Breast cancer is one of the most highly affecting tumors in women, with more than 48000 new cases per year in Italy. In the routine clinical practice mammography, ultrasound and magnetic resonance imaging (MRI) are commonly performed. Mammography is recognized as the screening technique for breast cancer detection worldwide. Ultrasound can figure out the solid or liquid features of a lesion, as well as discover new lesions missed out on mammography. MR is a second level technique using contrast media. Contrast enhanced mammography (CEM) is a recent imaging tool which uses contrast-enhanced recombined images for the assessment of tumour angiogenesis. CEM uses a dual energy technique to generate a high-resolution, low-energy digital mammography image and a high-energy contrast-enhanced image that provides information on lesion vascularity. Subsequently, these two images are recombined into one image, resulting in a digital subtracted image of the relative distribution of iodine in the breast. The aim of the study is to compare the measures of breast cancer taken on CEM with the ones given by the surgical specimens, analyzing factors causing discrepancies between the two techniques. The protocol was approved by the Ethics Commitee of Careggi University Hospital. Each patient provided written informed consent. From January 2016 to January 2018 263 consecutive patients (mean age 59.03 yrs) affected by a hystologically proven breast cancer underwent a CEM exam. Exclusion criteria were:patients undergoing primary systemic therapy; cases with index lesion out of CEM field of view; cases with post-core biopsy hematoma or background enhancement. CEM examination was performed on a Selenia mammography system (Hologic, Massachusetts USA) equipped with an automatic power injector (Bayer HealthCare, Berlin Germany). After a single injection of contrast medium (Iomeron 400 mg / mL, 1.5 mL / kg) a Cranio-Caudal and a Mediolateral Oblique view were acquired, starting from the affected breast. Three board-certificated radiologists with at least 25 years of experience in breast imaging evaluated the CEM exams on a dedicated workstation.They measured the maximum size on the early scans, observing wash-out in the subsequent ones, when occurred. To evaluate the type of lesion enhancement, MR morphology descriptors were adopted: Mass Enhancement (three-dimentional space occupying lesion), Non Mass Enhancement (NME, non-space occupying lesion), and Ring enhancement. Pearson’s Correlation analysis was performed to match virtual imaging and specimens (largest diameter) measures. Significance was considered when p < 0.01 level. For the agreement between measurements, the mean differences in size found by CEM vs histopathology findings were obtained to determine the limits of agreement by Bland Altman plots. A difference of 10 mm between CEM measures and histological ones was assumed as a critical cut-off, referring to the surgical margins and to the existing literature. Therefore the sample has been divided in two groups: group “A” including perfect congruence between imaging cancer diameter and histological cancer diameter, CEM diameter overestimation < = 10 mm and CEM diameter underestimation <10 mm, group B including CEM diameter overestimation >10 mm and CEM diameter underestimation >10 mm. The final series included 162 patients. A strong correlation (r = 0.852) between the two variables and a slight CEM cancer size overestimation were detected. The Bland-Altman agreement analysis of histology-CEM revealed a mean of 3,19 mm bias between the two measurements. Limits of agreement (LOA 95%) -12,08 mm to 18,48 mm. The current study shows an excellent CEM accuracy in the pre-operative evaluation of breast cancer size, with a slight CEM overestimation compared to histology. It could be a valid option when MRI can not be performed.
Il tumore al seno è una delle neoplasie di maggior riscontro nella donna, con più di 48.000 nuovi casi diagnosticati ogni anno in Italia. Ad oggi nella pratica clinica le metodiche utilizzate sono la mammografia, l’ ecografia e la risonanza magnetica (RM). La mammografia è l'unica metodica riconosciuta per lo screening del tumore al seno. L’ecografia consente di determinare la natura solida o liquida di una lesione e di individuare lesioni non visibili mammograficamente. La RM è un esame di secondo livello, complementare alla mammografia ed all’ecografia, che utilizza mezzo di contrasto. La mammografia con mezzo di contrasto (CEDM) è una recente tecnica di imaging che, evidenziando l’angiogenesi tumorale, fornisce due tipi di informazioni: una di tipo morfologico ed una di tipo funzionale. Tutti i mammografi attualmente in commercio utilizzano la tecnica a doppia energia, che sfrutta la dipendenza energetica dell’attenuazione subita dai raggi X dalle strutture che compongono la mammella. Una coppia di due immagini, a bassa ed alta energia, è ottenuta dopo infusione di mezzo di contrasto. Successivamente, le due immagini vengono elaborate per evidenziare le aree di contrast uptake. L’obiettivo di questo studio è confrontare le dimensioni tumorali misurate in CEDM con l’esame istologico, analizzando anche eventuali fattori responsabili delle differenza fra le due misurazioni. Il protocollo è stato approvato dal comitato etico dell’Azienda Ospedaliera Careggi. Tutte le pazienti hanno firmato il consenso informato prima di sottoporsi all’esame. Da Gennaio 2016 a Gennaio 2018, 263 pazienti (età media 59.03 aa) affette dal tumore al seno istologicamente comprovato sono state sottoposte ad esame CEDM. I criteri di esclusione sono stati: pazienti sottoposte a terapia sistemica; casi in cui la lesione indice fuoriusciva dal campo di vista; presenza di ematoma post-biopsia o significativo background enhancement. Gli esami CEDM sono stati effettuati con sistema mammografico Selenia (Hologic, Massachusetts USA), dotato di iniettore automatico (Bayer HealthCare, Berlin Germany). Dopo una singola iniezione di mezzo di contrasto iodato (Iomeron 400 mg / mL, 1.5 mL / kg) sono state acquisite due proiezioni, cranio-caudale e medio-laterale-obliqua. Tre radiologi con 25 anni di esperienza nell’imaging mammario hanno visualizzato gli esami su una workstation dedicata. Sono state misurate le dimensioni delle lesioni indice riportando il diametro massimo nelle scansioni precoci e cercando il wash-out in quelle tardive. Per valutare il tipo di enhancement sono stati utilizzati i criteri RM: Mass Enhancement, Non Mass Enhancement (NME), Ring enhancement. Il test di Pearson è stato eseguito per correlare le dimensioni virtuali con quelle dell’esame istologico. Valori di p<0.01 sono stati considerati statisticamente significativi. Le differenze medie riscontrate in CEDM rispetto ai risultati istologici sono state utilizzate per determinare i LOA attraverso i Bland-Altman plots. Una differenza di 10mm tra le dimensioni CEDM e quelle istologiche è stata assunta come cut-off. Il campione è stato suddiviso in due gruppi: gruppo A, che include i casi in cui si è riscontrata una perfetta congruenza tra i diametri misurati in CEDM e quelli istologici; i casi in cui si è osservata una sovrastima ed una sottostima all’imaging <10mm; gruppo B che ha incluso i casi in cui si è osservata una sovrastima ed una sottostima >10mm. Il campione finale comprendeva 162 pazienti. Si è osservata una forte correlazione (r=0.852) tra le due variabili ed una lieve sovrastima da parte della CEDM. L’analisi di concordanza di Bland-Altman ha rivelato un bias medio di 3.19mm tra le due misurazoni. LOA (95%) compresi tra -12.08mm e 18.48mm. Lo studio ha dimostrato un’ eccellente accuratezza della metodica nella valutazione preoperatoria del tumore al seno, con una lieve sovrastima da parte della CEDM. La CEDM potrebbe essere una valida alternativa alla RM nel bilancio di estensione della neoplasia, specialmente in presenza di controindicazioni all'esecuzione della stessa.
Ruolo della mammografia con mezzo di contrasto nella valutazione dimensionale pre-operatoria del tumore al seno
PICCOLO, Claudia Ludia
2020-05-11
Abstract
Breast cancer is one of the most highly affecting tumors in women, with more than 48000 new cases per year in Italy. In the routine clinical practice mammography, ultrasound and magnetic resonance imaging (MRI) are commonly performed. Mammography is recognized as the screening technique for breast cancer detection worldwide. Ultrasound can figure out the solid or liquid features of a lesion, as well as discover new lesions missed out on mammography. MR is a second level technique using contrast media. Contrast enhanced mammography (CEM) is a recent imaging tool which uses contrast-enhanced recombined images for the assessment of tumour angiogenesis. CEM uses a dual energy technique to generate a high-resolution, low-energy digital mammography image and a high-energy contrast-enhanced image that provides information on lesion vascularity. Subsequently, these two images are recombined into one image, resulting in a digital subtracted image of the relative distribution of iodine in the breast. The aim of the study is to compare the measures of breast cancer taken on CEM with the ones given by the surgical specimens, analyzing factors causing discrepancies between the two techniques. The protocol was approved by the Ethics Commitee of Careggi University Hospital. Each patient provided written informed consent. From January 2016 to January 2018 263 consecutive patients (mean age 59.03 yrs) affected by a hystologically proven breast cancer underwent a CEM exam. Exclusion criteria were:patients undergoing primary systemic therapy; cases with index lesion out of CEM field of view; cases with post-core biopsy hematoma or background enhancement. CEM examination was performed on a Selenia mammography system (Hologic, Massachusetts USA) equipped with an automatic power injector (Bayer HealthCare, Berlin Germany). After a single injection of contrast medium (Iomeron 400 mg / mL, 1.5 mL / kg) a Cranio-Caudal and a Mediolateral Oblique view were acquired, starting from the affected breast. Three board-certificated radiologists with at least 25 years of experience in breast imaging evaluated the CEM exams on a dedicated workstation.They measured the maximum size on the early scans, observing wash-out in the subsequent ones, when occurred. To evaluate the type of lesion enhancement, MR morphology descriptors were adopted: Mass Enhancement (three-dimentional space occupying lesion), Non Mass Enhancement (NME, non-space occupying lesion), and Ring enhancement. Pearson’s Correlation analysis was performed to match virtual imaging and specimens (largest diameter) measures. Significance was considered when p < 0.01 level. For the agreement between measurements, the mean differences in size found by CEM vs histopathology findings were obtained to determine the limits of agreement by Bland Altman plots. A difference of 10 mm between CEM measures and histological ones was assumed as a critical cut-off, referring to the surgical margins and to the existing literature. Therefore the sample has been divided in two groups: group “A” including perfect congruence between imaging cancer diameter and histological cancer diameter, CEM diameter overestimation < = 10 mm and CEM diameter underestimation <10 mm, group B including CEM diameter overestimation >10 mm and CEM diameter underestimation >10 mm. The final series included 162 patients. A strong correlation (r = 0.852) between the two variables and a slight CEM cancer size overestimation were detected. The Bland-Altman agreement analysis of histology-CEM revealed a mean of 3,19 mm bias between the two measurements. Limits of agreement (LOA 95%) -12,08 mm to 18,48 mm. The current study shows an excellent CEM accuracy in the pre-operative evaluation of breast cancer size, with a slight CEM overestimation compared to histology. It could be a valid option when MRI can not be performed.File | Dimensione | Formato | |
---|---|---|---|
Tesi_CL_Piccolo.pdf
accesso aperto
Descrizione: Tesi di Dottorato
Dimensione
3.15 MB
Formato
Adobe PDF
|
3.15 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.