- © 2007 by American Society of Clinical Oncology
Reproductive and Sexual Function After Platinum-Based Chemotherapy in Long-Term Ovarian Germ Cell Tumor Survivors: A Gynecologic Oncology Group Study
- David M. Gershenson,
- Anna M. Miller,
- Victoria L. Champion,
- Patrick O. Monahan,
- Qianqian Zhao,
- David Cella and
- Stephen D. Williams
- From the Department of Gynecologic Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, TX; Indiana University Cancer Center; School or Nursing; School of Medicine, Indiana University, Indianapolis, IN; and the Evanston Northwestern Healthcare, Northwestern University Feinberg School of Medicine, Evanston, IL
- Address reprint requests to David M. Gershenson, MD, The University of Texas, M.D. Anderson Cancer Center, PO Box 301439, Houston, TX 77230-1439; e-mail: dgershen{at}mdanderson.org
Abstract
Purpose To compare malignant ovarian germ cell tumor survivors with a matched control group of females on menstrual and reproductive outcomes, sexual functioning, and dyadic adjustment.
Patients and Methods Eligible patients met the following criteria: (1) history of malignant ovarian germ cell tumor; (2) treatment with surgery plus platinum-based chemotherapy; (3) age at least 18 years and continuously disease-free with minimum follow-up of 2 years; (4) capability of completing questionnaire and telephone interview; and (5) completion of written informed consent. The control group was drawn from acquaintances recommended by survivors and matched for age, race, and education. Scales with established reliability and validity were used to measure quality-of-life concepts of sexual functioning and social networks.
Results One hundred thirty-two survivors and 137 controls completed the study. Of 132 survivors, 71 (53.8%) had fertility-sparing surgery. Of fertile survivors, 62 (87.3%) reported still having menstrual periods. Twenty-four survivors reported 37 offspring after cancer treatment. Compared with controls, survivors had significantly greater reproductive concerns (P < .0001), less sexual pleasure (P = .003), and lower scores on the total Sexual Activity Scale Score (P = .001). However, survivors had better dyadic consensus (P = .004), dyadic satisfaction (P = .005), and dyadic cohesion (P = .014).
Conclusion Women who had fertility-sparing surgery were very likely to retain menstrual function and fertility after chemotherapy. Although there is some increase in gynecologic symptoms and diminution in sexual pleasure, survivors tended to have stronger, more positive relationships with significant others.
INTRODUCTION
Since the introduction of platinum-based combination chemotherapy for malignant ovarian germ cell tumors in the late 1970s, most patients can expect cure.1-6 Before the 1980s, conventional wisdom held that chemotherapy treatment of a female patient in childhood, adolescence, or young adulthood almost invariably resulted in infertility. Early reports of women receiving chemotherapy had revealed that the primary lesion was follicle destruction and ovarian stromal fibrosis.7-9 Factors such as cumulative drug dose, duration of therapy, and age at treatment were thought to be important in influencing the incidence of ovarian dysfunction.10-13 In addition, several studies seemed to indicate that the prepubertal ovary is more resistant to the adverse effects of chemotherapy.9,14-18
During the last two decades, several reports that detail successful pregnancies after treatment of a variety of malignancies, including Hodgkin's disease, lymphomas, breast cancer, malignant melanoma, and gynecologic cancers, have emerged. In a large multi-institutional study sponsored by the National Cancer Institute, Byrne et al19 interviewed 2,283 long-term survivors of childhood or adolescent cancer and 3,270 control subjects selected from among the survivors' siblings. They noted an overall fertility deficit among survivors of 15%.
However, subsequent studies have revealed a significant risk of premature menopause in young patients receiving chemotherapy compared with controls. In one study, the risk of premature menopause was associated with the type of therapy administered rather than the type of cancer.20 The highest risk occurred in women aged 21 to 25 years who received radiation therapy below the diaphragm combined with alkylating-agent chemotherapy.
Several reports have documented successful pregnancies in young patients who previously underwent fertility-sparing surgery and combination chemotherapy for malignant ovarian germ cell tumors.3,4,6,21-26 Gershenson27 reported a questionnaire study of 40 patients treated for malignant ovarian germ cell tumors in which all patients were successfully treated with fertility-sparing surgery followed by combination chemotherapy; most patients received nonplatinum-based chemotherapy. At the time of analysis, 33 patients (83%) were having regular menses. Premature menopause was documented in one patient. Of 16 patients who had attempted pregnancy since chemotherapy, 11 delivered 22 healthy infants, none of whom had major birth defects.
Brewer et al6 reported their experience with 26 patients treated with surgery plus platinum-based chemotherapy for ovarian dysgerminoma. Of the patients who underwent fertility-sparing surgery and chemotherapy, 71% maintained their normal menstrual function during and after chemotherapy, and 93% had returned to their prechemotherapy menstrual pattern at the time of the questionnaire. Three subsequent reports have detailed post-therapy reproductive function in patients with malignant ovarian germ cell tumors and have noted normal menstrual function in at least 80%.28-30 Several live births were reported in each of these series.
Although several studies have addressed physical and/or reproductive issues in young women receiving chemotherapy for ovarian germ cell tumor, missing in the literature is the impact of this chemotherapy on sexual functioning and subsequent ability to form close relationships (dyadic relationships). In 1998, our group of multi-institutional collaborators initiated this study of quality of life in survivors of malignant ovarian germ cell tumors. Patients were those treated on prospective clinical trials of the Gynecologic Oncology Group (GOG) or similar protocols at The University of Texas M.D. Anderson Cancer Center. Subsequently, a protocol was activated through the GOG. The purpose of this research was to compare prospectively malignant ovarian germ cell tumor survivors with a matched control group of females on menstrual and reproductive outcomes, sexual functioning, and dyadic adjustment. This article describes the findings of that portion of the study that assesses menstrual, reproductive, gynecologic, and sexual function in these individuals. Subsequent publications will report other aspects of physical, psychological, and emotional health.
PATIENTS AND METHODS
Eligible patients met the following criteria: (1) history of early or advanced malignant ovarian germ cell tumor; (2) treatment with surgery plus platinum-based chemotherapy on GOG protocols 45, 78, 90, or 116, or protocol at M.D. Anderson Cancer Center (patients with any prior chemotherapy or radiotherapy were excluded)1-3,6,31; (3) age at least 18 years old and continuously disease-free with minimum follow-up of 2 years at the time of the interview; (4) capability of completing a written questionnaire and telephone interview in English; and (5) completion of written informed consent. The control group was drawn from acquaintances recommended by survivors and matched for age, race, and education.
All patients except those entered on GOG protocols 90 and 116 were treated with three to six cycles of cisplatin, etoposide, and bleomycin (BEP) or cisplatin, vinblastine, and bleomycin (PVB).1-3,6,31 All had early or advanced ovarian germ cell tumors of various histologic types. Patients enrolled on GOG protocol 116 received three cycles of carboplatin and etoposide as adjuvant therapy for dysgerminoma.31 Some patients on GOG Protocol 90 also received up to three cycles of the combination of vincristine, dactinomycin, and cyclophosphamide (VAC). Patients who received prior irradiation were not eligible for this study.
Cancer diagnosis and treatment variables were obtained from existing databases of the GOG and M.D. Anderson Cancer Center. Other sociodemographic characteristics were collected through the background questionnaire. Patients were contacted first by the treatment site and gave informed consent according to institutional and federal guidelines before enrollment.
Acquaintance controls were selected from individuals without a history of cancer proposed by the participating patients and matched for age, race, and education. Patients and controls were contacted by trained interviewers from Indiana University. Those who agreed to participate were mailed an informed consent and background questionnaire (with cancer-related questions deleted for controls). Subsequently, a 60-minute computer-assisted telephone interview was conducted with both subjects and controls.
Fertility was defined in two ways for this study: (1) Survivors were coded as fertile if they had fertility-sparing surgery initially and no subsequent surgery that precluded pregnancy within the first 2 years after initial treatment. This definition was used because some survivors with initial fertility-sparing surgery reported having infertility-inducing surgery within 2 years of diagnosis, which they attributed to post-treatment problems. (2) Potential fertility was measured for both survivors and controls, based on their answer to the question, “Have you had difficulty becoming pregnant?” This definition incorporated subjective perceptions of fertility and allowed for comparisons between survivors and controls.
Scales with established reliability and validity were used to measure quality-of-life concepts of sexual functioning and social networks. Sexual functioning was assessed during the telephone interview, using four scales addressing sexual activity, gynecologic symptomatology, reproductive concerns, and sexual self concept.
The Sexual Activity Questionnaire (SAQ) was originally developed to investigate the impact of long-term preventive therapy for women at high risk for breast cancer.32 Factor analysis indicated that the pleasure and discomfort scales accounted for 54.6% of the variance. Validity was also demonstrated when the SAQ scale discriminated between sexual functioning of pre- and postmenopausal women. For this study, test-retest reliability was calculated by the kappa (κ) statistic and by Pearson's correlation. Correlations using the κ statistic ranged between 0.5 and 1.0 and for the Pearson's correlation ranged between 0.65 and 1.0. For this study, the alpha coefficient was .76, .90, and .63 for the total, Pleasure and Discomfort scales, respectively.
The Gynecological Symptom Scale includes specific questions about common gynecologic concerns, such as vaginal bleeding or dryness, and sexual intercourse problems.33 The alpha coefficient for this study was .78.
The 14-item Reproductive Concerns Scale was developed to assess reproductive problems or worries of ovarian and breast cancer survivors who were diagnosed and treated during their childbearing years.34 The alpha coefficient for this study was .85.
The 26-item Sexual Self Schema scale is designed to measure a cognitive self-view of both positive and negative aspects of sexuality.35 Positive aspects include an inclination to experience romantic/passionate emotions and a behavioral openness to sexual experiences and/or relationships, whereas the negative aspects of embarrassment and/or conservatism appear to be deterrents to sexual expression. Construct validity was demonstrated when the scale predicted sexual outcomes following cancer. Internal consistency reliabilities ranged from .66 to .81, and test-retest reliability was .89 at 2 weeks. For this study, the alpha coefficient was .72.
The Dyadic Adjustment Scale was used to assess the quality of marital or other dyadic relationships. The overall scale reliability was reported as 0.96 in 1976 and as 0.91 in 1982, with subscale reliabilities ranging from 0.73 to 0.94 in 1976 and 0.81 to 0.95 in 1982, using Cronbach's alpha coefficient.36,37 For this study, the overall alpha coefficient was .95; Dyadic Consensus, .94; Dyadic Satisfaction, .88; Dyadic Cohesion, .88; and Affectional Expression, .61.
Another report from our group focused more globally on quality-of-life issues in the same cohorts, including psychological well-being, social well-being, and spiritual well-being, has been submitted for publication and is currently under review.
Statistical analyses were performed using SAS version 8.2 (SAS Institute, Cary, NC). Multiple linear regression was used to statistically compare survivors and controls on outcomes after adjusting for important covariates. T tests were used to compare reproductive and sexual functioning scales between fertile and infertile survivors at diagnosis.
RESULTS
Demographics for survivors and controls are reported in Table 1. A total of 170 women were identified by the GOG and M.D. Anderson as eligible and available for contact. These patients were from institutions that were active GOG members at the time the protocol was initiated and who activated the protocol through their institutional review boards. However, 23 (14%) were lost to follow-up, primarily related to long-term follow-up and societal mobility. Thus, 147 women were contacted, and 142 (97%) agreed to participate. Ten failed to complete both parts of the study, resulting in a sample size of 132 survivors from 40 different sites and an acquaintance control total of 137. For controls, 247 acquaintance names were submitted by survivors; 58 were not able to be found, and a final pool of 160 were available for matching before enrolling 137 controls into the study.
Demographic Characteristics of Survivors and Controls
Median survivor age at study entry was 35.5 years (range, 19 to 64 years), and median control age was 34.5 years (range, 19 to 58 years), which was not statistically different. However, using the two-sided Fisher's exact test, controls (69%) were significantly more likely to be in a partnered relationship (married/relationship v others) than were survivors (57%; Table 1; P = .04). Additionally, using the Mantel test for linear trend, controls had higher incomes (Table 1; P = .01). The clearest split in income was that 43% of controls earned $65,000 or more compared with 30% of survivors (Table 1).
Physical Parameters of Survivors
Of the 132 survivors, 71 (53.8%) had fertility-sparing surgery as part of initial treatment. Of those fertile survivors, 62 (87.3%) reported that they were still having menstrual periods, compared with 114 (83.2%) controls, whereas nine (12.7%) fertile survivors reported no longer having periods. Twenty-three (37%) of the 62 fertile survivors who reported that they were still having menstrual periods were taking oral contraceptives. Survivors who had fertility-sparing surgery were significantly younger at diagnosis than were survivors who had surgically-induced infertility within 2 years postdiagnosis (median age, 21 v 29 years; P < .0001) and also were significantly younger at the time of study (median age, 32 v 39 years; P < .0001).
Of the nine fertile survivors who reported that they were not still having menstrual periods, five reported having had a hysterectomy later than 2 years post-treatment, and two reported menopause before age 40. The reasons for amenorrhea for the remaining two women were unclear (one reported menopause at age 26 before cancer treatment with PVB; the other woman, who was only 21 years old at the time of the interview, began menstrual periods at age 14, was diagnosed at age 16, was treated with a combination of etoposide and carboplatin, and reported not having gone through menopause).
Of the 61 infertile survivors, 34 were taking estrogen-replacement therapy, 20 were not, and seven were missing information.
Comparison of Survivors and Controls on Pregnancy
When the 132 survivors were asked about their desire for childbearing, 34 answered “not at all,” 13 answered “a little bit,” 17 “somewhat,” 19 “quite a bit,” and 49 answered “very much.” Answers to this question for the 137 controls were 18, 24, 18, 26, and 47, respectively (four controls did not answer this question).
At interview time, 24 survivors reported they had 37 children after cancer treatment. Fifteen survivors had one child, six had two children, two had three children, and one had four children. Of those 24 women, the first pregnancy after cancer treatment resulted in a live birth for 20 women (83%), and no major medical problems with the infants were reported. Seven survivors (5.3%) reported having spontaneous abortions after cancer treatment and 12 (9.1%) before cancer, for a total of 19 (14%) spontaneous abortions. Four survivors (3%) reported induced abortions after cancer treatment, and six (4.5%) reported induced abortions before cancer treatment.
By comparison, 92 controls reported that they had 186 children. Thirty-four controls had one child, 34 had two children, 16 had three children, five had four children, two had five children, and one had six children. Thirty-five controls (25.5%) reported having had spontaneous abortions, and 11 (8%) reported having had induced abortions. Distinction between therapeutic and elective abortions was not made in this study.
Ten survivors and three controls reported that they had undergone an infertility evaluation. Details of reproductive data are presented in Table 2.
Reproductive Profile of Survivors and Controls
Comparison of Survivors and Controls on Sexual Functioning and Dyadic Adjustment
For all comparisons, income, partnership (married/relationship v others), and perception of fertility were controlled, and the means for survivors and controls were adjusted for those three covariates (Table 3). Because survivors and controls were matched on age, race, and education, these variables were not controlled. Differences did emerge on several variables between survivors and controls. Compared with controls, survivors had significantly greater reproductive concerns (P < .0001), less sexual pleasure (P = .017), and lower scores on the total Sexual Activity Scale Score (P = .007). However, survivors had better dyadic consensus (perceived agreement with partner; P = .001), dyadic satisfaction (individual's adjustment to the partner relationship; P = .002), and dyadic cohesion (shared pleasant activities; P = .004; Table 3). Survivors did not differ significantly from controls on menstrual/gynecologic problems, sexual discomfort, or Sexual Self Schema score, but did differ significantly from controls on the total Dyadic Adjustment Scale score (P = .010).
Differences in Reproductive and Sexual Functioning Scales Between Survivors and Controls
Comparison of Fertile and Infertile Survivors
Comparisons also were made between fertile and infertile survivors, based on whether they had surgically induced infertility within 2 years of diagnosis. When controlling for current age (Table 4), infertile survivors reported significantly greater reproductive concerns (P < .0001) and sexual discomfort (P = .036). They also reported marginally significantly higher Sexual Self Schema total scores (P = .057), indicating a somewhat more positive sexual self-concept.
Differences in Reproductive and Sexual Functioning Scales Between Fertile and Infertile Survivors at Diagnosis
DISCUSSION
To our knowledge, the present cooperative group study is the largest and most comprehensive survey of survivors of malignant ovarian germ cell tumors. Our findings indicated that 87% of fertile survivors were still having menstrual function at the time of the study, confirming prior observations.6,27-30 Likewise, similar to the studies of Zanetta et al29 and Tangir et al,30 survivors reported a large number of healthy offspring.
Of the 71 patients in our study who underwent fertility-sparing surgery and platinum-based chemotherapy, we could document premature menopause in only two women (3%). Tangir et al30 reported that two (3%) of 64 patients who underwent fertility-sparing surgery plus chemotherapy developed premature menopause; however, neither of these patients received platinum-based chemotherapy.
Possible premature menopause was difficult to ascertain in the present study because of inconsistencies in self-reported data. Menopause was defined as “stopped having periods.”
Interestingly, some women who reported that they no longer had periods also answered “No” to a parallel question about whether they had gone through menopause. Possibly, young women might equate menopause with older women and not consider stopping menses as true menopause. Alternatively, they may not have had typical menopausal symptoms. Therefore, these results may somewhat underestimate the frequency of premature menopause. The median age of survivors at the time of the survey was 36 years; thus, it is possible that others will ultimately develop premature menopause.
As noted in Tables 2 and 3, a number of significant differences were observed between survivors and controls and between fertile and infertile survivors, respectively. Importantly, survivors had significantly greater reproductive concerns and less sexual pleasure than did controls. Interestingly, however, survivors had Dyadic Adjustment Scale scores that implied better marital or dyadic relationships than did controls. One might speculate that, because of their concerns and insecurities, survivors exert greater effort toward a healthy long-term relationship than do women without a history of cancer, but this is pure conjecture.
Predictably, infertile survivors reported significantly greater reproductive concerns and sexual discomfort compared with fertile survivors. Infertile survivors also had higher Sexual Self Schema total scores, which may reflect coming to terms with their infertility and having positive self-perceptions of themselves as women.
Some patients in this study were treated up to two decades beforehand. This no doubt explains the relatively high number of patients whose initial surgical procedure included hysterectomy and/or bilateral salpingo-oophorectomy. The favorable outcome of treatment of these patients, both in terms of survival and also quality of life and fertility, again underscores the importance of a vigorous attempt at fertility-sparing surgery for newly diagnosed patients.
In summary, the results of this large multicenter study confirm and expand our knowledge of the hopeful outlook regarding post-treatment reproductive and sexual function in survivors of malignant ovarian germ cell tumors. These findings lend some insight into issues that patients with malignant ovarian germ cell tumors should address with their physicians at the time of diagnosis and primary treatment. Survivors may have more concerns about reproductive and sexual outcomes although this can vary as a function of fertility. It seems that being in a partnered relationship protected survivors in terms of reproductive and sexual functioning. Thus, it may be particularly important for gynecologic oncologists to specifically discuss implications of fertility and sexual functioning with patients who are not partnered at the time of diagnosis. Women who had fertility-sparing surgery were very likely to retain menstrual function and fertility after chemotherapy. Although there is some increase in gynecologic symptoms and diminution in sexual pleasure, survivors tended to have stronger, more positive relationships with significant others. Future publications from this large study will focus on other aspects of physical and psychological health of these survivors.
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The author(s) indicated no potential conflicts of interest.
AUTHOR CONTRIBUTIONS
Conception and design: David M. Gershenson, Victoria L. Champion, David Cella, Stephen D. Williams
Financial support: Stephen D. Williams
Administrative support: Stephen D. Williams
Provision of study materials or patients: David M. Gershenson, Stephen D. Williams
Collection and assembly of data: David M. Gershenson, Anna M. Miller, Victoria L. Champion, Stephen D. Williams
Data analysis and interpretation: David M. Gershenson, Anna M. Miller, Victoria L. Champion, Patrick O. Monahan, Qianqian Zhao, David Cella, Stephen D. Williams
Manuscript writing: David M. Gershenson, Anna M. Miller, Victoria L. Champion, Patrick O. Monahan, David Cella
Final approval of manuscript: David M. Gershenson, Anna M. Miller, Victoria L. Champion, Stephen D. Williams
Other: Anna M. Miller
Acknowledgments
The following Gynecologic Oncology Group member institutions participated in this study: University of Alabama at Birmingham; Duke University Medical Center; Abington Memorial Hospital; University of Minnesota Medical School; University of Mississippi Medical Center; Colorado Gynecologic Oncology Group P.C.; Milton S. Hershey Medical Center; University of Cincinnati; University of North Carolina School of Medicine; University of Iowa Hospitals and Clinics; Indiana University Medical Center; Wake Forest University School of Medicine; University of California Medical Center at Irvine; Tufts-New England Medical Center; Rush-Presbyterian –St. Luke's Medical Center; SUNY Downstate Medical Center; University of Kentucky; Community Clinical Oncology Program; The Cleveland Clinic Foundation; Johns Hopkins Oncology Center; State University of New York at Stony Brook; Eastern Pennsylvania GYN/ONC Center, P.D.; Washington University School of Medicine; Memorial Sloan-Kettering Cancer Center; Cooper Hospital/University Medical Center; Columbus Cancer Council; The University of Texas M.D. Anderson Cancer Center; Fox Chase Cancer Center; University of Oklahoma; Tacoma General Hospital; Christiana Health Care; Carolinas Medical Care Center; Grant/Riverside Cancer Services; Hinsdale Hospital; Massachusetts General Hospital; Long Beach Memorial Medical Center; Miami Valley Hospital; University of New Mexico; University of Wisconsin; Women & Infants Hospital of Rhode Island; and Women's Hospital, Baton Rouge.
Footnotes
-
This study was supported by National Cancer Institute (NCI) R01 grant CA77470 and NCI grants to the Gynecologic Oncology Group Administrative Office (CA 27469) and the Gynecologic Oncology Group Statistical Office (CA 37517).
-
Presented in part at the 38th Annual Meeting of the of the American Society of Clinical Oncology, Orlando, FL, May 18-21, 2002.
-
Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.
- Received July 31, 2006.
- Accepted March 14, 2007.
