Collaboration with Virginia Commonwealth University Medical Physics Graduate Program (Samantha Conrad, Benjamin Lewis, Mark Ostyn, Siqiu Wang, Mattlew Riblett, Rebecca Mahon, and Patrick Brunick), Henry Ford Cancer Institute Residency Program (Akila Kumarasiri, Ilma Xhaferllari, Jennifer Dolan, Qixue Wu and Anthony Doemer), and William Beaumont – Dearborn (Cory Knill)

1. Sparing all salivary glands with IMRT for head and neck cancer: Longitudinal study of patient-reported xerostomia and head-and-neck quality of life
Hawkins et al. investigated how sparing all salivary glands affects the patient-reported outcome measurements (PROMs) of xerostomia. Xerostomia (dry mouth) is common in patients with head and neck cancer; 252 patients requiring treatment to the bilateral neck (with all-gland-sparing IMRT) answered approx. 600 questionnaires about their experience with xerostomia and their head and neck quality of life over a 60 month period. Utilizing both univariate analysis and multivariate analysis, results showed that reducing the dose to bilateral parotid gland (bPG), contralateral submandibular gland (cSMG), and oral cavity (OC) doses maximizes quality of life. Radiother Oncol. January 2018 Volume 126, Issue 1, Pages 68-74

2. Local recurrences after curative IMRT for HNSCC: Effect of different GTV to high-dose CTV margins
Zukauskaite et al. compared positions of local recurrences (LR) and failure rates of head and neck squamous cell carcinomas (HNSCC) in patients treated using IMRT with different GTV-CTV1 margins from three different clinics. 1576 patients treated with IMRT were followed-up with; after 41 months 272 patients had LR. The CT locations of the recurrences were compared to IMRT plans and the different GTV-CTV1 margins. The 95% prescription dose was evaluated as well. 51% of the LRs were located within the GTV, 83% received the prescribed dose regardless of the margins, and the local recurrence rate was not influenced by the differences in the margins; no correlations or statistical significances between the IMRT margins and LR locations were found. Radiother Oncol. January 2018 Volume 126, Issue 1, Pages 48-55

3. HyperArc VMAT planning for single and multiple brain metastases stereotactic radiosurgery: a new treatment planning approach
Ohira et al. investigated a new treatment planning approach available in the newly developed prototype treatment planning system from Varian. A retrospective study of 23 patients with 1–4 brain metastases who were treated with SRS was used to compare dosimetric parameters between retrospectively generated plans using the conventional VMAT (C-VMAT), or HyperArc VMAT (HA-VMAT). They reported that HA-VMAT provided higher homogeneity index, conformity index, and lower gradient index. Moderate-to-low dose volumes (4–16Gy) were significantly reduced but resulted in more complex MLC patterns and higher MU in HA-VMAT than C-VMAT. They demonstrated that the HA-VMAT planning method is superior in these dosimetric parameters to C-VMAT planning for patients with 1–4 brain metastases. Radiation Oncology. 2018 January; 13(13).

4. Evaluation of the tumor movement and the reproducibility of two different immobilization setups for image-guided stereotactic body radiotherapy of liver tumors
Dreher et al. evaluated the tumor movement and setup accuracy of patients using two types of immobilization setups during liver SBRT. 54 liver tumor patients were included in this study, 40 patients immobilized using a vacuum couch with low pressure foil, and 14 patients with abdominal compression. Immobilization efficacy was evaluated using the ratio of GTV to ITV, tumor movement in 4D CT scans, and movement in daily online adjustments after CBCT scans. The ratio of GTV to ITV was smaller for the low pressure foil system, indicating this method allows for more relative movement of the GTV in the ITV. The online adjustments and tumor movement in 4D CT scans were smaller using abdominal compression. The authors conclude that abdominal compression leads to a greater reduction in tumor motion than the low pressure foil, leading to higher accuracy during patient positioning. Radiation Oncology. 2018 January; 13(15).

5. Does intensity modulation increase target dose calculation errors of conventional algorithms for lung SBRT?
Zheng et al. investigated whether additional uncertainty was present in treatment planning lung SBRT IMRT cases with commonly used treatment planning algorithms compared to conformal techniques. Between two study groups, 20 cases were examined using type-A algorithms (pencil beam and ray tracing), and 20 were examined using type-B algorithms (collapsed cone, and analytical anisotropic). They found that for both types of algorithms, the difference between IMRT and conformal treatment was not statistically significant. The authors speculate that this is because the smallest fields (and most uncertain) are present at the center of the target (where the dose calculations algorithms are most certain). JACMP 2018 http://onlinelibrary.wiley.com/doi/10.1002/acm2.12266/full

6. Deep-inspiration breath-hold intensity modulated radiation therapy to the mediastinum for lymphoma patients: setup uncertainties and margins
Aristophanous et al. investigated the setup uncertainties and margins in IMRT for mediastinal lymphoma patients using DIBH. The residual errors were retrospectively measured, according to autoregistration, for the total PTV and 6 anatomic subregions in 3 directions. Large differences were found among various subregions and directions, most noticeably in the lower heart, neck, and axilla regions and in the superoinferior direction. In addition, 3 IGRT daily setup strategies were also examined: no IGRT, CBCT, and CT on rails (CTOR). Despite the measurable improvement in margin reduction from using IGRT, there was no clear conclusion whether CTOR offers an advantage over CBCT. Int J Radiat Oncol Biol Phys. 2018 Jan 1;100(1):254-62

7. Intensity modulated radiation therapy and second cancer risk in adults (Commentary)
Filippi et al. commented on the second cancer (SC) risk of IMRT in comparison with that of 3D-CRT. Citing recent radiobiological modeling studies, preclinical data, and preliminary clinical data, the authors suggest that there is at least an equivalence in SC induction risk between 3D-CRT and IMRT, with possible reduced risks of certain solid tumors for IMRT. The article also addresses the need for more clinical data in order to gain further insights. Int J Radiat Oncol Biol Phys. 2018 Jan 1;100(1):17-20

8. Technical Note: The impact of deformable image registration methods on dose warping
Qin et al. presented findings regarding the differences between organ doses warped by a purely image-based (IM-DIR) deformable image registration method and a novel biomechanical model-based (BM-DIR) deformable image registration method when applied for dose accumulation purposes. A retrospective study of 10 randomly selected patients, consisting of 5 Head and Neck (3x VMAT, 2x nine-field IMRT) and 5 Prostate cases (5x seven- or nine-field IMRT), compared the results of these two methods when used to register a CT taken after three weeks of treatment to the pre-treatment planning CT and subsequently warp the computed treatment doses from one image to the other. The researchers reported the dosimetric discrepancies between the two methods for soft-tissue organs (i.e. parotid, bladder and rectum) noting that greater discrepancies between the results of the conventional IM-DIR and BM-DIR approaches manifested in cases of large tumor volume shrinkage (e.g. the parotids). It was demonstrated that the BM-DIR method could produce a more realistic DVF around organs which experience large volume variation leading the researchers to suggest that radiotherapeutic applications which incorporate high dose gradients (e.g. IMRT and VMAT) to these sites could stand to benefit from its application. Medical Physics. January 2018 DOI: 10.1002/mp.12741.

9. Multi-GPU configuration of 4D intensity modulated radiation therapy inverse planning using global optimization
Hagan et al. designed a multi-GPU based particle swarm optimization algorithm for optimizing 4D-IMRT treatment plans using a vendor (Varian) specific GPU workstation and the Eclipse treatment planning system. The authors developed methods of applying deformable image registration on 10 phases with variable number of particles using sparse matrices to keep data within memory limitations of the system (CPU 256GB RAM and each GPU 12 GB RAM). They determined the maximum time benefit resulted from using 5 of the 8 available GPUs with increased down sampling and less particles in the swarm also reducing time. Time speed up beyond 5 GPUs was hampered by data transfer speeds and GPU clock rates. The 4D IMRT lung cancer plan using the authors approach (200 particles and 25 iterations) reduced dose to the organs at risk by as much as 26% of the max dose compared to the conventional IMRT plan. A plan with 5 GPUs and 50 particles took 35 minutes to plan. Phys in Med and Bio. 2018 Jan; 63(2)

10. The robustness of dual isocenter VMAT radiation therapy for bilateral lymph node positive breast cancer
Boman et al. compared VMAT treatment planning using dual isocenters vs. a single isocenter for treatment of lymph node positive bilateral breast cancer (BBC). The authors had previously noted that a split-arc VMAT technique reduced dose to the ipsilateral lung and heart at the expense of a higher dose to the contralateral breast when compared to conventional VMAT arcs or static fields for left or right sided treatments. The authors were interested in comparing the dosimetric differences between single and dual isocenter plans using this split-arc VMAT technique and investigating the robustness of the dual isocenter approach. The results showed a small dosimetric advantage when using the dual isocenter approach. Conformity index, D98PTV, V20LUNGS, V5LUNGS, DmeanLUNGS, and DmeanHEART all showed statistically significant differences in favor of dual isocenters. Dual isocenters did increase the number of MU's delivered by 16.6%. The robustness of dual isocenters was tested by applying 2 and 5 mm couch shift errors in the lateral, longitudinal, and vertical directions and looking at the CTV & PTV dose. This is to simulate potential errors when moving the couch from one isocenter to the other. The maximum decrease was 5% (2.5 Gy) for D98PTV and 1% (0.5 Gy) for D98CTV when 5 mm shift was applied. Physica Medica, December 2017, Volume 44, Pages 11-17

11. Auditing local methods for quality assurance in radiotherapy using the same set of predefined treatment plans
Seravalli et al. performed at TG-119 type study where they audited 21 Dutch radiotherapy centers. They extended the standardization past TG-119 by also creating standard radiotherapy plans, which were imported into the planning systems, recalculated, delivered, and analyzed using a standard auditing QA phantom(PTW's OCTAVIUS 3D) along with the local QA equipment. In 80% of the evaluated measurements the results of the local QA analysis matched the audit (both failed or both passed). The authors suggest the mismatches may have been largely caused by different QA measuring devices used by the institutions. One of the main hurdles in this study was developing a robust set of standardized plans that could be imported and delivered on the different institutional TPS-linac combinations, which eventually necessitated the generation of different audit plans for Varian and Elekta machines. https://www.sciencedirect.com/science/article/pii/S2405631617300799

12. Relationship between dosimetric leaf gap and dose calculation errors for high definition multi-leaf collimators in radiotherapy
Kim et al. compared the sweep-gap measured DLG values on a Varian Edge machine to the optimal DLG values found by minimizing the TPS-measured dose differences for 5 spine SRS cases. The calculated spine SRS doses were systematically lower when using the measured DLG values. Increasing the DLG value in the TPS resulted in average ion chamber measurement errors less than 1% [-2.2% to 2.3%] and film gamma pass rates (3%/3mm) greater than 97%. Furthermore, DLG values optimized using spine SRS cases also resulted in similar TPS-measurement agreements for TG119 test cases, lung and liver SBRT cases, and SRS brain cases. https://www.sciencedirect.com/science/article/pii/S2405631617300532

13. Retrospective dosimetry study of intensity-modulated radiation therapy for nasopharyngeal carcinoma: measurement-guided dose reconstruction and analysis
Sun et al. investigated a measurement-guided 3D dose reconstruction (3D-MGR) technique for QA'ing nasopharyngeal IMRT plans, and compared that with conventional phantom-based planar dosimetry (2D-PBD). 30 plans and their pre-treatment 2D-PBD data were analyzed. 3D-MGR achieved global gamma pass rates similar to conventional 2D-PBD. However, structure-specific gamma pass rates significantly decreased under stricter criteria, including the PTV. The average deviation of all inspected dose volumes and volumetric dose parameters ranged from − 2.93% to 1.17%, range [− 15.66%, +6.66%]. Consequently, authors caution that even upon passing the pre-treatment 2D-PBD QA, there could still be a risk of dose errors like under-dose in PTVs and overdose in critical structures, and 3D-MGR is recommended as the more clinically efficient verification for complicated nasopharyngeal IMRT. https://ro-journal.biomedcentral.com/articles/10.1186/s13014-018-0993-2

14. Fraction-variant beam orientation optimization for non-coplanar IMRT
O'Connor et al. compared fraction-variant to fraction invariant beam orientation optimization (BOO) for non-coplanar IMRT treatment planning. Fraction-variant plans were generated using a BOO formulated based on group sparsity that simultaneously optimizes the non-coplanar beam angles in all fractions. For each fraction, 500–700 candidate beams are used. The group sparsity encourages most candidate beams to be inactive to allow for different sets of beams to be active for different fraction. SBRT treatment plans consisting of five fractions were generated for a digital phantom, a prostate case, and a lung case, and a conventional thirty-fraction treatment plan was generated for a head and neck case. Non-coplanar IMRT requires a large number of beams to maximize dosimetric quality increasing treatment time. Fraction variant BOO addresses this disadvantage by considering fewer beams per fraction while a wide range of beam geometries are utilized over the course of treatment. Fraction variant BOO improved dosimetric quality of treatment plans by reducing the OAR mean dose and D2cm values on average by 3.3% and 3.8%, respectively and dose conformality increased or remained constant in all cases. Alternatively, dosimetric quality can be maintained while treatment time is reduced. https://www.ncbi.nlm.nih.gov/pubmed/29351088

15. Spatiotemporal radiotherapy planning using a global optimization approach
Adibi et al. investigated the therapeutic gain obtained by altering the radiation dose distribution for over different fractions using a spatiotemporally planning approach. Spatiotemporal treatment planning is achieved by integrating the fluence map optimization to maximize BED to the target volume and minimize the BED to critical structures. This optimization scheme led to a large scale non-convex problem solved using global optimization technique to obtain tight upper and lower bounds. Equivalent uniform BED was used to compare the results for two tumor sites, prostate and C-shape scenarios. During different fractions, different parts of the target volume are irradiated. Spatiotemporal treatment planning led to therapeutic gain by observing an increase in the BED to the target while the BED to normal tissue is reduced. https://www.ncbi.nlm.nih.gov/pubmed/29328046

16. Evaluation of a commercial automatic treatment planning system for liver stereotactic body radiation therapy treatments
Gallio et al. investigated the automated treatment planning module of the Pinnacle treatment planning system (TPS), specifically for liver stereotactic body radiation therapy (SBRT). Ten liver SBRT cases were studied via six treatment plans. Four of the treatment plans were manual in nature (two Pinnacle manual module and two Monaco TPS) and the remaining two were automated (Pinnacle auto-planning module). Two plans (from different planners) from each TPS were developed to study user dependency. The metrics showing statistically significant differences between Pinnacle and Monaco planning techniques were: the plan average beam irregularity, number of segments, and monitor units. The metrics showing statistically significant differences between manual and automated planning techniques were the spinal cord doses and human resource planning time. Ultimately, the authors concluded that the automated planning technique generated clinically acceptable plans, half of which were preferred to the manually generated plans with the benefits of decreased human resource planning time and increased consistency in plan quality. The most complex and challenging cases still required human skills and a manual approach. https://doi.org/10.1016/j.ejmp.2018.01.016

17. Quantifying the effect of 3T Magnetic Resonance Imaging residual system distortions and patient-induced susceptibility distortions on radiation therapy treatment planning for prostate cancer
Adjeiwaah et al. investigated the effect of magnetic resonance system and patient-induced susceptibility distortions from a 3T scanner on dose distributions for prostate cancer. Combined displacement fields from the distortions were used to distort 17 prostate patient CT images. VMAT dose plans were initially optimized on distorted CT images and the plan parameters transferred to the original patient CT images to calculate a new dose distribution. Maximum residual mean distortions of 3.19 mm at a radial distance of 25 cm and maximum mean patient-induced susceptibility shifts of 5.8 mm were found using the lowest bandwidth (122 Hz per pixel). Dose difference between the distorted and undistorted images was less than 0.5%. Adjeiwaah et al. concluded that Patient-induced susceptibility distortions at high field strengths in closed bore magnetic resonance scanners are larger than residual system distortions after using vendor-supplied 3-dimensional correction for the delineated regions studied. However, errors in dose due to disturbed patient outline and shifts caused by patient-induced susceptibility effects are below 0.5%. https://doi.org/10.1016/j.ijrobp.2017.10.021