify a priori the hypotheses and justification for the number of comparison groups in the study, and [5] base sample size on the live birth outcome and the number of study arms. If such studies continue to demonstrate effectiveness, then demonstrating cost-effectiveness and feasibility in clinical practice should be the final step before recommending routine use of acupuncture in IVF.