Clinical review

Screening for breast cancer

R W Blamey, A R M Wilson, J Patnick. 

Lack of knowledge of the pathogenesis of breast cancer means that primary prevention is currently a distant prospect for the majority of women. Early detection represents an alternative approach for reducing mortality from this disease.

	Methods of screening

Top Methods of screening Organisational aspects of... Recommendations for screening The basic screen Benefits and potential...

There is no evidence that clinical examination, breast ultrasonography, or teaching self examination of the breast are effective tools for early detection. However, randomised controlled trials have shown that screening by mammography can significantly reduce mortality from breast cancer by up to 40% in those who attend. The benefit is greatest in women aged 50-70 years. Published data from the combined Swedish trials showed an overall reduction in breast cancer mortality of 29% during 12 years of follow up in women aged over 50 who were invited for screening.

View larger version (100K): [in this window] [in a new window]   Small carcinoma found on screening (arrow)

Screening tests should be simple to apply, cheap, easy to perform, and easy and unambiguous to interpret and should identify those with disease and exclude those without. Film screen mammography requires high technology equipment, special film and dedicated processing, highly trained radiographers to perform the examinations, and highly trained readers to interpret the films. Mammography is at present the best screening tool available and was the first screening method for any malignancy which has been shown to be of value in randomised trials. The potential benefits of digital mammography remain to be evaluated.

View larger version (28K): [in this window] [in a new window]   Summary of 7-12 years' mortality data from randomised and case-control(*) studies of breast cancer screening. Points and lines represent absolute change in mortality and confidence interval

	Organisational aspects of screening

Top Methods of screening Organisational aspects of... Recommendations for screening The basic screen Benefits and potential...

Over 70% of the target population must accept the invitation to participate if screening is to reduce mortality significantly, and the cost per life year saved rises if fewer participate. To achieve optimal participation accurate lists of names, ages, and current addresses are required. Factors affecting attendance for screening include the level of encouragement by general practitioners, knowledge about the screening programme, and the views and experiences of family and friends. Screening programmes must include the initial screening process, assessment of screen detected abnormalities, and clearly defined treatment pathways.

Standards must be set to ensure that targets for mortality reduction are achieved and that there is quality assurance at each stage of the screening process. Screening and assessment should be carried out by multidisciplinary teams experienced in the management of breast disease. Specific training and regular education programmes related to screening should be mandatory for all professionals involved and there should be regular audit and review of individual and programme results and performance.

	Recommendations for screening

Top Methods of screening Organisational aspects of... Recommendations for screening The basic screen Benefits and potential...

Age range
Current data indicate that the reduction in mortality is greatest in women aged 50-70 (29%). A smaller reduction in mortality of 24% is achievable in younger women (40-50), but screening is less cost effective because of the lower incidence of breast cancer in these women. In Europe the consensus view is that mammographic screening of younger women on a population basis cannot be justified.

View larger version (12K): [in this window] [in a new window]   Rates of interval cancer after a negative screen in women aged 50-69

Frequency of screening
The interval between mammographic screens was selected from evidence from the Swedish studies. A UKCCCR trial comparing annual with standard triennial mammographic screens has shown a small and insignificant advantage to annual screening of women. For women aged 50 to 60, the appropriate screening interval is likely to be between two and three years. Screening in women aged under 50 may need to be repeated more frequently.

Screening method
There is clear evidence that two mammographic views of each breast (mediolateral oblique and craniocaudal) significantly improves both sensitivity, particularly for small breast cancers, and specificity. A comparison of performance in screening units in the UK demonstrated a 42% increase in the detection of carcinomas measuring <15 mm in those using two views. The additional radiation dose of two-view mammography is only of concern in the few women with large dense breasts. Data from the UK screening programme also indicate significant improvements in small cancer detection rates when the mammographic film density is between 1.4 and 1.8. Double reading of films improves sensitivity by 5-10%.

View larger version (99K): [in this window] [in a new window]   Discrete lesions identified on screening. Ultrasound of the lesion showed it to be benign

	The basic screen

Top Methods of screening Organisational aspects of... Recommendations for screening The basic screen Benefits and potential...

The first part of screening is the basic screen. The radiologist is responsible for ensuring appropriate levels of sensitivity and specificity. Among women aged 50-52, a minimum of 36 invasive cancers and four ductal in situ cancers (DCIS) should be detected for every 10 000 attenders at an initial (prevalent) screen. At subsequent screens (age 53-64) at least 40 screen detected invasive cancers and five DCIS per 10 000 are expected. More than 50% of all invasive cancers detected should be less than 15 mm in diameter (measured pathologically). Recall rates for assessment should be less than 7% among prevalent attendees and less than 5% at subsequent screens. Women with a "normal" screening outcome should be informed of their result by letter within two weeks. Patients judged to have an important abnormality require further assessment.

There are only two possible end points to assessment: no significant abnormality or a diagnosis of breast cancer.

Assessment should be by the triple approach combining further imaging (mammography and ultrasound) with clinical examination and proceeding to needle biopsy where indicated. Assessment is best carried out by a dedicated assessment team consisting of an experienced radiologist, surgeon, and pathologist supported by radiographers and a breast care nurse.

Approximately two thirds of screen detected abnormalities prove to be unimportant on further mammography or ultrasound examination. When a significant abnormality is thought to be present, diagnosis by either fine needle aspiration (FNA) or needle core biopsy should be attempted after clinical assessment. Automated wide bore (14 gauge) needle core biopsy provides a histological diagnosis which has the advantage of differentiating invasive from in situ disease, but unlike FNA the result cannot be made available immediately. An 11 gauge vacuum assisted biopsy device is now available which, because it provides more tissue, increases the diagnostic yield when biopsying microcalcification. Core biopsies should be x rayed to ensure sufficient calcification has been sampled.

View larger version (126K): [in this window] [in a new window]   Impalpable stellate lesion detected by screening. Lesion is either a radial scar or an invasive carcinoma, and so excision is required even if results of cytology or core biopsy are reported as benign

View larger version (109K): [in this window] [in a new window]   Histology of radial scar

Image guided biopsy of impalpable lesions using ultrasound, or x ray stereotaxis for abnormalities not visible on ultrasound, is highly accurate. Up to 70% of important abnormalities detected by screening are impalpable, and image guided fine needle aspiration or core biopsy is necessary. Impalpable lesions may be localised by ultrasonography if visible on this modality or by mammography. Ultrasound guided biopsy is the method of choice as it is more accurate, quicker, easier to perform, cheaper, and associated with less patient discomfort than x ray guided techniques. Ultrasound is also an accurate means of performing needle biopsy of palpable abnormalities. For a small number of lesions, such as calcifications and architectural distortions, neither FNA nor needle core biopsy provides a clear diagnosis, and in these cases vacuum core biopsy sampling (such as the Mammotome probe) or very wide bore biopsy (such as the ABBI system) may be considered. Stellate lesions should be excised even when the FNA at core biopsy indicates benign disease to ensure a cancer is not missed. The vast majority of benign lesions can be diagnosed by these techniques, and open surgery to establish a diagnosis should be avoided. For malignant lesions definitive preoperative diagnosis can be achieved in over 95% of invasive cancers. The minimum standard for preoperative diagnosis of cancers in the NHSBSP is 70%.

View larger version (104K): [in this window] [in a new window]   Patient with a stellate lesion seen on mammography (left). Diagnostic work up included a magnification mammograph (top right). The lesion was investigated and found to be a cancer and then excisedspecimen x ray showing complete excision (bottom right)

View larger version (236K): [in this window] [in a new window]   Fine needle aspiration: performed freehand (left) and guided by ultrasound image (right)

Palpable lesions
Fine needle aspiration of palpable lesions is usually carried out freehand but can be image guided if there is doubt that the palpable lesion coincides with the radiological abnormality. Image guided aspiration is of value if the first freehand aspiration fails to achieve a definitive diagnosis. There may be advantages to having the results of fine needle aspiration cytology reported immediately.

Multidisciplinary assessment
When results of all diagnostic procedures are available, they are discussed by the multidisciplinary team, who together decide on appropriate management. Preoperative diagnosis of cancer facilitates informed patient counselling and choice of treatments; it also allows the surgeon to plan definitive treatment as a one stage surgical procedure in most patients and avoids the need for frozen section.

View larger version (201K): [in this window] [in a new window]   A core biopsy low power (left), high power (right) showing an invasive lobular carcinoma

Localisation biopsy and excision
Impalpable lesions need to be localised for surgery. This can be achieved by placing a hooked wire under image guidance in the tissues adjacent to the lesion. The surgeon can then identify the site of the abnormality and excise it. Accurate placement of the localising wire is essential. A variety of wire localization systems are available.

View larger version (70K): [in this window] [in a new window]   Radiographs for stereotactic guided fine needle aspiration. Needle can be seen penetrating the lesion

View larger version (157K): [in this window] [in a new window]   Mammogram after placement of hooked wire adjacent to mammographic lesion. Lateral (left) and craniocaudal (right) views

	Benefits and potential drawbacks of screening

Top Methods of screening Organisational aspects of... Recommendations for screening The basic screen Benefits and potential...

Characteristics of screen detected cancers
Compared with symptomatic cancers, screen detected cancers are smaller and more likely to be non-invasive (in situ), while any invasive cancers detected are more likely to be better differentiated, of special type, and node negative. The ability of screening to influence mortality from breast cancer indicates that early diagnosis identifies breast cancers at an earlier stage in their evolution when the chances of metastatic disease being present is smaller.

Psychological morbidity induced by screening
No increase in anxiety has been found in women invited to attend breast screening. There does appear to be a short term increase in anxiety associated with recall for assessment, but, by three months after attending for assessment, women who are shown to have no important abnormality (false positives) are no more anxious than control women. It has been suggested that the excess years as a breast cancer patient caused by a cancer being diagnosed earlier might diminish a patient's quality of life, but the psychological morbidity in women with screen detected breast cancer has been reported to be similar to or less than that in age matched controls.

View larger version (100K): [in this window] [in a new window]   Cosmetic result of recent diagnostic excision biopsysmall scar and no loss of tissue

View larger version (18K): [in this window] [in a new window]   Relation between node positivity and tumour size for screen detected and symptomatic breast cancers

Risks of mammography
It has been calculated that for every two million women aged over 50 who have been screened by means of a single mammogram, one extra cancer a year after 10 years may be caused by the radiation delivered to the breast. Compared with an incidence of breast cancer that approaches 2000 in every million women aged 60, this risk is very small.

View larger version (148K): [in this window] [in a new window]   Specimen radiograph of therapeutic excision showing wide clearance margins around impalpable lesion. Ligaclips aid orientation1 anterior, 2 medial, 3 inferior

Unnecessary biopsies
A proportion of women who undergo biopsy will be found not to have cancer, but in Britain the number of women undergoing a biopsy for benign disease is small. The proportion of benign biopsies performed in a screening programme should be monitored and compared with that in an unscreened group of women of the same age. Women who require biopsy are likely to be extremely anxious, but there is no evidence that this anxiety is sustained if the results are benign.

	Acknowledgments

The sources of the data presented in illustrations are: J M Dixon and J R C Sainsbury, Handbook of Diseases of the Breast (Churchill Livingstone) 1993:86 for the graph of results of trials of screening; L Tabar et al, Br J Cancer 1987;55:547-51 for the graph of rates of interval cancers between screens; T J Anderson et al, Br J Cancer 1991;64:108-13 for the graph of node positivity and cancer size for screen detected and symptomatic cancers; and N E Day, Br Med Bull 1991;47:400-17 (copyright British Council) for the table of observed and expected detection of cancer by screening. The data are reproduced with permission of the journals or copyright holders.

	Footnotes

R W Blamey is professor of surgical science and A R M Wilson is consultant radiologist, Nottingham City Hospital, Nottingham, and J Patnick is national coordinator of the NHS Breast Screening Programme, Sheffield.

The ABC of breast diseases is edited by J Michael Dixon, consultant surgeon and senior lecturer in surgery, Edinburgh Breast Unit, Western General Hospital, Edinburgh.

This is one of three completely updated chapters from the new edition of the BMJ book ABC of Breast Diseases. The first was published last week, and the third will appear next week.