1.1 Clinical data
Eighty-seven patients with pathologically confirmed HER-2 low-expressing breast cancer and 51 HER-2-positive breast cancer within the First Affiliated Hospital of Hunan University of Traditional Chinese Medicine between 2014-01-01 and 2022-04-12 were included in this study and retrospectively analyzed. (Our Institutional Review Board and Ethics Committee approved the study protocol with the consent of all subjects.)Inclusion criteria: (1) Those who had not undergone surgery or radiotherapy before MRI examination was performed. (2) Those with pathological tests consistent with HER-2 low expression or HER-2-positive breast cancer. (3) Lesions with measurable maximum diameter ≥ 10 mm. exclusion criteria: (1) Patients with poor image quality and excessive artifacts that interfere with diagnostic analysis. (2) Incorrect or incomplete clinical and imaging data. (3) Patients in the stage of pregnancy or pregnancy preparation. 35 patients with HER-2 low expression breast cancer and 25 patients with HER-2-positive breast cancer were finally excluded. Fifty-two patients with HER-2 low-expressing breast cancer and 26 patients with HER-2-positive breast cancer were included.
1.2 Examination methods
The examination instruments were all GE 3.0T Signa HDxt superconducting MR scanner, which was equipped with an 8-channel breast-specific phased-array coil. Patients should be placed in prone position with both breasts naturally lowered in the breast coil. The scan sequence and parameters are set as follows: (1) cross-sectional diffusion-weighted imaging: matrix 128×128, 2 acquisitions, TE 62.5 ms, TR 6000 ms, FOV 32 cm×32 cm, layer spacing 1 mm, layer thickness 4 mm, b values 0 and 1000 mm2/s, respectively; (2) cross-sectional T2 asymmetric echo iterative water-fat separation sequence by least-squares estimation (3) cross-sectional Tl fast spin-echo sequence: matrix 320×192, acquired twice, TE 35 ms, TR 8200 ms, FOV 32 cm×32 cm, layer spacing 1 mm, layer thickness 4 mm; (3) cross-sectional Tl fast spin-echo sequence: matrix 320×192, acquired twice, TE 7.8 ms, TR 400 ms, FOV 32 cm×32 cm layer, thickness 4 mm, layer spacing 1 mm; (4) breast assessment volume imaging gradient echo sequence: matrix 352×320, TE 2.1 ms, TR 4 ms, layer thickness 1.4 mm, layer spacing 0.7 mm, FOV 37 cm×37 cm, inversion angle 12°. Pre-contrast cross-sectional pre-scan was performed first, followed by injection of contrast agent (Gadodiamide Injection) via the elbow vein, a total of 15 ml was given at a flow rate of 2 ml/s. The scan was started after 30 s. Digital subtraction was performed simultaneously with the scan, and six consecutive images were acquired with a scan time of 60 s in each time phase.
1.3 ADC value measurement and TIC curve generation
The MRI images of all patients were post-processed using Functool software, and the region of interest (ROI) was selected in the area with the highest parenchymal enhancement and the largest extent of the lesion. The average value of the three measurements is taken, and the TIC curve is automatically generated in the system.
1.4 Magnetic resonance characterization
Two highly qualified professional imaging physicians independently analyzed and recorded the MRI image characteristics of all patients, including the basic
Table 1 Comparison of MRI features between low expression group and positive group
|
Features
|
Low Expression
|
Positive
|
F/t
|
P值
|
Age
|
51±10.65
|
50.5±11.42
|
0.191
|
0.849
|
Histological grading
|
|
|
454
|
0.001
|
Class I
|
7
|
0
|
|
|
Class II
|
24
|
8
|
|
|
Class III
|
23
|
20
|
|
|
Lesion distribution
|
|
|
1.135
|
0.606
|
Unifocal lesions
|
41
|
18
|
|
|
Multifocal lesions
|
5
|
4
|
|
|
Multicentric lesions
|
6
|
4
|
|
|
Maximum diameter
|
|
|
5.375
|
0.02
|
≤2cm
|
24
|
5
|
|
|
>2cm
|
28
|
21
|
|
|
Morphology
|
|
|
2.786
|
0.246
|
Round-like
|
8
|
5
|
|
|
lobulated
|
32
|
19
|
|
|
Irregular
|
12
|
2
|
|
|
Edge
|
|
|
1.118
|
0.644
|
Smooth
|
8
|
5
|
|
|
Unclear
|
11
|
3
|
|
|
burr-like
|
33
|
18
|
|
|
Background Substance Enhancement
|
|
|
626
|
0.561
|
almost nothing
|
12
|
4
|
|
|
small amount
|
37
|
15
|
|
|
Medium amount
|
4
|
2
|
|
|
Significant
|
9
|
5
|
|
|
Internal reinforcement
|
|
|
3.9
|
0.048
|
Uniformity
|
0
|
0
|
|
|
Uneven
|
36
|
12
|
|
|
Ring
|
16
|
14
|
|
|
TIC Curve
|
|
|
1.321
|
0.434
|
Type I
|
1
|
1
|
|
|
Type II
|
5
|
4
|
|
|
Type III
|
48
|
22
|
|
|
ADC value (×10—3
|
0.767±0.144
|
0.862±0.150
|
2.714
|
0.008
|
s/mm2)
|
|
|
|
|
conditions of lesions (number, distribution, maximum diameter), MRI signs (internal strengthening form, morphology, margins, T2 signal condition, internal strengthening characteristics, ADC value, TIC curve type). If there was any dispute on the above recorded results, higher qualified physicians were invited for further discussion and analysis, and the final results were obtained by consultation among the three.
1.5 Immunohistochemical analysis
According to the newly released Chinese Society of Clinical Oncology Breast Cancer Treatment Guidelines 2021, HER-2 expression criteria were defined as follows.
The criteria for positive HER-2 expression were: (1) HER-2 immunohistochemistry result was (++++). (2) HER-2 immunohistochemistry result of (++) was followed by FISH test, and when the FISH test result was (+), HER-2 expression was positive.
The criteria for HER-2 low expression were: (1) HER-2 immunohistochemistry result of (+). (2) When the HER-2 immunohistochemistry result was (++), FISH test was performed, and when the FISH test result was (-), HER-2 was considered as low expression.
1.6 Statistical analysis
Data were analyzed for each clinicopathological and MRI characteristic index using SPSS 25.0 statistical analysis software. patient age, tumor histological grade, DCE-MRI morphological features, enhancement and DWI quantitative/semi-quantitative. parameters were compared in the HER-2-positive group with the HER-2 low expression group. For the count data, Fisher's exact probability method Pearson test or Yates corrected chi-square analysis could be used. Three methods of independent sample t-test, Wilcoxon signed rank sum test or independent sample approximate t-test could be used for the measurement data. Dichotomous logistic regression analysis was used for the multifactorial predictive differential analysis between the HER-2 low expression group and the HER-2 positive group. For continuous variables (ADC values), ROC curves were used for further analysis to determine the optimal threshold for differentiating the HER-2 low expression group from the HER-2 positive group according to the magnitude of sensitivity and specificity, and the area under the curve was calculated. The difference was considered statistically significant at P<0.05.