Andries G. Visser

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.

Portal dose measurement in radiotherapy using an electronic portal imaging device (EPID)

Physical characteristics of a commercially available electronic portal imaging device (EPID), relevant to dosimetric applications in high-energy photon beams, have been investigated. The EPID basically consists of a fluorescent screen, mirrors and a CCD camera. Image acquisition for portal dose measurement has been performed with a special procedure, written in the command language that comes with the system. The observed day-to-day variation in local EPID responses, i.e. measured grey scale value (EPID signal) per unit of delivered portal dose, is 0.4% (1 SD); day-to-day variation in relative EPID responses (e.g. normalized to the on-axis response) are within 0.2% (1 SD). Measured grey scale values are linearly proportional to transmitted portal doses with a proportionality constant which is independent of the thickness of a flat, water-equivalent absorber in the beam, but which does significantly depend on the size of the applied x-ray beam. It is shown that the observed increased in EPID response with increasing field size is mainly due to contributions to the EPID signals from scattered light: visible photons produced by the x-ray beam in a point of the fluorescent screen not only generate a grey scale value in the corresponding point of the EPID image, but also lead (due to scatter from components of the EPID structure onto the CCD chip) to an increased grey scale value at all other points of the image. A point spread function, derived from measured data and describing the increase in EPID response at the beam axis due to off-axis irradiation of the fluorescent screen, has been successfully applied to connect portal doses with grey scale values measured with the EPID.

Dosimetric verification of intensity modulated beams produced with dynamic multileaf collimation using an electronic portal imaging device.

Dose distributions can often be significantly improved by modulating the two-dimensional intensity profile of the individual x-ray beams. One technique for delivering intensity modulated beams is dynamic multileaf collimation (DMLC). However, DMLC is complex and requires extensive quality assurance. In this paper a new method is presented for a pretreatment dosimetric verification of these intensity modulated beams utilizing a charge-coupled devicecamera based fluoroscopic electronic portal imaging device(EPID). In the absence of the patient, EPIDimages are acquired for all beams produced with DMLC. These images are then converted into two-dimensional dose distributions and compared with the calculated dose distributions. The calculations are performed with a pencil beam algorithm as implemented in a commercially available treatment planning system using the same absolute beam fluence profiles as used for calculation of the patient dose distribution. The method allows an overall verification of (i) the leaf trajectory calculation (including the models to incorporate collimator scatter and leaf transmission), (ii) the correct transfer of the leaf sequencing file to the treatment machine, and (iii) the mechanical and dosimetrical performance of the treatment unit. The method was tested for intensity modulated 10 and 25 MV photon beams; both model cases and real clinical cases were studied. Dose profiles measured with the EPID were also compared with ionization chamber measurements. In all cases both predictions and EPID measurements and EPID and ionization chamber measurements agreed within 2% (1σ). The study has demonstrated that the proposed method allows fast and accurate pretreatment verification of DMLC.

Analysis and reduction of 3D systematic and random setup errors during the simulation and treatment of lung cancer patients with CT-based external beam radiotherapy dose planning

PURPOSE To determine the magnitude of the errors made in (a) the setup of patients with lung cancer on the simulator relative to their intended setup with respect to the planned treatment beams and (b) in the setup of these patients on the treatment unit. To investigate how the systematic component of the latter errors can be reduced with an off-line decision protocol for setup corrections. METHODS AND MATERIALS For 39 patients with CT planning, digitally-reconstructed radiographs (DRRs) were calculated for anterior-posterior and lateral beams. Retrospectively, the position of the visible anatomy relative to the planned isocenter was compared with the corresponding position on the digitized simulator radiographs using contour match software. The setup accuracy at the treatment unit relative to the simulator setup was measured for 40 patients for at least 5 fractions per patient in 2 orthogonal beams with the aid of an electronic portal imaging device (EPID). Setup corrections were applied, based on an off-line decision protocol, with parameters derived from knowledge of the random setup errors in the studied patient group. RESULTS The standard deviations (SD) of the simulator setup errors relative to the CT planning setup in the lateral, longitudinal, and anterior-posterior directions were 4.0, 2.8, and 2.5 mm, respectively. The SD of rotations around the anterior-posterior axis was 1.6 degrees and around the left-right axis 1.3 degrees. The setup error at the treatment unit had a small random component in all three directions (1 SD = 2 mm). The systematic components were larger, particularly in the longitudinal direction (1 SD = 3.6 mm), but were reduced with the decision protocol to 1 SD < 2 mm with, on average, 0.6 setup correction per patient. CONCLUSION Setup errors at the simulator, which become systematic errors if the simulation defines the reference setup, were comparable to the systematic setup errors at the treatment unit in case no off-line protocol would have been applied. Hence, the omission of a separate simulation step can reduce systematic errors as efficiently as the application of an off-line correction protocol during treatment. The random errors were sufficiently small to make an off-line protocol feasible.

Accuracy in radiation field alignment in head and neck cancer: A prospective study

A prospective study has been performed to determine the accuracy of radiation field alignment for a group of 22 patients with tumors in the head and neck. The accuracy was assessed by an analysis of 138 megavolt portal films in comparison to 55 simulation films. The distance (at the patient midplane) between corresponding points at the field edges on verification film and simulation film appeared to be 5 mm on the average and the standard deviation 5 mm. The analysis was extended by translational and rotational matching of the fields in order to separate each error in a translation error of the field with respect to the patient and an error in field size or shape. Translation errors appear to be somewhat larger than field size or shape errors. From an analysis of a series of megavolt films taken every third radiotherapy session, it was concluded that treatment-to-treatment variations are as large as the errors due to the transition from simulation to treatment situation. Further analysis showed that variation of the patients position within the cast is clearly one of the error sources.

Electronic portal image assisted reduction of systematic set-up errors in head and neck irradiation.

PURPOSE To quantify systematic and random patient set-up errors in head and neck irradiation and to investigate the impact of an off-line correction protocol on the systematic errors. MATERIAL AND METHODS Electronic portal images were obtained for 31 patients treated for primary supra-glottic larynx carcinoma who were immobilised using a polyvinyl chloride cast. The observed patient set-up errors were input to the shrinking action level (SAL) off-line decision protocol and appropriate set-up corrections were applied. To assess the impact of the protocol, the positioning accuracy without application of set-up corrections was reconstructed. RESULTS The set-up errors obtained without set-up corrections (1 standard deviation (SD)=1.5-2mm for random and systematic errors) were comparable to those reported in other studies on similar fixation devices. On an average, six fractions per patient were imaged and the set-up of half the patients was changed due to the decision protocol. Most changes were detected during weekly check measurements, not during the first days of treatment. The application of the SAL protocol reduced the width of the distribution of systematic errors to 1mm (1 SD), as expected from simulations. A retrospective analysis showed that this accuracy should be attainable with only two measurements per patient using a different off-line correction protocol, which does not apply action levels. CONCLUSIONS Off-line verification protocols can be particularly effective in head and neck patients due to the smallness of the random set-up errors. The excellent set-up reproducibility that can be achieved with such protocols enables accurate dose delivery in conformal treatments.

Reirradiation of recurrent head and neck cancers: external and/or interstitial radiation therapy

Recurrent cancer in the head and neck is not an uncommon clinical problem. The average cure rate of these patients has been reported to vary between 30 and 40% and most failures are due to locoregional relapses. After a previous full course of radiation, surgery is the salvage modality of choice; however, if surgery was not feasible for whatever reason, reirradiation has been offered to some patients. To establish the role of reirradiation in head and neck cancer, we analyzed a 13-year experience with patients reirradiated in the DDHCC. The reirradiation was performed between 1970 and 1980 by means of external beam radiation therapy (ERT series; n = 55) and between 1985 and 1988 by external radiation combined with interstitial radiation therapy (IRT +/- ERT series; n = 18). A minimum follow-up of 3 years was allowed for. An improvement in local control was observed (50% vs. 29%) for the IRT +/- ERT series and the ERT series, respectively. The improvement in local control was not reflected in a survival benefit; i.e. an actuarial overall survival of 20% at 5 years was observed in both series. No treatment-related deaths occurred. However, for the patients that were controlled at the reirradiated site, 28% (4/16 of the ERT series and 3/9 of the IRT +/- ERT series) did experience severe side effects.

Fractionated high-dose-rate and pulsed-dose-rate brachytherapy: First clinical experience in squamous cell carcinoma of the tonsillar fossa and soft palate

PURPOSE Fractionated high-dose-rate (fr.HDR) and pulsed-dose-rate (PDR) brachytherapy (BT) regimens, which simulate classical continuous low-dose-rate (LDR) interstitial radiation therapy (IRT) schedules, have been developed for clinical use. This article reports the initial results using these novel schedules in squamous cell carcinoma (SCC) of the tonsillar fossa (TF) and/or soft palate (SP). METHODS AND MATERIALS Between 1990 and 1994, 38 patients with TF and SP tumors (5 T1, 22 T2, 10 T3, and 1 T4) were treated by fr.HDR or PDR brachytherapy, either alone or in combination with external irradiation (ERT). Half of the patients were treated with fr.HDR, which entailed twice-daily fractions of > or = 3 Gy. The other 19 patients were administered PDR, which consisted of pulses of < or = 2 Gy delivered 4-8 times/day. The median cumulative dose of IRT +/- ERT series was 66 Gy (range 55-73). The results in these patients treated by brachytherapy were compared to 72 patients with similar tumors treated in our institute with curative intent, using ERT alone. The median cumulative dose of ERT-only series was 70 Gy (range 40-77). RESULTS Excellent locoregional control was achieved with the use of IRT +/- ERT, with only 13% (5 of 38) developing local failure, and salvage surgery being possible in three of the latter (60%). Neither BT scheme (fr.HDR vs. PDR) nor tumor site (TF vs. SP) significantly influenced local control rates. The type and severity of the side effects observed are comparable to those reported in the literature for LDR-IRT. These results contrast sharply with our ERT-only series, in which 39% of patients (28 of 72) developed local failure, with surgical salvage being possible only in three patients (11%). Taking the data set of 110 patients, in a univariate analysis IRT, T stage, N stage, overall treatment time (OTT), and BEDcor10 (biological effective dose with a correction for the OTT) were significant prognostic factors for local relapse-free survival (LRFS) and overall survival (OS) at 3 years. Using Cox proportional hazard analysis, only T stage and BEDcor10 remained significant for LRFS (p < 0.001 and 0.008, respectively), as well as for OS (p < 0.001 and 0.003, respectively). With regard to the current (IRT) and historical (ERT) series, for the LRFS at 3 years, dose-response relationships were established, significant, however, only for the BEDcor10 (p = 0.03). CONCLUSION The 3-year LRFS of approximately 90% for TF and SP tumors reported here is comparable with the best results in the literature, particularly given the fact that 30% of the patients (11 of 38) presented with T3/4 tumors. When compared with our historical (ERT-only) controls, the patients treated with IRT had superior local control. A dose-response relationship was established for the BEDcor10.

Quality assurance using portal imaging: The accuracy of patient positioning in irradiation of breast cancer☆

PURPOSE To study the accuracy of patient positioning in irradiation of breast cancer. METHODS AND MATERIALS Megavolt portal images were obtained using a fast electronic megavoltage radiotherapy imaging system in 17 breast cancer patients immobilized with plastic fixation masks on a flat board with arm support and in 14 patients positioned without a mask on either a flat or a wedge-shaped board. Quantitative analysis of 510 megavolt portal images and comparison to 66 digitized simulation films was performed. Differences between the positioning techniques were evaluated. RESULTS For the position of the patient in the field, standard deviations of the difference between simulation and treatment images were 3.2 mm and 4.6 mm for irradiation with and without masks, respectively. Larger standard deviations were found for the field width and length (5-7 mm), for collimator rotation (1.5-2 degrees), and for the position of the lung shielding block for patients positioned on the flat board (10-16 mm). The changes in field size and collimator rotation appeared to be largely due to the inclination of the technologists to slightly adapt fields in order to obtain a seemingly better congruity of the field with the skin or mask markings. Comparison of the accuracy of patient positioning with and without masks yielded similar error rates; standard deviations and extremes tended to be somewhat larger in positioning without a mask. The wedge-shaped board was preferred because of the ease of patient set-up and because the use of a lung block is avoided. The transition from simulation to treatment set-up yielded larger deviations than repeated treatment set-ups. CONCLUSION These results emphasize again the continuous need for focusing attention on the accuracy of patient positioning in order to achieve maximal precision in radiotherapy. The electronic portal imaging system is very suitable for both quick on-line treatment verification and off-line analyses.

Portal dose image (PDI) prediction for dosimetric treatment verification in radiotherapy. I. An algorithm for open beams

A method is presented for calculation of transmission functions for high energy photon beams through patients. These functions are being used in our clinic for prediction of portal dose images (PDIs) which are compared with PDIs measured with an electronic portal imaging device (EPID). The calculations are based on the planning CT-scan of the patient and on the irradiation geometry as determined in the treatment planning process. For each beam quality, the required input data for the algorithm for transmission prediction are derived from a limited number of measured beam data. The method has been tested for a PDI-plane at 160 cm from the focus, in agreement with the fixed focus-to-detector distance of our fluoroscopic EPIDs. For 6, 23 and 25 MV photon beams good agreement (approximately 1%) has been found between calculated and measured transmissions through anthropomorphic phantoms.