Screening outcomes by risk factor and age: evidence from BreastScreen WA for discussions of risk-stratified population screening.

OBJECTIVES: To estimate rates of screen-detected and interval breast cancers, stratified by risk factor, to inform discussions of risk-stratified population screening. DESIGN: Retrospective population-based cohort study; analysis of routinely collected BreastScreen WA program clinical and administrative data. SETTING, PARTICIPANTS: All BreastScreen WA mammography screening episodes for women aged 40 years or more during 1 July 2007 - 30 June 2017. MAIN OUTCOME MEASURES: Cancer detection rate (CDR) and interval cancer rate (ICR), by risk factor. RESULTS: A total of 323 082 women were screened in 1 026 137 screening episodes (mean age, 58.5 years; SD, 8.6 years). The overall CDR was 68 (95% CI, 67-70) cancers per 10 000 screens, and the overall ICR was 9.7 (95% CI, 9.2-10.1) cancers per 10 000 women-years. Interactions between the effects on CDR of age group and five risk factors were statistically significant: personal history of breast cancer (P = 0.039), family history of breast cancer (P = 0.005), risk-relevant benign conditions (P = 0.012), hormone-replacement therapy (P = 0.002), and self-reported symptoms (P < 0.001). The influence of these risk factors (except personal history) increased with age. For ICR, only the interaction between age and hormone-replacement therapy was significant (P < 0.001), although weak interactions between age and family history of breast cancer or having dense breasts were noted (each P = 0.07). The influence of family history on ICR was significant only for women aged 40-49 years. CONCLUSIONS: Screening CDR and (for some risk factors) ICR were higher for women in some age groups with personal histories of breast cancer or risk-relevant benign breast conditions or first degree family history of breast cancer, women with dense breasts or self-reported breast-related symptoms, and women using hormone-replacement therapy. Our findings could inform the evaluation of risk-based screening.

Agreement between MRI and pathologic breast tumor size after neoadjuvant chemotherapy, and comparison with alternative tests: individual patient data meta-analysis.

BACKGROUND: Magnetic resonance imaging (MRI) may guide breast cancer surgery by measuring residual tumor size post-neoadjuvant chemotherapy (NAC). Accurate measurement may avoid overly radical surgery or reduce the need for repeat surgery. This individual patient data (IPD) meta-analysis examines MRI's agreement with pathology in measuring the longest tumor diameter and compares MRI with alternative tests. METHODS: A systematic review of MEDLINE, EMBASE, PREMEDLINE, Database of Abstracts of Reviews of Effects, Heath Technology Assessment, and Cochrane databases identified eligible studies. Primary study authors supplied IPD in a template format constructed a priori. Mean differences (MDs) between tests and pathology (i.e. systematic bias) were calculated and pooled by the inverse variance method; limits of agreement (LOA) were estimated. Test measurements of 0.0 cm in the presence of pathologic residual tumor, and measurements >0.0 cm despite pathologic complete response (pCR) were described for MRI and alternative tests. RESULTS: Eight studies contributed IPD (N = 300). The pooled MD for MRI was 0.0 cm (LOA: +/-3.8 cm). Ultrasound underestimated pathologic size (MD: -0.3 cm) relative to MRI (MD: 0.1 cm), with comparable LOA. MDs were similar for MRI (0.1 cm) and mammography (0.0 cm), with wider LOA for mammography. Clinical examination underestimated size (MD: -0.8 cm) relative to MRI (MD: 0.0 cm), with wider LOA. Tumors "missed" by MRI typically measured 2.0 cm or less at pathology; tumors >2.0 cm were more commonly "missed" by clinical examination (9.3 %). MRI measurements >5.0 cm occurred in 5.3 % of patients with pCR, but were more frequent for mammography (46.2 %). CONCLUSIONS: There was no systematic bias in MRI tumor measurement, but LOA are large enough to be clinically important. MRI's performance was generally superior to ultrasound, mammography, and clinical examination, and it may be considered the most appropriate test in this setting. Test combinations should be explored in future studies.

Systematic review and meta-analysis of prognostic characteristics for breast cancers in populations with digital vs film mammography indicate the transition may have increased both early detection and overdiagnosis.

OBJECTIVES: Film mammography has been replaced by digital mammography in breast screening programs globally. This led to a small increase in the rate of detection, but whether the detection of clinically important cancers increased is uncertain. We aimed to assess the impact on tumor characteristics of screen-detected and interval breast cancers. STUDY DESIGN AND SETTING: We searched seven databases from inception to October 08, 2023, for publications comparing film and digital mammography within the same population of asymptomatic women at population (average) risk of breast cancer. We recorded reported tumor characteristics and assessed risk of bias using the Risk Of Bias In Non-randomised Studies - of Interventions tool. We synthesized results using meta-analyses of random effects. RESULTS: Eighteen studies were included in the analysis from 8 countries, including 11,592,225 screening examinations (8,117,781 film; 3,474,444 digital). There were no differences in tumor size, morphology, grade, node status, receptor status, or stage in the pooled differences for screen-detected and interval invasive cancer tumor characteristics. There were statistically significant increases in screen-detected ductal carcinoma in situ (DCIS) across all grades: 0.05 (0.00-0.11), 0.14 (0.05-0.22), and 0.19 (0.05-0.33) per 1000 screens for low, intermediate, and high-grade DCIS, respectively. There were similar (non-statistically significant) increases in screen-detected invasive cancer across all grades. CONCLUSION: The increased detection of all grades of DCIS and invasive cancer may indicate both increased early detection of more aggressive disease and increased overdiagnosis.

Impact of Full-Field Digital Mammography Versus Film-Screen Mammography in Population Screening: A Meta-Analysis.

BACKGROUND: Breast screening programs replaced film mammography with digital mammography, and the effects of this practice shift in population screening on health outcomes can be measured through examination of cancer detection and interval cancer rates. METHODS: A systematic review and random effects meta-analysis were undertaken. Seven databases were searched for publications that compared film with digital mammography within the same population of asymptomatic women and reported cancer detection and/or interval cancer rates. RESULTS: The analysis included 24 studies with 16 583 743 screening examinations (10 968 843 film and 5 614 900 digital). The pooled difference in the cancer detection rate showed an increase of 0.51 per 1000 screens (95% confidence interval [CI] = 0.19 to 0.83), greater relative increase for ductal carcinoma in situ (25.2%, 95% CI = 17.4% to 33.5%) than invasive (4%, 95% CI = -3% to 13%), and a recall rate increase of 6.95 (95% CI = 3.47 to 10.42) per 1000 screens after the transition from film to digital mammography. Seven studies (80.8% of screens) reported interval cancers: the pooled difference showed no change in the interval cancer rate with -0.02 per 1000 screens (95% CI = -0.06 to 0.03). Restricting analysis to studies at low risk of bias resulted in findings consistent with the overall pooled results for all outcomes. CONCLUSIONS: The increase in cancer detection following the practice shift to digital mammography did not translate into a reduction in the interval cancer rate. Recall rates were increased. These results suggest the transition from film to digital mammography did not result in health benefits for screened women. This analysis reinforces the need to carefully evaluate effects of future changes in technology, such as tomosynthesis, to ensure new technology leads to improved health outcomes and beyond technical gains.

Stillbirth risk prediction using machine learning for a large cohort of births from Western Australia, 1980-2015.

Quantification of stillbirth risk has potential to support clinical decision-making. Studies that have attempted to quantify stillbirth risk have been hampered by small event rates, a limited range of predictors that typically exclude obstetric history, lack of validation, and restriction to a single classifier (logistic regression). Consequently, predictive performance remains low, and risk quantification has not been adopted into antenatal practice. The study population consisted of all births to women in Western Australia from 1980 to 2015, excluding terminations. After all exclusions there were 947,025 livebirths and 5,788 stillbirths. Predictive models for stillbirth were developed using multiple machine learning classifiers: regularised logistic regression, decision trees based on classification and regression trees, random forest, extreme gradient boosting (XGBoost), and a multilayer perceptron neural network. We applied 10-fold cross-validation using independent data not used to develop the models. Predictors included maternal socio-demographic characteristics, chronic medical conditions, obstetric complications and family history in both the current and previous pregnancy. In this cohort, 66% of stillbirths were observed for multiparous women. The best performing classifier (XGBoost) predicted 45% (95% CI: 43%, 46%) of stillbirths for all women and 45% (95% CI: 43%, 47%) of stillbirths after the inclusion of previous pregnancy history. Almost half of stillbirths could be potentially identified antenatally based on a combination of current pregnancy complications, congenital anomalies, maternal characteristics, and medical history. Greatest sensitivity is achieved with addition of current pregnancy complications. Ensemble classifiers offered marginal improvement for prediction compared to logistic regression.

Meta-analysis of magnetic resonance imaging in detecting residual breast cancer after neoadjuvant therapy.

BACKGROUND: It has been proposed that magnetic resonance imaging (MRI) be used to guide breast cancer surgery by differentiating residual tumor from pathologic complete response (pCR) after neoadjuvant chemotherapy. This meta-analysis examines MRI accuracy in detecting residual tumor, investigates variables potentially affecting MRI performance, and compares MRI with other tests. METHODS: A systematic literature search was undertaken. Hierarchical summary receiver operating characteristic (HSROC) models were used to estimate (relative) diagnostic odds ratios ([R]DORs). Summary sensitivity (correct identification of residual tumor), specificity (correct identification of pCR), and areas under the SROC curves (AUCs) were derived. All statistical tests were two-sided. RESULTS: Forty-four studies (2050 patients) were included. The overall AUC of MRI was 0.88. Accuracy was lower for "standard" pCR definitions (referent category) than "less clearly described" (RDOR = 2.41, 95% confidence interval [CI] = 1.11 to 5.23) or "near-pCR" definitions (RDOR = 2.60, 95% CI = 0.73 to 9.24; P = .03.) Corresponding AUCs were 0.83, 0.90, and 0.91. Specificity was higher when negative MRI was defined as contrast enhancement less than or equal to normal tissue (0.83, 95% CI = 0.64 to 0.93) vs no enhancement (0.54, 95% CI = 0.39 to 0.69; P = .02), with comparable sensitivity (0.83, 95% CI = 0.69 to 0.91; vs 0.87, 95% CI = 0.80 to 0.92; P = .45). MRI had higher accuracy than mammography (P = .02); there was only weak evidence that MRI had higher accuracy than clinical examination (P = .10). No difference in MRI and ultrasound accuracy was found (P = .15). CONCLUSIONS: MRI accurately detects residual tumor after neoadjuvant chemotherapy. Accuracy was lower when pCR was more rigorously defined, and specificity was lower when test negativity thresholds were more stringent; these definitions require standardization. MRI is more accurate than mammography; however, studies comparing MRI and ultrasound are required.

Meta-analysis of the association of breast cancer subtype and pathologic complete response to neoadjuvant chemotherapy.

BACKGROUND: Pathologic complete response (pCR) is a surrogate end-point for prognosis in neoadjuvant chemotherapy (NAC) for breast cancer. We aimed to report summary estimates of the proportion of subjects achieving pCR (pCR%) by tumour subtype, and to determine whether subtype was independently associated with pCR, in a study-level meta-analysis. METHODS: We systematically identified NAC studies reporting pCR data according to tumour subtype, using predefined eligibility criteria. Descriptive, qualitative and quantitative data were extracted. Random effects logistic meta-regression examined whether pCR% was associated with subtype, defined using three categories for model 1 [hormone receptor positive (HR+/HER2-), HER2 positive (HER2+), triple negative (ER-/PR-/HER2-)] and 4 categories for model 2 [HER2+ further classified as HER2+/HR+ and HER2+/HR-]. Subtype-specific odds ratios (OR) were calculated and were adjusted for covariates associated with pCR in our data. RESULTS: In model 1, based on 11,695 subjects from 30 eligible studies, overall pooled pCR% was 18.9% (16.6-21.5%), and in model 2 (20 studies, 8095 subjects) pooled pCR% was 18.5% (16.2-21.1%); tumour subtype was associated with pCR% (P<0.0001) in both models. Subtype-specific pCR% (model 2) was: 8.3% (6.7-10.2%) in HR+/HER2- [OR 1/referent], 18.7% (15.0-23.1%) in HER2+/HR+ [OR 2.6], 38.9% (33.2-44.9%) in HER2+/HR- [OR 7.1] and 31.1% (26.5-36.1%) in triple negative [OR 5.0]; pCR% was significantly higher for the HER2+/HR- compared with the triple negative subtype, however pCR% was very similar for these subtypes (and OR=5.0 both subtypes) when studies using HER2-directed therapy with NAC were excluded from the model. Neither sensitivity analysis (excluding unknown subtypes), nor adjustment for associated covariates, substantially altered our findings. INTERPRETATION: This meta-analysis provides evidence of an independent association between breast cancer subtype and pCR; odds of pCR were highest for the triple negative and HER2+/HR- subtypes, with evidence of an influential effect on achieving pCR in the latter subtype through inclusion of HER2-directed therapy with NAC.