Mammography—Variables Associated with Accuracy

Patient Characteristics
Tumor Characteristics
Physician Characteristics
Facility Characteristics
International Comparisons
Prevalent Versus Subsequent Examination and the Interval Between Exams
Digital Mammography
Mammography and CAD

Several characteristics of women being screened that are associated with the accuracy of mammography include age, breast density, whether it is the first or subsequent exam, and time since last mammogram. Younger women have lower sensitivity and higher false-positive rates on screening mammography than do older women (refer to the Breast Cancer Surveillance Consortium performance measures by age for more information).

For women of all ages, high breast density is associated with 10% to 29% lower sensitivity.[1] High breast density is an inherent trait, which can be familial [2,3] but also may be affected by age, endogenous [4] and exogenous [5,6] hormones,[7] selective estrogen receptor modulators such as tamoxifen,[8] and diet.[9] Hormone therapy is associated with increased breast density and is associated not only with lower sensitivity but also with an increased rate of interval cancers.[10]

The Million Women Study in the United Kingdom revealed three patient characteristics that were associated with decreased sensitivity and specificity of screening mammograms in women aged 50 to 64 years: use of postmenopausal hormone therapy, prior breast surgery, and body mass index below 25.[11] In addition, a longer interval since the last mammogram increases sensitivity, recall rate, and cancer detection rate and decreases specificity.[12]

Strategies have been proposed to improve mammographic sensitivity by altering diet, timing mammograms with menstrual cycles, interrupting hormone therapy before the examination, or using digital mammography machines.[13] Obese women have more than a 20% increased risk of having false-positive mammography results compared with underweight and normal weight women, although sensitivity is unchanged.[14]

Some cancers are more easily detected by mammography than other cancers are. In particular, mucinous, lobular, and rapidly growing cancers can be missed because their appearance on x-rays is similar to that of normal breast tissue.[15] Medullary carcinomas may be similarly missed.[16] Some cancers, particularly those associated with BRCA 1/2 mutations, masquerade as benign tumors.[17,18]

Radiologist performance is critical to assessing mammographic interpretive performance, yet there is substantial, well-documented variability among radiologists. Factors that influence radiologists’ performance include their level of experience and the volume of mammograms they interpret.[19] There is often a trade-off between sensitivity and specificity, such that higher sensitivity may be associated with lower specificity. Radiologists in academic settings have a higher positive predictive value (PPV) of their recommendations to undergo biopsy than do community radiologists.[20] Fellowship training in breast imaging may lead to improved cancer detection, but it is associated with higher false-positive rates.[13]

After controlling for patient and radiologist characteristics, screening mammography interpretive performance (specificity, PPV, area under the curve [AUC]) varies by facility and is associated with facility-level characteristics. Higher interpretive accuracy of screening mammography was seen at facilities that offered screening examinations alone, included a breast imaging specialist on staff, did single as opposed to double readings, and reviewed interpretive audits two or more times each year.[21]

False-positive rates vary significantly between facilities performing diagnostic mammography and are higher at facilities where concern about malpractice is high.[22] False-positive rates are also higher at facilities serving vulnerable women (women of racial or ethnic minorities and women with lower educational attainment, limited household income, or rural residence) than at facilities serving nonvulnerable women, perhaps because of poorer compliance with recommendations for follow-up examinations.[23] Analyses that do not adjust for important patient characteristics may falsely conclude that there is more facility variation in overall accuracy than actually exists.[22]

International comparisons of screening mammography have found higher specificity in countries with more highly centralized screening systems and national quality assurance programs.[24,25] For example, one study reported that the recall rate is twice as high in the United States as it is in the United Kingdom, yet there is no difference in the rate of cancers detected. Such comparisons may be confounded by social, cultural, and economic factors.[25]

The likelihood of diagnosing cancer is highest with the prevalent (first) screening examination, ranging from 9 to 26 cancers per 1,000 screens, depending on the woman’s age. The likelihood decreases for follow-up examinations, ranging from 1 to 3 cancers per 1,000 screens.[26] The optimal interval between screening mammograms is unknown. In particular, the breast cancer mortality-focused, randomized, controlled trials (RCTs) used single screening intervals with little variability across the trials. A prospective United Kingdom trial randomly assigned women aged 50 to 62 years to receive mammograms annually or at the standard 3-year interval. Although the grade and node status were similar in both groups, more cancers of slightly smaller size were detected in the annual screening group, with a lead time of approximately 7 months in comparison with triennial screening.[27]

A large observational study found a slightly increased risk of late-stage disease at diagnosis for women in their 40s who were adhering to a 2-year versus a 1-year schedule (28% vs. 21%; odds ratio = 1.35; 95% confidence interval [CI], 1.01–1.81), but no difference was seen for women in their 50s or 60s.[28,29]

A Finnish study of 14,765 women aged 40 to 49 years assigned women born in even-numbered years to annual screens and women born in odd-numbered years to triennial screens. The study was small in terms of number of deaths, with low power to discriminate breast cancer mortality between the two groups. There were 18 deaths from breast cancer in 100,738 life-years in the triennial screening group and 18 deaths from breast cancer in 88,780 life-years in the annual screening group (hazard ratio, 0.88; 95% CI, 0.59–1.27).[30]

Digital mammography is more expensive than screen-film mammography (SFM) but is more amenable to data storage and sharing. Performance of both technologies has been compared directly in several trials yielding similar results.

A large cohort of women undergoing both types of mammography was evaluated at 33 U.S. centers in the Digital Mammographic Imaging Screening Trial (DMIST), showing no differences in mammographic sensitivity and specificity. Digital mammography had a higher sensitivity in premenopausal and perimenopausal women, in women younger than 50 years, and in women with dense breasts, according to a planned subset analysis.[13] Digital mammography was associated with lower sensitivity among women older than 65 years.

This approach, in which two tests are applied to the same individuals in a single arm, may yield a biased estimate of the relative sensitivity of the tests. If one of the tests detects overdiagnosed cancers to a greater extent than the other, the test that detects fewer cancers may appear to have lower sensitivity even though it may result in less harm and provide a better cost-to-benefit ratio. An alternative design would randomize individuals to two arms with each test performed in only one arm, compare the interval cancer rates in each arm, and compare the relative sensitivities of the two tests.

An Italian trial of parallel cohorts of 14,385 women matched for age and interpreting radiologist were screened by either full-field digital mammography or SFM. Recall rate and cancer detection rate, especially for clustered microcalcifications, were higher for digital mammography, whereas the recall rate for poor technical quality was higher for SFM. PPV—the probability that an individual with a positive screening result has the disease—was the same.[31]

The Oslo II Study randomly assigned women to screening by digital mammography (n = 6,944) or SFM (n = 16,985) with soft-copy double reading by experienced radiologists. Recall and cancer detection rates were higher for digital mammography, but there was no difference in PPV or incidence of interval cancers.[32]

A study of a single-center population-based screening program in the Netherlands compared women aged 50 to 75 years screened by full-field digital mammography (FFDM) that included computer-aided detection (CAD) with women screened by SFM. In 5 years, 311,082 screening examinations were done by SFM and 56,518 by FFDM. The groups were assembled without obvious bias but without randomization. The recall rate was higher in the FFDM group, but there was no difference in detection of invasive breast cancer. There was higher detection of ductal carcinoma in situ (DCIS) in the FFDM group related to increased detection of clustered microcalcifications.[33]

A review of ten controlled studies of various designs found that, overall, digital mammography increases breast cancer detection (combining invasive cancer and DCIS) and that recall rates are not consistently better with either technology.[34]

CAD systems are designed to help radiologists read mammograms by highlighting suspicious regions such as clustered microcalcifications and masses.[35] Generally, CAD systems increase sensitivity and decrease specificity [36] and increase detection of DCIS.[37] Several CAD systems are in use. One large population-based study comparing recall rates and breast cancer detection rates before and after the introduction of CAD systems found no change in either rate.[35,38] Another large study noted an increase in recall rate and increased DCIS detection but no improvement in cancer detection rate.[37]

1. Rosenberg RD, Hunt WC, Williamson MR, et al.: Effects of age, breast density, ethnicity, and estrogen replacement therapy on screening mammographic sensitivity and cancer stage at diagnosis: review of 183,134 screening mammograms in Albuquerque, New Mexico. Radiology 209 (2): 511-8, 1998.  [PUBMED Abstract]
   
   
2. Pankow JS, Vachon CM, Kuni CC, et al.: Genetic analysis of mammographic breast density in adult women: evidence of a gene effect. J Natl Cancer Inst 89 (8): 549-56, 1997.  [PUBMED Abstract]
   
   
3. Boyd NF, Dite GS, Stone J, et al.: Heritability of mammographic density, a risk factor for breast cancer. N Engl J Med 347 (12): 886-94, 2002.  [PUBMED Abstract]
   
   
4. White E, Velentgas P, Mandelson MT, et al.: Variation in mammographic breast density by time in menstrual cycle among women aged 40-49 years. J Natl Cancer Inst 90 (12): 906-10, 1998.  [PUBMED Abstract]
   
   
5. Harvey JA, Pinkerton JV, Herman CR: Short-term cessation of hormone replacement therapy and improvement of mammographic specificity. J Natl Cancer Inst 89 (21): 1623-5, 1997.  [PUBMED Abstract]
   
   
6. Laya MB, Larson EB, Taplin SH, et al.: Effect of estrogen replacement therapy on the specificity and sensitivity of screening mammography. J Natl Cancer Inst 88 (10): 643-9, 1996.  [PUBMED Abstract]
   
   
7. Baines CJ, Dayan R: A tangled web: factors likely to affect the efficacy of screening mammography. J Natl Cancer Inst 91 (10): 833-8, 1999.  [PUBMED Abstract]
   
   
8. Brisson J, Brisson B, Coté G, et al.: Tamoxifen and mammographic breast densities. Cancer Epidemiol Biomarkers Prev 9 (9): 911-5, 2000.  [PUBMED Abstract]
   
   
9. Boyd NF, Greenberg C, Lockwood G, et al.: Effects at two years of a low-fat, high-carbohydrate diet on radiologic features of the breast: results from a randomized trial. Canadian Diet and Breast Cancer Prevention Study Group. J Natl Cancer Inst 89 (7): 488-96, 1997.  [PUBMED Abstract]
   
   
10. Crouchley K, Wylie E, Khong E: Hormone replacement therapy and mammographic screening outcomes in Western Australia. J Med Screen 13 (2): 93-7, 2006.  [PUBMED Abstract]
    
    
11. Banks E, Reeves G, Beral V, et al.: Influence of personal characteristics of individual women on sensitivity and specificity of mammography in the Million Women Study: cohort study. BMJ 329 (7464): 477, 2004.  [PUBMED Abstract]
    
    
12. Yankaskas BC, Taplin SH, Ichikawa L, et al.: Association between mammography timing and measures of screening performance in the United States. Radiology 234 (2): 363-73, 2005.  [PUBMED Abstract]
    
    
13. Pisano ED, Gatsonis C, Hendrick E, et al.: Diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med 353 (17): 1773-83, 2005.  [PUBMED Abstract]
    
    
14. Elmore JG, Carney PA, Abraham LA, et al.: The association between obesity and screening mammography accuracy. Arch Intern Med 164 (10): 1140-7, 2004.  [PUBMED Abstract]
    
    
15. Porter PL, El-Bastawissi AY, Mandelson MT, et al.: Breast tumor characteristics as predictors of mammographic detection: comparison of interval- and screen-detected cancers. J Natl Cancer Inst 91 (23): 2020-8, 1999.  [PUBMED Abstract]
    
    
16. Wallis MG, Walsh MT, Lee JR: A review of false negative mammography in a symptomatic population. Clin Radiol 44 (1): 13-5, 1991.  [PUBMED Abstract]
    
    
17. Tilanus-Linthorst M, Verhoog L, Obdeijn IM, et al.: A BRCA1/2 mutation, high breast density and prominent pushing margins of a tumor independently contribute to a frequent false-negative mammography. Int J Cancer 102 (1): 91-5, 2002.  [PUBMED Abstract]
    
    
18. Ganott MA, Harris KM, Klaman HM, et al.: Analysis of False-Negative Cancer Cases Identified with a Mammography Audit. Breast J 5 (3): 166-175, 1999.  [PUBMED Abstract]
    
    
19. Elmore JG, Jackson SL, Abraham L, et al.: Variability in interpretive performance at screening mammography and radiologists' characteristics associated with accuracy. Radiology 253 (3): 641-51, 2009.  [PUBMED Abstract]
    
    
20. Meyer JE, Eberlein TJ, Stomper PC, et al.: Biopsy of occult breast lesions. Analysis of 1261 abnormalities. JAMA 263 (17): 2341-3, 1990.  [PUBMED Abstract]
    
    
21. Taplin S, Abraham L, Barlow WE, et al.: Mammography facility characteristics associated with interpretive accuracy of screening mammography. J Natl Cancer Inst 100 (12): 876-87, 2008.  [PUBMED Abstract]
    
    
22. Jackson SL, Taplin SH, Sickles EA, et al.: Variability of interpretive accuracy among diagnostic mammography facilities. J Natl Cancer Inst 101 (11): 814-27, 2009.  [PUBMED Abstract]
    
    
23. Goldman LE, Walker R, Miglioretti DL, et al.: Accuracy of diagnostic mammography at facilities serving vulnerable women. Med Care 49 (1): 67-75, 2011.  [PUBMED Abstract]
    
    
24. Smith-Bindman R, Chu PW, Miglioretti DL, et al.: Comparison of screening mammography in the United States and the United kingdom. JAMA 290 (16): 2129-37, 2003.  [PUBMED Abstract]
    
    
25. Elmore JG, Nakano CY, Koepsell TD, et al.: International variation in screening mammography interpretations in community-based programs. J Natl Cancer Inst 95 (18): 1384-93, 2003.  [PUBMED Abstract]
    
    
26. Kerlikowske K, Grady D, Barclay J, et al.: Positive predictive value of screening mammography by age and family history of breast cancer. JAMA 270 (20): 2444-50, 1993.  [PUBMED Abstract]
    
    
27. The Breast Screening Frequency Trial Group.: The frequency of breast cancer screening: results from the UKCCCR Randomised Trial. United Kingdom Co-ordinating Committee on Cancer Research. Eur J Cancer 38 (11): 1458-64, 2002.  [PUBMED Abstract]
    
    
28. White E, Miglioretti DL, Yankaskas BC, et al.: Biennial versus annual mammography and the risk of late-stage breast cancer. J Natl Cancer Inst 96 (24): 1832-9, 2004.  [PUBMED Abstract]
    
    
29. Mandelblatt JS, Cronin KA, Bailey S, et al.: Effects of mammography screening under different screening schedules: model estimates of potential benefits and harms. Ann Intern Med 151 (10): 738-47, 2009.  [PUBMED Abstract]
    
    
30. Parvinen I, Chiu S, Pylkkänen L, et al.: Effects of annual vs triennial mammography interval on breast cancer incidence and mortality in ages 40-49 in Finland. Br J Cancer 105 (9): 1388-91, 2011.  [PUBMED Abstract]
    
    
31. Del Turco MR, Mantellini P, Ciatto S, et al.: Full-field digital versus screen-film mammography: comparative accuracy in concurrent screening cohorts. AJR Am J Roentgenol 189 (4): 860-6, 2007.  [PUBMED Abstract]
    
    
32. Skaane P, Hofvind S, Skjennald A: Randomized trial of screen-film versus full-field digital mammography with soft-copy reading in population-based screening program: follow-up and final results of Oslo II study. Radiology 244 (3): 708-17, 2007.  [PUBMED Abstract]
    
    
33. Karssemeijer N, Bluekens AM, Beijerinck D, et al.: Breast cancer screening results 5 years after introduction of digital mammography in a population-based screening program. Radiology 253 (2): 353-8, 2009.  [PUBMED Abstract]
    
    
34. Skaane P: Studies comparing screen-film mammography and full-field digital mammography in breast cancer screening: updated review. Acta Radiol 50 (1): 3-14, 2009.  [PUBMED Abstract]
    
    
35. Gur D, Sumkin JH, Rockette HE, et al.: Changes in breast cancer detection and mammography recall rates after the introduction of a computer-aided detection system. J Natl Cancer Inst 96 (3): 185-90, 2004.  [PUBMED Abstract]
    
    
36. Ciatto S, Del Turco MR, Risso G, et al.: Comparison of standard reading and computer aided detection (CAD) on a national proficiency test of screening mammography. Eur J Radiol 45 (2): 135-8, 2003.  [PUBMED Abstract]
    
    
37. Fenton JJ, Taplin SH, Carney PA, et al.: Influence of computer-aided detection on performance of screening mammography. N Engl J Med 356 (14): 1399-409, 2007.  [PUBMED Abstract]
    
    
38. Elmore JG, Carney PA: Computer-aided detection of breast cancer: has promise outstripped performance? J Natl Cancer Inst 96 (3): 162-3, 2004.  [PUBMED Abstract]