Screening for Breast Cancer


Breast cancer is the third most common cause of death in women and excluding skin cancer is the most common cancer in women.

There were an estimated 16,300 new cases of breast cancer in Canada in 1993 and an estimated 5,400 deaths. Over the last 20 years the incidence rate has increased by about 15% whereas the mortality has remained relatively stable.

There is an east-west gradient with lower rates in the east. Risk factors for breast cancer include hormonal, dietary and hereditary factors. Early menarche, late menopause and delayed first pregnancy are associated with higher risk.

There is some evidence linking high intake of dietary fat to risk of breast cancer; family history, obesity, alcohol use, ionizing radiation and post-menopausal estrogen replacement therapy (see Chapter 52) have also been associated with increased risk, while the evidence for oral contraceptives is more controversial.

There are three maneuvers to be considered. They are clinical examination of the breasts, mammography, and self-examination of the breasts. Some of the seven RCTs carried out these screening maneuvers in combination, and some separately; the prescribed frequencies of the mammographic screenings varied from 12 to 33 months.

The sensitivities and specificities of detection varied widely between the trials depending on the maneuver(s) that were employed, the length of the interval between screenings, the underlying incidence of the disease and the method of calculating estimates of the screening proficiency.

For sensitivity the range was 46-88%; for specificity, it was 82-99.9%. In the Canadian trial the sensitivity, using the ratio of screen-detected cases to all cases, of annual mammography plus clinical examination was 88% in women 50-59 and 81% in the women 40-49. Specificity, using surgical biopsy as the definition of a positive case, ranged from 96.5% to 99.9%.

For the younger women, the ratio of benign biopsy to malignant biopsy was about 9:1 on the first screen, dropping to 6:1 on later screens. For women 50 and over, it was about 5:1 on the initial screen, dropping to 3:1 for later screens.

Effectiveness of Prevention and Treatment

Trials for different age

The seven screening trials enrolled women whose age at entry ranged from 40 to 74. In three of the trials individuals were the unit of randomization; in the the other four the units were neighbourhood or practice clusters. The four trials which were located in Sweden investigated only the benefits of mammography.

The original trial, that of the Health Insurance Plan (HIP) of New York, demonstrated a significant mortality reduction with a relative risk (RR) of 0.45 five years after entry in women aged 50 to 59. After nine years this rose to 0.67. The combined Swedish trials, after 7 to 12 years of follow-up, also showed a significant benefit extending to the age of 69 (RR=0.71).

The Edinburgh trial at 10 years of follow-up produced a non-significant benefit (RR=0.85). Compliance in this trial was poor though and the study and control groups were found to differ on factors that could have affected survival. The Canadian trial, which in the 50-59 year old age group looked only at the benefits of mammography over and above that of annual clinical examination, found an improved survival rate (RR=0.97), but at 7 years of follow-up the improvement was not statistically significant.

Most interest in breast screening has centered about benefit for the 40-49-year-old women. In this age group the HIP showed a non-None significant decrease in mortality (RR=0.95) five years after entry. At nine years this had dropped to 0.81 but it was still not significant.

However, the number of women in the group was not large and consequently the power of the test was low. Because of these two factors the Task Force in its 1986 report gave a C Recommendation for the 40-49 year olds. Now with seven trials reporting, and two of them, the Swedish Two County trial and the Canadian trial, having large numbers of compliant enrollees, there is now a considerable amount of evidence.

None of the trials showed a significant benefit. The relative risks ranged from 0.51 to 1.36. In the five trials that had reported early age-specific follow-up, increased mortality occurred in this younger age group in all of the studies (in the Malmo trial the younger age group was composed of women 45-54). In the HIP the excess of deaths disappeared after 3 years but in the other trials it lasted for seven or eight years. Of all the trials, the Canadian study showed the greatest excess, RR=1.36 (95% confidence interval: 0.84-2.21); this was after seven years of follow-up.

This is not really surprising since the Canadian trial was the only one of the seven which was an efficacy trial and the results of efficacy trials are expected to be more extreme than those of effectiveness trials.

Breast Self-Examination

The result of studies

Before the introduction of mass screening programs, the vast majority of tumours were reputed to have been detected by the women themselves. As a consequence of this, breast self-examination(BSE) was and is advocated by various bodies and organizations in the hopeful expectation that early detection will result in improved survival.

Five studies have shown an association between the practice of BSE and factors associated with better survival, such as stage, tumour size or axillary node involvement, but other studies have shown no benefit.

Four studies compared the survival rates from breast cancer in women who had been taught or practiced BSE and in those who had not been instructed or did not practice it. Foster looked at those who had performed it regularly with those who had not.

At five years the respective rates were 75% and 57%. Locker compared all those invited to attend an instructive course in BSE with an historical group of cases. The latter had slightly better survival despite having poorer prognostic indicators.

However, after seven years of follow-up those in the instructed group who attended had a significantly lower mortality in contrast with those who did not attend. Le Geyte compared those who practiced BSE with those who had never been taught it. After 6 years of follow-up the respective survival rates were 73.1% and 66.1% (p=0.07).