Explaining variance in the cumulus mammographic measures that predict breast cancer risk: a twins and sisters study.

BACKGROUND: Mammographic density, the area of the mammographic image that appears white or bright, predicts breast cancer risk. We estimated the proportions of variance explained by questionnaire-measured breast cancer risk factors and by unmeasured residual familial factors. METHODS: For 544 MZ and 339 DZ twin pairs and 1,558 non-twin sisters from 1,564 families, mammographic density was measured using the computer-assisted method Cumulus. We estimated associations using multilevel mixed-effects linear regression and studied familial aspects using a multivariate normal model. RESULTS: The proportions of variance explained by age, body mass index (BMI), and other risk factors, respectively, were 4%, 1%, and 4% for dense area; 7%, 14%, and 4% for percent dense area; and 7%, 40%, and 1% for nondense area. Associations with dense area and percent dense area were in opposite directions than for nondense area. After adjusting for measured factors, the correlations of dense area with percent dense area and nondense area were 0.84 and -0.46, respectively. The MZ, DZ, and sister pair correlations were 0.59, 0.28, and 0.29 for dense area; 0.57, 0.30, and 0.28 for percent dense area; and 0.56, 0.27, and 0.28 for nondense area (SE = 0.02, 0.04, and 0.03, respectively). CONCLUSIONS: Under the classic twin model, 50% to 60% (SE = 5%) of the variance of mammographic density measures that predict breast cancer risk are due to undiscovered genetic factors, and the remainder to as yet unknown individual-specific, nongenetic factors. IMPACT: Much remains to be learnt about the genetic and environmental determinants of mammographic density.

Examining the sensitivity of ultrasound-guided large core biopsy for invasive breast carcinoma in a population screening programme.

INTRODUCTION: To evaluate the sensitivity of ultrasound-guided core-needle biopsy (UCB) in invasive breast carcinoma and to establish causes of false-negative biopsy in a population screening programme. METHOD: We identified 571 consecutive women diagnosed with surgically proven invasive breast cancer. Histology from 14-gauge UCB was compared with surgical histology to identify true-positive and false-negative ultrasound core biopsies. True-positive and false-negative groups were compared for tumour size and histology. On blinded review of UCB images and pathology reports from false negative (n = 20) and a random sample of true-positive cases (n = 80), we compared core sample number and needle visualisation in the lesion. RESULTS: Of 571 carcinomas sampled with UCB, 551 (96.5%) were true positive and 20 (3.5%) were false negative. The mean core number was 2.0 (range 1-3) for false negatives and 2.25 (range 1-4) for true positives (P = 0.27). Mean tumour sizes were 13.3 and 16.2 mm for the false-negative and true-positive groups, respectively (P = 0.25). Tubular carcinomas represented 30% (6/20) of false-negative cases compared with 5.1% (28/551) of the true-positive cases (P < 0.001). On blinded review, needle visualisation within the lesion was demonstrated in 47.4% (9/19) of false-negative cases and 76.3% (61/80) of true-positive cases (P = 0.02). CONCLUSION: We demonstrated a sensitivity of 96.5% with a mean of 2.21 cores. False-negative results were more likely in the absence of post-fire needle position verification and with tubular carcinomas. Neither tumour size nor core number predicted diagnostic accuracy.

Cancer on a mammogram is not memorable: readers remember their recalls and not cancers.

AIM: To determine if presence of cancer on a mammogram makes that mammogram more memorable. MATERIALS AND METHODS: A total of 100 mammograms (25 cancers) were grouped into 5 sets of 20 cases. Set pairs were presented in five reads to eight radiologist readers. Readers were asked to 'clear' or 'call back' cases, and at post-baseline reads to indicate whether each case was 'new' or 'old' (remembered from prior read). Two sets were presented only at baseline, to calculate each reader's false recollection rate. For cases presented more than once ('old' cases, 100 presentations) readers could have 'correct memory' or 'memory loss'. Memory performance was defined as odds ratio of correct memory to memory loss. Multivariate logistic data regression analysis identified predictors of memory performance from: reader, set, time since last read, presence of cancer, and whether the case was called back at the last read. RESULTS: Memory performance differed markedly between readers and reader identity was a highly significant predictor of memory performance. Presence of cancer was not a significant predictor of memory performance (odds ratio 0.77, 95% CI: 0.49-1.21). Whether the case was called back at the last read was a highly significant predictor (odds ratio 4.22, 95% CI: 2.70-6.61) for the model incorporating reader variability, and also the model without reader variability (odds ratio 2.67, 95% CI: 1.74-4.08). CONCLUSION: The only statistically significant predictor of radiologist memory for a mammogram was whether the radiologist 'called it back' at a prior reading round. Presence of cancer on a mammogram did not make it memorable.

Inference about causation from examination of familial confounding: application to longitudinal twin data on mammographic density measures that predict breast cancer risk.

BACKGROUND: Mammographic density is a strong risk factor for breast cancer. It is unknown whether there are different causes of variation in mammographic density at different ages. METHODS: Mammograms and questionnaires were obtained on average 8 years apart from 327 Australian female twin pairs (204 monozygous and 123 dizygous). Mammographic dense area and percentage dense area were measured using a computer-assisted method. The correlational structure of the longitudinal twin data was estimated under a multivariate normal model using FISHER. Inference about causation from examination of familial confounding was made by regressing each twin's recent mammographic density measure against one or both of her and her co-twin's past measures. RESULTS: For square root dense area and percentage dense area (age- and body mass index-adjusted), the correlations over time within twins were 0.86 and 0.82, and the cross-twin correlations were 0.71 and 0.65 for monozygous pairs and 0.25 and 0.20 for dizygous pairs, respectively. As a predictor of a twin's recent dense area, the regression coefficient (SE) for the co-twin's past dense area reduced after adjusting for her own past measure from 0.84 (0.03) to 0.09 (0.03) for monozygous pairs and from 0.63 (0.04) to 0.04 (0.03) for dizygous pairs. Corresponding estimates for percentage dense area were 0.73 (0.04), 0.10 (0.03), 0.42 (0.05), and 0.03 (0.03). CONCLUSION: Mammographic density measures are highly correlated over time and the familial/genetic components of their variation are established before mid-life. IMPACT: Mammographic density of young women could provide a means for breast cancer control.

Are genetic and environmental components of variance in mammographic density measures that predict breast cancer risk independent of within-twin pair differences in body mass index?

To understand the role of mammographic density on breast cancer risk, it is important to take into account body mass index (BMI). As with age, BMI is negatively confounded with mammographic density, and a previous US twin study found that the covariance structure of mammographic density depended on the extent to which pairs differ in BMI. Using a computerised thresholding technique, we measured mammographic dense area for 571 monozygous (MZ) and 380 dizygous (DZ) twin pairs aged 40-70 years from Australia and North America. After adjusting for age and BMI, we calculated estimates of variance, covariance, correlation and, under the assumptions of the classic twin model, additive genetic (A), common environment (C) and person-specific environmental (E) components of variance. Analyses were conducted both within and across categories of within-pair differences in BMI, under a bivariate normal model using the software FISHER. The variance, MZ and DZ correlations, and the differences between MZ and DZ correlations and covariances were not constant across absolute within-pair differences in BMI (for the DZ correlation, P = 0.04, all other P < 0.001). No model involving a combination of one or more of A, C and E gave an acceptable fit. The interpretation of these observations is not straightforward. They, and other data, challenge the assumptions of the classic twin model for mammographic density and suggest that an insightful test of those assumptions can be made by testing the stability of correlations, covariances and variance components across absolute within-pair differences in potential mediators.

Image-guided sampling reveals increased stroma and lower glandular complexity in mammographically dense breast tissue.

Mammographic density (MD) adjusted for age and body mass index (BMI) is a strong heritable breast cancer risk factor; however, its biological basis remains elusive. Previous studies assessed MD-associated histology using random sampling approaches, despite evidence that high and low MD areas exist within a breast and are negatively correlated with respect to one another. We have used an image-guided approach to sample high and low MD tissues from within individual breasts to examine the relationship between histology and degree of MD. Image-guided sampling was performed using two different methodologies on mastectomy tissues (n = 12): (1) sampling of high and low MD regions within a slice guided by bright (high MD) and dark (low MD) areas in a slice X-ray film; (2) sampling of high and low MD regions within a whole breast using a stereotactically guided vacuum-assisted core biopsy technique. Pairwise analysis accounting for potential confounders (i.e. age, BMI, menopausal status, etc.) provides appropriate power for analysis despite the small sample size. High MD tissues had higher stromal (P = 0.002) and lower fat (P = 0.002) compositions, but no evidence of difference in glandular areas (P = 0.084) compared to low MD tissues from the same breast. High MD regions had higher relative gland counts (P = 0.023), and a preponderance of Type I lobules in high MD compared to low MD regions was observed in 58% of subjects (n = 7), but did not achieve significance. These findings clarify the histologic nature of high MD tissue and support hypotheses regarding the biophysical impact of dense connective tissue on mammary malignancy. They also provide important terms of reference for ongoing analyses of the underlying genetics of MD.

Variation in mammographic appearance between projections of small breast cancers compared with radial scars.

AIMS: The study aims to assess variation in appearance between mammographic projections (conspicuity variation) for invasive breast cancers (IBCs) compared with radial scars (RS). Conspicuity variation has been previously described as characteristic of RS. The lesions were also compared with respect to breast density and the proportion of cases detected by one of two readers and required a third (consensus) read. MATERIALS AND METHODS: The study was approved by the BreastScreen Victoria research committee. Mammograms of 75 randomly selected invasive breast cancers, with histological diameter ≤10 mm (IBC), were mixed with 67 consecutively detected RS, all from a double-reading population-based breast cancer screening programme. On blinded review, these 142 lesions were classified for mammographic findings and assessed for marked or minor conspicuity variation between views. We assessed the associations between lesion type, lesion spicules and centres, breast density, conspicuity variation and proportion detected by one reader only. RESULTS: Marked conspicuity variation was common, but not statistically different for IBC and RS (64% vs. 66%, χ(2) = 0.8, P = 0.04). Conspicuity variation did not correlate with spiculation type (long, fine or short, broad based) or lesion centres (lucent or dense) (ρ < 0.05, P = 0.5), and showed no significant change with increasing Breast Imaging Reporting and Data System breast density (IBC, χ(2) = 2.3, P = 0.5; RS, χ(2) = 0.95, P = 0.6). Density did not vary by lesion type. In the screening programme, 29% of IBC (125 of 431) versus 43% of RS (32 of 75) had been detected by one of two readers (χ(2) = 2.7, P = 0.098). CONCLUSIONS: Two-thirds of small IBCs displayed marked conspicuity variation, similar to RS. Therefore, conspicuity variation does not discriminate between IBC and RS.

Malignancy detection in digital mammograms: important reader characteristics and required case numbers.

RATIONALE AND OBJECTIVES: To determine the relationship between heightened levels of reader performance and reader practice in terms of number of cases read and previous experience. MATERIALS AND METHODS: A test set of mammograms was developed comprising 50 cases. These cases consisted of 15 abnormals (biopsy proven) and 35 normals (confirmed at subsequent rescreen). Sixty-nine breast image readers reviewed these cases independently and their performance was measured by recording their individual receiver operating characteristic score (area under the curve), sensitivity, and specificity. These measures of performance were then compared to a range of factors relating to the reader such as years of certification and reporting, number of cases read per year, previous experiences, and satisfaction levels. Correlation analyses using Spearman methods were performed along with the Mann-Whitney test to detect differences in performance between specific reader groups. RESULTS: Improved reader performance was found for years certified (P = .004), years of experience (P = .0001), and hours reading per week (P = .003) shown by positive statistical significant relationships with Az values (area under receiver operating characteristic curve). Statistical comparisons of Az values scored for individuals who read varying number of cases per year showed that those individuals whose annual mammographic case load was 5000 or more (P = .03) or between 2000 and 4999 (P = .05), had statistically significantly higher scores than those who read less than 1000 cases per year. CONCLUSION: The results of this study have shown variations in reader performance relating to parameters of reader practice and experience. Levels of variance are shown and potential acceptance levels for diagnostic efficacy are proposed which may inform policy makers, judicial systems and public debate.

Common genetic variants associated with breast cancer and mammographic density measures that predict disease.

Mammographic density for age and body mass index (BMI) is a heritable risk factor for breast cancer. We aimed to determine if recently identified common variants associated with small gradients in breast cancer risk are associated with mammographic density. We genotyped 497 monozygotic and 330 dizygotic twin pairs and 634 of their sisters from 903 families for 12 independent variants. Mammographic dense area, percent dense area, and nondense area were measured by three observers using a computer-thresholding technique. Associations with mammographic density measures adjusted for age, BMI, and other determinants were estimated (a) cross-sectionally using a multivariate normal model for pedigree analysis (P(x)), (b) between sibships, and (c) within sibships using orthogonal transformations of outcomes and exposures. A combined test of association (P(c)) was derived using the independent estimates from b and c. We tested if the distributions of P values across variants differed from the uniform distribution (P(u)). For dense area and percent dense area, the distributions of P(c) values were not uniform (both P(u) <0.007). Consistent with their breast cancer associations, rs3817198 (LSP1) and rs13281615 (8q) were associated with dense area and percent dense area (all P(x) and P(c) <0.05), and rs889312 (MAP3K1), rs2107425 (H19), and rs17468277 (CASP8) were marginally associated with dense area (some P(x) or P(c) <0.05). All associations were independent of menopausal status. At least two common breast cancer susceptibility variants are associated with mammographic density measures that predict breast cancer. These findings could help elucidate how those variants and mammographic density measures are associated with breast cancer susceptibility.

Predictors of mammographic density: insights gained from a novel regression analysis of a twin study.

Understanding which factors influence mammographically dense and nondense areas is important because percent mammographic density adjusted for age is a strong, continuously distributed risk factor for breast cancer, especially when adjusted for weight or body mass index. Using computer-assisted methods, we measured mammographically dense areas for 571 monozygotic and 380 dizygotic Australian and North American twin pairs ages 40 to 70 years. We used a novel regression modeling approach in which each twin's measure of dense and nondense area was regressed against one or both of the twin's and co-twin's covariates. The nature of changes to regression estimates with the inclusion of the twin and/or co-twin's covariates can be evaluated for consistency with causal and/or other models. By causal, we mean that if it were possible to vary a covariate experimentally then the expected value of the outcome measure would change. After adjusting for the individual's weight, the co-twin associations with weight were attenuated, consistent with a causal effect of weight on mammographic measures, which in absolute log cm(2)/kg was thrice stronger for nondense than dense area. After adjusting for weight, later age at menarche, and greater height were associated with greater dense and lesser nondense areas in a manner inconsistent with causality. The associations of dense and nondense areas with parity are consistent with a causal effect and/or within-person confounding. The associations between mammographic density measures and height are consistent with shared early life environmental factors that predispose to both height and percent mammographic density and possibly breast cancer risk.

Using mammographic density to improve breast cancer screening outcomes.

It is possible that the performance of mammographic screening would be improved if it is targeted at women at higher risk of breast cancer or who are more likely to have their cancer missed at screening, through more intensive screening or alternative screening modalities. We conducted a case-control study within a population-based Australian mammographic screening program (1,706 invasive breast cancers and 5,637 randomly selected controls). We used logistic regression to examine the effects of breast density, age, and hormone therapy use, all known to influence both breast cancer risk and the sensitivity of mammographic screening, on the risk of small (15 mm) screen-detected and interval breast cancers. The risk of small screen-detected cancers was not associated with density, but the risk of large screen-detected cancers was nearly 3-fold for the second quintile and approximately 4-fold for the four highest density categories (third and fourth quintiles and the two highest deciles) compared with the lowest quintile. The risk of interval cancers increased monotonically across the density categories [highest decile odds ratio (OR), 4.65; 95% confidence interval (95% CI), 2.96-7.31]. The risk of small and large screen-detected cancers, but not interval cancers, increased with age. After adjusting for age and density, hormone therapy use was associated with a moderately elevated risk of interval cancers (OR, 1.43; 95% CI, 1.12-1.81). The effectiveness of the screening program could be improved if density were to be used to identify women most likely to have poor screening outcomes. There would be little additional benefit in targeting screening based on age and hormone therapy use.

Imaging surveillance of the breast in a patient diagnosed with scleroderma after breast-conserving surgery and radiotherapy.

Collagen vascular disease (CVD), particularly scleroderma, is a contraindication to radiation therapy because of increased risk of fibrosis. We report a patient with early stage breast cancer diagnosed with scleroderma after breast-conserving surgery and radiation. She developed marked breast fibrosis, rendering mammographic, sonographic, and clinical surveillance ineffective. She has subsequently been followed with magnetic resonance imaging (MRI) of the breast. We illustrate this case and review the literature relating to CVD and radiation therapy. MRI may be a suitable surveillance method in this situation.

Use of breast cancer screening and treatment services by Australian women aged 25-44 years following Kylie Minogue's breast cancer diagnosis.

OBJECTIVE: To examine the effects of the publicity surrounding Kylie Minogue's diagnosis with breast cancer on doctor-referred breast imaging, image-guided biopsy, and cancer excisions among a low-risk population of women in Australia. Method We examine changes in unilateral and bilateral breast imaging, image-guided breast biopsies, and surgical excisions of breast cancer before and after the announcement of Kylie Minogue's diagnosis with breast cancer in May 2005. The study included procedures provided through the Australian public health system to women aged 25-44 years from October 2004 and June 2006. RESULTS: The odds of women aged 25-44 years undergoing imaging procedures increased by 20% in the first and second quarters after the Minogue publicity, compared to the preceding two quarters. The volume of biopsies als increased but the biopsy rate, measured as a proportion of imaging procedures, did not change among women aged 25-34 years and decreased among women aged 35-44 years. The volume of operations to excise breast cancers did not change for either age group. Compared to the 6 month period before the publicity, there was a large and significant decrease in the odds that an excision would follow biopsy (25-34 years: OR 95% CI=0.69, 0.48-0.98; 35-44 years: OR 95% CI=0.83, 0.72-0.95). CONCLUSIONS: High-publicised illnesses may affect both consumer and provider behaviour. Although they present opportunities to improve public health, they also have the potential to adversely impact the appropriateness and cost-effectiveness of service delivery.

Can mammographic findings help discriminate between atypical ductal hyperplasia and ductal carcinoma in situ after needle core biopsy?

In a screening population of women, the mammographic characteristics for 68 cases of atypical ductal hyperplasia (ADH) diagnosed by needle core biopsy (NCB) were reviewed to seek mammographic findings which differentiate between ductal carcinoma in situ (DCIS) and ADH. A blinded analysis by two radiologists was performed for 48 cases with microcalcification. The mammographic findings were correlated with the surgical histological results of benign non-atypical, ADH and carcinoma (DCIS or invasive) to identify features which were associated with a higher or lower odds ratio (OR) for malignancy. Underestimates for malignancy occurred in 14 of 29 cases with granular calcification form (OR 7.9, 95% confidence interval (CI) 1.5-41) and 6 of 8 cases with segmental/linear branching distribution (OR 9.0, 95%CI 1.6-52). No malignancy was found at surgical excision in 16 cases with fine, rounded calcification. In conclusion, detailed assessment of calcification distribution and form gave helpful predictors for malignancy. Lesions with fine rounded calcification were always benign.

Mammographic densities during the menopausal transition: a longitudinal study of Australian-born women.

OBJECTIVE: The objective of this study was to investigate hormonal and other factors associated with mammographic density during the menopausal transition and in postmenopause. DESIGN: Mammograms were obtained from 252 participants in the Melbourne Women's Midlife Health Project-a longitudinal population-based study that included annual interviews, blood collection, and physical measurements; 869 original films of the right craniocaudal view were digitized. Total area of the breast and the area of dense tissue were measured, and the percentage of mammographically dense tissue (PMD) was calculated. Data were analyzed using time-series regression models. RESULTS: Of the 252 women, 87% had more than one mammogram, and the mean age was 56.0 (SD 3.6) years (range 45-67); 129 women who had never used hormone therapy were included in the analyses. The mean nondense breast tissue area increased through the menopausal transition (P for trend=0.01), there was no significant trend in the mean dense breast tissue area, and mean PMD decreased (P for trend=0.004). Multivariate analysis showed that increasing age (P<0.005) and body mass index (BMI) (P<0.05), having had children (P<0.05), and higher than average free testosterone levels (P<0.05) (or lower than average sex hormone-binding globulin levels) were associated with increased area of nondense tissue. Increasing age (P<0.05) and BMI (P<0.05) were associated with decreased PMD. There was a tendency for higher than average free testosterone levels (P<0.07) and having had children (P=0.07) to be associated with lower PMD. After controlling for age, there were no significant associations with the area of dense tissue. CONCLUSIONS: This longitudinal observational study has shown that after controlling for age, there was no apparent effect of menopausal change on the area of dense breast tissue. Aging and increasing BMI through the menopausal transition were associated with increased nondense breast tissue and explain a small, but statistically significant, portion of the variation in PMD tissue.

Margins and outcome of screen-detected breast cancer with extensive in situ component.

BACKGROUND: In situ disease surrounding invasive tumours is an important consideration in the management of patients with early breast cancer. This study of screen-detected breast cancers assessed the influence of in situ disease including an extensive in situ component (defined as ductal carcinoma in situ involving more than 25% of the area within the invasive tumour) on surgical management, local recurrence and survival of a group of patients. METHODS: A total of 595 cases of invasive breast cancer detected at St Vincent's BreastScreen were retrospectively reviewed to determine presence and extent of in situ disease, the surgical procedure and adequacy of excision. Outcome was examined in a cohort of 126 cases. RESULTS: A total of 438 (74%) patients had in situ foci in or around the invasive tumour and 107 (18%) were defined as extensive in situ component (EIC)-positive. The initial procedure was mastectomy in 20% of the cases and breast-conserving surgery in 80% including 18% who underwent further surgery. Re-excision (P = 0.02) or mastectomy (P = 0.01) was more often required in patients with EIC. After definitive local excision, margins were close or involved with invasive disease in 3% but the patients with EIC were more likely to have margins close or involved with in situ disease (16 vs 2%; P = 0.001). There were seven deaths and one local invasive recurrence in the follow-up group and none of the deaths were in patients who were EIC-positive. CONCLUSIONS: EIC predicts for a higher rate of re-excision and/or mastectomy. For patients with EIC, there is an acceptably low risk of local recurrence if margins are clear.