br One possibility is to compare the
One possibility is to compare the apparent diffusion coefficient (ADC) obtained by diffusion-weighted imag-ing before and after chemotherapy. This type of magnetic resonance imaging (MRI) can provide information about tumor texture, cellularity and proliferation potential, mak-ing it more effective than physical examination or other imaging modalities for evaluating breast cancer (19). Some studies have shown a negative correlation between ADC and the Ki-67 index (20,21), but this work did not systematically compare pre- and post-therapy situations. Therefore, we undertook the present study to examine whether change in ADC as a result of NAC might serve as a surrogate for change in Ki-67 index in patients with invasive breast cancer.
MATERIALS AND METHODS
This retrospective study was approved by our Institutional Review Board, and the requirement to obtain informed con-sent was waived. A clinical database of 154 breast cancer patients who underwent NAC in our unit between February 2016 and October 2017 were reviewed. Eighty-seven patients met the following criteria were included in our study: (1) pCR was not achieved after NAC, (2) MRI data were available pre- and post-NAC, (3) specimens for detect-ing Ki-67 were taken by core needle biopsy before NAC and by surgery after NAC within 7 days after MRI, and (4) lesions found by MRI were >0.5 cm before and after NAC. Forty-three patients were excluded because they had achieved pCR after NAC (n = 17), had not undergone MRI both before and after NAC (n = 14), had not got specimens for detecting Ki-67 both before and after NAC (n = 5), or had the lesion <0.5 cm after NAC (n = 7). And 24 patients who had undergone previous treatment or biopsy at an out-side institution were also excluded. For the patients had mul-tiple breast carcinoma, we evaluated only one lesion with a largest diameter per patient although multiple tumors may be heterogeneous.
IHC was used to determine the Ki-67 index before NAC (Ki-67pre) and afterwards (Ki-67post). All immunohistochemi-cal analyses were carried out in the same reference laboratory using anti-Ki-67 antibody (Clone MIB1, Dako, Glostrup, Denmark; 1:50 dilution), and stained sections were examined independently by two pathologists blinded to patient
treatment response. Cells that stained positively for Ki-67 were counted and expressed as a percentage of all GelGreen examined. Two parameters were calculated by comparing the Ki-67 index before and after NAC: one was the change in Ki-67 index, defined as DKi-67 = Ki-67post ¡ Ki-67pre; the other was the percentage change in Ki-67 index, defined as DKi-67% = [Ki-67post ¡ Ki-67pre]/Ki-67pre.
Estrogen receptor and human epidermal growth factor receptor-2 status were also evaluated before NAC by IHC using antibodies (M7047 and polyclonal, respec-tively, Dako). The cutoff value for estrogen receptor posi-tivity was 10% positive tumor cells with nuclear staining. And HER2 positivity was defined as 3+, complete membrane staining or fluorescence in situ hybridization amplification.
Breast tissue was imaged by MR using a 1.5-T system (Mag-netom Avanto, Siemens Healthcare, Germany) equipped with a dedicated, 8-channel phased array breast coil. Subjects lay in the prone position during scanning, and the scan parameters were as described (19).
Diffusion-weighted MR images were acquired in the axial plane using an echo-planar imaging sequence, parallel imaging with sensitivity encoding (acceleration factor: 2), fat suppression (in a spectral selective attenuated inversion-recovery sequence), volume shimming, and the following quantitative parameters: b values, 0 and 800 s/mm2; repetition time, 5800 ms; echo time, 86 ms; inversion time, 180 ms; section thickness, 6 mm; intersection gap, 0.2 mm; field of view, 32 £ 32 cm; and matrix dimensions, 323 £ 448. ADC maps were automatically created by the system from trace-weighted images with b val-ues of 0 and 800. ADC was calculated according to the follow-ing formula: ADC = ¡(1/b) ln (S2/S1), where S2 and S1 were the signal intensities at respective b values of 800 and 0.
For imaging with multiphase dynamic contrast-enhancement, images were acquired before contrast agent injection and approximately 15 seconds afterwards. The contrast agent was Gd-DTPA (Magnevist meglumine, Bayer Health Care Pharmaceuticals, Germany), which was injected at a dose of 0.1 mmol/kg body weight at the rate of 2.0 ml/s, followed by a 20-ml saline flush. A total of 8 10 phases were continuously collected and each phase took 55 seconds to acquire.
Analysis of MRI
To reduce the effects of partial volume artifacts, the smallest lesion for ADC calculation was set to a diameter of 0.5 cm. Diffusion-weighted images and dynamic contrast-enhanced images were reviewed on a postprocessing workstation (Syngo MR B17, Siemens Systems) by two experienced radi-ologists blinded to pathology findings. The tumor diameters before NAC (Dpre) and afterwards (Dpost) were measured using dynamic contrast-enhanced images. The tumors were