Purpose:To demonstrate a patient specific, image‐guided quality assurance method that tests both dosimetric and geometric accuracy for single‐isocenter multiple‐target VMAT radiosurgery (SIMT‐VMAT‐SRS)Method:We used a new film type, EBT‐XD (optimal range 0.4–40Gy), and an in‐house PMMA phantom having a coronal plane for film and a 0.125 cm3 ionization chamber (IC). The phantom contained fiducial features for kV image guided setup and for accurate film marking. Five patient plans with multiple targets sizes ranging from 3 to 21mm in diameter and prescribed doses from 14 to 18 Gy were selected. Two verification plans were generated for each case with the film plane passing through the center of the largest and smallest targets. For the four largest targets we obtained an IC measurement. For each case, a calibration film was irradiated using a custom designed step pattern. The films were scanned using a flatbed color scanner and converted to dose using the calibration film and the three channel calibration method. Image registration was performed between film and treatment planning system calculations to evaluate the geometric accuracy.Results:The mean registration vector had an average magnitude of 0.47 mm (range from 0.13mm to 0.64 mm). For the four largest targets, the mean ratio of the IC and film measurement to expected dose was 0.990 (range 0.968 to 1.009) and 1.032 (1.021 to 1.046), respectively. The fraction of pixels having gamma index < 1 for criteria of 3%/3mm, 3%/2mm, 3%/1mm was 98.8%, 97.5% and 87.2% before geometric registration and 99.1%, 98.3% and 94.8% after registration.Conclusion:We have demonstrated an image‐guided QA method can assess both geometric and dosimetric accuracy. The phantom was positioned with sub‐millimeter accuracy. Absolute film dosimetry using EBT‐XD film was sufficiently accurate for assessment of dose to multi‐targets too small for IC measurement in SRS VMAT plans.

Purpose:To quantified the dosimetric characteristics of a Calypso compatible radiotherapy couch using continuously scanned transmission factor as a function of gantry angle.Methods:Radiation dose as a function of gantry angle was measured with an ionization chamber at the center of a 30 cm diameter cylindrical phantom placed on Qfix kVue Calypso compatible couch. A 3‐D scanner was employed to continuously scan the dose during an arc rotation delivery. Nearly continuous transmission curves of 0.03° step size were obtained from IEC gantry angle 90° to180° for 6 MV and 15 MV photon beams at three couch longitudinal positions (due to longitudinally non‐uniform shape of the couch rails) with rails both at the center and on sides. The transmission curves were evaluated to assess the attenuation by the couch. They were also compared to those of a carbon fiber couch top.Results:The scanned transmission curves of the Calypso comptible couch showed fine details of the couch transmission as a function of gantry angle. The attenuation by the couch varied from 2.4% to 12.9% for the 6 MV photo beam and from 1.8% to 8.9% for the 15 MV photon beam depending on the gantry angle. The attenuation by the flat couch top is only approximately 2 to 3%, but by the rail is much more significant. Because the rails are not uniformly shaped in the longitudinal direction, attenuation by the rail for the same gantry angle varied with the rail longitudinal position. The transmission curves of the carbon fiber couch top are nearly identical to those of the Calypso compatible couch top.Conclusion:The Calypso compatible couch attenuates radiation by 2.4%–12.9% and 1.8%–8.9% for 6 MV and 15 MV photon beams, respectively. The attenuation by the rails is much more significant than the flat couch top.

Purpose: To evaluate the accuracy of single-isocenter multiple-target VMAT radiosurgery (SIMT-VMAT-SRS) by analysis of pre-treatment verification measurements. Methods: Our QA procedure used a phantom having a coronal plane for EDR2 film and a 0.125 cm3 ionization chamber. Film measurements were obtained for the largest and smallest targets for each plan. An ionization chamber measurement (ICM) was obtained for sufficiently large targets. Films were converted to dose using a patient-specific calibration curve and compared to treatment planning system calculations. Alignment error was estimated using image registration. The gamma index was calculated for 3%/3 and 3%/1 mm criteria. The median dose in the target region and, for plans having an ICM, the average dose in the central 5 mm was calculated. Results: The average equivalent target diameter of the 48 targets was 15 mm (3–43 mm). Twenty of the 24 plans had an ICM for the plan corresponding to the largest target (diameter 11–43 mm) with a mean ratio of chamber reading to expected dose (ED) and the mean ratio of film to ED (averaged over the central 5 mm) was 1.001 (0.025 SD) and 1.000 (0.029 SD), respectively. For all plans, the mean film to ED (from the median dose in the target region) was 0.997 (0.027 SD). The mean registration vector was (0.15,0.29) mm, with an average magnitude of 0.96 mm. Before (after) registration, the average fraction of pixels having gamma < 1 was 99.3% (99.6%) and 89.1% (97.6%) for 3%/3mm and 3%/1mm, respectively. Conclusion: Our results demonstrate dosimetric accuracy of SIMT-VMAT-SRS for targets as small as 3 mm. Film dosimetry provides accurate assessment of the absolute dose delivered to targets too small for an ionization chamber measurement; however, the relatively large registration vector indicates that image-guidance should replace laser-based setup for patient-specific evaluation of geometric accuracy.

PurposeFiducial‐based imaging is often used in IGRT. Traditional gold fiducial marker often has substantial reconstruction artifacts. These artifacts Result in poor image quality of DRR for online kV‐to‐DRR matching. This study evaluated the image quality of PEEK in DRR in static and moving phantom.MethodsCT scan of the Gold and PEEK fiducial (both 1×3 mm) was acquired in a 22 cm cylindrical phantom filled with water. Image artifacts was evaluated with maximum CT value deviated from water due to artifacts; volume of artifacts in 10×10 cm in the center slice; maximum length of streak artifacts from the fiducial. DRR resolution were measured using FWHM and FWTM. 4DCT of PEEK fiducial was acquired with the phantom moving sinusoidally in superior‐inferior direction. Motion artifacts were assessed for various 4D phase angles.ResultsThe maximum CT value deviation was −174 for Gold and −24 for PEEK. The volume of artifacts in a 10x10 cm 3 mm slice was 0.369 for Gold and 0.074 cm3 for PEEK. The maximum length of streak artifact was 80mm for Gold and 7 mm for PEEK. PEEK in DRR, FWHM was close to actual (1.0 mm for Gold and 1.1 mm for PEEK). FWTM was 1.8 mm for Gold and 1.3 mm for PEEK in DRR. Barrel motion artifact of PEEK fiducial was noticeable for free‐breathing scan. The apparent PEEK length due to residual motion was in close agreement with the calculated length (13 mm for 30–70 phase, 10 mm in 40–60 phase).ConclusionStreak artifacts on planning CT associated with use of gold fiducial can be significantly reduced by PEEK fiducial, while having adequate kV image contrast. DRR image resolution at FWTM was improved from 1.8 mm to 1.3 mm. Because of this improvement, we have been routinely use PEEK for liver IGRT.

Purpose:To measure the spatial accuracy of dose delivery under conebeam CT guidance to small targets using static arcs with MLC‐defined fields.Methods:A plastic (PMMA) phantom simulating a small brain lesion was constructed. The lesion was a cylindrical air cavity having 3 mm length and 3.175 mm diameter. Gafchromic EBT 2 film passing through the cavity was pin pricked at the exact geometric cavity center. Following CT simulation treatment plans involving static arcs with MLC‐defined fields were designed, similar to radiotherapy with circular tertiary cones. 5×2 mm fields were generated by opening two centrally located leaf pairs of a high‐definition MLC to 2.0 mm. The target was irradiated with 6 MV photons on a STx linac under conebeam CT image guidance to 600 cGy at isocenter. Films from coronal and sagittal planes were scanned and evaluated using the ImageJ software. The distance between the centroid of the 50% isodose line and the center of the pin prick was the measure of geographic accuracy.Results:Single arcs resulted in ellipsoidal 50% isodose surfaces having diameters of 5.0 mm along the gantry axis and 7.7 mm in the cross‐sectional directions. 4‐arc delivery at couch angles of 0, 45, 90 and 315 degrees produced 50% isodose surfaces having diameters of 5.6, 5.7 and 7.5 mm diameters along the lateral, longitudinal and anterior‐posterior directions, respectively. In any of the experiments, the largest distance by which the centroid of the 50% isodose line missed the center of the target was 0.36 mm or less.Conclusion:MLC‐defined static arcs delivered under conebeam CT guidance provide precise irradiation of small targets. The method is very efficient since no special attachments or modifications of the accelerator are required. Since MLC leaves are stationary throughout the treatment, patient‐specific QA involving phantoms should not be required.

Purpose:To determine the coefficients of bi‐ and tri‐exponential functions for the best fit of radial dose functions of the new iodine brachytherapy source: Iodine‐125 Seed AgX‐100.Methods:The particle swarm optimization (PSO) method was used to search for the coefficients of the biand tri‐exponential functions that yield the best fit to data published for a few selected radial distances from the source. The coefficients were encoded into particles, and these particles move through the search space by following their local and global best‐known positions. In each generation, particles were evaluated through their fitness function and their positions were changed through their velocities. This procedure was repeated until the convergence criterion was met or the maximum generation was reached. All best particles were found in less than 1,500 generations.Results:For the I‐125 seed AgX‐100 considered as a point source, the maximum deviation from the published data is less than 2.9% for bi‐exponential fitting function and 0.2% for tri‐exponential fitting function. For its line source, the maximum deviation is less than 1.1% for bi‐exponential fitting function and 0.08% for tri‐exponential fitting function.Conclusion:PSO is a powerful method in searching coefficients for bi‐exponential and tri‐exponential fitting functions. The bi‐ and tri‐exponential models of Iodine‐125 seed AgX‐100 point and line sources obtained with PSO optimization provide accurate analytical forms of the radial dose function. The tri‐exponential fitting function is more accurate than the bi‐exponential function.

Purpose: Although radiation attenuation by the treatment couch can be included in the calculation of radiotherapy dose, difference between planned and actual treatment couch positions can generate significant dose discrepancies. We propose a method to predict and correct the dosimetric effect of the couch in actual treatment position. Methods: The couch transmission factor, T, varies with beam angle, G, couch lateral position, x, and vertical position, y, i.e., T=T(x,y,G). If T(x,y,G) is known for a fixed couch vertical position y=h, the transmission of central‐axis beam (CAX) T(x,y,G) can be obtained by T(x,y,G)=T(x + ,h,G), where x + =x‐(y‐h)tan(G) and G is the angle between the beam and the vertical axis. Similarly, the transmission of any off‐CAX point can be obtained using a similar formula. We measured CAX couch transmission at a fixed couch vertical position over the couch lateral motion range for all gantry angles by continuously scanning rotating arc beams. A 2D couch transmission correction matrix can thus be generated from T(x,h,G) for each treatment field for the actual couch position. By applying the transmission correction matrix to the planned field dose, the couch effect can be predicted and corrected. To verify this method, we measured couch transmission T(x, y=10cm, G=225°)(225°=IEC 135°) and compared to that obtained from equivalent T(x + , y=3cm, G=225°) over the range of lateral motion with a step size of 2 cm . Results: The measured couch transmission factors T(x, y=10cm, G=225°) are in excellent agreement with those obtained from the equivalent T(x + , y=3cm, G=225°). The mean difference is 0.00406±0.00135. Conclusion: The couch transmission correction matrix for any couch position and beam angle can be obtained from one set of scanning measurements at a fixed couch vertical position. The dosimetric effect of the treatment couch can be predicted and corrected by applying the couch transmission correction to the planned dose.

Deep inspiration breath hold (DIBH) can reduce irradiated cardiac volume during tangential left breast radiation therapy. Commonly the setup position for DIBH is based on gated on-board orthogonal kV imaging (OBI) of the rib cage. Position of tumor bed can vary relative to the rib cage, especially for large breasts. Accurate tumor bed localization is critical for a lumpectomy bed boost in which a conformal approach is preferred. The current study assessed position and target coverage variation of the tumor bed in boost treatment aligned to the rib cage using gated orthogonal kV image in DIBH radiation therapy. At a single institution, 12 consecutive patients underwent left breast radiation therapy of 45 Gy and followed by a boost dose of 16 Gy to the tumor bed. After setup positions were aligned to the rib cage using gated orthogonal kV imaging, feedback-guided (video-based RPM system) CBCTs were acquired in two 30 sec or three 20 sec DIBH on the first day of boost treatment. The reference planning CT and CBCT was aligned to the rib cage, and positional difference of the tumor bed between the two was recorded. The distance between target center and lung/chest wall boundary ("target distance") was determined by 3D margin expansion in planning CT. Changes in target volume covered by 95% of prescription dose (TV95), due to positional changes, were calculated. Statistical difference between the two TV95 was evaluated using paired t-test. Mean left breast volume was 1391 (range 611-3113) cm3. Mean "target distance" was 4.5 (1.3-9.4) cm. Mean positional change in tumor bed was 0.6 (0-1.5) cm lateral, 0.7 (0-2.1) cm vertical, and 0.4 (0-1.0) cm longitudinal, resulting in 1.1 (0.4-2.5) cm 3D displacement. Mean reduction of TV95 due to displacement was 8.4% (0.0-38.3%) and was statistically significant (p < 0.03). Magnitude of the tumor bed displacement tend to increase with the breast volume (r = 0.75) and with the "target distance" (r = 0.76). Positional change greater than 1 cm occurred in lateral or vertical direction when breast volume was greater than 2000 cm3 or "target distance" was greater than 6 cm. Position of tumor bed can vary relative to the rib cage in DIBH radiation therapy, and significantly affect boost dose to the tumor bed. For patients with breast volumes greater than 2000 cm3 or "target distance" greater than 6 cm, setup based on feedback-guided CBCT or implanted fiducial markers is recommended; otherwise a PTV margin expansion of 2cm in lateral and vertical direction may be needed for the tumor bed.

Purpose:Dose calculation software is thoroughly evaluated when it is commissioned; however, evaluation of periodic software updates is typically limited in scope due to staffing constraints and the need to quickly return the treatment planning system to clinical service. We developed a tool for quickly and comprehensively testing and documenting dose calculation software against measured data.Methods:A tool was developed using MatLab (The MathWorks, Natick, MA) for evaluation of dose calculation algorithms against measured data. Inputs to the tool are measured data, reference DICOM RT PLAN files describing the measurements, and dose calculations in DICOM format. The tool consists of a collection of extensible modules that can perform analysis of point dose, depth dose curves, and profiles using dose difference, distance‐to‐agreement, and the gamma‐index. Each module generates a report subsection that is incorporated into a master template, which is converted to final form in portable document format (PDF).Results:After each change to the treatment planning system, a report can be generated in approximately 90 minutes. The tool has been in use for more than 5 years, spanning 5 versions of the eMC and 4 versions of the AAA. We have detected changes to the algorithms that affected clinical practice once during this period.Conclusion:Our tool provides an efficient method for quality assurance of dose calculation software, providing a complete set of tests for an update. Future work includes the addition of plan level tests, allowing incorporation of, for example, the TG‐119 test suite for IMRT, and integration with the treatment planning system via an application programming interface. Integration with the planning system will permit fully‐automated testing and reporting at scheduled intervals.

Purpose: Current QA procedures verify accuracy of individual equipment parameters, but may not include CT and MRI localizers. This study uses an end-to-end approach to measure the overall targeting errors in individual patients previously treated for trigeminal neuralgia. Methods: The trigeminal nerve is simulated by a 3 mm long, 3.175 mm (1/8 inch) diameter MRI contrast-filled cavity embedded within a PMMA plastic capsule. The capsule is positioned within the head frame such that the cavity position matches the Gamma Knife coordinates of 10 previously treated patients. Gafchromic EBT2 film is placed at the center of the cavity in coronal and sagittal orientations. The films are marked with a pin prick to identify the cavity center. Treatments are planned for delivery with 4 mm collimators using MRI and CT scans acquired with the clinical localizer boxes and acquisition protocols. Coordinates of shots are chosen so that the cavity is centered within the 50% isodose volume. Following irradiation, the films are scanned and analyzed. Targeting errors are defined as the distance between the pin prick and the centroid of the 50% isodose line. Results: Averaged over 10 patient simulations, targeting errors along the x, y and z coordinates (patient left-to-right, posterior-anterior, head-to-foot) were, respectively, −0.060 +/− 0.363, −0.350 +/− 0.253, and 0.364 +/− 0.191 mm when MRI was used for treatment planning. Planning according to CT exhibited generally smaller errors, namely 0.109 +/− 0.167, −0.191 +/− 0.144, and 0.211 +/− 0.94 mm. The largest errors in MRI and CT planned treatments were, respectively, y = −0.761 and x = 0.428 mm. Conclusion: Unless patient motion or stronger MRI image distortion in actual treatments caused additional errors, all patients received the prescribed dose, i.e., the targeted section of the trig±eminal nerve was contained within the 50% isodose surface in all cases.

Purpose: The dosimetric effect of the treatment couch is non-negligible in today's radiotherapy treatment. To accurately include couch in dose calculation, we investigated the dependence of couch transmission factors on field size and couch-isocenter distance. Methods: Couch transmission factors for Varian Exact Couch were determined by taking the ratios of ionization of a posterior-anterior beam with and without the couch in the beam path. Measurements were performed at the isocenter using a PTW cylindrical ionization chamber (Model 31030) with an Aluminum buildup cap of 1.1 cm thick for the 6 MV photon beam. Ionization readings for beam sizes ranging from 2 × 2 cm2 to 40 × 40 cm2 were taken. Transmission factors for couch-isocenter distances ranging from 3 cm to 20 cm were also investigated. Results: The couch transmission factors increased with the field size approximately in an exponential manner. For the field sizes that we tested, the transmission factor ranged from 0.976 to 0.992 for couch-isocenter distance of 3 cm. The transmission factor was also monotonically dependent on couch-isocenter separation distance, but in a lighter magnitude. For the tested couch heights, the transmission factor ranged from 0.974 – 0.972 for 2 × 2 cm2 field size and 0.992 – 0.986 for 40 × 40 cm2 field size. The dependence on couch-isocenter distance is stronger for larger field size. Conclusions: The transmission factor of a radiotherapy treatment couch increases with field size of the radiation beam and its distance from the isocenter. Such characterization of the couch transmission factor helps improve the accuracy of couch modeling for radiotherapy treatment planning.

Purpose:To develop a practical device having sufficient accuracy for daily QA tests of accelerators used for SRS and SBRT.Methods:The UAB (Universal Alignment Ball) consists of a 6.35 mm (1/4 inch) diameter tungsten sphere located concentrically within a 25.4 mm (1 inch) diameter acrylic plastic (PMMA) sphere. The spheres are embedded in polystyrene foam, which, in turn, is surrounded by a cylindrical PMMA shell. The UAB is placed on the couch and aligned with wall lasers according to marks that have known positions in relation to the center of the spheres. Using planar and cone beam images the couch is shifted till the surface of the PMMA sphere matches Eclipse‐generated circular contours. Anterior and lateral MV images taken with small MLC openings allow measurement of distance between kV and MV isocenter, laser and MLC alignment. Measurements were taken over a one‐month period.Results:Artifacts from the tungsten sphere were confined within the PMMA sphere and did not affect cone beam localization of the sphere boundary, allowing 0.1 mm precise alignment with a computer‐generated circle centered at kV isocenter. In tests extending over a one‐month period, the distance between kV and MV isocenters along the vertical, longitudinal and lateral directions was 0.125 +/−0.06, 0.19 +/−0.08, and 0.02 +/−0.08 mm, respectively. Laser misalignment along these directions was 0.34 +/‐ 0.15, 0.74 +/−0.29, and 0.49 +/−0.22 mm. Automated couch shifts moved the spheres to within 0.1 mm of the selected position. The center of a 1cmx1cm MLC‐defined field remained within +/−0.2 mm of the tungsten sphere center as the gantry was rotated.Conclusion:The UAB is practical for daily end‐to‐end QA tests of accelerator alignment. It provides tenths‐mm accuracy for measuring agreement of kV and MV isocenters, couch motions, gantry flex and laser alignment.

Purpose:HDR brachytherapy using interstitial needle template for cervical cancer is commonly delivered in 4‐5 fractions. Routine verification of needle positions before each fraction is often based on radiographic imaging of implanted fiducial markers. The current study evaluated interfractional displacement of implanted fiducial markers using CT images.Methods:9 sequential patients with cervical interstitial needle implants were evaluated. The superior and inferior borders of the target volumes were defined by fiducial markers in planning CT. The implant position was verified with kV orthogonal images before each fraction. A second CT was acquired prior 3rd fraction (one or 2 days post planning CT). Distances from inferior and superior fiducial markers to pubic symphysis plane (perpendicular to vaginal obtulator)were measured. Distance from needle tip of a reference needle (next to the inferior marker) to the pubic symphysis plane was also determined. The difference in fiducial marker distance or needle tip distance between planning CT and CT prior 3rd fraction were measured to assess markers migration and needle displacement.Results:The mean inferior marker displacement was 4.5 mm and ranged 0.9 to 11.3 mm. The mean superior marker displacement was 2.7 mm and ranged 0 to 10.4 mm. There was a good association between inferior and superior marker displacement (r=0.95). Mean averaged inferior and superior marker displacement was 3.3 mm and ranged from 0.1 to 10.9 mm, with a standard deviation of 3.2 mm. The mean needle displacement was 5.6 mm and ranged 0.2 to 15.6 mm. Needle displacements were reduced (p<0.05) after adjusting according to needle‐to‐fiducials distance.Conclusion:There were small fiducial marker displacements between HDR fractions. Our study suggests a target margin of 9.7 mm to cover interfractional marker displacements (in 95% cases) for pretreatment verification based on radiographic imaging.

Purpose: To design and implement a small-field IMRT/SRS/SBRT dosimetry measurement with a regular-size ion chamber and films. Methods: An acrylic phantom was constructed and commissioned to sandwich a Kodak EDR2 radiographic film. After a patient QA plan was delivered, the phantom was shifted superiorly by 10 cm and a reference plan was delivered on the same film. The reference plan has four-field-box beam geometry with 4 × 4 cm2 field size. Since the dose distribution of the reference plan was uniform and large, the absolute dose of the reference plan can be accurately measured separately with a regular-size ion chamber (diameter: 0.6 cm). After normalization, a two-dimensional absolute dose distribution can be obtained and compared to that calculated by Eclipse Treatment Planning System. An in-house software written in MATLAB was used to analyze films. Two-dimensional gamma indexes were calculated to evaluate patient QA plans. Results: Three patient-specific IMRT/SRS/SBRT QA plans were used to verify the feasibility of the method. The prescription dose of the reference plan was 3.5 Gy. The scatter dose from one plan to the other plan (which is 10 cm or more away) is very small (<1 cGy) and can be ignored. Compared with Eclipse calculation, the measured dose errors of the reference plan were −0.1−, −0.3%, and −0.8%, respectively. For the three patient-specific QA plans, the point dose measurement errors were 2.1%, 0.8% and 0.2%, respectively, and the γ>1 failure rates were 0.2%, 0.2%, and 0.1%, respectively. Conclusion: Since the scatter dose from another plan 10 cm (or more) away was very small, this method allows one to measure the dose of any small target with a regular-size ion chamber and films without volume limitation.

PURPOSE To evaluate accuracy of dose calculation for multiple small targets treated simultaneously using single-isocenter VMAT. METHODS 3, 5, 7, 10, 15, and two 20 mm diameter targets were contoured on a plane in a 15×20×20 cm3 phantom. Radiosurgery plans using coplanar and non-coplanar arcs were developed using a 6 MV beam with a 5 mm MLC (6X-STD) and a 10 MV flattening filter free beam with a 2.5 mm MLC (10FFF-HD). A uniformity goal was applied to one of the 20 mm targets to create a volume suitable for measurement using an ionization chamber. Dose was calculated on a 1×1×1 mm3 grid using the analytic anisotropic algorithm (AAA). Plans were delivered to a water-equivalent phantom and the dose measured using an ionization chamber and radiographic film. RESULTS For the 6X-STD plans, the chamber measurements were 1.7% less than calculation for both plans. For the remaining targets, the film doses ranged from 1.2% to 4.2% lower than calculation for the targets larger than 3 mm, whereas doses were 8.9% and 7.0% lower than calculation for the 3 mm target for the coplanar and non-coplanar plans, respectively. For the 10FFF-HD plans, the chamber measurements were 2.6% and 1.5% higher than calculation for the coplanar and non-coplanar plans, respectively. The film doses for the remaining targets ranged from 1.1% lower to 3.1% higher than calculation, with the largest differences noted for the 5 mm target. CONCLUSION With renormalization, the present results suggest dose accuracy within 3% for targets at least 3 mm in diameter for the 10FFF-HD beam and for targets of at least 5 mm diameter for the 6X-MLC beam. Work is ongoing to identify the source of the larger differences observed for the 3 mm target and the 6X-STD beam. Author has research support from Varian Medical Systems.

Purpose: HDR Brachytherapy using interstitial needle templates for gynecological cancer is commonly delivered in 4–5 fractions. Potential movement of afterloading needles between fractions requires verification before each treatment. The current study analyzed the interfractional displacement of the implant needles relative to implanted fiducial markers (for tumor border) positions. Methods: 15 patients with interstitial needle implants were reviewed. Treatment plans were based on CT images and the superior and inferior borders of the target volumes were defined by fiducial markers. The implant position was verified with kV orthogonal images before each treatment. The 3D positions of the needle tip and fiducial markers were determined from the orthogonal images. Needle displacement was defined by the change of the 3D distance between needle tip and 2 fiducial markers between planning and before the last fraction. The relative 3D distances between 2 fiducial markers were also measured to assess possible markers migration. Results: The median needle was 1.1 mm and the displacement ranged from 0.2–7.8 mm for the needle that near superior fiducial marker, and was typically in the inferior direction. One patient had significant needle displacement (7.8 mm) and positions were corrected before treatment. The needle displacements in the remaining patients were less than 4 mm and were not corrected because of adequate treatment margins. The displacement between 2 fiducial markers changed from 0.2–4.9 mm with a median change of 1 mm. Conclusion: There was detectable displacement of implant needles between HDR fractions, typically in the inferior direction. Pretreatment verification using orthogonal imaging is required and used to correct needle position based on implanted fiducial markers and needle tip positions.

PET staging of cervical cancer has increased identification of patients with para-aortic lymph node (PALN) metastasis. IMRT enables dose escalation in this area, but matching IMRT fields with traditional whole pelvis (WP) fields presents a challenge. We report institutional outcomes for PALN positive cervical cancer (CC) patients treated with a novel dynamic-field matching technique (DMT), which was developed to minimize dose inhomogeneities at the WP and IMRT field junction. From 2003 to 2012, 20 patients with CC and PALN metastasis were treated utilizing the DMT. As opposed to single-isocenter half-beam junction techniques, the DMT employs wedge-shaped dose junctions for the abutment of WP and IMRT fields. During treatment, leaves which define the superior border of the WP move continuously from a position 1.5 cm above to 1.5 cm below the superior border of the field, creating a wedge-shaped dose distribution. Para-aortic IMRT fields are then optimized to create a complementary wedge, accounting for dose contribution from WP fields. We reviewed the records of all patients who completed treatment with DMT and abstracted treatment, toxicity, and disease-related outcome data for analysis. Overall survival (OS) and disease-free survival (DFS) estimates were determined by the Kaplan-Meier method. All patients were treated with concurrent platinum-based chemotherapy and 8 underwent a prior staging or debulking operation. Median prescribed dose to the WP field was 45 Gy (range, 45-49.6 Gy) and para-aortic IMRT field 50.4 Gy (range, 45-55.5 Gy). Median parametrial boost dose was 5.4 Gy. All but 3 patients underwent HDR (13 pts) or LDR (4 pts) brachytherapy, the majority receiving 8 Gy x 3 fractions. All patients developed lower GI toxicity; 10 grade 1, 9 grade 2, and 1 grade 4 (enterovaginal fistula). The fistula was diagnosed 6 months after completion of radiation therapy with 45 Gy WP, 45 Gy to PALN PTV, and LDR Syed needle brachytherapy. It developed within WP fields but well away from any dose contribution by IMRT fields. Median follow-up was 15.9 months (range, 4.4-114.7 months). Median DFS was 18.7 months (95% CI, 8.4-29.0) with 1 and 2-year DFS 62.7% and 40.2%, respectively. One-year OS was 82.8% and 2-year OS 63.1%; median OS has not yet been reached. Of the 9 patients who developed recurrence, first site of recurrence was within the WP field for 1 patient, within the field and distant for 1 patient, and distant in 7 patients. There were no failures at the dynamic junction between IMRT and WP fields. DMT provides a means for joining WP and para-aortic IMRT fields that substantially reduces dose deviations at the junction due to field mismatch. Treatment with DMT is simple, effective, and tolerated with no apparent increase in toxicity.