Summary

METHODS
The study included 209 patients with invasive BC at the time of initial diagnosis. Patients underwent whole-body ¹⁸F-fluorodeoxyglocose positron emission tomography/computed tomography and breast magnetic resonance imaging including diffusion weighted imaging. Histologic grade (HG), histological type, human epidermal growth factor 2 (HER-2), Ki-67, estrogen receptor (ER), and progesterone receptor (PR) markers of the breast tumor were evaluated in pathological samples. Tumor-node-metastasis (TNM) staging was performed based on clinical, pathological, and imaging findings.

RESULTS
HER-2 positivity and PR positivity demonstrated a strong correlation with distant metastasis (p=0.00040 ve 0.00045). ER positivity was positively correlated with SUV_max (p=0.0001) and SUV_max/ADC_mean (p=0.006). PR was positively correlated with ADC_mean (0.028). SUV_max was correlated with the tumor size (p=0.008), TNM stage (p=0.022 and r=0.159), and HG (p<0.0001 and r=0.347).

CONCLUSION
Both SUV_max and ADC_mean are helpful parameters in determining prognosis in BC. HER-2 and PR positivity, and tumor size can be used as revealing and useful parameters in determining distant metastases.

Introduction

¹⁸F-FDG positron emission tomography/computed tomography (¹⁸F-FDG-PET/CT) and diffusion-weighted image (DWI) are imaging methods that give indirect information of the biological properties of cancer. Apparent diffusion coefficient (ADC) and maximum standardized uptake value (SUV_max) values measured from breast mass have the potential to be used as prognostic biomarkers.[5-13]

DWI is an magnetic resonance imaging (MRI) technique based on thermal energy-dependent random movements (Brownian motion) of water molecules in biological tissues, and its quantitative parameter is ADC. In high cellular malignant tissue, low ADC values are expected due to restricted fluid diffusion in the relatively decreased extracellular space and an increase in the nucleus/cytoplasm ratio.[11,12]

¹⁸F-FDG-PET/CT is a widely used diagnostic method in the diagnosis, systemic staging, detecting recurrence and evaluation of response to treatment, as well as distinguishing malignant from benign lesions, which enables image acquisition by using high glucose metabolism in cancer cells and therefore increased FDG uptake.[4-7] FDG uptake is quantified as the SUV max, and this numerical value is generally associated with the biological aggressiveness of the tumor. Although PET/CT has a high sensitivity ranged from 81 to 99 % in tumors above 2 cm in initial staging, studies have shown that diagnostic accuracy is quite limited in small tumors (<1 cm) which results in a false negative PET/CT.[7,14,15]

Many biological factors, including molecular subtypes, histologic grade (HG), estrogen receptor (ER), progesterone receptor (PR), human epidermal growth factor receptor 2 (HER-2), and Ki-67 and also age and axillary lymph node (LN) involvement, have been used to determine the prognosis.[16]

Studies are showing that ADC[11,12,16] and SUV_max [7-9] or both of these values[5,13,14] obtained from the primary tumor in BC have a significant relationship with pathological prognostic factors. In our study, we investigate the relationship between SUV_max, ADC values and prognostic factors and also the contribution of these parameters of the primary tumor to the prediction of distant metastasis in BC patients.

Methods

¹⁸F-FDG-PET/CT
All the patients were imaged using an FDG-PET/CT scanner with 16-multi-detector CT (mCT 20 ultra HD LSO PET/CT, (Siemens molecular imaging, Hoffmann Estates, Illinois, USA).

Patients were fasted for at least 6 h before the PETCT procedure and all the patients" blood glucose levels measured below 150 mg/dL. All patients were administered intravenously with ¹⁸F-FDG radiopharmaceutical, calculated from 0.15 mCi/kg based on their body weight. Following the injection, the patients were rested in the half-lying position for 50-60 min in a silent room. At the end of the rest period, combined image acquisition began unenhanced CT scan (3.5 mm slice thickness, 120 kV tube and up to 80 mA s) and subsequent 3D mode PET scan (5-7 bed positions, 3 min per bed position) between vertex and upper femur at the supine position.

1.5-Tesla Breast MRI
The MRI were acquired in the prone position, using a 1.5-Tesla scanner (Signa HDi; GE Healthcare, Milwaukee, WI) with a dedicated bilateral breast phased-array coil. Before the examination, a catheter was placed through the antecubital vein to the patients. Standard protocol with contrast and DWI sequence was used in all examinations.

MR imaging protocol included axial, coronal, and sagittal turbo spin-echo T2 weighted (3D) sequence with 2 mm slice thickness and 1 mm slice spacing, axial fat-suppressed T2 sequence with slice thickness 2 mm and slice thickness 3 mm (fat suppression technique was the short-tau inversion recovery), axial DWI sequence with single-shot EPI; b value = 0 and 800 s/mm² with 3 mm slice thickness and slice spacing; after pre-contrast T1-weighted axial 3D dynamic gradient echo fat suppressed images, a bolus of 0.1 mmol/L per kilogram of body weight contrast agent gadoterate meglumine and 20 mL saline injection was administered with an automated contrast injector. In the post-contrast phase, six phased axial T1-weighted 3D dynamic gradient-echo fat-suppressed consecutive serial images were obtained with a maximum of 60-second intervals. Post-processing, maximum-intensity projection, subtraction, and ADC maps were obtained. All lesions were seen and evaluated at MRI.

Imaging Analysis
All ¹⁸F-FDG-PET/CT images were analyzed by two nuclear medicine physicians with 5 and 10 years of experience in PET/CT. The readers were blinded to the histopathologic diagnosis and had any knowledge of other quantitative imaging data. The volume of interest (VOI) was determined as the area where FDG uptake was most intense on the relevant primary breast tumor, and FDG uptake was semiquantitatively analyzed in this area. SUV_max was considered the voxel with the highest SUV in the VOI examined and was used to measure and record uptake.

Two radiology physicians with 5 and 12 years of experience in breast MRI evaluated ADC maps retrospectively. The readers had no knowledge of histopathologic diagnosis and quantitative imaging data. On the ADC map, multiple uniform circular 20 mm2 region of interests placed within the primary breast tumor. ADC measurements were made only from a solid portion of the tumor and dynamic contrast-enhanced images were used as a reference to avoid measuring from cystic, hemorrhagic, or necrotic areas. The average of the ADC values was calculated and noted as ADC_mean.

Figures 1 and 2 show the symbolic images of the ADC and the SUV_max measurement.

Fig. 1. 47 years old woman with Luminal B type invasive mucinous breast cancer, histologic grade 3, Immunohistochemical staining revealed ER+, PR+, HER-2 -, and Ki-67 %65, stage 4 with bone metastases. On ¹⁸F-FDG-PET/CT images, a primary tumor (SUV_max:11.9) of approximately 5 cm in diameter is observed in the upper outer quadrant of the left breast (first row). Multiple metastatic lymph nodes (SUV_max:11.6) are seen on axillary sections (second row). There is an osteolytic metastatic lesion (SUV_max:12.3) in the L3 vertebral body (third row).
ER: Estrogen receptor; PR: Progesterone receptor; HER-2: Human epidermal growth factor receptor 2; ¹⁸F-FDG-PET/CT: ¹⁸F-FDG positron emission tomography/computed tomography; SUV: Standardized uptake value.

Fig. 2. Same patient. (a) Post-contrast T1 weighted axial image shows high intensity left upper outer quadrant lesion. (b) Post-contrast subtraction image. (c) Axial diffusion-weighted image with b value of 800 shows restricted diffusion in a 5 cm mass. (d) ADC map shows restricted diffusion (ADC_mean: 830×10^-6 mm²/s).
ADC: Apparent diffusion coefficient.

Histological Evaluation
In all patients, first the histological grade (HG) and type of the tumor were determined in the pathology specimens obtained by core needle biopsy or surgery. Subtyping was made according to the markers ER, PR, HER-2, and Ki-67 with immunohistochemical tests. The ER and PR results were determined to be positive according to the proportion of positively stained cell nuclei was higher than 10% and negative when less than 10%. HER-2 expression was determined with fluorescence in situ hybridization. Score +2 and +3 are defined as positive. If the Ki-67 proliferation index was over 15%, it was considered positive and below 15% was considered negative. LN status was evaluated mainly with the imaging techniques and also fine-needle aspiration and sentinel LN biopsy. TNM staging of all the patients was evaluated with initial imaging, biopsy, and pre-treatment clinical staging.

Statistical Analysis
Statistical analyzes were performed using IBM SPSS Statistics 25.0 (Armonk, NY: IBM Corp.) package program. Normality assumption of quantitative data was checked by Shapiro-Wilk test. Since the normality assumption was not provided, the Mann-Whitney U test and the Kruskal-Wallis test were used for assessing the relationship between SUV_max, ADC_mean, SUV_max/ADC values, and the clinicopathological parameters. Dunn test for pairwise comparison and Bonferroni correction was applied to the results. Correlation of quantitative data with each other (SUV_max, ADC, SUV_max/ ADC) or with HG and clinical stage was evaluated with Spearman"s Rho correlation coefficient. Logistic regression analysis was used to develop a model that can predict the presence of distant metastasis in the patient with prognostic factors and imaging data, variables with p<0.10 in univariate logistic regression analysis and variables considered to be clinically significant were evaluated by multiple logistic regression analysis.

Results

Table 1 Patient characteristics

Comparison of Groups According to the Presence of Distant Metastasis
In the comparison of metastatic and non-metastatic groups, clinicopathologic factors and quantitative imaging data were compared in univariate analysis, and those with p<0.1 were evaluated by logistic regression tests in multivariate analysis. Tumor size, HER-2 overexpression, and PR positivity were determined as independent variables affecting the presence of metastasis in multivariate analysis (Table 2).

Table 2 Multiple logistic regression analysis for distant metastasis

Distant metastasis was present in 83 (39.7%) of the patients. There was no significant difference in mean age, histopathological diagnosis, HG, ADC_mean, SUV_max, SUV_max/ADC_mean, and molecular subtypes between metastatic and non-metastatic groups (p>0.05). The distribution of ADC_mean, SUV_max ve SUV_max/ADC_mean values of the metastatic and non-metastatic groups did not differ significantly (p>0.05) (Table 3).

Table 3 Comparison of the ADC_mean, SUV_max and SUVmax/ADC_mean depending on distant metastasis

Univariate analysis showed that a large tumor size (>2 cm) was significantly correlated with the presence of distant metastasis (p<0.005). Immunohistochemistry receptor positivity was present for HER-2, ER, and PR and p values between metastatic and nonmetastatic groups were p=0.0004, 0.803, and 0.00045, respectively. HER-2 positivity and PR positivity demonstrated a strong correlation with distant metastasis. There was no difference for molecular subtypes (luminal A, luminal B, triple-negative, and HER2) and a high Ki-67 index between metastatic and nonmetastatic groups (Table 4).

Table 4 Comparison of the prognostic factors depending on the distant metastasis

Correlation of SUV_max and ADC_mean
The mean SUV_max, ADC_mean, and SUV_max/ADC_mean values were 12.16±8.54 (range, 1.6-52.5), 962±206×10^-6 mm²/s (range, 464-1980×10^-6), and 0.13±0.009, respectively. There was no correlation between ADC_mean and SUV_max (correlation coefficient r=-0.017, p=0.805).

Relationships between Prognostic Factors, SUV_max, ADC_mean and SUV_max/ADC_mean
Univariate analysis showed that SUV_max and SUV_max/ ADC_mean were significantly correlated with large tumor size (p=0.008 and p=0.002), Ki67 status (p<0.0001, p<0.0001). In terms of hormone receptor status; ER positivity was positively correlated with high SUV_max (p=0.0001) and SUV_max/ADC_mean (p=0.006). PR was positively correlated with ADC_mean (0.028). HER-2 positivity had no significant correlation (p>0.1) with SUV_max and ADC_mean (Table 5).

Table 5 Associations of SUV_max, ADC_mean, and SUV_max/ADC_mean with clinicopathologic prognostic factors

The relationships between clinical stage and HG, ADC_mean, SUV_max, and SUV_max/ADC_mean were evaluated with Spearman's Rho coefficient. SUV_max and SUV_max/ADC_mean had a positive significant association with the clinical stage (p=0.022 and r=0.159) and HG (p<0.0001 and r=0.347) (Fig. 3). However, ADC_mean had no significant correlation with clinical stage and HG. When histopathological subtypes were evaluated, IDC had higher SUV_max and SUV_max/ ADC_mean values compared to invasive lobular carcinoma (ILC) (p=0.048).

Fig. 3. Histogram plot demonstrated the relationships between HG, ADC_mean, SUV_max, and SUV_max/ ADCmean. In Spearman's Rho coefficient, SUV_max (a) and SUV_max/ADC_mean (b) had a positive significant correlation with HG (p<0.0001 and r=0.347), however, ADC_mean had no significant correlation (p=0.719) with clinical stage and HG (c).
HG: Histologic grade; ADC: Apparent diffusion coefficient; SUV: Standardized uptake value.

Discussion

Relationships between SUV_max and ADC_mean
Our study with 209 invasive BC showed no correlation between SUV_max and ADC_mean. Similar to our study, many of the studies evaluating this relationship were also found no correlation between ADC_mean and SUV_max. [14,17,18] In contrast, few previous studies[5,13] showed that SUV_max and ADC_mean were inversely correlated. Among these studies, only Kitajima et al.[5] had a larger sample size (214 IDC patients) compared to our study and in that study, ADC_mean showed a weak inverse correlation with SUV_max.

In the study of Baba et al.[17] including malignant and benign breast tumors, a linear inverse correlation was found between SUV_max and ADC_mean. However, no correlation was observed when only malignant tumors were considered. Another finding determined by the author is that very high ADC values and high SUV_max do not show an inverse correlation in mucinous tumors and may affect the statistics. In these studies, we thought that differences between study subjects; sample size, histopathological subtypes, the inclusion of benign and in situ cases, may cause differences in the SUV_max and ADC_mean correlation results.

Relationships between SUV_max and Prognostic Factors
In our study, the relationships between SUV_max and ADC_mean values of primary breast tumor and clinicopathological prognostic factors were evaluated. We observed that high SUV_max was correlated with tumor size, ER, Ki-67, high HG, advanced TNM stage, histological subtype, and molecular subtype. These results are similar to previously published studies; SUV_max had a positive correlation with tumor size and high HG.[13,17,18] Contrary to our study, there is also a study showing that there is no relationship between HG and SUV_max.[19]

IDC is the most common type of invasive BC and the second most common tumor type is ILC. We observed a significant difference between SUV_max of the two groups of histologic types in line to previous studies.[20,21]

Higher SUV values were seen in tumors with a triple-negative hormonal profile in the current study, a finding consistent with the previous studies.[7,14,17] The majority of triple-negative tumors (80%) are the intrinsic basal type and this molecular subtype is associated with a poor prognosis due to the lack of hormonal markers used in targeted hormonal therapy.[14,17]

Ki-67 proliferation index is a marker of high mitotic activity and is useful for evaluating the degree of cellularity. Our study confirms a highly significant positive relationship between Ki-67 and enhanced glycolysis as determined by the measure of SUV_max, as observed previously.[14,20] Our study also demonstrated a strong correlation between ER negativity and high SUV_max values. SUV_max showed no significant correlation with HER-2 overexpression and PR positivity, as observed previously.[7,13,14,20,22,23]

Axillary LN positivity was not correlated with SUV_max . This finding was consistent with a previous study[17] but inconsistent with other studies.[13,18] However, our study had a larger sample size of all these studies.

Relationship between ADC and Prognostic Factors
In many studies comparing ADC in malignant and benign lesions of the breast, significantly lower ADC values and diffusion restriction were observed in malignant lesions. In vivo, perfusion is also important factor as microscopic motion that affects ADC. Increasing microvessels due to tumor angiogenesis in malignant lesions may cause an increase in ADC due to the perfusion effect.[24] Studies are showing that ER positivity causes high tumor cellularity and ER positivity causes low ADC values due to decreased intra-tumor perfusion by blocking the angiogenic pathway.[25,26]

In our study, we found a significant association between the low ADC values and ER positivity which is consistent with many of the previous studies.[11,18,24,27-30] We also observed a significant association between low ADC for PR-positive carcinomas as compared to PR-negative cancers (p=0.028). No relationship was found between Ki-67 and ADC values, supporting the similar studies.[25,30] HER-2 overexpression is associated with poor prognosis which is accompanied by angiogenesis and cellularity. Because of the increased cellularity, it can be expected low ADC values in HER-2 positive cases. [31] But in most of the studies investigating the relationship between ADC and HER-2, no correlation was found in line with our study.[11,18,28]

Ipsilateral axillary LN metastasis is the main predictor of long-term survival.[3,4] The presence of LN metastasis is very important in the decision to proceed with conservative therapy and ADC value would help staging of axillary LN non-inasively.[18] However, in the present study, ADC_mean was not correlated with LN metastasis in consistence with the previous studies. [17,18,24] A few studies with small sample sizes reported that a lower ADC_mean was associated with positive LN metastasis.[13,32]

Factors Affecting the Presence of Metastasis
In our study, tumor size, HER-2 overexpression, and PR positivity were determined as independent variables affecting the presence of metastasis in multivariate analysis.

Tumor size is a well-known and important prognostic factor of BC and increasing size of the breast tumor is associated with high metastatic potential and decreased overall survival.[4] Our study confirmed that the risk of distant metastasis is significantly higher in tumors larger than 2 cm, at the initial presentation.

The majority of BC cases are of the luminal-A subtype, which are hormone receptor-positive tumors (ER and/or PR) and these types of tumors are sensitive to hormonal therapies. However, cases with luminal BC constitute the majority of patients with distant metastases at the time of diagnosis and are frequently incurable. Recent studies revealed that invasiveness and metastasis of luminal BC are supported by the two isoforms of PR (PR-A and PR-B), in two different pathways. As a result of these studies, PR-A has the main responsibility in promoting invasiveness and metastasis by suppressing estrogen/ER action. Overexpression of PR-A is associated with increased invasiveness of BC and decreased disease-free survival.[33,34] In our study, a highly significant positive correlation was found between PR positivity in the breast tumor and the presence of distant metastasis at the time of diagnosis.

HER-2 oncoprotein is responsible for cancer development by stimulating cell proliferation. HER-2 also increased angiogenesis through up-regulation of vascular endothelial growth factor (VEGF) and increasing microvascular density of the tumor accompanied by increasing invasion and metastasis.[17,18,30] Our study showed that HER-2 overexpression of BC seen in 49 patients was positively and strongly associated with distant metastasis (Table 4). HG has been accepted as a prognostic factor for metastasis in the previous studies depending on tumor size.[4] In this study, HG was included in the multiple logistic regression analysis since p<0.1 in univariate analysis, but it was not an independent risk factor for the presence of distant metastases.

SUV_max/ADC_mean is a combination of these parameters and it was more accurate than either SUV_max and ADC_mean for demonstrating the relationships with prognostic factors. SUV_max/ADC_mean was previously used in the study of Baba et al.[17] and was found useful in differentiating benign from malignant breast tumors. In our study, this ratio did not differ significantly between patients with and without distant metastases (Table 3).

This study has some limitations. First, it was performed retrospectively in a single institution. Second, metastases were evaluated only at the time of diagnosis, and metastases developed during treatment and maintenance were not evaluated due to the short follow-up period.

Third, since FDG-PET/CT is only applied to advanced- stage BC patients in our center, the number of early-stage patients was relatively low.

Conclusion

Peer-review: Externally peer-reviewed.

Conflict of Interest: All authors declared no conflict of interest.

Ethics Committee Approval: The study was approved by the University of Health Sciences İstanbul Training and Research Hospital Clinical Research Ethics Committee (no: 2279, date: 08/05/2020).

Financial Support: None declared.

Authorship contributions: Concept - K.E.Ç., M.A.N.; Design - K.E.Ç., M.A.N.; Supervision - M.B.U., E.A.; Funding - K.E.Ç., M.A.N.; Materials - K.E.Ç., M.A.N.; Data collection and/or processing - K.E.Ç., M.A.N.; Data analysis and/ or interpretation - K.E.Ç., M.A.N., M.B.U.; Literature search - K.E.Ç., M.A.N., M.B.U., E.A.; Writing - K.E.Ç., M.B.U.; Critical review - E.A.

References

1) Feng Y, Spezia M, Huang S, Yuan C, Zeng Z, Zhang L, et al. Breast cancer development and progression: Risk factors, cancer stem cells, signaling pathways, genomics, and molecular pathogenesis. Genes Dis 2018;5(2):77-106.

2) Bitencourt AGV, Andrade WP, Cunha RR da, Conrado JLF de A, Lima ENP, Barbosa PNVP, et al. Detection of distant metastases in patients with locally advanced breast cancer: role of 18F-fluorodeoxyglucose positron emission tomography/computed tomography and conventional imaging with computed tomography scans. Radiol Bras 2017;50(4):211.

3) Redig AJ, McAllister SS. Breast cancer as a systemic disease: a view of metastasis. J Intern Med 2013;274(2):113-26.

4) Weigelt B, Peterse JL, van "t Veer LJ. Breast cancer metastasis: markers and models. Nat Rev Cancer 2005;5(8):591-602.

5) Kitajima K, Yamano T, Fukushima K, Miyoshi Y, Hirota S, Kawanaka Y, et al. Correlation of the SUVmax of FDG-PET and ADC values of diffusion-weighted MR imaging with pathologic prognostic factors in breast carcinoma. Eur J Radiol 2016;85(5):943-9.

6) Gallivanone F, Panzeri MM, Canevari C, Losio C, Gianolli L, De Cobelli F, et al. Biomarkers from in vivo molecular imaging of breast cancer: pretreatment 18F-FDG PET predicts patient prognosis, and pretreatment DWI-MR predicts response to neoadjuvant chemotherapy. MAGMA 2017;30(4):359-73.

7) Groheux D, Giacchetti S, Moretti JL, Porcher R, Espié M, Lehmann-Che J, et al. Correlation of high 18FFDG uptake to clinical, pathological and biological prognostic factors in breast cancer. Eur J Nucl Med Mol Imaging 2011;38(3):426-35.

8) Wang CL, MacDonald LR, Rogers JV, Aravkin A, Haseley DR, Beatty JD. Positron emission mammography: correlation of estrogen receptor, progesterone receptor, and human epidermal growth factor receptor 2 status and 18F-FDG. AJR Am J Roentgenol 2011;197(2):W247-55.

9) Koolen BB, Vrancken Peeters MJ, Wesseling J, Lips EH, Vogel WV, Aukema TS, et al. Association of primary tumour FDG uptake with clinical, histopathological and molecular characteristics in breast cancer patients scheduled for neoadjuvant chemotherapy. Eur J Nucl Med Mol Imaging 2012;39(12):1830-8.

10) Song SE, Shin SU, Moon HG, Ryu HS, Kim K, Moon WK. MR imaging features associated with distant metastasis-free survival of patients with invasive breast cancer: a case-control study. Breast Cancer Res Treat 2017;162(3):559-69.

11) Kamitani T, Matsuo Y, Yabuuchi H, Fujita N, Nagao M, Jinnouchi M, et al. Correlations between apparent diffusion coefficient values and prognostic factors of breast cancer. Magn Reson Med Sci 2013;12(3):193-9.

12) Cipolla V, Santucci D, Guerrieri D, Drudi FM, Meggiorini ML, de Felice C. Correlation between 3T apparent diffusion coefficient values and grading of invasive breast carcinoma. Eur J Radiol 2014;83(12):2144-50.

13) Nakajo M, Kajiya Y, Kaneko T, Kaneko Y, Takasaki T, Tani A, et al. FDG PET/CT and diffusion-weighted imaging for breast cancer: prognostic value of maximum standardized uptake values and apparent diffusion coefficient values of the primary lesion. Eur J Nucl Med Mol Imaging 2010;37(11):2011-20.

14) Miyake KK, Nakamoto Y, Kanao S, Tanaka S, Sugie T, Mikami Y, et al. Journal Club: Diagnostic value of (18) F-FDG PET/CT and MRI in predicting the clinicopathologic subtypes of invasive breast cancer. AJR Am J Roentgenol 2014;203(2):272-9.

15) Kitajima K, Miyoshi Y. Present and future role of FDG-PET/CT imaging in the management of breast cancer. Jpn J Radiol 2016;34(3):167-80.

16) Boria F, Tagliati C, Baldassarre S, Ercolani P, Marconi E, Simonetti BF, et al. Morphological MR features and quantitative ADC evaluation in invasive breast cancer: Correlation with prognostic factors. Clin Imaging 2018;50:141-6.

17) Baba S, Isoda T, Maruoka Y, Kitamura Y, Sasaki M, Yoshida T, et al. Diagnostic and prognostic value of pretreatment SUV in 18F-FDG/PET in breast cancer: comparison with apparent diffusion coefficient from diffusion-weighted MR imaging. J Nucl Med 2014;55(5):736-42.

18) Choi BB, Kim SH, Kang BJ, Lee JH, Song BJ, Jeong SH, et al. Diffusion-weighted imaging and FDG PET/CT: predicting the prognoses with apparent diffusion coefficient values and maximum standardized uptake values in patients with invasive ductal carcinoma. World J Surg Oncol 2012;10:126.

19) Choi JH, Lim I, Noh WC, Kim HA, Seong MK, Jang S, et al. Prediction of tumor differentiation using sequential PET/CT and MRI in patients with breast cancer. Ann Nucl Med 2018;32(6):389-97.

20) Gil-Rendo A, Martínez-Regueira F, Zornoza G, García- Velloso MJ, Beorlegui C, Rodriguez-Spiteri N. Association between [18F]fluorodeoxyglucose uptake and prognostic parameters in breast cancer. Br J Surg 2009;96(2):166-70.

21) Jung NY, Kim SH, Choi BB, Kim SH, Sung MS. Associations between the standardized uptake value of 18FFDG PET/CT and the prognostic factors of invasive lobular carcinoma: in comparison with invasive ductal carcinoma. World J Surg Oncol 2015;13:113

22) Osborne JR, Port E, Gonen M, Doane A, Yeung H, Gerald W, et al. 18F-FDG PET of locally invasive breast cancer and association of estrogen receptor status with standardized uptake value: Microarray and immunohistochemical analysis. J Nucl Med 2010;51(4):543-50.

23) Mavi A, Cermik TF, Urhan M, Puskulcu H, Basu S, Yu JQ, et al. The effects of estrogen, progesterone, and CerbB- 2 receptor states on 18F-FDG uptake of primary breast cancer lesions. J Nucl Med 2007;48(8):1266-72.

24) Choi SY, Chang Y-W, Park HJ, Kim HJ, Hong SS, Seo DY. Correlation of the apparent diffusion coefficiency values on diffusion-weighted imaging with prognostic factors for breast cancer. Br J Radiol 2012;85(1016):e474.

25) Sung HK, Eun SC, Hyeon SK, Bong JK, Jae JC, Ji HJ, et al. Diffusion-weighted imaging of breast cancer: Correlation of the apparent diffusion coefficient value with prognostic factors. J Magn Reson Imaging 2009;30(3):615-20.

26) Ludovini V, Sidoni A, Pistola L, Bellezza G, De Angelis V, Gori S, et al. Evaluation of the prognostic role of vascular endothelial growth factor and microvessel density in stages I and II breast cancer patients. Breast Cancer Res Treat 2003;81(2):159-68.

27) Kim EJ, Kim SH, Park GE, Kang BJ, Song BJ, Kim YJ, et al. Histogram analysis of apparent diffusion coefficient at 3.0t: Correlation with prognostic factors and subtypes of invasive ductal carcinoma. J Magn Reson Imaging 2015;42(6):1666-78.

28) Martincich L, Deantoni V, Bertotto I, Redana S, Kubatzki F, Sarotto I, et al. Correlations between diffusion- weighted imaging and breast cancer biomarkers. Eur Radiol 2012;22(7):1519-28.

29) Henry NL, Hayes DF. Uses and abuses of tumor markers in the diagnosis, monitoring, and treatment of primary and metastatic breast cancer. Oncologist 2006;11(6):541-52.

30) Jeh SK, Kim SH, Kim HS, Kang BJ, Jeong SH, Yim HW, et al. Correlation of the apparent diffusion coefficient value and dynamic magnetic resonance imaging findings with prognostic factors in invasive ductal carcinoma. J Magn Reson Imaging 2011;33(1):102-9.

31) Yoshikawa MI, Ohsumi S, Sugata S, Kataoka M, Takashima S, Mochizuki T, et al. Relation between cancer cellularity and apparent diffusion coefficient values using diffusion-weighted magnetic resonance imaging in breast cancer. Radiat Med - Med Imaging Radiat Oncol 2008 May;26(4):222-6.

32) Razek AAKA, Gaballa G, Denewer A, Nada N. Invasive ductal carcinoma: Correlation of apparent diffusion coefficient value with pathological prognostic factors. NMR Biomed 2010;23(6):619-23.

33) Rosati R, Oppat K, Huang Y, Kim S, Ratnam M. Clinical association of progesterone receptor isoform A with breast cancer metastasis consistent with its unique mechanistic role in preclinical models. BMC Cancer 2020;20(1):512.

34) McFall T, McKnight B, Rosati R, Kim S, Huang Y, Viola- Villegas N, et al. Progesterone receptor A promotes invasiveness and metastasis of luminal breast cancer by suppressing regulation of critical microRNAs by estrogen. J Biol Chem 2018;293(4):1163-77.