Pre-Chemoradiotherapy FDG PET/CT cannot Identify Residual Metabolically-Active Volumes within Individual Esophageal Tumors.

OBJECTIVE: To study whether subvolumes with a high pre-chemoradiotherapy (CRT) FDG uptake could identify residual metabolically-active volumes (MAVs) post-CRT within individual esophageal tumors. Accurate identification will allow simultaneous integrated boost to these subvolumes at higher risk to improve clinical outcomes. METHODS: Twenty patients with esophageal cancer were treated with CRT plus surgery and underwent FDG PET/CT scans before and after CRT. The two scans were rigidly registered. Seven MAVs pre-CRT and four MAVs post-CRT within a tumor were defined with various SUV thresholds. The similarity and proximity between the MAVs pre-CRT and post-CRT were quantified with three metrics: fraction of post-CRT MAV included in pre-CRT MAV, volume overlap and centroid distance. RESULTS: Eight patients had no residual MAV. Six patients had local residual MAV (SUV ≥2.5 post-CRT) within or adjoining the original MAV (SUV ≥2.5 pre-CRT). On average, less than 65% of any post-CRT MAVs was included in any pre-CRT MAVs, with a low volume overlap <45%, and large centroid distance >8.6 mm. In general, subvolumes with higher FDG-uptake pre-CRT or post-CRT had lower volume overlap and larger centroid distance. Six patients had new distant MAVs that were determined to be inflammation from radiation therapy. CONCLUSIONS: Pre-CRT PET/CT cannot reliably identify the residual MAVs within individual esophageal tumors. Simultaneous integrated boost to subvolumes with high FDG uptake pre-CRT may not be feasible.

Beam controlled arc therapy--a delivery concept for stationary targets.

Volumetric modulated arc therapy (VMAT) presupposes that it is beneficial to deliver radiation from all beam angles as the gantry rotates, requiring the multi-leaf collimator to maintain continuity in shape from one angle to another. In turn, radiation from undesirable beam angles could compromise the dose distribution. In this work, we challenge the notion that the radiation beam must be held on as the gantry rotates around the patient. We propose a new approach for delivering intensity-modulated arc therapy, beam-controlled arc therapy (BCAT), during which the radiation beam is controlled on or off and the dose rate is modulated while the gantry rotates around the patient. We employ linear-programming-based dose optimization to each aperture weight, resulting in some zero weight apertures. During delivery, the radiation beam is held off at control points with zero weights as the MLC shape transits to the next non-zero weight shape. This was tested on ten head and neck cases. Plan quality and delivery efficiency were compared with VMAT. Improvements of up to 17% (p-value 0.001) and 57% (p-value 0.018) in organ-at-risk sparing and target dose uniformity, respectively, were achieved. Compared to the fixed number of apertures used in single-arc and double-arc VMAT, the BCAT used 109 and 175 apertures on average, respectively. The difference in total MUs for VMAT and BCAT plans was less than 4%. Plan quality improvement was confirmed after delivery with γ analysis resulting in over 99% agreement, or 4 in 1099 points that failed.

A surrogate-based metaheuristic global search method for beam angle selection in radiation treatment planning.

An important element of radiation treatment planning for cancer therapy is the selection of beam angles (out of all possible coplanar and non-coplanar angles in relation to the patient) in order to maximize the delivery of radiation to the tumor site and minimize radiation damage to nearby organs-at-risk. This category of combinatorial optimization problem is particularly difficult because direct evaluation of the quality of treatment corresponding to any proposed selection of beams requires the solution of a large-scale dose optimization problem involving many thousands of variables that represent doses delivered to volume elements (voxels) in the patient. However, if the quality of angle sets can be accurately estimated without expensive computation, a large number of angle sets can be considered, increasing the likelihood of identifying a very high quality set. Using a computationally efficient surrogate beam set evaluation procedure based on single-beam data extracted from plans employing equallyspaced beams (eplans), we have developed a global search metaheuristic process based on the nested partitions framework for this combinatorial optimization problem. The surrogate scoring mechanism allows us to assess thousands of beam set samples within a clinically acceptable time frame. Tests on difficult clinical cases demonstrate that the beam sets obtained via our method are of superior quality.

Magnetic collimation and metal foil filtering for electron range and fluence modulation.

We investigated the use of magnetically collimated electron beams together with metal filters for electron fluence and range modulation. A longitudinal magnetic field collimation method was developed to reduce skin dose and to improve the electron beam penumbra. Thin metal foils were used to adjust the energies of magnetically collimated electrons. The effects for different types of foils such as Al, Be, Cu, Pb, and Ti were studied using Monte Carlo calculations. An empirical pencil beam dose calculation model was developed to calculate electron dose distributions under magnetic collimation and foil modulation. An optimization method was developed to produce conformal dose distributions for simulated targets such as a horseshoe-shaped target. Our results show that it is possible to produce an electron depth dose enhancement peak using similar techniques of producing a spread-out Bragg peak. In conclusion, our study demonstrates new aspects of using magnetic collimation and foil filtration for producing fluence and range modulated electron dose distributions.

Dose homogeneity as a function of source activity in optimized I-125 prostate implant treatment plans.

PURPOSE: In conventional treatment planning for permanent I-125 prostate implants, it has been suggested that lower seed activities result in more homogeneous dose distributions and also less overdose of the critical structures. We sought to determine if this hypothesis holds by analyzing treatment plans constructed using an automated optimized approach. METHODS AND MATERIALS: We studied treatment plans for 10 patients using mixed-integer programming and the branch-and-bound method. Two mixed-integer models (that yielded somewhat different treatment plans) were developed: a "basic" model and a "dose homogeneity" model. For each resulting treatment plan, we examined dose homogeneity (by evaluating the dose non-uniformity ratio [DNR] and the full-width half-maximum [FWHM] of the differential dose-volume histogram [DVH]) as a function of three different source activities (0.35 mCi, 0.44 mCi, and 0.66 mCi). In addition, target coverage and critical structure dose distributions were evaluated. Plans using multiple source activities were also evaluated for resulting dose inhomogeneities. RESULTS: The homogeneity model results in a more homogeneous dose distribution than the basic model. DNR is lowered by an average of 42% (standard deviation [SD] = 19%), 39% (SD = 21%), and 33% (SD = 21%) for the 0.35 mCi, 0.44 mCi, and 0.66 mCi seeds, respectively, when the homogeneity model is employed over the basic model. Corresponding average decreases in the FWHM of the DVH for 0.35 mCi, 0.44 mCi, and 0.66 mCi, respectively, are 29 Gy (SD = 28 Gy), 24 Gy (SD = 22 Gy), and 27 Gy (SD = 13 Gy). Seeds of 0.35 mCi and 0.44 mCi result in the lowest DNR and narrower FWHM of the DVH relative to 0.66 mCi seeds. In general, the 0.44 mCi seeds produce greater target coverage and require fewer seeds and needles than the 0.35 mCi seeds. Although 0.66 mCi seeds result in the greatest target coverage, they yield highest critical structure doses. They also yield solutions requiring the least number of seeds and needles. However, the dose distributions from 0.66 mCi seeds are highly inhomogeneous. Multiple source activities in the same treatment plan produce dose distributions that are comparable in homogeneity to 0.44 mCi seed implants. CONCLUSIONS: Even when an optimization model that seeks to minimize dose inhomogeneity is employed, all factors involved in seed implants make 0.44 mCi the best activity choice in comparison with 0.35 mCi and 0.66 mCi.

Tissue mimicking materials for a multi-imaging modality prostate phantom.

Materials that simultaneously mimic soft tissue in vivo for magnetic resonance imaging (MRI), ultrasound (US), and computed tomography (CT) for use in a prostate phantom have been developed. Prostate and muscle mimicking materials contain water, agarose, lipid particles, protein, Cu++, EDTA, glass beads, and thimerosal (preservative). Fat was mimicked with safflower oil suffusing a random mesh (network) of polyurethane. Phantom material properties were measured at 22 degrees C. (22 degrees C is a typical room temperature at which phantoms are used.) The values of material properties should match, as well as possible, the values for tissues at body temperature, 37 degrees C. For MRI, the primary properties of interest are T1 and T2 relaxations times, for US they are the attenuation coefficient, propagation speed, and backscatter, and for CT, the x-ray attenuation. Considering the large number of parameters to be mimicked, rather good agreement was found with actual tissue values obtained from the literature. Using published values for prostate parenchyma, T1 and T2 at 37 degrees C and 40 MHz are estimated to be about 1,100 and 98 ms, respectively. The CT number for in vivo prostate is estimated to be 45 HU (Hounsfield units). The prostate mimicking material has a T1 of 937 ms and a T2 of 88 ms at 22 degrees C and 40 MHz; the propagation speed and attenuation coefficient slope are 1,540 m/s and 0.36 dB/cm/MHz, respectively, and the CT number of tissue mimicking prostate is 43 HU. Tissue mimicking (TM) muscle differs from TM prostate in the amount of dry weight agarose, Cu++, EDTA, and the quality and quantity of glass beads. The 18 microm glass beads used in TM muscle increase US backscatter and US attenuation; the presence of the beads also has some effect on T1 but no effect on T2. The composition of tissue-mimicking materials developed is such that different versions can be placed in direct contact with one another in a phantom with no long term change in US, MRI, or CT properties. Thus, anthropomorphic phantoms can be constructed.

An iterative sequential mixed-integer approach to automated prostate brachytherapy treatment plan optimization.

Conventional treatment planning for interstitial prostate brachytherapy is generally a 'trial and error' process in which improved treatment plans are generated by iteratively changing, via expert judgement, the configuration of sources within the target volume in order to achieve a satisfactory dose distribution. We have utilized linear mixed-integer programming (MIP) and the branch-and-bound method, a deterministic search algorithm, to generate treatment plans. The rapidity of dose falloff from an interstitial radioactive source requires fine sampling of the space in which dose is calculated. This leads to a large and complex model that is difficult to solve as a single 3D problem. We have therefore implemented an iterative sequential approach that optimizes pseudo-independent 2D slices to achieve a fine-grid 3D solution. Using our approach, treatment plans can be generated in 20-45 min on a 200 MHz processor. A comparison of our approach with the manual 'trial and error' approach shows that the optimized plans are generally superior. The dose to the urethra and rectum is usually maintained below harmful levels without sacrificing target coverage. In the event that the dose to the urethra is undesirably high, we present a refined optimization approach that lowers urethra dose without significant loss in target coverage. An analysis of the sensitivity of the optimized plans to seed misplacement during the implantation process is also presented that indicates remarkable stability of the dose distribution in comparison with manual treatment plans.