Difficulties in Translating Relative Risks Into Absolute Risk

PARIS--When counseling women about breast cancer risk, physiciansface the difficulty of translating relative risks into real-lifeprospects, Michael Baum, CHM, FRCS, of the Royal Marsden Hospital,London, said in a presentation at the Eighth Annual European CancerConference (ECCO-8).

The lay public's perception is that about one in 12 women willdevelop breast cancer at some point in life, Prof. Baum said,and, thus, a woman who is told that she has a threefold increasein risk may go through life believing that her chances of developingbreast cancer in any one year are 25%.

"This is cruel and we have ourselves to blame for all thefear that's out there in the community," he said. "Thepublic has difficulty in translating relative risk into absoluterisk."

Prof. Baum's counseling strategy is to advise women not to thinkabout their lifetime risk, but about what may happen in the nextdecade. For example, he said, a normal 50-year-old woman has a2% chance of developing breast cancer over the next decade, and,thus, even a threefold increase in the odds means only a 6% risk.

"Counseled that way, most of my patients can live with thatrisk without being forced into inappropriate and experimentalrisk avoidance schemes," Prof. Baum said.

It may be more difficult, he acknowledged, to counsel the youngpatient with a suggestive family history. He noted that computerprograms are now available that can calculate the probabilityof penetrance of a dominant gene.

A woman with a family history of breast cancer may have a rareinherited gene with a high penetrance, such as BRCA1, BRCA2, andp53; a common inherited gene with low penetrance; or sporadicmutations due to an inherited deficiency of DNA repair mechanisms,Prof. Baum commented.

"It is likely that the majority of breast cancers are dueto sporadic mutations with hormonal and environmental promotionalfactors, such as reproductive history and diet," he said.

To calculate the relative risk of breast cancer, Prof. Baum said,it is best to rely on observational rather than experimental approaches.

Factors associated with a fourfold or greater increase in risk,he noted, include age, a history of early breast cancer in twofirst-degree relatives, and a history of breast cancer in thecontralateral breast.

On the other hand, many factors are associated with a less thantwofold increase in risk, including age at menarche, age at firstpregnancy, age at menopause, oral contraceptive use, and use ofhormone replacement therapy.

He advised physicians not to focus undue attention on these latterrisk factors, since relative risks of two or less are within theexperimental error of the statistical technique.

Prof. Baum emphasized that no intervention, including prophylacticmastec-tomy, can guarantee 100% protection, and that even thevalue of mammographic screening and follow-up must be consideredunproven and experimental.

Dual Contraception-Prevention?

The possibility of chemoprevention in women at high or moderatelyincreased risk, he said, would be practical only if incidentalto another health-promoting activity, such as contraception orhormone replacement therapy.

Prof. Baum believes it would be entirely plausible to developa contraceptive regimen that would couple a gonadotropin-releasinghormone (GnRH) agonist, to suppress LH/FSH production and ovulation,and also protect the breast, with a low-dose of estradiol sufficientto protect the skeleton without stimulating the epithelium ofthe breast or endometrium. Such a strategy for high-risk premenopausalwomen is now being piloted at the Royal Marsden Hospital.

Prof. Baum also reminded the audience that the use of tamoxifen(Nolvadex) to prevent breast cancer in healthy women is stillunder study and cannot be recommended outside the context of clinicaltrials.

Chance of Developing Breast Cancer by Age (Average-Risk)

By age 25 .......................... one in 19,608

By age 30 .......................... one in 2,525

By age 35 .......................... one in 622

By age 40 .......................... one in 217

By age 45 .......................... one in 93

By age 50 .......................... one in 50

By age 55 .......................... one in 33

By age 60 .......................... one in 24

By age 65 .......................... one in 17

By age 70 .......................... one in 14

By age 75 .......................... one in 11

By age 80 .......................... one in 10

By age 85 .......................... one in 9

Ever .................................. one in 8

Source: NCI Surveillance Program

Who Gets Breast Cancer?

Estimated New Breast Cancer Cases in Women by Age, 1996

Age Estimate* Percent

20-29 510 0.3%

30-39 8,700 4.7%

40-49 33,400 18.1%

50-59 30,900 16.8%

60-69 40,000 21.7%

70-79 44,700 24.3%

80+ 26,000 14.1%

Total 184,300 100.0%

*Estimates may not add to total due to rounding.

Source: American Cancer Society, Surveillance Research, 1995.Data from SEER, 1995.