Radiation-induced pulmonary toxicity: a dose-volume histogram analysis in 201 patients with lung cancer.

PURPOSE: To relate lung dose-volume histogram-based factors to symptomatic radiation pneumonitis (RP) in patients with lung cancer undergoing 3-dimensional (3D) radiotherapy planning. METHODS AND MATERIALS: Between 1991 and 1999, 318 patients with lung cancer received external beam radiotherapy (RT) with 3D planning tools at Duke University Medical Center. One hundred seventeen patients were not evaluated for RP because of <6 months of follow-up, development of progressive intrathoracic disease making scoring of pulmonary symptoms difficult, or unretrievable 3D dosimetry data. Thus, 201 patients were analyzed for RP. Univariate and multivariate analyses were performed to test the association between RP and dosimetric factors (i.e., mean lung dose, volume of lung receiving >or=30 Gy, and normal tissue complication probability derived from the Lyman and Kutcher models) and clinical factors, including tobacco use, age, sex, chemotherapy exposure, tumor site, pre-RT forced expiratory volume in 1 s, weight loss, and performance status. RESULTS: Thirty-nine patients (19%) developed RP. In the univariate analysis, all dosimetric factors (i.e., mean lung dose, volume of lung receiving >or=30 Gy, and normal tissue complication probability) were associated with RP (p range 0.006-0.003). Of the clinical factors, ongoing tobacco use at the time of referral for RT was associated with fewer cases of RP (p = 0.05). These factors were also independently associated with RP according to the multivariate analysis (p = 0.001). Models predictive for RP based on dosimetric factors only, or on a combination with the influence of tobacco use, had a concordance of 64% and 68%, respectively. CONCLUSIONS: Dosimetric factors were the best predictors of symptomatic RP after external beam RT for lung cancer. Multivariate models that also include clinical variables were slightly more predictive.

A practical and easy method to locate the first three internal mammary lymph node-bearing areas.

PURPOSE: To assess the distance from a clinically recognized anatomic landmark to the different costocondral interspaces in female patients to facilitate the design of radiation fields intended to include specific internal mammary nodal areas. METHODS AND MATERIALS: The distance from the suprasternal notch (SSN) to the caudal portion of the first through fourth interspace was measured on a computer display of the chest skeleton of 65 female patients with left-sided breast cancer. The relationship between these distances and bone size (sternal length and standing height) was assessed via linear regression. In 21 of the 65 patients where myocardial perfusion imaging of the heart was available, the relationship between the location of the 3rd costochondral interspace and the left ventricle was assessed. RESULTS: In 90% of patients (59/65), the first, second, third, and fourth interspace were within 5, 8.5, 11, and 14 cm of the SSN, respectively. The SSN-interspace distances did not correlate well with sternal length (r = 0.28) or standing height (r = 0.31). In 20 of 21 patients (95%), the third interspace "shadowed" the cephalad aspect of the left heart ventricle. Median "shadowing" was 3 cm (range 0.5-6 cm). CONCLUSION: The caudal portion of the third costochondral interspace is < or = 11 cm caudal to the SSN in 90% of patients. These measurements can be used to clinically design radiation therapy fields intended to treat the upper three interspaces. The distance from the SSN to the 1st through 4th interspaces is not related to sternal length or to standing height. In patients with left-sided breast cancer, radiation treatment fields designed to include the internal mammary lymph nodes in the upper three interspaces may incidentally include a portion of the heart.

Cardiac perfusion changes in patients treated for breast cancer with radiation therapy and doxorubicin: preliminary results.

PURPOSE: To determine the incidence and dose dependence of regional cardiac perfusion abnormalities in patients with left-sided breast cancer treated with radiation therapy (RT) with and without doxorubicin (Dox). METHODS: Twenty patients with left-sided breast cancer underwent cardiac perfusion imaging using single photon emission computed tomography (SPECT) prechemotherapy, pre-RT, and 6 months post-RT. SPECT perfusion images were registered onto 3-dimensional (3D) RT dose distributions. The volume of heart in the RT field was quantified, and the regional RT dose was calculated. A decrease in regional cardiac perfusion was assessed subjectively by visual inspection and objectively using image fusion software. Ten patients received Dox-based chemotherapy (total dose 120-300 mg/m(2)), and 10 patients had no chemotherapy. RT was delivered by tangent beams in all patients to a total dose of 46-50 Gy. RESULTS: Overall, 60% of the patients had new visible perfusion defects 6 months post-RT. A dose-dependent perfusion defect was seen at 6 months with minimal defect appreciated at 0-10 Gy, and a 20% decrease in regional perfusion at 41-50 Gy. One of 20 patients had a decrease in left ventricle ejection fraction (LVEF) of greater than 10% at 6 months; 2/20 patients had developed transient pericarditis. No instances of myocardial infarction or congestive heart failure (CHF) have occurred. CONCLUSIONS: RT causes cardiac perfusion defects 6 months post-RT in most patients. Long-term follow-up is needed to assess whether these perfusion changes are transient or permanent and to determine if these findings are associated with changes in overall cardiac function and clinical outcome.

Functional imaging of normal tissues with nuclear medicine: applications in radiotherapy.

Functional imaging techniques are gaining significant interest from radiation oncologists. Many now claim the need for physical and physiological information during both treatment planning and in the study of normal tissue injury. Toward this goal, the nuclear medicine functional imaging modalities, single-photon emission computed tomography and positron-emission computed tomography, have been used. This article reviews the studies performed in radiotherapy that used these modalities, and attempts to stimulate further interest in this topic.

Variability of the depth of supraclavicular and axillary lymph nodes in patients with breast cancer: is a posterior axillary boost field necessary?

PURPOSE: To determine the variability of the depth of supraclavicular (SC) and axillary (AX) lymph nodes in patients undergoing radiation therapy for breast cancer and to relate this variability with the patient's anterior/posterior (A/P) diameter. The dosimetric consequences of the variability in depth are explored and related to the need for a posterior axillary boost field. METHOD AND MATERIALS: In 49 patients undergoing treatment-planning computed tomography (CT) scanning in the treatment position, the maximum depth of the SC and AX lymph nodes was measured on CT images. The A/P diameter was measured at the location of the SC and AX, respectively. The relationship between the SC/AX lymph node depth and patient diameter was determined using linear regression. For an anterior SC and AX field, the relative dose to the SC and AX lymph nodes were calculated for a 6 MV photon beam. RESULTS: The maximum depth of the SC lymph nodes ranged from 2.4 to 9.5 cm (median, 4.3 cm). The depth was less than 3 cm in 4 patients, 3-6 cm in 39 (80%), and greater than 6 cm in 6 patients. There was a linear relationship between the SC lymph node depth and the A/P diameter. The depth of the SC lymph nodes in cm equals approximately one-half of the A/P diameter minus 3.5 (r(2) = 0.69). In 94% (46 of 49) of patients, the SC lymph node depth was between one-fifth and one-half of the A/P diameter. The depth of the axillary lymph nodes ranged from 1.4 to 8 cm (median, 4.3 cm). The depth was less than 3 cm in 8 patients, 3-6 cm in 32 (65%), and greater than 6 cm in 9 patients. The AX lymph node depth in cm equals approximately one-half of the A/P diameter minus 3 (r(2) = 0.81). In all patients, the AX lymph nodes were shallower than mid-depth. The depth of the SC and AX lymph nodes was within +/- 1 cm in 53% (26 of 49) of patients. The AX lymph nodes were located at >/= 1 cm shallower or greater depth than the SC in 24.5% (12 of 49) and 22.5% (11 of 49) of patients, respectively. If an anterior 6-MV beam only is used to treat the SC and AX lymph nodes in these 49 patients, the dose to the AX is within +/- 5% of the SC dose in 53% (26 of 49) patients and is 90% or more of the dose delivered in the SC in 90% (44 of 49) of patients. CONCLUSION: The maximum depth of the SC and AX lymph nodes varies widely and is related to the patient's size represented by the A/P diameter. In most patients, the AX lymph nodes lie at approximately the same depth or shallower than the SC. Therefore, the rationale for a posterior axillary boost field needs to be further assessed. When the AX and SC lymph nodes are deep, opposed supraclavicular and axillary fields and/or the use of a higher energy beam might be reasonable.

Can angiotensin-converting enzyme inhibitors protect against symptomatic radiation pneumonitis?

This study was designed to determine whether patients taking angiotensin-converting enzyme (ACE) inhibitors while receiving radiation therapy for lung cancer are protected from developing symptomatic radiation pneumonitis. The records of 213 eligible patients receiving thoracic irradiation for lung cancer with curative intent at Duke University Medical Center from 1994-1997 were reviewed. Of the 213 patients, 26 (12.2%) were on ACE inhibitors (usually for the management of hypertension) during radiotherapy (group 1); the remaining 187 patients (group 2) were not. Patients were irradiated, with fields shaped to protect normal tissues, with total doses of 50-80 Gy. After treatment, patients were generally followed every 3 months for 2 years, then every 6 months thereafter. Symptomatic radiation pneumonitis was scored according to modified National Cancer Institute Common Toxicity Criteria (i.e., radiographic changes alone were not sufficient for the diagnosis of pneumonitis). There was no difference in the incidence of pneumonitis between the two groups (P = 0.75). Fifteen percent of the patients on ACE inhibitors (group 1) developed symptomatic radiation-induced lung injury compared to 12% of the patients not receiving these drugs (group 2). Although patients in group 1 tended to develop pneumonitis slightly sooner than did patients in group 2, this difference also was not significant (P = 0. 8). Within the dose range prescribed for treating hypertension, ACE inhibitors do not appear to either decrease the incidence or delay the onset of symptomatic radiation pneumonitis among lung cancer patients receiving thoracic irradiation.

The effect of pressure from the table top and patient position on pelvic organ location in patients with prostate cancer.

PURPOSE: To assess the impact of pressure from the table top and patient position on the relationship of the prostate, rectum, and bladder to the bony pelvis. METHODS AND MATERIALS: In 9 patients with prostate cancer (3 status postprostatectomy), computed tomography (CT) scans were obtained in four positions: supine with and without false table top under the buttocks, prone with and without false table top under the lower abdomen. In four patients, a fifth scan was obtained in the first position (supine with table top in place) to assess the impact of changes in bladder/rectal fullness over time. Urination and defecation were not permitted between scans. For each patient, the four (or five) CT scans were registered to each other. RESULTS: The anal canal and the rectum caudal to the coccyx shifted posteriorly in 7/9 patients when the support under the buttocks was removed in the supine position. When pressure from the table top was removed in the prone position, the anterior bladder extension increased. The superior rectum was adjacent to the prostate in all scans and the prostate/superior rectum/bladder generally moved together. Rectal fullness changed with time and rectal gas position was gravity-dependent and shifted with patient position. Bladder volume increased with time. Organs had shifted and/or changed fullness between the first and fifth scan obtained in the same patient position approximately 90 min apart, mostly due to increase in bladder volume. All patients found the supine position most comfortable. CONCLUSIONS: The bladder and rectal fullness vary with time, confounding the ability to attribute changes in organ location to positional factors. Pressure from the table top affects the relative location of pelvic organs and, in part, is responsible for changes previously attributed to position/gravity.

The effect of patient-specific factors on radiation-induced regional lung injury.

PURPOSE: To assess the impact of patient-specific factors on radiation (RT)-induced reductions in regional lung perfusion. METHODS: Fifty patients (32 lung carcinoma, 7 Hodgkin's disease, 9 breast carcinoma and 2 other thoracic tumors) had pre-RT and > or = 24-week post-RT single photon emission computed tomography (SPECT) perfusion images to assess the dose dependence of RT-induced reductions in regional lung perfusion. The SPECT data were analyzed using a normalized and non-normalized approach. Furthermore, two different mathematical methods were used to assess the impact of patient-specific factors on the dose-response curve (DRC). First, DRCs for different patient subgroups were generated and compared. Second, in a more formal statistical approach, individual DRCs for regional lung injury for each patient were fit to a linear-quadratic model (reduction = coefficient 1 x dose + coefficient 2 x dose2). Multiple patient-specific factors including tobacco history, pre-RT diffusion capacity to carbon monoxide (DLCO), transforming growth factor-beta (TGF-beta), chemotherapy exposure, disease type, and mean lung dose were explored in a multivariate analysis to assess their impact on the coefficients. RESULTS: None of the variables tested had a consistent impact on the radiation sensitivity of regional lung (i.e., the slope of the DRC). In the formal statistical analysis, there was a suggestion of a slight increase in radiation sensitivity in the dose range >40 Gy for nonsmokers (vs. smokers) and in those receiving chemotherapy (vs. no chemotherapy). However, this finding was very dependent on the specific statistical and normalization method used. CONCLUSION: Patient-specific factors do not have a dramatic effect on RT-induced reduction in regional lung perfusion. Additional studies are underway to better clarify this issue. We continue to postulate that patient-specific factors will impact on how the summation of regional injury translates into whole organ injury. Refinements in our methods to generate and compare SPECT scans are needed.

A neural network to predict symptomatic lung injury.

A nonlinear neural network that simultaneously uses pre-radiotherapy (RT) biological and physical data was developed to predict symptomatic lung injury. The input data were pre-RT pulmonary function, three-dimensional treatment plan doses and demographics. The output was a single value between 0 (asymptomatic) and 1 (symptomatic) to predict the likelihood that a particular patient would become symptomatic. The network was trained on data from 97 patients for 400 iterations with the goal to minimize the mean-squared error. Statistical analysis was performed on the resulting network to determine the model's accuracy. Results from the neural network were compared with those given by traditional linear discriminate analysis and the dose-volume histogram reduction (DVHR) scheme of Kutcher. Receiver-operator characteristic (ROC) analysis was performed on the resulting network which had Az = 0.833 +/- 0.04. (Az is the area under the ROC curve.) Linear discriminate multivariate analysis yielded an Az = 0.813 +/- 0.06. The DVHR method had Az = 0.521 +/- 0.08. The network was also used to rank the significance of the input variables. Future studies will be conducted to improve network accuracy and to include functional imaging data.

Acute and late morbidity of using a breast positioning ring in women with large/pendulous breasts.

PURPOSE: To assess acute and late effects of radiation therapy in women with breast cancer treated with a breast positioning ring. MATERIALS AND METHODS: Fifty-six patients with large and/or pendulous breasts were irradiated using a breast positioning ring. The incidence of acute morbidity was correlated with patient weight and breast 'size'. Cosmesis was scored at > or = 1 year following radiation therapy by the patients. Dose changes in the buildup region under the ring were measured using a computer-controlled scanning system. RESULTS: Moist desquamation (MD) occurred in 60.7% (34/56) of patients treated with the breast ring. The incidence of MD was more common in patients with larger breasts (P = 0.08), the severity necessitating a treatment interruption in 5 out of 56 (9%) patients. Cosmesis at > or = 1 year following radiation therapy was scored as > or = good by all patients. The surface dose under the ring was approximately 85% of the Dmax dose. CONCLUSIONS: The incidence or severity of acute MD in patients treated with a breast positioning ring appears high in patients with large pendulous breasts, and might be related in part, to the increased skin dose due to the positioning ring. To date, there appears to be no significant late normal tissue effects in patients treated with the positioning ring. Additional follow up is needed to assess the long-term consequences of the ring on cosmesis.

Postmastectomy radiotherapy: toxicities and techniques to reduce them.

The role of locoregional radiation therapy after mastectomy is controversial. It reduces the risk of tumor relapse, improves breast cancer-specific survival and possibly overall survival, but has potential morbidity. This article reviews the technical aspects of postmastectomy radiation therapy and its associations with treatment-related morbidity. We consider common problems that arise in the technical setup of radiation fields. Adverse effects of postmastectomy radiation therapy may be reduced or prevented by careful radiation treatment planning.

The effect of beam divergence on target coverage.

The well-known fact that radiation beams diverge is frequently not considered during the treatment planning process. Complacency with respect to beam divergence can, in some situations, lead to inappropriate field design. In this review, the potential problems arising from failure to adequately account for beam divergence in treatment planning are outlined, and commonly encountered clinical examples are illustrated.

Incorporation of functional status into dose-volume analysis.

The dose-volume histogram (DVH) has gained wide acceptance as a mechanism for reducing the voluminous data of a three-dimensional dose distribution into a two-dimensional graph. These graphs are often converted to a single figure of merit. This data reduction technique is used both for clinical treatment plan evaluation and as part of proposed systems for estimating control and complication probabilities. It has long been recognized that a major shortcoming of the DVH as an analysis tool is that all spatial information is discarded. A subtler problem, which is addressed in this work, is that the DVH also implies homogeneity of biological consequence of irradiation in what may be a functionally heterogeneous volume of tissue. An extension to the DVH, the functional dose-volume histogram, or dose-function histogram (DFH), is proposed, that explicitly includes quantitative three-dimensional functional information. The concept is illustrated by the use of SPECT imaging to assess the functional status of irradiated lung.

Differences in gross target volumes on contrast vs. noncontrast CT scans utilized for conformal radiation therapy treatment planning for prostate carcinoma.

PURPOSE: To compare the gross target volumes (GTVs) (prostate and seminal vesicles) defined on noncontrast and contrast-enhanced computed tomography (CT) images used for three-dimensional conformal treatment planning (3DCRT). METHODS AND MATERIALS: From 1993 to 1996, 39 patients referred for radiation therapy for adenocarcinoma of the prostate underwent pretreatment pelvic CT scanning with and without intravenous (i.v.) contrast for treatment planning purposes. Seven patients were excluded because of incomplete data sets. The prostate and seminal vesicles were outlined by the same physician on all images of 32 patients. On 18 CT exams, the prostate and seminal vesicles were blindly outlined a second time by the same physician to evaluate intraphysician consistency. Discrepancies between the GTVs outlined with and without contrast and between the first and second outline on the same study were assessed by calculating the projected area in the anterior-to-posterior (AP) and right lateral (RLAT) beam's-eye view (BEV). To assess the magnitude, frequency, and direction of discrepancies between the two GTVs, the extension of the GTVs in six directions (right, left, anterior, posterior, cephalad, and caudal) was determined. RESULTS: The GTV outlined with contrast was larger in all directions, except caudal, in the majority of patients. The change in the GTV with contrast was significant in the cephalad (p=0.0003) and right (p=0.0007) directions, but not in the other directions. Although the increase with contrast in any direction was usually small (average < or =5 mm), these changes resulted in a significant increase in GTV area in both the AP and RLAT BEV (9.0%, p=0.0017 and 8.2%, p=0.023, respectively). The intraphysician variability in outlining the prostate/ seminal vesicles was minimal. CONCLUSIONS: The addition of i.v. contrast does appear to make a significant difference in how the prostate and seminal vesicles are outlined by an experienced observer. The increase in area of the target, found when contrast is used, should be taken into consideration when designing the treatment fields for patients with carcinoma of the prostate.

Standard vs conformal radiation therapy for adenocarcinoma of the prostate: no difference.

Objective: To compare results of treatment of adenocardinoma of the prostate using Standard (2D) vs Conformal (3D) treatment planning. Methods: The records of all patients with adenocarcinoma of the prostate treated curatively with radiation therapy alone from July 1991 to June 1994 were reviewed. Acute and late complications were scored by the RTOG criteria. Biochemical failure was defined as a rising PSA of at least 10% on two measurements separated >/=1 month or either a PSA nadir >4 ng/ml or >1 ng/ml. Disease free survival (DFS) was defined as no evidence of local, distant, or biochemical failure. 2D planning included standard simulation with target volume drawn from the treatment planning or diagnostic CT. 3D planning included a CT in the treatment position with computer simulation using beam's-eye-view for field design. Results: Two-hundred and seventeen 2D and 45 3D patients had similar median age and pre-treatment PSA, T-stage, and dose to the prostate. The median follow-up periods for the 2D and 3D groups were 32.0 and 21.5 months, respectively. The two-year actuarial survival, local or biochemical control, and DFS were not different. The 3D group had a significantly higher incidence of acute bladder side effects of all grades and acute grade 1/2 rectal complications. There were no differences in the incidence of late bladder or rectal complications. Conclusions: Careful 2D planning for the treatment of localized adenocarcinoma of the prostate is an acceptable means of treatment. Within the dose range of 64-70 Gy, this preliminary analysis demonstrated no reduction in complications nor improvement in local or biochemical control, or DFS was seen with the the use of 3D treatment planning.

Physical and biological predictors of changes in whole-lung function following thoracic irradiation.

PURPOSE: To develop methods of predicting the pulmonary consequences of thoracic irradiation (RT) by prospectively studying changes in pulmonary function following RT. METHODS AND MATERIALS: 100 patients receiving incidental partial-lung irradiation during treatment of tumors in or adjacent to the thorax had whole-lung function assessed via symptoms and pulmonary function tests (PFTs: FEV1-forced expiratory volume 1 s; DLCO-diffusion capacity) before and repeatedly 6-48 months following RT. All had computed tomography-based three-dimensional (3D) dose calculations with lung density heterogeneity corrections for dose-volume histogram (DVH) and normal tissue complication probability (NTCP) calculations. Functional DVHs (DVfH) based on SPECT (single photon emission computed tomography) lung perfusion scans, and serial transforming growth factor-beta (TGF-beta1) levels were available in 50 and 48 patients, respectively. The incidence and severity of changes in whole-lung function were correlated with clinical, physical, and biological factors. Exploratory statistical analyses were performed using chi-square, Pearson correlations, logistic regression, and multiple linear regression. RESULTS: RT-induced symptoms developed in 21 patients. In the overall group, the single best predictor for the development of symptoms was the NTCP (p < 0.05). Pre-RT PFTs alone were less predictive (p = 0.1 for FEV1, p = 0.08 for DLCO). A multivariate model based on pre-RT DLCO and CT-based NTCP was strongly predictive for the development of symptoms (p < 0.001). NTCPs based on SPECT-derived DVf Hs and TGF-beta1 levels did not appear to provide additional predictive value. The presence or absence of pulmonary symptoms was correlated with the decline in PFT 6 months following RT (p < 0.05). In the overall group, the degree of decline in PFTs was not well correlated with any of the dose-volume variables considered. In patients with "good" pre-RT PFTs, there was a relationship between the percent reduction in PFT and dose-volume parameters such as the percent of lung volume receiving > 30 Gy (p < 0.05). CONCLUSION: The extent of alteration in whole-lung function (symptoms or PFT changes) appears to be related to both dose-volume and pre-RT PFT parameters. The data suggest that no one variable is likely to be an adequate predictor and that multivariate predictive models will be needed. Additional studies are underway to develop better predictive models that consider physical factors such as the DVH and regional perfusion, as well as biological/clinical factors such as pre-RT PFTs and TGF-beta1.

Comparison of two repositioning devices used during radiation therapy for Hodgkin's disease.

PURPOSE: Patients irradiated for Hodgkin's disease are fixed in an immobilization cradle to improve repositioning. In the early 1990s, we changed our cradle system from a "short" upper torso cradle to an extended near-total body cradle that also includes the lower torso and thighs. In this study, we assess the impact of the extended cradle on the reproducibility of patient repositioning during irradiation of Hodgkin's disease. METHODS AND MATERIALS: A total of 782 port films of 56 patients treated immediately before and after the change-over were studied to assess positioning reproducibility. Patients treated prior to 1993 were positioned in the short cradle, while those treated 1993 and later were positioned in the extended cradle. All treatment were delivered via anterior and posterior fields and treatment areas above and below the diaphragm were considered separately and together. All treatment fields were simulated and the field shape was designed on anterior and posterior radiographs. Discrepancies in field placement between the simulation radiographs and subsequent port films were noted by a radiation oncologist and requests for position adjustment (both translational and rotational shifts) were noted. The number, magnitude, and direction of any physician-requested position adjustment on port films were retrospectively reviewed. For the purpose of scoring the frequency of field misplacements, when an adjustment was noted on two port films taken during the same treatment session (i.e., a left shift on both an anterior and a posterior port film), it was scored as only one event. A two-tailed chi-square test was used to compare the differences in requested shifts in the two patient groups. RESULTS: The study population consisted of 56 patients (31 short and 25 extended cradle) representing 92 treatment sites. A total of 782 port films representing 450 treatment setups were analyzed (292 above and 158 below the diaphragm). When all port films above the diaphragm (mostly mantle fields) are considered, position adjustments were requested in 13.4% (21 out of 157) of treatment setups with the upper torso cradle and in 5.9% (8 out of 135) of treatment setups with the extended cradle (p = 0.054). When all port films below the diaphragm (mostly paraaortic/spleen and pelvic fields) are considered, position adjustments were requested in 33.8% (27 out of 80) of treatment setups with the upper torso cradle and in 16.7 % (13 out of 78) of treatment setups with the extended cradle (p = 0.056). A reduction in the frequency of both translational and rotational adjustments were seen. When both treatment sites are combined, position adjustments were requested in 20.3% (48 out of 237) of treatment setups with the upper torso cradle and in 9.9% (21 out of 213) of treatment setups when the extended cradle was used (p = 0.0086). CONCLUSIONS: The extended cradle provides superior repositioning of patients undergoing radiation therapy for Hodgkin's disease. Differences observed in setup accuracy in this study underscore the importance of aggressive immobilization of patients with Hodgkin's disease. Increased accuracy of daily setup may provide an opportunity to improve the therapeutic ratio both by increased likelihood of tumor control and decreased risk of normal tissue complications.

Comparison of two head and neck immobilization systems.

PURPOSE: Accurate and reproducible patient positioning is fundamental to the success of fractionated radiotherapy. Concurrent with the introduction of three-dimensional treatment planning capabilities at our institution, a head and neck immobilization system consisting of a standard foam rubber head support and three casting strips was replaced by a customized mask-based device. This study was performed to analyze the impact of the customized immobilization system on the reproducibility of patient setup during irradiation of head and neck and brain tumors. METHODS AND MATERIALS: Patients treated from 1989-1991 were immobilized with the strip system while those treated from 1991-1995 were immobilized with the mask. All treatment fields were simulated and were treated on a 4 MV (where the strip, but not the mask, system was fixed to the treatment couch) or > or = 6 MV (where both the strip and the mask systems were fixed to the couch) accelerator. Port films were taken on the initial treatment day, routinely during treatment, and following shifts (requested). The number, magnitude, and direction of any isocenter shifts were retrospectively reviewed. A two-tailed chi square test was used to compare the differences in requested shifts in the strip and mask groups. RESULTS: The study population consisted of 69 brain tumor (35 strip, 34 mask) and 71 head and neck (37 strip, 34 mask) patients. A total of 1575 port films representing 1070 isocenter placements were analyzed. No differences between the immobilization systems was seen on the 4-MV accelerator (where the mask system was not fixed to the couch). On the > or = 6-MV units, the frequency of shifts was 16.1% versus 6.2% (p = 0.002) with the strips and mask, respectively. Almost all of the benefit was seen in the routine films, where the corresponding rates were 13.2% and 4.1% (p = 0.007). For the mask system, the rate of requested shifts on routine films was 4.1% (8/197) for the > or = 6-MV units and 14.5% (24/166) for the 4-MV unit (p = 0.001). CONCLUSION: Using the frequency of physician-requested isocenter shifts as an indicator of the accuracy of patient repositioning, the newer mask system appears to be an improvement over the previously used strip system, provided that the immobilization device is secured to the treatment couch. Increased accuracy of daily setup provides an opportunity to improve the therapeutic ratio both by increased likelihood of tumor control and decreased risk of normal tissue complications.

Impact of cradle immobilization on setup reproducibility during external beam radiation therapy for lung cancer.

PURPOSE: To compare the setup accuracy during fractionated radiation therapy for two patient groups with lung cancer treated with and without an immobilization cradle. METHODS: Three hundred ninety-seven port films from 30 patients immobilized in the Alpha Cradle were compared with 329 port films from 30 patients who were not immobilized with the cradle. All patients were treated with curative intent for nonmetastatic lung cancer. The frequency of physician-requested isocenter shifts were compared in the two groups using a two-tailed chi-square test. Initial port films taken on the first day of treatment, routine films taken usually weekly during radiation therapy, and requested films taken after a requested shift were considered separately. The immobilization device consisted of a custom-made foam cradle that extended from above the head to the knees. Patients were generally treated with their arms above their heads, and treatment setup marks in the immobilized patients were placed on both the patients' skin and the immobilization cradle. For the noncradle patients, setup marks were placed only on the patients' skin. RESULTS: For the routine films, the frequency of physician-requested isocenter shifts was lower in immobilized patients than in the nonimmobilized group (p = 0.139). Most of this reduction was seen on oblique fields (p = 0.038). No benefits were seen among initial or requested films. The two groups were well balanced with regard to stage, age, field size, and total dose. CONCLUSIONS: The use of aggressive immobilization improves the setup reproducibility in patients receiving external beam radiation therapy for lung cancer, especially during treatment with oblique fields. This improvement in treatment accuracy might improve the therapeutic ratio.

Quantification of radiation-induced regional lung injury with perfusion imaging.

PURPOSE: To better understand the dose and time dependence of radiation therapy (RT)-induced regional lung dysfunction as assessed by changes in regional lung perfusion. METHODS AND MATERIALS: Patients who were to receive RT for tumors in and around the thorax, wherein portions of healthy lung would be incidentally irradiated, were prospectively studied. Regional function was assessed pre- and post-RT with single photon emission computed tomography (SPECT) lung perfusion scans, obtained following the intravenous administration of approximately 4 mCi of technetium-99m macroaggregated albumin. Pre-RT computed tomography (CT) scans were used to calculate the three-dimensional (3D) dose distribution, reflecting tissue density inhomogeneity corrections. Each SPECT scan was correlated with the pre-RT CT scan, and the 3D dose distribution. Changes in regional lung perfusion were correlated with regional RT dose, at various time intervals following radiation. RESULTS: The data from 20 patients (7 breast cancer, 5 lymphoma, 1 esophagus, 1 sarcoma, and 6 lung cancer) have been analyzed. Patients with gross intrathoracic lung cancers causing obstruction of regional pulmonary arteries were not included. For most patients, there is a statistically significant dose-dependent reduction in regional blood flow at all time points following radiation. While a time dependence is suggested in the high dose range, the limited amount of data prevents meaningful statistical evaluation. CONCLUSIONS: Radiation therapy-induced regional lung dysfunction occurs in a dose-dependent manner and develops within 3-6 months following radiation. In contrast to classical "sigmoid" dose-response curves, described mainly for changes following whole lung irradiation, these data suggest a more gradual relationship between regional dysfunction and RT dose. Retraction of irradiated lung with secondary movement of unirradiated lung into the "3D-defined irradiated volume" may have introduced inaccuracies into this analysis. Additional studies are currently underway to assess this possibility and better refine this dose-response curve. Studies are underway to determine if changes in assessments of whole lung function, such as pulmonary function tests, can be predicted by summing the regional changes observed.