Diffusion-weighted imaging of suspicious (BI-RADS 4) breast lesions: stratification based on histopathology.

OBJECTIVE: To test the use of diffusion-weighted imaging (DWI) in stratifying suspicious breast lesions (BI-RADS 4), correlating them with histopathology. We also investigated the performance of DWI related to the main enhancement patterns (mass and non-mass) and tested its reproducibility. MATERIALS AND METHODS: Seventy-six patients presented 92 lesions during the sampling period. Two independent examiners reviewed magnetic resonance imaging studies, described the lesions, and determined the apparent diffusion coefficient (ADC) values. Differences among benign, indeterminate- to high-risk, and malignant findings, in terms of the ADCs, were assessed by analysis of variance. Using receiver operating characteristic (ROC) curves, we compared the performance of ADC values in masses and non-mass lesions, and tested the reproducibility of measurements by determining the coefficient of variation and smallest real difference. RESULTS: Among the 92 lesions evaluated, the histopathology showed that 37 were benign, 11 were indeterminate- to high-risk, and 44 were malignant. The mean ADC differed significantly among those histopathological groups, the value obtained for the malignant lesions (1.10 × 10-3 mm2/s) being significantly lower than that obtained for the other groups (p < 0.001). ROC curves demonstrated that DWI performed better when applied to masses than when applied to non-mass lesions (area under the curve, 0.88 vs. 0.67). Reproducibility was good (coefficient of variation, 7.03%; and smallest real difference, ± 0.242 × 10-3 mm2/s). CONCLUSION: DWI can differentiate between malignant and nonmalignant (benign or indeterminate- to high-risk) lesions, showing better performance for masses. Nevertheless, stratification based on histopathological criteria that are more refined has yet to be achieved.

OBJETIVO: Testar a sequência ponderada em difusão (SPD) para estratificação de lesões suspeitas (BI-RADS 4) por ressonância magnética em correlação com a histopatologia. Também investigamos o desempenho da SPD relacionada a padrões de realce (nódulo e não nódulo) e testamos sua reprodutibilidade. MATERIAIS E MÉTODOS: Setenta e seis pacientes apresentaram 92 lesões durante o período amostral. Dois examinadores independentes revisaram os estudos, descreveram as lesões e mediram os coeficientes de difusão aparente (CDAs). Diferenças de CDA entre achados benignos, indeterminados/de alto risco e malignos foram avaliadas por análise de variância. Comparamos o desempenho dos CDAs em nódulos e não nódulos por curvas receiver operating characteristic (ROC) e testamos a reprodutibilidade das mensurações pelo coeficiente de variação e menor diferença real. RESULTADOS: Obtivemos 37 lesões benignas, 11 indeterminadas/de alto-risco e 44 cânceres. As médias dos CDAs desses grupos histopatológicos foram significativamente diferentes (p < 0,001), devido aos valores mais baixos em achados malignos (1,10 × 10-3 mm2/s). Curvas ROC demonstraram melhor desempenho da SPD aplicada a nódulos (área sob a curva de 0,88 contra 0,67 para não nódulos). A reprodutibilidade foi boa (coeficiente de variação de 7,03% e menor diferença real de ± 0,242 × 10-3 mm2/s). CONCLUSÃO: A SPD pode diferenciar achados malignos de não malignos, com melhor desempenho para nódulos. Entretanto, a estratificação baseada em critérios histopatológicos mais refinados ainda não foi alcançada.

Diffusion-weighted imaging of suspicious (BI-RADS 4) breast lesions: stratification based on histopathology / Estudo de lesões mamárias suspeitas (BI-RADS 4) por sequência ponderada em difusão: estratificação baseada na histopatologia

Abstract Objective: To test the use of diffusion-weighted imaging (DWI) in stratifying suspicious breast lesions (BI-RADS 4), correlating them with histopathology. We also investigated the performance of DWI related to the main enhancement patterns (mass and non-mass) and tested its reproducibility. Materials and Methods: Seventy-six patients presented 92 lesions during the sampling period. Two independent examiners reviewed magnetic resonance imaging studies, described the lesions, and determined the apparent diffusion coefficient (ADC) values. Differences among benign, indeterminate- to high-risk, and malignant findings, in terms of the ADCs, were assessed by analysis of variance. Using receiver operating characteristic (ROC) curves, we compared the performance of ADC values in masses and non-mass lesions, and tested the reproducibility of measurements by determining the coefficient of variation and smallest real difference. Results: Among the 92 lesions evaluated, the histopathology showed that 37 were benign, 11 were indeterminate- to high-risk, and 44 were malignant. The mean ADC differed significantly among those histopathological groups, the value obtained for the malignant lesions (1.10 × 10-3 mm2/s) being significantly lower than that obtained for the other groups (p < 0.001). ROC curves demonstrated that DWI performed better when applied to masses than when applied to non-mass lesions (area under the curve, 0.88 vs. 0.67). Reproducibility was good (coefficient of variation, 7.03%; and smallest real difference, ± 0.242 × 10-3 mm2/s). Conclusion: DWI can differentiate between malignant and nonmalignant (benign or indeterminate- to high-risk) lesions, showing better performance for masses. Nevertheless, stratification based on histopathological criteria that are more refined has yet to be achieved.

Resumo Objetivo: Testar a sequência ponderada em difusão (SPD) para estratificação de lesões suspeitas (BI-RADS 4) por ressonância magnética em correlação com a histopatologia. Também investigamos o desempenho da SPD relacionada a padrões de realce (nódulo e não nódulo) e testamos sua reprodutibilidade. Materiais e Métodos: Setenta e seis pacientes apresentaram 92 lesões durante o período amostral. Dois examinadores independentes revisaram os estudos, descreveram as lesões e mediram os coeficientes de difusão aparente (CDAs). Diferenças de CDA entre achados benignos, indeterminados/de alto risco e malignos foram avaliadas por análise de variância. Comparamos o desempenho dos CDAs em nódulos e não nódulos por curvas receiver operating characteristic (ROC) e testamos a reprodutibilidade das mensurações pelo coeficiente de variação e menor diferença real. Resultados: Obtivemos 37 lesões benignas, 11 indeterminadas/de alto-risco e 44 cânceres. As médias dos CDAs desses grupos histopatológicos foram significativamente diferentes (p < 0,001), devido aos valores mais baixos em achados malignos (1,10 × 10-3 mm2/s). Curvas ROC demonstraram melhor desempenho da SPD aplicada a nódulos (área sob a curva de 0,88 contra 0,67 para não nódulos). A reprodutibilidade foi boa (coeficiente de variação de 7,03% e menor diferença real de ± 0,242 × 10-3 mm2/s). Conclusão: A SPD pode diferenciar achados malignos de não malignos, com melhor desempenho para nódulos. Entretanto, a estratificação baseada em critérios histopatológicos mais refinados ainda não foi alcançada.

Predictive performance of BI-RADS magnetic resonance imaging descriptors in the context of suspicious (category 4) findings.

OBJECTIVE: To determine the positive predictive value (PPV) and likelihood ratio for magnetic resonance imaging (MRI) characteristics of category 4 lesions, as described in the Breast Imaging Reporting and Data System (BI-RADS(®)) lexicon, as well as to test the predictive performance of the descriptors using multivariate analysis and the area under the curve derived from a receiver operating characteristic (ROC) curve. MATERIALS AND METHODS: This was a double-blind review study of 121 suspicious findings from 98 women examined between 2009 and 2013. The terminology was based on the 2013 edition of the BI-RADS. RESULTS: Of the 121 suspicious findings, 53 (43.8%) were proven to be malignant lesions, with no significant difference between mass and non-mass enhancement (p = 0.846). The PPVs were highest for masses with a spiculated margin (71%) and round shape (63%), whereas segmental distribution achieved a high PPV (80%) for non-mass enhancement. Kinetic analyses performed poorly, except for type 3 curves applied to masses (PPV of 73%). Logistic regression models were significant for both patterns, although the results were better for masses, particularly when kinetic assessments were included (p = 0.015; pseudo R(2) = 0.48; area under the curve = 90%). CONCLUSION: Some BI-RADS MRI descriptors have high PPV and good predictive performance-as demonstrated by ROC curve and multivariate analysis-when applied to BI-RADS category 4 findings. This may allow future stratification of this category.

OBJETIVO: Determinar o valor preditivo positivo (VPP) e a razão de verossimilhança positiva de características de ressonância magnética (RM) de lesões da categoria 4, como descritas no léxico do Breast Imaging Reporting and Data System (BI-RADS®), e testar o desempenho preditivo dos descritores por meio de análise multivariada e área sob a curva derivada da curva receiver operating characteristic (ROC). MATERIAIS E MÉTODOS: Foi realizado um estudo revisional duplo-cego de 121 achados suspeitos em 98 mulheres examinadas entre 2009 e 2013. A terminologia foi baseada na edição de 2013 do BI-RADS. RESULTADOS: Dos 121 achados suspeitos, 53 (43,8%) eram de fato lesões malignas, sem diferença significativa entre nódulos e realce não nodular (p = 0,846). Nódulos com margem espiculada (71%) e forma redonda (63%) apresentaram os maiores VPPs, ao passo que a distribuição segmentar teve alto VPP para realce não nodular (80%). Apenas a curva cinética do tipo 3 teve bom desempenho quando aplicada a nódulos (VPP = 73%). Modelos de regressão logística foram significantes para os dois padrões principais, embora os nódulos tenham apresentado resultados melhores, particularmente com a introdução da análise cinética (p = 0,015; pseudo-R2 = 0,48; área sob a curva = 90%). CONCLUSÃO: Alguns descritores de RM do BI-RADS têm alto VPP e bom desempenho preditivo - demonstrado por curva ROC e análise multivariada - quando aplicados a achados da categoria 4 segundo o BI-RADS. Isso pode permitir futura estratificação dessa categoria.

Predictive performance of BI-RADS magnetic resonance imaging descriptors in the context of suspicious (category 4) findings / Desempenho preditivo dos descritores de ressonância magnética do BI-RADS no contexto de achados suspeitos (categoria 4)

Abstract Objective: To determine the positive predictive value (PPV) and likelihood ratio for magnetic resonance imaging (MRI) characteristics of category 4 lesions, as described in the Breast Imaging Reporting and Data System (BI-RADS®) lexicon, as well as to test the predictive performance of the descriptors using multivariate analysis and the area under the curve derived from a receiver operating characteristic (ROC) curve. Materials and Methods: This was a double-blind review study of 121 suspicious findings from 98 women examined between 2009 and 2013. The terminology was based on the 2013 edition of the BI-RADS. Results: Of the 121 suspicious findings, 53 (43.8%) were proven to be malignant lesions, with no significant difference between mass and non-mass enhancement (p = 0.846). The PPVs were highest for masses with a spiculated margin (71%) and round shape (63%), whereas segmental distribution achieved a high PPV (80%) for non-mass enhancement. Kinetic analyses performed poorly, except for type 3 curves applied to masses (PPV of 73%). Logistic regression models were significant for both patterns, although the results were better for masses, particularly when kinetic assessments were included (p = 0.015; pseudo R2 = 0.48; area under the curve = 90%). Conclusion: Some BI-RADS MRI descriptors have high PPV and good predictive performance-as demonstrated by ROC curve and multivariate analysis-when applied to BI-RADS category 4 findings. This may allow future stratification of this category.

Resumo Objetivo: Determinar o valor preditivo positivo (VPP) e a razão de verossimilhança positiva de características de ressonância magnética (RM) de lesões da categoria 4, como descritas no léxico do Breast Imaging Reporting and Data System (BI-RADS®), e testar o desempenho preditivo dos descritores por meio de análise multivariada e área sob a curva derivada da curva receiver operating characteristic (ROC). Materiais e Métodos: Foi realizado um estudo revisional duplo-cego de 121 achados suspeitos em 98 mulheres examinadas entre 2009 e 2013. A terminologia foi baseada na edição de 2013 do BI-RADS. Resultados: Dos 121 achados suspeitos, 53 (43,8%) eram de fato lesões malignas, sem diferença significativa entre nódulos e realce não nodular (p = 0,846). Nódulos com margem espiculada (71%) e forma redonda (63%) apresentaram os maiores VPPs, ao passo que a distribuição segmentar teve alto VPP para realce não nodular (80%). Apenas a curva cinética do tipo 3 teve bom desempenho quando aplicada a nódulos (VPP = 73%). Modelos de regressão logística foram significantes para os dois padrões principais, embora os nódulos tenham apresentado resultados melhores, particularmente com a introdução da análise cinética (p = 0,015; pseudo-R2 = 0,48; área sob a curva = 90%). Conclusão: Alguns descritores de RM do BI-RADS têm alto VPP e bom desempenho preditivo - demonstrado por curva ROC e análise multivariada - quando aplicados a achados da categoria 4 segundo o BI-RADS. Isso pode permitir futura estratificação dessa categoria.

Simple magnetic resonance imaging criteria can differentiate ductal carcinomasin situ from invasive carcinomas / Critérios simples de imagem por ressonância mangética podem diferenciar carcinomas ductais in situ de carcinomas invasivos

Objective: To investigate if apparent diffusion coefficient (ADC) values can discriminate ductal carcinomas in situ (DCIS) from invasive carcinomas and to test the incremental gain of a model combining these measurements to dynamic contrast-enhanced (DCE) main pattern (mass versus non-mass). Methods: Forty-four lesions (12 DCIS and 32 invasive cancers) were reviewed by two examiners, their ADCs were averaged, and they were classified according to enhancement patterns. A logistic regression model with ADC values and enhancement patterns was devised. Receiver operating characteristic (ROC) curves were used to compare the discriminative performance of isolated ADCs to the regression model by their areas under the curve (AUCs). Results: ADC values were significantly different between lesion types (p=0.034), with mean of 1.23x10-3 mm2/s for DCIS and 1.05x10-3 mm2 /s for invasive cancers. The model grouping enhancement patterns and ADC values had better performance AUC=0.80) than isolated ADCs (AUC=0.71), though the difference was not statistically significant (p=0.105). Conclusion: ADC measurements of pre-invasive breast lesions are substantially different from those of invasive cancers. When ADC measurements are associated with main enhancement patterns, the performance of the technique is increased.

Objetivo: Investigar se valores de coeficiente de difusão aparente (ADC) podem discriminar carcinomas ductais in situ (CDIS) de carcinomas invasivos e testar o ganho incremental de modelo combinando tais medidas ao padrão principal de realce (nódulo versus realce não nodular) do estudo contrastado dinâmico (ECD). Métodos: Quarenta e quatro lesões (12 CDIS e 32 cânceres invasivos) foram revisadas por dois examinadores, seus ADCs médios calculados e elas foram classificadas de acordo com padrões de realce. Um modelo de regressão logística com valores de ADC e padrões de realce foi delineado. Curvas receiver operating characteristic (ROC) foram utilizadas para comparar a performance discriminativa dos ADCs isolados ao modelo de regressão através de suas áreas sob a curva (AUCs). Resultados: Os valores de ADC foram significantemente diferentes entre tipos de lesão (p=0,034), com média de 1,23x10-3 mm2 /s para CDIS e 1,05x10-3 mm2 /s para cânceres invasivos. O modelo agrupando padrões de realce e valores de ADC teve melhor performance (AUC=0,80) do que ADCs isolados (AUC=0,71), ainda que a diferença não tenha sido estatisticamente significante (p=0,105). Conclusão: Medidas de ADC de lesões mamárias pré-invasivas são substancialmente diferentes daquelas de cânceres invasivos. Quando medidas de ADC são associadas aos principais padrões de realce, a performance da técnica é aumentada.

Subcategorization of Suspicious Breast Lesions (BI-RADS Category 4) According to MRI Criteria: Role of Dynamic Contrast-Enhanced and Diffusion-Weighted Imaging.

OBJECTIVE: The purposes of this study were to investigate whether dynamic contrast-enhanced MRI is adequate for subcategorization of suspicious lesions (BI-RADS category 4) and to evaluate whether use of DWI improves diagnostic performance. MATERIALS AND METHODS: The study group was composed of 103 suspicious lesions found in 83 subjects. Patient ages and lesion sizes were compiled, and two radiologists reanalyzed the images; subcategorized the findings as BI-RADS 4A, 4B, or 4C; and calculated apparent diffusion coefficient (ADC) values. The stratified variables were tested by univariate analysis and inserted in two multivariate predictive models, which were used to generate ROC curves and compare AUCs. Positive predictive values (PPVs) for each subcategory and ADC level were calculated, and interobserver agreement was tested. RESULTS: Forty-four (42.7%) suspicious findings proved malignant. Except for age (p = 0.08), all stratified predictor variables were significant in univariate analyses (p < 0.01). Logistic regression models did not differ substantially after comparison of the ROC curves (p = 0.09), but the one including ADC values was slightly better: AUC of 0.89 (95% CI, 0.82-0.95) against AUC of 0.85 (95% CI, 0.78-0.93). PPV increased progressively in each BI-RADS 4 subcategory (4A, 0.15; 4B, 0.37; 4C, 0.84). ADC values of 1.10 × 10(-3) mm(2)/s or less had the second highest PPV (0.77). Interobserver agreement was substantial at a kappa value of 0.80 (95% CI, 0.70-0.90; p < 0.01). CONCLUSION: Risk stratification of suspicious lesions (BI-RADS category 4) can be satisfactorily performed with DCE-MRI and slightly improved when DWI is introduced.