Cancer incidence following treatment for infertility at a clinic in the UK

Pat Doyle1,5, Noreen Maconochie1, Valerie Beral2, Anthony J. Swerdlow3 and S.L. Tan4

1 Epidemiology Unit, Department of Epidemiology and Population Health, London School of Hygiene and Tropical Medicine, University of London, Keppel St, London WC1E 7HT, 2 Imperial Cancer Research Fund, University of Oxford, Gibson Building, Radcliffe Infirmary, Oxford OX2 6HE, 3 Section of Epidemiology, Institute of Cancer Research, Sutton, Surrey SM2 5NG, UK and 4 Department of Obstetrics and Gynaecology, McGill University, Royal Victoria Hospital, Montreal, Canada


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
BACKGROUND: There is concern about the long-term health impact of ovarian stimulation treatment for infertility, in particular the effect on cancer risk. The aim of this study was to investigate the incidence of cancer in a cohort of women attending a large infertility clinic in the UK. METHODS: Women who were UK residents attending the clinic between January 1, 1975 and December 31, 1989 were identified for the study. The cohort was followed-up and cancer incidence rates calculated. RESULTS: The study cohort was made up of 5556 women of whom 75% had received ovarian stimulation drug treatment at the clinic. A total of 118 cancers (including 55 breast, four corpus uteri and six ovarian) were incident in the cohort from the beginning of 1990 to the end of 1997. The incidence rates of cancer of the breast, corpus uteri and ovary were not significantly different from expectation based on national cancer rates, and were similar for women who had received hormonal treatment to stimulate their ovaries and those who had not. CONCLUSIONS: These data do not support a hypothesis linking infertility treatment involving ovarian stimulation with increased breast, uterine and ovarian cancer over the follow-up period studied.

Key words: cancer incidence/infertility treatment/ovarian stimulation


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
In the 1980s, case reports of ovarian cancer in women undergoing assisted conception raised concern about the long-term effects of infertility treatment (Fishel and Jackson, 1989Go). Subsequent epidemiological studies reported positive associations between exposure to fertility drugs used to stimulate ovulation and the risk of ovarian cancer (Whittemore et al., 1992Go; Rossing et al., 1994Go; Shushan et al., 1996Go). However, more recent studies have not confirmed this association, nor have they found an association between ovarian stimulation and the overall risk of breast cancer (Bristow and Karlan, 1996Go; Modan et al., 1998Go; Potashnik et al., 1999Go; Ricci et al., 1999Go; Venn et al., 1999Go; Dor et al., 2002Go; Ness et al., 2002Go). Raised risks for cancer of the corpus uteri have been observed in cohorts of infertile women, but no clear association with ovarian stimulation has been demonstrated (Modan et al., 1998Go; Venn et al., 1999Go). The study described here assesses cancer risk in a group of women treated for infertility at a clinic in the UK.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
The design was a cohort study of cancer incidence in a group of women attending the Hallam Medical Centre, a large infertility clinic in London, UK. Women who had had at least one cycle of infertility treatment between January 1, 1975 (when the clinic opened) and December 31, 1989, and who were resident in the UK and aged >=20 years at time of treatment, were identified for the study. The women were divided into two groups: those who had received drugs to stimulate ovulation (stimulated group or `exposed' group) and those who had not (unstimulated group or `unexposed' group). We did not have full clinical details for each patient on the database, but drug regimens used at the clinic over the period of study included clomiphene citrate alone, clomiphene citrate in combination with HMG and HCG, and from 1985 onwards GnRH agonist (e.g. buserelin). Treatment for women who did not receive ovarian stimulation was primarily artificial insemination by husband (AIH) or by donor (DI).

Details of all women in the cohort were submitted for tracing to the National Health Service Central Register (NHSCR) in Southport, UK, who provided ongoing notifications of emigrations, deaths and cancer registrations. The completeness of clinic records for women who had emigrated, died or been diagnosed with cancer was uncertain before 1990, when we compiled the dataset. For this reason we investigated cancer incidence from 1990 onwards in women known to be alive, cancer-free and still resident in the UK on January 1, 1990 (5556 women). Restricting the follow-up period to 1990 onwards, rather than date of first treatment onwards, resulted in the exclusion of 74 women (451 person-years) who had died, emigrated or been diagnosed with cancer before 1990.

Women in the cohort were followed-up from 1990 until cancer registration, death, emigration or December 31, 1997 (considered the most recent date for complete cancer registration in England and Wales), whichever was earliest. Person-years at risk (from January 1, 1990), stratified by 5 year age band and single calendar year, were calculated using the Stata computer package (Statacorp, 1997Go). Age and calendar year-specific cancer incidence rates for England and Wales (Office for National Statistics, 1997–2001Go) were used to calculate expected numbers of cancers and the standardized incidence ratio (SIR), calculated as the ratio of observed to expected cancers, summed over age group and period strata. Exact 95% confidence intervals (CI) and statistical significance levels were obtained using the Poisson distribution (Bailar and Ederer, 1964Go). To evaluate the effect of ovarian stimulation on risk within the cohort, hazard (rate) ratios (RR) were estimated using the Cox proportional hazards model, adjusting for current age (as the time variable), current calendar year, parity following the last treatment cycle and year of first clinic visit (Breslow and Day, 1987Go). Statistical significance was assessed using the likelihood ratio test.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Characteristics of the cohort
Among the 5556 fully traced and eligible women in the cohort, 4188 (75%) had had at least one stimulated cycle at the clinic (Table IGo). The distributions of year of first clinic visit were similar in both treatment groups (P = 0.95), as was the length of time that had elapsed between first visit and 1st January 1990 when the cohort was identified (median in both groups 7.5 years, P = 0.92). On average the group who received ovarian stimulation was slightly older than the group who did not, and a higher proportion was still nulliparous after the last treatment cycle at this clinic (both P < 0.001). During the follow-up period 59 women died and 33 (0.6%) were lost to follow-up.


View this table:
[in this window]
[in a new window]
 
Table I. Characteristics of women in the cohort of 5556 women alive and cancer-free at 31st December 1989, by whether or not they received ovarian stimulation during treatment at the clinica
 
Cancer incidence
A total of 118 women were registered with cancer up to 31st December 1997. One woman was registered with both ovarian and corpus uterine cancer at the same time, and was included in each of the analyses of cancer of the ovary and cancer of the uterus, but once only in the analysis of all cancers. A total of 55 breast cancers, four corpus uterine cancers (hereafter referred to as `uterine' cancers) and six ovarian cancers occurred in the cohort (Table IIGo). There was no significant difference between stimulated and unstimulated groups in the incidence of any of the cancer sites studied (P = 0.89, 0.07 and 0.53 for cancers of the breast, uterus and ovary respectively). The length of time between first treatment and development of cancer was also similar in the two groups for all sites studied (P = 0.36, 0.18 and 0.35 for breast, uterine and ovarian cancer respectively; Table IGo). Compared with the general population, the numbers of cancer of the breast and uterus were higher than expected in both stimulated and unstimulated groups, but not significantly so (all P > 0.38; Table IIGo). Ovarian cancer incidence appeared lower than expected in the stimulated group and higher in the unstimulated group, but the numbers of cases were too small to draw any conclusions (P = 0.97 and 0.67 respectively). Incidence of cancer of the cervix was significantly lower than expected overall (P = 0.02), and no cases were observed in the unstimulated group. Cancers other than breast, uterus, cervix and ovary showed higher incidence in both groups (P = 0.02 overall) compared with national rates. Diagnoses contributing to this increase were 13 cases of malignant melanoma (overall SIR 223; 95% CI 119–382; P = 0.01) and 11 cases of colorectal cancer (overall SIR 189; 95% CI 94–338; P = 0.07). However, there were no significant differences in risk between stimulated and unstimulated groups for either of these cancer sites [adjusted relative risks 0.64 (95% CI 0.12–2.43) for melanoma and 0.87 (95% CI 0.19–4.05) for colorectal cancer; P = 0.48 and 0.86 respectively]. Nine of the 13 malignant melanoma cases were in the stimulated group (SIR 197; 95% CI 90–374; P = 0.09) and four in the unstimulated group (SIR 358; 95% CI 97–916; P = 0.05). For colorectal cancer, nine of the 11 cases were in the stimulated group (SIR 197; 95% CI 90–374; P = 0.09) and two in the unstimulated group (SIR 179; 95% CI 22–646; P = 0.61).


View this table:
[in this window]
[in a new window]
 
Table II. Cancer incidence in women treated for infertility at the clinic, differentiated according to whether they received ovarian stimulation or not during treatment at the clinic
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Overall, the incidence of breast, uterine and ovarian cancers in 5556 women treated at the Hallam Medical Centre was no greater than expected based on national rates over the period of follow-up. In addition, we found no evidence for a link between ovarian stimulation treatment and increased cancer incidence. Although the numbers of cancers observed here were small (except for breast cancer) the results are consistent with the overall findings of the majority of cohort studies in this area (Modan et al., 1998Go; Potashnik et al., 1999Go; Ricci et al., 1999Go; Venn et al., 1999Go; Dor et al., 2002Go). Raised incidence of melanoma and colorectal cancer (and reduced incidence of cancer of the cervix) were detected in the cohort as a whole compared with the general population, but there was no evidence of an association with ovarian stimulation treatment. With the data available to us we were unable to investigate this finding further, although we note that patients receiving infertility treatment may be of higher socio-economic status, on average, than the general population, and both melanoma and colorectal cancer have a higher incidence in people with higher socio-economic status. The decreased risk of cancer of the cervix in the whole cohort compared with the general population may reflect increased surveillance within this group of women.

To the best of our knowledge this is the first epidemiological report of cancer incidence following ovarian stimulation treatment of UK residents. Although the results presented here are largely negative, they must be interpreted with care. First, we started the follow-up in 1990 which was, on average, 7.5 years since the patient's first clinic visit (Table IGo). Thus we could not examine cancer incidence shortly after treatment, and therefore could not address the finding of Venn et al. of a raised incidence of breast and uterine cancer within 12 months of exposure to fertility drugs (Venn et al., 1999Go). Secondly, we were unable to make adjustments for potential confounding factors, such as age at first birth or use of oral contraceptives, in the analyses because data were not available. Also, the women in the cohort were relatively young. Although by the end of the follow-up period a median of 15.5 years had elapsed since first clinic visit, the average age of these women was still only 46 years and it is possible that ovarian stimulation may have adverse effects later in life in exposed women. Furthermore, despite the relatively large cohort size, few ovarian and uterine cancers were incident over the follow-up period, resulting in limited power to detect statistically significant differences in the incidence of these rarer tumours. This problem is common to almost all the epidemiological studies on this topic to date. In order to improve statistical power we plan to continue follow-up: the number of expected cancers will increase markedly as the cohort ages. We would also recommend combining data from different studies in order to conduct a pooled analysis.


    Acknowledgements
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
We wish to thank Moya Simmons, Elizabeth Hilton, Bonita Peachey, Mary Sewell, Juliet Jain and Graham Davies for their work on the project. We are indebted to Bridgett Mason and Malcolm Pike for their invaluable support in the early stages of the project.

The study was funded by the Imperial Cancer Research Fund.


    Notes
 
5 To whom correspondence should be addressed. E-mail: pat.doyle{at}lshtm.ac.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 Acknowledgements
 References
 
Bailar, J.C. and Ederer, F. (1964) Significance factors for the ratio of a Poisson variable to its expectation. Biometrics, 20, 639–643.[ISI]

Breslow, N.E. and Day, N.E. (1987) Statistical Methods in Cancer Research Vol 2. The Design and Analysis of Cohort Studies. International Agency for Research on Cancer (IARC scientific publications No. 82), Lyons, France

Bristow, R.E. and Karlan, B.Y. (1996) Ovulation induction, infertility, and ovarian cancer risk. Fertil. Steril., 66, 499–507.[ISI][Medline]

Dor, J., Lerner-Geva, L., Rabinovici, J., Chetrit, A., Levran, D., Lunenfeld, B., Mashiach, S. and Modan, B. (2002) Cancer incidence in a cohort of infertile women who underwent in-vitro fertilization. Fertil. Steril., 77, 324–327.[ISI][Medline]

Fishel, S. and Jackson, P. (1989) Follicular stimulation for high tech pregnancies: are we playing it safe? Br. Med. J., 299, 309–311.[ISI][Medline]

Modan, B., Ron, E., Lerner-Geva, L., Blumstein, T., Menczer, J., Rabinovici, J., Oelsner, G., Freedman, L., Mashiach, S. and Lunenfeld, B. (1998) Cancer incidence in a cohort of infertile women. Am. J. Epidemiol., 147, 1038–1042.[Abstract]

Ness, R.B., Cramer, D.W., Goodman, M.T., Kjaer, S.K., Mallin, K., Mosgaard, B.J., Purdie, D.M., Risch, H.A., Vergona, R. and Wu, A.H. (2002) Infertility, fertility drugs, and ovarian cancer: a pooled analysis of case–control studies. Am. J. Epidemiol., 155, 217–224.[Abstract/Free Full Text]

Office for National Statistics (1997–2001) Series MB1: Cancer statistics, Registrations. Numbers 23–28. The Stationery Office, UK.

Potashnik, G., Lerner-Geva, L., Genkin, L., Chetrit, A., Lunenfeld, E. and Perath, A. (1999) Fertility drugs and risk of breast and ovarian cancers: results of a long-term follow-up study. Fertil. Steril., 71, 853–859.[ISI][Medline]

Ricci, E., Parazzini, F., Negri, E., Marsico, S. and La Vecchia, C. (1999) Fertility drugs and the risk of breast cancer. Hum. Reprod., 14, 1653–1655.[Abstract/Free Full Text]

Rossing, M.A., Daling, J.R., Weiss, N.S., Moore, D.E. and Self, S.G. (1994) Ovarian tumors in a cohort of infertile women. NEJM, 331, 771–776.[Abstract/Free Full Text]

Shushan, A., Elchalal, U., Paltiel, O., Peretz, T., Iscovich, J. and Schenker, J.G. (1996) Human menopausal gonadotrophin and the risk of epithelial ovarian cancer. Fertil. Steril., 65, 13–18.[ISI][Medline]

Statacorp (1997) Stata Statistical Software: Release 5.0. Stata Corporation, College Station, Texas.

Venn, A., Watson, L., Bruinsma, F., Giles, G. and Healy, D. (1999) Risk of cancer after use of fertility drugs with in-vitro fertilisation. Lancet, 354, 1586–1590.[ISI][Medline]

Whittemore, A.S., Harris, R., Itnyre, J. and Collaborative Ovarian Cancer Group (1992) Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case–control studies. II. Invasive epithelial ovarian cancers in white women. Am. J. Epidemiol., 136, 1184–1203.[Abstract]

Submitted on February 25, 2002; accepted on April 17, 2002.