William B. Harms

A technique for the quantitative evaluation of dose distributions

The commissioning of a three-dimensional treatment planning system requires comparisons of measured and calculated dose distributions. Techniques have been developed to facilitate quantitative comparisons, including superimposed isodoses, dose-difference, and distance-to-agreement (DTA) distributions. The criterion for acceptable calculation performance is generally defined as a tolerance of the dose and DTA in regions of low and high dose gradients, respectively. The dose difference and DTA distributions complement each other in their useful regions. A composite distribution has recently been developed that presents the dose difference in regions that fail both dose-difference and DTA comparison criteria. Although the composite distribution identifies locations where the calculation fails the preselected criteria, no numerical quality measure is provided for display or analysis. A technique is developed to unify dose distribution comparisons using the acceptance criteria. The measure of acceptability is the multidimensional distance between the measurement and calculation points in both the dose and the physical distance, scaled as a fraction of the acceptance criteria. In a space composed of dose and spatial coordinates, the acceptance criteria form an ellipsoid surface, the major axis scales of which are determined by individual acceptance criteria and the center of which is located at the measurement point in question. When the calculated dose distribution surface passes through the ellipsoid, the calculation passes the acceptance test for the measurement point. The minimum radial distance between the measurement point and the calculation points (expressed as a surface in the dose-distance space) is termed the gamma index. Regions where gamma > 1 correspond to locations where the calculation does not meet the acceptance criteria. The determination of gamma throughout the measured dose distribution provides a presentation that quantitatively indicates the calculation accuracy. Examples of a 6 MV beam penumbra are used to illustrate the gamma index.

CLINICAL DOSE-VOLUME HISTOGRAM ANALYSIS FOR PNEUMONITIS AFTER 3D TREATMENT FOR NON-SMALL CELL LUNG CANCER (NSCLC)

PURPOSE To identify a clinically relevant and available parameter upon which to identify non-small cell lung cancer (NSCLC) patients at risk for pneumonitis when treated with three-dimensional (3D) radiation therapy. METHODS AND MATERIALS Between January 1991 and October 1995, 99 patients were treated definitively for inoperable NSCLC. Patients were selected for good performance status (96%) and absence of weight loss (82%). All patients had full 3D treatment planning (including total lung dose-volume histograms [DVHs]) prior to treatment delivery. The total lung DVH parameters were compared with the incidence and grade of pneumonitis after treatment. RESULTS Univariate analysis revealed the percent of the total lung volume exceeding 20 Gy (V20), the effective volume (Veff) and the total lung volume mean dose, and location of the tumor primary (upper versus lower lobes) to be statistically significant relative to the development of > or = Grade 2 pneumonitis. Multivariate analysis revealed the V20 to be the single independent predictor of pneumonitis. CONCLUSIONS The V20 from the total lung DVH is a useful parameter easily obtained from most 3D treatment planning systems. The V20 may be useful in comparing competing treatment plans to evaluate the risk of pneumonitis for our individual patient treatment and may also be a useful parameter upon which to stratify patients or prospective dose escalation trials.

Preliminary report of toxicity following 3D radiation therapy for prostate cancer on 3DOG/RTOG 9406

PURPOSE A prospective Phase I dose escalation study was conducted to determine the maximally-tolerated radiation dose in men treated with three-dimensional conformal radiation therapy (3D CRT) for localized prostate cancer. This is a preliminary report of toxicity encountered on the 3DOG/RTOG 9406 study. METHODS AND MATERIALS Each participating institution was required to implement data exchange with the RTOG 3D quality assurance (QA) center at Washington University in St. Louis. 3D CRT capabilities were strictly defined within the study protocol. Patients were registered according to three stratification groups: Group 1 patients had clinically organ-confined disease (T1,2) with a calculated risk of seminal vesicle invasion of < 15%. Group 2 patients had clinical T1,2 disease with risk of SV invasion > or = 15%. Group 3 (G3) patients had clinical local extension of tumor beyond the prostate capsule (T3). All patients were treated with 3D techniques with minimum doses prescribed to the planning target volume (PTV). The PTV margins were 5-10 mm around the prostate for patients in Group 1 and 5-10 mm around the prostate and SV for Group 2. After 55.8 Gy, the PTV was reduced in Group 2 patients to 5-10 mm around the prostate only. Minimum prescription dose began at 68.4 Gy (level I) and was escalated to 73.8 Gy (level II) and subsequently to 79.2 Gy (level III). This report describes the acute and late toxicity encountered in Group 1 and 2 patients treated to the first two study dose levels. Data from RTOG 7506 and 7706 allowed calculation of the expected probability of observing a > or = grade 3 late effect more than 120 days after the start of treatment. RTOG toxicity scores were used. RESULTS Between August 23, 1994 and July 2, 1997, 304 Group 1 and 2 cases were registered; 288 cases were analyzable for toxicity. Acute toxicity was low, with 53-54% of Group 1 patients having either no or grade 1 toxicity at dose levels I and II, respectively. Sixty-two percent of Group 2 patients had either none or grade 1 toxicity at either dose level. Few patients (0-3%) experienced a grade 3 acute bowel or bladder toxicity, and there were no grade 4 or 5 toxicities. Late toxicity was very low in all patient groups. The majority (81-85%) had either no or mild grade 1 late toxicity at dose level I and II, respectively. A single late grade 3 bladder toxicity in a Group 2 patient treated to dose level II was recorded. There were no grade 4 or 5 late effects in any patient. Compared to historical RTOG controls (studies 7506, 7706) at dose level I, no grade 3 or greater late effects were observed in Group 1 and Group 2 patients when 9.1 and 4.8 events were expected (p = 0.003 and p = 0.028), respectively. At dose level II, there were no grade 3 or greater toxicities in Group 1 patients and a single grade 3 toxicity in a Group 2 patient when 12.1 and 13.0 were expected (p = 0.0005 and p = 0.0003), respectively. Multivariate analysis demonstrated that the relative risk of developing acute bladder toxicity was 2.13 if the percentage of the bladder receiving > or = 65 Gy was more than 30% (p = 0.013) and 2.01 if patients received neoadjuvant hormonal therapy (p = 0.018). The relative risk of developing late bladder complications also increased as the percentage of the bladder receiving > or = 65 Gy increased (p = 0.026). Unexpectedly, there was a lower risk of late bladder complications as the mean dose to the bladder and prescription dose level increased. This probably reflects improvement in conformal techniques as the study matured. There was a 2.1 relative risk of developing a late bowel complication if the total rectal volume on the planning CT scan exceeded 100 cc (p = 0.019). CONCLUSION Tolerance to high-dose 3D CRT has been better than expected in this dose escalation trial for Stage T1,2 prostate cancer compared to low-dose RTOG historical experience. With strict quality assurance standards and review, 3D CRT can be safely studied in a co

Dose-volume histograms

A plot of a cumulative dose-volume frequency distribution, commonly known as a dose-volume histogram (DVH), graphically summarizes the simulated radiation distribution within a volume of interest of a patient which would result from a proposed radiation treatment plan. DVHs show promise as tools for comparing rival treatment plans for a specific patient by clearly presenting the uniformity of dose in the target volume and any hot spots in adjacent normal organs or tissues. However, because of the loss of positional information in the volume(s) under consideration, it should not be the sole criterion for plan evaluation. DVHs can also be used as input data to estimate tumor control probability (TCP) and normal tissue complication probability (NTCP). The sensitivity of TCP and NTCP calculations to small changes in the DVH shape points to the need for an accurate method for computing DVHs. We present a discussion of the methodology for generating and plotting the DVHs, some caveats, limitations on their use and the general experience of four hospitals using DVHs.

Preliminary results of a prospective trial using three dimensional radiotherapy for lung cancer

PURPOSE To evaluate the preliminary results of a prospective trial using three-dimensional (3D) treatment for lung cancer. METHODS AND MATERIALS Seventy patients with inoperable Stage I through IIIB lung cancer were treated with three-dimensional thoracic irradiation with or without chemotherapy (35% received chemotherapy). Total prescribed dose to the tumor ranged from 60-74 Gy (uncorrected for lung density). All patients were evaluated for local control, survival, and development of pneumonitis. These parameters were evaluated in respect to and compared with three-dimensional parameters used in their treatment planning. RESULTS With a minimum follow-up of 6 to 30 months, the 2-year cause-specific survival rate for Stages I and II was 90% and 53% for Stage III (no difference between Stages IIIA and IIIB). Patients with local tumor control had a better 2-year overall survival rate (47%) than those with local failure (31%). Volumetrically heterogeneously calculated doses were important to the accurate delineation of dose-volume coverage as there was a wide range of discrepancies between a homogeneously prescribed point dose calculation and the heterogeneously calculated volume coverage of that prescription. High-grade pneumonitis was correlated with the location of the tumor with lower lobe tumors having a much higher risk than those with upper lobe tumors. A critical volume effect and threshold dose were apparent in the development of high-grade pneumonitis. CONCLUSIONS Three-dimensional therapy for lung cancer has been practically implemented at the Mallinckrodt Institute of Radiology and shows promising results in our preliminary analysis. The incidence of high-grade pneumonitis, however, warrants careful selection of patients for future dose escalation. Future dose escalation trials in lung cancer should be directed to volumes that limit the amount of elective nodal irradiation. However, the volume of necessary elective nodal irradiation remains unknown and should be studied prospectively. Dose escalation trials are indicated and may be facilitated by smaller target volumes.

A software tool for the quantitative evaluation of 3D dose calculation algorithms

Current methods for evaluating modern radiation therapy treatment planning (RTP) systems include the manual superposition of calculated and measured isodose curves and the comparison of a limited number of calculated and measured point doses. Both techniques have significant limitations in providing quantitative evaluations of the large number of dose data generated by modern RTP systems. More sophisticated comparison techniques have been presented in the literature, including dose-difference and distance-to-agreement (DTA) analyses. A software tool has been developed that uses superimposed isodose plots, dose-difference, and DTA distributions to quantify errors in computed dose distributions. Dose-difference and DTA analyses are overly sensitive in regions of high- and low-dose gradient, respectively. The logical union of locations that fail both dose-difference and DTA acceptance criteria, termed the composite evaluation, is calculated and displayed. The composite evaluation provides a method for the physicist to efficiently identify regions that fail both the dose-difference and DTA acceptance criteria. The tool provides a computer platform for the quantitative comparison of calculated and measured dose distributions.

Verification data for electron beam dose algorithms

The Collaborative Working Group (CWG) of the National Cancer Institute (NCI) electron beam treatment planning contract has performed a set of 14 experiments that measured dose distributions for 28 unique beam-phantom configurations that simulated various patient anatomic structures and beam geometries. Multiple dose distributions were measured with film or diode detectors for each configuration, resulting in 78, 2-D planar dose distributions and one, 1-D depth-dose distribution. Measurements were made for 9- and 20-MeV electron beams, using primarily 6 x 6- and 15 x 15-cm applicators at several SSDs. Dose distributions were measured for shaped fields, irregular surfaces, and inhomogeneities (1-D, 2-D, and 3-D), which were designed to simulate many clinical electron treatments. The data were corrected for asymmetries, and normalized in an absolute manner. This set of measured data can be used for verification of electron beam dose algorithms and is available to others for that purpose.

Clinical implementation of a commercial multileaf collimator: Dosimetry, networking, simulation, and quality assurance

PURPOSE Clinical implementation of multileaf collimation (MLC) includes commissioning (including leaf calibration), dosimetric measurements (penumbra, transmission, calculation parameters), shaping methods, networking for file transfer, verification simulation, and development of a quality assurance (QA) program. Differences of MLC and alloy shaping in terms of penumbra and stair-step effects must be analyzed. METHODS AND MATERIALS Leaf positions are calibrated to light field. The resultant decrement line, penumbras, leaf transmission data, and isodoses in various planes were measured with film. Penumbra was measured for straight edges and corners, in various media. Ion chambers were used to measure effects of MLC on output, scatter, and depth dose. We maintain midleaf intersection criteria. MLC fields are set 7 mm beyond planning target volumes. After shaping by vendor software or by our three-dimensional planning system, files are transferred to the MLC workstation by means of sharing software, interface cards, and cabling. A MLC emulator was constructed for simulation. Our QA program includes file checks, monthly checks (leaf position accuracy and interlock tests), and annual review. RESULTS We found the MLC leaf position (light field) corresponds to decrement lines ranging from 50 to 59%. Transmission through MLC (1.5-2.5%) is less than alloy (3.5%). Multileaf penumbra is slightly wider than for alloy. Relative penumbra did not increase in the lung, and composite field dosimetry exhibited negligible differences compared with alloy. Verification simulations provide diagnostic image quality hard copies of the MLC fields. Monitor unit parameters used for alloy held for MLC. DISCUSSION Clinical implementation for MLC as a block replacement was conducted on a site-by-site basis. Time studies indicate significant (25%) in-room time reductions. Through imaging and dosimetric analysis, the accuracy of field delivery has increased with MLC. The most significant impact of MLC is the ability to increase the number of daily treatment fields, thereby reducing normal tissue dosing, which is vital for dose escalation.

THREE-DIMENSIONAL RADIATION TREATMENT PLANNING STUDY FOR PATIENTS WITH CARCINOMA OF THE LUNG

PURPOSE Several reports in the literature suggest that local-regional control and possibly survival could be improved for inoperable nonsmall cell lung cancer if the radiation dose to the target volume could be increased. Higher doses, however, bring with them the potential for increased side effects and complications of normal tissues. Three-dimensional treatment planning has shown significant potential for improving radiation treatment planning in several sites, both for tumor coverage and for sparing of normal tissue from high doses of radiation and, thus, has the potential of developing radiation therapy techniques that result in uncomplicated local-regional control of lung cancer. We have studied the feasibility of large-scale implementation of true three-dimensional technologies in the treatment of patients with cancers of the thorax. METHODS AND MATERIALS CT scans were performed on 10 patients with inoperable nonsmall cell lung cancer to obtain full volumetric image data, and therapy was planned on our three-dimensional radiotherapy treatment planning system. Target volumes were determined using the new ICRU nomenclature--Gross Tumor Volume, Clinical Target Volume, and Planning Target Volume. Plans were performed according to our standard treatment policies based on traditional two-dimensional radiotherapy treatment planning methodologies and replanned using noncoplanar three-dimensional beam techniques. The results were quantitatively compared using dose-volume histograms, dose-surface displays, and dose statistics. RESULTS Target volume delineation remains a difficult problem for lung cancer. Defining Gross Tumor Volume and Clinical Target Volume may depend on window and level settings of the three-dimensional radiotherapy treatment planning system, suggesting that target volume delineation on hard copy film is inadequate. Our study shows that better tumor coverage is possible with three-dimensional plans. Dose to critical structures (e.g., the heart) could often be reduced (or at least remain acceptable) using noncoplanar beams even with dose escalation to 75 to 80 Gy for the planning volume surrounding the Gross Target Volume. CONCLUSION Commonly used beam arrangements for treatment of lung cancer appear to be inadequate to safely deliver tumor doses of higher than 70 Gy. Although conventional treatment techniques may be adequate for tumor coverage, they are inadequate for sparing of normal tissues when the prescription dose is escalated. The ability to use noncoplanar fields for such patients is a major advantage of three-dimensional planning. This capability led to better tumor coverage and reduced dose to critical normal tissues. However, this advantage was achieved at the expense of a greater time commitment by the treatment planning staff (particularly the radiation oncologist) and a greater complexity of treatment delivery. In summary, three-dimensional radiotherapy treatment planning appears to provide the radiation oncologist with the necessary tools to increase tumor dose, which may lead to increased local-regional control in patients with lung cancer while maintaining normal tissue doses at acceptable tolerance levels.

Dosimetry of therapeutic photon beams using an extended dose range film

For intensity modulated radiation therapy (IMRT) dose distribution verification, multidimensional measurements are required to quantify the steep dose-gradient regions. High resolution, two-dimensional dose distributions can be measured using radiographic film. However, the photon energy response of film is known to be a function of depth, field size, and photon beam energy, potentially reducing the accuracy of dose distribution measurements. The dosimetric properties of the recently developed Kodak EDR2 film were investigated and compared to those of Kodak XV film. The dose responses of both film types to 6 MV and 18 MV photon beams were investigated for depths of 5 cm, 10 cm, and 15 cm and field sizes of 4x4 cm2 and 15x15 cm2. This analysis involved the determination of sensitometric curves for XV and EDR2 films, the determination of dose profiles from exposed XV and EDR2 films, and comparison of the film-generated dose profiles to ionization chamber measurements. For the combinations of photon beam energy, depth, and field size investigated here, our results indicate that the sensitometric curves are nearly independent of field size and depth of calibration. For a field size of 4x4 cm2, a single sensitometric curve for either EDR2 and XV film can be used for the determination of relative dose profiles. For the larger field size, the sensitometric curve for EDR2 film is superior to XV film in regions where the dose falls below 20% of the central axis dose, due to the effects that the increased low energy scattered photon contributions have on film response. The limited field size and depth dependence of sensitometric data measured using EDR2 film, along with the inherently wide linear dose-response range of EDR2 film, makes it better suited to the verification of IMRT dose distributions.