4615
Cellular microstructural mapping (cell size imaging) by time-dependent diffusion MRI for prediction of prognostic factors in breast cancer1The First Affiliated Hospital of Zhengzhou University, Zhengzhou, China, 2Philips Healthcare, Beijing, China
Synopsis
Keywords: Breast, Diffusion/other diffusion imaging techniques, time-dependent diffusion weighted imaging
Motivation: Recent advances in time-dependent diffusion MRI (td-dMRI) for microstructural modeling provide the opportunity to characterize cancer pathology in vivo.
Goal(s): This study aims to evaluate the accuracy of td-dMRI–based microstructural mapping for noninvasively characterizing of breast tumors, and further to evaluated whether the tumor microstructural properties could be used to distinguish prognostic factors in breast cancer.
Approach: All patients underwent T1-weighted imaging, T2-weighted imaging, diffusion weighted imaging, td-dMRI, and dynamic contrast enhancement scans on a 3T scanner.
Results: Results showed that td-MRI parameters showed significant differences between benign and malignant breast tumors, and can also be used for prediction of prognostic factors in breast cancer.
Impact: The cellular microstructural mapping by time-dependent diffusion MRI show great potential for noninvasive evaluation of pathologic characteristics in breast cancer in a clinical setting.
Introduction
The biological heterogeneity of breast cancer has been revealed at the genetic level [1]. The estrogen receptor (ER), progesterone receptor (PR), Her-2, and Ki-67 expression status defines breast cancer subtypes and is closely related to tumor cellularity, vascularity, and aggressiveness[2], and they are commonly identifed as vital immunohistochemical (IHC) markers for treatment decisions and prognosis. Recent advances in diffusion MRI-based microstructural modeling provide the opportunity to characterize cancer pathology in vivo. By incorporating the time-dependency of water molecule diffusion with specifc biophysical models, we can estimate important microstructural properties such as cell size, cell volume fraction, and transcytolemmal water exchange etc.[3, 4], which are closely related to the pathological changes of tumor. This study aims to investigate the feasibility of time-dependent diffusion MRI (td-dMRI)-based microstructural mapping for cellular properties to breast cancer, and further to evaluate whether the microstructural properties could be used to distinguish prognostic factors in breast cancer.Materials and Methods
This prospective study collected 200 patients with suspected breast tumors from the First Affiliated Hospital of Zhengzhou University from March to August 2023. All patients underwent T1-weighted imaging, T2-weighted imaging, diffusion weighted imaging, td-dMRI, and dynamic contrast enhancement scans on a 3T scanner (Ingenia Elition, Philips Healthcare, Best, the Netherlands). The td-dMRI scan included acquisitions of diffusion MRI with both oscillating (OGSE) and pulsed (PGSE) gradient encoding using the oscillating frequencies up to 33 Hz. Data were fitted with a two-compartment model (IMPULSE)[5] to estimate mean cell diameter (dmean), intracellular fraction (fin), extracellular difusivity (Dex), and cellularity (fin/d). The apparent diffusion coefficient (ADC) were calculated from conventional diffusion weighted imaging. The receiver operating characteristic (ROC) curve was used to access the diagnostic performance of dmean, fin, Dex, cellularity and ADC values. P < 0.05 indicated that the difference was statistically significant. The independent samples t test was used to compared the dmean, fin, Dex, cellularity, and ADC values between benign and malignant breast tumors, between breast cancer with different histological grading, as well as between breast cancer with positive and negative expression of ER, PR, Her-2, and Ki-67, respectively. The receiver operating characteristic (ROC) curve was used to access the diagnostic performance of dmean, fin, Dex, and ADC values in differentiation between benign and malignant tumors, as well as in recognization of different breast cancer risk factors. P < 0.05 indicated that the difference was statistically significant.Results and Discussion
195 patients (5 cases were excluded due to the bad image quality) with195 lesions were included and divided into malignant (n=115) and benign lesion group (n=80) according to the histopathological results. Figures 1 and 2 show representative images of patients with benign and malignant lesions. The dmean, fin and cellularity values of malignant lesions were significantly higher than those of benign lesions (15.74±2.69 vs. 14.28±4.70 μm, 0.348±0.128 vs. 0.265±0.202, 25.14±11.00 vs. 18.88±14.38 ×10-3 um-1), and the Dex and ADC of malignant lesions were significantly lower than those of benign lesions (2.121±0.398 vs. 2.375±0.327 um2/ms, 0.879±0.171 vs. 1.457±0.353 um2/ms). (Table 1 and Figure 3) For differentiation between benign and malignant breast lesions ADC showed the highest AUC (0.9280) and specificity (90.38) than other parameters, and the fin showed relatively high AUC (0.749) and sensitivity (85.86) than the d, Dex and cellularity. (Figure 4)The td-MRI parametersshowed no significant difference between breast tumors with high and low expression of Ki-67. The finvalues of Her-2(+) tumors were significantly lower than those of Her-2(-) tumors, and the ADCvalues of Her-2(+) tumors were significantly higher than those of Her-2(-) tumors.The finandcellularityvalues of PR(+) tumors were significantly higher than those of PR(-) tumors.The Dexvalues of ER(+) tumors were significantly lower than those of ER(-) tumors. (Table 1) The Dex showed a AUC of 0.650 for differentiation between ER(+) and ER(-). fin and cellularity showed AUCs of 0.709 and 0.662 in differentiation of PR(+) and PR(-), respectively. finandADC showed AUCs of 0.524 and 0.558 in differentiation of Her-2(+) and Her-2(-), respectively.
Conclusion
The td-dMRI–based microstructural mapping demonstrates promise for characterizing breast cancer, and may be helpful for prognostic evaluation.Key words
time-dependent diffusion weighted imaging, breast tumor, benign and malignant, molecular prognostic biomarkerAcknowledgements
No acknowledgment.
References
[1] Perou CM, Sørlie T, Eisen MB et al (2000) Molecular portraits of human breast tumours. Nature 406:747–752
[2] Goldhirsch A, Wood WC, Coates AS, Gelber RD, Thürlimann B, Senn HJ (2011) Strategies for subtypes–dealing with the diversity of breast cancer: highlights of the St. Gallen International Expert Consensus on the Primary Therapy of Early Breast Cancer 2011. Ann Oncol 22:1736–1747
[3] Reynaud O (2017) Time-dependent difusion MRI in cancer: Tissue modeling and applications. Front Phys 5. https://doi.org/10.3389/fphy.2017.00058
[4] Stepišnik J (1993) Time-dependent self-difusion by NMR spinecho. Physica B 183:343–350
[5] Xu J, Jiang X, Li H, et al. Magnetic resonance imaging of mean cell size in human breast tumors[J]. Magnetic resonance in medicine, 2020, 83(6): 2002-2014.
Figures
Table 1 Comparison of td-MRI parameters between benign and malignant breast lesions, and between different subtypes or histological grades of breast cancer
Figure 1a-g Fibroadenoma of the right breast. 1a showed sagittal contrast-enhanced T1WI as reference showing the uneven enhancement in lesion. 1b showed fusion image,shows the lesion's fusion with OGSE. 1c showed diffusion-weighted imaging (DWI) with obvious limited diffusion. 1d -1g including the cell diameter, intracellular fraction (f in), extracellular diffusivity (Dex) ,and cellularity, fitted from the imaging microstructural parameters using limited spectrally edited diffusion.
Figure 2a-g right invasive breast carcinoma. 1a showed sagittal T1-enhanced, shows the lesion's uneven enhancement as reference. 1b showed fusion image,shows the lesion's fusion with OGSE.1c showed diffusion-weighted imaging (DWI) with obvious limited diffusion. 1d -1g including the cell diameter, intracellular fraction (f in), extracellular diffusivity (Dex) ,and cellularity, fitted from the imaging microstructural parameters using limited spectrally edited diffusion.
Figure 3. microstructural parameters by td-dMRI with significant differences between benign and malignant tumors, or between different subtypes of breast cancer.
Figure 4. ROC curves for different parameters (dmeam, fin, Dex, cellularity, ADC) for differentiation between malignant and benign breast lesions (A), or between different subtypes of breast cancer (B. ER; C. PR; and D. Her-2).