The institutional review board approved this retrospective study in which we reviewed 71 breast cancers in 60 patients waiving the requirement to obtain any informed consent.
All patients underwent DM,
DBT,
and MRI between January 2015 and March 2017,
before definitive surgery.
The patients had a mean age of 52.5 years (range,
34-71 years).
Patients undergone to neoadjuvant chemotherapy were excluded from the study.
All patients meeting the following inclusion criteria (approved by the local ethics committee) in accordance with the literature and the European Society of Breast Cancer Specialists (EUSOMA) guidelines3,6.
The 60 patients included in our study underwent preoperative MRI staging for: (1) newly diagnosed invasive lobular carcinoma; (2) high risk of breast cancer; (3) being eligible candidates for partial breast irradiation (PBI) on the basis on conventional imaging.
Others EUSOMA recommendations4, considered in our study,
included patients: (1) with dense breasts; (2) with multifocal,
multicentric or bilateral cancer (invasive and/ or ductal carcinoma in situ,
DCIS) demonstrated on conventional imaging and confirmed by pathology; (3) with unilateral unifocal pure DCIS at conventional imaging (to exclude synchronous ipsilateral or contralateral invasive carcinoma); (4) candidate for total skin-sparing mastectomy (evaluation of nipple-areola complex).
Furthermore all patients underwent combined DM and DBT (combo mode): (1) women proposed for needle biopsy; (2) asymptomatic women between 30 and 80 years.
Before to be undergoing the mammographic examination,
all patients were informed about the technical aspects,
benefits and risks of the DBT,
and they subsequently gave their informed consent.
Combo mode is performed through the acquisition of standard bilateral two-view mammograms (cranio-caudal and medio-lateral oblique) once in DBT and then whitin a single compression for each projection in DM.
The mammography machine was Senographe Pristina 3D.
Nine low-dose DBT projection images are acquired during movement of the x-ray tube through a 25° arc (±23.4°).
Overall scanning time is about 5 seconds and reconstruction time about 10 seconds.
All patients were subsequently underwent preoperative MRI (Area 1.5 T,
Siemens) using a dedicated 4-channel phased-array coil. The DBT,
DM (evaluated individually or in combination) and MR images were reviewed for retrospective independent interpretation by two experienced radiologists in breast imaging (with more than 2 years of experience in DBT); the radiologists were aware of the study purpose,
but were blinded to the histological lesion size.
Images of each modality were presented randomly to each reader in separate sessions and any discrepancy in opinion was resolved by consensus.
For each lesion visible on DM,
DBT and MRI,
the the major axis of the tumour size using dedicated software on the workstation.
All patients underwent surgical excision of the primary tumour and sentinel lymph node biopsy and/or axillary dissection.
The breast sample were sent to the Pathology Department laboratories of the of our hospital,
after verification of current presence of the lesion in the surgical specimen by radiographic evaluation. Each sample,
after macroscopic measurement,
was cut into multiple 4-mm-thick sections from the deep margin to the superficial edge towards the skin.
Each slice of breast tissue was placed on sheet of blotting paper,
taking care to maintain the orientation established by the surgeons.
If the lesion was visible,
it was macroscopically measured along the two main axes,
and its distance from each surgical resection margin was recorded.
When the tumour was not macroscopically identified,
sampling was performed following the indications of the radiologist and showing the exact position of the lesion on the radiograph of the piece,
received at the Pathology Department.
Moreover,
in each individual case the opportunity to perform macrosements was evaluated,
which were set up especially in cases where the tumor was (i) particularly extended,
(ii) not delimitable with certainty for the presence of spicule or thickening around to the neoplasm,
(iii) not visible macroscopically,
as in the case of in situ tumors with intra-adipose growth and diffusion and identifiable only for the presence of microcalcifications.
The sampling phase was carried out using a photographic equipment,
installed under the extraction hood,
which allows images of each level to be acquired,
as well as marking on them the exact location of the single sampling.
The samples were placed in special plastic biocassettes and then in the automatic processor where they underwent passages in alcohols,
xylene and paraffin.
After processing,
the tissue in the biocassettes was included in paraffin and were cut from this section 4 micron thickness,
which were placed on slides and stained with hematoxylin and eosin.
The microscopic histological dimensions of the lesions were made on the slides using a metric eyepiece.
Moreover,
in cases that are more difficult to delimit,
given the presence of multiple foci in different areas,
the overall extension has also been obtained through 3D reconstructions.
In fact,
if the visualization of the neoplasia occurred on sections at different levels,
being aware of the thickness of the single level,
it was possible to obtain a three-dimensional measurement of the lesion.
Statistical Analysis: for each lesion the difference in millimeters was calculated between the measurement detected by each imaging method and the histological examination,
in order to evaluate the concordance and to reproduce the dispersion of the sample.
Measurements were considered consistent with histology if included within ± 5 mm,
underestimated and overestimated respectively if <5 mm and> 5 mm compared to histological examination.
In the assessment of concordance,
the lesions were also stratified according to the dimensions 2 cm and> 2 cm.
The statistically significant differences between the values observed in the study were calculated using the Student's t test,
using a value of p <0.01 as significance threshold.