Introduction

Obtaining biopsy specimens from tumors in the brain and deeper layers of organs, and from gynecological tumors such as ovarian cancer, involves invasive techniques. MRI is very useful as a non-invasive cancer diagnosis method. Diffusion MRI is a technique to observe the diffusion of water molecules, which is restricted by the cell membrane and intracellular organelles, and thus reflects the structural information of the cells and surrounding tissues (Fig. 1)^1,2. Since the diffusion distance of water molecules depends on the diffusion time, an appropriate diffusion time must be set for the size of each structure. Accordingly, we utilized a wide range of diffusion times. In addition to middle~longer diffusion times (of 7.5–100 ms) using Pulse Gradient Spin Echo (PGSE), shorter diffusion times (of 1.875–7.5 ms) using Oscillating Gradient Spin Echo (OGSE) were applied³. There are over 100 types (about 40 malignant types) of subcutaneous soft-tissue tumors, a type of peripheral cancer, and tissue diagnosis can be difficult, especially with deeper localization. Diffusion MRI may provide important information to distinguish between cystic tumors with hemorrhage and necrosis, cellularly-rich tumors, fatty tumors, and tumors with mucous humor, based on water diffusion. The frequency of subcutaneous soft tissue tumors is low, complicating systematic human studies. In this study, LLC- and B16-bearing mice were used as subcutaneous tumor models. The former is accompanied with inflammatory changes such as hemorrhage and necrosis, while the latter is rich in cellular components. Based on the Apparent Diffusion Coefficient (ADC) values from the diffusion time measurements, the ability to discriminate between normal and tumor tissues was evaluated. We also analyzed the ability to discriminate between the two cancer cell lines.

Methods

Mice (male C57BL/6, 9–12 weeks old) were inoculated subcutaneously with either LLC or B16 cells into the right dorsal area. When each tumor reached 300–500 mm³, MR imaging was performed. Mice were anesthetized with 2% isoflurane before and during scanning with a 7T MRI (Bruker BioSpec). The imaging parameters for PGSE and OGSE acquisitions were as follows: TR/TE 5000/80 ms; FOV = 20×20 mm²; resolution = 200×200 μm²; slice thickness = 1 mm; number of slices = 10; b value = 100–500 s/mm²; diffusion time = 7.5–100 ms for PGSE, 1.875–7.5 ms for OGSE. The diffusion time in the OGSE method can measure 3–5 μm structures, and that in the PGSE method can measure 5–20 μm structures (Fig. 2). Averaged ADC values, obtained by averaging nine ADC calculation regions (six in tumor tissues and three in normal skin tissues), were set up on the images. Based on the ADC values from the diffusion time measurements, the ability to discriminate between normal and tumor tissues was evaluated. We also analyzed the ability to discriminate between the two cancer cell lines. Since the ADC values comprehensively reflect not only cell density but also cell size and intra- and extracellular structures, we used the IMPULSED model fitting tool (MRI signals are defined as signals arising from the intracellular and extracellular spaces, and cells are modeled as opaque spheres for simplicity) to calculate the parameters of the cell properties⁴. In the future, we will modify the IMPULSED model into a fitting model that can successfully explain our MRI data and improve MRI diagnostic accuracy.

Results and discussion

The relationships between diffusion times and ADC values measured by OGSE and PGSE sequences are graphed as a scatter diagram (Fig. 3). For both tumor-bearing mice (LLC and B16), the ADC values were significantly lower in tumor tissues than in normal skin tissues for lower diffusion times (1.875–6.75 ms, p < 0.001). The water diffusion appears to be more restricted in tumor tissues, resulting in lower ADC values (Fig. 1)⁵. Our MRI data indicated that tumor tissues can be discriminated from normal tissues, based on the ADC values (Fig. 3). The ADC values for LLC mice tended to be smaller than those for B16 mice. The central parts of the tumor sites in LLC mice were inflamed and ulcerated. The chronic inflammatory responses may induce fibrosis of the surrounding tissues, resulting in the diffusion disturbance of water molecules and the ADC reduction⁶. Our MRI data indicated that the two tumor cell lines may be discriminated according to their ADC values.

Conclusions

Based on the ADC values from the diffusion time measurements (OGSE and PGSE) for subcutaneous tumor model mice, tumor tissues can be discriminated from normal tissues. In addition, two types of tumors can be discriminated.

Acknowledgements

No acknowledgement found.

References

1. Le Bihan, D. et al. Radiology 161, 401–407 (1986)

2. Martin, M. Magn. Reson. Insights 6, 59–64 (2013)

3. Wu, D. et al. Magn. Reson. Med. 72, 1366–1374 (2014)

4. Jiang, X. et al. Magn. Reson. Med. 75, 1076–1085 (2016)

5. Shrinkhin, E. G. et al. Int. J. Mol. Sci. 23, 15780 (2022)

6. Kim, S. et al. Nature 457, 102106 (2009)