This prospective double-institution observational study included all consecutive female patients with pathologically proven breast cancer with a decision to start NAC after being reviewed by each institution’s Breast Unit multidisciplinary team. All patients were initially diagnosed with baseline mammography and breast ultrasound followed by a core needle biopsy under sonographic guidance from the radiologically suspicious masses for histopathological confirmation. Data were collected from January 2017 to February 2021. The study was approved by the institutions’ Ethics Committees. Written informed consent was obtained from all patients who agreed to take part in this study. Inclusion criteria included female patients’ ages more than 18 years, breast cancer diagnosis based on core needle biopsy, completion of institutional NAC treatment protocol followed by surgery and histopathology as well as completion of pre- and post-NAC CESM examinations. Exclusion criteria included contrast media reaction, renal impairment, pregnancy, incomplete NAC treatment or CESM examinations, and lack of informed consent.
All mammographic examinations were performed at the institutions’ Radiology Departments using the GE Healthcare machine (Chalfont St. Giles, UK) allowing dual-energy CESM acquisitions with some specific hardware and software enhancements for image processing and acquisition.
Unlike conventional mammography, the X-ray spectrum was tailored to produce energies just above the K-edge of iodine (33.2 keV) to increase the sensitivity to low concentrations of iodine to optimize the visualization of iodine. Typically, for a 50% glandular and 5-cm-thick breast, exposure times were around 1 s and 3 s for low and high-energy images, respectively.
An intravenous injection of 1.5 ml/kg body weight of a non-ionic contrast medium (Omnipaque) was done manually at one shot with a flow rate of 3 ml/s followed by a flush of saline. After 2 min of contrast medium injection, the patient was positioned, and the exposure was done in the four standard mammographic views: two mediolateral oblique (MLO) and two craniocaudal (CC) views. In each view, the breast was subjected to a combination of low and high-energy exposures, while the breast was still compressed to minimize motion blur keeping the image details seen with the standard mammography. The subtracted images were automatically generated after each low- and high-energy exposure for each view. The average examination time was 10 min. After the examination, the patients were monitored for around 30 min for any potential contrast medium adverse reactions.
Two sets of CESM were evaluated: one before the start of the NAC treatment and the second after the last dose of chemotherapy with a maximum interval of 14 days from the elective surgery. In our institutions, CESM is displayed paired to allow for comparison. The MLO is assessed followed by CC views with the patient’s right breast on the left side of the screen and the left breast on the left side of the screen. In the CC views, the outer aspect of the breasts is displayed superiorly.
Interpretation of CESM studies was performed via one of the participating radiologists, and in cases of interobserver disagreement, the case was discussed, and a joint consensus was reached. CESM images were evaluated in accordance with the Breast Imaging Reporting and Data System (BI-RADS). The largest dimension of the lesion was recorded from the two consecutive CESMs taken before the start and at the end of NAC.
The Response Evaluation Criteria in Solid Tumors (RECIST) 1.1 criteria were used in our study to evaluate the response to NAC where the response was classified as follows:
“Complete response (CR, disappearance of all lesion/s)
Partial response (PR, ≥ 30% reduction in longest diameter)
Stable disease (SD, < 30% reduction/ < 20% increase in longest diameter)
Progressive disease (PD, ≥ 20% increase in longest diameter)” [11, 27]
We compared the largest diameter of the residual lesion taken from the CESM, 2 weeks before surgery to that in the postoperative histopathological study. The efficacy of CESM in assessing residual disease after NAC was compared to the histological analysis, which was considered the gold standard.
Following surgery, all surgical specimens were sent to the institutions’ laboratories for histopathological confirmation. Apart from the macroscopic measurement, the tumor’s longest diameter, which is required for identifying the T stage in the TNM classification, was confirmed histopathologically.
Data were statistically described in terms of mean ± standard deviation (± SD), range, or frequencies (number of cases), and percentages when appropriate. A comparison of tumor size between CESM and histopathology was done using Wilcoxon signed rank test for paired (matched) samples. For comparing the response between the different pathological types, chi-square ((χ2) test was performed. Correlation between various variables was done using Pearson moment correlation equation for linear relation of normally distributed variables and Spearman rank correlation equation for non-normal variables/nonlinear monotonic relation. Accuracy was represented using the terms sensitivity, specificity, + ve predictive value, and − ve predictive value. Two-sided p values less than 0.05 were considered statistically significant. The diagnostic performance index of CESM for CR was tested using the Clopper–Pearson test with 95% confidence intervals. All statistical calculations were done using the computer program IBM SPSS (Statistical Package for the Social Science; IBM Corp, Armonk, NY, USA) release 22 for Microsoft Windows.