Predicting survival in patients with glioblastoma using MRI radiomic features extracted from radiation planning volumes

BACKGROUND

Quantitative image analysis using pre-operative magnetic resonance imaging (MRI) has been able to predict survival in patients with glioblastoma (GBM). The study explored the role of postoperative radiation (RT) planning MRI-based radiomics to predict the outcomes, with features extracted from the gross tumor volume (GTV) and clinical target volume (CTV).

METHODS

Patients with IDH-wildtype GBM treated with adjuvant RT having MRI as a part of RT planning process were included in the study. 546 features were extracted from each GTV and CTV. A LASSO Cox model was applied, and internal validation was performed using leave-one-out cross-validation with overall survival as endpoint. Cross-validated time-dependent area under curve (AUC) was constructed to test the efficacy of the radiomics model, and clinical features were used to generate a combined model. Analysis was done for the entire group and in individual surgical groups-gross total excision (GTR), subtotal resection (STR), and biopsy.

RESULTS

235 patients were included in the study with 57, 118, and 60 in the GTR, STR, and biopsy subgroup, respectively. Using the radiomics model, binary risk groups were feasible in the entire cohort (p < 0.01) and biopsy group (p = 0.04), but not in the other two surgical groups individually. The integrated AUC (iAUC) was 0.613 for radiomics-based classification in the biopsy subgroup, which improved to 0.632 with the inclusion of clinical features.

CONCLUSION

Imaging features extracted from the GTV and CTV regions can lead to risk-stratification of GBM undergoing biopsy, while the utility in other individual subgroups needs to be further explored.

Correlation of hyperpolarized 13 C-MRI data with tissue extract measurements

Hyperpolarized (HP) 13C MRI provides the means to monitor lactate metabolism noninvasively in tumours. Since 13C -lactate signal levels obtained from HP 13C imaging depend on multiple factors, such as the rate of 13C substrate delivery via the vasculature, the expression level of monocarboxylate transporters (MCTs) and lactate dehydrogenase (LDH), and the local lactate pool size, the interpretation of HP 13C metabolic images remains challenging. In this study, ex vivo tissue extract measurements (i.e., NMR isotopomer analysis, western blot analysis) derived from an MDA-MB-231 xenograft model in nude rats were used to test for correlations between the in vivo 13C data and the ex vivo measures. The lactate-to-pyruvate ratio from HP 13C MRI was strongly correlated with [1- 13C ]lactate concentration measured from the extracts using NMR (R = 0.69, p < 0.05), as well as negatively correlated with tumour wet weight (R = -  0.60, p < 0.05). In this tumour model, both MCT1 and MCT4 expressions were positively correlated with wet weight ( ρ = 0.78 and 0.93, respectively, p < 0.01). Lactate pool size and the lactate-to-pyruvate ratio were not significantly correlated.

Partial Fourier reconstruction for improved resolution in 3D hyperpolarized 13 C EPI

PURPOSE

Asymmetric in-plane k-space sampling of EPI can reduce the minimum achievable TE in hyperpolarized 13 C with spectral-spatial radio frequency pulses, thereby reducing T 2 * weighting and signal-losses. Partial Fourier image reconstruction exploits the approximate Hermitian symmetry of k-space data and can be applied to asymmetric data sets to synthesize unmeasured data. Here we tested whether the application of partial Fourier image reconstruction would improve spatial resolution from hyperpolarized [1- 13 C ]pyruvate scans in the human brain.

METHODS

Fifteen healthy control subjects were imaged using a volumetric dual-echo echo-planar imaging sequence with spectral-spatial radio frequency excitation. Images were reconstructed by zero-filling as well as with the partial Fourier reconstruction algorithm projection-on-convex-sets. Resulting images were quantitatively evaluated with a no-reference image quality assessment.

RESULTS

The no-reference image sharpness metric agreed with perceived improvements in image resolution and contrast. The [1- 13 C ]lactate images benefitted most, followed by the [1- 13 C ]pyruvate images. The 13 C -bicarbonate images were improved by the smallest degree, likely owing to relatively lower SNR.

CONCLUSIONS

Partial Fourier imaging and reconstruction were shown to improve the sharpness and contrast of human HP 13 C brain data and is a viable method for enhancing resolution.

Noninvasive Interrogation of Cancer Metabolism with Hyperpolarized 13C MRI

This review will highlight recent advances in hyperpolarized ¹³C MR spectroscopic imaging, which can be used to noninvasively interrogate tumor metabolism. After providing an overview of MR and hyperpolarization, we will discuss the latest advances in data acquisition techniques. Next, we will shift our focus to hyperpolarized probe design and provide an overview of the latest hyperpolarized ¹³C MR spectroscopic imaging probes developed in the last several years.

Dual-Echo EPI sequence for integrated distortion correction in 3D time-resolved hyperpolarized 13 C MRI

PURPOSE

To provide built-in off-resonance correction in time-resolved, volumetric hyperpolarized ¹³ C metabolic imaging by implementing a novel dual-echo 3D echo-planar imaging (EPI) sequence and reconstruction.

METHODS

A spectral-spatial pulse for single-resonance excitation followed by a dual-echo 3D EPI readout was implemented to provide 64 × 8 × 6 cm³ coverage at 5 × 5 × 5 mm³ nominal resolution. Multiple sources of EPI distortions were encoded using a multi-echo ¹ H EPI reference scan. Phase maps computed from the reference scans were combined with a bulk ¹³ C frequency offset encoded in the dual-echo [1-¹³ C]pyruvate images to correct geometric distortion and improve spatial registration. The proposed scheme was validated in a phantom study, and in vivo [1-¹³ C]pyruvate and [1-¹³ C]lactate rat images were acquired with intentional transmit frequency deviations to assess the dual-echo 3D EPI sequence.

RESULTS

The phantom study demonstrated improved spatial registration in off-resonance corrected images. Close agreement was observed between metabolic kidney signal and the underlying anatomy in rat imaging experiments. Relative to a single-echo acquisition, the coherent addition of the two corrected echoes provided the expected increase in signal-to-noise ratio by approximately 2.

CONCLUSION

A novel dual-echo 3D EPI acquisition sequence for integrated off-resonance correction in hyperpolarized ¹³ C imaging was developed and demonstrated. The proposed sequence offers clear advantages over flyback EPI for time-resolved metabolic mapping. Magn Reson Med 79:643-653, 2018. © 2017 International Society for Magnetic Resonance in Medicine.

Hyperpolarized 13C Metabolic MRI of the Human Heart: Initial Experience

Rationale:

Altered cardiac energetics is known to play an important role in the progression toward heart failure. A noninvasive method for imaging metabolic markers that could be used in longitudinal studies would be useful for understanding therapeutic approaches that target metabolism.

Objective:

To demonstrate the first hyperpolarized ¹³C metabolic magnetic resonance imaging of the human heart.

Methods and Results:

Four healthy subjects underwent conventional proton cardiac magnetic resonance imaging followed by ¹³C imaging and spectroscopic acquisition immediately after intravenous administration of a 0.1 mmol/kg dose of hyperpolarized [1-¹³C]pyruvate. All subjects tolerated the procedure well with no adverse effects reported ≤1 month post procedure. The [1-¹³C]pyruvate signal appeared within the chambers but not within the muscle. Imaging of the downstream metabolites showed ¹³C-bicarbonate signal mainly confined to the left ventricular myocardium, whereas the [1-¹³C]lactate signal appeared both within the chambers and in the myocardium. The mean ¹³C image signal:noise ratio was 115 for [1-¹³C]pyruvate, 56 for ¹³C-bicarbonate, and 53 for [1-¹³C]lactate.

Conclusions:

These results represent the first ¹³C images of the human heart. The appearance of ¹³C-bicarbonate signal after administration of hyperpolarized [1-¹³C]pyruvate was readily detected in this healthy cohort (n=4). This shows that assessment of pyruvate metabolism in vivo in humans is feasible using current technology.

Clinical Trial Registration:

URL: https://www.clinicaltrials.gov. Unique identifier: NCT02648009.

Accelerated 3D echo-planar imaging with compressed sensing for time-resolved hyperpolarized 13 C studies

PURPOSE

To enable large field-of-view, time-resolved volumetric coverage in hyperpolarized ¹³ C metabolic imaging by implementing a novel data acquisition and image reconstruction method based on the compressed sensing framework.

METHODS

A spectral-spatial pulse for single-resonance excitation followed by a symmetric echo-planar imaging (EPI) readout was implemented for encoding a 72 × 18 cm² field of view at 5 × 5 mm² resolution. Random undersampling was achieved with blipped z-gradients during the ramp portion of the echo-planar imaging readout. The sequence and reconstruction were tested with phantom studies and consecutive in vivo hyperpolarized ¹³ C scans in rats. Retrospectively and prospectively undersampled data were compared on the basis of structural similarity in the reconstructed images and the quantification of the lactate-to-pyruvate ratio in rat kidneys.

RESULTS

No artifacts or loss of resolution are evident in the compressed sensing reconstructed images acquired with the proposed sequence. Structural similarity analysis indicate that compressed sensing reconstructions can accurately recover spatial features in the metabolic images evaluated.

CONCLUSION

A novel z-blip acquisition sequence for compressed sensing accelerated hyperpolarized ¹³ C 3D echo-planar imaging was developed and demonstrated. The close agreement in lactate-to-pyruvate ratios from both retrospectively and prospectively undersampled data from rats shows that metabolic information is preserved with acceleration factors up to 3-fold with the developed method. Magn Reson Med 77:538-546, 2017. © 2016 International Society for Magnetic Resonance in Medicine.

Intensity correction for multichannel hyperpolarized 13C imaging of the heart

PURPOSE

Develop and test an analytic correction method to correct the signal intensity variation caused by the inhomogeneous reception profile of an eight-channel phased array for hyperpolarized (13) C imaging.

THEORY AND METHODS

Fiducial markers visible in anatomical images were attached to the individual coils to provide three dimensional localization of the receive hardware with respect to the image frame of reference. The coil locations and dimensions were used to numerically model the reception profile using the Biot-Savart Law. The accuracy of the coil sensitivity estimation was validated with images derived from a homogenous (13) C phantom. Numerical coil sensitivity estimates were used to perform intensity correction of in vivo hyperpolarized (13) C cardiac images in pigs.

RESULTS

In comparison to the conventional sum-of-squares reconstruction, improved signal uniformity was observed in the corrected images.

CONCLUSION

The analytical intensity correction scheme was shown to improve the uniformity of multichannel image reconstruction in hyperpolarized [1-(13) C]pyruvate and (13) C-bicarbonate cardiac MRI. The method is independent of the pulse sequence used for (13) C data acquisition, simple to implement and does not require additional scan time, making it an attractive technique for multichannel hyperpolarized (13) C MRI.