Abstract
Background/Aim: This study investigated the feasibility and efficacy of multiparametric magnetic resonance imaging (MRI)-guided dose-escalated hypofractionated intensity-modulated radiation therapy with simultaneous integrated boost (IMRT-SIB) for glioblastoma. Patients and Methods: Eighteen patients underwent postoperative IMRT-SIB for glioblastoma using three MRI sequences: double inversion recovery (DIR), diffusion tensor imaging (DTI), and post-gadolinium T1-weighted imaging. Prescribed doses were 60 Gy and 40 Gy in 15 fractions for residual enhancing lesions and surrounding tumor-infiltrating areas, respectively. For surrounding tumor-infiltrating areas, asymmetric margins were set with reference to DTI imaging. Results: The 1-year overall survival rate was 58.0%, and the 1-year local control rate for the residual enhancing lesions was 76.2%, while that for surrounding tumor-infiltrating areas was 39.4%. One patient (6%) developed grade 2 cerebral radiation necrosis 10 months after IMRT-SIB, but there was no grade 3 or higher adverse event. Conclusion: Multiparametric MRI-guided dose-escalated IMRT-SIB with DIR and DTI imaging has the potential to improve local control rates without increasing adverse events.
Current standard of care for newly diagnosed glioblastoma is surgery followed by concurrent radiotherapy and daily temozolomide, and then followed by six cycles of temozolomide (1). Conventional 2 Gy daily fractions to a total dose of 60 Gy is the standard dose fractionation regimen, however hypofractionated radiotherapy using 40 Gy in 15 fractions over 3 weeks is also acceptable for patients ≥60 years older (2). An analysis of treatment failure patterns for glioblastoma shows that recurrence within the planning target volume (PTV) is highest and marginal recurrence is infrequent (6%) even if radiation therapy is given with a limited margin (3).
Currently, a single clinical target volume (CTV) defined as the resection cavity and residual enhancing regions with the addition of a 20 mm isotropic margin is recommended (4). However, the proportion of cells that makes up the tumor is different between the enhancing regions and the surrounding tumor-infiltrating areas. That is, the coexistence of normal neurons and tumor cells makes radiotherapy difficult. We considered that local control rates could be improved without increasing adverse events if higher doses of radiation could be administered separately to the residual enhancing areas and the surrounding tumor-infiltrating areas. To achieve this goal, we used double inversion recovery (DIR) imaging and diffusion tensor imaging (DTI) as adjuncts for target delineation in addition to contrast-enhanced T1-weighted imaging. DIR MRI pulse sequence includes two inversion recovery pulses suppressing the signal from both cerebrospinal fluid and normal white matter and results in improved delineation of white matter lesions (5). DTI can visualize the white matter tracts and predict the direction of tumor cell infiltration (6). Glioblastoma cells will infiltrate along the axial direction of the neurons, while they are less likely to infiltrate perpendicular to the axis of the neurons. We hypothesized that we could provide more optimized radiotherapy by detecting tumor cells mixed with normal neurons with DIR images, and by evaluating the direction of tumor cell infiltration with DTI. We set a larger margin in the axial direction of the neurons and a smaller margin in the direction perpendicular to the axis. Furthermore, we considered it is necessary to prescribe different doses for the gross tumor volume (GTV) on contrast-enhanced T1-weighted images and the tumor-infiltrating area on DIR images.
The purpose of this study was to investigate the feasibility and efficacy of multiparametric MRI-guided dose-escalated hypofractionated intensity-modulated radiation therapy (IMRT) with simultaneous integrated boost (SIB) for glioblastoma, and to present a proof of concept using DIR and DTI images.
Patients and Methods
Patients. Between January 2011 and December 2019, 51 patients underwent postoperative radiotherapy for glioblastoma at our institution. Of the 51 patients, 18 (male/female: 13/5) who were followed for more than 3 months were retrospectively evaluated. This study was approved by our institutional ethics committee and written informed consent from all subjects was obtained. The mean age and standard deviation (SD) of the 18 patients was 73.5 (range=24-82 years) and 2.1 years, respectively. Recursive partitioning analysis (RPA) class distribution (7) was: class III- 2 (11%); class IV-2 (11%); class V- 14 (78%).
Computed tomography (CT) scan. The patients were immobilized using the individually formed masks and CT images were acquired. Multidetector-row CT was performed on a 64-slice GE Discovery CT 750HD scanner (GE Healthcare, Waukesha, WI, USA) with a slice thickness of 2.5 mm with no interslice gap. Contrast agent was not used, and the whole procedure generally lasted about 20 min. Multiparametric MRI protocol
Multiparametric MRI was performed using either a 1.5T MRI unit (Signa HDxt, GE Healthcare) or a 3.0T unit (Siemens, Magnetom Skyra, Munich, Germany). Images were obtained using dedicated head coils. Three imaging sequences were used for the target volume delineation: axial DIR, axial DTI, and axial post-gadolinium T1-weighted images. The sequences were obtained with a 2.5-mm slice thickness.
Target volume delineation. CT was used for dose calculation, and multiparametric MRI was used with CT to identify brain structures and tumors. A rigid fusion of both CT and multiparametric MRI was performed by the Monaco software (Elekta, Stockholm, Sweden). GTV was defined on the post-gadolinium T1-weighted images (Figure 1A) and CTV was defined on the DIR images (Figure 1B). Two different PTVs were designated as follows: PTVh defined as GTV plus 1-2 mm margin; PTVl encompassed PTVh and CTV plus asymmetrical 5-20 mm margin excluding PTVh. The asymmetric margin of PTVl was set with reference to DTI imaging. When creating PTVl, a large margin of 10-20 mm was set in the axial direction of the nerve fibers, and a smaller margin of 3-5 mm was set in the direction perpendicular to the axial direction (Figure 1C).
Multiparametric magnetic resonance imaging (MRI)-based contouring of gross tumor volume (GTV), clinical target volume (CTV), and planning target volume (PTV). (A) Post-gadolinium T1-weighted imaging. GTV (blue area) was defined on the post-gadolinium T1-weighted images. PTVh was generated by expanding the GTV by 1-2 mm. (B) Double inversion recovery imaging. CTV (red area) was defined on the double inversion recovery images. (C) Diffusion tensor imaging. PTVl (yellow area) encompassed PTVh and CTV plus asymmetrical 5-20 mm margin excluding PTVh. PTVl was set with reference to diffusion tensor imaging. A large margin of 10-20 mm was set in the axial direction of the nerve fibers, and a smaller margin of 3-5 mm was set in the direction perpendicular to the axial direction.
Dose prescription. All patients underwent IMRT with SIB (IMRT-SIB) using helical tomotherapy (Hi-ART system, Accuray, Sunnyvale, CA, USA). Prescribed doses were 60 Gy and 40 Gy in 15 fractions for PTVh and PTVl, respectively, and each dose was defined as the minimum dose received by 95% volume (D95%) of PTVh or PTVl. No limit was set on the maximum dose for PTVh. D0.1cc (the top dose delivered to a 0.1-ml volume) of the brainstem and optic nerve was set to be less than 41 Gy.
Treatment. IMRT-SIB was performed in 15 daily fractions over a period of 3 weeks. Concurrent temozolomide was administered orally at a dose of 75 mg per square meter of body-surface area per day for 21 consecutive days from day 1 until the final day of radiotherapy. Adjuvant temozolomide was administered at a dose of 150 to 200 mg per square meter per day for 5 consecutive days of a 28-day cycle for up to 12 cycles or until disease progression.
Patient evaluation and follow-up. Before IMRT-SIB, all the patients underwent physical assessment, complete blood counts, blood chemical analyses including tests of renal and hepatic function, and multiparametric MRI. During IMRT-SIB, patients were assessed for adverse events weekly. After IMRT-SIB, gadolinium-enhanced MRI as well as physical assessment was performed every 2-3 months until tumor progression. Adverse events were assessed and graded by physicians according to the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 4.0. Response was assessed on gadolinium-enhanced MRI with the use of serial measures of the product of the two largest cross-sectional diameters. Local progression was defined as an increase in tumor size by at least 25% or the development of a new lesion.
Statistical analysis. The primary end point was overall survival (OS), measured from the last date of IMRT-SIB until death or censoring at the last day that the patient was known to be alive. Local progression-free survival (PFS) was measured from the last date of IMRT-SIB until disease progression within the PTV or death or until the last evaluation date. The Kaplan–Meier curves were generated for OS and PFS, and the generalized Wilcoxon test or the log-rank test was used for comparing the Kaplan–Meier curves. p-Values less than 0.05 were considered statistically significant.
Results
OS and PFS. At a median follow-up of 12 months, the Kaplan–Meier estimate of median OS for all 18 patients was 19 months, which was significantly longer than the RPA-adjusted historical controls (19 months vs. 7.5 months, p=0.001) (Figure 2A). The 1-year OS rate of all 18 patients was 58.0% [95%confidence interval (CI)=27.3%-79.6%]. The 1-year survival rate of 14 patients with RPA class V was 53.0% (95%CI=23.3%-75.9%), which was significantly longer than the historical controls (p=0.003) (Figure 2B). Median overall survival was not reached in 14 patients with RPA class V. The median PFS of all 18 patients was 9 months and the 1-year PFS was 43.9% (95%CI=15.3%-69.7%) (Figure 2C). The 1-year local PFS rate for PTVh was 76.2% (95%CI=38.2%-92.6%), while that for PTVl was 39.4% (95%CI=11.3%-67.2%), significantly higher than that for PTVl (p=0.03) (Figure 2D). When the tumor recurred, re-irradiation was repeated regardless of the history of radiation therapy until grade 3 or higher adverse events occurred.
Kaplan–Meier curves of postoperative patients with glioblastoma. (A) Kaplan–Meier estimates of probabilities of overall survival. The analysis included all the patients (solid line) who underwent multiparametric magnetic resonance imaging (MRI)-guided intensity-modulated radiation therapy with simultaneous integrated boost for glioblastoma. The median overall survival was significantly longer than that of the historical controls (dashed line) adjusted by the recursive partitioning analysis (19 months vs. 7.5 months, p=0.001). (B) Kaplan–Meier estimates for overall survival in patients with recursive partitioning analysis class V. The 1-year survival rate (solid line) was 53.0%, which was significantly longer than the historical controls (dashed line) (p=0.003). (C) Kaplan–Meier estimates of probabilities of progression-free survival. Considering all recurrences, the median progression-free survival was 9 months and the 1-year progression-free survival was 43.9% (95%CI=15.3%-69.7%). (D) Kaplan–Meier estimates of probabilities of local progression-free survival. The 1-year local progression-free survival rate for PTVh was 76.2%, while that for PTVl was 39.4%, which was significantly higher than that for PTVl (p=0.03).
Adverse events. During the radiotherapy period, all patients tolerated the treatment without grade 2 or higher adverse events. One patient (6%) developed grade 2 cerebral radiation necrosis 10 months after IMRT-SIB, which was treated with oral steroids and resolved without symptomatic neurological deficits.
Discussion
In this study, we demonstrated the feasibility and potential efficacy of multiparametric MRI-guided dose-escalated hypofractionated IMRT-SIB for glioblastoma. The high local control rate (76.2% after 1 year) for the enhancing residual tumors and the long OS rate (58.0% after 1 year) are promising. Our cohort is compatible with the previous phase 2 trial of hypofractionated IMRT (8). In the previous study, 8 fractions of irradiation were performed with total doses of 68, 40, and 32 Gy for three layered PTVs. The 2-year local control rate was 63.9% and the 1-year OS was 69.6% (8). The difference between the previous study and ours is that the total dose was lower in this study, the margins were smaller, and asymmetric margins were set using DIR and DTI images. A moderately high dose of 60 Gy in 15 fractions is likely to control tumors enhanced on post-gadolinium T1-weighted images, even with a small margin of 1-2 mm. On the other hand, the surrounding tumor-infiltrating areas, which are hyperintense on DIR images, are unlikely to be controlled by a dose of 40 Gy in 15 fractions. The 1-year local PFS rate for the surrounding tumor-infiltrating areas (PTVl) was 39.4% (95%CI=11.3%-67.2%). This implies that higher doses of radiation are required even in the surrounding infiltrates that are not contrast-enhanced.
The advantage of asymmetric margins using DIR and DTI images is that radiation-induced adverse events can be avoided, leaving the possibility of re-irradiation in case of recurrence. The smaller the margin and the lower the radiation dose to the normal brain, the higher the chance of salvage radiotherapy and the longer the survival rate. At present, there is no consensus on the optimal radiotherapy dose, or on the optimal margin settings. However, if we can image tumor-infiltrating areas by DIR and set the optimal margin using DTI images, we may be able to provide radiotherapy with higher conformality. To the best of our knowledge, this is the first report showing the potential effectiveness of multiparametric MRI-guided hypofractionated IMRT-SIB using DIR and DTI images (9).
There are several limitations in this study. First, IDH1 mutation and MGMT gene promoter methylation were not evaluated in this study (10). Since the purpose of this study was to present a proof of concept of multiparametric MRI-guided IMRT-SIB using DIR and DTI images, the absence of data on IDH1 mutation and MGMT gene promoter methylation is not a limitation. Further studies incorporating IDH1 mutation and MGMT gene promoter methylation would ensure the feasibility and efficacy of multiparametric MRI-guided IMRT-SIR. Second, this is a retrospective single institutional experience with glioblastoma with different RPA classes. A large-scale prospective randomized controlled trial will provide more detailed insights into the multiparametric MRI-guided IMRT-SIR.
Conclusion
In conclusion, a single-institutional cohort study cannot be generalized to others without further investigations, however, multiparametric MRI-guided dose-escalated IMRT-SIB with DIR and DTI imaging has the potential to improve local control rates without increasing adverse events.
Footnotes
Authors’ Contributions
Study concepts, study design, data analysis and interpretation, data acquisition, quality control of data, and manuscript preparation: Hama and Tate. Statistical analysis: Hama. Manuscript editing and review: Hama and Tate.
Conflicts of Interest
None of the Authors have any conflicts of interest to declare regarding this study.
- Received October 28, 2021.
- Revision received November 15, 2021.
- Accepted November 16, 2021.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.