Rapid Responses to:

EDITORIALS: Olav Meirik and Timothy M M Farley Risk of cancer and the oral contraceptive pill BMJ 2007; 335: 621-622 [Full text]

Rapid Responses: Submit a response to this article

Rapid Responses published:

CNS and pituitary cancers increased by oral contraceptives

Ellen C Grant   (17 September 2007)

Pituitary adenomas are not cancers

Richard Quinton, Andrew Advani   (3 October 2007)

Clinically meaningful temporal span for follow up

Grazyna T Adamiak   (26 October 2007)

CNS and pituitary cancers increased by oral contraceptives 17 September 2007
Ellen C Grant, physician and medical gynaecologist Kingston-upon-Thames, KT2 7JU Send response to journal: Re: CNS and pituitary cancers increased by oral contraceptives	Meirik and Farley write that long term follow-up of women taking the pill in the UK RCGP study shows no increased risk of cancer.1,2 Unfortunately, the study is deeply flawed. One in 3 women was lost to follow-up. Also most cancers were recorded in older women after 1998 when HRT records were no longer kept. Such fundamental problems underestimate the large cancer increases caused by taking hormones for different reasons at different ages, especially if hormone users remain listed as never users. One clear result from the muddled RCGP pill study is a large increased risk of CNS and pituitary tumours. The relative risk with current use, or less than 60 months since last use, was 6.47 (1.55 - 27.07). Meirik and Farley suggest that this may be due to prescription bias because contraceptive pills are prescribed for menstrual irregularities and women with irregular bleeding may already have pituitary disease. However, amenorrhoea or irregular cycles are likely to be due to measurable and rectifiable nutritional deficiencies in my experience and nutritional deficiencies commonly caused irregularities in athletes.3 Furthermore, laboratory animals fed progestins and oestrogens develop pituitary tumours. In 1944 Valzquez-Lopez and colleagues described tumours in the pituitary gland and related hypothalamic centres following prolonged treatment with oestrogen.4 In 1966 Poel induced pituitary tumours in mice after prolonged feeding of synthetic progestins.5 In 1975 Jacobi and colleagues induced pituitary tumours in male rats with a single dose of oestrogen.6 What are the implications for the use of large doses of hormones as morning-after pills? All hormonal contraceptives act predominantly like progesterone whether or not oestrogen is also added. High levels of progesterone in pregnancy inhibit ovulation. Adolescence is when the biological reproductive clock starts ticking. This may be a particularly vulnerable time for the functions of the hypothalamus and pituitary to be blocked by hormone use. Girls starting the pill in adolescence are likely to keep taking hormonal contraceptives for longer than women in the RCGP study, especially if they are given implants. There is often a liaison with schools and clinics to give hormones to adolescents without the family general practitioner or parents being informed. Pituitary tumours are likely to become commoner as the starting age for hormonal contraception falls to as young as 13. 1 Meirik O, Farley TMM. Risk of cancer and the oral contraceptive pill BMJ, Sep 2007; doi:10.1136/bmj.39336.503067.BE 2 Hannaford P C, Selvaraj S, Elliott AM, Angus V, Iversen L, Lee AJ. Cancer risk among users of oral contraceptives: cohort data from the Royal College of General Practitioner's oral contraception study. BMJ 2007; 0: bmj.39289.649410.55v1 3 Kopp-Woodroffe SA, Manore MM, Dueck CA, Skinner JS, Matt KS. Energy and nutrient status of amenorrheic athletes participating in a diet and exercise training intervention program. Int J Sport Nutr. 1999 ;9:70-88. 4 Valzquez-Lopez E. The reaction of the pituitary gland and related hypothalamic centres in the hamster to prolonged treatment with oestrogens. J Pathol Bacteriol 1944 ; 56: 1-7. 5 Poel WE. Pituitary tumors in mice after prolonged feeding of synthetic progestins. Science 1966; 154:402-03. 6. Jacobi J, Lloyd HM, Mears JD. Induction of pituitary tumors in male rats by a single dose of estrogen. 1975:7:228-30. Competing interests: None declared
Pituitary adenomas are not cancers 3 October 2007
Richard Quinton, Consultant & Senior Lecturer in Endocrinology Royal Victoria Infirmary & University of Newcastle-upon-Tyne, NE1 4LP, UK, Andrew Advani Send response to journal: Re: Pituitary adenomas are not cancers	Dear Sir, In their Editorial discussion of the UKRCGP study results, Meirik & Farley (1) repeatedly make the same factual error as the study authors themselves (2) by referring to "cancers.....of the pituitary". For the record, pituitary adenomas are benign tumours in that they (a) do not metastasise, (b) do not invade adjacent tissues and (c) are exceptionally slow-growing (3). The great majority of pituitary tumours are hormonally inactive microadenomas (<1cm diameter), which are of no clinical significance. Although every macroademona must have started off as a "micro", it is estimated that only 10% of microadenomas have the potential to evolve into macroadenomas (4). Carcinomas of the pituitary gland are so rare that an Endocrinologist could easily practice for an entire lifetime in a specialist pituitary referral centre and never see a single case (4). Despite not being cancerous, adverse events do occur in relation to benign pituitary adenomas due to (a) local mass effect (typically on the overlying optic chiasm to cause bitemporal visual field loss), (b) systemic effects of secreted products, PRL causing hypogonadism in prolactinomas, GH in acromegaly and ACTH in Cushing's disease, or (c) hypopituitarism resulting from tumour mass itself, or more commonly, its treatment by surgery and/or fractionated external beam radiotherapy (4). Pitfalls regularly encountered by the non-specialist include (a) performing unnecessary neurosurgery on patients whose pituitary adenomas were PRL-secreting and would therefore best been treated with oral dopamine-agonist therapy and (b) automatically assuming that high PRL levels are diagnostic of a prolactinoma and thereby failing the allow for the possibility of a non-secretory adenoma (causing hyperprolactinaemia via pituitary stalk compression) expanding to compress the optic chiasm despite normalisation of PRL levels on dopamine-agonist therapy. Outcomes can be improved by ensuring that patients are treated by clinicians with the appropriate expertise in the field, working as part of a multidisciplinary team (Endocrinology, specialist Pituitary Surgery, Neuroradiology and Clinical Oncology) (4). Yours sincerely, Andrew Advani & Richard Quinton Refs: (1). Meirik & Farley Editorial in BMJ 12/9/07 (2). Hannaford et al, BMJ 12/9/07 (3). Any textbook of pathology or even Wikipedia (4). "Pituitary Tumours: epidemiology, pathogenesis & Management". Edited by Susan Webb, Bioscientifica, Bristol, 1998. ISBn 1 901978 04 4 Competing interests: None declared
Clinically meaningful temporal span for follow up 26 October 2007
Grazyna T Adamiak, PhD, MA, MH&W Unemployed Send response to journal: Re: Clinically meaningful temporal span for follow up	In the Editorial, BMJ, September 29, Olav Meirik and Timothy M M Farley comment on the results reported by Hannaford et al regarding the effects of oral contraceptives on cancer risks. Merik and Farley conclude that the pill is a safe contraceptive method with respect to cancer. However, there are some methodological problems related to the temporal assumptions made by Hannaford and colleagues. In a time to event analysis it is possible to identify a temporal span within which most cancers occur and the cancer cases decline after a top or slowly continue to rise. A long-term cohort study of the risk of cancer needs to define a hypothetical time span within which most cancers in the population at risk are assumed to develop and demonstrate clinically significant and traceable symptoms. In the study of Hannaford et al a definition of a sufficient or risk generating temporal span is lacking or even discussed. The Authors report results from two studies, the general practitioners subset completed in 1996 and the final data set completed in 2004. Whereas the general practitioners' dataset comprises a period from the initial inclusion in 1968 during 14 months to 1996 the final and second study included all the participants that the Authors were able to follow until 2004. This means that all those dying from cancer or other reasons during the whole period were excluded from the final analysis. This could weaken the final results on the risk and the resulting conclusions on the safety might be unjustified. The general practitioners’ dataset leads namely to opposite warrants on the “safety” of oral contraceptives as this analysis shows that there were clear trends of increasing rates of cervical and central nervous system or pituitary cancer with longer duration of oral contraceptives use. Thus, a hypothetic definition of the temporal span within which cancers are expected to develop such as they could be clinically traced, appears to be necessary. There is obviously need of demarcation of a clinically meaningful time period of follow-up of the cancer risk due to a defined exposure. The longer period of follow-up the larger numbers of those dying in cancers or other reasons and these women are excluded from the final analysis at the same time as more unrelated or confounding factors might intervene on the final results. Competing interests: None declared