- Email: [email protected]
Worry and risk perception of breast cancer in a prevention trial of low dose tamoxifen in midlife postmenopausal hormone users
The Breast 34 (2017) 108e114
Contents lists available at ScienceDirect
The Breast journal homepage: www.elsevier.com/brst
Original article
Worry and risk perception of breast cancer in a prevention trial of low dose tamoxifen in midlife postmenopausal hormone users Gabriella Rondanina a, Matteo Puntoni b, Aliana Guerrieri-Gonzaga c, Domenico Marra d, Bernardo Bonanni c, Andrea DeCensi d, e, * a
Psychology Center Piccapietra, Genoa, Italy Office of the Scientific Director, E.O. Ospedali Galliera, Genoa, Italy Division of Cancer Prevention and Genetics, European Institute of Oncology, Milan, Italy d Division of Medical Oncology, E.O. Ospedali Galliera, Genoa, Italy e Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, UK b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 March 2017 Received in revised form 28 April 2017 Accepted 16 May 2017
Objective: There is increasing interest in combining postmenopausal hormone therapy (HT) and SERMs in midlife women. We previously showed that refusal to participate in a prevention trial of low dose tamoxifen in HT users was associated with higher worry about breast cancer. Given this counterintuitive finding, we studied which factors influenced worry and risk perception of breast cancer. Methods: We assessed the relationships of breast cancer worry and risk perception with age, age at menopause, Gail risk, education, adherence to mammographic screening, BMI, smoking, physical activity, alcohol use, anxiety and depression in 457 midlife HT users who were eligible to participate in the trial. Results: Women with menopause <48 years were more worried about breast cancer than women with menopause >52 years (OR ¼ 5.0, 95% CI, 1.2e21.1). Worry was also associated with high absolute risk perception and former smoking. Factors associated with higher risk perception were age>60 years, atrisk life style, worry about breast cancer and depression. Conclusions: The inverse association between early menopause and worry about breast cancer is in contrast with the known protective effect of early menopause on breast cancer risk and seems to reflect a feeling of aging and disease vulnerability. Our findings indicate that worry about cancer has an affective construct which is independent of breast cancer biology but is engaged in health decision making. Increasing breast cancer risk awareness in subjects high in worry without a plan of emotional coping may therefore be counterproductive because of avoidant attitudes. © 2017 Published by Elsevier Ltd.
Keywords: Cancer worry Cancer risk perception Breast cancer Tamoxifen Chemoprevention Postmenopausal hormone therapy
1. Introduction Postmenopausal hormone therapy (HT) at the onset of menopause is an example of the paradigm shift towards a proactive behavior of health promotion inasmuch as users are required to possess a good sense of awareness and ability to carefully assess risks and benefits [1]. This is true also for the use of tamoxifen in primary prevention since the trade-off between risks and benefits is subject to continuous debate [2e4]. However, the decisionmaking process about participating in a breast cancer therapeutic
* Corresponding author. Division of Medical Oncology, E.O. Ospedali Galliera, Mura delle Cappuccine 14, 16128, Genoa, Italy. E-mail address: [email protected] (A. DeCensi). http://dx.doi.org/10.1016/j.breast.2017.05.008 0960-9776/© 2017 Published by Elsevier Ltd.
prevention trial is poorly understood. Despite the highest evidence of efficacy ensuing from large trials, the uptake of selective estrogen receptor modulators such as tamoxifen and raloxifene [5] or aromatase inhibitors [6e8] has so far been very limited in clinical practice [9,10], mainly because of the risk of endometrial cancer and venous thromboembolism, partly attenuating the 50% reduction of breast cancer risk observed in the NSABP-P1 trial [11]. Likewise, the use of HT has dramatically dropped [12] after the initial publication of the WHI trial [13], notwithstanding the evidence of a timing effect with a mortality reduction in women aged 50 to 59 in subsequent analyses [1,14]. While initial reports linked a decrease in breast cancer incidence to the HT drop after the WHI trial results [15], most recent evidence on the association between breast cancer incidence and the rate of MHT use during the past years in Europe suggests in fact that breast cancer incidence in
G. Rondanina et al. / The Breast 34 (2017) 108e114
women aged 45e64 has not decreased in parallel with the HT drop, indicating the role of additional confounding factors [16]. While attention has mainly been focused on medical variables discouraging the use of tamoxifen for prevention, including concerns about adverse events [17] and lack of demonstrated mortality reduction [2,3], other studies have pointed to risk perception, worry about breast cancer and lived experiences as factors affecting the decision to participate in a preventive therapy trial for breast cancer [18e20]. This is not surprising given the notion that adherence to breast cancer screening programs is influenced by sociodemographic as well as emotional factors [21]. We pioneered the concept of combining postmenopausal HT and low dose tamoxifen to retain the benefits while reducing the risks of either agent in a phase III trial (HOT trial) in 1884 midlife women [22,23]. The results showed a non-significant 20% reduction of breast cancer in the tamoxifen arm compared with the placebo arm [23]. Importantly, tamoxifen showed favorable trends in women on HT for <5 years (adjusted RR ¼ 0.35; 95% CI 0.15e0.85) and in estrogen-alone users (RR ¼ 0.26, 95% CI 0.05e1.28), but no effect in the combined estro-progestin subgroup. Combining HT to tamoxifen was also associated with a lower incidence of menopausal symptoms relative to tamoxifen alone [24]. More recently, the combination of conjugated estrogen with the SERM bazedoxifene has proven to be beneficial on menopausal quality of life [25] without an increase in mammographic density [26]. Participation in the HOT trial was higher in women at younger age and in those satisfied with health care providers, whereas it was lower in women with breast cancer worry, suggesting a condition of psychological well-being as a promoting factor for the trial participation [27]. Given the counterintuitive finding and the discrepancy with some prior studies on the role of breast cancer worry in adopting prevention strategies, here we searched in depth which factors explained the worry about breast cancer and the risk perception of breast cancer. These factors may be important to better explain health promotion attitudes and therefore enhance compliance to therapeutic prevention in midlife postmenopausal women.
109
available to address initial questions and to fix an appointment with the study personnel, mostly physicians. Counseling about the trial included an evaluation of benefits and risks of HT and tamoxifen given alone, and description of the study procedures. Participant-physician communication was based on a scripted protocol. After counseling and subsequent decision making about the trial, all women received by mail a preprinted, self-report questionnaire based on previous research on medical decision making in the area of breast cancer prevention [29]. 2.3. Psychological factors Worry about developing breast cancer was the study's main outcome measure and was assessed by the following question: “How worried are you about getting breast cancer in the next 5 years?” Responses were graded using a Likert scale: “not at all worried”, “slightly worried”, “somewhat worried”, “worried” and “very worried.” Risk perception of breast cancer were assessed as previously described [19]. Perception of absolute breast cancer risk was assessed using standard verbal and numerical measures within the span of 5 years. For the verbal measures, women were asked, “How likely are you to get breast cancer in the next 5 years?”.Responses were “very unlikely”,“unlikely”,“50/50 chance”, “likely” and “very likely.” For the numerical measures, women were asked, “On a scale from 0 to 100, with 0 indicating certain not to happen and 100 indicating certain to happen, how likely are you to get breast cancer in the next 5 years?”. Perceptions of comparative breast cancer risk were assessed through the following question: “Compared with other women of your age, what are the odds that you will get breast cancer in the next 5 years?”. Responses were “much below average”, “below average”, “same average risk as other women my age”, “above average”, and “much above average”. Although the level of breast cancer risk knowledge was not formally measured, the results of the Gail model and the risk associated with HT use were illustrated to the screened women at face to face or telephone counseling. Anxiety and depression were assessed using the Hospital Anxiety and Depression Scale [30].
2. Materials and methods 2.4. Sociodemographic, health-related and lifestyle variables 2.1. Participants Participants in the psychosocial study were postmenopausal healthy women currently on or about to start HT for the treatment of menopausal symptoms or health promotion who were eligible for a double-blind, placebo-controlled, phase III trial to assess the efficacy of low dose tamoxifen (5 mg/day) administered for 5 years in women undergoing different types of HT. A randomized phase II trial established the optimal dose of tamoxifen in HT users [28]. The main results of the phase III trial (the HOT study, HT Opposed by Low-Dose Tamoxifen, Clinical Trials.gov NCT01579734) have previously been published [23]. The current psychosocial study was conducted in two centers, the European Institute of Oncology, Milan, and the Galliera Hospital, Genoa, Italy. The institutional review board of each site approved the protocol, and all women gave their written informed consent. 2.2. Study procedures Study procedures were previously described [27]. Briefly, participants were aware of the trial's launch through different means, including a mass media campaign (television, press, and web site announcements) or through direct contact with the study personnel during clinical consultations. A toll-free number was
We assessed age, education, marital status, and employment, whereas race was not included because all but three women were Caucasian. In addition, we included age at menopause, 5-year Gail risk of breast cancer (%), using 1.3% as cut off point for higher risk based on the Italian incidence data [31], body mass index (kg/m2), current use, type and duration of HT, frequency of mammographic screening (never, one to two times, or every 1e2 years), smoking habit (never, former, or current), physical activity (<3, 3 to 6,>6 h/ week), use of any medication (sometimes vs regular), and alcohol use (<five, five to 10, 11 to 20, or>20 glasses a week). 2.5. Sample size and statistical analyses In our initial decision making study [27], we had assumed the hypothesis (which turned out to be wrong) of a prevalence of breast cancer worry (worried þ very worried) of 5% in trial decliners and 15% in trial participants. With 450 women, including 250 trial participants and 200 decliners, we would be able to detect such a difference with 90% power at 5% two-sided statistical significance. Descriptive statistics used for continuous factors were mean and standard deviation (SD) or median and interquartile range (IQR); for categorical factors, absolute and relative (%) frequencies were adopted. Boxplots were used to visualize the associations between
110
G. Rondanina et al. / The Breast 34 (2017) 108e114
categorical and continuous factors and Spearman's rho was adopted to measure correlation among parameters and a nonparametric test for trend across ordered groups was calculated. The relationships between the study outcomes (breast cancer worry and absolute risk perception of breast cancer) and socio-demographic, health-related and psychological factors and response to the trial (participation vs. refusal) were examined by a logistic regression modeling. Two separate multivariate models were created: one setting worry about breast cancer as dependent variable (coded as worried/very worried vs. not/slightly/somewhat), the second setting absolute risk perception as dependent variable (coded as likely/very likely vs. 50-50/unlikely/very unlikely). HT use was categorized in three groups according to its duration (<5, 5e10, 10 þ years). For multivariate variable selection modeling we chose a manual approach (instead of stepwise selection algorithms), considering all clinically and statistically significant factors from univariate analyses and maintaining in the final models only those reaching a likelihood ratio test p-value <0.3. Forest plots (instead of tables) were employed to visually represent Odds Ratios (OR) and 95% confidence intervals of factors included in the final models. All analyses were conducted using STATA software (version 13; STATA Corp, College Station, TX, USA). Two-tailed probabilities were reported, and p-value ¼ 0.05 was used to define nominal statistical significance.
3. Results A total of 1457 women received counseling about the trial characteristics, 496 (34.0%) were entered onto the trial, whereas 961 (66%) refused to participate. After trial entry or refusal, all 1457 women received by mail a preprinted, self-reported psychosocial questionnaire, to which 457 (31.4%) replied. A total of 265 replies (53.4%) were from women who participated in the trial, and 192 (20.0%) were from women who refused to participate in the trial. The participant flow diagram is depicted in Fig. 1. The main subject characteristics of the 457 participants are summarized in Table 1. The mean age was 57 ± 5 years, mean age at menopause was 49 ± 4, and 28% were overweight (BMI>25) or obese (BMI>30). Approximately 20% were de novo HRT users, over 55% were on HRT for <5 years, <20% started HRT >5 years since menopause and 4% started HRT >10 years since menopause (overall
Fig. 1. Participant flow diagram.
Table 1 Main characteristics of study participants (n ¼ 457). n (%)a Age, years Mean (SD) Median (IQR) Age at menopause, years Mean (SD) Median (IQR) Body Mass Index <18.5 18.5e24.9 25.0e29.9 30þ Parity No Yes Mammography screening 1 or 2 Regularly 5-year Gail risk (%) <1.3 1.3 Smoking Habit Non smoker Smoker Former Daily Alcohol consumption No Yes Physical Activity No Yes Education < High school High school University degree Marital status Single Married Divorced Widow
57 (5.0) 57 (54e59) 49 (4.3) 50 (47e52) 16 (3.5) 314 (68.7) 105 (23.0) 22 (4.8) 54 (12.7) 371 (87.3) 35 (7.7) 422 (92.3) 105 (24.8) 319 (75.2) 235 (55.2) 96 (22.5) 95 (22.3) b
415 (92.4) 34 (7.6) 222 (48.8) 233 (51.2) 110 (24.1) 242 (53.0) 105 (23.0) 31 (6.8) 342 (74.9) 50 (10.9) 34 (7.4)
Abbreviations: SD ¼ standard deviation; IQR ¼ interquartile range. a Percentages were adjusted for missing data. b More than 10 glasses/week.
mean ± SD, 3.0 ± 3.5 years). Route of HT was oral in 47% and transdermal in 53% of the cases. Combined HT was continuous in 28% and sequential in 49%, whereas 19% were on unopposed estrogen. Estrogen therapy included 92% on estradiol, of whom 10% at low dose (1 mg), CEE users were 8%, of whom 3% at low dose (0.3 mg). Most used progestins were progesterone in 20% and medroxyprogesterone acetate in 15%. Nearly all women underwent regular mammographic screening programs, their 5 year Gail risk was 1.3% or greater in 75% of subjects, 55% were never smokers, half were regular exercisers, three quarters had high school or university degree and were married. The main findings of psychological questionnaire are summarized in Table 2. Nearly 12% reported to be worried or very worried of breast cancer, and approximately 20% reported their absolute perceived risk of breast cancer to be likely or very likely or to be higher relative to a woman of the same age (perceived relative risk). Over 60% of the women reported no anxiety and 70% no depression symptoms, whereas 17% and 13% had moderate to severe anxiety or depression, respectively. The factors independently associated with worry about breast cancer in a multivariate model are reported in Fig. 2. There was a statistically significant inverse association between age at menopause and breast cancer worry, as women with early menopause (<48 years) were 5 times more worried about breast cancer
G. Rondanina et al. / The Breast 34 (2017) 108e114 Table 2 Main psychological factors of study population (n ¼ 457).a n (%) Worry about breast cancer Not at all 109 (24.1) Slightly 242 (53.5) Somewhat 48 (10.6) Worried 41 (9.1) Very worried 12 (2.7) Perceived absolute breast cancer risk (verbal assessment) Very unlikely 31 (7.2) Unlikely 148 (34.3) 50e50 chance 162 (37.6) Likely 87 (20.2) Very likely 3 (0.7) (numerical, 0e100 scale) Mean (SD) 31.8 (21.2) Median (Range) 30 (0e100) Perceived relative breast cancer risk Much below average 25 (6.0) Below average 133 (32.0) Same risk 165 (39.8) Above average 89 (21.4) Much above average 3 (0.7) Anxiety, score (HADS-A) Normal (0e7) 286 (63.0) Mild (8e10) 91 (20.0) Moderate (11e14) 56 (12.3) Severe (15e21) 21 (4.6) Depression, score (HADS-D) Normal (0e7) 320 (70.6) Mild (8e10) 74 (16.3) Moderate (11e14) 52 (11.5) Severe (15e21) 7 (1.5) a
111
significantly less worried of breast cancer than never users. Longer duration of HT was not associated with elevated worry or risk perception, nor was there any association with HT type or route. Former smokers were more worried than never smokers, whereas current smokers were not significantly more worried. Alcohol use, lack of physical exercise and regular screening were not significantly associated with worry. The relationship between age at menopause and worry about breast cancer is further depicted in Fig. 3. There was a significant inverse linear trend between age at menopause (as continuous variable) and breast cancer worry (p-trend ¼ 0.001). The factors independently associated with a likely or very likely absolute risk perception of developing breast cancer are shown in Fig. 4. Women older than 60 years and women with moderate or severe depression had a higher increased perception of breast cancer risk, whereas women with moderate or severe anxiety had a borderline significant trend to a lower absolute risk perception. Women with heightened breast cancer worry had a higher risk perception (p-trend<0.001 using the scale variable, data not shown). Life style attitudes at increased risk such as smoking, overweight and no exercise were associated with a 3-fold absolute risk perception of breast cancer. Use of HT and parity were associated with a trend to an increased risk perception. The correlation between the Gail risk level and absolute risk perception was weak (Spearman's rho ¼ 0.12) and not significant in the multivariate model. The variable “response to the trial” (participation vs. refusal) added in the models did not affect the OR estimates of other factors included and was not significantly associated with worry and risk perception (p ¼ 0.4 and 0.3, respectively).
Percentages were adjusted for missing data.
4. Discussion compared with women who had menopause after 52 years (OR ¼ 5.0, 95% CI: 1.2e21.1). Women who had a higher absolute risk perception of developing breast cancer were more worried. There was no clear association between anxiety or/depression and breast cancer worry (p ¼ 0.1). Users of HT were slightly albeit not
There is a lot of interest in understanding the reasons why therapeutic prevention of breast cancer has very little uptake [4,9,10,32] despite strong evidence of efficacy from large randomized clinical trials [5e7]. Several studies have pointed to medical variables, such as fear of serious adverse events and increase of
Fig. 2. Forest plot of the factors associated with worry about breast cancer (HT¼Hormone Replacement Therapy).
112
G. Rondanina et al. / The Breast 34 (2017) 108e114
Fig. 3. Box plot of the relationship between age at menopause and worry about breast cancer.
Fig. 4. Forest plot of the factors associated with absolute risk perception of breast cancer (HT¼Hormone Replacement Therapy).
menopausal and sexual symptoms [17,19] or immature mortality data [2,3] as important factors for refusing therapeutic prevention with SERMs. Predictors of low adherence also include early menopausal and sexual side effects during therapeutic prevention [33]. In our previous study [27], the most frequent reasons for trial entry were willingness to participate in a research program (60%), the need/desire to receive frequent medical care (58%) and the desire to contribute to medical knowledge (44%); whereas reasons for refusal included fear of medication abuse (33%), concern about adverse effects (31%), and physician advice against enrollment (24%). So the different uptake to the questionnaire reflects the lowest motivation to the research contribution in the decliner group. In addition, it should be taken into account that tamoxifen is still off label in Italy for prevention purposes. Other studies have underlined psychological factors such as worry about breast cancer and risk perception in influencing the
decision to participate or refuse preventive therapy trials [17,20,27]. This picture is similar to the issue of prescribing HT as a part of a prevention strategy for women at the onset of menopause [1]. Increased participation and retention to breast cancer therapeutic prevention as well as to HT is therefore likely to be the result of a better understanding of factors such as breast cancer worry and risk perception. In a previous study [27], we had shown that women who were very worried had a higher rate of refusal to participate into the trial. Given this counterintuitive finding, we searched in more depth which factors influenced worry and risk perception of breast cancer. The most striking finding from our study is the strong inverse association between early menopause and worry about breast cancer. This is in contrast with the known protective effect of a shorter life-time hormonal exposure of early menopause on breast cancer risk [34]. Considering the wide CI and the multiple testing, however, a false positive result cannot be excluded. This finding
G. Rondanina et al. / The Breast 34 (2017) 108e114
suggests that early menopause probably reflects a feeling of premature aging and higher vulnerability to disease and specifically to breast cancer. Interestingly, the borderline significant inverse association between HT use and worry about breast cancer is again in contrast with the clinical evidence of an increase risk of developing breast cancer during combined HT use [13]. A number of mechanisms may be advocated to explain this counter-intuitive relationship, including a favorable modulation of mood and well being by the HT as well as a delayed aging [25,35]. Worry about breast cancer was also significantly associated with increased risk perception and a trend to moderate or severe anxiety, two factors which may potentially contribute to avoidant behaviors towards health promotion programs. Indeed, recent studies have shown that women with high levels of worry and anxiety tend to adopt avoidant behaviors towards health promotion particularly if the cognitive construct of high risk perception makes them perceive the risk of disease very likely [36]. Thus, increasing risk perception and awareness in very worried women and high level of anxiety may be counterproductive if not accompanied by strategies of emotional improvement. For instance, Ferrer et al. [37] have shown that greater worry appeared to reverse the conventional relationship between self-efficacy and dietary restriction inasmuch as those who were self-efficacious and worried were less likely to restrict unhealthy foods. More importantly, among subjects higher in worry, higher levels of risk perception were associated with unhealthy life style such as lower vegetable consumption and exercise [38]. These results suggest the hypothesis that, among people high in worry, attempts to increase risk perception without a concomitant strategy of emotional coping could be counterproductive because of avoidant attitudes. It is also true that extra screening is more likely to be acceptable than taking a medication such as tamoxifen, as reflected by the main reasons for refusal in our study. Interestingly, we found that former smokers were more worried than never smokers, whereas current smokers were not significantly more worried than never smokers. While smoking is only mildly associated with increased breast cancer risk [39], former smokers have a more in-depth cognitive processing of risk information than current smokers, who in fact tend to minimize risk through a reduction dissonance process [40]. Also, smokers hold more pessimistic and avoidant beliefs about cancer, which have been shown to deter early-detection and prevention behaviors [41]. Taken together, all the above mentioned findings support the notion of the affective nature of worry. Conversely, the predominant cognitive nature of risk perception is underlined by the strong associations with biological factors, including older age and unhealthy life style, although the affective component of risk perception is illustrated by the association with worry and depression. The strong association between worry and risk perception describes the strict interdependence between the two constructs which are engaged in the decision making process, where the affective factors act directly to respond to internal needs and promote action, whereas the cognitive factors provide the different options to plan a decision program [42]. This strict crosstalk suggests that for any prevention plan to become effective health care providers should focus not only on risks and benefits of treatments but also to improve the understanding of the affective factors implicated in the decision making process of health behavior changes. Increasing risk perception in women who are very worried and anxious might in fact determine avoidant behaviors and be counterproductive [37]. Our study population had a prevalence of psychological disorders which is in line with the expected figures in the same age range [43,44] as only 12% of our population reported to be worried or very worried of breast cancer, and approximately 20% reported
113
their absolute perceived risk of breast cancer to be likely or very likely or to be higher relative to a woman of the same age. Also, only 17% and 13% of our women had moderate to severe anxiety or depression, respectively. All women in our study were assessed before trial acceptance or refusal and before allocation to tamoxifen or placebo. Also, 75% of our study population were at increased risk for breast cancer according to the Gail model adjusted for the Italian incidence data. In addition, most women in the study were taking combined HT, a known risk factor for breast cancer, so our population was mostly composed of high risk women based on the Gail model and the use of combined HT. Therefore, our findings are applicable to a vast population of midlife women at increased risk for breast cancer who are willing to receive HT for menopausal disorder relief or health promotion but are concerned about breast cancer risk. In conclusion, our findings show that the affective construct of worry about breast cancer is different from the cognitive construct of risk perception, which is more adherent to biological factors. A better understanding of the emotional factors that modulate the decision making process can potentially improve adherence to HT use and breast cancer therapeutic prevention programs. Clinical trials are underway to assess the long term efficacy and safety of low dose tamoxifen in midlife women on HT (Clinical Trials.gov NCT01579734) or because of prior intraepithelial neoplasia (ClinicalTrials.gov NCT01357772). Financial support Supported by the Italian Foundation for Cancer Research (A. DeCensi), the Italian Association for Cancer Research (AIRC) (A. DeCensi), Avon Italia (A. DeCensi), the Italian League against Cancer (LILT project number 51/2005) (A. DeCensi), the American Italian Cancer Foundation, the Italian Ministry of Health, RFPS-2006-1339898 (A. DeCensi). Tamoxifen and placebo were donated by FIDIA Farmaceutici S.p.a, Abano Terme, Italy. We are grateful to the study participants for their heart and energy. Disclosure of conflicts of interest The authors declare that there are no conflicts of interest. References [1] Manson JE, Kaunitz AM. Menopause managementegetting clinical care back on track. N Engl J Med 2016;374:803e6. [2] Cameron DA. Breast cancer chemoprevention: little progress in practice? Lancet 2014;383. 10e1020. [3] Narod SA. Tamoxifen chemoprevention-end of the road? JAMA Oncol 2015;1: 1033e4. [4] DeCensi A, Thorat MA, Bonanni B, Smith SG, Cuzick J. Barriers to preventive therapy for breast and other major cancers and strategies to improve uptake. Ecancermedicalscience 2015;9:595. [5] Cuzick J, Sestak I, Bonanni B, Costantino JP, Cummings S, DeCensi A, et al. Selective oestrogen receptor modulators in prevention of breast cancer: an updated meta-analysis of individual participant data. Lancet 2013;25: 1827e34. s-Martínez JE, Cheung AM, Chlebowski RT, Wactawski[6] Goss PE, Ingle JN, Ale Wende J, et al. Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med 2011;364:2381e91. [7] Cuzick J, Sestak I, Forbes JF, Dowsett M, Knox J, Cawthorn S, et al. Anastrozole for prevention of breast cancer in high-risk postmenopausal women (IBIS-II): an international, double-blind, randomized placebo-controlled trial. Lancet 2014;383:1041e8. [8] Dunn BK, Cazzaniga M, DeCensi A. Exemestane: one part of the chemopreventive spectrum for ER-positive breast cancer. Breast 2013;22:225e37. [9] Waters EA, McNeel TS, Stevens WM, Freedman AN. Use of tamoxifen and raloxifene for breast cancer chemoprevention in 2010. Breast Cancer Res Treat 2012;134:875e80. [10] Smith SG, Sestak I, Forster A, Partridge A, Side L, Wolf MS, et al. Factors affecting uptake and adherence to breast cancer chemoprevention: asystematic review and meta-analysis. Ann Oncol 2016;27:575e90.
114
G. Rondanina et al. / The Breast 34 (2017) 108e114
[11] Fisher B, Costantino JP, Wickerham DL, Cecchini RS, Cronin WM, Robidoux A, et al. Tamoxifen for prevention of breast cancer: report of the national surgical adjuvant breast and bowel project P-1 study. J Natl Cancer Inst 1998;16(90): 1371e88. [12] Jewett PI, Gangnon RE, Trentham-Dietz A, Sprague BL. Trends of postmenopausal estrogen plus progestin prevalence in the United States between 1970 and 2010. Obstet Gynecol 2014;124:727e33. [13] Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA 2002;288:321e33. [14] Manson JE, Chlebowski RT, Stefanick ML, Aragaki AK, Rossouw JE, Prentice RL, et al. Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women's Health Initiative randomized trials. JAMA 2013;310:1353e68. [15] Daubisse-Marliac L, Delafosse P, Boitard JB, Poncet F, Grosclaude P, Colonna M. Breast cancer incidence and time trend in France from 1990 to 2007: a population-based study from two French cancer registries. Ann Oncol 2011;22:329e34. [16] Antoine C, Ameye L, Paesmans M, de Azambuja E, Rozenberg S. Menopausal hormone therapy use in relation to breast cancer incidence in 11 European countries. Maturitas 2016;84:81e8. [17] Bober SL, Hoke LA, Duda RB, Regan MM, Tung NM. Decision-making about tamoxifen in women at high risk for breast cancer: clinical and psychological factors. J Clin Oncol 2004;22:4951e7. [18] Ropka ME, Keim J, Philbrick JT. Patient decisions about breast cancer chemoprevention: a systematic review and meta-analysis. J Clin Oncol 2010;28: 3090e5. [19] Bastian LA, Lipkus IM, Kuchibhatla MN, Weng HH, Halabi S, Ryan PD, et al. Women's interest in chemoprevention for breast cancer. Arch Intern Med 2001;161:1639e44. [20] Holmberg C, Waters EA, Whitehouse K, Daly M, McCaskill-Stevens W. My lived experiences are more important than your probabilities: the role of individualized risk estimates for decision making about participation in the study of tamoxifen and raloxifene (STAR). Med Decis Mak 2015;35:1010e22. [21] Katapodi MC, Lee KA, Facione NC, Dodd MJ. Predictors of perceived breast cancer risk and the relation between perceived risk and breast cancer screening: a meta-analytic review. Prev Med 2004;38:388e402. [22] Decensi A, Galli A, Veronesi U. HRT opposed to low-dose tamoxifen (HOT study): rationale and design. Recent Results Cancer Res 2003;163(104e11): 264e6. discussion. [23] DeCensi A, Bonanni B, Maisonneuve P, Serrano D, Omodei U, Varricchio C, et al. Italian HOT Study Group. A phase-III prevention trial of low-dose tamoxifen in postmenopausal hormone replacement therapy users: the HOT study. Ann Oncol 2013;24:2753e60. [24] Veronesi U, Maisonneuve P, Rotmensz N, Bonanni B, Boyle P, Viale G, et al. Tamoxifen for the prevention of breast cancer: late results of the Italian Randomized Tamoxifen Prevention Trial among women with hysterectomy. J Natl Cancer Inst 2007;99:727e37. [25] Pinkerton JV, Pan K, Abraham L, Racketa J, Ryan KA, Chines AA, et al. Sleep parameters and health-related quality of life with bazedoxifene/conjugated estrogens: a randomized trial. Menopause 2014;21:252e9. [26] Harvey JA, Pinkerton JV, Baracat EC, Shi H, Chines AA, Mirkin S. Breast density changes in a randomized controlled trial evaluating bazedoxifene/conjugated estrogens. Menopause 2013;20:138e45.
[27] Rondanina G, Puntoni M, Severi G, Varricchio C, Zunino A, Feroce I, et al. Psychological and clinical factors implicated in decision making about a trial of low-dose tamoxifen in hormone replacement therapy users. J Clin Oncol 2008;26:1537e43. [28] Decensi A, Gandini S, Serrano D, Cazzaniga M, Pizzamiglio M, Maffini F, et al. Randomized dose-ranging trial of tamoxifen at low doses in hormone replacement therapy users. J Clin Oncol 2007;25:4201e9. [29] Lerman C, Daly M, Sands C, Balshem A, Lustbader E, Heggan T, et al. Mammography adherence and psychological distress among women at risk for breast cancer. J Natl Cancer Inst 1993;85:1074e80. [30] Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the Hospital Anxiety and Depression Scale: an updated literature review. J Psychosom Res 2002;52:69e77. [31] Guerrieri-Gonzaga A, Robertson C, Bonanni B, Serrano D, Cazzaniga M, Mora S. Preliminary results on safety and activity of a randomized, double-blind, 2 x 2 trial of low-dose tamoxifen and fenretinide for breast cancer prevention in premenopausal women. J Clin Oncol 2006;24:129e35. [32] Meyskens Jr FL, Mukhtar H, Rock CL, Cuzick J, Kensler TW, Yang CS, et al. Cancer prevention: obstacles, challenges and the road ahead. J Natl Cancer Inst 2015;108. pii: djv309. [33] Land SR, Walcott FL, Liu Q, Wickerham DL, Costantino JP, Ganz PA. Symptoms and QOL as predictors of chemoprevention adherence in NRG oncology/ NSABP trial P-1. J Natl Cancer Inst 2015;108. pii: djv365. [34] Collaborative Group on Hormonal Factors in Breast Cancer. Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol 2012;13:1141e51. [35] Welton AJ, Vickers MR, Kim J, Ford D, Lawton BA, MacLennan AH, et al. Health related quality of life after combined hormone replacement therapy: randomised controlled trial. BMJ 2008;337:a1190. [36] Persoskie A, Ferrer RA, Klein WM. Association of cancer worry and perceived risk with doctor avoidance: an analysis of information avoidance in a nationally representative US sample. J Behav Med 2014;37:977e87. [37] Ferrer RA, Bergman HE, Klein WM. Worry as a predictor of nutrition behaviors: results from a nationally representative survey. Health Educ Behav 2013;40:88e96. [38] Ferrer RA, Portnoy DB, Klein WM. Worry and risk perceptions as independent and interacting predictors of health protective behaviors. J Health Commun 2013;18:397e409. [39] Gaudet MM, Gapstur SM, Sun J, Diver WR, Hannan LM, Thun MJ. Active smoking and breast cancer risk: original cohort data and meta-analysis. J Natl Cancer Inst 2013;105:515e25. [40] Kaufman AR, Koblitz AR, Persoskie A, Ferrer RA, Klein WM, Dwyer LA, et al. Factor structure and stability of smoking-related health beliefs in the national lung screening trial. Nicot Tob Res 2016;18:321e9. [41] Quaife SL, McEwen A, Janes SM, Wardle J. Smoking is associated with pessimistic and avoidant beliefs about cancer: results from the international cancer benchmarking partnership. Br J Cancer 2015;112:1799e804. € berg L. Worry and risk perception. Risk Anal 1998;18:85e93. [42] Sjo [43] Tangen T, Mykletun A. Depression and anxiety through the climacteric period: an epidemiological study (HUNT-II). J Psychosom Obstet Gynaecol 2008;29: 125e31. [44] Burt VK, Altshuler LL, Rasgon N. Depressive symptoms in the perimenopause: prevalence, assessment, and guidelines for treatment. Harv Rev Psychiatry 1998;6:121e32.
Contents lists available at ScienceDirect
The Breast journal homepage: www.elsevier.com/brst
Original article
Worry and risk perception of breast cancer in a prevention trial of low dose tamoxifen in midlife postmenopausal hormone users Gabriella Rondanina a, Matteo Puntoni b, Aliana Guerrieri-Gonzaga c, Domenico Marra d, Bernardo Bonanni c, Andrea DeCensi d, e, * a
Psychology Center Piccapietra, Genoa, Italy Office of the Scientific Director, E.O. Ospedali Galliera, Genoa, Italy Division of Cancer Prevention and Genetics, European Institute of Oncology, Milan, Italy d Division of Medical Oncology, E.O. Ospedali Galliera, Genoa, Italy e Wolfson Institute of Preventive Medicine, Queen Mary University of London, London, UK b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 13 March 2017 Received in revised form 28 April 2017 Accepted 16 May 2017
Objective: There is increasing interest in combining postmenopausal hormone therapy (HT) and SERMs in midlife women. We previously showed that refusal to participate in a prevention trial of low dose tamoxifen in HT users was associated with higher worry about breast cancer. Given this counterintuitive finding, we studied which factors influenced worry and risk perception of breast cancer. Methods: We assessed the relationships of breast cancer worry and risk perception with age, age at menopause, Gail risk, education, adherence to mammographic screening, BMI, smoking, physical activity, alcohol use, anxiety and depression in 457 midlife HT users who were eligible to participate in the trial. Results: Women with menopause <48 years were more worried about breast cancer than women with menopause >52 years (OR ¼ 5.0, 95% CI, 1.2e21.1). Worry was also associated with high absolute risk perception and former smoking. Factors associated with higher risk perception were age>60 years, atrisk life style, worry about breast cancer and depression. Conclusions: The inverse association between early menopause and worry about breast cancer is in contrast with the known protective effect of early menopause on breast cancer risk and seems to reflect a feeling of aging and disease vulnerability. Our findings indicate that worry about cancer has an affective construct which is independent of breast cancer biology but is engaged in health decision making. Increasing breast cancer risk awareness in subjects high in worry without a plan of emotional coping may therefore be counterproductive because of avoidant attitudes. © 2017 Published by Elsevier Ltd.
Keywords: Cancer worry Cancer risk perception Breast cancer Tamoxifen Chemoprevention Postmenopausal hormone therapy
1. Introduction Postmenopausal hormone therapy (HT) at the onset of menopause is an example of the paradigm shift towards a proactive behavior of health promotion inasmuch as users are required to possess a good sense of awareness and ability to carefully assess risks and benefits [1]. This is true also for the use of tamoxifen in primary prevention since the trade-off between risks and benefits is subject to continuous debate [2e4]. However, the decisionmaking process about participating in a breast cancer therapeutic
* Corresponding author. Division of Medical Oncology, E.O. Ospedali Galliera, Mura delle Cappuccine 14, 16128, Genoa, Italy. E-mail address: [email protected] (A. DeCensi). http://dx.doi.org/10.1016/j.breast.2017.05.008 0960-9776/© 2017 Published by Elsevier Ltd.
prevention trial is poorly understood. Despite the highest evidence of efficacy ensuing from large trials, the uptake of selective estrogen receptor modulators such as tamoxifen and raloxifene [5] or aromatase inhibitors [6e8] has so far been very limited in clinical practice [9,10], mainly because of the risk of endometrial cancer and venous thromboembolism, partly attenuating the 50% reduction of breast cancer risk observed in the NSABP-P1 trial [11]. Likewise, the use of HT has dramatically dropped [12] after the initial publication of the WHI trial [13], notwithstanding the evidence of a timing effect with a mortality reduction in women aged 50 to 59 in subsequent analyses [1,14]. While initial reports linked a decrease in breast cancer incidence to the HT drop after the WHI trial results [15], most recent evidence on the association between breast cancer incidence and the rate of MHT use during the past years in Europe suggests in fact that breast cancer incidence in
G. Rondanina et al. / The Breast 34 (2017) 108e114
women aged 45e64 has not decreased in parallel with the HT drop, indicating the role of additional confounding factors [16]. While attention has mainly been focused on medical variables discouraging the use of tamoxifen for prevention, including concerns about adverse events [17] and lack of demonstrated mortality reduction [2,3], other studies have pointed to risk perception, worry about breast cancer and lived experiences as factors affecting the decision to participate in a preventive therapy trial for breast cancer [18e20]. This is not surprising given the notion that adherence to breast cancer screening programs is influenced by sociodemographic as well as emotional factors [21]. We pioneered the concept of combining postmenopausal HT and low dose tamoxifen to retain the benefits while reducing the risks of either agent in a phase III trial (HOT trial) in 1884 midlife women [22,23]. The results showed a non-significant 20% reduction of breast cancer in the tamoxifen arm compared with the placebo arm [23]. Importantly, tamoxifen showed favorable trends in women on HT for <5 years (adjusted RR ¼ 0.35; 95% CI 0.15e0.85) and in estrogen-alone users (RR ¼ 0.26, 95% CI 0.05e1.28), but no effect in the combined estro-progestin subgroup. Combining HT to tamoxifen was also associated with a lower incidence of menopausal symptoms relative to tamoxifen alone [24]. More recently, the combination of conjugated estrogen with the SERM bazedoxifene has proven to be beneficial on menopausal quality of life [25] without an increase in mammographic density [26]. Participation in the HOT trial was higher in women at younger age and in those satisfied with health care providers, whereas it was lower in women with breast cancer worry, suggesting a condition of psychological well-being as a promoting factor for the trial participation [27]. Given the counterintuitive finding and the discrepancy with some prior studies on the role of breast cancer worry in adopting prevention strategies, here we searched in depth which factors explained the worry about breast cancer and the risk perception of breast cancer. These factors may be important to better explain health promotion attitudes and therefore enhance compliance to therapeutic prevention in midlife postmenopausal women.
109
available to address initial questions and to fix an appointment with the study personnel, mostly physicians. Counseling about the trial included an evaluation of benefits and risks of HT and tamoxifen given alone, and description of the study procedures. Participant-physician communication was based on a scripted protocol. After counseling and subsequent decision making about the trial, all women received by mail a preprinted, self-report questionnaire based on previous research on medical decision making in the area of breast cancer prevention [29]. 2.3. Psychological factors Worry about developing breast cancer was the study's main outcome measure and was assessed by the following question: “How worried are you about getting breast cancer in the next 5 years?” Responses were graded using a Likert scale: “not at all worried”, “slightly worried”, “somewhat worried”, “worried” and “very worried.” Risk perception of breast cancer were assessed as previously described [19]. Perception of absolute breast cancer risk was assessed using standard verbal and numerical measures within the span of 5 years. For the verbal measures, women were asked, “How likely are you to get breast cancer in the next 5 years?”.Responses were “very unlikely”,“unlikely”,“50/50 chance”, “likely” and “very likely.” For the numerical measures, women were asked, “On a scale from 0 to 100, with 0 indicating certain not to happen and 100 indicating certain to happen, how likely are you to get breast cancer in the next 5 years?”. Perceptions of comparative breast cancer risk were assessed through the following question: “Compared with other women of your age, what are the odds that you will get breast cancer in the next 5 years?”. Responses were “much below average”, “below average”, “same average risk as other women my age”, “above average”, and “much above average”. Although the level of breast cancer risk knowledge was not formally measured, the results of the Gail model and the risk associated with HT use were illustrated to the screened women at face to face or telephone counseling. Anxiety and depression were assessed using the Hospital Anxiety and Depression Scale [30].
2. Materials and methods 2.4. Sociodemographic, health-related and lifestyle variables 2.1. Participants Participants in the psychosocial study were postmenopausal healthy women currently on or about to start HT for the treatment of menopausal symptoms or health promotion who were eligible for a double-blind, placebo-controlled, phase III trial to assess the efficacy of low dose tamoxifen (5 mg/day) administered for 5 years in women undergoing different types of HT. A randomized phase II trial established the optimal dose of tamoxifen in HT users [28]. The main results of the phase III trial (the HOT study, HT Opposed by Low-Dose Tamoxifen, Clinical Trials.gov NCT01579734) have previously been published [23]. The current psychosocial study was conducted in two centers, the European Institute of Oncology, Milan, and the Galliera Hospital, Genoa, Italy. The institutional review board of each site approved the protocol, and all women gave their written informed consent. 2.2. Study procedures Study procedures were previously described [27]. Briefly, participants were aware of the trial's launch through different means, including a mass media campaign (television, press, and web site announcements) or through direct contact with the study personnel during clinical consultations. A toll-free number was
We assessed age, education, marital status, and employment, whereas race was not included because all but three women were Caucasian. In addition, we included age at menopause, 5-year Gail risk of breast cancer (%), using 1.3% as cut off point for higher risk based on the Italian incidence data [31], body mass index (kg/m2), current use, type and duration of HT, frequency of mammographic screening (never, one to two times, or every 1e2 years), smoking habit (never, former, or current), physical activity (<3, 3 to 6,>6 h/ week), use of any medication (sometimes vs regular), and alcohol use (<five, five to 10, 11 to 20, or>20 glasses a week). 2.5. Sample size and statistical analyses In our initial decision making study [27], we had assumed the hypothesis (which turned out to be wrong) of a prevalence of breast cancer worry (worried þ very worried) of 5% in trial decliners and 15% in trial participants. With 450 women, including 250 trial participants and 200 decliners, we would be able to detect such a difference with 90% power at 5% two-sided statistical significance. Descriptive statistics used for continuous factors were mean and standard deviation (SD) or median and interquartile range (IQR); for categorical factors, absolute and relative (%) frequencies were adopted. Boxplots were used to visualize the associations between
110
G. Rondanina et al. / The Breast 34 (2017) 108e114
categorical and continuous factors and Spearman's rho was adopted to measure correlation among parameters and a nonparametric test for trend across ordered groups was calculated. The relationships between the study outcomes (breast cancer worry and absolute risk perception of breast cancer) and socio-demographic, health-related and psychological factors and response to the trial (participation vs. refusal) were examined by a logistic regression modeling. Two separate multivariate models were created: one setting worry about breast cancer as dependent variable (coded as worried/very worried vs. not/slightly/somewhat), the second setting absolute risk perception as dependent variable (coded as likely/very likely vs. 50-50/unlikely/very unlikely). HT use was categorized in three groups according to its duration (<5, 5e10, 10 þ years). For multivariate variable selection modeling we chose a manual approach (instead of stepwise selection algorithms), considering all clinically and statistically significant factors from univariate analyses and maintaining in the final models only those reaching a likelihood ratio test p-value <0.3. Forest plots (instead of tables) were employed to visually represent Odds Ratios (OR) and 95% confidence intervals of factors included in the final models. All analyses were conducted using STATA software (version 13; STATA Corp, College Station, TX, USA). Two-tailed probabilities were reported, and p-value ¼ 0.05 was used to define nominal statistical significance.
3. Results A total of 1457 women received counseling about the trial characteristics, 496 (34.0%) were entered onto the trial, whereas 961 (66%) refused to participate. After trial entry or refusal, all 1457 women received by mail a preprinted, self-reported psychosocial questionnaire, to which 457 (31.4%) replied. A total of 265 replies (53.4%) were from women who participated in the trial, and 192 (20.0%) were from women who refused to participate in the trial. The participant flow diagram is depicted in Fig. 1. The main subject characteristics of the 457 participants are summarized in Table 1. The mean age was 57 ± 5 years, mean age at menopause was 49 ± 4, and 28% were overweight (BMI>25) or obese (BMI>30). Approximately 20% were de novo HRT users, over 55% were on HRT for <5 years, <20% started HRT >5 years since menopause and 4% started HRT >10 years since menopause (overall
Fig. 1. Participant flow diagram.
Table 1 Main characteristics of study participants (n ¼ 457). n (%)a Age, years Mean (SD) Median (IQR) Age at menopause, years Mean (SD) Median (IQR) Body Mass Index <18.5 18.5e24.9 25.0e29.9 30þ Parity No Yes Mammography screening 1 or 2 Regularly 5-year Gail risk (%) <1.3 1.3 Smoking Habit Non smoker Smoker Former Daily Alcohol consumption No Yes Physical Activity No Yes Education < High school High school University degree Marital status Single Married Divorced Widow
57 (5.0) 57 (54e59) 49 (4.3) 50 (47e52) 16 (3.5) 314 (68.7) 105 (23.0) 22 (4.8) 54 (12.7) 371 (87.3) 35 (7.7) 422 (92.3) 105 (24.8) 319 (75.2) 235 (55.2) 96 (22.5) 95 (22.3) b
415 (92.4) 34 (7.6) 222 (48.8) 233 (51.2) 110 (24.1) 242 (53.0) 105 (23.0) 31 (6.8) 342 (74.9) 50 (10.9) 34 (7.4)
Abbreviations: SD ¼ standard deviation; IQR ¼ interquartile range. a Percentages were adjusted for missing data. b More than 10 glasses/week.
mean ± SD, 3.0 ± 3.5 years). Route of HT was oral in 47% and transdermal in 53% of the cases. Combined HT was continuous in 28% and sequential in 49%, whereas 19% were on unopposed estrogen. Estrogen therapy included 92% on estradiol, of whom 10% at low dose (1 mg), CEE users were 8%, of whom 3% at low dose (0.3 mg). Most used progestins were progesterone in 20% and medroxyprogesterone acetate in 15%. Nearly all women underwent regular mammographic screening programs, their 5 year Gail risk was 1.3% or greater in 75% of subjects, 55% were never smokers, half were regular exercisers, three quarters had high school or university degree and were married. The main findings of psychological questionnaire are summarized in Table 2. Nearly 12% reported to be worried or very worried of breast cancer, and approximately 20% reported their absolute perceived risk of breast cancer to be likely or very likely or to be higher relative to a woman of the same age (perceived relative risk). Over 60% of the women reported no anxiety and 70% no depression symptoms, whereas 17% and 13% had moderate to severe anxiety or depression, respectively. The factors independently associated with worry about breast cancer in a multivariate model are reported in Fig. 2. There was a statistically significant inverse association between age at menopause and breast cancer worry, as women with early menopause (<48 years) were 5 times more worried about breast cancer
G. Rondanina et al. / The Breast 34 (2017) 108e114 Table 2 Main psychological factors of study population (n ¼ 457).a n (%) Worry about breast cancer Not at all 109 (24.1) Slightly 242 (53.5) Somewhat 48 (10.6) Worried 41 (9.1) Very worried 12 (2.7) Perceived absolute breast cancer risk (verbal assessment) Very unlikely 31 (7.2) Unlikely 148 (34.3) 50e50 chance 162 (37.6) Likely 87 (20.2) Very likely 3 (0.7) (numerical, 0e100 scale) Mean (SD) 31.8 (21.2) Median (Range) 30 (0e100) Perceived relative breast cancer risk Much below average 25 (6.0) Below average 133 (32.0) Same risk 165 (39.8) Above average 89 (21.4) Much above average 3 (0.7) Anxiety, score (HADS-A) Normal (0e7) 286 (63.0) Mild (8e10) 91 (20.0) Moderate (11e14) 56 (12.3) Severe (15e21) 21 (4.6) Depression, score (HADS-D) Normal (0e7) 320 (70.6) Mild (8e10) 74 (16.3) Moderate (11e14) 52 (11.5) Severe (15e21) 7 (1.5) a
111
significantly less worried of breast cancer than never users. Longer duration of HT was not associated with elevated worry or risk perception, nor was there any association with HT type or route. Former smokers were more worried than never smokers, whereas current smokers were not significantly more worried. Alcohol use, lack of physical exercise and regular screening were not significantly associated with worry. The relationship between age at menopause and worry about breast cancer is further depicted in Fig. 3. There was a significant inverse linear trend between age at menopause (as continuous variable) and breast cancer worry (p-trend ¼ 0.001). The factors independently associated with a likely or very likely absolute risk perception of developing breast cancer are shown in Fig. 4. Women older than 60 years and women with moderate or severe depression had a higher increased perception of breast cancer risk, whereas women with moderate or severe anxiety had a borderline significant trend to a lower absolute risk perception. Women with heightened breast cancer worry had a higher risk perception (p-trend<0.001 using the scale variable, data not shown). Life style attitudes at increased risk such as smoking, overweight and no exercise were associated with a 3-fold absolute risk perception of breast cancer. Use of HT and parity were associated with a trend to an increased risk perception. The correlation between the Gail risk level and absolute risk perception was weak (Spearman's rho ¼ 0.12) and not significant in the multivariate model. The variable “response to the trial” (participation vs. refusal) added in the models did not affect the OR estimates of other factors included and was not significantly associated with worry and risk perception (p ¼ 0.4 and 0.3, respectively).
Percentages were adjusted for missing data.
4. Discussion compared with women who had menopause after 52 years (OR ¼ 5.0, 95% CI: 1.2e21.1). Women who had a higher absolute risk perception of developing breast cancer were more worried. There was no clear association between anxiety or/depression and breast cancer worry (p ¼ 0.1). Users of HT were slightly albeit not
There is a lot of interest in understanding the reasons why therapeutic prevention of breast cancer has very little uptake [4,9,10,32] despite strong evidence of efficacy from large randomized clinical trials [5e7]. Several studies have pointed to medical variables, such as fear of serious adverse events and increase of
Fig. 2. Forest plot of the factors associated with worry about breast cancer (HT¼Hormone Replacement Therapy).
112
G. Rondanina et al. / The Breast 34 (2017) 108e114
Fig. 3. Box plot of the relationship between age at menopause and worry about breast cancer.
Fig. 4. Forest plot of the factors associated with absolute risk perception of breast cancer (HT¼Hormone Replacement Therapy).
menopausal and sexual symptoms [17,19] or immature mortality data [2,3] as important factors for refusing therapeutic prevention with SERMs. Predictors of low adherence also include early menopausal and sexual side effects during therapeutic prevention [33]. In our previous study [27], the most frequent reasons for trial entry were willingness to participate in a research program (60%), the need/desire to receive frequent medical care (58%) and the desire to contribute to medical knowledge (44%); whereas reasons for refusal included fear of medication abuse (33%), concern about adverse effects (31%), and physician advice against enrollment (24%). So the different uptake to the questionnaire reflects the lowest motivation to the research contribution in the decliner group. In addition, it should be taken into account that tamoxifen is still off label in Italy for prevention purposes. Other studies have underlined psychological factors such as worry about breast cancer and risk perception in influencing the
decision to participate or refuse preventive therapy trials [17,20,27]. This picture is similar to the issue of prescribing HT as a part of a prevention strategy for women at the onset of menopause [1]. Increased participation and retention to breast cancer therapeutic prevention as well as to HT is therefore likely to be the result of a better understanding of factors such as breast cancer worry and risk perception. In a previous study [27], we had shown that women who were very worried had a higher rate of refusal to participate into the trial. Given this counterintuitive finding, we searched in more depth which factors influenced worry and risk perception of breast cancer. The most striking finding from our study is the strong inverse association between early menopause and worry about breast cancer. This is in contrast with the known protective effect of a shorter life-time hormonal exposure of early menopause on breast cancer risk [34]. Considering the wide CI and the multiple testing, however, a false positive result cannot be excluded. This finding
G. Rondanina et al. / The Breast 34 (2017) 108e114
suggests that early menopause probably reflects a feeling of premature aging and higher vulnerability to disease and specifically to breast cancer. Interestingly, the borderline significant inverse association between HT use and worry about breast cancer is again in contrast with the clinical evidence of an increase risk of developing breast cancer during combined HT use [13]. A number of mechanisms may be advocated to explain this counter-intuitive relationship, including a favorable modulation of mood and well being by the HT as well as a delayed aging [25,35]. Worry about breast cancer was also significantly associated with increased risk perception and a trend to moderate or severe anxiety, two factors which may potentially contribute to avoidant behaviors towards health promotion programs. Indeed, recent studies have shown that women with high levels of worry and anxiety tend to adopt avoidant behaviors towards health promotion particularly if the cognitive construct of high risk perception makes them perceive the risk of disease very likely [36]. Thus, increasing risk perception and awareness in very worried women and high level of anxiety may be counterproductive if not accompanied by strategies of emotional improvement. For instance, Ferrer et al. [37] have shown that greater worry appeared to reverse the conventional relationship between self-efficacy and dietary restriction inasmuch as those who were self-efficacious and worried were less likely to restrict unhealthy foods. More importantly, among subjects higher in worry, higher levels of risk perception were associated with unhealthy life style such as lower vegetable consumption and exercise [38]. These results suggest the hypothesis that, among people high in worry, attempts to increase risk perception without a concomitant strategy of emotional coping could be counterproductive because of avoidant attitudes. It is also true that extra screening is more likely to be acceptable than taking a medication such as tamoxifen, as reflected by the main reasons for refusal in our study. Interestingly, we found that former smokers were more worried than never smokers, whereas current smokers were not significantly more worried than never smokers. While smoking is only mildly associated with increased breast cancer risk [39], former smokers have a more in-depth cognitive processing of risk information than current smokers, who in fact tend to minimize risk through a reduction dissonance process [40]. Also, smokers hold more pessimistic and avoidant beliefs about cancer, which have been shown to deter early-detection and prevention behaviors [41]. Taken together, all the above mentioned findings support the notion of the affective nature of worry. Conversely, the predominant cognitive nature of risk perception is underlined by the strong associations with biological factors, including older age and unhealthy life style, although the affective component of risk perception is illustrated by the association with worry and depression. The strong association between worry and risk perception describes the strict interdependence between the two constructs which are engaged in the decision making process, where the affective factors act directly to respond to internal needs and promote action, whereas the cognitive factors provide the different options to plan a decision program [42]. This strict crosstalk suggests that for any prevention plan to become effective health care providers should focus not only on risks and benefits of treatments but also to improve the understanding of the affective factors implicated in the decision making process of health behavior changes. Increasing risk perception in women who are very worried and anxious might in fact determine avoidant behaviors and be counterproductive [37]. Our study population had a prevalence of psychological disorders which is in line with the expected figures in the same age range [43,44] as only 12% of our population reported to be worried or very worried of breast cancer, and approximately 20% reported
113
their absolute perceived risk of breast cancer to be likely or very likely or to be higher relative to a woman of the same age. Also, only 17% and 13% of our women had moderate to severe anxiety or depression, respectively. All women in our study were assessed before trial acceptance or refusal and before allocation to tamoxifen or placebo. Also, 75% of our study population were at increased risk for breast cancer according to the Gail model adjusted for the Italian incidence data. In addition, most women in the study were taking combined HT, a known risk factor for breast cancer, so our population was mostly composed of high risk women based on the Gail model and the use of combined HT. Therefore, our findings are applicable to a vast population of midlife women at increased risk for breast cancer who are willing to receive HT for menopausal disorder relief or health promotion but are concerned about breast cancer risk. In conclusion, our findings show that the affective construct of worry about breast cancer is different from the cognitive construct of risk perception, which is more adherent to biological factors. A better understanding of the emotional factors that modulate the decision making process can potentially improve adherence to HT use and breast cancer therapeutic prevention programs. Clinical trials are underway to assess the long term efficacy and safety of low dose tamoxifen in midlife women on HT (Clinical Trials.gov NCT01579734) or because of prior intraepithelial neoplasia (ClinicalTrials.gov NCT01357772). Financial support Supported by the Italian Foundation for Cancer Research (A. DeCensi), the Italian Association for Cancer Research (AIRC) (A. DeCensi), Avon Italia (A. DeCensi), the Italian League against Cancer (LILT project number 51/2005) (A. DeCensi), the American Italian Cancer Foundation, the Italian Ministry of Health, RFPS-2006-1339898 (A. DeCensi). Tamoxifen and placebo were donated by FIDIA Farmaceutici S.p.a, Abano Terme, Italy. We are grateful to the study participants for their heart and energy. Disclosure of conflicts of interest The authors declare that there are no conflicts of interest. References [1] Manson JE, Kaunitz AM. Menopause managementegetting clinical care back on track. N Engl J Med 2016;374:803e6. [2] Cameron DA. Breast cancer chemoprevention: little progress in practice? Lancet 2014;383. 10e1020. [3] Narod SA. Tamoxifen chemoprevention-end of the road? JAMA Oncol 2015;1: 1033e4. [4] DeCensi A, Thorat MA, Bonanni B, Smith SG, Cuzick J. Barriers to preventive therapy for breast and other major cancers and strategies to improve uptake. Ecancermedicalscience 2015;9:595. [5] Cuzick J, Sestak I, Bonanni B, Costantino JP, Cummings S, DeCensi A, et al. Selective oestrogen receptor modulators in prevention of breast cancer: an updated meta-analysis of individual participant data. Lancet 2013;25: 1827e34. s-Martínez JE, Cheung AM, Chlebowski RT, Wactawski[6] Goss PE, Ingle JN, Ale Wende J, et al. Exemestane for breast-cancer prevention in postmenopausal women. N Engl J Med 2011;364:2381e91. [7] Cuzick J, Sestak I, Forbes JF, Dowsett M, Knox J, Cawthorn S, et al. Anastrozole for prevention of breast cancer in high-risk postmenopausal women (IBIS-II): an international, double-blind, randomized placebo-controlled trial. Lancet 2014;383:1041e8. [8] Dunn BK, Cazzaniga M, DeCensi A. Exemestane: one part of the chemopreventive spectrum for ER-positive breast cancer. Breast 2013;22:225e37. [9] Waters EA, McNeel TS, Stevens WM, Freedman AN. Use of tamoxifen and raloxifene for breast cancer chemoprevention in 2010. Breast Cancer Res Treat 2012;134:875e80. [10] Smith SG, Sestak I, Forster A, Partridge A, Side L, Wolf MS, et al. Factors affecting uptake and adherence to breast cancer chemoprevention: asystematic review and meta-analysis. Ann Oncol 2016;27:575e90.
114
G. Rondanina et al. / The Breast 34 (2017) 108e114
[11] Fisher B, Costantino JP, Wickerham DL, Cecchini RS, Cronin WM, Robidoux A, et al. Tamoxifen for prevention of breast cancer: report of the national surgical adjuvant breast and bowel project P-1 study. J Natl Cancer Inst 1998;16(90): 1371e88. [12] Jewett PI, Gangnon RE, Trentham-Dietz A, Sprague BL. Trends of postmenopausal estrogen plus progestin prevalence in the United States between 1970 and 2010. Obstet Gynecol 2014;124:727e33. [13] Rossouw JE, Anderson GL, Prentice RL, LaCroix AZ, Kooperberg C, Stefanick ML, et al. Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women's Health Initiative randomized controlled trial. JAMA 2002;288:321e33. [14] Manson JE, Chlebowski RT, Stefanick ML, Aragaki AK, Rossouw JE, Prentice RL, et al. Menopausal hormone therapy and health outcomes during the intervention and extended poststopping phases of the Women's Health Initiative randomized trials. JAMA 2013;310:1353e68. [15] Daubisse-Marliac L, Delafosse P, Boitard JB, Poncet F, Grosclaude P, Colonna M. Breast cancer incidence and time trend in France from 1990 to 2007: a population-based study from two French cancer registries. Ann Oncol 2011;22:329e34. [16] Antoine C, Ameye L, Paesmans M, de Azambuja E, Rozenberg S. Menopausal hormone therapy use in relation to breast cancer incidence in 11 European countries. Maturitas 2016;84:81e8. [17] Bober SL, Hoke LA, Duda RB, Regan MM, Tung NM. Decision-making about tamoxifen in women at high risk for breast cancer: clinical and psychological factors. J Clin Oncol 2004;22:4951e7. [18] Ropka ME, Keim J, Philbrick JT. Patient decisions about breast cancer chemoprevention: a systematic review and meta-analysis. J Clin Oncol 2010;28: 3090e5. [19] Bastian LA, Lipkus IM, Kuchibhatla MN, Weng HH, Halabi S, Ryan PD, et al. Women's interest in chemoprevention for breast cancer. Arch Intern Med 2001;161:1639e44. [20] Holmberg C, Waters EA, Whitehouse K, Daly M, McCaskill-Stevens W. My lived experiences are more important than your probabilities: the role of individualized risk estimates for decision making about participation in the study of tamoxifen and raloxifene (STAR). Med Decis Mak 2015;35:1010e22. [21] Katapodi MC, Lee KA, Facione NC, Dodd MJ. Predictors of perceived breast cancer risk and the relation between perceived risk and breast cancer screening: a meta-analytic review. Prev Med 2004;38:388e402. [22] Decensi A, Galli A, Veronesi U. HRT opposed to low-dose tamoxifen (HOT study): rationale and design. Recent Results Cancer Res 2003;163(104e11): 264e6. discussion. [23] DeCensi A, Bonanni B, Maisonneuve P, Serrano D, Omodei U, Varricchio C, et al. Italian HOT Study Group. A phase-III prevention trial of low-dose tamoxifen in postmenopausal hormone replacement therapy users: the HOT study. Ann Oncol 2013;24:2753e60. [24] Veronesi U, Maisonneuve P, Rotmensz N, Bonanni B, Boyle P, Viale G, et al. Tamoxifen for the prevention of breast cancer: late results of the Italian Randomized Tamoxifen Prevention Trial among women with hysterectomy. J Natl Cancer Inst 2007;99:727e37. [25] Pinkerton JV, Pan K, Abraham L, Racketa J, Ryan KA, Chines AA, et al. Sleep parameters and health-related quality of life with bazedoxifene/conjugated estrogens: a randomized trial. Menopause 2014;21:252e9. [26] Harvey JA, Pinkerton JV, Baracat EC, Shi H, Chines AA, Mirkin S. Breast density changes in a randomized controlled trial evaluating bazedoxifene/conjugated estrogens. Menopause 2013;20:138e45.
[27] Rondanina G, Puntoni M, Severi G, Varricchio C, Zunino A, Feroce I, et al. Psychological and clinical factors implicated in decision making about a trial of low-dose tamoxifen in hormone replacement therapy users. J Clin Oncol 2008;26:1537e43. [28] Decensi A, Gandini S, Serrano D, Cazzaniga M, Pizzamiglio M, Maffini F, et al. Randomized dose-ranging trial of tamoxifen at low doses in hormone replacement therapy users. J Clin Oncol 2007;25:4201e9. [29] Lerman C, Daly M, Sands C, Balshem A, Lustbader E, Heggan T, et al. Mammography adherence and psychological distress among women at risk for breast cancer. J Natl Cancer Inst 1993;85:1074e80. [30] Bjelland I, Dahl AA, Haug TT, Neckelmann D. The validity of the Hospital Anxiety and Depression Scale: an updated literature review. J Psychosom Res 2002;52:69e77. [31] Guerrieri-Gonzaga A, Robertson C, Bonanni B, Serrano D, Cazzaniga M, Mora S. Preliminary results on safety and activity of a randomized, double-blind, 2 x 2 trial of low-dose tamoxifen and fenretinide for breast cancer prevention in premenopausal women. J Clin Oncol 2006;24:129e35. [32] Meyskens Jr FL, Mukhtar H, Rock CL, Cuzick J, Kensler TW, Yang CS, et al. Cancer prevention: obstacles, challenges and the road ahead. J Natl Cancer Inst 2015;108. pii: djv309. [33] Land SR, Walcott FL, Liu Q, Wickerham DL, Costantino JP, Ganz PA. Symptoms and QOL as predictors of chemoprevention adherence in NRG oncology/ NSABP trial P-1. J Natl Cancer Inst 2015;108. pii: djv365. [34] Collaborative Group on Hormonal Factors in Breast Cancer. Menarche, menopause, and breast cancer risk: individual participant meta-analysis, including 118 964 women with breast cancer from 117 epidemiological studies. Lancet Oncol 2012;13:1141e51. [35] Welton AJ, Vickers MR, Kim J, Ford D, Lawton BA, MacLennan AH, et al. Health related quality of life after combined hormone replacement therapy: randomised controlled trial. BMJ 2008;337:a1190. [36] Persoskie A, Ferrer RA, Klein WM. Association of cancer worry and perceived risk with doctor avoidance: an analysis of information avoidance in a nationally representative US sample. J Behav Med 2014;37:977e87. [37] Ferrer RA, Bergman HE, Klein WM. Worry as a predictor of nutrition behaviors: results from a nationally representative survey. Health Educ Behav 2013;40:88e96. [38] Ferrer RA, Portnoy DB, Klein WM. Worry and risk perceptions as independent and interacting predictors of health protective behaviors. J Health Commun 2013;18:397e409. [39] Gaudet MM, Gapstur SM, Sun J, Diver WR, Hannan LM, Thun MJ. Active smoking and breast cancer risk: original cohort data and meta-analysis. J Natl Cancer Inst 2013;105:515e25. [40] Kaufman AR, Koblitz AR, Persoskie A, Ferrer RA, Klein WM, Dwyer LA, et al. Factor structure and stability of smoking-related health beliefs in the national lung screening trial. Nicot Tob Res 2016;18:321e9. [41] Quaife SL, McEwen A, Janes SM, Wardle J. Smoking is associated with pessimistic and avoidant beliefs about cancer: results from the international cancer benchmarking partnership. Br J Cancer 2015;112:1799e804. € berg L. Worry and risk perception. Risk Anal 1998;18:85e93. [42] Sjo [43] Tangen T, Mykletun A. Depression and anxiety through the climacteric period: an epidemiological study (HUNT-II). J Psychosom Obstet Gynaecol 2008;29: 125e31. [44] Burt VK, Altshuler LL, Rasgon N. Depressive symptoms in the perimenopause: prevalence, assessment, and guidelines for treatment. Harv Rev Psychiatry 1998;6:121e32.