3649

Whole-Tumor Histogram Analysis of Monoexponential and Advanced Diffusion-weighted Imaging for Sinonasal Malignant Tumors: Correlations with Histopathologic Features

Zebin Xiao¹, Zuohua Tang², Jing Zhang³, Guang Yang³, Wenjiao Zeng⁴, Jianfeng Luo⁵, Rong Wang², Linying Guo², and Zhongshuai Zhang⁶

¹Eye & ENT Hospital of Fudan University, Shanghai, China, ²Radiology, Eye & ENT Hospital of Fudan University, Shanghai, China, ³Shanghai Key Laboratory of Magnetic Resonance, East China Normal University, Shanghai, China, ⁴Pathology, School of Basic Medical Sciences, Fudan University, Shanghai, China, ⁵Biostatistics, School of Public Health, Fudan University, Shanghai, China, ⁶MR Scientific Marketing, Siemens Healthcare, Shanghai, China

Synopsis

This is the first study with a large sample size to systematically investigate the correlation of monoexponential diffusion-weighted imaging (DWI) and advanced DWI (intravoxel incoherent motion [IVIM] and diffusion kurtosis imaging [DKI]) parameters with histopathologic features of sinonasal malignant tumors using whole-tumor histogram analysis, which could improve the interpretation of DWI findings and promote the use of these diffusion methods in clinical practice. In comparison with monoexponential DWI and biexponential DWI (IVIM), histogram metrics derived from DKI may better reflect the microstructure of sinonasal malignant tumors, including the cellular, stromal and nuclear fractions.

Background and purpose

Monoexponential DWI, IVIM and DKI are increasingly used in the evaluation of sinonasal malignant tumors. Nevertheless, their histopathologic basis with regard to tissue characterization of sinonasal malignant tumors is still unclear. Region of interest (ROI) measurements are the most common used method in histological correlation studies, but it cannot reflect the heterogeneity of sinonasal malignant tumors comprehensively^1-3. Whole-tumor histogram analysis may be a more integrated method to investigate the histopathologic basis of monoexponential and advanced models of DWI in the characterization of sinonasal malignant tumors.⁴ Thus, the purpose of this study was to correlate histogram parameters derived from monoexponential DWI and advanced DWI (including IVIM and DKI) with histopathologic features of sinonasal malignant tumors.

Methods

Seventy-six patients with sinonasal malignant tumors who underwent multi-b-value DWI scans on a 3T MR scanner (MAGNETOM Verio, Siemens Healthcare, Erlangen, Germany) were enrolled. The detailed DWI parameters were as follows: TR/TE = 5200/83 msec, δ = 27.4 msec, Δ = 39.4 msec, number of averages = 2, acquisition matrix = 120 × 120; field of view (FOV) = 220×220 mm2, slice thickness = 5 mm, intersection gap = 5 mm, parallel imaging acceleration factor = 2; 14 different b values ranging from 0 to 2500 sec/mm2 were used (b = 0, 50, 100, 150, 200, 250, 300, 350, 400, 800, 1000, 1500, 2000, and 2500 sec/mm2). The estimation of three DWI models were performed using custom-written scripts in MATLAB (version R2016a; MathWorks, Natick, Mass) to provide ADC, D, D*, f, Dk and K parametric maps on a pixel-by-pixel basis.⁵ The whole-tumor histogram metrics were calculated on these parametric maps using PyRadiomics (version 1.3.0; http://github.com/Radiomics/pyradiomics) based on Python (version 3.5.4; http://www.python.org).⁶ Spearman correlations and stepwise multiple linear regression analyses were performed to determine the correlations between histogram metrics and histopathologic features, including nuclear, cytoplasmic, cellular and stromal fractions, as well as the nuclear-to-cytoplasmic (N/C) ratio.

Results

Most histogram metrics of ADC, Dk and f showed significant correlations with the investigated histopathologic features (P < .05). One example of the calculated conventional and advanced DWI parameters are shown in Figure 1, the corresponding whole-tumor histogram distributions are displaced in Figure 2, and the Histopathologic processing of an H&E stained slice is shown in Figure 3. Several D and K histogram metrics significantly correlated with cellular, stromal and nuclear fractions (all P < .05). Significant correlations between the 75th percentile of D and the cytoplasmic fraction and as well as the kurtosis of K and the N/C ratio were also found (all P < .05). The skewness of Dk, K, and the 75th percentile of D were independently associated with cellular and nuclear fractions, and the skewness of Dk and K were independently associated with the stromal fraction (all P < .05).

Conclusion

Some histogram metrics derived from monoexponential DWI and advanced DWI (IVIM, DKI) showed significant correlations with histopathologic features in sinonasal malignant tumors, suggesting that these parameters could reflect detailed microstructural information, including cellular, stromal, nuclear and cytoplasmic fractions and the N/C ratio. Moreover, the histogram metrics obtained from DKI were the most efficient indicators for the characterization of tumor microstructure. Therefore, whole-tumor histogram analysis of IVIM and DKI is useful for planning the treatment and predicting the prognosis of the patients with sinonasal malignant tumors.

Keywords

Neoplasms; Magnetic Resonance Imaging; Histopathology; Diffusion Magnetic Resonance Imaging; Histogram Analysis

Acknowledgements

This work was supported by the Grant of Science and Technology Commission of Shanghai Municipality (Grant number: 17411962100) and Key Project of the National Natural Science Foundation of China (Grant number: 61731009).

References

1.Su SY, Kupferman ME, DeMonte F, et al. Endoscopic resection of sinonasal cancers. Current oncology reports 2014;16:369

2.Eggesbo HB. Imaging of sinonasal tumours. Cancer imaging : the official publication of the International Cancer Imaging Society 2012;12:136-152

3.Koeller KK. Radiologic Features of Sinonasal Tumors. Head and neck pathology 2016;10:1-12

4.Chandarana H, Rosenkrantz AB, Mussi TC, et al. Histogram analysis of whole-lesion enhancement in differentiating clear cell from papillary subtype of renal cell cancer. Radiology 2012;265:790-798

5.Xiao Z, Zhong Y, Tang Z, et al. Standard diffusion-weighted, diffusion kurtosis and intravoxel incoherent motion MR imaging of sinonasal malignancies: correlations with Ki-67 proliferation status. European radiology 2018;28:2923-2933

6.van Griethuysen JJM, Fedorov A, Parmar C, et al. Computational Radiomics System to Decode the Radiographic Phenotype. Cancer research 2017;77:e104-e107

Figures

Representative monoexponential and advanced DWI parametric maps in a 58-year-old man with sinonasal squamous cell carcinoma. The mass is located in the right ethmoidal sinus and nasal cavity with the involvement of the right orbit (yellow polygonal ROI), showing heterogeneous hypointensity on ADC (a), D (b) and Dk (e) maps, hypo- to hyperintensity on D* (c) and f (d) maps, and hyperintensity on K (f) map. Graphs show signal intensity versus b value fits in single pixels of sinonasal squamous cell carcinoma with the IVIM (g) and DKI (h) models.

The whole-tumor histogram distributions of monoexponential and advanced DWI parameters obtained from the same case as in Figure 2. (a) ADC, (b) D, (c) D*, (d) f, (e) Dk and (f) K histograms.

Histopathologic processing of an H&E stained slice from the same case as in Figure 2. (a) A digitized H&E stained image was obtained at ×20 magnification. (b) The nuclear (blue area, Positive Pixel Count Algorithm setting 1), (c) cellular (red area, Positive Pixel Count Algorithm setting 2) and stromal (blue area, Positive Pixel Count Algorithm setting 2) fractions were segmented and calculated using the two algorithm settings. With these, the cytoplasmic fraction (red area on c minus blue area on b) and the N/C ratio were also calculated.

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)

3649