Abstract
Keywords
The shock of diagnosis, existential concerns, the burden of treatment and changes in body and self-image present a substantial challenge to adaptation and coping for women with breast cancer. For many of them, the emotional distress reaches the level of a diagnosable psychiatric disorder.
Few studies have utilized structured psychiatric interviews to assess psychiatric morbidity in women with breast cancer [1], [2] and none have compared morbidity rates in early stage versus advanced disease [3–7]. Although body image is better preserved after conservation surgery [8], [9], studies of women undergoing lumpectomy compared with those experiencing mastectomy have not shown differences in morbidity [10], [11], highlighting, instead, the existential impact of the diagnosis itself. Thus high rates of morbidity persisting across the course of the illness would not be surprising.
Three hundred and three women were recruited (between October 1994 and March 1997) to a randomised controlled trial of cognitive-existential group therapy for women receiving adjuvant chemotherapy for early stage breast cancer [12]. A further 200 women, with advanced breast cancer, were recruited (between March 1996 and July 2001) into a randomised controlled trial of the effect of supportive-expressive group therapy on quality of life and survival. The baseline psychiatric morbidity and quality of life data of these two samples was compared, to gain insight into the relative levels of morbidity found at differing stages of the illness trajectory.
Method
Sample
Inclusion criteria for the early stage patients were a recent diagnosis of stage II (TNM classification, T1 N1 M0; T2 N1 M0; T2 N0 M0) breast cancer [13] or stage I (T1 N0 M0) breast cancer and a poor prognostic factor (hormone receptor status or histology grade such that adjuvant chemotherapy was recommended). Exclusion criteria were age over 65 years, geographical inaccessibility, a prior diagnosis of cancer (except basal cell skin cancer), minimal English, dementia, psychosis and intellectual disability.
The comparable inclusion criterion for women with advanced breast cancer was the diagnosis of stage IV disease (using the TNM system) [13]; exclusion criteria were age over 70 years, geographical inaccessibility, a previous diagnosis of cancer (except basal cell skin cancer), minimal English, dementia, psychosis and intellectual disability.
Design
The two multicentre trials were approved by the ethics committees of all participating institutions; both samples were recruited from the oncology services in seven general, metropolitan hospitals and a range of private practitioners. Research staff consulted with clinicians to confirm eligibility prior to approaching the women to obtain informed consent. The staff then conducted a structured psychiatric interview and administered self-report questionnaires covering psychosocial state and quality of life. Only the preintervention aspects of these studies are covered in this report.
Measures
The following measures were used:
1. The Monash Interview for Liaison Psychiatry (MILP) [14], a structured psychiatric interview for use in the medically ill.
The MILP enables standardized DSM-IV [15] psychiatric diagnoses to be made covering mood, anxiety, somatizing and substance abuse disorders. Interrater reliability and procedural validity have been well demonstrated [14].
2. The European Organization for Research and Treatment of Cancer (EORTC) quality of life questionnaire QLQ-C30 [16] and the EORTC QLQ-BR23 breast module [17].
The QLQ-C30 is a validated, self-report measure of quality of life covering physical, role, cognitive, emotional and social domains; the QLQ-BR23 module adds the domains of body image, sexual functioning, perspectives on the future, arm and breast symptoms, and side-effects of cancer treatments. Responses are rated on a four-point scale as ‘not at all’, ‘a little’, ‘quite a bit’ or ‘very much’.
3. Other self-report scales for psychological studies.
The scales used were the Hospital Anxiety and Depression Scale (HADS) [18], the Affects Balance Scale (ABS) [19], the Mental Adjustment to Cancer (MAC) Scale [20] and the Medical Coping Modes Questionnaire (MCMQ) [21]. All are well validated and commonly used.
Statistical analysis
The Statistical Package for the Social Sciences (SPSS for Windows) [22] was used to analyze the data. Confidence intervals (95%) for diagnostic rates were computed using the CIA computer program [23] (which employs the method of Wilson [24], as recommended by Newcombe and Altman [25]). Pearson correlations and two-tailed t-tests were utilized to examine factors associated with psychiatric disorders and quality of life [22].
Results
Samples
Of 491 eligible women approached for the early breast cancer study, 303 were enrolled, a response rate of 62%. This contrasts with a response rate of 46% for the metastatic breast cancer study where 200 out of 440 eligible women agreed to participate. Reasons for refusal are shown in Table 1. Given that the studies were of group therapies, not liking group approaches was only identified by approximately 10% of refusers in each study. Although similar proportions of women in each trial identified themselves as coping satisfactorily, or being well supported, particular reasons for the lower recruitment rate in the advanced cancer study related to the burden of medical treatments (18%) and practical impediments such as transport difficulties (15%).
Response rates and reasons for refusal to participate in trials for women with early and advanced breast cancer
The mean age of the women was 46 years (SD = 8) in the early breast cancer sample and 51 years (SD = 9) in the metastatic breast cancer sample. Most of the women in both of the cohorts were married, Australian-born and educated to senior secondary school or beyond. Nearly half (147 49%) of the women with early stage cancer were in paid employment and 49 (16%) were not working due to illness. Of the women with advanced cancer just over a third (69 35%) were working while a quarter 48 (24%) were too ill to work. We could not obtain sociodemographic profiles of study refusers.
Interviews for the early stage women took place a mean 102 days (SD = 56) after breast surgery at primary diagnosis. For the metastatic cohort, interviews occurred a mean 63 months (SD = 44) after initial diagnosis (excluding women who were stage IV at initial breast cancer diagnosis) and a mean of 10 months (SD = 12) after metastatic diagnosis. Excluding those women who were stage IV at diagnosis (33 16%), the mean duration from diagnosis to relapse for the metastatic sample was 52 months (SD = 41).
In the metastatic sample, local recurrence had occurred in 37 (18%) and regional recurrence in 18 (9%) of the women (in addition to their distant metastases). Metastases were visceral in 106 (53%) and nonvisceral in 94 (47%); the sites were bone in 138 (69%), lung in 64 (32%), liver in 63 (31%), supraclavicular node in 37 (18%), brain in 4 (2%), skin in 4 (2%) and other in 16 (8%). Many women had more than one site of metastatic disease with 33 (16%) having more than three sites.
With respect to cancer treatments, mastectomy as a primary surgical intervention was more common than lumpectomy for the advanced disease group, consistent with higher numbers of larger tumours and the presence of a subsample with stage IV cancer at initial diagnosis. Most of the women with early stage breast cancer had chemotherapy (as per the eligibility criteria for the study). Of the women with metastatic cancer, 103 (51%) had chemotherapy at the primary stage of their disease plus a range of further hormonal therapy, chemotherapy, and radiotherapy in response to their advanced disease. Specific hormonal therapies at relapse included tamoxifen in 93 women (46%); medroxyprogesterone acetate in 27 (13%); anastrozole in 15 (8%); aminoglutethimide in 15 (8%); and letrozole in 13 (7%). Chemotherapy for recurrence comprised doxorubicin/cyclophosphamide in 54 women (27%); epirubicin/cyclophosphamide in 22 (11%); cyclophosphamide/ methotrexate/fluorouracil in 20 (10%); docetaxel in 20 (10%); paclitaxel in 16 (8%); high dose chemotherapy with stem cell rescue in eight (4%); vinorelbine in five (3%) and capecitabine in four (2%). While the majority (167 84%) needed no further surgery, the procedures occurring, following recurrence, were local excisions in nine women (5%); skeletal repairs in seven (3.5%); bilateral oophorectomy in six (3%); pleurodesis in three (1.5%) and neurosurgery in two (1%).
Psychiatric disorders
No statistically significant differences were found between the rates of psychiatric disorders in the two cohorts (see Table 2). A current DSM-IV psychiatric disorder was diagnosed in 135 (45%) women with early stage cancer and in 83 (42%) of their metastatic counterparts. A proportion of each group (early: 20%; metastatic: 8%) showed comorbidity. Adjustment disorder (including anxious and depressed moods) was the most common psychiatric diagnosis in early stage women (8728.7%) as well as in the advanced group (5326.5%). Nicotine dependence was the most common second diagnosis in both samples.
Comparison of psychiatric disorders for women with early and advanced breast cancer†
Depression (major depression, dysthymia or adjustment disorder) was diagnosed in 111 (37%) of women with early stage disease and 62 (31%) of the women with advanced disease. Furthermore, of those early stage breast cancer patients with a current diagnosis of depression, 35 (31.5%) had a past history of this condition and this association was significant (p < 0.001). Of those patients in the advanced group with a current depressive disorder, 22 (35.5%) had a positive past history (p = 0.002). Thus rates of a past history of depression were comparable.
Simple phobic disorders included a needle phobia. Patients with posttraumatic stress disorders had suffered from longstanding problems. Stressors identified included a child's death, fire, assaults and motor car accidents. (Respondents were asked to identify a stressor, but this did not include the impact of the breast cancer itself.)
Factors associated with psychiatric morbidity
Factors, from quality of life and other self-report scales, which are significantly associated with depression and anxiety, in both cohorts, are shown in Table 3. Additionally, while age was not significantly related to anxiety or depression for the women with early stage breast cancer, younger age was associated with depression in women with metastatic disease (mean age of depressed women, 47.6 years [SD = 8.3] and of non-depressed women, 53.2 years [SD = 8.9], t-test, p = 0.000). Marital status, hormone receptor status, type of surgery, chemotherapy, hormonal therapy or radiotherapy were not associated with a diagnosis of depression or anxiety in either sample.
The factors from quality of life and self-report scales which were significantly associated with a DSM-IV diagnosis of a depressive or anxiety disorder for women with early and advanced breast cancer†
Quality of life
On the two EORTC quality of life measures, the responses were dichotomised with ‘none’ and ‘a little’ combined as ‘negligible’, and ‘quite a bit’ and ‘very much’ combined as ‘substantial’(see Table 4). Hair loss through chemotherapy caused less distress in the advanced disease patients, who experienced more hot flushes and were more dissatisfied with body image. The prevalence of lymphoedema increased with the duration of the illness.
Comparative quality of life responses on the European Organization for Research and Treatment of Cancer QLQ-C30 and BR23 for women with early and advanced breast cancer
Both cohorts carried very high rates of low libido as well as negligible enjoyment when sexually active. Higher educational level (p < 0.01) and younger age (p < 0.01) correlated with higher sexual functioning scores for the early stage women. In addition, for this group, significantly better body image scores were associated with breast conservation surgery (p < 0.01), marital status (p < 0.01) and older age (p < 0.001). In contrast, for women with advanced disease, the type of surgery was not associated with body image scores on the EORTC QLQ-BR23, although younger women rated their body image more poorly than did older women (p < 0.01).
Discussion
Excluding non-melanocytic skin cancers, breast cancer is the most common cancer diagnosed and the most common cause of cancer death among Australian women [26]. It has been estimated to effect one in 11 Australian women who live to the age of 75, and its incidence continues to increase by an average of 1.5% per year [27]. Accordingly, the implications of research findings in the breast cancer population have the potential to impact on a large number of women. Women may live with breast cancer for many years and attention to their psychosocial concerns, alongside their physical treatments, is critical to their quality of life.
In 1997, women in the Australian community had a 12 month prevalence of major depression of 6.8%, dysthymia 1.3%, anxiety disorder 12.1% and substance abuse disorder 4.5% [28]. The overall level of psychiatric morbidity in this sample was 18% [28]. However, our groups of women with early and advanced breast cancer had substantially greater morbidity (45% and 42%, respectively). These findings have major relevance for clinical services seeking to support women living with such illnesses.
At the time when the initial shock and existential threat of diagnosis first confronts patients, the women with early breast cancer, in our study, had a rate of major depression (9.6%) which is higher than community norms. This rate is also remarkably similar to rates found in two British studies that used the Present State Examination (a structured interview) 3 months after surgery [1], [2]. Following metastatic recurrence and associated treatment, for women in our study, the proportion with major depression (6.5%) resumed a community-like level. Interestingly, the rate of psychiatric disorder is not more prevalent in women with advanced disease despite the existential challenges that could accompany this illness. The ongoing surveillance of patients during management of their metastatic breast cancer may account for this diminished rate. Indeed, certain forms of distress (for instance, those that are associated with hair loss) decrease, possibly as patients adjust to the reality of illness. Nonetheless, global psychiatric morbidity remains high, probably reflecting the stress of living with a serious illness; hence the high prevalence of adjustment disorder which afflicts one-quarter of patients across the illness trajectory in our samples. These women have to adjust to many losses which may affect their mood, but anxiety about what the future may hold seems to be a less apparent initial consequence of diagnosis with primary breast cancer, or with metastases. Such conditions represent a substantial clinical need which may not subside without appropriate intervention.
Cognitive attitudes such as hopelessness, helplessness or resignation to one's fate are significantly associated with depression. As such, these predisposing factors may be useful in alerting health care professionals to potential psychomorbidity. Similarly, our finding that current depression is significantly associated with a past history of depression suggests that some women are more vulnerable by virtue of their past history. Having a depressed mood may interfere with a patient's coping, or even complying, with treatment regimens, thus creating a further barrier to dealing with the cancer. Early detection and management of altered moods is critical.
A measure of caution is necessary in interpreting our data. They have been derived from prepsychological treatment cohorts who had consented to participate in randomised controlled trials of group therapy. As such, our eligibility criteria, recruitment rate and non-random sampling design are sources of bias. One in three women who refused to join the metastatic breast cancer study cited treatment, health or practical reasons (compared with none in these categories in the early stage sample). Those refusers, who may have been experiencing major effects from their illness or lacking in support, could have had depression that is not reflected in our prevalence rates. However, those who refused to participate had not given informed consent, so clinical information was not accessible in relation to their physical health or their psychological wellbeing. Despite not being designed as a prevalence study per se, the size of these two convenience samples bolsters confidence that the findings are clinically relevant. Additionally, the consistency of the 95% confidence intervals for psychiatric diagnoses is noteworthy. Furthermore, our adoption of a standardized psychiatric interview (which has been shown to be able to detect minor mood disorders with greater sensitivity than some other structured approaches [14]) has provided strength in the psychiatric assessment that is not possible with dimensional mood scales alone.
Breast cancer and its treatment inevitably affect quality of life. Fatigue and emotional functioning influence the likelihood of psychiatric disorders in these patients. Moreover, the considerable impact on selfimage and substantial decrease in sexual functioning provide challenges to self-esteem and intimate relationships at a time when personal support is critical. Decreased libido and sexual enjoyment may not be surprising given the impact of breast cancer and the concomitant treatment; however, the enhancement of relationships through attention to intimacy and affection warrants encouragement. Clinicians need to be watchful for these concerns and supportive of women struggling to adapt to such potentially devastating changes.
We suggest that our findings reinforce the best practice guidelines endorsed by Australia's National Health and Medical Research Council for the management of early stage [29] and advanced [30] breast cancer, as well as integral psychosocial care [31]. In short, dedicated support by a multidisciplinary team is essential, at all stages of the disease, to assuage the chronic morbidity associated with breast cancer.
Conclusion
Psychiatric morbidity is clearly a significant aspect of the experience of women with breast cancer, both in the early and advanced stages of the disease. Comprehensive supportive care is not yet uniformly available to women who have the misfortune to develop breast cancer. Clinical services therefore face the challenge to respond with appropriate programs of psychosocial treatments. The increasing sophistication in chemotherapy, radiotherapy and hormone therapy must be matched by innovative therapies to help women who 326 PSYCHIATRIC DISORDER IN BREAST CANCER suffer psychologically as part of their encounter with breast cancer.
Footnotes
Acknowledgements
The National Health and Medical Research Council of Australia, Cancer Council of Victoria, Pratt Foundation and the Kathleen Cuningham National Breast Cancer Foundation funded the studies. We thank the clinicians who facilitated access to patients; Mary Harvey, Jill Ikin, Caroline Long, Nadia Ranieri and Kathryn Watty who assisted with recruitment; Dean McKenzie for statistical advice.