GIRADS: método de clasificación de masas anexiales, revisión sistemática y metanálisis. 10 años de experiencia / GIRADS: adnexal masses classification method, systematic review and metanalysis. 10 years of experience

OBJETIVO: evaluar la experiencia en la utilización del método GIRADS para clasificar masas anexiales a diez años de su primera publicación. MÉTODO: Se realizó búsqueda de estudios que utilizan el sistema GIRADS: Medline (Pubmed), Google Scholar y Web of Science, desde enero de 2009 hasta diciembre de 2019. Se calculó la sensibilidad y especificidad agrupada, Likelihood ratio (LR) (+) y LR (-) y Odds ratio de diagnóstico (DOR). La calidad de los estudios se evaluó con QUADAS-2. RESULTADOS: Se identificaron 15 estudios y se incluyeron 13 de ellos con 4473 masas, 878 de ellas malignas. La prevalencia media de malignidad ovárica fue del 23 % y la agrupada de 19.6%. El riesgo de sesgo fue alto en cuatro estudios para el dominio "selección de pacientes" y fue bajo en todos en todos los estudios para los dominios "prueba índice" y "prueba de referencia". La sensibilidad, especificidad, LR (+) y LR (-) agrupadas y el DOR del sistema GIRADS para clasificar las masas anexiales fueron: 96.8% (intervalo de confianza [IC] 95% = 94% - 98%), 91.2 % (IC 95 % = 85% - 94%), 11.0 (IC 95% = 6.9 -13.4) y 0.035 (IC 95% = 0.02- 0.09), y 209 (IC 95% = 99-444), respectivamente. La heterogeneidad fue alta para la sensibilidad y especificidad. De acuerdo a la metaregresión, la heterogeneidad entre los estudios se explica por la prevalencia de malignidad, múltiples observadores y la ausencia de diagnóstico histopatológico para todos los casos incluidos en un determinado estudio. CONCLUSIÓN: el sistema GIRADS tiene un buen rendimiento diagnóstico para clasificar masas anexiales.

OBJECTIVE: to evaluate the experience of using GIRADS method to classify adnexal masses ten years after its publication. METHOD: A search was carried out for studies reporting on the use of the GIRADS system in the Medline (Pubmed), Google Scholar and Web of Science databases, from January 2009 to December 2019. Pooled sensitivity and specificity, Likelihood ratio (LR) (+) and LR (-) and Diagnostic Odds ratio (DOR) were calculated. The quality of the studies was assessed by QUADAS-2. RESULTS: 15 studies were identified, and 13 of them were included with 4473 masses, of which 878 were malignant. The mean prevalence of ovarian malignancy was 23% and the prevalence pooled. of 19.6%. The risk of bias was high in four studies for the domain 'patient selection' and low for all studies for the domains 'index test' and 'reference test'. The sensitivity, specificity, pooled LR (+) and LR (-) and the DOR of the GIRADS system to classify adnexal masses were 96.8% (95% confidence interval [CI] = 94% -98%), 91.2 % (95% CI = 85% -94%), 11.0 (95% CI = 6.9-13.4) and 0.035 (95% CI = 0.02-0.09), and 209 (95% CI = 99-444), respectively. Heterogeneity was high for both sensitivity and specificity. According to meta-regression, this heterogeneity was explained by the prevalence of malignancy, the use of multiple observers, and the absence of histopathological diagnosis for all cases included in a given study. CONCLUSION: the GIRADS system has a good diagnostic performance to classify adnexal masses.

Gynecologic imaging reporting and data system combined with three-dimensional contrast-enhanced ultrasonography for differential diagnosis of benign and malignant ovarian masses / 中国医学影像技术

Objective To evaluate the efficacy of the combination of gynecologic imaging reporting and data system (GI-RADS) uhrasonographic stratification and 3D contrast-enhanced ultrasonography (3D-CEUS) in identifying malignant lesions from benign ovarian masses.Methods Both of 2D ultrasound (2D-US) and 3D-CEUS were performed on 102 patients with ovarian masses.The perfusion characteristics of ovarian masses were observed with 3D-CEUS,and the 2D-US features of ovarian masses were analyzed based on GI-RADS.Simple and multiple Logistic regression analysis were used to investigate whether the independent risk predictors in differential diagnosis of benign and malignant ovarian could be confirmed.In addition,ROC curves were drawn.The diagnostic efficacy of GI-RADS combined with 3D-CEUS scoring system was evaluated and compared with that of only GI-RADS.Results Simple and multiple Logistic regression analysis confirmed that there were 8 independent predictors of malignant masses,including large papillary projections (≥7 mm),separated or wall thickness ≥3 mm,central blood flow,the proportion of solid part ≥50％,combination of ascites,high level enhancement,uneven distribution of contrast media in enhanced solid part and the vascular with characteristics as dense,tortuous and anfractuous.When using 4 points as the cut-off,the area under the curve (AUC) of GI-RADS combined with 3D-CEUS scoring system in identifying malignant ovarian masses was 0.969,higher than that of only GI-RADS (0.839;Z=1.64,P=0.029).Furthermore,the scoring system showed higher sensitivity,specificity,positive predictive value,negative predictive value and accuracy (all P＜0.001).Conclusion The combination of GI-RADS with 3D-CEUS can be more effective to distinguish malignant lesions from benign ovarian masses.