REVIEW
The efficacy and tolerability of SSRI/SNRIs in the treatment of vasomotor symptoms in menopausal women: A systematic review Amy P. Handley, MSN, FNP-C (Associate Professor and Associate Dean)1 & Mary Williams, PhD, RN (Nurse Practitioner)2 1 2
College of Nursing, Brigham Young University, Provo, Utah Graduate Studies and Faculty, College of Nursing, Brigham Young University, Provo, Utah
Keywords Hot flashes; vasomotor symptoms; menopause; selective serotonin reuptake inhibitors (SSRI); hormone replacement therapy. Correspondence Amy Handley, Lakeview Family Medicine, 811 N 900 W, Orem, UT 84057. Tel: 801-812-3200; E-mail: [email protected] Received: 19 September 2012; accepted: 25 February 2013 doi: 10.1002/2327-6924.12137 Disclosures: The authors report no conflicting interests.
Abstract Purpose: To systematically review the evidence related to the efficacy and tolerability of selective serotonin reuptake inhibitors (SSRIs) and serotonin/norepinephrine reuptake inhibitors (SNRIs) used for the treatment of vasomotor symptoms in perimenopausal and postmenopausal women. Data sources: Medline, CINAHL, and the Cochrane Library databases were searched to identify randomized controlled trials (RCTs). Eighteen trials met the criteria for review. Conclusions: Results from these trials indicate that paroxetine, citalopram, escitalopram, venlafaxine, and desvenlafaxine are effective in reducing the frequency and severity of hot flashes. Fluoxetine and sertraline appear to be less effective and should be considered second-line options for treatment. Implications for practice: The SSRIs and SNRIs can reduce hot flashes by 65% and begin working within the first week. Patient response is variable and if one drug does not improve hot flashes, another can be tried after a 1- to 2-week drug trial. Paroxetine, citalopram, and escitalopram appear to have the fewest adverse effects. Considering cost, paroxetine and citalopram are the most cost-efficient.
Vasomotor symptoms are the most commonly reported symptoms during the menopausal transition (Speroff, Gass, Constantine, & Olivier, 2008). Hot flashes, one of the most common vasomotor symptoms, are most prevalent in the late menopausal transition and occur in about 65% of all women (Grady, 2006). These symptoms usually resolve within 4–5 years; however, 10%–15% of symptomatic women continue to have hot flashes for several years longer (Grady, 2006). Vasomotor symptoms disrupt sleep, reduce quality of life, increase irritability, and are the menopausal symptoms that most cause women to seek treatment (Archer, Dupont, Constantine, Pickar, & Olivier, 2009). Hormone therapy (HT) is the treatment of choice for menopause-related vasomotor symptoms, reducing frequency by approximately 75% (MacLennan, Broadbent, Lester, & Moore, 2004). The current recommendation is to prescribe the lowest effective dose for the shortest time period, although extended therapy 54
is acceptable under individualized circumstances (North American Menopause Society, 2012). However, current evidence indicates that HT is not appropriate for all women. In general, HT is not recommended in women with a history of endometrial cancer or breast cancer because of possible increased risk for recurrence (North American Menopause Society, 2012). Obese women, those with factor V Leiden mutation, and women with a history of venous thromboembolism are at increased risk for venous thromboembolism during HT use (North American Menopause Society, 2012). The known risks of HT use in healthy women aged 50–59 are low; however, long-term HT and initiation of therapy in women aged 60 and older are associated with greater risks (North American Menopause Society, 2012). This information has led practitioners and women transitioning to postmenopause to seek viable nonhormonal alternatives for the relief of vasomotor symptoms (Carroll & Kelley, 2009).
Journal of the American Association of Nurse Practitioners 27 (2015) 54–61 C 2014 American Association of Nurse Practitioners
Vasomotor symptoms in menopausal women
A. P. Handley & M. Williams
Because the mechanism of hot flashes is not fully understood, research has been conducted on a variety of nonhormonal treatments to alleviate vasomotor symptoms (Hall, Frey, & Soares, 2011). During the transition to postmenopause, changes in the neuroendocrine system are thought to disrupt thermoregulation in the hypothalamus, resulting in body core temperature changes. The body responds through perspiration and vasodilation, producing a hot flash (Carroll & Kelley, 2009). Also during this transition, estrogen and progesterone levels decline. Although the role of estrogens in the production of hot flashes is unclear, evidence suggests that estrogen withdrawal, not low levels of estrogen, leads to hot flashes (Pachman, Jones, & Loprinzi, 2010). With estrogen level decline, endorphin production is decreased reducing norepinephrine and serotonin levels and increasing norepinephrine and serotonin receptors in the hypothalamus. It is theorized that this upregulation narrows the set point in the thermoregulatory nucleus and triggers hot flashes (Pachman et al., 2010). Because selective serotonin reuptake inhibitors (SSRIs) and serotonin/norepinephrine reuptake inhibitors (SNRIs) increase serotonin and norepinephrine levels systemically, they have the potential to reduce hot flashes and be considered an alternative to HT for reduction of vasomotor symptoms (Carroll & Kelley, 2009). Evidence now supports the efficacy of SSRIs and SNRIs in reducing hot flashes. Research began with breast cancer patients and has expanded to the general population of healthy peri- and postmenopausal women. Most healthcare providers will care for women experiencing vasomotor symptoms and will need the best evidence to adequately advise women. Practitioners are generally familiar with the risks in using HT, but may not be fully aware of effective nonhormonal treatment options (Hickey, Emery, Gregson, Doherty, & Saunders, 2010). The purpose of this article is to systematically review the evidence related to the efficacy and tolerability of SSRIs and SNRIs used for the treatment of vasomotor symptoms in peri- and postmenopausal women. Recommendations for clinical practice are based on the results of the evidence and analysis of the outcomes.
Methods Identification of studies An electronic search was conducted to identify randomized controlled trials (RCTs) from 2007 to 2012 in the following databases: Medline, CINAHL, and the Cochrane Library. To increase the quality of the review, research articles dating back to 2000 were included. The articles’ reference lists were examined for additional publications.
Search terms used were antidepressant, venlafaxine, fluoxetine, sertraline, paroxetine, citalopram, escitalopram, desvenlafaxine, SSRI, SNRI, and hot flash. Only peerreviewed, RCT articles published in English were included. Articles that were included evaluated the efficacy and tolerability of SSRIs and SNRIs in peri- and postmenopausal women experiencing vasomotor symptoms who did not have other health-related conditions, such as cancer, at the time of treatment. Pilot studies and trials using patients with depression or anxiety were excluded. A total of 18 trials met the criteria for review.
Data extraction Two independent reviewers (AH and MW) extracted data from the articles for the systematic review. Data were extracted into eight categories, including study design, treatment, length and location of trial, number of participants, patient characteristics, outcomes, and adverse effects. Differences were resolved by consensus between reviewers after consulting the article.
Quality of the studies Assessment of the studies’ methodological rigor (Table S1) was achieved by using the Critical Appraisal Skills Programme Questionnaire (CASP), which uses 10 questions to evaluate study design, participant allocation, blinding, follow-up, power calculation, results, and application (Guyatt, Sackett, & Cook, 1994). The questions are scored by answering yes, no, or cannot tell.
Results Patient characteristics RCTs in this review selected healthy peri- and postmenopausal women aged 27–78 experiencing hot flashes. The inclusion requirement for number of hot flashes per week ranged from a minimum of none to at least 50 per week. Of the studies that reported the mean number per week, the range was 46–76 hot flashes/week. The majority of participants were Caucasian. Women using psychoactive drugs, HT, antidepressants, chemotherapy, or radiation were excluded. Some studies included women with a history of cancer and allowed stable tamoxifen or aromatase inhibitor use; patients with active cancer at time of treatment were excluded.
Study characteristics A total of 18 trials were evaluated (Table S1). All studies were randomized controlled, double-blinded with the 55
Vasomotor symptoms in menopausal women
exception of Kalay, Demir, Haberal, Kalay, and Kandemir (2007), which was single-blinded. Four studies were crossover trials (Carpenter et al., 2007; Gordon, Kerwin, Boesen, & Senf, 2006; Loprinzi et al., 2002; Stearns et al., 2005). Suvanto-Luukkonen et al. (2005) and Boekhout et al. (2011) compared two drugs to placebo.
A. P. Handley & M. Williams
Suvanto-Luukkonen et al. (2005) studied 150 women and compared fluoxetine and citalopram to placebo. All three groups had a significant reduction in hot flash frequency but there was no significant difference between the groups. The dropout rates (14% at 3 months and 36% at 9 months) were not significant between groups. The majority of withdrawals (58%) were due to ineffectiveness of treatment rather than adverse effects (28%).
Measurements All studies evaluated hot flashes through a validated, self-reported, daily hot flash diary, recording frequency and severity. Severity was measured from 0 to 4 and was based on hot flash duration. Composite scores, the product of the frequency and severity ratings, were then calculated. Higher scores indicated more severe symptoms. Five studies added the Green Climacteric Scale (GCS) to assess four menopausal symptoms and quality of life components: psychologic, somatic, vasomotor, and sexual (Archer, Dupont, et al., 2009; Archer, Seidman, Constantine, Pickar, & Olivier, 2009; Grady et al., 2007; Soares et al., 2008; Stearns, Beebe, Iyengar, & Dube, 2003). Scores on the GCS range from 0 to 63 with higher scores indicating more severe symptoms. Two studies used the Hot Flash Related Daily Interference Scale, a validated scale that measures the degree to which hot flashes interfere with daily activities and quality of life (Barton et al., 2010; Carpenter et al., 2007). One study measured physiological hot flash frequency using an ambulatory sternal skin conductance monitor (Carpenter et al., 2007). Two studies used the modified Kupperman index that grades menopausal symptoms on a scale of 0–3 and weights the most common symptoms, multiplying hot flashes by 4 and sweating and insomnia by 2 (Kalay et al., 2007; Suvanto-Luukkonen et al., 2005).
Outcomes Selective serotonin reuptake inhibitors.
Paroxetine All three trials studying paroxetine (Paxil) found significant reductions in hot flash frequency and/or composite scores (Soares et al., 2008; Stearns et al., 2003, 2005). Soares et al. (2008) studied 56 women and reported a 6.1/week reduction in hot flash frequency on paroxetine controlled release (CR) compared to 2.8/week on placebo (p = .03). Stearns et al.’s (2005) trial of 151 women on two doses of paroxetine reported a reduced frequency of 40.6% on 10 mg compared to 13.7% on placebo (p = .0006) and 51.7% on 20 mg compared to 26.6% placebo (p = .002). Stearns et al. (2003) studied 165 women on two doses of paroxetine CR (12.5 and 25 mg) and reported hot flash score reductions of 62% and 64% compared to 38% on placebo (p = .007 and .03). Each study addressed dosing by including a low-dose and high-dose arm or by allowing a dose increase during the study. The low dosage had less toxicity (Stearns et al., 2003, 2005); however, the majority of subjects titrated up to the 25 mg dose in the Soares et al. (2008) study. Hot flash score reductions were similar between the low and high doses in the Stearns et al. (2003) study, suggesting that the lower dose is an adequate starting dose that can be increased if the patient does not respond. Stearns et al. (2003) suggest to use the controlled release form of paroxetine because it is better tolerated. The only statistically significant adverse effect compared to placebo was nausea at the 20 mg dose (Stearns et al., 2005).
Fluoxetine
Sertraline
Two trials studied the use of fluoxetine (Prozac) in the treatment of hot flashes. Loprinzi et al. (2002) studied 81 women in a crossover trial and reported a 50% reduction of hot flash scores on fluoxetine compared to 36% reduction on placebo, a nonsignificant difference (p = .35). The crossover analysis found improved reductions of 24% for hot flash scores (p = .02) and 19% for hot flash frequency (p = .01) compared with placebo. The trial’s 18% withdrawal rate created a sample size too small to produce adequate power. There were no differences in adverse symptoms between groups.
Three studies addressed the use of sertraline (Zoloft) to alleviate hot flashes (Gordon et al., 2006; Grady et al., 2007; Wu et al., 2009). Wu et al. (2009) studied 46 women and Grady et al. (2007) had 99 participants and both trials failed to find statistically significant decreases in hot flash frequencies or scores. Yet, both studies were underpowered. Gordon et al. (2006) studied 97 women and reported a modest effect of five fewer hot flashes/week on sertraline compared to placebo (p = .002). However, reduction in composite severity scores was nonsignificant.
56
Vasomotor symptoms in menopausal women
A. P. Handley & M. Williams
The dropout rate in all three trials was 10%–11%. Wu et al. (2009) did not follow-up on reasons for withdrawals although no serious drug-related adverse events were reported. In the Grady et al. (2007) trial, participants in the sertraline group reported increased nausea and decreased sexual function compared to those on placebo (p = .005, .001). In the Gordon et al. (2006) study, 49% of the participants reported nausea during week 1 compared to 19% on placebo (p < .001). However, after the first week there were no differences in symptoms between groups. All three trials using sertraline to reduce hot flashes have significant limitations. Each failed to reach the number of subjects to have adequate power. Wu et al. (2009) had a small sample size and did not explain the method or reasoning for titrating doses from 25 to 100 mg/day. Gordon et al. (2006) had no minimum requirement of hot flashes for study inclusion and although the mean of 45/week is similar to other trials, the range was as low as 2/week, making it difficult for subjects to achieve a significant reduction in frequency.
Citalopram A trial of 254 women conducted by Barton et al. (2010) found citalopram (Celexa) significantly reduced hot flash scores by 49%–55% depending upon dosage compared to 23% reduction on placebo (p ࣘ .002). There were no significant differences in adverse effects between the citalopram and placebo groups. However, the authors provided no discussion regarding the 20% withdrawal rate in the study or whether the dropouts were related to adverse effects. In evaluating dosing response, the authors suggest beginning with the 10 mg dose and increasing to 20 mg if necessary. The 30 mg dose was not recommended because of higher toxicity with minimal increased hot flash improvement compared to lower doses. Kalay et al. (2007) studied 100 participants and found citalopram significantly reduced hot flash scores by 37% compared to 13% placebo (p = .0001). No participants withdrew from the study because of adverse effects. However, a limitation to this study was the single-blind design to allow doctors to titrate the dose up to 40 mg if subject was not responding to the 20 mg dose.
Escitalopram Freeman et al. (2011) conducted a study of 205 women and found escitalopram (Lexapro) significantly reduced hot flash frequency and severity compared to placebo (47% vs. 33%, p < .001; 24% vs. 14%, p < .001). This trial examined racial differences in treatment response; approximately half of the subjects were black. Results found that race did not significantly affect treatment re-
sponse. The trial had a low withdrawal rate of 4%, no serious adverse events were reported, and there were no significant differences in symptoms between treatment groups.
Serotonin–norepinephrine reuptake inhibitors. Venlafaxine Four studies addressed the use of venlafaxine (Effexor) for treatment of vasomotor symptoms. Boekhout et al. (2011) utilized 102 participants to compare venlafaxine and clonidine to placebo. Both drugs significantly reduced hot flash scores, but venlafaxine created a more immediate effect and stronger reduction (41%, 26% vs. placebo [p < .001, .045]). The adverse effects of nausea and constipation were significantly higher in the venlafaxine group compared to placebo. Researchers suggest that practitioners start at a lower 37.5 mg dose and only increase as needed in order to decrease the potential for side effects. The trial by Carpenter et al. (2007) is the only study in this review to measure physiological hot flashes using an ambulatory sternal skin conductance monitor. This study of 70 participants found that physiological frequency was reduced by 22% compared to 0% placebo (p = .001) at the 37.5 mg dose and by 14% compared to a 13% increase on placebo (p = .013) at the 75 mg dose. Diary reported frequency was reduced by 42% compared to 18% placebo (p = .001) at the low dose and by 25% compared to 4% placebo (p = .001) at the high dose. Patients taking 75 mg reported significantly less trouble sleeping, but significantly more constipation and dry mouth (p = .034, .032, .002, respectively). Patients on the low dose had significantly more constipation, headaches, and dry mouth compared to placebo (p = .001, .007,
The efficacy and tolerability of SSRI/SNRIs in the treatment of vasomotor symptoms in menopausal women: A systematic review Amy P. Handley, MSN, FNP-C (Associate Professor and Associate Dean)1 & Mary Williams, PhD, RN (Nurse Practitioner)2 1 2
College of Nursing, Brigham Young University, Provo, Utah Graduate Studies and Faculty, College of Nursing, Brigham Young University, Provo, Utah
Keywords Hot flashes; vasomotor symptoms; menopause; selective serotonin reuptake inhibitors (SSRI); hormone replacement therapy. Correspondence Amy Handley, Lakeview Family Medicine, 811 N 900 W, Orem, UT 84057. Tel: 801-812-3200; E-mail: [email protected] Received: 19 September 2012; accepted: 25 February 2013 doi: 10.1002/2327-6924.12137 Disclosures: The authors report no conflicting interests.
Abstract Purpose: To systematically review the evidence related to the efficacy and tolerability of selective serotonin reuptake inhibitors (SSRIs) and serotonin/norepinephrine reuptake inhibitors (SNRIs) used for the treatment of vasomotor symptoms in perimenopausal and postmenopausal women. Data sources: Medline, CINAHL, and the Cochrane Library databases were searched to identify randomized controlled trials (RCTs). Eighteen trials met the criteria for review. Conclusions: Results from these trials indicate that paroxetine, citalopram, escitalopram, venlafaxine, and desvenlafaxine are effective in reducing the frequency and severity of hot flashes. Fluoxetine and sertraline appear to be less effective and should be considered second-line options for treatment. Implications for practice: The SSRIs and SNRIs can reduce hot flashes by 65% and begin working within the first week. Patient response is variable and if one drug does not improve hot flashes, another can be tried after a 1- to 2-week drug trial. Paroxetine, citalopram, and escitalopram appear to have the fewest adverse effects. Considering cost, paroxetine and citalopram are the most cost-efficient.
Vasomotor symptoms are the most commonly reported symptoms during the menopausal transition (Speroff, Gass, Constantine, & Olivier, 2008). Hot flashes, one of the most common vasomotor symptoms, are most prevalent in the late menopausal transition and occur in about 65% of all women (Grady, 2006). These symptoms usually resolve within 4–5 years; however, 10%–15% of symptomatic women continue to have hot flashes for several years longer (Grady, 2006). Vasomotor symptoms disrupt sleep, reduce quality of life, increase irritability, and are the menopausal symptoms that most cause women to seek treatment (Archer, Dupont, Constantine, Pickar, & Olivier, 2009). Hormone therapy (HT) is the treatment of choice for menopause-related vasomotor symptoms, reducing frequency by approximately 75% (MacLennan, Broadbent, Lester, & Moore, 2004). The current recommendation is to prescribe the lowest effective dose for the shortest time period, although extended therapy 54
is acceptable under individualized circumstances (North American Menopause Society, 2012). However, current evidence indicates that HT is not appropriate for all women. In general, HT is not recommended in women with a history of endometrial cancer or breast cancer because of possible increased risk for recurrence (North American Menopause Society, 2012). Obese women, those with factor V Leiden mutation, and women with a history of venous thromboembolism are at increased risk for venous thromboembolism during HT use (North American Menopause Society, 2012). The known risks of HT use in healthy women aged 50–59 are low; however, long-term HT and initiation of therapy in women aged 60 and older are associated with greater risks (North American Menopause Society, 2012). This information has led practitioners and women transitioning to postmenopause to seek viable nonhormonal alternatives for the relief of vasomotor symptoms (Carroll & Kelley, 2009).
Journal of the American Association of Nurse Practitioners 27 (2015) 54–61 C 2014 American Association of Nurse Practitioners
Vasomotor symptoms in menopausal women
A. P. Handley & M. Williams
Because the mechanism of hot flashes is not fully understood, research has been conducted on a variety of nonhormonal treatments to alleviate vasomotor symptoms (Hall, Frey, & Soares, 2011). During the transition to postmenopause, changes in the neuroendocrine system are thought to disrupt thermoregulation in the hypothalamus, resulting in body core temperature changes. The body responds through perspiration and vasodilation, producing a hot flash (Carroll & Kelley, 2009). Also during this transition, estrogen and progesterone levels decline. Although the role of estrogens in the production of hot flashes is unclear, evidence suggests that estrogen withdrawal, not low levels of estrogen, leads to hot flashes (Pachman, Jones, & Loprinzi, 2010). With estrogen level decline, endorphin production is decreased reducing norepinephrine and serotonin levels and increasing norepinephrine and serotonin receptors in the hypothalamus. It is theorized that this upregulation narrows the set point in the thermoregulatory nucleus and triggers hot flashes (Pachman et al., 2010). Because selective serotonin reuptake inhibitors (SSRIs) and serotonin/norepinephrine reuptake inhibitors (SNRIs) increase serotonin and norepinephrine levels systemically, they have the potential to reduce hot flashes and be considered an alternative to HT for reduction of vasomotor symptoms (Carroll & Kelley, 2009). Evidence now supports the efficacy of SSRIs and SNRIs in reducing hot flashes. Research began with breast cancer patients and has expanded to the general population of healthy peri- and postmenopausal women. Most healthcare providers will care for women experiencing vasomotor symptoms and will need the best evidence to adequately advise women. Practitioners are generally familiar with the risks in using HT, but may not be fully aware of effective nonhormonal treatment options (Hickey, Emery, Gregson, Doherty, & Saunders, 2010). The purpose of this article is to systematically review the evidence related to the efficacy and tolerability of SSRIs and SNRIs used for the treatment of vasomotor symptoms in peri- and postmenopausal women. Recommendations for clinical practice are based on the results of the evidence and analysis of the outcomes.
Methods Identification of studies An electronic search was conducted to identify randomized controlled trials (RCTs) from 2007 to 2012 in the following databases: Medline, CINAHL, and the Cochrane Library. To increase the quality of the review, research articles dating back to 2000 were included. The articles’ reference lists were examined for additional publications.
Search terms used were antidepressant, venlafaxine, fluoxetine, sertraline, paroxetine, citalopram, escitalopram, desvenlafaxine, SSRI, SNRI, and hot flash. Only peerreviewed, RCT articles published in English were included. Articles that were included evaluated the efficacy and tolerability of SSRIs and SNRIs in peri- and postmenopausal women experiencing vasomotor symptoms who did not have other health-related conditions, such as cancer, at the time of treatment. Pilot studies and trials using patients with depression or anxiety were excluded. A total of 18 trials met the criteria for review.
Data extraction Two independent reviewers (AH and MW) extracted data from the articles for the systematic review. Data were extracted into eight categories, including study design, treatment, length and location of trial, number of participants, patient characteristics, outcomes, and adverse effects. Differences were resolved by consensus between reviewers after consulting the article.
Quality of the studies Assessment of the studies’ methodological rigor (Table S1) was achieved by using the Critical Appraisal Skills Programme Questionnaire (CASP), which uses 10 questions to evaluate study design, participant allocation, blinding, follow-up, power calculation, results, and application (Guyatt, Sackett, & Cook, 1994). The questions are scored by answering yes, no, or cannot tell.
Results Patient characteristics RCTs in this review selected healthy peri- and postmenopausal women aged 27–78 experiencing hot flashes. The inclusion requirement for number of hot flashes per week ranged from a minimum of none to at least 50 per week. Of the studies that reported the mean number per week, the range was 46–76 hot flashes/week. The majority of participants were Caucasian. Women using psychoactive drugs, HT, antidepressants, chemotherapy, or radiation were excluded. Some studies included women with a history of cancer and allowed stable tamoxifen or aromatase inhibitor use; patients with active cancer at time of treatment were excluded.
Study characteristics A total of 18 trials were evaluated (Table S1). All studies were randomized controlled, double-blinded with the 55
Vasomotor symptoms in menopausal women
exception of Kalay, Demir, Haberal, Kalay, and Kandemir (2007), which was single-blinded. Four studies were crossover trials (Carpenter et al., 2007; Gordon, Kerwin, Boesen, & Senf, 2006; Loprinzi et al., 2002; Stearns et al., 2005). Suvanto-Luukkonen et al. (2005) and Boekhout et al. (2011) compared two drugs to placebo.
A. P. Handley & M. Williams
Suvanto-Luukkonen et al. (2005) studied 150 women and compared fluoxetine and citalopram to placebo. All three groups had a significant reduction in hot flash frequency but there was no significant difference between the groups. The dropout rates (14% at 3 months and 36% at 9 months) were not significant between groups. The majority of withdrawals (58%) were due to ineffectiveness of treatment rather than adverse effects (28%).
Measurements All studies evaluated hot flashes through a validated, self-reported, daily hot flash diary, recording frequency and severity. Severity was measured from 0 to 4 and was based on hot flash duration. Composite scores, the product of the frequency and severity ratings, were then calculated. Higher scores indicated more severe symptoms. Five studies added the Green Climacteric Scale (GCS) to assess four menopausal symptoms and quality of life components: psychologic, somatic, vasomotor, and sexual (Archer, Dupont, et al., 2009; Archer, Seidman, Constantine, Pickar, & Olivier, 2009; Grady et al., 2007; Soares et al., 2008; Stearns, Beebe, Iyengar, & Dube, 2003). Scores on the GCS range from 0 to 63 with higher scores indicating more severe symptoms. Two studies used the Hot Flash Related Daily Interference Scale, a validated scale that measures the degree to which hot flashes interfere with daily activities and quality of life (Barton et al., 2010; Carpenter et al., 2007). One study measured physiological hot flash frequency using an ambulatory sternal skin conductance monitor (Carpenter et al., 2007). Two studies used the modified Kupperman index that grades menopausal symptoms on a scale of 0–3 and weights the most common symptoms, multiplying hot flashes by 4 and sweating and insomnia by 2 (Kalay et al., 2007; Suvanto-Luukkonen et al., 2005).
Outcomes Selective serotonin reuptake inhibitors.
Paroxetine All three trials studying paroxetine (Paxil) found significant reductions in hot flash frequency and/or composite scores (Soares et al., 2008; Stearns et al., 2003, 2005). Soares et al. (2008) studied 56 women and reported a 6.1/week reduction in hot flash frequency on paroxetine controlled release (CR) compared to 2.8/week on placebo (p = .03). Stearns et al.’s (2005) trial of 151 women on two doses of paroxetine reported a reduced frequency of 40.6% on 10 mg compared to 13.7% on placebo (p = .0006) and 51.7% on 20 mg compared to 26.6% placebo (p = .002). Stearns et al. (2003) studied 165 women on two doses of paroxetine CR (12.5 and 25 mg) and reported hot flash score reductions of 62% and 64% compared to 38% on placebo (p = .007 and .03). Each study addressed dosing by including a low-dose and high-dose arm or by allowing a dose increase during the study. The low dosage had less toxicity (Stearns et al., 2003, 2005); however, the majority of subjects titrated up to the 25 mg dose in the Soares et al. (2008) study. Hot flash score reductions were similar between the low and high doses in the Stearns et al. (2003) study, suggesting that the lower dose is an adequate starting dose that can be increased if the patient does not respond. Stearns et al. (2003) suggest to use the controlled release form of paroxetine because it is better tolerated. The only statistically significant adverse effect compared to placebo was nausea at the 20 mg dose (Stearns et al., 2005).
Fluoxetine
Sertraline
Two trials studied the use of fluoxetine (Prozac) in the treatment of hot flashes. Loprinzi et al. (2002) studied 81 women in a crossover trial and reported a 50% reduction of hot flash scores on fluoxetine compared to 36% reduction on placebo, a nonsignificant difference (p = .35). The crossover analysis found improved reductions of 24% for hot flash scores (p = .02) and 19% for hot flash frequency (p = .01) compared with placebo. The trial’s 18% withdrawal rate created a sample size too small to produce adequate power. There were no differences in adverse symptoms between groups.
Three studies addressed the use of sertraline (Zoloft) to alleviate hot flashes (Gordon et al., 2006; Grady et al., 2007; Wu et al., 2009). Wu et al. (2009) studied 46 women and Grady et al. (2007) had 99 participants and both trials failed to find statistically significant decreases in hot flash frequencies or scores. Yet, both studies were underpowered. Gordon et al. (2006) studied 97 women and reported a modest effect of five fewer hot flashes/week on sertraline compared to placebo (p = .002). However, reduction in composite severity scores was nonsignificant.
56
Vasomotor symptoms in menopausal women
A. P. Handley & M. Williams
The dropout rate in all three trials was 10%–11%. Wu et al. (2009) did not follow-up on reasons for withdrawals although no serious drug-related adverse events were reported. In the Grady et al. (2007) trial, participants in the sertraline group reported increased nausea and decreased sexual function compared to those on placebo (p = .005, .001). In the Gordon et al. (2006) study, 49% of the participants reported nausea during week 1 compared to 19% on placebo (p < .001). However, after the first week there were no differences in symptoms between groups. All three trials using sertraline to reduce hot flashes have significant limitations. Each failed to reach the number of subjects to have adequate power. Wu et al. (2009) had a small sample size and did not explain the method or reasoning for titrating doses from 25 to 100 mg/day. Gordon et al. (2006) had no minimum requirement of hot flashes for study inclusion and although the mean of 45/week is similar to other trials, the range was as low as 2/week, making it difficult for subjects to achieve a significant reduction in frequency.
Citalopram A trial of 254 women conducted by Barton et al. (2010) found citalopram (Celexa) significantly reduced hot flash scores by 49%–55% depending upon dosage compared to 23% reduction on placebo (p ࣘ .002). There were no significant differences in adverse effects between the citalopram and placebo groups. However, the authors provided no discussion regarding the 20% withdrawal rate in the study or whether the dropouts were related to adverse effects. In evaluating dosing response, the authors suggest beginning with the 10 mg dose and increasing to 20 mg if necessary. The 30 mg dose was not recommended because of higher toxicity with minimal increased hot flash improvement compared to lower doses. Kalay et al. (2007) studied 100 participants and found citalopram significantly reduced hot flash scores by 37% compared to 13% placebo (p = .0001). No participants withdrew from the study because of adverse effects. However, a limitation to this study was the single-blind design to allow doctors to titrate the dose up to 40 mg if subject was not responding to the 20 mg dose.
Escitalopram Freeman et al. (2011) conducted a study of 205 women and found escitalopram (Lexapro) significantly reduced hot flash frequency and severity compared to placebo (47% vs. 33%, p < .001; 24% vs. 14%, p < .001). This trial examined racial differences in treatment response; approximately half of the subjects were black. Results found that race did not significantly affect treatment re-
sponse. The trial had a low withdrawal rate of 4%, no serious adverse events were reported, and there were no significant differences in symptoms between treatment groups.
Serotonin–norepinephrine reuptake inhibitors. Venlafaxine Four studies addressed the use of venlafaxine (Effexor) for treatment of vasomotor symptoms. Boekhout et al. (2011) utilized 102 participants to compare venlafaxine and clonidine to placebo. Both drugs significantly reduced hot flash scores, but venlafaxine created a more immediate effect and stronger reduction (41%, 26% vs. placebo [p < .001, .045]). The adverse effects of nausea and constipation were significantly higher in the venlafaxine group compared to placebo. Researchers suggest that practitioners start at a lower 37.5 mg dose and only increase as needed in order to decrease the potential for side effects. The trial by Carpenter et al. (2007) is the only study in this review to measure physiological hot flashes using an ambulatory sternal skin conductance monitor. This study of 70 participants found that physiological frequency was reduced by 22% compared to 0% placebo (p = .001) at the 37.5 mg dose and by 14% compared to a 13% increase on placebo (p = .013) at the 75 mg dose. Diary reported frequency was reduced by 42% compared to 18% placebo (p = .001) at the low dose and by 25% compared to 4% placebo (p = .001) at the high dose. Patients taking 75 mg reported significantly less trouble sleeping, but significantly more constipation and dry mouth (p = .034, .032, .002, respectively). Patients on the low dose had significantly more constipation, headaches, and dry mouth compared to placebo (p = .001, .007,
