help button home button The Oncologist http://theoncologist.alphamedpress.org/subscriptions/etoc.dtl
HOME HELP CONTACT US SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The Oncologist, Vol. 12, No. 1, 124-135, January 2007; doi:10.1634/theoncologist.12-1-124
© 2007 AlphaMed Press

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow CME: Take the course for this article:
 Randomized, Double-Blind, Placebo-Controlled Crossover Trials of Venlafaxin...
Right arrow eLetters: Submit a response to this article
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article link to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carpenter, J. S.
Right arrow Articles by Shelton, R. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carpenter, J. S.
Right arrow Articles by Shelton, R. C.

Symptom Management and Supportive Care

Randomized, Double-Blind, Placebo-Controlled Crossover Trials of Venlafaxine for Hot Flashes After Breast Cancer

Janet S. Carpentera, Anna Maria Stornioloa, Shelley Johnsa, Patrick O. Monahana, Faouzi Azzouza, Julie L. Elama, Cynthia S. Johnsona, Richard C. Sheltonb

a Indiana University, Indianapolis, Indiana, USA; b Vanderbilt University, Nashville, Tennessee, USA

Key Words. Antidepressive agents • Serotonin reuptake inhibitors • Hot flashes • Menopause

Janet S. Carpenter, Ph.D., R.N., Indiana University, 1111 Middle Drive NU340D, Indianapolis, Indiana 46202, USA. Telephone: 317-278-6093; Fax: 317-274-2021; e-mail: carpentj{at}iupui.edu; Web site: http://expertise.cos.com

Received June 29, 2006; accepted for publication September 29, 2006.


   LEARNING OBJECTIVES
Top
 Learning Objectives
Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Disclosure of Potential...
 References
 
After completing this course, the reader will be able to:

  1. Discuss the efficacy of venlafaxine in alleviating hot flashes and improving secondary outcomes.
  2. Identify the week of treatment that venlafaxine was most effective.
  3. List three side effects associated with venlafaxine.

Access and take the CME test online and receive 1 AMA PRA Category 1 CreditTM at CME.TheOncologist.com


   ABSTRACT
Top
 Learning Objectives
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Disclosure of Potential...
 References
 
Background. Although venlafaxine reduces self-reported hot flashes, no data have established the drug’s impact on physiologically documented hot flashes. Two randomized, double-blind, placebo-controlled crossover trials examined the efficacy of two doses of venlafaxine in relation to physiological and self-reported hot flashes and other outcomes, including negative affect, fatigue, sleep, and quality of life.

Methods. Sample: 57 breast cancer survivors in the low-dose study; 20 in the high-dose study. Setting: university cancer clinics in the Southeast and Midwest. Intervention: 37.5 mg of venlafaxine (low-dose study) or 75 mg of venlafaxine (high-dose study). Measures: hot flash frequency (physiological monitor, diary, and event marker), hot flash severity (diary), hot flash bother (diary), and questionnaires for hot flash impact on daily life, negative affect, fatigue, sleep, and quality of life.

Results. Subjective but not physiological hot flash measures showed placebo effects. Venlafaxine resulted in modest decreases in hot flashes, but only hot flash interference improved differentially at the higher dose. The timing of venlafaxine’s effects on hot flashes varied by dose. Only women with a >50% decrease in physiological hot flashes experienced significant improvement in fatigue, sleep quality, and quality of life. Although side effects were mild, most patients discontinued venlafaxine long-term.

Conclusions. Although venlafaxine resulted in modest and acute reductions in hot flashes with few side effects, it may not be tolerable to some patients long-term. At least 50% relief in physiological hot flashes may be needed for patients to demonstrate improvement in other outcomes, including decreased fatigue, improved sleep, and improved quality of life.


   INTRODUCTION
Top
 Learning Objectives
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Conclusion
 Disclosure of Potential...
 References
 
Frequent, severe, and bothersome hot flashes are well documented in breast cancer survivors [1–5]. Approximately 65% of survivors experience hot flashes, with 59% rating the symptom as severe and 44% rating the symptom as extremely bothersome. In breast cancer survivors, unrelieved hot flashes are related to negative affect, fatigue, sleep difficulties, and overall poor quality of life [6].

Serotonin and/or norepinephrine reuptake inhibitors (SSRIs/SNRIs) are widely used in the treatment of hot flashes in breast cancer survivors. Controlled trials indicate that the SNRI venlafaxine (Effexor XR; Wyeth Pharmaceuticals, Madison, NJ) and the SSRI paroxetine (Paxil; Glaxo-SmithKline, Philadelphia) are more effective than placebo in decreasing patient reports of the number and severity of hot flashes [7, 8]. Paroxetine is a powerful inhibitor of cytochrome P450 2D6 (CYP2D6), which metabolizes tamoxifen to the potent metabolite endoxifen [9–11]. In contrast, venlafaxine is a weak inhibitor of CYP2D6 and only slightly reduces plasma concentrations of endoxifen, making it preferable to paroxetine.

Two randomized, controlled trials of venlafaxine have been reported. In one, women with a history of breast cancer or who refused estrogen replacement for fear of cancer completed a 1-week baseline and were randomized to 4 weeks of placebo or 37.5 mg, 75 mg, or 150 mg of venlafaxine per day (n = 191) [7]. After stratifying by age, hot flash frequency, tamoxifen use, and duration of hot flashes, hot flash frequency and severity decreased significantly with all doses (37%–67%) compared with placebo (27%) (p <.01); however, it is unclear whether effects varied significantly by dose. The study was not a true dose-response study since different patients received different doses [7], and as pointed out in a recent review article, the data analysis and presentation of results do not clearly indicate whether dose differences were statistically significant [12]. In another study, 80 healthy women were randomized to placebo or 37.5 mg of venlafaxine for 1 week followed by 75 mg of venlafaxine for another 11 weeks [13]. Compared with placebo, venlafaxine did not significantly change hot flashes but did improve mental health and vitality [13].

These trials did not include physiological tests of the effect of venlafaxine on hot flashes because they did not include a physiological measure of hot flashes [7, 13]. Change in self-reported hot flashes may not be synonymous with physiological change since self-reports are known to underestimate physiologically documented hot flashes [14–18]. The discrepancy between physiological and self-reported hot flashes may relate to several factors, including the expectation of benefit from treatment and/or poor compliance with diaries, particularly at night [17, 19]. Understanding venlafaxine’s role in alleviating physiological hot flashes is important for elucidating underlying biological mechanisms so that better therapies can be developed [3].

The objectives of the study were to evaluate venlafaxine’s (a) efficacy for self-reported and physiological hot flashes, (b) efficacy for other related symptoms (negative affect, fatigue, sleep difficulties, and overall quality of life), and (c) side effects. Compared with placebo, we hypothesized that breast cancer survivors treated with venlafaxine would exhibit (a) less frequent subjective and objective hot flashes and less subjective hot flash severity, bother, and interference (primary outcomes); (b) less negative affect, fatigue, sleep difficulties, and improved quality of life (secondary outcomes); and (c) minimal side effects with a desire to stay on venlafaxine long-term. To meet the objectives, we performed two sequential clinical trials: a 37.5-mg low-dose study and a 75-mg high-dose study.


   MATERIALS AND METHODS
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Conclusion
 Disclosure of Potential...
 References
 
Sample and Setting
Breast cancer survivors were recruited between 2000 and 2004 with follow-up continuing through October 2005 from cancer center clinics in the Southeast (low-dose study) and Midwest (high-dose study). Both studies were approved by local institutional review boards, and all patients provided written informed consent.

Inclusion criteria were (a) adult women with a history of breast cancer, (b) no other cancer, (c) disease-free and functioning independently at the time of study enrollment, (d) at least 4 weeks removed from completing local therapy (chemotherapy or radiation), (e) experiencing daily hot flashes (≥1 per day), (f) desirous of treatment for hot flashes but not currently using any other hot flash treatments, (g) post-menopausal or using a clinically acceptable method of birth control throughout the study to prevent pregnancy (e.g., birth control pills or intrauterine device), (h) living within 60 miles of the study site to enable access to the hot flash monitor, and (i) verified as being nondepressed by the study psychologist through structured clinical interview [20]. Excluded were those (a) on tamoxifen or aromatase inhibitor for < 6 weeks, (b) on antidepressants, (c) receiving hot flash treatment within the past 4 weeks (e.g., soy supplements, botanicals, vitamin E, and prescription medications), or (d) pregnant or lactating.

Design and Intervention
Each study was a 14-week, randomized, double-blind, placebo-controlled crossover trial. Treatment in the low-dose study consisted of 6 weeks of 37.5 mg of venlafaxine daily (n = 64). The high-dose treatment consisted of 1 week of 37.5 mg of venlafaxine daily + 4 weeks of 75 mg of venlafaxine daily + 1 week of 37.5 mg of venlafaxine daily (n = 20). Extended release formulations were used. There were no washout periods because of the relatively short half-life of the drug (5 hours) and its major metabolite (11 hours).

Consenting women were scheduled for 14 weekly visits. Weeks 1 and 2 provided baseline information (B1 and B2), and weeks 3 through 14 included 6 weeks of treatment (T1–T6) and 6 weeks of placebo (P1–P6). Thus, low-dose study patients received 37.5 mg each day during T1–T6, whereas high-dose study patients received 75 mg each day during T2–T5 and 37.5 mg at T1 and T6 as they tapered on and off the drug. Trained study nurses met patients in the clinic or at their homes or places of work to reduce subject burden and maintain consistent follow-up. The average time per visit was 10 minutes. In addition, patients were telephoned 1, 6, and 12 months after completing the weekly visits to assess continued venlafaxine use.

Assignment and Blinding
Randomization and blinding were carried out by the university’s investigational drug services pharmacy. This group computer-generated the randomization sequence (without blocking or stratification), purchased the study drug and placebo, and blindly dispensed study medication and placebo. After nurses determined eligibility and obtained consent, the patient’s study number was sent to the pharmacist, who then randomized the patient to one of two different sequence groups: drug/placebo or placebo/drug. Patients, nurses, and investigators were blinded. Placebo and treatment capsules were opaque, identical in size, color, and shape, and placed in standard medicine bottles. For the low-dose study, two bottles were used (placebo and treatment). For the high-dose study, three bottles were used during each study arm to blind patients to the taper-on, treatment, and taper-off doses. Each bottle was dispensed individually by the study nurse at the appropriate patient visit. Unblinding was done at the end of each study, except for one high-dose study participant who had a significant adverse event (hypertension).

Measures
A study psychologist verified the absence of depressive symptoms using two measures: the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders [20] and the well-validated 17-item Hamilton Rating Scale-Depression (Ham-D) [21, 22]. Both were administered in person prior to randomization at B1 or B2 and again at P6 and T6. Interviews lasted ≤60 minutes and were conducted in the psychologist’s office. In addition, the Ham-D was administered over the telephone throughout placebo (P2 and P4) and treatment arms (T2 and T4) to evaluate for depression throughout the duration of the trial.

Adherence to treatment was assessed using capsule counts and weekly written verification by patients that no new hot flash treatments were started during the study. At the end of the study, women were asked whether the first treatment or second treatment was placebo. Their responses were compared to the randomization schema to determine whether they knew when they had received placebo (i.e., if they were correct, they were not successfully blinded).

Physiological hot flash frequency was evaluated using weekly 24-hour ambulatory sternal skin conductance monitoring as previously described and validated [15, 23–25]. Sternal skin conductance monitoring is more specific in detecting hot flashes than other measures of core or peripheral temperature (for review, see [3]) and is highly correlated with self-reported hot flashes under controlled conditions [16]. Using this technology, hot flashes can be differentiated from sweating during exercise or other conditions.

Self-reported hot flash frequency was tabulated using electronic event markers and written diaries that were completed during one 24-hour period each week. With each hot flash, women pressed two red buttons on the hot flash monitor and recorded time, severity, and bother in the paper diary. Severity and bother were rated using two separate 10-point numeric rating scales (0 = not at all and 10 = extremely severe or 10 = extremely bothersome). Hot flash interference, the degree to which hot flashes interfered with daily activities, and quality of life were evaluated using the validated Hot Flash-Related Daily Interference Scale [26].

A negative affect index was calculated as the combination of standardized scores on four questionnaires: the Profile of Mood States Short Form total mood disturbance score (excluding fatigue) [27], the negative affect subscale of the Positive and Negative Affect Scale [28], the Center for Epidemiological Studies Depression Scale [29], and the Ham-D described above. This was done because negative affect is considered to be a conglomerate of anxiety, depression, and other negative mood states [30, 31]. All of these scales are widely used and psychometrically sound.

Fatigue was measured using the Profile of Mood States Short Form fatigue subscale, consistent with other studies on fatigue in cancer patients [27, 32, 33]. The Pittsburgh Sleep Quality Index provided a global score ranging from 0 to 16, with scores ≥5 indicating poor sleep quality and high sleep disturbance [34–36]. Quality of life was assessed using the well-validated 36-item Medical Outcomes Survey (MOS) with its eight subscales (physical function, social function, role limitations due to physical health problems, role limitations due to emotional problems, bodily pain, vitality, mental health, and general health perceptions) [37, 38].

Side effects were assessed with a symptom checklist [39, 40], weekly blood pressure monitoring, and continued venlafaxine use at follow-up. For the symptom checklist, patients indicated whether they had each of the 36 symptoms during the previous week and, if yes, rated severity using a five-point scale (0 = not at all severe, 1 = a little, 2 = moderately, 3 = quite a bit, 4 = extremely severe). The total number of side effects present and total severity of side effects were computed. Blood pressures were monitored during patient visits by study nurses using Gamma 200 ambulatory blood pressure cuffs (R-00.09.010; Standris Medical Supply, Inc., Coatesville, PA, http://www.standris.com). Follow-up questions at 1, 6, and 12 months asked whether patients were taking venlafaxine and, if so, at what dose.

Statistical Analysis
Analyses were performed separately for the two crossover studies (low-dose vs. placebo and high-dose vs. placebo) using mixed linear models with random intercepts. The significance level was .05 for all analyses. All available cases at each time point were used.

In crossover studies, the treatment effect is a within-person effect [41]. Therefore, power for the obtained sample size was estimated using the paired t test. This is a conservative estimate, because the mixed linear analysis will have slightly greater power resulting from reduced residual variance after adjusting for baseline measures and from multiple repeated measures. This power calculation was made even more conservative by using only those patients who completed all 14 study weeks as the sample size (Figs. 1Go, 2Go), whereas mixed linear models use all available data. For our low-dose study (n = 45 at week 14), the two-sided paired t test ({alpha} = 0.05) had 86% power to detect a difference in means equal to 0.50 standard deviation (medium effect size). For our high-dose study (n = 15 at week 14), the two-sided paired t test ({alpha} = 0.05) had 43% power to detect a medium (0.50) effect size and 80% power to detect a large (0.78) effect size.


Figure 1
View larger version (40K):
[in this window]
[in a new window]

  		Figure 1. Low-dose study accrual and retention.*, missed patients (lost to follow-up or unable to contact at one or more follow-ups). Abbreviations: B, baseline; n, number of patients completing the assessment; P, placebo; SE, side effects; T, treatment.

 

Figure 2
View larger version (33K):
[in this window]
[in a new window]

  		Figure 2. High-dose study accrual and retention. Abbreviations: B, baseline; n, number of patients completing the assessment; P, placebo; SE, side effects; T, treatment.

 
To evaluate efficacy, we first examined whether hot flashes differed between placebo and baseline. Because the overall time effect was significant (indicating a placebo effect), we then compared B1 and B2 to each of the placebo weeks to determine when the placebo effect occurred. Second, we evaluated the main treatment effect by comparing drug to placebo and including baseline assessments of the hot flash outcome as two separate covariates in each model. Two separate baseline measures are generally thought to produce better precision and power than a single measure or averaging two measures into one variable. According to the classic analysis of crossover studies, the treatment effect (drug vs. placebo) was a within-person effect averaged over the two randomized groups. Specifically, a classic mixed linear analysis for crossover trials was performed in which the main test of interest is the within-person treatment effect (drug vs. placebo) adjusted for the sequence effect, the period effect, and the two baseline measures to reduce bias [41]. The sequence effect refers to the order in which the patient receives treatment or placebo. Patients were randomized to receive either treatment first then placebo, or placebo first then treatment. Thus, the effect of "sequence" refers to the comparison of the two randomized groups that differ on their sequence (treatment/placebo vs. placebo/ treatment). The period effect refers to the time period while on treatment or placebo (i.e., the first period of 6 weeks vs. the second period of the next 6 weeks).

Although it is impossible [41] to accurately test for the presence of a "carryover" effect (i.e., whether the effect of treatment "carries over" and impacts results while on placebo) in a traditional 2 x 2 crossover design (two groups, two periods) such as this one, one can test the effects of sequence and period-by-treatment interaction, which are possible indicators of a carryover effect. The effects of sequence and period-by-treatment interaction were not significant for any outcomes. Therefore, the period-by-treatment interaction was not included in these models [41]. However, consistent with standard analyses for crossover studies, the sequence effect was included in all models [41]. To help interpret this effect, the adjusted outcome means were reported for drug versus placebo. These adjusted means were calculated from the mixed linear models and adjusted for period, sequence (i.e., averaged over the two sequence groups), and B1 and B2 measures. Effect sizes were calculated based on the difference between the adjusted means divided by the standard deviation. These effect sizes are consistent with the mixed linear analyses; however, for an additional effect size, we reported the percentage change of the adjusted mean from baseline separately for treatment and placebo. For the one outcome that showed no placebo effect (monitor hot flash frequency), we compared each of the six treatment weeks to baseline (B1) to determine the timing of effects.

For secondary outcomes, we evaluated the main treatment effect as described above by comparing drug to placebo, including baseline assessment of the secondary outcome in each model. We also combined women from the low- and high-dose studies to identify a subgroup of 15 women (12 low-dose, 3 high-dose) who obtained a ≥50% reduction in monitor hot flashes while on treatment (B1–T6). This was done because (a) we were interested in whether outcomes improved regardless of dose and (b) the monitor was the only measure that did not show a placebo effect; t tests were used to examine whether the 15 women who obtained the ≥50% reduction demonstrated a significant change from baseline to T6 in secondary outcomes. We used B2 for this comparison as these measures were completed at B2 and not B1.

Finally, we evaluated the main treatment effect as described above for side effects. We evaluated the total number of side effects, total side-effect severity, and specific known side effects of the drug, including diarrhea, trouble sleeping, constipation, headaches, dizziness, dry mouth, nausea, food cravings, and hypertension.


   RESULTS
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Conclusion
 Disclosure of Potential...
 References
 
Low-dose study accrual and retention are shown in Figure 1Go. Baseline demographic and disease/treatment variables were comparable between low-dose study groups. These patients were mainly non-Hispanic whites (91%), married (81%), and working full time (60%). Median education was 15 years, mean age was 50.5 (SD = 8.7), and mean body mass index was 26.9 (SD = 7.6). Low-dose study patients were mostly stage II or less at diagnosis (96%) and had received radiation (68%), chemotherapy (70%), or both (51%). Mean months post-treatment was 41.5 (SD = 39.4), and 51% were taking endocrine therapy.

High-dose study accrual and retention appear in Figure 2Go. Groups were comparable on baseline demographic and disease/treatment variables. These patients were mostly non-Hispanic whites (90%), married (53%), working full time (63%), with a median education of 16 years, a mean age of 53.0 (SD = 8.9), and mean body mass index of 27.9 (SD = 13.0). High-dose study patients were mostly stage II or less at diagnosis (94%), were taking endocrine therapy (63%), were a mean of 26.0 months post-treatment (SD = 17.0), and had received radiation (58%), chemotherapy (63%), or both (38%).

Both accrual figures indicate women withdrew prior to randomization due to not meeting inclusion criteria (e.g., depression and no hot flashes) or lack of interest. Following randomization, women withdrew due to side effects, unknown reasons, or for failing to take approved contraceptives in the low-dose study, whereas in the high-dose study, withdrawals were related to side effects.

Placebo Effect
There was no placebo effect for the physiological hot flash monitor in the low-dose study (p= .52). There was a placebo effect for self-reported hot flashes. Compared with B1, low-dose patients recorded significantly (a) fewer diary hot flashes at P1, P3, P4, P5, and P6 (p< .05) (Fig. 3Go); (b) greater diary hot flash severity during P3, P4, and P6 (p< .003) (Fig. 4Go); and (c) greater bother during P1, P3, P4, P5, and P6 (p < .04) (Fig. 4Go). Findings were similar when using B2 as the comparison.


Figure 3
View larger version (12K):
[in this window]
[in a new window]

  		Figure 3. Low-dose study placebo and treatment effects for hot flash frequency. These are plotted as the average for the week over the two sequence groups (placebo/treatment vs. treatment/placebo) regardless of period (treatment first vs. second) since these were not significant in the mixed linear models. For simplicity, placebo is plotted before treatment. For placebo effects, compare diary at B1 to P1–P6. For treatment effects, compare P1–P6 to T1–T6. Abbreviations: B, baseline; P, placebo; T, treatment.

 

Figure 4
View larger version (11K):
[in this window]
[in a new window]

  		Figure 4. Low-dose study placebo and treatment effects for hot flash severity, bother, and interference. Worsening diary severity is seen by comparing B1 to P3, P4, and P6, and worsening bother is seen by comparing B1 to P1, P3, P4, P5, and P6. For treatment effects, compare P1–P6 to T1–T6. These are plotted as the average for the week over the two sequence groups (placebo/treatment vs. treatment/placebo) regardless of period (treatment first vs. second) since these were not significant in the mixed linear models. For simplicity, placebo is plotted before treatment. Abbreviations: B, baseline; T, treatment; P, placebo.

 
Similarly, in the high-dose study, there was no placebo effect for physiological hot flashes (p < .71), but there was a placebo effect for self-reported hot flashes. Compared with B1, high-dose patients recorded significantly (a) fewer diary hot flashes (P1–P6, p < .05) (Fig. 5Go); (b) fewer hot flash event markers (P1–P6, p < .05) (Fig. 5Go); (c) lower diary hot flash severity at P6 (p= .02) (Fig. 6Go); and (d) lower diary hot flash bother at P6 (p< .05) (Fig. 6Go). Findings for B2 were similar.


Figure 5
View larger version (13K):
[in this window]
[in a new window]

  		Figure 5. High-dose study placebo and treatment effects on hot flash frequency, with fewer monitor flashes at T5 compared with B1. For placebo effects, compare diary and event marker hot flashes at P1–P6 to B1. These are plotted as the average for the week over the two sequence groups (placebo/ treatment vs. treatment/placebo) regardless of period (treatment first vs. second) since these were not significant in the mixed linear models. For simplicity, placebo is plotted before treatment. Abbreviations: B, baseline; T, treatment; P, placebo.

 

Figure 6
View larger version (11K):
[in this window]
[in a new window]

  		Figure 6. High-dose study placebo and treatment effects for hot flash severity, bother, and interference. Placebo effects for diary severity and bother are seen by comparing B1 to P6. Treatment effects are observed by comparing P1–P6 to T1–T6. These are plotted as the average for the week over the two sequence groups (placebo/treatment vs. treatment/placebo) regardless of period (treatment first vs. second) since these were not significant in the mixed linear models. For simplicity, placebo is plotted before treatment. Abbreviations: B, baseline; T, treatment; P, placebo.

 
In addition, at week 14 of each study, 74% of patients could correctly identify the weeks they had received placebo. These data suggested that, by the end of the study, patients were not blinded to placebo.

Treatment Effect for Hot Flashes
Both doses reduced hot flashes (Tables 1Go, 2Go). Compared with placebo, both doses reduced physiological hot flash frequency, self-reported frequency (diary and event marker), and hot flash severity and bother. With the higher dose, hot flash interference was also significantly lower during drug versus placebo. According to adjusted means in Table 1Go, physiological hot flashes decreased by 1.7 hot flashes per day (24 hours) or 22% during low-dose treatment compared with baseline. In addition, diary hot flashes decreased by 1.06 per day during placebo and 2.52 per day during treatment compared with baseline (or 1.46 per day compared with placebo). Similarly, as shown in Table 2Go, physiological hot flashes decreased by 1.03 per day or 14% with treatment compared with baseline, and diary hot flashes decreased by 25% with treatment.


View this table:
[in this window]
[in a new window]

  		Table 1. Efficacy of drug on hot flash and secondary outcomes for low dose

 

View this table:
[in this window]
[in a new window]

  		Table 2. Efficacy of drug on hot flash and secondary outcomes for high dose

 
Venlafaxine’s time to effect varied between doses. In the low-dose study, there were fewer monitor hot flashes at T1 than B1 (p= .003) and at T1 than T2, T3, and T6 (p< .05) (Fig. 3Go). For the high-dose study, there were fewer monitor flashes at T5 than B1 (p = .02) (Fig. 5Go). Results for B2 were similar.

Treatment Effect for Secondary Outcomes
There were no significant treatment effects for secondary outcomes at either dose (Tables 1Go, 2Go). Overall, negative affect, fatigue, sleep quality, and quality of life did not improve with treatment compared with placebo. However, the subgroup of 15 women who experienced a ≥50% decrease in monitor hot flashes showed significant improvement from baseline to T6 in Profile of Mood States Short Form fatigue (p = .007), sleep disturbance (p = .03), MOS vitality (p = .048), and MOS mental health (p = .02). No change in negative affect or other MOS subscales were noted.

Side Effects and Follow-Up
The total number and severity of side effects at each dose were not significantly different from placebo (Tables 1Go, 2Go). However, compared with placebo, low-dose study patients reported significantly more severe constipation, headaches, and dry mouth while on drug compared with placebo (Table 1Go); however, the mean severity for each of these side effects was reported as between not at all (0) to slightly (1) severe. In the high-dose study, women reported significantly less trouble sleeping and more severe constipation and dry mouth on venlafaxine (Table 2Go). Mean severity for each side effect was not at all (0) to slightly (1) severe. In addition, blood pressure was not affected with either treatment dose.

After collapsing across randomization sequences, low-dose patients who chose to continue venlafaxine were 55% at month 1, 36% at month 6, and 42% at month 12 because some patients who stopped taking the drug by month 6 had restarted it by month 12. The top reasons women discontinued venlafaxine at each time point were side effects (21%–22%) and feeling like the medication had stopped working (21%–22%); however, Figure 1Go indicates that many patients were missed in the low-dose follow-up analysis. Similarly, high-dose patients who continued on venlafaxine were 27% at month 1, 20% at month 6, and 23% at month 12. Side effects were again one of the top two reasons women discontinued venlafaxine at each time point (17%–20%); the other reason was feeling too busy to take medications each day (17%–20%).


   DISCUSSION
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Conclusion
 Disclosure of Potential...
 References
 
These trials indicate that, when compared with placebo, venlafaxine modestly reduced physiological and self-reported hot flashes. Both doses reduced physiological and subjective hot flash frequency as well as hot flash severity and bother. Hot flash interference improved only at the 75-mg dose. At the 37.5-mg daily dose, effects on physiological hot flashes were greatest at 1 week of treatment, whereas at the 75-mg daily dose, effects were greatest at 5 weeks of treatment. Among women who achieved a ≥50% reduction in physiological hot flashes, venlafaxine also improved secondary outcomes, including fatigue, sleep quality, vitality, and mental health. Mild side effects were seen at both doses and were one reason women discontinued taking the drug long-term. Thus, although venlafaxine resulted in modest and acute reductions in hot flashes with few side effects, patients may not tolerate the drug long-term.

Consistent with other controlled studies (for reviews, see [12, 42]), we found significant placebo effects with self-reported hot flash measures. In both studies, the number of reported hot flashes decreased during placebo. In addition, in the high-dose study, severity and bother also decreased with placebo. These decreases may represent (a) patients’ perception that the placebo was effective and/or (b) patients becoming less compliant with diaries over time. It is interesting that diary severity and bother actually increased significantly during the low-dose study, although Figure 4Go shows these changes to be less than one point on the 0- to 10-point rating scales. The lack of placebo effects with the physiological monitor suggests greater accuracy in calculating treatment effect sizes with the monitor than self-reports (e.g., diaries and event marker).

Although we found venlafaxine to be significantly more effective than placebo, the change in hot flash frequency was modest. There was a reduction of 1.7 hot flashes at the low dose (22% reduction) and 2.0 at the high dose (14% reduction). Whether or not this change is clinically meaningful is debatable. In this study, secondary outcomes, including fatigue, sleep disturbance, vitality, and mental health, improved only when a ≥50% reduction in physiological hot flashes was achieved. It is possible that a 50% reduction is needed to reach a clinically meaningful effect. Including the physiological monitor in future studies could help to validate this 50% threshold as a standard for evaluating other pharmacological and/or behavioral therapies.

In the low-dose study, the greatest effects were seen after 1 week, consistent with other hot flash studies (for review, see [12]) but not consistent with the antidepressant effects, which can take several weeks. In the high-dose study, the greatest effects were seen after 5 weeks. These results may indicate the low-dose mechanism of action on hot flashes may be different from the antidepressant one. For example, the mechanism by which SSRIs/SNRIs work to alleviate premenstrual dysphoric disorder is through allopregnanolone and {gamma}-aminobutyric acid [43, 44]. However, the mechanism of venlafaxine’s effects on hot flashes remains unclear.

That a 50% or more decrease in physiological hot flashes was associated with improved secondary outcomes supports the existence of a symptom cluster or menopausal syndrome as proposed by others [5, 45–47]. Because secondary outcomes were linked to physiological hot flashes, these improvements are not an artifact of positive reporting bias. In addition, it remains unclear whether venlafaxine had an indirect or direct effect on secondary outcomes. Others suggest treatments improve hot flashes, and this subsequently improves other outcomes [13]. It is also plausible that venlafaxine improved hot flashes and secondary outcomes directly and concurrently. Further research to clarify this relationship is needed.

Consistent with other studies [7, 13], side effects were greater during treatment but mild in severity. Compared with placebo, dry mouth and constipation increased with both doses, but headaches increased only with the low dose. In addition, improvement in the single-item trouble sleeping side-effect question was consistent with the subset data showing improvement in the Pittsburgh Sleep Quality Index.

Despite the low side-effect profile, follow-up assessments indicated side effects were a top reason women discontinued venlafaxine. These results may indicate (a) perceived treatment benefits did not outweigh the mild side effects, (b) side effects were mild but prolonged and thus intolerable, or (c) side effects changed or increased over the months of follow-up. Unfortunately, we did not assess side-effect severity at the 1-, 6-, and 12-month follow-up and therefore are unable to address which possibility might have occurred.

Limitations
Main study limitations were (a) racially and ethnically homogeneous samples, (b) small sample sizes, (c) limited treatment time, and (d) lack of pharmacogenetic data. Because of the lack of sample diversity and lack of a noncancer comparison group, it is unclear whether our findings can be generalized to a broader population. Accrual and retention were low, possibly because of the intense weekly assessment schedule; however, although sample sizes were small, they were large enough to obtain significant findings for efficacy. The studies were underpowered to detect differences in secondary outcomes, and this may be a reason we saw improvement only in those outcomes among the subset that achieved a ≥50% reduction in physiological hot flashes. Treatment time was limited because of the crossover nature of the study. Although treatment effects for depression are typically not seen for 8–10 weeks of treatment [9, 48], a shorter treatment period seemed reasonable given that previous studies had found significant improvement in hot flashes after 4–6 weeks [7, 49, 50]. Finally, genetic polymorphisms that might have helped explain response to venlafaxine, such as allelic variations in serotonin transporter genes [51], were not evaluated. These may have led to subtherapeutic dosing, particularly in the low-dose study.

There are limitations involved in assessing hot flashes. First, we have shown herein that subjective hot flash measures are prone to placebo effects. Because of this, using baseline to calculate a percentage reduction with treatment may overestimate efficacy compared with using placebo to calculate a percentage reduction with treatment. Second, subjective hot flash measures as well as secondary outcome measures may both be subject to a positive reporting bias. This potential reporting bias interferes with the ability to assess whether improvement in physiological hot flashes is more clinically relevant than improvement in subjective hot flashes.


   CONCLUSION
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
Disclosure of Potential...
 References
 
Evidence-based treatments are needed for breast cancer survivors’ frequent, severe, and bothersome hot flashes. Venlafaxine (37.5 mg and 75 mg) significantly decreased physiological hot flashes as well as self-reported hot flash frequency, severity, bother, and interference (75 mg only). Effects were timed differently for each dose, suggesting a mechanism of action other than the antidepressant one. Reductions of 50% or more in physiological hot flashes were associated with improved fatigue, sleep disturbance, vitality, and mental health, but only a small subset of women achieved this degree of relief. Side effects, though mild and expected, were one of the main reasons women discontinued venlafaxine long-term. Findings help clarify the role of venlafaxine in treating hot flashes among breast cancer patients.


   DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Disclosure of Potential...
References
 
R.C.S. has received grant/research support from Eli Lilly and Company, GlaxoSmithKline, Janssen Pharmaceutica, Pfizer Inc., Rhône-Poulenc Rorer Pharmaceutica, Sanofi Pharmaceutica, SmithKline Beecham Pharmaceuticals, Wyeth-Ayerst Laboratories, Astra Zeneca Pharmaceutica, and Abbott Laboratories; has acted as a consultant for Pfizer Inc. and Janssen Pharmaceutica; and is on the speakers bureau for Bristol-Myers Squibb Company, Eli Lilly and Company, Janssen Pharmaceutica, Pfizer Inc., SmithKline Beecham Pharmaceuticals, Solvay Pharmaceuticals, Inc., Wyeth-Ayerst Laboratories, and Abbott Laboratories.


   ACKNOWLEDGMENT
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Disclosure of Potential...
 References
 
This work was supported by the NIH, National Institute of Nursing Research Grant NR01 NR05261 (to J.S. Carpenter, principal investigator). We thank Pam Carney, Gloria Cherry, Heather Cucullu, Bill Sayre, Dr. David Johnson, Dr. Sheila Ridner, Dr. Laurel Brown, and Dr. Liguo Yu.


   REFERENCES
Top
 Learning Objectives
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Conclusion
 Disclosure of Potential...
 References
 

  1. Canney PA, Hatton MQ. The prevalence of menopausal symptoms in patients treated for breast cancer. Clin Oncol (R Coll Radiol) 1994;6:297–299.
  2. Harris PF, Remington PL, Trentham-Dietz A et al. Prevalence and treatment of menopausal symptoms among breast cancer survivors. J Pain Symptom Manage 2002;23:501–509.[CrossRef][Medline]
  3. Carpenter JS. State of the science: Hot flashes and cancer. Part 1: Definition, scope, impact, physiology, and measurement. Oncol Nurs Forum 2005;32:959–968.[CrossRef][Medline]
  4. Carpenter JS. State of the science: Hot flashes and cancer. Part 2: Management and future directions. Oncol Nurs Forum 2005;32:969–978.[CrossRef][Medline]
  5. Gupta P, Sturdee DW, Palin SL et al. Menopausal symptoms in women treated for breast cancer: The prevalence and severity of symptoms and their perceived effects on quality of life. Climacteric 2006;9:49–58.[CrossRef][Medline]
  6. Carpenter JS, Johnson D, Wagner L et al. Hot flashes and related outcomes in breast cancer survivors and matched comparison women. Oncol Nurs Forum 2002;29:E16–E25.[Medline]
  7. Loprinzi CL, Kugler JW, Sloan JA et al. Venlafaxine in management of hot flashes in survivors of breast cancer: A randomised controlled trial. Lancet 2000;356:2059–2063.[CrossRef][Medline]
  8. Stearns V, Beebe KL, Iyengar M et al. Paroxetine controlled release in the treatment of menopausal hot flashes: A randomized controlled trial. JAMA 2003;289:2827–2834.[Abstract/Free Full Text]
  9. Hiemke C, Hartter S. Pharmacokinetics of selective serotonin reuptake inhibitors. Pharmacol Ther 2000;85:11–28.[CrossRef][Medline]
  10. Stearns V, Johnson MD, Rae JM et al. Active tamoxifen metabolite plasma concentrations after coadministration of tamoxifen and the selective serotonin reuptake inhibitor paroxetine. J Natl Cancer Inst 2003;95:1758–1764.[Abstract/Free Full Text]
  11. Jin Y, Desta Z, Stearns V et al. CYP2D6 genotype, antidepressant use, and tamoxifen metabolism during adjuvant breast cancer treatment. J Natl Cancer Inst 2005;97:30–39.[Abstract/Free Full Text]
  12. Nelson HD, Vesco KK, Haney E et al. Nonhormonal therapies for menopausal hot flashes: Systematic review and meta-analysis. JAMA 2006;295: 2057–2071.[Abstract/Free Full Text]
  13. Evans ML, Pritts E, Vittinghoff E et al. Management of postmenopausal hot flushes with venlafaxine hydrochloride: A randomized, controlled trial. Obstet Gynecol 2005;105:161–166.[Abstract/Free Full Text]
  14. Carpenter JS, Monahan PO, Azzouz F. Accuracy of subjective hot flush reports compared with continuous sternal skin conductance monitoring. Obstet Gynecol 2004;104:1322–1326.[Abstract/Free Full Text]
  15. Carpenter JS, Andrykowski MA, Freedman RR et al. Feasibility and psychometrics of an ambulatory hot flash monitoring device. Menopause 1999;6:209–215.[Medline]
  16. Freedman RR. Laboratory and ambulatory monitoring of menopausal hot flashes. Psychophysiology 1989;26:573–579.[Medline]
  17. Savard J, Davidson JR, Ivers H et al. The association between nocturnal hot flashes and sleep in breast cancer survivors. J Pain Symptom Manage 2004; 27:513–522.[CrossRef][Medline]
  18. Thurston RC, Blumenthal JA, Babyak MA et al. Emotional antecedents of hot flashes during daily life. Psychosom Med 2005;67:137–146.[Abstract/Free Full Text]
  19. Pedhazur EJ, Schmelkin L. Measurement, Design, and Analysis: An Integrated Approach. Hillsdale, NJ: Lawrence Erlbaum, 1991.
  20. First MB, Gibbon M, Spitzer RL et al. User’s Guide for the Structured Interview for DSM-IV Axis I Disorders – Research Version. New York: Biometrics Research, 1996.
  21. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry 1960;23:56–62.[Medline]
  22. Williams JB. A structured interview guide for the Hamilton Depression Rating Scale. Arch Gen Psychiatry 1988;45:742–747.[Abstract]
  23. Carpenter JS. Physiological monitor for assessing hot flashes. Clin Nurse Spec 2005;19:8–10.[Medline]
  24. Carpenter JS, Gautam S, Freedman RR et al. Circadian rhythm of objectively recorded hot flashes in postmenopausal breast cancer survivors. Menopause 2001;8:181–188.[CrossRef][Medline]
  25. Carpenter JS, Gilchrist JM, Chen K et al. Hot flashes, core body temperature, and metabolic parameters in breast cancer survivors. Menopause 2004;11:375–381.[CrossRef][Medline]
  26. Carpenter JS. The Hot Flash Related Daily Interference Scale: A tool for assessing the impact of hot flashes on quality of life following breast cancer. J Pain Symptom Manage 2001;22:979–989.[CrossRef][Medline]
  27. McNair DM, Lorr M, Droppelman LF. POMS Manual for the Profile of Mood States. San Diego: Educational and Industrial Testing Service, 1971.
  28. Watson D, Clark LA, Tellegen A. Development and validation of brief measures of positive and negative affect: The PANAS scales. J Pers Soc Psychol 1988;54:1063–1070.[CrossRef][Medline]
  29. Radloff IS. The CES-D Scale: A self-report depression scale for research in the general population. Appl Psych Meas 1977;1:385–401.[CrossRef]
  30. Watson D, Clark LA. Negative affectivity: The disposition to experience aversive emotional states. Psychol Bull 1984;96:465–490.[CrossRef][Medline]
  31. Watson D, Clark LA, Carey G. Positive and negative affectivity and their relation to anxiety and depressive disorders. J Abnorm Psychol 1988;97: 346–353.[CrossRef][Medline]
  32. Curran SL, Andrykowski M, Studts JL. Short form of the profile of mood states (POMS-SF): Psychometric information. Psychol Assess 1995;7:80–83.
  33. Meek PM, Nail LM, Barsevick A et al. Psychometric testing of fatigue instruments for use with cancer patients. Nurs Res 2000;49:181–190.[CrossRef][Medline]
  34. Carpenter JS, Andrykowski MA. Psychometric evaluation of the Pittsburgh Sleep Quality Index. J Psychosom Res 1998;45:5–13.[CrossRef][Medline]
  35. Buysse DJ, Reynolds CF 3rd, Monk TH et al. The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Res 1989;28:193–213.[CrossRef][Medline]
  36. Berger AM, Parker KP, Young-McCaughan S et al. Sleep wake disturbances in people with cancer and their caregivers: State of the science. Oncol Nurs Forum 2005;32:E98–E126.[CrossRef][Medline]
  37. Ware JE Jr, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992;30: 473–483.[Medline]
  38. Ware JE, Snow KK, Kosinski M et al. SF-36 Health Survey: Manual and Interpretation Guide. Boston: The Health Institution New England Medical Center, 1993.
  39. Carpenter JS, Andrykowski MA. Menopausal symptoms in breast cancer survivors. Oncol Nurs Forum 1999;26:1311–1317.[Medline]
  40. Carpenter JS, Andrykowski MA, Cordova MJ et al. Do participants’ reports of symptom prevalence or severity vary by interviewer gender? Nurs Res 1999;48:276–279.[CrossRef][Medline]
  41. Senn S. Cross-over Trials in Clinical Research. 2nd ed. Chichester, U.K.: John Wiley & Sons, Inc., 2002.
  42. Sloan JA, Loprinzi CL, Novotny PJ et al. Methodologic lessons learned from hot flash studies. J Clin Oncol 2001;19:4280–4290.[Abstract/Free Full Text]
  43. Guidotti A, Costa E. Can the antidysphoric and anxiolytic profiles of selective serotonin reuptake inhibitors be related to their ability to increase brain 3{alpha}, 5{alpha}-tetrahydroprogesterone (allopregnanolone) availability? Biol Psychol 1998;44:865–873.
  44. Griffin LD, Mellon SH. Selective serotonin reuptake inhibitors directly alter activity of neurosteroidogenic enzymes. Proc Natl Acad Sci U S A 1999; 96:13512–13517.[Abstract/Free Full Text]
  45. Beck SL, Dudley WN, Barsevick A. Pain, sleep disturbance, and fatigue in patients with cancer: Using a mediation model to test a symptom cluster. Oncol Nurs Forum 2005;32:E48–E55.[CrossRef]
  46. National Institutes of Health State-of-the-Science Conference statement: Management of menopause-related symptoms. Ann Intern Med 2005;142: 1003–1013.[Medline]
  47. Dodd MJ, Miaskowski C, Paul SM. Symptom clusters and their effect on the functional status of patients with cancer. Oncol Nurs Forum 2001;28: 465–470.[Medline]
  48. Beique JC, de Montigny C, Blier P et al. Venlafaxine: Discrepancy between in vivo 5-HT and NE reuptake blockade and affinity for reuptake sites. Synapse 1999;32:198–211.[CrossRef][Medline]
  49. Loprinzi CL, Pisansky TM, Fonseca R et al. Pilot evaluation of venlafaxine hydrochloride for the therapy of hot flashes in cancer survivors. J Clin Oncol 1998;16:2377–2381.[Abstract]
  50. Quella SK, Loprinzi CL, Sloan J et al. Pilot evaluation of venlafaxine for the treatment of hot flashes in men undergoing androgen ablation therapy for prostate cancer. J Urol 1999;162:98–102.[CrossRef][Medline]
  51. Pollock BG, Ferrell RE, Mulsant BH et al. Allelic variation in the serotonin transporter promoter affects onset of paroxetine treatment response in late-life depression. Neuropsychopharmacology 2000;23:587–590.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 JCOHome page
J. S. Carpenter and D. Flockhart
Flash Points
J. Clin. Oncol., December 10, 2007; 25(35): 5546 - 5547.
[Full Text] [PDF]

Home page
 JCOHome page
G. Deng, A. J. Vickers, K. S. Yeung, G. M. D'Andrea, H. Xiao, A. S. Heerdt, S. Sugarman, T. Troso-Sandoval, A. D. Seidman, C. A. Hudis, et al.
Randomized, Controlled Trial of Acupuncture for the Treatment of Hot Flashes in Breast Cancer Patients
J. Clin. Oncol., December 10, 2007; 25(35): 5584 - 5590.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow CME: Take the course for this article:
 Randomized, Double-Blind, Placebo-Controlled Crossover Trials of Venlafaxin...
Right arrow eLetters: Submit a response to this article
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article link to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Reprints/Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Carpenter, J. S.
Right arrow Articles by Shelton, R. C.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Carpenter, J. S.
Right arrow Articles by Shelton, R. C.

HOME HELP CONTACT US SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
THE ONCOLOGIST STEM CELLS CME ALPHAMED PRESS JOURNALS