Venlafaxine is superior to clonidine as treatment of hot flashes in breast cancer patients a double-blind, randomized study

Transcription

1 Annals of Oncology 18: , 2007 doi: /annonc/mdl478 Published online 17 January 2007 is superior to as treatment of hot flashes in breast cancer patients a double-blind, randomized study S. Loibl 1 *, K. Schwedler 1, G. von Minckwitz 1,2, R. Strohmeier 1, K. M. Mehta 2 & M. Kaufmann 1 1 Department of Obstetrics and Gynecology, University Hospital Frankfurt am Main, Frankfurt am Main, Germany; 2 German Breast Group, Schleussnerstr. 42, Neu-Isenburg, Germany Received 16 October 2006; revised 22 November 2006; accepted 30 November 2006 Background: Classical hormone replacement therapy for hot flashes is contraindicated in breast cancer especially in endocrine responsive disease. Patients and methods: In a double-blind, randomized phase III study, breast cancer patients suffering from hot flashes at least twice a day, who were not taking any medication against hypertension and depression received either mg twice a day or 37.5 mg twice a day for 4 weeks. The primary end point was defined as the frequency of hot flashes after 4 weeks of treatment. A self-reported 1-week hot flash and other symptom questionnaire were kept before the start of treatment until the end of treatment course. Results: From April 2002 to October 2004, 80 patients were recruited of whom 64 were assessable for efficacy analyses. Thirty-three received and 31, nine patients stopped early because of side-effects and seven withdrew consent. At the end of treatment week 4, the median hot flash frequency dropped by 7.6 hot flashes per day for patients receiving and 4.85 hot flashes per day for those receiving (P = 0.025). Conclusion: is significantly more effective in reducing the frequency of hot flashes in breast cancer patients than. Key words: breast cancer,, hot flashes, introduction Hot flashes can significantly impair quality of life in postmenopausal patients, especially when estrogens are contraindicated or rejected. About two-thirds of postmenopausal breast cancer patients report hot flashes [1]. Several reasons support the theory that hot flashes are more troublesome in breast cancer patients than in others [2]. Premenopausal women treated for breast cancer with chemotherapy or ovarian ablative procedures undergo premature menopause and suffer as well from hot flashes and other climacteric symptoms [3, 4]. Postmenopausal women treated with tamoxifen or more recently with aromatase inhibitors (ARIs) suffer from hot flashes [5]. The sudden interruption of the hormone replacement therapy (HRT) after the diagnosis of breast cancer causes a rebound phenomenon with even more severe hot flashes [6]. HRT is the most effective treatment of treating hot flashes. It is, however, not recommended in breast cancer patients due to an increased risk of relapse [7, 8]. Especially, for tamoxifen-treated patients HRT might not be useful [9]. Therefore, other nonhormonal agents have been investigated to reduce hot flashes in cancer patients [10]. *Correspondence to: Dr S. Loibl, J.W. Goethe-University Hospital, Department of Obstetrics and Gynecology, Theodor-Stern-Kai 7, Frankfurt am Main, Germany. Tel: ; Fax: ; loibl@em.uni-frankfurt.de, a centrally active a-adrenergic agonist, that reduces vascular reactivity, and which is widely used as a nonestrogenic treatment of hot flashes in breast cancer patients does decrease hot flashes to a moderate degree but it is also accompanied by several toxic effects [11]. Oral and transdermal formulations have both been tested [12]. It has been reported that will reduce hot flashes by up to 46% in postmenopausal women [13]. is an antidepressant that inhibits neuronal serotonin and norepinephrine reuptake, a so-called selective serotonin norepinephrine reuptake inhibitor (SNRI). Recommended doses for treating depressions range from 75 to 375 mg/d. It has been demonstrated in several trials that and other SNRI or selective serotonin-reuptake inhibitors (SSRI) such as fluoxetine or paroxetine are effective in reducing hot flashes in cancer patients by at least 50% [14 16]. This double-blind, randomized clinical trial was conducted to compare to another nonhormonal agent in the treatment of hot flashes of breast cancer patients. patients and methods patients Only women 18 years of age or older with primary breast cancer were considered eligible for this clinical trial. They must have had bothersome hot flashes at least 14 times a week or must have been seeking help for hot original article ª 2007 European Society for Medical Oncology

2 Annals of Oncology flashes. The hot flashes had to be present for at least 4 weeks before study entry. A predefined menopausal status was not required. Tamoxifen, gonadotropine-releasing agonists or ARIs, were allowed as long as the patients had been on such therapy for at least a month and were planning to continue the therapy for the duration of the study. Eastern Cooperative Oncology Group performance status had to be zero to one. All patients had to give written informed consent approved by the local regulatory board. Previous treatment with and as well as estrogens, progestogens, or androgens for hot flashes was an exclusion criterion. Further exclusion criteria were current treatment with hypertensive or antidepressant agents, other nonhormonal agents for hot flashes such as black cohosh, isoflavone, and vitamine E. Patients with hypertension or hypotension, peripheral or cardiovascular diseases, or symptomatic cardiac diseases were excluded. Metastatic disease was an exclusion criterion. n=40 n=4 toxicities n=2 consent withdrawn n=1 incomplete diary n=14 Screened n=121 Randomized n=80 Cross-over n=39 n=15 n=40 n=6 toxicities n=3 consent withdrawn treatment Eligible patients were stratified according to age ( 50 years versus >50 years) and adjuvant therapy (endocrine versus no endocrine therapy) and were randomized in a double-blinded fashion to receive twice a day either mg or 37.5 mg as tablets for 4 weeks. Stratified randomization lists were prepared using Clinstat. The sealed lists were kept at the study pharmacist who allocated the treatment and prepared the medication. After 4 weeks of treatment, patients were planned to crossover to the other treatment option. The crossover part of the trial was abandoned because 26% of the patients completing the trial at that time were not assessable due to incomplete evaluation forms, either because of treatment discontinuation or poor reporting. The simplification of the trial increased the percentage of patients completing the treatment and evaluation forms (Figure 1). evaluation For better comparison, the methodology and the questionnaire was adapted with permission from the North Central Cancer Treatment Group [17]. A self-reported 1-week hot flash and other symptom questionnaire were kept 1 week before the start of treatment until the end of the treatment course. The frequency of hot flashes as well as the severity of the hot flashes had to be noted daily. To evaluate for toxicity/symptoms the patients were asked to note whether they were having one of the following symptoms during each study week: loss of appetite, mouth dryness, nausea, tiredness, constipation, restless sleep, nervousness, sweating, dizziness, mood disorder, diarrhea, sleeplessness, or other symptoms. The symptoms/toxic effects were not graded. statistical considerations The primary end point of the study was the hot flash frequency (number of hot flashes reported per week) in both treatment groups after 4 weeks of treatment. Secondary end points included the hot flash severity score which was adapted from the North Central Cancer Treatment Group by scoring each individual hot flash from one point for a mild hot flash to four points for a very severe hot flash and then multiplying the severity of each hot flash with the frequency and summing it up (i.e. one mild hot flash, three severe hot flashes, two very severe hot flashes a day results in a score of 18). Another secondary objective was the incidence of toxic effects reported at baseline throughout the treatment period. These toxicity data were analyzed with a v 2 test. A (two-sided) of <0.05 was considered statistically significant. Eighty percent of the data of each patient had to be available for the patient to be eligible for evaluation. Missing scattered data in otherwise complete diaries were imputed by taking the average of the day before and the day after or a similar rule for two missing days or a missing day at the end of a sequence. The analysis of the crossover part is done in a descriptive fashion. evaluable n=33 A sample size of 34 in each group would have an 80% power to detect a difference of 20% between the two treatment arms assuming that the common standard deviation (SD) is 0.65 using a two-group t-test with 0.05 one-sided significance level [18]. results evaluable n=31 Figure 1. Study population and design. The crossover part was stopped after 39 patients had been included. Thereafter, patients were simply randomized and did not crossover. From April 2002 to October 2004, 80 patients with primary breast cancer had been included into the trial in the University Hospital Frankfurt, Germany. Sixty-four of them were assessable for efficacy and toxicity. One patient did not fill in the baseline evaluation for hot flashes, 10 patients stopped treatment because of side-effects (four in the group and six in the group). Five patients withdrew consent and did not send back the evaluation form. Sixty-three patients were assessable for the score because one patient reported only the total number of hot flashes. Twenty-nine (74%) of the 39 patients entering the crossover part were assessable for the crossover part with complete questionnaires for both therapies and baseline. Fourteen patients started with and crossed over to and 15 patients vice versa. The baseline characteristics and hot flashes are well balanced between the two groups (Table 1 and 2). Most of the patients (n = 55) had been concurrently taking an endocrine treatment of breast cancer tamoxifen (n = 42), tamoxifen and goserelin (n = 6), an ARI (n = 5), and an ARI with goserelin (n = 2). efficacy The primary end point of hot flash frequency demonstrated that decreased the frequency of hot flashes significantly compared with baseline (100%) by 57% [standard error (SE) 7.4]. decreased the frequency by 37% (SE 4.7). The median hot flash frequency dropped by 7.6 hot flashes per day for patients receiving and 4.85 hot flashes per day 690 Loibl et al. Volume 18 No. 4 April 2007

3 Annals of Oncology original article Table 1. Patient characteristics at baseline 100 for those receiving. Changes in hot flash frequency favor significantly (P = 0.025) (Figure 2). A 75% drop in hot flash frequency was reported by 29% (n = 9) of the patients in the group compared with 12% (n = 4) in the group. An increase in hot flashes was reported by 10% (n = 3) in the group and by 6% (n =2)in the group (Table 3). A complete reduction of hot flashes was only reported in the group (n = 6). The hot flash score, a secondary end point, decreased by 11.4 units per day [39% reduction (SE) 5.4] in the group and by 8.9 units per day [57% (SE) 6.2] in the group (P = 0.043) (Figure 3). Patients who started with and crossed over to had a less effective reduction in hot flash frequency after the 8 weeks of treatment (63% reduction from baseline) than the patients who started on first and crossed over to (42% reduction from baseline). toxicity The patients were asked to report weekly to evaluate toxicity. In both therapy groups the most common toxic effects were mouth dryness, tiredness, and restless sleep (Tables 4 and 5). Nausea occurred in 39% of the patients in the group in the first treatment week and decreased thereafter to 19%. Nausea was significantly more frequently reported in the group compared with the group (P = 0.05). Mouth dryness, constipation, and restless sleep were more commonly reported by patients treated with but were not significantly different (Table 5). The frequency of patients reporting nervousness or moodiness was significantly less while taking compared with baseline (P = and 0.05, respectively). discussion (n = 33) n (%) (n = 31) n (%) Age (39) 12 (39) >50 20 (61) 19 (61) Current antihormone therapy 0.99 Yes 27 (82) 28 (90) Tamoxifen ARI 4 3 No 6 (18) 3 (10) ARI, aromatase inhibitors. The results confirm earlier trials where both drugs, and, have been studied in randomized fashions and proved to be superior to placebo [12, 14]. Here they are compared for the first time head to head in breast cancer patients. Despite the assessable sample size being less than anticipated, the results are nevertheless statistically significant, because the SD (0.33) is less than assumed and the SE small. Hot flashes in % Baseline Time in weeks Table 2. Baseline week: hot flash frequency, scoring, and severity n =33 P=0.025 Figure 2. Median hot flash frequency per day from baseline to treatment week 4 for the first part of the trial (n = 64). There was a 57% reduction with compared with 37% with (P = 0.025; two-sided t-test). Dotted lines represent and solid lines. n =31 Daily hot flash frequency 0.42 Mean Median Range Daily hot flash score 0.77 Mean Median Range Daily hot flash severity 0.78 Mean Median Range is one of the most extensively studied SSRI/SNRI for the treatment of hot flashes in cancer patients and was therefore chosen in our trial as a comparator to which is considered today to be the standard nonhormonal agent in Germany. Two trials with have demonstrated its effect in male and female cancer patients suffering from hormone deprivation. The pilot trial revealed a 50% reduction in hot flashes [19]. A subsequent dose-evaluating placebocontrolled trial demonstrated 75 mg/d of to be the optimum for treating hot flashes leading to a median hot flash score reduction of 61% and reduction in frequency of 46% which compares favorably to our results [14]. in a dosage of 0.1 mg/d seems effective against hot flashes with modest effect [12]. This trial was carried out with 0.15 mg because of the ready availability of the drug. It has not yet been ruled out, whether the effect of is dependent on the dosage. Ginsburg et al. [20] did postulate that there will be no difference between 0.1 versus 0.15 mg whereas Laufer et al. [13] could demonstrate a dose-dependent effect of Volume 18 No. 4 April 2007 doi: /annonc/mdl

4 Hot flashes in % Baseline Time in weeks P=0.043 Figure 3. Median hot flash severity score per day from baseline to treatment week 4 for the first part of the trial. (n = 63). There is a significant difference in the severity score between the two groups (P = 0.043; two-sided t-test). Dotted lines represent and solid lines. Table 3. Basic hot flash daily frequency change from baseline to the end of therapy % of baseline after treatment week 4 %(n = 33) %(n = 31) (4) 29 (9) (8) 26 (8) (11) 19 (6) (8) 16 (5) >100 6 (2) 10 (3) from 0.1 to 0.2, and 0.4 mg per day. Instead of using extended release 75 mg per day, we used 37.5 mg twice a day to provide two tablets with treatment in both groups instead of a placebo in the group. The effects of hormonal usage against hot flashes may take several weeks to manifest themselves whereas the effects of, as reported in earlier trials, occurred during the first week. Therefore, it seemed justifiable to use a 4-week treatment as former investigators did, despite hot flashes being a long-term problem. Due to the short half-life of the drugs, a wash out period is not necessary and was not incorporated in our as well as in previous trials by other groups. This trial as well as the others might have been strengthened if the follow-up period after stopping the treatment had been continued to gain more information on the hot flashes. The only long-term trial investigating the efficacy of citalopram and fluoxetine (two antidepressants) in a 9-month placebo-controlled trial did not find any difference between the three treatment arms, even on increasing the dose after 6 months [21]. This might be explained by the different characteristics and severity of menopausal symptoms in cancer and noncancer patients [22]. Data regarding the exact menopausal status were not collected, because all patients had climacteric symptoms due to various reasons. The evaluation of the hot flash frequency demonstrated that the patients older than 50 years of age responded better to than patients 50 years of age or Table 4. Symptoms per patients and therapy at baseline Symptoms n = 33 (%) n = 31 (%) Loss of appetite 1 (3.0) 1 (3.2) 0.95 Sleeplessness 9 (27.3) 8 (25.8) 0.96 Nausea 0 (0) 0 (0) 1 Drowsiness 4 (12.1) 4 (12.9) 0.99 Tiredness 7 (21.1) 8 (25.8) 0.86 Sweating 9 (27.3) 9 (29.0) 0.95 Constipation 1 (3.0) 2 (6.5) 0.25 Restless sleep 11 (33.3) 11 (35.5) 0.79 Nervousness 6 (18.1) 5 (16.1) 0.77 Moodiness 7 (21.1) 6 (19.4) 0.81 Mouth dryness 5 (15.2) 6 (19.4) 0.52 Table 5. Symptoms after 4 weeks of treatment in each treatment group Symptoms n = 33 (%) n = 31 (%) Annals of Oncology Loss of appetite 1 (3.0) 1 (3.2) >0.9 Sleeplessness 10 (30.3) 9 (29.0) >0.9 Nausea 2 (6.1) 6 (19.4) 0.05 Drowsiness 6 (19.4) 5 (16.1) >0.9 Tiredness 14 (42.4) 11 (35.5) >0.9 Sweating 11 (33.3) 10 (32.3) >0.9 Constipation 10 (30.3) 5 (16.1) 0.08 Restless sleep 17 (51.5) 11 (35.5) Nervousness 5 (15.1) 2 (6.5) >0.1 Moodiness 5 (15.1) 3 (9.7) >0.1 Mouth dryness 17 (51.1) 11 (35.5) younger. In earlier trials with SSRI/SNRI, an influence of age on the efficacy could not be demonstrated [15]. Numbers in the two groups were, however, small and the effect should not be overestimated. The trial was not limited to tamoxifen-induced hot flashes. Other groups did not report on any differences on the basis of whether patients were receiving tamoxifen or not. Most of the patients were taking an endocrine treatment, either tamoxifen or an ARI. Due to the small number of nonendocrine-treated patients, a separate analysis was not possible. One week of baseline evaluation and 4 weeks of treatment was chosen as in the earlier carried out trials [17]. Besides vasomotor symptoms, hot flashes are accompanied by a variety of other symptoms such as sleeplessness, tiredness, restless sleep, moodiness, or mouth dryness. The latter is a common complaint in menopausal women because the composition of saliva in the pre- and postmenopausal women is estrogen dependent [23]. Both therapies resulted in various side-effects although seems to be more efficacious than. Nevertheless, the same number of patients wished to continue with and. More patients, however, stopped treatment with at the beginning of the therapy mainly due to nausea. The preference of a therapy to reduce hot flashes depends not only on the efficacy but also on 692 Loibl et al. Volume 18 No. 4 April 2007

5 Annals of Oncology the side-effects. Despite experiencing side-effects while being treated with, other symptoms beside hot flashes, e.g. nervousness and moodiness, improved. These positive effects might have influenced the decision to continue the treatment besides the reduction of hot flashes and hot flash score. The ideal scenario would be a complete reduction in hot flashes; however, nonhormonal agents do manage to reduce them by up to 60% [10]. Therefore, the side-effects experienced are of the utmost importance when choosing the correct treatment. To avoid treatment discontinuation, is recommended to be introduced in a dosage of 37.5 mg per day in the first week and then increased to 75 mg extended release once a day. The trial has several limitations. A larger sample size might have revealed an even more profound difference in frequency and severity of the hot flashes. Subgroups such as pre- or postmenopausal patients could have been investigated in more depth. The grading of the toxic effects could have given more detailed information. A longer follow-up might have given information about treatment effect and side-effects even after stopping the treatment. and other SSRI/SNRI such as paroxetine are regarded as the most promising nonhormonal agents for treating hot flashes in cancer and noncancer patients [19, 24]. Some concerns, however, have been raised because it could be demonstrated that, when co-administered with tamoxifen, paroxetine reduces the active metabolite endoxifen by inhibiting cytochrome P450 2D6 (CYP2D6) [25]. Another study found that to be a class effect but is the less potent CYP2D6 inhibitor, which might have, according to a new study, no effect on CYP2D6 [26, 27]. This is the first report comparing a SSRRI/SNRI antidepressant with for the treatment of hot flashes in breast cancer patients. The trial indicates that is more effective in terms of reducing the frequency and severity of hot flashes compared with. acknowledgements This paper was presented in part at the 41st annual meeting of the American Society of Clinical Oncology, 13 17th May 2005, Orlando, FL. references 1. Carpenter JS, Andrykowski MA, Cordova M et al. Hot flashes in postmenopausal women treated for breast carcinoma: prevalence, severity, correlates, management, and relation to quality of life. Cancer 1998; 82: Couzi RJ, Helzlsouer KJ, Fetting JH. Prevalence of menopausal symptoms among women with a history of breast cancer and attitudes toward estrogen replacement therapy. J Clin Oncol 1995; 13: Rose DP, Davis TE. Ovarian function in patients receiving adjuvant chemotherapy for breast cancer. Lancet 1977; 1: de Haes H, Olschewski M, Kaufmann M et al. Zoladex Early Breast Cancer Research Association Trialists Group. Quality of life in goserelin-treated versus cyclophosphamide + methotrexate + fluorouracil-treated premenopausal and perimenopausal patients with node-positive, early breast cancer: the Zoladex Early Breast Cancer Research Association Trialists Group. J Clin Oncol 2003; 21: original article 5. Fallowfield L, Cella D, Cuzick J, et al. Quality of life of postmenopausal women in the Arimidex, Tamoxifen, Alone or in Combination (ATAC) Adjuvant Breast Cancer Trial. J Clin Oncol 2004; 22: Jensen J, Christiansen C. Dose response and withdrawal effects on climacteric symptoms after hormone replacement therapy. A placebo-controlled therapeutic trial. Maturitas 1983; 5: Rossouw JE, Anderson GL, Prentice RL et al. Risk and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women s Health Initiative randomized controlled trial. JAMA 2002; 288: Holmberg L, Anderson H. HABITS: (hormone replacement therapy after breast cancer is it safe?), a randomised comparison: trial stopped. Lancet 2003; 363: Sestak I, Kealy R, Edwards et al. Influence of hormone replacement therapy on tamoxifen-induced vasomotor symptoms. J Clin Oncol 2006; 24: Stearns V, Ullmer L, Lopez JF et al. Hot flushes. Lancet 2001; 360: Goldberg RM, Loprinzi CL, O Fallon JR et al. Transdermal for ameliorating tamoxifen-induced hot flashes. J Clin Oncol 1994; 12: Pandya KJ, Raubertas RF, Flynn PJ et al. Oral in postmenopausal patients with breast cancer experiencing tamoxifen induced hot flashes: a university of Rochester Cancer Center Community Clinical Oncology Program Study. Ann Intern Med 2000; 132: Laufer LR, Erlik Y, Meldrum DR et al. Effect of on hot flashes in postmenopausal women. Obstet Gynecol 1982; 60: Loprinzi CL, Kugler JW, Sloan JA et al. in management of hot flashes in survivors of breast cancer: a randomized controlled trial. Lancet 2000; 356: Loprinzi CL, Sloan JA, Perez EA et al. Phase III evaluation of fluoxetine for treatment of hot flashes. J Clin Oncol 2002; 20: Stearns V, Slack R, Greep N et al. Paroxetine is an effective treatment for hot flashes: results from a prospective randomized clinical trial. J Clin Oncol 2005; 23: Sloan JA, Loprinzi CL, Novotny PJ et al. Methodologic lessons learned form hot flash studies. J Clin Oncol 2001; 19: O Brien RG, Muller KE. Applied Analysis of Variance on Behavioural Sciences. New York, NY: Marcel Dekker 1983; Loprinzi CL, Pisansky TM, Fonseca R et al. Pilot evaluation of hydrochloride for the therapy of hot flashes in cancer survivors. J Clin Oncol 1998; 16: Ginsburg J, O Reilly B, Swinhoe J. Effect of oral on human cardiovascular responsiveness: a possible explanation of the therapeutic action of the drug in menopausal flushing and migraine. Br J Obstet Gynaecol 1985; 92: Suvanto-Luukkonen E, Koivunen R, Sundstrom H et al. Citalopram and fluoxetine in the treatment of postmenopausal symptoms: a prospective, randomized, 9- month, placebo-controlled, double-blind study. Menopause 2005; 12: Tchen N, Mar Fan H, Yi Q et al. Do women with primary breast cancer who experience chemotherapy-induced menopause have worse symptoms than those due to natural menopause? Proc Am Soc Clin Oncol 2005 (Abstr 8016). 23. Asplund R and Aberg E. Oral dryness, nocturia and the menopause. Maturitas 2005; 50: Stearns V, Beebe KL, Iyenger M et al. Paroxetine controlled release in the treatment of menopausal hot flash: a randomized controlled trial. JAMA 2003; 289: 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: 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: Borges S, Desta Z. Quantitative effect of CYP2D6 genotype and inhibitors on tamoxifen metabolism: implication for optimizing of breast cancer treatment. Clin Pharmacol Ther 2006; 80: Volume 18 No. 4 April 2007 doi: /annonc/mdl