Judy Caines

A Review of Interval Breast Cancers Diagnosed among Participants of the Nova Scotia Breast Screening Program

PURPOSE To conduct a radiologic review of interval breast cancer cases to determine rates of true interval and missed cancers in Nova Scotia, Canada. MATERIALS AND METHODS This quality assurance project was exempt from institutional review board approval. Interval cancer cases were identified among women aged 40-69 years who were participants in the Nova Scotia Breast Screening Program from 1991 to 2004. For each case, the index negative screening mammogram was reviewed blindly by three radiologists from a pool of experienced radiologists. Cases were identified as those with normal or abnormal findings, the latter being a case that required further investigation. True interval cases were identified as cases in which a minimum of two radiologists reviewed the findings as normal. True interval and missed cancer rates were calculated separately for women according to age group and screening interval (for ages 40-49 years, a 1-year interval; for ages 50-69 years, a 1-year and a 2-year interval). RESULTS The rate of missed cancers per 1000 women screened was one-half of the true interval rate among women screened annually (for ages 40-49 years, 0.45 vs 0.93; for ages 50-69 years, 1.08 vs 2.22). Among women aged 50-69 years who were screened biennially, the rate of missed cancers per 1000 women screened was one-third of the true interval rate (0.90 vs 3.15). Similarly, the rate of missed cancers per 10,000 screening examinations was one-half of the true interval rate among those 40-49 years old (1.95 vs 3.99) and one-third of the true interval rate among those 50-69 years old (3.34 vs 10.44). CONCLUSION In screening programs, true interval cancer rates should be differentiated from missed cancer rates as part of ongoing quality assurance.

The Burden of False-Positive Results in Analog and Digital Screening Mammography: Experience of the Nova Scotia Breast Screening Program

Purpose The Canadian Task Force on Preventive Health Care released recommendations for breast cancer screening, in part, based on harms associated with screening. The purpose of this study was to describe the rate of false-positive (FP) screening mammograms and to describe the extent of the investigations after an FP. Methods A cohort was identified that consisted of all screening mammograms performed through the Screening Program (2000-2011) with patients ages 40-69 years at screening. Rates of FP screening mammograms were calculated as well as rates of further investigations required, including additional imaging, needle core biopsy, and surgery. Analyses were stratified by 10-year age group, screening status (first vs rescreen), and technology. Results A total of 608,088 screening mammograms were included. The FP rate varied by age group, and decreased with increasing age (digital, 40-49 years old, FP = 8.0%; 50-59 years old, FP = 6.3%; 60-69 years old, FP = 4.6%). The FP rate also varied by screening status (digital, first screen, FP = 12.0%; rescreen, FP = 5.6%), and this difference was consistent across age groups. The need for further investigation varied by age group, with invasive procedures being more heavily used as women age (digital, rescreen group, surgery: 40-49 years old, 1.1%; 50-59 years old 1.6%, 60-69 years old, 1.8%). Conclusions Both the FP screening mammogram rate and the degree to which further investigation was required varied by age group and screening status. Reporting on these rates should form part of the evaluation of screening performance.

Utility of relative and absolute measures of mammographic density vs clinical risk factors in evaluating breast cancer risk at time of screening mammography

OBJECTIVE Various clinical risk factors, including high breast density, have been shown to be associated with breast cancer. The utility of using relative and absolute area-based breast density-related measures was evaluated as an alternative to clinical risk factors in cancer risk assessment at the time of screening mammography. METHODS Contralateral mediolateral oblique digital mammography images from 392 females with unilateral breast cancer and 817 age-matched controls were analysed. Information on clinical risk factors was obtained from the provincial breast-imaging information system. Breast density-related measures were assessed using a fully automated breast density measurement software. Multivariable logistic regression was conducted, and area under the receiver-operating characteristic (AUROC) curve was used to evaluate the performance of three cancer risk models: the first using only clinical risk factors, the second using only density-related measures and the third using both clinical risk factors and density-related measures. RESULTS The risk factor-based model generated an AUROC of 0.535, while the model including only breast density-related measures generated a significantly higher AUROC of 0.622 (p < 0.001). The third combined model generated an AUROC of 0.632 and performed significantly better than the risk factor model (p < 0.001) but not the density-related measures model (p = 0.097). CONCLUSION Density-related measures from screening mammograms at the time of screen may be superior predictors of cancer compared with clinical risk factors. ADVANCES IN KNOWLEDGE Breast cancer risk models based on density-related measures alone can outperform risk models based on clinical factors. Such models may support the development of personalized breast-screening protocols.

Breast density scales: the metric matters

OBJECTIVE Measures of percent mammographic density (PMD) are often categorized using various density scales. The purpose of this study was to examine information loss associated with the use of categorical density scales. METHODS Baseline PMD was assessed at 1% precision for 2,374 females. The data were used to create 21-category, 4-category and 2-category density scales. R-squared and root mean square error were used to evaluate the effect of categorizing PMD. The area under the receiver operator characteristic curves were compared between cancer risk models employing solely categorical PMD scales and solely baseline PMD for a subset of females (424 cases, 848 controls). RESULTS R-squared value decreased from 1.00 (1% PMD) to 0.56 (2-category scale), while root mean square error increased from 0.00 (1% PMD) to 10.83 (2-category scale). The area under the receiver operator characteristic curve decreased from 0.64 for a cancer risk model using 1% PMD to 0.58 for a risk model using a 21-category density scale (p < 0.0001), 0.55 for a 4-category Breast Imaging, Reporting and Data System-like scale (p < 0.0001) and 0.50 for a 2-category Breast Imaging, Reporting and Data System-like scale (high vs low) (p < 0.0001). CONCLUSION Categorizing PMD measures into categorical density scales leads to a significant loss of information. Indeed, a simple high versus low split of PMD using a 50% cut point yields a cancer risk model with no discriminatory power. Advances in knowledge: Use of categorical mammographic density scales rather than continuous percent mammographic density measures leads to significant loss of information. Breast cancer risk models using categorical mammographic density scales perform more poorly than models using continuous PMD measures.

Consistency of visual assessments of mammographic breast density from vendor-specific “for presentation” images

Abstract. Discussions of percent breast density (PD) and breast cancer risk implicitly assume that visual assessments of PD are comparable between vendors despite differences in technology and display algorithms. This study examines the extent to which visual assessments of PD differ between mammograms acquired from two vendors. Pairs of “for presentation” digital mammography images were obtained from two mammography units for 146 women who had a screening mammogram on one vendor unit followed by a diagnostic mammogram on a different vendor unit. Four radiologists independently visually assessed PD from single left mediolateral oblique view images from the two vendors. Analysis of variance, intra-class correlation coefficients (ICC), scatter plots, and Bland–Altman plots were used to evaluate PD assessments between vendors. The mean radiologist PD for each image was used as a consensus PD measure. Overall agreement of the PD assessments was excellent between the two vendors with an ICC of 0.95 (95% confidence interval: 0.93 to 0.97). Bland–Altman plots demonstrated narrow upper and lower limits of agreement between the vendors with only a small bias (2.3 percentage points). The results of this study support the assumption that visual assessment of PD is consistent across mammography vendors despite vendor-specific appearances of “for presentation” images.