This article is mentioned in:

Introduction

Invasive ductal carcinoma (IDC), a heterogeneous disease, is the most common pathological type of breast cancer (1). There are three main prognostic determinants for breast cancer, including the lymph node status, tumor size and histological grade (2). The prognostic value of histological grade is considered to be equivalent to that of the lymph node status (3), whereas it is greater than that of tumor size (4). According to the World Health Organization (5), IDC can be classified into grades I, II and III, depending on the nuclear features, tubular formation and mitotic count. Compared with grades I and II, IDC of grade III is associated with a reduced time to relapse or mortality due to breast cancer (6). Therefore, the accurate determination of the IDC histological grade is particularly important for selection of the appropriate treatment and prediction of the disease prognosis.

Diffusion weighted imaging (DWI) is an advanced functional magnetic resonance imaging (MRI) technique, which is based on the measurement of water molecule diffusion in tissues (7,8). Diffusion is quantified by the apparent diffusion coefficient (ADC), with low ADC values suggesting restricted diffusion (8). DWI has been initially applied for the diagnosis of acute stroke in clinical practice (9). With the rapid development of MRI techniques over the past years, the clinical application of DWI has been greatly extended for the imaging of various tumors (7–10). DWI is particularly helpful in the evaluation of breast masses by providing information on tumor behavior (8).

There have been numerous studies regarding the application of DWI in the diagnosis of breast lesions, and in the differential diagnosis between benign and malignant breast tumors (11,12). The association between DWI and the histological grade of IDC has also been reported, although differential findings have been observed due to the use of a different region of interest (ROI) and measurement methods (13–20). The majority of previous studies have applied the mean ADC value to indicate the pathological characteristics of tumors. However, the minimum ADC (ADCmin) value is considered to be able to reflect the most malignant portions of tumors (21). In particular, the ADCmin value would be helpful for the selection of treatment plans if information on the IDC histological grade was also provided.

In the present retrospective study, the aim was to investigate the correlation of the ADCmin value with the histological grade and the aggressiveness of breast IDC.

Materials and methods

Study patients

In total, 281 patients with suspected with breast cancer were examined by breast MRI at the Second Hospital of Shandong University, (Jinan, Shandong, China) between May 2013 and July 2016 were reviewed. Among these cases, 152 cases were excluded due to receiving chemotherapy or radiotherapy (n=25), lack of surgical confirmation (n=26), benign lesions (n=76) or suffering from other types of malignant tumors (n=25). The remaining 129 patients with pathologically-diagnosed invasive ductal carcinoma according to the World Health Organization classification of tumors of the breast (5), who were all females, aged between 27 to 72 years old (median age of 48 years), with a mean age of 47.42±10.26 years, were included into the analysis. Following MRI examination, all lesions were pathologically verified by lumpectomy, mastectomy or biopsy at the Institute of Pathology at the Second Hospital of Shandong University (Jinan, China). In order to avoid misdiagnosis caused by tumor heterogeneity, the biopsy was performed under the guidance of MRI inspection. Prior written informed consent was obtained from each patient, and the study was approved by the ethics committee of the Second Hospital of Shandong University.

MRI examination

All breast MRI examinations were performed on a 3.0-T system (GE Discovery MR750; GE Healthcare, Milwaukee, WI, USA), with an eight-channel dedicated breast coil. Patients were in the prone position, and both breasts were imaged simultaneously. For premenopausal females, imaging was performed between day 7 and day 14 of the menstrual cycle. An MRI plain scan was performed with the axial T2-weighted single-shot fast spin echo sequence using a modified Dixon technique (T2 IDEAL) for intravoxel fat-water separation, using the following parameters: Repetition time (TR), 2,500 msec; echo time (TE), 53.5 msec; slice thickness, 6 mm; layer spacing, 1.0 mm; field of view, 360×360 mm; matrix size, 320×192 pixels; and number of excitations (NEX), 3. Similarly, the parameters for the axial T1-weighted fast spin-echo were as follows: TR, 569 msec; TE, 15.6 msec; slice thickness, 6 mm; layer spacing, 1.0 mm; field of view, 360×360 mm; matrix size, 256×192; NEX, 4.

In order to obtain dynamic contrast-enhanced (DCE) MRI scans, a dynamic examination was performed using the axial T1-weighted 3D dynamic gradient echo fat sequence. The parameters for this examination were as follows: TR/TE, 3.9/1.7; flip angle, 5°; field of view, 360×360 mm; matrix size, 348×348; and slice thickness, 1.8 mm. Subsequent to unenhanced acquisition, Gadodiamide (0.2 mmol/kg body weight, GE Healthcare Life Sciences, Little Chalfont, UK) was intravenously injected at the rate of 2 ml/sec, followed by 20 ml saline flush. DCE image acquisition was initiated immediately after the saline injection. The sequence was repeated for seven times without time gaps, and each sequence lasted for 60 sec.

DWI was performed with an axial single-shot fat suppressed echo-planar diffusion weighted sequence (TR, 3,000 msec; TE, 49.5 msec; slice thickness, 6 mm; layer spacing, 1.0 mm; field of view, 360×360 mm; matrix size, 128×96; and NEX, 4). The diffusion-sensitizing gradient was applied along the x, y and z axes, while b-values of 0 and 800 sec/mm2 were used.

Image analysis

MRI scans were independently reviewed by two experienced radiologists at a workstation (Advantage Windows Workstation 4.6; GE Healthcare) in a blinded manner. For each case, the final decision was made only upon agreement between these two radiologists. ADC measurement was performed in the GE workstation software. A ROI with a mean size of 25.7 mm2 (ranging between 8.0 and 79.0 mm2) was placed on the highest-signal focal in the DWI images that corresponded to the lowest-signal area in the ADC maps. Subsequently, the ADC values were automatically calculated on the ADC maps. The ROI was smaller than the lesion size and was placed in the solid part of IDC, avoiding the necrotic and hemorrhagic regions. The ROI size of each lesion was consistent for multiple measurements, in which the lowest of three measurements was accepted as the minimum ADC (ADCmin) value.

Histopathological analysis

The histological grade of the tumors was assessed using the Nottingham modification of the Bloom-Richardson system (22), considering the following three parameters: i) Tubular formation (1 point, tubular formation in >75% of the tumor; 2 points, tubular formation in 10–75%; and 3 points, tubular formation in <10%); ii) nuclear pleomorphism (1 point, nuclei with minimal variation in size and shape; 2 points, moderate nuclear variation; and 3 points, marked nuclear variation); and iii) mitotic count (1 point, 0–11 mitotic counts; 2 points, 12–22 mitotic counts; and 3 points, >23 mitotic counts), calculated using a light microscope (BX43; Olympus Corporation, Tokyo, Japan) at magnification of ×40 with a field diameter of 0.63 mm and a field area of 0.312 mm2. The final decision on the histological grade was established only upon agreement of the investigators (two pathologists). Scores of 3–5, 6–7 and 8–9 were considered to indicate histological grades I, II and III, respectively (22).

Statistical analysis

Data are expressed as the mean ± standard deviation. SPSS software (version 18.0; SPSS, Inc., Chicago, IL, USA) was used for statistical analysis. Due to the fact that the minimum ADCmin or maximum ADCmin represented a specific value for each patient, no statistical analysis could be performed on these data. In the present study, the mean ADCmin value was calculated by averaging the values of all the patients. The mean ADCmin value difference was compared with the histological grade using the one-way analysis of variance and least significant difference test. Independent sample (Student's t-test) and receiver operating characteristic (ROC) curve analysis were used to analyze the diagnostic value of the mean ADCmin value in differentiating less and more aggressive IDC cases. The optimal cutoff point of the mean ADCmin value, and the corresponding sensitivity and specificity were determined. P<0.05 was considered to indicate differences that were statistically significant.

Results

Histological grade and ADCmin values in IDC patients

Among the 129 breast IDC cases included in the present study, pathological analysis revealed that there were 17 (13.18%) cases of histological grade I, 79 (61.24%) cases of histological grade II, and 33 (25.58%) cases of histological grade III (Figs. 1–3). Furthermore, the mean ADCmin value of all IDC cases was (0.97±0.15)×10−3 mm2/sec. The minimum, maximum and mean ADCmin values of IDC obtained by histological grade are presented in Table I.

Figure 1. Representative imaging examination of histological grade I breast invasive ductal carcinoma in a 44-year-old female. (A) Axial contrast-enhanced image demonstrating the 1.2-cm round mass with heterogeneous enhancement in the lower-outer quadrant of the right breast (arrow). (B) Axial ADC map of the breast demonstrating a low signal with a low ADCmin value (1.20×10−3 mm2/sec) (arrow). ADCmin, minimum apparent diffusion coefficient.

Figure 3. Representative imaging examination of histological grade III breast invasive ductal carcinoma in a 53-year-old female. (A) Axial contrast-enhanced image demonstrating the 1.6-cm irregular mass with heterogeneous enhancement in the central area of the left breast (arrow). (B) Axial ADC map of the breast demonstrating a low signal with a low ADCmin value (0.78×10−3 mm2/sec) (arrow). ADCmin, minimum apparent diffusion coefficient.

Table I. Range and mean values of ADCmin of IDC according to the different histological grades.

Table I.

Range and mean values of ADCmin of IDC according to the different histological grades.

Grade	n	ADCmin range (×10−3 mm2/sec)	Mean ADCmin (×10−3 mm2/sec)	P-value
I	17	0.95–1.34	1.14±0.11	<0.01
II	79	0.72–1.30	0.99±0.12	<0.01
III	33	0.50–1.10	0.86±0.13	<0.01

[i] P-values were obtained from comparison with the mean ADC_min groups. ADC_min, minimum apparent diffusion coefficient.

When considering the different histological grades of the included IDC patients, the mean ADCmin values were (1.14±0.11)×10−3, (0.99±0.12)×10−3 and (0.86±0.13)×10−3 mm2/sec for patients with grade I, II and III disease, respectively. The corresponding ranges of these values were (0.95–1.34)×10−3, (0.72–1.30)×10−3 and (0.50–1.10)×10−3 mm2/sec, respectively. These measurement results revealed that different ADCmin values corresponded to IDC cases with different histological grades. Compared with cases of higher grades, tumors of lower grades exhibited significantly higher ADC values (F=33.49; P<0.01). In particular, there was a significant difference in the ADCmin value between grade II and III tumors (P<0.01), as well as between grade I and II tumors (P<0.01; Fig. 4). Taken together, these results suggest that, the ADCmin value is inversely correlated with the histological grade of IDC.

Figure 4. Box plot demonstrating the correlation between the ADCmin value and the histological grade of IDC. Mean ADCmin values for IDC cases of histological grades I, II and III were (1.14±0.11)×10−3, (0.99±0.12)×10−3 and (0.86±0.13)×10−3 mm2/sec, respectively (P<0.01). ADCmin, minimum apparent diffusion coefficient; IDC, invasive ductal carcinoma.

ADCmin value and disease aggressiveness in IDC patients

To evaluate the role of the ADCmin value in determining the aggressiveness of IDC, the cases were divided into the less aggressive (grades I and II) and more aggressive (grade III) groups. The results revealed that the mean ADCmin values in the less aggressive group was (1.01±0.13)×10−3 mm2/sec, while this value was (0.86±0.13)×10−3 mm2/sec in the more aggressive group (t=5.76, P<0.01). These results suggest that the ADCmin value is inversely correlated with the aggressiveness of IDC.

ROC analysis of ADCmin value in IDC diagnosis

ROC analysis demonstrated that the ADCmin value was a significant parameter in the diagnosis of less aggressive IDC, with an area under the curve (AUC) of 0.81. The ADCmin threshold value of 0.90×10−3 mm2/sec corresponded to a sensitivity of 86.5% and a specificity of 72.7% in the detection of less aggressive tumors (Fig. 5). These results suggest the importance of the ADCmin value in diagnosing less aggressive IDC.

Figure 5. Receiver operating characteristic curve analysis of ADCmin values in the diagnosis of IDC. The ADCmin value exhibited great significance in the diagnosis of less aggressive IDC cases, with an area under the curve of 0.81. The ADCmin threshold value of 0.90×10−3 mm2/sec corresponded to a sensitivity of 86.5% and a specificity of 72.7% for the detection of less aggressive tumors. ADCmin, minimum apparent diffusion coefficient; IDC, invasive ductal carcinoma.

Discussion

The histological grade represents an important prognostic factor for tumors in clinical practice, which is helpful in evaluating the tumor behavior (2). In addition, DWI is quick examination that does not require the use of a contrast agent and is quantified by the apparent diffusion coefficient. Moreever, the ADC value is quantitative and, therefore, an objective calculation (8,23). Previous studies have examined the association between the ADC value and tumor grade. For instance, certain studies have demonstrated that the ADC value is inversely correlated with the tumor grade (13–15). Cipolla et al (16) also revealed that the ADC values were significantly higher in G1 tumors as compared with G3 tumors, while there was no statistically significant difference upon comparison of G1 and G3 tumors with G2 tumors. By contrast, several other studies have indicated that no correlation exists between the ADC value and tumor grade (17–20).

In the present study, the results demonstrated that the mean ADCmin values for IDC of grades I, II and III were (1.14±0.11)×10−3, (0.99±0.12)×10−3 and (0.86±0.13)×10−3 mm2/sec, respectively, with the corresponding ranges of (0.95–1.34)×10−3, (0.72–1.30)×10−3 and (0.50–1.10)×10−3 mm2/sec. Tumors of lower grades were observed to exhibit significantly higher ADC values compared with tumors of higher grades. In addition, there was significant difference in the ADCmin value between tumors of grades II and III, as well as between tumors of grades I and II. This phenomenon may be attributed to the higher cellular density and smaller extracellular space in IDC of higher grades, which results in lower ADC values (24). Additionally, the overlapping of the ADCmin for the three grades of IDC is possibly due to tumor heterogeneity (1).

The majority of previous studies (13–20) have used the mean ADC value when investigating its correlation with the tumor grade. However, in the present study, the ADCmin values were used instead, and significant differences were observed in these values between different tumor grades. IDC is a heterogeneous tumor consisting of invasive cancer nests, fibrosis and necrosis. In MRI scans, the regions of the maximum ADC (ADCmax) values reflect the lowest cellular zone, while the regions of the ADCmin values reflect the highest cellular zone composed of stroma (21). Furthermore, the presence of fibrosis and necrosis may affect the ADC values, particularly the ADCmax values. Therefore, there will always be differences in the ADCmax values in the local measurement and pathological characterization of tumors. In the present study, the ADCmin value was used for tumor pathological characterization, which has been suggested to reflect the most malignant portions of tumors (21).

There have been several studies investigating the application of the ADCmin values in differentiating the benign and malignant breast masses (as well as in differentiating breast cancer subtypes), and in detecting the invasive component in ductal carcinoma in situ (21,25,26). Byun et al (27) have measured the mean ADC and the mean ADCmin values in the regions with the highest fluorodeoxyglucose (FDG) uptake using sequential 18F-FDG positron-emission tomography and MRI, and examined the correlation of the corresponding ADC values with the histological grade of IDC. The majority of these aforementioned studies have applied the multiple ROI method in breast MRI, and the ADCmin value represented the lowest mean ADC value among multiple small ROIs within the lesion. However, this method is time-consuming, limiting its clinical application (21,25,27). In the present study, the single ROI was smaller in size when compared with the lesion size, and was placed in the solid part of the IDC, avoiding the necrotic and hemorrhagic regions. The ROI size of each lesion was consistent for multiple measurements, and the lowest ADC value among three measurements was determined as the ADCmin for each lesion. This method has been previously reported, with considerable feasibility in clinical settings (26). Furthermore, the results of the current study were consistent with previous findings (13–15), with a simpler and more feasible method used, and confirmed that the ADCmin value was inversely correlated with the histological grade of breast IDC. Further studies with larger sample sizes are required to evaluate the clinical application of the ADCmin value in tumor grading.

Biological evaluation of tumors is important for the selection of treatment options. Different tumor cell densities may indicate different histological structures and biological invasions. The ADC values for IDC lesions were lower as compared with those of other malignant tumors, which may be due to the densely packed tumor cells, restricting effective motion and diffusion of water molecules (28). Costantini et al (13) have identified that the mean ADC value for IDC was 1.03×10−3 mm2/sec, while the mean ADC value for ductal carcinoma in situ was 1.05×10−3 mm2/sec. In the present study, the mean ADCmin value for IDC was (0.97±0.15)×10−3 mm2/sec, while the values for the less aggressive (grades I and II) and more aggressive (grade III) tumor groups were (1.01±0.13)×10−3 and (0.86±0.13)×10−3 mm2/sec, respectively. Thus, a significant difference was observed in the mean ADCmin value between the less and more aggressive IDC groups, which was in line with previous observations (13). Furthermore, a cutoff point for the ADCmin threshold value of 0.90×10−3 mm2/sec was used to detect the less aggressive tumors in the present study, corresponding to a sensitivity of 86.5% and a specificity of 72.7%. The results of the ROC curve analysis further revealed that the ADCmin value was important in the diagnosis of less aggressive IDC cases, with an AUC of 0.81.

There are also certain limitations in the present study. Firstly, the sample size was relatively small; therefore, the findings need to be validated in studies using larger sample sizes. In addition, all lesions herein were IDCs, appearing as a mass-like enhancement. Therefore, it was not able to evaluate the ADCmin value of other types of breast carcinoma appearing as non mass-like enhancement with respect to the pathological grade. Furthermore, the measurement of ADCmin value was relatively subjective, which may lead to observational bias. Another limitation is that the association between the histological grade and the morphological alterations, as well as the enhancement pattern, were not examined in the present study. Routine imaging findings combined with the ADC value would improve the accuracy of the preoperative assessment of the histological grade for IDC, which would assist in the selection of the appropriate treatment options for breast cancer. Additionally, the association between the ADCmin values and other pathological characteristics was not investigated herein, which should be considered in further in-depth studies in the future. According to a previous study (29) and clinical practice, only two b-values (0 and 800 mm2/sec) were applied in the present study; thus, DWI images with more b values would be required in the future to obtain accurate ADC values. Finally, the present study was a retrospective study, and therefore, the ADCmin value was not considered in the clinical decisions.

In conclusion, the results of the current study indicated that the ADCmin value was correlated with the histological grade of IDC. Lower ADCmin values were associated with higher histological grades. These findings suggest that the ADCmin value may be considered as a promising prognostic parameter in identifying tumor aggressiveness.

Acknowledgements

Not applicable.

Funding

This study was supported by the Key Research and Development Plan of Shandong Province (grant no. 2015GSF119034) and the Seed Fund of the Second Hospital of Shandong University (grand no. S2014010020).

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

GS and WG designed the study. SZ, RT, PC, ZL, and FS performed the experiments. SZ analysed the data.

Ethics approval and consent to participate

The present study was approved by the ethics committee of the Second Hospital of Shandong University.

Consent for publication

Prior written informed consent was obtained from each patient.

Competing interests

The authors declare that they have no competing interests.

References