12 August, 2013

Meningiomas, which arise from tissues that cover the brain and spinal cord, represent about a fifth of all brain tumors, and they affect women more often than men [1]. The reason for a female predominance is not known, although most meningiomas express receptors for progesterone and some also express estrogen or androgen receptors [2]. Most arise from the arachnoid villi in association with the dura mater or intracranial venous sinuses. Meningiomas are typically slow growing and may compress the brain, but they usually do not invade neural tissues or metastasize to distant sites. Many meningiomas are asymptomatic, discovered only as incidental findings at autopsy or on brain imaging studies performed for some other reason (e.g. ordered because of headache or light-headedness) [3].

In a nationwide case-control study, Danish investigators recently reported associations between menopausal hormone therapy (HT) and meningioma [4]. Their findings, which are largely consistent with other observational research, are also notable from the perspectives of interpretation, perception, and validity.

The authors searched the Danish Cancer Registry for new cases of intracranial meningioma occurring between 2000 and 2009. They identified 924 cases among women aged 55–84 years without prior cancer diagnoses. These women were compared to similar women without meningioma. HT use was ascertained in a prescription registry established in 1995. The primary finding is that women who filled at least two prescriptions for HT during the time under consideration faced a 30% increased risk of meningioma compared to women who filled no prescriptions or just one prescription (odds ratio (OR) 1.3, 95% confidence interval (CI) 1.1–1.5). After at least 10 years of cumulative hormone use, risk was increased by 70% (95% CI 1.2–2.3). A heightened risk was observed among current as well as past users and among estrogen-only as well as estrogen–progestogen users [4].

Comment

These findings provide new evidence for the hypothesis that HT increases risk for this form of brain tumor. Other observational research generally supports a similar inference but is not fully consistent. In a multicenter European study [5], the Nurses’ Health Study [6], a nationwide Finnish cohort study [7], and the Million Women Study [8], meningioma risk was elevated to a similar extent, at least among subgroups of HT users. Risk was limited to current HT users in the European and Nurse’s Health cohorts and to estrogen-only users in the Finnish cohort and the Million Women Study. Other studies report no relation between HT and meningioma risk [9,10].

Associations reported by Andersen and colleagues [4] were sensationally misinterpreted by several news organizations, illustrating difficulties in translating scientific findings into a format easily understood by medical writers and health consumers. On July 2, 2013, the Daily Mail summarized this research for their readers, reporting that women prescribed HT 'for a decade have a 70% chance of developing a tumor' [11]. Wording in The Sun on the same date was nearly identical [12]. It is obvious that the reporter has misconstrued relative risk (odds ratio of 1.7 among women who used HT for at least a decade) for absolute risk. Even so, a relative risk of 70% seems frightfully high. After all, no one wants a brain tumor.

The 2013 International Menopause Society recommendations on menopausal hormone therapy point out patient counseling should consider benefits and risks 'in clear and comprehensible terms, e.g. as absolute numbers rather than, or in addition to, ... relative risk' (p. 303) [13]. The annual incidence of meningioma per 10,000 women (all ages) is about 0.2–0.7 [1]. If, for purposes of illustration, we assume that the incidence of meningioma is about 0.5 per 10,000 in the age group studied by Andersen and colleagues and we accept their estimates of relative risk, we can provide a rough estimate of absolute numbers for women using HT for different lengths of time.

As shown in the Table below, Danish women using any type of HT for 10 years or longer have a relative risk of 1.7; in other words, their risk is increased by 70%. However, this risk turns out to be very rare (less than one additional case per 10,000 women per year). Thus, each year one woman among 11,000 long-term HT users might develop a meningioma (calculated as 10,000 divided by 0.85); among women who never used HT, each year one woman in 20,000 might develop a meningioma (10,000 divided by 0.5). When considered over a longer period of time, for 10,000 women followed for 20 years (200,000 person years), the difference between never users and long-term users would be equivalent to seven additional cases of meningioma (Table). Many of these meningiomas would be expected to be asymptomatic; some would be more serious [3]. This very rare risk is quite different from the alarmist interpretation that women using HT for a decade have a 70% chance of developing a brain tumor, or the alarming perception that a 70% increase in risk represents imminent danger. Importantly also, very few women using HT for its primary indication of moderate to severe vasomotor symptoms will use HT for as long as 10 years.

Table Any type of hormone therapy (HT) and risk of meningioma, per 10,000 women

See text for details; relative risks (odds ratios) are from Table 4 of Andersen et al.⁴. Actual incidences represent estimates under the assumption of HT use for the durations shown

Odds ratios for the association between HT use and meningioma reported by Andersen and colleagues were statistically significant but of modest strength (1.3 for any HT use and 1.7 for use of 10 or more years). In large clinical trials, where the randomization process balances out both recognized and unrecognized sources of confounding, these risks would be interpreted as valid and clinically meaningful, even though findings might still have occurred on the basis of chance alone. This interpretation differs from that of similar associations in observational research. For relatively small associations – say, significant risk reductions above 0.5 or significant risk elevations below 2, within the range reported by Andersen and colleagues [4] – the validity is uncertain.

In the Danish case-control study, detection (surveillance) bias is an unsettling concern: were women who did not use HT, compared to women who received HT prescriptions, more likely to have small, undetected, incidental meningiomas? The answer may well be, 'yes'.

Study investigators correctly point out that free access to health care in Denmark reduces the likelihood of detection bias, but free access does not fully level the playing field. In these analyses, HT users were better educated than non-users [4]. Education may influence decisions to use HT and to seek medical assistance for a variety of symptoms and, hence, to undergo diagnostic evaluation [14,15].

In their careful consideration of potential confounders, the authors point out potential links between meningiomas and allergies, diabetes, stroke, and certain medications (statins, aspirin, NSAIDs, anti-asthma drugs, and antihistamines) [4]. Each of these potential confounders was indeed found significantly more often in their cases than controls. It is possible that women with meningiomas had poorer health (e.g. more often diabetic or hyperlipidemic) than other women. It is also possible that they had more health-seeking behaviors that contributed both to HT use and also to use of imaging procedures. The authors adjusted for recognized confounders, but they were unable to adjust for confounders they did not identify.

These considerations highlight some of the inherent difficulties in conducting even high-quality observational research. They do not mean that the Danish investigators are wrong or should have approached the question differently. They do reduce the certainty of any conclusion regarding a causal role for HT in determining meningioma risk.

Victor W. Henderson
Departments of Health Research & Policy and of Neurology & Neurological Sciences, Stanford University, Stanford, California, United States

References

1. Longstreth WT Jr, Dennis LK, McGuire VM, Drangsholt MT, Koepsell TD. Epidemiology of intracranial meningioma. Cancer 1993;72:639-48. 
http://www.ncbi.nlm.nih.gov/pubmed/8334619

2. Leães CG, Meurer RT, Coutinho LB, Ferreira NP, Pereira-Lima JF, da Costa Oliveira M. Immunohistochemical expression of aromatase and estrogen, androgen and progesterone receptors in normal and neoplastic human meningeal cells. Neuropathology 2010;30:44-9.
http://www.ncbi.nlm.nih.gov/pubmed/19703265

3. Chamoun R, Krisht KM, Couldwell WT. Incidential meningiomas. Neurosurg Focus 2011;31:E19. 
http://www.ncbi.nlm.nih.gov/pubmed/22133182

4. Andersen L, Friis S, Hallas J, Ravn P, Schrøder HD, Gaist D. Hormone replacement therapy increases the risk of cranial meningioma. Eur J Cancer 2013 June 22. Epub ahead of print.
http://www.ncbi.nlm.nih.gov/pubmed/23800670

5. Michaud DS, Gallo V, Schlehofer B, et al. Reproductive factors and exogenous hormone use in relation to risk of glioma and meningioma in a large European cohort study. Cancer Epidemiol Biomarkers Prev 2010;19:2562-9.
http://www.ncbi.nlm.nih.gov/pubmed/20802020

6. Jhawar BS, Fuchs CS, Colditz GA, Stampfer MJ. Sex steroid hormone exposures and risk for meningioma. J Neurosurg 2003;99:848-53.
http://www.ncbi.nlm.nih.gov/pubmed/14609164

7. Korhonen K, Auvinen A, Lyytinen H, Ylikorkala O, Pukkala E. A nationwide cohort study on the incidence of meningioma in women using postmenopausal hormone therapy in Finland. Am J Epidemiol 2012;175:309-14.
http://www.ncbi.nlm.nih.gov/pubmed/22287638

8. Benson VS, Pirie K, Green J, et al. Hormone replacement therapy and incidence of central nervous system tumours in the Million Women Study. Int J Cancer 2010;127:1692-8.
http://www.ncbi.nlm.nih.gov/pubmed/20091865

9. Johnson DR, Olson JE, Vierkant RA, et al. Risk factors for meningioma in postmenopausal women: results from the Iowa Women's Health Study. Neuro Oncol 2011;13:1011-19.
http://www.ncbi.nlm.nih.gov/pubmed/21750006

10. Claus EB, Calvocoressi L, Bondy ML, Wrensch M, Wiemels JL, Schildkraut JM. Exogenous hormone use, reproductive factors, and risk of intracranial meningioma in females. J Neurosurg 2013;118:649-56.
http://www.ncbi.nlm.nih.gov/pubmed/23101448

11. Daily Mail Article Accessed July 3, 2013 

12. The Sun Article. Accessed July 10, 2013 

13. de Villiers TJ, Pines A, Panay N, et al. Updated 2013 International Menopause Society recommendations on menopausal hormone therapy and preventive strategies for midlife health. Climacteric 2013;16:316-37.
http://www.ncbi.nlm.nih.gov/pubmed/23672656

14. Keating NL, Cleary PD, Rossi AS, Zaslavsky AM, Ayanian JZ. Use of hormone replacement therapy by postmenopausal women in the United States. Ann Intern Med 1999;130:545-53.
http://www.ncbi.nlm.nih.gov/pubmed/10189323

15. Hansen AH, Halvorsen PA, Ringberg U, Førde OH. Socio-economic inequalities in health care utilisation in Norway: a population based cross-sectional survey. BMC Health Serv Res 2012;12:336.
http://www.ncbi.nlm.nih.gov/pubmed/23006844