Performance sensitivity analysis of brain metastasis stereotactic radiosurgery outcome prediction using MRI radiomics.

Recent studies have used T1w contrast-enhanced (T1w-CE) magnetic resonance imaging (MRI) radiomic features and machine learning to predict post-stereotactic radiosurgery (SRS) brain metastasis (BM) progression, but have not examined the effects of combining clinical and radiomic features, BM primary cancer, BM volume effects, and using multiple scanner models. To investigate these effects, a dataset of n = 123 BMs from 99 SRS patients with 12 clinical features, 107 pre-treatment T1w-CE radiomic features, and BM progression determined by follow-up MRI was used with a random decision forest model and 250 bootstrapped repetitions. Repeat experiments assessed the relative accuracy across primary cancer sites, BM volume groups, and scanner model pairings. Correction for accuracy imbalances across volume groups was investigated by removing volume-correlated features. We found that using clinical and radiomic features together produced the most accurate model with a bootstrap-corrected area under the receiver operating characteristic curve of 0.77. Accuracy also varied by primary cancer site, BM volume, and scanner model pairings. The effect of BM volume was eliminated by removing features at a volume-correlation coefficient threshold of 0.25. These results show that feature type, primary cancer, volume, and scanner model are all critical factors in the accuracy of radiomics-based prognostic models for BM SRS that must be characterised and controlled for before clinical translation.

A generalized parametric response mapping method for analysis of multi-parametric imaging: A feasibility study with application to glioblastoma.

PURPOSE: Parametric response map (PRM) analysis of functional imaging has been shown to be an effective tool for early prediction of cancer treatment outcomes and may also be well-suited toward guiding personalized adaptive radiotherapy (RT) strategies such as sub-volume boosting. However, the PRM method was primarily designed for analysis of longitudinally acquired pairs of single-parameter image data. The purpose of this study was to demonstrate the feasibility of a generalized parametric response map analysis framework, which enables analysis of multi-parametric data while maintaining the key advantages of the original PRM method. METHODS: MRI-derived apparent diffusion coefficient (ADC) and relative cerebral blood volume (rCBV) maps acquired at 1 and 3-months post-RT for 19 patients with high-grade glioma were used to demonstrate the algorithm. Images were first co-registered and then standardized using normal tissue image intensity values. Tumor voxels were then plotted in a four-dimensional Cartesian space with coordinate values equal to a voxel's image intensity in each of the image volumes and an origin defined as the multi-parametric mean of normal tissue image intensity values. Voxel positions were orthogonally projected onto a line defined by the origin and a pre-determined response vector. The voxels are subsequently classified as positive, negative or nil, according to whether projected positions along the response vector exceeded a threshold distance from the origin. The response vector was selected by identifying the direction in which the standard deviation of tumor image intensity values was maximally different between responding and non-responding patients within a training dataset. Voxel classifications were visualized via familiar three-class response maps and then the fraction of tumor voxels associated with each of the classes was investigated for predictive utility analogous to the original PRM method. Independent PRM and MPRM analyses of the contrast-enhancing lesion (CEL) and a 1 cm shell of surrounding peri-tumoral tissue were performed. Prediction using tumor volume metrics was also investigated. Leave-one-out cross validation (LOOCV) was used in combination with permutation testing to assess preliminary predictive efficacy and estimate statistically robust P-values. The predictive endpoint was overall survival (OS) greater than or equal to the median OS of 18.2 months. RESULTS: Single-parameter PRM and multi-parametric response maps (MPRMs) were generated for each patient and used to predict OS via the LOOCV. Tumor volume metrics (P ≥ 0.071 ± 0.01) and single-parameter PRM analyses (P ≥ 0.170 ± 0.01) were not found to be predictive of OS within this study. MPRM analysis of the peri-tumoral region but not the CEL was found to be predictive of OS with a classification sensitivity, specificity and accuracy of 80%, 100%, and 89%, respectively (P = 0.001 ± 0.01). CONCLUSIONS: The feasibility of a generalized MPRM analysis framework was demonstrated with improved prediction of overall survival compared to the original single-parameter method when applied to a glioblastoma dataset. The proposed algorithm takes the spatial heterogeneity in multi-parametric response into consideration and enables visualization. MPRM analysis of peri-tumoral regions was shown to have predictive potential supporting further investigation of a larger glioblastoma dataset.

Evaluation of CT Perfusion Biomarkers of Tumor Hypoxia.

BACKGROUND: Tumor hypoxia is associated with treatment resistance to cancer therapies. Hypoxia can be investigated by immunohistopathologic methods but such procedure is invasive. A non-invasive method to interrogate tumor hypoxia is an attractive option as such method can provide information before, during, and after treatment for personalized therapies. Our study evaluated the correlations between computed tomography (CT) perfusion parameters and immunohistopathologic measurement of tumor hypoxia. METHODS: Wistar rats, 18 controls and 19 treated with stereotactic radiosurgery (SRS), implanted with the C6 glioma tumor were imaged using CT perfusion on average every five days to monitor tumor growth. A final CT perfusion scan and the brain were obtained on average 14 days (8-22 days) after tumor implantation. Tumor hypoxia was detected immunohistopathologically with pimonidazole. The tumor, necrotic, and pimonidazole-positive areas on histology samples were measured. Percent necrotic area and percent hypoxic areas were calculated. Tumor volume (TV), blood flow (BF), blood volume (BV), and permeability-surface area product (PS) were obtained from the CT perfusion studies. Correlations between CT perfusion parameters and histological parameters were assessed by Spearman's ρ correlation. A Bonferroni-corrected P value < 0.05 was considered significant. RESULTS: BF and BV showed significant correlations with percent hypoxic area ρ = -0.88, P < 0.001 and ρ = -0.81, P < 0.001, respectively, for control animals and ρ = -0.7, P < 0.001 and ρ = -0.6, P = 0.003, respectively, for all animals, while TV and BV were correlated (ρ = -0.64, P = 0.01 and ρ = -0.43, P = 0.043, respectively) with percent necrotic area. PS was not correlated with either percent necrotic or percent hypoxic areas. CONCLUSIONS: Percent hypoxic area provided significant correlations with BF and BV, suggesting that CT perfusion parameters are potential non-invasive imaging biomarkers of tumor hypoxia.

Survival prediction in high-grade gliomas using CT perfusion imaging.

Patients with high-grade gliomas usually have heterogeneous response to surgery and chemoirradiation. The objectives of this study were (1) to evaluate serial changes in tumor volume and perfusion imaging parameters and (2) to determine the value of these data in predicting overall survival (OS). Twenty-nine patients with World Health Organization grades III and IV gliomas underwent magnetic resonance (MR) and computed tomography (CT) perfusion examinations before surgery, and 1, 3, 6, 9, and 12 months after radiotherapy. Serial measurements of tumor volumes and perfusion parameters were evaluated by receiver operating characteristic analysis, Cox proportional hazards regression, and Kaplan-Meier survival analysis to determine their values in predicting OS. Higher trends in blood flow (BF), blood volume (BV), and permeability-surface area product in the contrast-enhancing lesions (CEL) and the non-enhancing lesions (NEL) were found in patients with OS < 18 months compared to those with OS ≥ 18 months, and these values were significant at selected time points (P < 0.05). Only CT perfusion parameters yielded sensitivities and specificities of ≥ 70% in predicting 18 and 24 months OS. Pre-surgery BF in the NEL and BV in the CEL and NEL 3 months after radiotherapy had sensitivities and specificities >80% in predicting 24 months OS in patients with grade IV gliomas. Our study indicated that CT perfusion parameters were predictive of survival and could be useful in assessing early response and in selecting adjuvant treatment to prolong survival if verified in a larger cohort of patients.

Dynamic perfusion CT in brain tumors.

Dynamic perfusion CT (PCT) is an imaging technique for assessing the vascular supply and hemodynamics of brain tumors by measuring blood flow, blood volume, and permeability-surface area product. These PCT parameters provide information complementary to histopathologic assessments and have been used for grading brain tumors, distinguishing high-grade gliomas from other brain lesions, differentiating true progression from post-treatment effects, and predicting prognosis after treatments. In this review, the basic principles of PCT are described, and applications of PCT of brain tumors are discussed. The advantages and current challenges, along with possible solutions, of PCT are presented.

CT perfusion imaging as an early biomarker of differential response to stereotactic radiosurgery in C6 rat gliomas.

BACKGROUND: The therapeutic efficacy of stereotactic radiosurgery for glioblastoma is not well understood, and there needs to be an effective biomarker to identify patients who might benefit from this treatment. This study investigated the efficacy of computed tomography (CT) perfusion imaging as an early imaging biomarker of response to stereotactic radiosurgery in a malignant rat glioma model. METHODS: Rats with orthotopic C6 glioma tumors received either mock irradiation (controls, N = 8) or stereotactic radiosurgery (N = 25, 12 Gy in one fraction) delivered by Helical Tomotherapy. Twelve irradiated animals were sacrificed four days after stereotactic radiosurgery to assess acute CT perfusion and histological changes, and 13 irradiated animals were used to study survival. Irradiated animals with survival >15 days were designated as responders while those with survival ≤15 days were non-responders. Longitudinal CT perfusion imaging was performed at baseline and regularly for eight weeks post-baseline. RESULTS: Early signs of radiation-induced injury were observed on histology. There was an overall survival benefit following stereotactic radiosurgery when compared to the controls (log-rank P<0.04). Responders to stereotactic radiosurgery showed lower relative blood volume (rBV), and permeability-surface area (PS) product on day 7 post-stereotactic radiosurgery when compared to controls and non-responders (P<0.05). rBV and PS on day 7 showed correlations with overall survival (P<0.05), and were predictive of survival with 92% accuracy. CONCLUSIONS: Response to stereotactic radiosurgery was heterogeneous, and early selection of responders and non-responders was possible using CT perfusion imaging. Validation of CT perfusion indices for response assessment is necessary before clinical implementation.

Improving quantitative CT perfusion parameter measurements using principal component analysis.

RATIONALE AND OBJECTIVES: To evaluate the improvements in measurements of blood flow (BF), blood volume (BV), and permeability-surface area product (PS) after principal component analysis (PCA) filtering of computed tomography (CT) perfusion images. To evaluate the improvement in CT perfusion image quality with poor contrast-to-noise ratio (CNR) in vivo. MATERIALS AND METHODS: A digital phantom with CT perfusion images reflecting known values of BF, BV, and PS was created and was filtered using PCA. Intraclass correlation coefficients and Bland-Altman analysis were used to assess reliability of measurements and reduction in measurement errors, respectively. Rats with C6 gliomas were imaged using CT perfusion, and the raw CT perfusion images were filtered using PCA. Differences in CNR, BF, BV, and PS before and after PCA filtering were assessed using repeated measures analysis of variance. RESULTS: From simulation, mean errors decreased from 12.8 (95% confidence interval [CI] = -19.5 to 45.0) to 1.4 mL/min/100 g (CI = -27.6 to 30.4), 0.2 (CI = -1.1 to 1.4) to -0.1 mL/100 g (CI = -1.1 to 0.8), and 2.9 (CI = -2.4 to 8.1) to 0.2 mL/min/100 g (CI = -3.5 to 3.9) for BF, BV, and PS, respectively. Map noise in BF, BV, and PS were decreased from 51.0 (CI = -3.5 to 105.5) to 11.6 mL/min/100 g (CI = -7.9 to 31.2), 2.0 (CI = 0.7 to 3.3) to 0.5 mL/100 g (CI = 0.1 to 1.0), and 8.3 (CI = -0.8 to 17.5) to 1.4 mL/min/100 g (CI = -0.4 to 3.1), respectively. For experiments, CNR significantly improved with PCA filtering in normal brain (P < .05) and tumor (P < .05). Tumor and brain BFs were significantly different from each other after PCA filtering with four principal components (P < .05). CONCLUSIONS: PCA improved image CNR in vivo and reduced the measurement errors of BF, BV, and PS from simulation. A minimum of four principal components is recommended.

The effect of scan duration on the measurement of perfusion parameters in CT perfusion studies of brain tumors.

RATIONALE AND OBJECTIVES: The aim of this study was to investigate the effect of scan duration on the measurement of blood flow (BF), blood volume (BV), and permeability-surface area product (PS) in patients undergoing computed tomography (CT) perfusion for brain tumors. MATERIALS AND METHODS: CT perfusion scans were performed in 14 patients with malignant glioma. Patients were scanned for 150 seconds, and BF, BV, and PS were assessed for scan durations of 150, 120, 90, and 60 seconds. Systematic, random, and percentage errors associated with shorter scan durations were calculated. Repeated-measures analyses of variance with paired t tests were used to compare the perfusion values measured from different scan durations. Systematic and random errors were correlated with scan duration. RESULTS: No effect of scan duration on BF and BV values was noted (P > .05). PS values were not affected by scan duration except in the tumor rim, in which they were significantly higher at 60 seconds (P < .01). Median percentage error was highest at 60 seconds for tumor core PS (median, 32.1%; interquartile range, 16.5%-43.0%). Tumor rim BV and PS and tumor core BF were correlated with scan durations (r = 0.42, -0.50, and -0.50, respectively; P < .01). Random errors were negatively correlated with scan durations for all tissue types (P ≤ .01) except for white matter BV. CONCLUSIONS: A scan duration of ≤60 seconds is not warranted for the measurement of PS in brain tumors. A scan duration of ≥90 seconds is recommended.