Catching errors with in vivo EPID dosimetry.

The potential for detrimental incidents and the ever increasing complexity of patient treatments emphasize the need for accurate dosimetric verification in radiotherapy. For this reason, all curative treatments are verified, either pretreatment or in vivo, by electronic portal imaging device (EPID) dosimetry in the Radiation Oncology Department of The Netherlands Cancer Institute-Antoni van Leeuwenhoek hospital, Amsterdam, The Netherlands. Since the clinical introduction of the method in January 2005 until August 2009, treatment plans of 4337 patients have been verified. Among these plans, 17 serious errors were detected that led to intervention. Due to their origin, nine of these errors would not have been detected with pretreatment verification. The method is illustrated in detail by the case of a plan transfer error detected in a 5 x 5 Gy intensity-modulated radiotherapy (IMRT) rectum treatment. The EPID reconstructed dose at the isocenter was 6.3% below the planned value. Investigation of the plan transfer chain revealed that due to a network transfer error, the plan was corrupted. 3D analysis of the acquired EPID data revealed serious underdosage of the planning target volume: On average 11.6%, locally up to 20%. This report shows the importance of in vivo (EPID) dosimetry for all treatment plans as well as the ability of the method to assess the dosimetric impact of deviations found.

Detection of internal organ movement in prostate cancer patients using portal images.

Previous research has indicated that the appearance of large gas pockets in portal images of prostate cancer patients might imply internal prostate motion. This was verified with simulations based on multiple computed tomography (CT) data for 15 patients treated in supine position. Apart from the planning CT scan, three extra scans were made during treatment. The clinical target volume (CTV) and the rectum were outlined in all scans. Lateral portal images were simulated from the CT data and difference images were calculated for all possible combinations of CT scans per patient; each scan was used both as reference and repeat scan but gas pockets in the reference scan were removed. Gas pockets in a repeat CT scan then show up as black areas in a difference image. Due to gravity, they normally appear in the ventral part of the rectum. The distances between the ventral edge of a gas pocket in a difference image and the projection of the delineated ventral rectum wall in the reference scan were calculated. These distances were correlated with the "true" rectum wall shifts (determined from direct comparison of the rectum delineations in reference and repeat scan) and with CTV movements determined by three-dimensional chamfer matching. Gas pockets occurred in 23% of cases. Nevertheless, about 50% of rectum wall shifts larger than 5 mm could be detected because they were associated with gas pockets with a lateral diameter > 2 cm. When gas pockets were visible in the repeat scan, rectum wall shifts could be accurately detected by the ventral gas pocket edge in the difference images (r= 0.97). The shift of the rectum wall as detected from gas pockets also correlated significantly with the anterior-posterior shift of the center of mass of the CTV (r=0.88). In conclusion, the simulations showed that lateral pelvic images contain more information than the bony structures that are normally used for setup verification. If large gas pockets appear in those images, a quantitative estimate of the position of prostate and rectum wall can be obtained by determination of the ventral edge of the gas pocket.

On-line set-up corrections during radiotherapy of patients with gynecologic tumors.

PURPOSE: Positioning of patients with gynecologic tumors for radiotherapy has proven to be relatively inaccurate. To improve the accuracy and reduce the margins from clinical target volume (CTV) to planning target volume (PTV), on-line set-up corrections were investigated. METHODS AND MATERIALS: Anterior-posterior portal images of 14 patients were acquired using the first six monitor units (MU) of each irradiation fraction. The set-up deviation was established by matching three user-defined landmarks in portal and simulator image. If the two-dimensional deviation exceeded 4 mm, the table position was corrected. A second portal image was acquired using 30 MU of the remaining dose. This image was analyzed off-line using a semiautomatic contour match to obtain the final set-up accuracy. To verify the landmark match accuracy, the contour match was retrospectively performed on the six MU images as well. RESULTS: The standard deviation (SD) of the distribution of systematic set-up deviations after correction was < 1 mm in left-right and cranio-caudal directions. The average random deviation was < 2 mm in these directions (1 SD). Before correction, all standard deviations were 2 to 3 mm. The landmark match procedure was sufficiently accurate and added on average 3 min to the treatment time. The application of on-line corrections justifies a CTV-to-PTV margin reduction to about 5 mm. CONCLUSIONS: On-line set-up corrections significantly improve the positioning accuracy. The procedure increases treatment time but might be used effectively in combination with off-line corrections.

Comparison of prostate cancer treatment in two institutions: a quality control study.

PURPOSE: To minimize differences in the treatment planning procedure between two institutions within the context of a radiotherapy prostate cancer trial. PATIENTS AND METHODS: Twenty-two patients with N0 M0 prostate cancer underwent a computed tomography (CT) scan for radiotherapy treatment planning. For all patients, the tumor and organs at risk were delineated, and a treatment plan was generated for a three-field technique giving a dose of 78 Gy to the target volume. Ten of the 22 cases were delineated and planned in the other institution as well. The delineated volumes and dose distributions were compared. RESULTS: All treatments fulfilled the trial criteria. The mean volume ratio of the gross tumor volumes (GTVs) in both institutions was 1.01, while the mean volume ratio of the planning target volumes (PTVs) was 0.88. The three-dimensional (3D) PTV difference was 3 mm at the prostate apex and 6-8 mm at the seminal vesicles. This PTV difference was mainly caused by a difference in the method of 3D expansion, and disappeared when applying an improved algorithm in one institution. The treated volume (dose > or =95% of isocenter dose) reflects the size of the PTV and the conformity of the treatment technique. This volume was on average 66 cm3 smaller in institution A than in institution B; the effect of the PTV difference was 31 cm3 and the difference in technique accounted for 36 cm3. The mean delineated rectal volume including filling was 112 cm3 and 125 cm3 for institution A and B, respectively. This difference had a significant impact on the relative dose volume histogram (DVH) of the rectum. CONCLUSION: Differences in GTV delineation were small and comparable to earlier quantified differences between observers in one institution. Different expansion methods for generation of the PTV significantly influenced the amount of irradiated tissue. Strict definitions of target and normal structures are mandatory for reliable trial results.

Internal organ motion in prostate cancer patients treated in prone and supine treatment position.

BACKGROUND AND PURPOSE: To compare supine and prone treatment positions for prostate cancer patients with respect to internal prostate motion and the required treatment planning margins. MATERIALS AND METHODS: Fifteen patients were treated in supine and fifteen in prone position. For each patient, a planning computed tomography (CT) scan was used for treatment planning. Three repeat CT scans were made in weeks 2, 4, and 6 of the radiotherapy treatment. Only for the planning CT scan, laxation was used to minimise the rectal content. For all patients, the clinical target volume (CTV) consisted of prostate and seminal vesicles. Variations in the position of the CTV relative to the bony anatomy in the four CT scans of each patient were assessed using 3D chamfer matching. The overall variations were separated into variations in the mean CTV position per patient (i.e. the systematic component) and the average 'day-to-day' variation (i.e. the random component). Required planning margins to account for the systematic and random variations in internal organ position and patient set-up were estimated retrospectively using coverage probability matrices. RESULTS: The observed overall variation in the internal CTV position was larger for the patients treated in supine position. For the supine and prone treatment positions, the random components of the variation along the anterior-posterior axis (i.e. towards the rectum) were 2.4 and 1.5 mm (I standard deviation (1 SD)), respectively; the random rotations around the left-right axis were 3.0 and 2.9 degrees (1 SD). The systematic components of these motions (1 SD) were larger: 2.6 and 3.3 mm, and 3.7 and 5.6 degrees, respectively. The set-up variations were similar for both treatment positions. Despite the smaller overall variations in CTV position for the patients in prone position, the required planning margin is equal for both groups (about 1 cm except for 0.5 cm in lateral direction) due to the larger impact of the systematic variations. However, significant time trends cause a systematic ventral-superior shift of the CTV in supine position only. CONCLUSIONS: For internal prostate movement, it is important to distinguish systematic from random variations. Compared to patients in supine position, patients in prone position had smaller random but somewhat larger systematic variations in the most important coordinates of the internal CTV position. The estimated planning margins to account for the geometrical uncertainties were therefore similar for the two treatment positions.

Acute morbidity reduction using 3DCRT for prostate carcinoma: a randomized study.

PURPOSE: To study the effects on gastrointestinal and urological acute morbidity, a randomized toxicity study, comparing conventional and three-dimensional conformal radiotherapy (3DCRT) for prostate carcinoma was performed. To reveal possible volume effects, related to the observed toxicity, dose-volume histograms (DVHs) were used. METHODS AND MATERIALS: From June 1994 to March 1996, 266 patients with prostate carcinoma, stage T1-4N0M0 were enrolled in the study. All patients were treated to a dose of 66 Gy (ICRU), using the same planning procedure, treatment technique, linear accelerator, and portal imaging procedure. However, patients in the conventional treatment arm were treated with rectangular, open fields, whereas conformal radiotherapy was performed with conformally shaped fields using a multileaf collimator. All treatment plans were made with a 3D planning system. The planning target volume (PTV) was defined to be the gross target volume (GTV) + 15 mm. Acute toxicity was evaluated using the EORTC/RTOG morbidity scoring system. RESULTS: Patient and tumor characteristics were equally distributed between both study groups. The maximum toxicity was 57% grade 1 and 26% grade 2 gastrointestinal toxicity; 47% grade 1, 17% grade 2, and 2% grade > 2 urological toxicity. Comparing both study arms, a reduction in gastrointestinal toxicity was observed (32% and 19% grade 2 toxicity for conformal and conventional radiotherapy, respectively; p = 0.02). Further analysis revealed a marked reduction in medication for anal symptoms: this accounts for a large part of the statistical difference in gastrointestinal toxicity (18% vs. 14% [p = ns] grade 2 rectum/sigmoid toxicity and 16% vs. 8% [p < 0.0001] grade 2 anal toxicity for conventional and conformal radiotherapy, respectively). A strong correlation between exposure of the anus and anal toxicity was found, which explained the difference in anal toxicity between both study arms. No difference in urological toxicity between both treatment arms was found, despite a relatively large difference in bladder DVHs. CONCLUSIONS: The reduction in gastrointestinal morbidity was mainly accounted for by reduced toxicity for anal symptoms using 3DCRT. The study did not show a statistically significant reduction in acute rectum/sigmoid and bladder toxicity.

Inclusion of geometrical uncertainties in radiotherapy treatment planning by means of coverage probability.

PURPOSE: Following the ICRU-50 recommendations, geometrical uncertainties in tumor position during radiotherapy treatments are generally included in the treatment planning by adding a margin to the clinical target volume (CTV) to yield the planning target volume (PTV). We have developed a method for automatic calculation of this margin. METHODS AND MATERIALS: Geometrical uncertainties of a specific patient group can normally be characterized by the standard deviation of the distribution of systematic deviations in the patient group (Sigma) and by the average standard deviation of the distribution of random deviations (sigma). The CTV of a patient to be planned can be represented in a 3D matrix in the treatment room coordinate system with voxel values one inside and zero outside the CTV. Convolution of this matrix with the appropriate probability distributions for translations and rotations yields a matrix with coverage probabilities (CPs) which is defined as the probability for each point to be covered by the CTV. The PTV can then be chosen as a volume corresponding to a certain iso-probability level. Separate calculations are performed for systematic and random deviations. Iso-probability volumes are selected in such a way that a high percentage of the CTV volume (on average > 99%) receives a high dose (> 95%). The consequences of systematic deviations on the dose distribution in the CTV can be estimated by calculation of dose histograms of the CP matrix for systematic deviations, resulting in a so-called dose probability histogram (DPH). A DPH represents the average dose volume histogram (DVH) for all systematic deviations in the patient group. The consequences of random deviations can be calculated by convolution of the dose distribution with the probability distributions for random deviations. Using the convolved dose matrix in the DPH calculation yields full information about the influence of geometrical uncertainties on the dose in the CTV. RESULTS: The model is demonstrated to be fast and accurate for a prostate, cervix, and lung cancer case. A CTV-to-PTV margin size which ensures at least 95% dose to (on average) 99% of the CTV, appears to be equal to about 2Sigma + 0.7sigma for three all cases. Because rotational deviations are included, the resulting margins can be anisotropic, as shown for the prostate cancer case. CONCLUSION: A method has been developed for calculation of CTV-to-PTV margins based on the assumption that the CTV should be adequately irradiated with a high probability.

Multiple two-dimensional versus three-dimensional PTV definition in treatment planning for conformal radiotherapy.

PURPOSE: To demonstrate the need for a fully three-dimensional (3D) computerized expansion of the gross tumour volume (GTV) or clinical target volume (CTV), as delineated by the radiation oncologist on CT slices, to obtain the proper planning target volume (PTV) for treatment planning according to the ICRU-50 recommendations. MATERIALS AND METHODS: For 10 prostate cancer patients two PTVs have been determined by expansion of the GTV with a 1.5 cm margin, i.e. a 3D PTV and a multiple 2D PTV. The former was obtained by automatically adding the margin while accounting in 3D for GTV contour differences in neighbouring slices. The latter was generated by automatically adding the 1.5 cm margin to the GTV in each CT slice separately; the resulting PTV is a computer simulation of the PTV that a radiation oncologist would obtain with (the still common) manual contouring in CT slices. For each patient the two PTVs were compared to assess the deviations of the multiple 2D PTV from the 3D PTV. For both PTVs conformal plans were designed using a three-field technique with fixed block margins. For each patient dose-volume histograms and tumour control probabilities (TCPs) of the (correct) 3D PTV were calculated, both for the plan designed for this PTV and for the treatment plan based on the (deviating) 2D PTV. RESULTS: Depending on the shape of the GTV, multiple 2D PTV generation could locally result in a 1 cm underestimation of the GTV-to-PTV margin. The deviations occurred predominantly in the cranio-caudal direction at locations where the GTV contour shape varies significantly from slice to slice. This could lead to serious underdosage and to a TCP decrease of up to 15%. CONCLUSIONS: A full 3D GTV-to-PTV expansion should be applied in conformal radiotherapy to avoid underdosage.

Field margin reduction using intensity-modulated x-ray beams formed with a multileaf collimator.

PURPOSE: In axial, coplanar treatments with multiple fields, the superior and inferior ends of a planning target volume (PTV) are at risk to get underdosed due to the overlapping penumbras of all treatment fields. We have investigated a technique using intensity modulated x-ray beams that allows the use of small margins for definition of the superior and inferior field borders while still reaching a minimum PTV-dose of 95% of the isocenter dose. METHODS AND MATERIALS: The applied intensity modulated beams, generated with a multileaf collimator, include narrow (1.1-1.6 cm) boost fields to increase the dose in the superior and inferior ends of the PTV. The benefits of this technique have been assessed using 3D treatment plans for 10 prostate cancer patients. Treatment planning was performed with the Cadplan 3D planning system (Varian-Dosetek). Dose calculations for the narrow boost fields have been compared with measurements. The application of the boost fields has been tested on the MM50 Racetrack Microtron (Scanditronix Medical AB), which allows fully computer-controlled setup of all involved treatment fields. RESULTS: Compared to our standard technique, the superior-inferior field length can be reduced by 1.6 cm, generally yielding smaller volumes of rectum and bladder in the high dose region. For the narrow boost fields, calculated relative dose distributions agree within 2% or 0.2 cm with measured dose distributions. For accurate monitor unit calculations, the phantom scatter table used in the Cadplan system had to be modified using measured data for square fields smaller than 4 x 4 cm2. The extra time needed at the MM50 for the setup and delivery of the boost fields is usually about 1 min. CONCLUSION: The proposed use of intensity modulated beams yields improved conformal dose distributions for treatment of prostate cancer patients with a superior-inferior field size reduction of 1.6 cm. Treatments of other tumor sites can also benefit from the application of the boost fields.

Automatic calculation of three-dimensional margins around treatment volumes in radiotherapy planning.

Following the publication of ICRU Report 50, the concepts of GTV (gross tumour volume). CTV (clinical target volume) and PTV (planning target volume) are being used in radiotherapy planning with increasing frequency. In 3D planning, the GTV (or CTV) is normally outlined by the clinician in CT or MRI slices. The PTV is determined by adding margins to these volumes. Since manual drawing of an accurate 3D margin in a set of 2D slices is extremely time consuming, software has been developed to automate this step in the planning. The target volume is represented in a 3D matrix grid with voxel values one inside and zero outside the target volume. It is expanded by centering an ellipsoid at every matrix element within the volume. The shape of the ellipsoid reflects the size of the margins in the three main orthogonal directions. Finally, the PTV contours are determined from the 50% iso-value lines of the expanded volume. The software tool has been in clinical use since the end of 1994 and has mostly been applied to the planning of prostate irradiations. The accuracy is better than can be achieved manually and the workload has been reduced considerably (from 4 h manually to approximately 1 min automatically).

Physical characteristics of a commercial electronic portal imaging device.

An electronic portal imaging device (EPID) for use in radiotherapy with high energy photons has been under development since 1985 and has been in clinical use since 1988. The x-ray detector consists of a metal plate/fluorescent screen combination, which is monitored by a charge-coupled device (CDD)-camera. This paper discusses the physical quantities governing image quality. A model which describes the signal and noise propagation through the detector is presented. The predicted contrasts and signal-to-noise ratios are found to be in agreement with measurements based on the EPID images. Based on this agreement the visibility of low contrast structures in clinical images has been calculated with the model. Sufficient visibility of relevant structures (4-10 mm water-equivalent thickness) has been obtained down to a delivered dose of 4 cGy at dose maximum. It is found that the described system is not limited by quantum noise but by camera read-out noise. In addition we predict that with a new type of CCD sensor the signal-to-noise ratio can be increased by a factor of 5 at small doses, enabling high quality imaging, for most relevant clinical situations, with a patient dose smaller than 4 cGy. The latter system would be quantum noise limited.

High-precision prostate cancer irradiation by clinical application of an offline patient setup verification procedure, using portal imaging.

PURPOSE: To investigate in three institutions, The Netherlands Cancer Institute (Antoni van Leeuwenhoek Huis [AvL]), Dr. Daniel den Hoed Cancer Center (DDHC), and Dr, Bernard Verbeeten Institute (BVI), how much the patient setup accuracy for irradiation of prostate cancer can be improved by an offline setup verification and correction procedure, using portal imaging. METHODS AND MATERIALS: The verification procedure consisted of two stages. During the first stage, setup deviations were measured during a number (Nmax) of consecutive initial treatment sessions. The length of the average three dimensional (3D) setup deviation vector was compared with an action level for corrections, which shrunk with the number of setup measurements. After a correction was applied, Nmax measurements had to be performed again. Each institution chose different values for the initial action level (6, 9, and 10 mm) and Nmax (2 and 4). The choice of these parameters was based on a simulation of the procedure, using as input preestimated values of random and systematic deviations in each institution. During the second stage of the procedure, with weekly setup measurements, the AvL used a different criterion ("outlier detection") for corrective actions than the DDHC and the BVI ("sliding average"). After each correction the first stage of the procedure was restarted. The procedure was tested for 151 patients (62 in AvL, 47 in DDHC, and 42 in BVI) treated for prostate carcinoma. Treatment techniques and portal image acquisition and analysis were different in each institution. RESULTS: The actual distributions of random and systematic deviations without corrections were estimated by eliminating the effect of the corrections. The percentage of mean (systematic) 3D deviations larger than 5 mm was 26% for the AvL and the DDHC, and 36% for the BVI. The setup accuracy after application of the procedure was considerably improved (percentage of mean 3D deviations larger than 5 mm was 1.6% in the AvL and 0% in the DDHC and BVI), in agreement with the results of the simulation. The number of corrections (about 0.7 on the average per patient) was not larger than predicted. CONCLUSION: The verification procedure appeared to be feasible in the three institutions and enabled a significant reduction of mean 3D setup deviations. The computer simulation of the procedure proved to be a useful tool, because it enabled an accurate prediction of the setup accuracy and the required number of corrections.