Eva Oberacker1, Cecilia Diesch1, Jacek Nadobny2, Andre Kuehne3, Thomas Wilhelm Eigentler1, Pirus Ghadjar2, Peter Wust2, and Thoralf Niendorf1,3,4
1Berlin Ultrahigh Field Facility (B.U.F.F.), Max Delbrück Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany, 2Clinic for Radiation Oncology, Charité Universitätsmedizin, Berlin, Germany, 3MRI.TOOLS GmbH, Berlin, Germany, 4Experimental and Clinical Research Center (ECRC), joint cooperation between the Charité Medical Faculty and the Max-Delbrück-Center for Molecular Medicine in the Helmholtz Association, Berlin, Germany
Synopsis
Ultrahigh-field MR employs higher radio frequencies
(RF) than conventional MR and has unique potential to provide focal temperature
manipulation and high resolution imaging in an integrated device (ThermalMR).
The advantage of integrated therapy monitoring and guidance benefits thermal
interventions in brain tumor treatments. To approach this goal this work
examines the inter-patient variability in a heterogeneous group of glioma multiforme
patients using EMF simulations. Our findings
provide useful indicators as potential patient inclusion criteria for thermal
treatment of brain tumors and form the technological basis for treatment
planning and RF applicator developments en
route to clinical applications of Thermal MR.
Introduction
Hyperthermia has proven beneficial as an adjunct treatment
of Glioblastoma Multiforme (GBM)1. Advances in RF based thermal
interventions have evoked developments in the design of annular2,3
or helmet shaped4 RF applicators for the treatment of brain tumors.
Optimization algorithms used to confine the RF power deposition to the target
volume (TV) are under constant revision5,6. The treatability of
tumors strongly depends on a successful employment, monitoring and guidance of
treatment planning algorithms as well as on suitable RF applicator design. Recognizing
this challenge and opportunity, this EMF simulation study investigates the
subject dependent thermal intervention plans (TIP) for six GBM patient-based realistic
voxel models. For this purpose ten RF applicator configurations7 are
examined using two phase and amplitude optimization algorithms8 for ThermalMR
at 297 MHz.Methods
Six realistic patient voxel models were generated based on the
delineation of the gross tumor volume (GTV), the clinical target volume (CTV),
major anatomical structures and organs at risk
9. Electromagnetic
field simulations were performed for each combination of patient model and RF
applicator design using Sim4Life
10.
From the channel-wise E- and H-field data, SAR
10g11
matrices were calculated
12. The two algorithms for phase and
amplitude optimization aim at maximizing the RF power delivered to the target
volume (TV), P
TV 8. For algorithm 1,
maximizing P
TV is the sole optimization goal (Power Optimization) while algorithm 2 has the additional constraint
to homogenize the RF power distribution over the TV (Uniformity Optimization).
The RF applicator configurations
7 under
investigation (Figure1) differ in a) the number of RF channels available for
shaping the RF power deposition [8/16/32], b) the coverage of the head and thus
focusing capabilities along the head-feet direction (z-direction) [planar/interleaved/two
rows], c) the conformity to the human head [ring/elliptical] and d) the
presence/absence of a water bolus for RF coupling from the antennae to the head.
The metrics chosen to analyze the TIP are
- SAR10g,max(TV): maximum
SAR10g(TV)
- SAF:
SAR amplification factor = ratio between SAR10g,mean(TV) and SAR10g,mean(healthy)
- TCSAR>Lim:
target coverage with SAR10g≥40W/kg
- P(TV)/V(TV):
power delivered to the TV per ml target volume
- THQ:
target-to-hotspot-quotient = ratio between SAR10g,mean(TV) and
highest 1vol% SAR10g(healthy).
For all metrics, higher values indicate a better prospective
treatment plan. The TIPs for each patient model were analyzed regarding the i)
performance of the optimization algorithms, ii) number of RF channels and iii) brain
coverage in z-direction. In order to draw conclusions about potential inclusion
criteria for given brain tumor sites and sizes, all values are normalized to
the inter-patient maximum of the single metric.
Results & Discussion
The patient-specific TIPs obtained for all RF applicator
configurations for Power and Uniformity Optimization are summarized in Figure2a)
with the absolute values of the metrics investigated being summarized in Table1.
The Power Optimization (red) shows an overall better performance than the
Uniformity Optimization (blue), often even for TCSAR>Lim. Further
analysis will thus concentrate on the Power Optimization.
The intra-patient dependency on the RF applicator design is
addressed in Figure2b). It can be appreciated that coverage of the head and
focusing capabilities in z-direction are not sufficient for the planar, single
ring RF applicator designs (1.1.R and 2.1.R, red), leading to a suboptimal TIP.
Moving from the 16-channel arrangements with large brain coverage (2.3.R/2.3.E)
to the fully stacked 32-channel designs (3.1.R/3.1.E) did not enhance the TIP
for any of the patient models. This finding benefits RF applicator developments
because engineering constraints would be substantially offset by limiting the RF
channel count to 16.
Figure3 shows the comparison of the 16 channel designs for
all tumor models. The larger the TV, the more use can be made of the vertical
extension of the RF applicators (green and blue lines for patients3,5&6).
While the presence of a water bolus (orange) seems to be advantageous for the
superficial TV in Patient5, it yields worst TIP for the deep seated tumor in
Patient3.
A particular challenge was faced when computing the TIP for
Patient6 who exhibits two separate TVs in close proximity. For this case, Uniformity
Optimization yielded a solution of two excitation vectors, causing separate SAR10g
maxima but poor coverage of the larger TV. Power Optimization resulted in a
much better coverage of the larger TV but did not achieve any increased RF
exposure in the smaller TV. Treating the TVs separately did not enable better
TIPs.Conclusion
Including a heterogeneous group of patients in the TIP and a
range of RF applicator configurations into our EMF simulations has generated
valuable insights. Our results demonstrate that i) the Power Optimization
consistently outperforms the Uniformity Optimization, ii) planar, single ring arrangements
of RF antennae do not provide sufficient control over the RF power delivery to
treat high risk regions such as the brain and iii) the TIP enhancement effect
of increasing the RF channel count plateaus, after which it does not outweigh
the engineering effort.
To conclude, our optimization algorithms and RF applicator
designs favor target volumes where the TV’s extent along the head-feet
direction approaches half of the wavelength (λ/2) in brain or tumor tissue. For
the TIP of small or separate target volumes it is conceptually appealing to
pursue the use of multiple and/or higher RF frequencies to enhance thermal
intervention planning.Acknowledgements
This project has received funding from the European Research
Council (ERC) under the European Union's Horizon 2020 research and innovation
program under grant agreement No 743077 (ThermalMR).References
1. Sneed,
P.K., et al., Survival benefit of
hyperthermia in a prospective randomized trial of brachytherapy
boost±hyperthermia for glioblastoma multiforme. International Journal of
Radiation Oncology* Biology* Physics, 1998. 40(2): p. 287-295.
2. Winter,
L., et al., Design and Evaluation of a
Hybrid Radiofrequency Applicator for Magnetic Resonance Imaging and RF Induced
Hyperthermia: Electromagnetic Field Simulations up to 14.0 Tesla and
Proof-of-Concept at 7.0 Tesla. PLoS One, 2013. 8(4): p. e61661.
3. Winter,
L., et al., Thermal magnetic resonance:
physics considerations and electromagnetic field simulations up to 23.5 Tesla
(1GHz). Radiation Oncology, 2015. 10(1):
p. 1.
4. Dobsicek
Trefna, H., et al. Antenna Applicator for
Microwave Hyperthermia Treatment of Pediatric Brain Cancer. in 8th European Conference on Antennas and
Propagation, EuCAP 2014, The Hague, The Netherlands 6-11 April 2014. 2014.
5. Rijnen,
Z., et al., Clinical integration of
software tool VEDO for adaptive and quantitative application of phased array
hyperthermia in the head and neck. International Journal of Hyperthermia,
2013. 29(3): p. 181-193.
6. Bellizzi,
G.G., et al., Selecting the Optimal
Subset of Antennas in Hyperthermia Treatment Planning. IEEE Journal of
Electromagnetics, RF and Microwaves in Medicine and Biology, 2019: p. 1-1.
7. Oberacker E. et al., Radiofrequency applicator concepts for
simultaneous MR imaging and hyperthermia treatment of glioblastoma multiforme:
A 7.0 T (298 MHz) study, Proceedings of the 25th ISMRM Annual Meeting and
Exhibition, Honolulu, Hawaii, USA
8. Oberacker E. et al., Power Considerations for Radiofrequency
Applicator Concepts for Thermal Magnetic Resonance Interventions in the Brain
at 297 MHz, Proceedings of the 27th ISMRM Annual Meeting and
Exhibition, Montreal, Canada, 2019
9. Nadobny, J., et al. Fast and efficient generation of patient
models for hyperthermia based on radiation therapy contours. in 32nd Annual
Meeting of the European Society for Hyperthermic Oncology. 2018. Berlin:
Strahlentherapie und Onkologie.
10. Sim4Life V3.4,
ZurichMedTech, Zurich, Switzerland
11. IEEE, Recommended Practice for
Determining the Spatial‐Peak Specific Absorption Rate (SAR) in the human body
Due to Wireless Communications Devices: Measurement Techniques. 2003..
12. MatLab, The MathWorks
Inc., Natick, Massachusetts, US