Analysis of patient specific dosimetry quality assurance measurements in intensity modulated radiotherapy: a multi centre study.

AIM OF THE STUDY: Statistical analysis of pre-treatment dose verification of intensity-modulated radiotherapy (IMRT) fields to assess the quality of the IMRT practice at different radiotherapy centers. MATERIALS AND METHODS: The dose verification data acquired by the institutional physicist of 10 different hospitals for various types of patients were collected and analyzed for mean, median, standard deviation (SD), range, minimum and maximum % deviation. The percentage of cases having positive and negative dose differences as well dose differences within ± 3% were also determined. RESULTS: The mean values of percentage variation in difference between treatment planning systems calculated dose and difference between measured dose (D(TPS) and D(Meas)) are found to be from -1.79 to 1.48 and median from -1.79 to 1.51. The SDs are found to be from 0.76 to 3.70. The range of variation at these centers varies from 3.99 to 16.45 while minimum and maximum values of percentage variation in difference between D(TPS) and D(Meas) ranges from -10.33 to 13.38. The percentage of cases having positive dose difference ranges from 8 to 94 and cases having negative dose difference ranges from 6 to 92. The percentage of cases having dose difference within ± 3% varies from 57 to 100. CONCLUSION: IMRT centers are having random and biased (skewed towards over or under dose) distribution of the percentage variation in difference between measured and planned doses. The analysis of results of the IMRT pre-treatment dose verification reveals that there are systematic errors in the chain of IMRT treatment process at a few centers. The dosimetry quality audit prior to commissioning of IMRT may play an important role in avoiding such discrepancies.

The impact of new technologies on radiation oncology events and trends in the past decade: an institutional experience.

PURPOSE: To review the type and frequency of patient events from external-beam radiotherapy over a time period sufficiently long to encompass significant technology changes. METHODS AND MATERIALS: Ten years of quality assurance records from January 2001 through December 2010 were retrospectively reviewed to determine the frequency of events affecting patient treatment from four radiation oncology process steps: simulation, treatment planning, data entry/transfer, and treatment delivery. Patient events were obtained from manual records and, from May 2007 onward, from an institution-wide database and reporting system. Events were classified according to process step of origination and segregated according to the most frequently observed event types. Events from the institution-wide database were evaluated to determine time trends. RESULTS: The overall event rate was 0.93% per course of treatment, with a downward trend over time led by a decrease in treatment delivery events. The frequency of certain event types, particularly in planning and treatment delivery, changed significantly over the course of the study, reflecting technologic and process changes. Treatments involving some form of manual intervention carried an event risk four times higher than those relying heavily on computer-aided design and delivery. CONCLUSIONS: Although the overall event rate was low, areas for improvement were identified, including manual calculations and data entry, late-day treatments, and staff overreliance on computer systems. Reducing the incidence of pretreatment events is of particular importance because these were more likely to occur several times before detection and were associated with larger dosimetric impact. Further improvements in quality assurance systems and reporting are imperative, given the advent of electronic charting, increasing reliance on computer systems, and the potentially severe consequences that can arise from mistakes involving complex intensity-modulated or image-guided treatments.

Evaluation of tumor motion effects on dose distribution for hypofractionated intensity-modulated radiotherapy of non-small-cell lung cancer.

Respiration-induced tumor motion during intensity-modulated radiotherapy (IMRT) of non-small-cell lung cancer (NSCLC) could cause substantial differences between planned and delivered doses. While it has been shown that, for conventionally fractionated IMRT, motion effects average out over the course of many treatments, this might not be true for hypofractionated IMRT (IMHFRT). Numerical simulations were performed for nine NSCLC patients (11 tumors) to evaluate this problem. Dose distributions to the Clinical Target Volume (CTV) and Internal Target Volume (ITV) were retrospectively calculated using the previously-calculated leaf motion files but with the addition of typical periodic motion (i.e. amplitude 0.36-1.26cm, 3-8sec period). A typical IMHFRT prescription of 20Gy x 3 fractions was assumed. For the largest amplitude (1.26 cm), the average +/- standard deviation of the ratio of simulated to planned mean dose, minimum dose, D95 and V95 were 0.98+/-0.01, 0.88 +/- 0.09, 0.94 +/- 0.05 and 0.94 +/- 0.07 for the CTV, and 0.99 +/-0.01, 0.99 +/- 0.03, 0.98 +/- 0.02 and 1.00 +/- 0.01 for the ITV, respectively. There was minimal dependence on period or initial phase. For typical tumor geometries and respiratory amplitudes, changes in target coverage are minimal but can be significant for larger amplitudes, faster beam delivery, more highly-modulated fields, and smaller field margins.

Evaluation of respiration-correlated digital tomosynthesis in lung.

Digital tomosynthesis (DTS) with a linear accelerator-mounted imaging system provides a means of reconstructing tomographic images from radiographic projections over a limited gantry arc, thus requiring only a few seconds to acquire. Its application in the thorax, however, often results in blurred images from respiration-induced motion. This work evaluates the feasibility of respiration-correlated (RC) DTS for soft-tissue visualization and patient positioning. Image data acquired with a gantry-mounted kilovoltage imaging system while recording respiration were retrospectively analyzed from patients receiving radiotherapy for non-small-cell lung carcinoma. Projection images spanning an approximately 30 degrees gantry arc were sorted into four respiration phase bins prior to DTS reconstruction, which uses a backprojection, followed by a procedure to suppress structures above and below the reconstruction plane of interest. The DTS images were reconstructed in planes at different depths through the patient and normal to a user-selected angle close to the center of the arc. The localization accuracy of RC-DTS was assessed via a comparison with CBCT. Evaluation of RC-DTS in eight tumors shows visible reduction in image blur caused by the respiratory motion. It also allows the visualization of tumor motion extent. The best image quality is achieved at the end-exhalation phase of the respiratory motion. Comparison of RC-DTS with respiration-correlated cone-beam CT in determining tumor position, motion extent and displacement between treatment sessions shows agreement in most cases within 2-3 mm, comparable in magnitude to the intraobserver repeatability of the measurement. These results suggest the method's applicability for soft-tissue image guidance in lung, but must be confirmed with further studies in larger numbers of patients.

Accurate positioning for head and neck cancer patients using 2D and 3D image guidance.

Our goal is to determine an optimized image-guided setup by comparing setup errors determined by two-dimensional (2D) and three-dimensional (3D) image guidance for head and neck cancer (HNC) patients immobilized by customized thermoplastic masks. Nine patients received weekly imaging sessions, for a total of 54, throughout treatment. Patients were first set up by matching lasers to surface marks (initial) and then translationally corrected using manual registration of orthogonal kilovoltage (kV) radiographs with DRRs (2D-2D) on bony anatomy. A kV cone beam CT (kVCBCT) was acquired and manually registered to the simulation CT using only translations (3D-3D) on the same bony anatomy to determine further translational corrections. After treatment, a second set of kVCBCT was acquired to assess intrafractional motion. Averaged over all sessions, 2D-2D registration led to translational corrections from initial setup of 3.5 ± 2.2 (range 0-8) mm. The addition of 3D-3D registration resulted in only small incremental adjustment (0.8 ± 1.5 mm). We retrospectively calculated patient setup rotation errors using an automatic rigid-body algorithm with 6 degrees of freedom (DoF) on regions of interest (ROI) of in-field bony anatomy (mainly the C2 vertebral body). Small rotations were determined for most of the imaging sessions; however, occasionally rotations > 3° were observed. The calculated intrafractional motion with automatic registration was < 3.5 mm for eight patients, and < 2° for all patients. We conclude that daily manual 2D-2D registration on radiographs reduces positioning errors for mask-immobilized HNC patients in most cases, and is easily implemented. 3D-3D registration adds little improvement over 2D-2D registration without correcting rotational errors. We also conclude that thermoplastic masks are effective for patient immobilization.

The effect of significant tumor reduction on the dose distribution in intensity modulated radiation therapy for head-and-neck cancer: a case study.

We present a unique case in which a patient with significant tissue loss was monitored for dosimetric changes using weekly cone beam computed tomography (CBCT) scans. A previously treated nasopharynx patient presented with a large, exophytic, recurrent left neck mass. The patient underwent re-irradiation to 70 Gy using intensity modulated radiation therapy (IMRT) with shielding blocks over the spinal cord and brain stem. Weekly CBCT scans were acquired during treatment. Target contours and treatment fields were then transferred from the original treatment planning computed tomography (CT) to the CBCT scans and dose calculations were performed on all CBCT scans and compared to the planning doses. In addition, a "research" treatment plan was created that assumed the patient had not been previously treated, and the above analysis was repeated. Finally, to remove the effects of setup error, the outer contours of 2 CBCT scans with significant tumor reductions were transferred to the planning scan and dose in the planning scan was recalculated. Planning treatment volume (PTV) decreased 45% during treatment. Spinal cord D05 differed from the planned value by 3.5 +/- 9.8% (average + standard deviation). Mean dose to the oral cavity and D05 of the mandible differed from the planned value by 0.9 +/- 2.1% and 0.6 +/- 1.5%, respectively. Results for the research plan were comparable. Target coverage did not change appreciably (-0.2 +/- 2.5%). When the planning scan was recalculated with the reduced outer contour from the CBCT, spinal cord D05 decreased slightly due to the reduction in scattered dose. Weekly imaging provided us the unique opportunity to use different methods to examine the dosimetric effects of an unusually large loss of tissue. We did not see that tissue loss alone resulted in a significant effect on the dose delivered to the spinal cord for this case, as most fluctuation was due to setup error. In the IGRT era, delivered dose distributions can be more readily determined during treatment, and this information can be useful in deciding whether replanning is necessary.

New technologies in radiation therapy: ensuring patient safety, radiation safety and regulatory issues in radiation oncology.

New technologies such as intensity modulated and image guided radiation therapy, computer controlled linear accelerators, record and verify systems, electronic charts, and digital imaging have revolutionized radiation therapy over the past 10-15 y. Quality assurance (QA) as historically practiced and as recommended in reports such as American Association of Physicists in Medicine Task Groups 40 and 53 needs to be updated to address the increasing complexity and computerization of radiotherapy equipment, and the increased quantity of data defining a treatment plan and treatment delivery. While new technology has reduced the probability of many types of medical events, seeing new types of errors caused by improper use of new technology, communication failures between computers, corrupted or erroneous computer data files, and "software bugs" are now being seen. The increased use of computed tomography, magnetic resonance, and positron emission tomography imaging has become routine for many types of radiotherapy treatment planning, and QA for imaging modalities is beyond the expertise of most radiotherapy physicists. Errors in radiotherapy rarely result solely from hardware failures. More commonly they are a combination of computer and human errors. The increased use of radiosurgery, hypofractionation, more complex intensity modulated treatment plans, image guided radiation therapy, and increasing financial pressures to treat more patients in less time will continue to fuel this reliance on high technology and complex computer software. Clinical practitioners and regulatory agencies are beginning to realize that QA for new technologies is a major challenge and poses dangers different in nature than what are historically familiar.

Comparison of kilovoltage cone-beam computed tomography with megavoltage projection pairs for paraspinal radiosurgery patient alignment and position verification.

PURPOSE: Implanted gold markers and megavoltage (MV) portal imaging are commonly used for setup verification of paraspinal tumors treated with high-dose, single-fraction radiotherapy. We investigated whether the use of kilovoltage cone-beam computed tomography (CBCT) imaging eliminates the need for marker implantation. METHODS AND MATERIALS: Patients with paraspinal disease who were eligible for single-fraction stereotactic body radiotherapy were accrued to an institutional review board-approved protocol. Each of 16 patients underwent implantation of fiducial markers near the target. The markers were visible on the MV images. Three MV image pairs were acquired for each patient (initial, verification, and final) and were registered to the reference images. Every MV pair was complemented by a CBCT scan. CBCT image registration was performed automatically by maximizing the mutual information using a region of interest that excluded the markers. The corrections, as determined from the MV images, were compared with these from CBCT and were used for actual patient setup. RESULTS: The mean and standard deviation of the absolute values of the differences between the CBCT and MV corrections were 1.0 +/- 0.7, 1.0 +/- 0.6, and 1.0 +/- 0.8 mm for the left-right, anteroposterior, and superoinferior directions, respectively. The absolute differences between the corresponding pre- and post-treatment kilovoltage CBCT image registration were 0.6 +/- 0.5, 0.6 +/- 0.5, and 1.0 +/- 0.8 mm. CONCLUSION: The setup corrections found using CBCT without the use of implanted markers were consistent with the marker registration on MV projections. CBCT has additional advantages, including better positioning precision and robust automatic three-dimensional registration, as well as eliminating the need for invasive marker implantation. We have adopted CBCT for the setup of all single-fraction paraspinal patients. Our data have also demonstrated that target displacements during treatment are insignificant.

Incidence of late rectal and urinary toxicities after three-dimensional conformal radiotherapy and intensity-modulated radiotherapy for localized prostate cancer.

PURPOSE: To report the incidence and predictors of treatment-related toxicity at 10 years after three-dimensional conformal radiotherapy (3D-CRT) and intensity-modulated radiotherapy (IMRT) for localized prostate cancer. METHODS AND MATERIALS: Between 1988 and 2000, 1571 patients with stages T1-T3 prostate cancer were treated with 3D-CRT/IMRT with doses ranging from 66 to 81 Gy. The median follow-up was 10 years. Posttreatment toxicities were all graded according to the National Cancer Institute's Common Terminology Criteria for Adverse Events. RESULTS: The actuarial likelihood at 10 years for the development of Grade>or=2 GI toxicities was 9%. The use of IMRT significantly reduced the risk of gastrointestinal (GI) toxicities compared with patients treated with conventional 3D-CRT (13% to 5%; p<0.001). Among patients who experienced acute symptoms the 10-year incidence of late toxicity was 42%, compared with 9% for those who did not experience acute symptoms (p<0.0001). The 10-year incidence of late Grade>or=2 genitourinary (GU) toxicity was 15%. Patients treated with 81 Gy (IMRT) had a 20% incidence of GU symptoms at 10 years, compared with a 12% for patient treated to lower doses (p=0.01). Among patients who had developed acute symptoms during treatment, the incidence of late toxicity at 10 years was 35%, compared with 12% (p<0.001). The incidence of Grade 3 GI and GU toxicities was 1% and 3%, respectively. CONCLUSIONS: Serious late toxicity was unusual despite the delivery of high radiation dose levels in these patients. Higher doses were associated with increased GI and GU Grade 2 toxicities, but the risk of proctitis was significantly reduced with IMRT. Acute symptoms were a precursor of late toxicities in these patients.

Radiation therapy of large intact breasts using a beam spoiler or photons with mixed energies.

Radiation treatment of large intact breasts with separations of more than 24 cm is typically performed using x-rays with energies of 10 MV and higher, to eliminate high-dose regions in tissue. The disadvantage of the higher energy beams is the reduced dose to superficial tissue in the buildup region. We evaluated 2 methods of avoiding this underdosage: (1) a beam spoiler: 1.7-cm-thick Lucite plate positioned in the blocking tray 35 cm from the isocenter, with 15-MV x-rays; and (2) combining 6- and 15-MV x-rays through the same portal. For the beam with the spoiler, we measured the dose distribution for normal and oblique incidence using a film and ion chamber in polystyrene, as well as a scanning diode in a water tank. In the mixed-energy approach, we calculated the dose distributions in the buildup region for different proportions of 6- and 15-MV beams. The dose enhancement due to the beam spoiler exhibited significant dependence upon the source-to-skin distance (SSD), field size, and the angle of incidence. In the center of a 20 x 20-cm(2) field at 90-cm SSD, the beam spoiler raises the dose at 5-mm depth from 77% to 87% of the prescription, while maintaining the skin dose below 57%. Comparison of calculated dose with measurements suggested a practical way of treatment planning with the spoiler--usage of 2-mm "beam" bolus--a special option offered by in-house treatment planning system. A second method of increasing buildup doses is to mix 6- and 15-MV beams. For example, in the case of a parallel-opposed irradiation of a 27-cm-thick phantom, dose to D(max) for each energy, with respect to midplane, is 114% for pure 6-, 107% for 15-MV beam with the spoiler, and 108% for a 3:1 mixture of 15- and 6-MV beams. Both methods are practical for radiation therapy of large intact breasts.

Using an onboard kilovoltage imager to measure setup deviation in intensity-modulated radiation therapy for head-and-neck patients.

The purpose of the present study was to use a kilovoltage imaging device to measure interfractional and intrafractional setup deviations in patients with head-and-neck or brain cancers receiving intensity-modulated radiotherapy (IMRT) treatment. Before and after IMRT treatment, approximately 3 times weekly, 7 patients were imaged using the Varian On-Board Imager (OBI: Varian Medical Systems, Palo Alto, CA), a kilovoltage imaging device permanently mounted on the gantry of a Varian 21EX LINAC (Varian Medical Systems). Because of commissioning of the remote couch correction of the OBI during the study, online setup corrections were performed on 2 patients. For the other 5 patients, weekly corrections were made based on a sliding average of the measured data. From these data, we determined the interfractional setup deviation (defined as the shift from the original setup position suggested by the daily image), the residual error associated with the weekly correction protocol, and the intrafractional setup deviation, defined as the difference between the post-treatment and pretreatment images. We also used our own image registration software to determine interfractional and intrafractional rotational deviations from the images based on the template-matching method. In addition, we evaluated the influence of inter-observer variation on our results, and whether the use of various registration techniques introduced differences. Finally, translational data were compared with rotational data to search for correlations. Translational setup errors from all data were 0.0 +/- 0.2 cm, -0.1 +/- 0.3 cm, and -0.2 +/- 0.3 cm in the right-left (RL), anterior-posterior (AP), and superior-inferior (SI) directions respectively. Residual error for the 5 patients with a weekly correction protocol was -0.1 +/- 0.2 cm (RL), 0.0 +/- 0.3 cm (AP), and 0.0 +/- 0.2 cm (SI). Intrafractional translation errors were small, amounting to 0.0 +/- 0.1 cm, -0.1 +/- 0.2 cm, and 0.0 +/- 0.1 cm in the RL, AP, and SI directions respectively. In the sagittal and coronal views respectively, interfractional rotational errors were -1.1 +/- 1.7 degrees and -0.5 +/- 0.9 degrees, and intrafractional rotational errors were 0.3 +/- 0.6 degrees and 0.2 +/- 0.5 degrees. No significant correlation was seen between translational and rotational data. The OBI image data were used to study setup error in the head-and-neck patients. Nonzero systematic errors were seen in the interfractional translational and rotational data, but not in the intrafractional data, indicating that the mask is better at maintaining head position than at reproducing it.

Validation of GATE Monte Carlo simulations of the GE Advance/Discovery LS PET scanners.

The recently developed GATE (GEANT4 application for tomographic emission) Monte Carlo package, designed to simulate positron emission tomography (PET) and single photon emission computed tomography (SPECT) scanners, provides the ability to model and account for the effects of photon noncollinearity, off-axis detector penetration, detector size and response, positron range, photon scatter, and patient motion on the resolution and quality of PET images. The objective of this study is to validate a model within GATE of the General Electric (GE) Advance/Discovery Light Speed (LS) PET scanner. Our three-dimensional PET simulation model of the scanner consists of 12 096 detectors grouped into blocks, which are grouped into modules as per the vendor's specifications. The GATE results are compared to experimental data obtained in accordance with the National Electrical Manufactures Association/Society of Nuclear Medicine (NEMA/SNM), NEMA NU 2-1994, and NEMA NU 2-2001 protocols. The respective phantoms are also accurately modeled thus allowing us to simulate the sensitivity, scatter fraction, count rate performance, and spatial resolution. In-house software was developed to produce and analyze sinograms from the simulated data. With our model of the GE Advance/Discovery LS PET scanner, the ratio of the sensitivities with sources radially offset 0 and 10 cm from the scanner's main axis are reproduced to within 1% of measurements. Similarly, the simulated scatter fraction for the NEMA NU 2-2001 phantom agrees to within less than 3% of measured values (the measured scatter fractions are 44.8% and 40.9 +/- 1.4% and the simulated scatter fraction is 43.5 +/- 0.3%). The simulated count rate curves were made to match the experimental curves by using deadtimes as fit parameters. This resulted in deadtime values of 625 and 332 ns at the Block and Coincidence levels, respectively. The experimental peak true count rate of 139.0 kcps and the peak activity concentration of 21.5 kBq/cc were matched by the simulated results to within 0.5% and 0.1% respectively. The simulated count rate curves also resulted in a peak NECR of 35.2 kcps at 10.8 kBq/cc compared to 37.6 kcps at 10.0 kBq/cc from averaged experimental values. The spatial resolution of the simulated scanner matched the experimental results to within 0.2 mm.

Intensity-modulated radiotherapy as the boost or salvage treatment of nasopharyngeal carcinoma: the appropriate parameters in the inverse planning and the effect of patient's anatomic factors on the planning results.

The current study demonstrates that the large increase in normal tissue penalty often degrades target dose uniformity without a concomitant large improvement in normal tissue dose, especially in anatomically unfavorable patients. The excessively large normal tissue penalties do not improve treatment plans for patients having unfavorable geometry.

Technological advances in external-beam radiation therapy for the treatment of localized prostate cancer.

The relative inability of conventional radiotherapy to control localized prostate cancer results from resistance of subpopulations of tumor clonogens to dose levels of 65 to 70 Gy, the maximum feasible with traditional two-dimensional (2D) treatment planning and delivery techniques. Several technological advances have enhanced the precision and improved the outcome of external-beam radiotherapy. The three-dimensional conformal radiotherapy (3D-CRT) approach has permitted significant increases in the tumor dose to levels beyond those feasible with conventional techniques. Intensity-modulated radiotherapy (IMRT), an advanced form of conformal radiotherapy, has resulted in reduced rectal toxicity, permitting tumor dose escalation to previously unattainable levels with a concomitant improvement in local tumor control and disease-free survival. The combination of androgen deprivation and conventional-dose radiotherapy, tested mainly in patients with locally advanced disease, has also produced significant outcome improvements. Whether androgen deprivation will preclude the need for dose escalation or whether high-dose radiotherapy will obviate the need for androgen deprivation remains unknown. In some patients, both approaches may be necessary to maximize the probability of cure. In view of the favorable benefit-risk ratio of high-dose IMRT, the design of clinical trials to resolve these critical questions is essential.

CT image-guided intensity-modulated therapy for paraspinal tumors using stereotactic immobilization.

PURPOSE: To design and implement a noninvasive stereotactic immobilization technique with daily CT image-guided positioning to treat patients with paraspinal lesions accurately and to quantify the systematic and random patient setup errors occurring with this method. METHODS AND MATERIALS: A stereotactic body frame (SBF) was developed for "rigid" immobilization of paraspinal patients. The inherent accuracy of this system for stereotactic CT-guided treatment was evaluated with phantom studies. Seven patients with thoracic and lumbar spine lesions were immobilized with the SBF and positioned for 33 treatment fractions using daily CT scans. For all 7 patients, the daily setup errors, as assessed from the daily CT scans, were corrected at each treatment fraction. A retrospective analysis was also performed to assess what the impact on patient treatment would have been without the CT-based corrections (i.e., if patient setup had been performed only with the SBF). RESULTS: The average magnitude of systematic and random errors from uncorrected patient setups using the SBF was approximately 2 mm and 1.5 mm (1 SD), respectively. For fixed phantom targets, the system accuracy for the SBF localization and treatment was shown to be within 1 mm (1 SD) in any direction. Dose-volume histograms incorporating these uncertainties for an intensity-modulated radiotherapy plan for lumbar spine lesions were generated, and the effects on the dose-volume histograms were studied. CONCLUSION: We demonstrated a very accurate and precise method of patient immobilization and treatment delivery based on a noninvasive SBF and daily image guidance for paraspinal lesions. The SBF provides excellent immobilization for paraspinal targets, with setup accuracy better than 2 mm (1 SD). However, for highly conformal paraspinal treatments, uncorrected systematic and random errors of 2 mm in magnitude can result in a significantly greater (>100%) dose to the spinal cord than planned, even though the planned target coverage may not change substantially. With daily CT guidance using the SBF, we showed that the maximal spinal cord dose is ensured to be within 10-15% of the planned value.

Evaluation of concave dose distributions created using an inverse planning system.

PURPOSE: To evaluate and develop optimum inverse treatment planning strategies for the treatment of concave targets adjacent to normal tissue structures. METHODS AND MATERIALS: Optimized dose distributions were designed using an idealized geometry consisting of a cylindrical phantom with a concave kidney-shaped target (PTV) and cylindrical normal tissues (NT) placed 5-13 mm from the target. Targets with radii of curvature from 1 to 2.75 cm were paired with normal tissues with radii between 0.5 and 2.25 cm. The target was constrained to a prescription dose of 100% and minimum and maximum doses of 95% and 105% with relative penalties of 25. Maximum dose constraint parameters for the NT varied from 10% to 70% with penalties from 10 to 1000. Plans were evaluated using the PTV uniformity index (PTV D(max)/PTV D(95)) and maximum normal tissue doses (NT D(max)/PTV D(95)). RESULTS: In nearly all situations, the achievable PTV uniformity index and the maximum NT dose exceeded the corresponding constraints. This was particularly true for small PTV-NT separations (5-8 mm) or strict NT dose constraints (10%-30%), where the achievable doses differed from the requested by 30% or more. The same constraint parameters applied to different PTV-NT separations yielded different dose distributions. For most geometries, a range of constraints could be identified that would lead to acceptable plans. The optimization results were fairly independent of beam energy and radius of curvature, but improved as the number of beams increased, particularly for small PTV-NT separations or strict dose constraints. CONCLUSION: Optimized dose distributions are strongly affected by both the constraint parameters and target-normal tissue geometry. Standard site-specific constraint templates can serve as a starting point for optimization, but the final constraints must be determined iteratively for individual patients. A strategy whereby NT constraints and penalties are modified until the highest acceptable PTV uniformity index is achieved is discussed. This strategy can be used, in simple patient geometries, to ensure the lowest possible normal tissue dose. Strategies for setting the optimum dose constraints and penalties may vary for different optimization algorithms and objective functions. Increasing the number of beams can significantly improve normal tissue dose and target uniformity in situations where the PTV-NT separation is small or the normal tissue dose limits are severe. Setting unrealistically severe constraints in such situations often results in dose distributions that are inferior to plans achieved with more lenient constraints.