1963

Histogram model of DCE-MRI in predicting progression of enhancing non-measurable disease after chemoradiotherapy in high-grade glioma

Haimei Cao¹, Zhousan Huang¹, Ruowei Qiu¹, Zhiyong Li², Kan Deng³, Jay J Pillai⁴, Guanglong Huang², Yikai Xu¹, Jun Hua^5,6, and Yuankui Wu¹
¹Department of Medical Imaging, Nanfang Hospital, Southern Medical University, Guangzhou, China, ²Department of Neurosurgery, Nanfang Hospital, Southern Medical University, Guangzhou, China, ³Philips Healthcare, Guangzhou, China, ⁴Division of Neuroradiology, Russell H. Morgan Department of Radiology and Radiological Science, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ⁵Neurosection, Division of MRI Research, Department of Radiology, Johns Hopkins University School of Medicine, Baltimore, MD, United States, ⁶F.M. Kirby Research Center for Functional Brain Imaging, Kennedy Krieger Institute, Baltimore, MD, United States

Synopsis

Keywords: Tumors, Perfusion, enhancing non-measurable disease (NMD)

Early prediction of disease progression is of potential clinical significance for the management of high-grade glioma (HGG) patients. We investigated the value of histogram models based on volume transfer constant (K^trans) between the plasma and extravascular extracellular space and extravascular volume (Ve) in predicting the progression of enhancing non-measurable diseases (NMD) of HGG after chemoradiotherapy. Our results showed that histogram models based on K^trans and Ve can accurately predict the progression of enhancing NMD of HGG following chemoradiotherapy, and combining K^trans and Ve helps improve the prediction performance.

Introduction

Early prediction of disease progression is of potential clinical significance for the management of high-grade glioma (HGG) patients ¹. Follow-up is usually recommended for enhancing non-measurable disease (NMD), i.e, with a diameter of less than 10 mm. Dynamic contrast-enhanced (DCE) MRI can provide information on the dynamic characteristics of tumor angiogenesis and microcirculation, which may be useful in the evaluation of disease progression ^2-4. Histogram analysis is a robust method that can provide quantitative information on tissue characteristics ^{5, 6}. However, the use of histogram features of DCE-MRI perfusion parameters to assess the disease progression of enhancing NMD in HGG has not yet been reported. Here, we aimed to investigate the value of histogram models based on volume transfer constant (K^trans) between the plasma and extravascular extracellular space and extravascular volume (Ve) in predicting the progression of enhancing NMD in HGG after completion of chemoradiotherapy.

Materials and Methods

The DCE images of patients who underwent temozolomide-based chemoradiation after surgery from January 2016 to May 2022 were analyzed retrospectively. MRI data were acquired on a 3.0T scanner (Ingenia, Philips Healthcare, Best, The Netherlands) with a 16-channel head coil. Post-processing of DCE-MRI was performed using dedicated post-processing software (IntelliSpace Portal V9.0, Philips). We calculated K^trans and Ve based on the dual compartment extended Tofts pharmacokinetic model ^{7, 8}. Then, histogram features including 10th percentile, 90th percentile, Energy, Entropy, InterquartileRange (IR), Kurtosis, Maximum, Mean, MeanAbsoluteDeviation (MAD), Median, Minimum, Range, RobustMeanAbsoluteDeviation (RMAD), RootMeanSquared (RMS), Skewness, TotalEnergy, Uniformity and Variance of K^trans and Ve of enhancing NMD were extracted and compared between the progression and non-progression groups using the Mann-Whitney U test. Histogram features with significant differences (at P < 0.05) were included in binary logistic regression to construct histogram models of perfusion parameters individually or combined to predict progression in 2-3 months. Receiver operating characteristic curves were used to evaluate the prediction performance of different models (K^trans, Ve and K^trans + Ve).

Results

A total of 89 HGG patients (mean age ± standard deviation, 47 years ± 12; 67 men) were enrolled, including 50 in the progression group and 39 in the non-progression group. The numbers of histogram feature with significant differences between the two groups were 10 and 13 for K^trans and Ve, respectively. For K^trans, Mean, Median, Minimum, 10th percentile, 90th percentile, RMS and Uniformity were significantly higher in the progression group, while Entropy, Kurtosis and Skewness were significantly lower in the progression group. For Ve, Mean, Median, Minimum, 10th percentile, 90th percentile, Entropy, IR, MAD, RMAD and RMS were significantly higher in the progression group, while Kurtosis, Skewness and Uniformity were significantly lower in the progression group. The histogram model of K^trans (10th percentile + 90th percentile + Skewness) showed an area under the curve (AUC) of 0.927 in predicting progression. The models of Ve (10th percentile + Minimum + Skewness) had an AUC of 0.917. Combining K^trans and Ve, the model (Ve_10th percentile + Ve_Skewness + K^trans_ RootMeanSquared) had an AUC of 0.953. Table 1 summarizes the predictive performance of different models. Two representative cases of the progressive and non-progressive diseases are shown in Figures 1 and 2.ROC curves of the models of K^trans, Ve and K^trans + Ve in predicting progression are plotted in Figure 3.

Discussion

Our results demonstrated that the histogram model based on K^trans and Ve individually or combined could accurately predict the progression of enhancing NMD.
Previous studies have used DCE perfusion parameters to predict glioma progression. Yoo et al ⁹ found that the Mean of K^trans and Ve and the 5th percentile of K^trans could predict the progression of measurable disease (MD) in glioblastoma patients after chemoradiotherapy. However, their research did not analyze NMD. Of note, enhancing lesions smaller than 10 mm in diameter might be an early stage of glioma progression ¹⁰. Therefore, accurate identification of progression at the stage of NMD would provide earlier evidence for individualized clinical decision-making as compared with at the stage of MD.
Yun et al ⁴ investigated the diagnostic performance of DCE in differentiating true progression from pseudoprogression and found that Mean and 10th percentile of K^trans, Mean and 5th percentile of Ve were higher in true progression. The present study showed that histogram features of K^trans and Ve of enhancing NMD, such as 10th percentile, 90th percentile, Mean, Median RMS and Minimum, were higher in the progression group, which was basically consistent with the study by Yun et al ⁴. Of note, the progression group showed lower Skewness in this study, which was consistent with the study by Baek et al ¹¹. Generally, the decrease in Skewness represents an increase in perfusion parameters.

Conclusion

The histogram models based on K^trans and Ve can predict the progression of enhancing nonmeasurable disease in high-grade glioma following chemoradiotherapy 2-3 months in advance, and combining K^trans and Ve helps improve the prediction performance.

Acknowledgements

References

1. Khalifa J, Tensaouti F, Chaltiel L, et al. Identification of a candidate biomarker from perfusion MRI to anticipate glioblastoma progression after chemoradiation. Eur Radiol, 2016;26(11):4194-4203.

2. Kim H S, Goh M J, Kim N, et al. Which combination of MR imaging modalities is best for predicting recurrent glioblastoma? Study of diagnostic accuracy and reproducibility. Radiology, 2014;273(3):831-843.

3. Shin K E, Ahn K J, Choi H S, et al. DCE and DSC MR perfusion imaging in the differentiation of recurrent tumour from treatment-related changes in patients with glioma. Clinical Radiology, 2014;69(6):e264-e272.

4. Yun T J, Park C K, Kim T M, et al. Glioblastoma treated with concurrent radiation therapy and temozolomide chemotherapy: differentiation of true progression from pseudoprogression with quantitative dynamic contrast-enhanced MR imaging. Radiology, 2015;274(3):830-840.

5. Davnall F, Yip C S, Ljungqvist G, et al. Assessment of tumor heterogeneity: an emerging imaging tool for clinical practice?. Insights Imaging, 2012;3(6):573-589.

6. He W, Li X, Hua J, et al. Noninvasive Assessment of O(6)-Methylguanine-DNA methyltransferase ppromoter methylation status in world health organization grade II-IV glioma using histogram analysis of inflow-based vascular-space-occupancy combined with structural magnetic resonance imaging. J Magn Reson Imaging, 2021;54(1):227-236.

7. Hu L S, Baxter L C, Pinnaduwage D S, et al. Optimized preload leakage-correction methods to improve the diagnostic accuracy of dynamic susceptibility-weighted contrast-enhanced perfusion MR imaging in posttreatment gliomas. AJNR Am J Neuroradiol, 2010;31(1):40-48.

8. Tofts P S, Kermode A G. Measurement of the blood-brain barrier permeability and leakage space using dynamic MR imaging. 1. Fundamental concepts. Magn Reson Med, 1991;17(2):357-367.

9. Yoo R, Choi S H, Kim T M, et al. Dynamic contrast-enhanced MR imaging in predicting progression of enhancing lesions persisting after standard treatment in glioblastoma patients: a prospective study. European Radiology, 2017;27(8):3156-3166.

10. Young R J, Gupta A, Shah A D, et al. Potential utility of conventional MRI signs in diagnosing pseudoprogression in glioblastoma. Neurology, 2011;76(22):1918-1924.

11. Baek H J, Kim H S, Kim N, et al. Percent change of perfusion skewness and kurtosis: a potential imaging biomarker for early treatment response in patients with newly diagnosed glioblastomas. Radiology, 2012;264(3):834-843.

Figures

Table 1. Predictive Performance of Different Models in Predicting Progression

Note. — CI, confidence interval; Se, sensitivity; Sp, specificity; ROC, receiver operating characteristic; AUC, area under the ROC curve; RMS = RootMeanSquared.

Figure 1. A representative case of progression in a 35-year-old man with glioblastoma. An enhancing lesion (circle) with a size of 9× 6 × 7 mm³ was noted in the right frontal lobe on CE-T1W 6 months after chemoradiotherapy. The histogram of K^trans shows a relatively wide distribution in the enhancing lesion with a Kurtosis of 0.561 and a Skewness of 0.763, with the peak located at around 0.01 min^-1. The follow-up MRIs 3 months later demonstrated an avidly enhancing mass and the patient showed severe clinical deterioration and was diagnosed as having a progressive disease.

Figure 2. A representative case of non-progression in a 48-year-old woman with glioblastoma. An enhancing nodule (circle) with a size of 8 × 7 × 8 mm³ was noted in the right occipital lobe on CE-T1W 3 months after chemoradiotherapy. The histogram of K^trans shows a relatively narrow distribution in the enhancing nodule with a Kurtosis of 1.881 and a Skewness of 1.246, with the peak located at around 0.003 min^-1. The enhancing nodule measured 4 × 3 × 3 mm³ on the follow-up MRIs 3 months later and the patient was clinically stable and was diagnosed as having a non-progression disease.

Figure 3. Receiver operating characteristic curves of histogram models of single and combined perfusion parameters in predicting progression. The areas under the curves are 0.953 for K^trans + Ve, which is better than those of single perfusion parameter.

Proc. Intl. Soc. Mag. Reson. Med. 31 (2023)

1963

DOI: https://doi.org/10.58530/2023/1963