Psychologic changes after a gynecologic cancer

Blackwell Science, LtdOxford, UKJOGThe Journal of Obstetrics and Gynaecology Research1341-80762005 Asia and Oceania Federation of Obstetrics and GynaecologyApril 2005312152157Original ArticlePsychologic changes after gynecologic cancerR. W. Petersen et al. J. Obstet. Gynaecol. Res. Vol. 31, No. 2: 152–157, April 2005 Psychologic changes after a gynecologic cancer Rodney W. Petersen, Gaynor Graham and Julie A. Quinlivan Department of Obstetrics and Gynecology, The University of Melbourne, Royal Women’s Hospital, Melbourne, Australia Abstract Aim: The aim of this study was to explore the wider psychologic symptomatology experienced by women with a new diagnosis of a gynecologic cancer at the point of diagnosis and 6 weeks later. Methods: A prospective cohort study was carried out with ethics committee approval and informed consent. Women were recruited from three tertiary hospitals in Australia over an 8-month period. In order to cover a diverse range of potential symptomatology, we utilized the Hopkins Symptom Checklist (HSCL)-90, which covers 90 separate psychologic symptoms that can then be coded into eight domains. Women also completed questions relating to their perceived level of social support, and demographic data were collated separately. Results: Key findings were that levels of symptomatology remained uniform across the first 6 weeks following the diagnosis of the cancer regardless of the site of the cancer. Across the spectrum of symptomatology domains, the median scores were all higher in women with poor social supports compared with those with higher social support levels at 6 weeks. Statistically significant differences were observed in the domains of phobic-anxiety, retarded depression, and agitated depression at 6 weeks’ follow up. Conclusion: Women with a new diagnosis of a gynecologic cancer experience diverse psychologic symptomatology. Symptoms persist over the first 6 weeks and are higher in women with poor social supports. Screening of women for adequate social support structure and targeted interventions to resolve symptomatology need to be tailored to the type of symptoms experienced. Key words: anxiety, depression, endometrial cancer, Hopkins Symptom Checklist-90, ovarian cancer. Introduction Although several studies have examined the prevalence of anxiety and depression in women with gynecologic cancer, a recent review concluded that many were poorly designed and that therefore the prevalence of psychologic symptomatology in this population was poorly defined.1 Furthermore, most studies focused exclusively on symptoms of anxiety and depression, and failed to address other possible domains of symptomatology that can emerge under duress, such as obsessive–compulsive traits, somatization or even psychotism. In the 13 available studies, the prevalence of depression ranged from 4 to 90%, indicating levels of symp- tomatology that are both lower and higher than population norms.1 In contrast, levels of anxiety were uniformly reported to be higher than levels reported for non-cancer population norms. Data on more diverse symptomatology and the relation between psychologic symptoms and its effect upon satisfaction with the health care provided, the ability to cope with stress and the influence upon survival were not available.1 The lack of data regarding the explicit nature of psychologic symptomatology is reflected in the lack of randomized trials that have explored the efficacy of strategies to reduce levels of adverse psychologic symptomatology in women with a new diagnosis of a gynecologic cancer. Only three randomized trials exist, Received: August 27 2004. Accepted: December 8 2004. Reprint request to: Dr Rodney Petersen, University Department of Obstetrics and Gynecology, Royal Women’s Hospital, 132 Grattan St, Carlton, Melbourne, Victoria 3053, Australia. Email: r.petersen@unimelb.edu.au 152 Psychologic changes after gynecologic cancer with varying reports of efficacy.2–4 This is in stark contrast to the plethora of randomized data dealing with women with breast cancer. It is important to detect and treat any psychologic symptoms, as they independently cause distress to the woman and decrease quality of life.2,5 As treatments improve and survival increases, quality of life issues assume increasing importance in management planning. The aim of this study was to explore the wider psychologic symptomatology experienced by women with a new diagnosis of a gynecologic cancer at the point of diagnosis and 6 weeks later. In order to cover a diverse range of potential symptomatology, we utilized the Hopkins Symptom Checklist (HSCL)-90.6 This questionnaire covers 90 separate psychologic symptoms that can then be coded into eight domains of symptomatology. Secondary exploratory hypotheses were that women with a diagnosis of ovarian cancer and those with poor perceived social supports might report higher and more persistent levels of symptomatology and might benefit from additional resources compared with other women. Methods A prospective cohort study was carried out with multi-institutional ethics committee approval and informed patient consent in accordance with the standards of the Australian National Health and Medical Research Council. Patients were recruited from three tertiary hospitals in Australia that provided gynecologic oncology services over an 8-month period. The trial population included women with a new diagnosis of gynecologic cancer who were English-speaking, did not have a previous psychiatric history, and who had undergone surgery as their primary form of treatment. Women were provided with information sheets in the postoperative period and were invited to participate in the study. Once consent was obtained, the baseline questionnaire was completed before discharge. The questionnaire used was the HSCL-90.6 At this point only, patients also completed questions relating to their perceived level of social support, and of the time periods between onset of symptoms and presentation for investigation, diagnosis and treatment. Demographic data and tumor information were collated independently after discharge. Women were provided with all normal postoperative support, counseling and information services provided by the respective hospitals and their multidisciplinary care teams. Six weeks following surgery, at the time of first outpatient review, women again completed only the HSCL-90 questionnaire, which was distributed by clinic staff who were independent to the study investigators. The study investigators were blinded to the baseline results. Outcome measures The primary outcome measure was the difference between the baseline and the follow-up HSCL-90 subscale scores.6 The HSCL-90 is a patient self-rating symptom inventory, originally devised by researchers at the Johns Hopkins University in the 1950s. The initial questionnaire had 41 questions; however, it has been developed over the years and now contains 90 questions. The questionnaire has been extensively reviewed as a tool for measuring clinical change and is both valid and reliable. There are eight dimensions of symptomatology, which are called domains: • Interpersonal Sensitivity (domain 1), for example feeling critical of others, feeling inferior to others. • Phobic-anxiety (domain 2), for example being afraid of open spaces, being afraid to travel on public transport. • Retarded Depression (domain 3), for example loss of sexual interest, thoughts of death or dying. • Anger–Hostility (domain 4), for example easily annoyed, urges to injure people. • Somatization (domain 5), for example soreness of muscles, heavy arms and legs. • Obsessive–Compulsive (domain 6), for example check and double-check things, unwanted thoughts. • Agitated Depression (domain 7), for example nervousness, crying easily. • Psychoticism (domain 8), for example controlling thoughts, hearing voices. Each domain has a series of questions related to that dimension interspersed within the 90 questions. The questionnaire asked subjects to answer questions based on how much that problem has bothered them in the last 7 days including today. Answers are recorded by circling a number between 0 and 4 where 0 = not at all, 1 = a little bit, 2 = moderately, 3 = quite a bit, and 4 = extremely. Secondary outcomes included the impact of tumor type and perceived social support on outcome variability. 153 R. W. Petersen et al. Data collection and entry Once the questionnaires were completed, the responses were coded to produce a result for each domain. Demographic data on cancer site, stage, grade, patient age and perceived social support were added to the statistical database. Perceived Social Support was assessed using a Likert scale. Each woman was asked by a medical staff member to rate her level of social support along a scale of 1–5, where 1 = very poor, 3 = average and 5 = excellent. Statistical analysis Descriptive statistics were applied to evaluate outcomes. As data relating to HSCL-90 domains were not normally distributed, results are presented as median and interquartile range (IQR; 25th, 75th). Comparisons of baseline and follow-up paired data were compared using the Wilcoxon signed-rank test, and data related to domain outcomes in women with different tumor types and social support levels were analyzed using the Mann–Whitney U-test. A P-value of 0.05 was considered significant. Results Of 29 women approached to participate, consent was obtained from 27. One woman subsequently moved interstate. Therefore complete paired data were available for 26 women. The average age of the women in the study was 61 years (SD 13 years). Of the 26 patients, 10 had endometrial, nine ovarian, three cervix, and four other gynecologic cancers. The majority were stage 1 (61%), with 12% and 27% having stage 2 and 3 cancers, respectively. There were no stage 4 cancers. More than half the tumors were well differentiated (54%), with 23% being moderately and 23% poorly differentiated. Women had presented in diverse ways for medical care; the majority presented following postmenopausal bleeding (38%), but others with abdominal distension (15%), following routine screening(8%), with pelvic discomfort (8%) and post-coital or intermenstrual bleeding (8%). The median time from onset of symptomatology to presentation for medical care was 60 days (IQR 30–98). The median time from onset of symptoms to diagnosis was 62 days (IQR 42–134). Most women had waited only 15 days following presumptive diagnosis until achieving a definitive diagnosis and receiving treatment initiation. All women received surgical treatment, which included transabdominal hysterectomy and bilateral salpingooophorectomy (62%), pelvic lymph node dissection (58%), omentectomy (42%), or modified or radical hysterectomy (19%). Other procedures such as large bowel resection, vulval surgery, or groin node dissection were carried out less frequently. Overall, 73% of women required adjuvant therapy following surgery. Although most women reported high levels of support (median 5 of 5 [IQR 4–5]), nearly one-quarter reported low levels of support (scores £3). Table 1 summarizes the baseline and follow-up HSCL-90 data for the cohort. Of interest, levels of symptomatology did not markedly alter across the first 6-weeks. Table 2 summarizes the time-course changes observed in women with ovarian cancer as compared with women with other cancer types across the time span. There were no significant differences in the levels of symptomatology at baseline or 6 weeks as a result of the cancer site. Table 3 summarizes the levels of psychologic symptomatology reported by women with high perceived social support levels (Likert score >3) compared with those with poor perceived social support (Likert score £3) at baseline and 6 weeks. The median score for each domain was higher in women with poor supports Table 1 Hopkins Symptom Checklist-90 data at baseline and 6 weeks. Data are presented as median and interquartile (25th, 75th) ranges Domain Baseline median (IQR) 6-weeks median (IQR) P-value Interpersonal Sensitivity (domain 1) Phobic-anxiety (domain 2) Retarded Depression (domain 3) Anger–Hostility (domain 4) Somatization (domain 5) Obsessive–Compulsive (domain 6) Agitated Depression (domain 7) Psychoticism (domain 8) 0.14 (0–0.43) 0.11 (0–0.22) 0.42 (0.09–0.83) 0.21 (0–0.40) 0.48 (0.21–0.77) 0.41 (0–0.88) 0.67 (0.33–1.40) 0 (0–0.25) 0.14 (0–0.29) 0 (0–0.11) 0.39 (0.15–0.87) 0.17 (0–0.42) 0.36 (0.18–0.72) 0.33 (0.17–0.82) 0.59 (0.22–1.33) 0 (0–0) 0.21 0.16 0.36 0.72 0.99 0.79 0.19 0.05 IQR, interquartile range. 154 Psychologic changes after gynecologic cancer Table 2 Hopkins Symptom Checklist-90 domain outcomes of women with ovarian cancer compared with other gynecologic types. Data are presented as median and interquartile (25th, 75th) ranges Domain Baseline Interpersonal Sensitivity (domain 1) Phobic-anxiety (domain 2) Retarded Depression (domain 3) Anger–Hostility (domain 4) Somatization (domain 5) Obsessive–Compulsive (domain 6) Agitated Depression (domain 7) Psychoticism (domain 8) 6-weeks’ follow-up Interpersonal Sensitivity (domain 1) Phobic-anxiety (domain 2) Retarded Depression (domain 3) Anger–Hostility (domain 4) Somatization (domain 5) Obsessive–Compulsive (domain 6) Agitated Depression (domain 7) Psychoticism (domain 8) Site Scores median (IQR) P-value Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others 0.14 (0–0.57) 0.14 (0–0.42) 0 (0–0.22) 0.11 (0–0.33) 0.75 (0.08–1.19) 0.31 (0.08–0.77) 0.33 (0.08–0.46) 0.17 (0–0.25) 0.29 (0.21–0.64) 0.57 (0.23–0.79) 0.33 (0.17–0.5) 0.17 (0–1.0) 1 (0.33–1.56) 0.56 (0.33–1.11) 0.06 (0–0.31) 0 (0–0.03) 0.98 Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others Ovary Others 0.14 (0–0.29) 0 (0–0.28) 0 (0–0) 0.01 (0–0.12) 0.31 (0.23–0.92) 0.38 (0.15–0.65) 0.33 (0.04–0.63) 0.08 (0–0.33) 0.36 (0.04–1.75) 0.36 (0.23–0.72) 0.33 (0.33–1.17) 0.33 (0.17–0.78) 0.33 (0.11–0.78) 0.62 (0.22–1.16) 0 (0–0.12) 0 (0–0) 0.75 0.67 0.40 0.39 0.65 0.75 0.63 0.85 0.27 0.99 0.48 0.94 0.95 0.48 0.47 IQR, interquartile range. compared with those with better supports in five of the eight domains at baseline, and in every domain at 6 weeks. Several of these differences were statistically significant. At 6 weeks’ follow up, scores remained significantly higher in the domains of phobic-anxiety, and retarded and agitated depression. Discussion The findings indicate that levels of symptomatology remain quite uniform across the first 6 weeks following a diagnosis of gynecologic cancer. The highest levels of symptomatology were observed in the domain of agitated depression. The next highest scoring domains were retarded depression, obsessive– compulsive symptoms and somatization. Scores in the domains of interpersonal sensitivity, phobic-anxiety, anger–hostility, and psychoticism were low. The findings support previous work that suggests that the early weeks following the diagnosis of a gynecologic cancer are critical, with symptomatology persisting unchanged for at least 6 weeks. In a controlled prospective study of women with gynecologic cancer, women were assessed for emotional distress at diagnosis and 4, 8, and 12 months after treatment. It was reported that those women with cancer suffered depressed, anxious and confused moods, whereas those with benign disease suffered only anxiety. However, in both groups symptoms had largely resolved by 4 months.7 Given the ease with which patients, partners and providers can access negative information about 155 R. W. Petersen et al. Table 3 Hopkins Symptom Checklist-90 domain outcomes of women with low compared with high levels of support. Data are presented as median and interquartile (25th, 75th) ranges Domain Baseline Interpersonal Sensitivity (domain 1) Phobic-anxiety (domain 2) Retarded Depression (domain 3) Anger–Hostility (domain 4) Somatization (domain 5) Obsessive–Compulsive (domain 6) Agitated Depression (domain 7) Psychoticism (domain 8) 6-weeks’ follow-up Interpersonal Sensitivity (domain 1) Phobic-anxiety (domain 2) Retarded Depression (domain 3) Anger–Hostility (domain 4) Somatization (domain 5) Obsessive–Compulsive (domain 6) Agitated Depression (domain 7) Psychoticism (domain 8) Level of support Scores median (IQR) P-value Low High Low High Low High Low High Low High Low High Low High Low High 0.29 (0.11–0.68) 0 (0–0.43) 0.11 (0–0.42) 0.07 (0–0.31) 0.58 (0.17–1.04) 0.31 (0.08–1.08) 0.13 (0–0.98) 0.17 (0–0.40) 0.64 (0.21–0.96) 0.32 (0–0.91) 0.17 (0–0.63) 0.42 (0–1.13) 0.67 (0.22–1.22) 0.61 (0.36–1.53) 0.0 (0–0.25) 0.0 (0–0.44) 0.22 Low High Low High Low High Low High Low High Low High Low High Low High 0.14 (0–0.29) 0.07 (0–0.45) 0.12 (0.11–0.47) 0 (0–0.01) 0.81 (0.06–1.06) 0.23 (0.08–0.62) 0.29 (0–0.69) 0.08 (0–0.40) 0.68 (0.43–1.13) 0.28 (0.09–0.68) 0.70 (0.28–0.92) 0.33 (0.17–1.00) 1.44 (0.97–1.86) 0.44 (0.11–1.19) 0 (0–0.16) 0 (0–0) 0.98 0.76 0.99 0.48 0.54 0.58 0.61 0.92 0.007 0.02 0.54 0.16 0.41 0.03 0.67 IQR, interquartile. ovarian cancer survival, we hypothesized that those with ovarian cancer would report higher levels of symptomatology compared with women with other gynecologic cancers. However, no differences in overall levels or patterns of symptomatology were observed. The findings contrast to recent papers that have suggested that there is an increased risk of psychologic morbidity in ovarian cancer patients due to the late stage of disease at diagnosis and aggressive multimodal treatment.8,9 It might be that the impact of prognosis is not apparent during the first 6 weeks following diagnosis, even though all women received at lease two cycles of chemotherapy in this period. It might be that differences in symptomatology do not become apparent until there is actual disease progression. One study supports this premise with a finding 156 that disease progression over a 2-year period was associated with increased emotional distress.10 The second hypothesis was that women who perceived their level of social support to be low would report higher levels of psychologic symptoms compared with women who perceived their social support to be high. The results supported this hypothesis, with the median of 13 of 16 symptom domains being higher in women with poor social supports. At 6 weeks, some of these differences were statistically significant in the domains of phobic-anxiety, and retarded and agitated depression. This suggests that those women who perceive their social support to be poor are particularly at risk of developing anxiety and depressive symptoms and should be monitored closely. Psychologic changes after gynecologic cancer While the median scores were higher than for those women with benign disease, the medians for each domain calculated from this data were very much lower than those reported in the validation study of the HSCL-90, presumably because of the different study populations.6,7 The validation population were subjects who had been reviewed by a psychiatrist and diagnosed with a psychiatric disorder. A metaanalytical review of 58 papers from 1980 to 1994 on psychologic and psychiatric problems in patients with cancer concluded that cancer patients might not be more anxious or distressed than the general population, but are significantly more depressed.11 Cancer patients were not, however, more anxious, distressed or depressed than psychiatric patient reference groups, and compared with groups of patients with other diagnoses, they were frequently less anxious. However, female cancer patient populations reported higher levels of distress than male populations.11 It might be that a number of women in the study were in denial or using behavioral disengagement as a coping response. Behavioral strategies were not directly measured. However, in one study of quality of life in women with gynecologic cancers over the first year of life after diagnosis, women who used behavioral disengagement were at risk for poor quality of life at 12 months.5 The small number of women and the short followup period in this prospective observational cohort limits the conclusions that can be drawn. Larger prospective studies that explore a range of symptomatology might be indicated. There have only been three randomized controlled trials of psychologic interventions in women with gynecologic cancer 2–4 with two of these using group counseling and obtaining contradictory results.3,4 The most recent study trialled individual guided imagery, relaxation, and counseling, and found a significant reduction in anxiety at 6 weeks’ post intervention.2 The results of the present study suggest that screening of women for social supports and targeted interventions might help reduce the burden of morbidity associated with the diagnosis of a gynecologic cancer. References 1. Thomson DS, Shear MK. Psychiatric disorders and gynaecological oncology. A review of the literature. Gen Hosp Psychiatry 1998; 20: 241–247. 2. Petersen RW, Quinlivan JA. Preventing anxiety and depression in gynaecological cancer – a randomised controlled trial. Br J Obstetrics Gynaecol 2002; 109: 386–394. 3. Cain EN, Kohorn EI, Quinlian DM et al. Psychosocial benefits of a cancer support group. Cancer 1986; 57: 183–189. 4. Wenzel LB, Robinson SE, Blake DD. The effects of problemfocussed group counselling for early stage gynaecologic cancer patients. J Mental Health Counseling 1995; 17: 81–93. 5. Lutgendorf SK, Anderson B, Ullrich P et al. Quality of Life and mood in women with gynaecologic cancer: A one year prospective study. Cancer 2002; 94: 131–140. 6. Lipman RS, Covi L, Shapiro AK. The Hopkins Symptom Checklist (HSCL). Factors derived from the HSCL-90. J Affect Disord 1979; 1: 9–24. 7. Andersen B, Anderson B, deProsse C. Controlled prospective longitudinal study of women with cancer: II. Psychological outcomes. J Consult Clin Psychol 1989; 57: 692–697. 8. McCorkle R, Pasacretta J, Tzuh Tang S. The silent killer: Psychological issues ovarian cancer. Holistic Nursing Prac 2003; 17: 300–308. 9. Bodurka-Bevers D, Basen-Engquist K, Camack CL et al. Depression, anxiety and quality of life in patients with epithelial ovarian cancer. Gynecol Oncol 2000; 78: 302–308. 10. Guido F. Living with ovarian cancer. Gynecol Oncol 1993; 50: 202–207. 11. van’t Spijker A, Trisjsburg RW, Duivenvoorden HJ. Psychological sequelae of cancer diagnosis: A meta-analytical review of 58 studies after 1980. Psychosom Med 1997; 59: 280– 293. 157