A BOOST TECHNIQUE FOR IRRADIATION OF MALIGNANT CANINE NASAL TUMORS

Eighteen dogs with malignant nasal cavity tumors were treated with radiation therapy, including a boost technique. Three 3:0 Gy boost doses were added to a treatment protocol consisting of sixteen 3.0 Gy daily fractions, bringing the total dose to 57 Gy. This boost technique was implemented without an associated increase in overall treatment time by giving the boost doses on a twice-a-day basis. Boost doses were given during the first half of the radiation therapy period. The treatment was completed as planned in 16 of the 18 dogs; two dogs received lower doses (51 and 54 Gy). Median survival was 177 days, poorer than in some other reported studies of nasal tumor irradiation. Acute effects were unacceptable, with 11 of the 18 dogs developing severe mucositis, desquamation, edema, swelling, and pruritus. The extensive nature of the acute reactions compromised assessment of the effect of the increased radiation dose on the tumor. Although there is justification for assessing more aggressive radiation protocols in canine nasal tumor patients, total doses approximating 60 Gy can not be given as described because of the inability of acutely responding normal tissues to compensate.     

COMBINATION OF RADIATION THERAPY AND FIROCOXIB FOR THE TREATMENT OF CANINE NASAL CARCINOMA

Carcinomas represent two‐thirds of canine nasosinal neoplasms. Although radiation therapy (RT) is the standard of care, the incidence of local recurrence following treatment is high. Cyclooxygenase‐isoform‐2 (COX‐2) is expressed in 71–95% of canine nasal carcinomas and has been implicated in tumor growth and angiogenesis. Accordingly, COX‐2 inhibition seems rational to improve outcome. Dogs with histologically confirmed, previously untreated nasal carcinomas were randomized to receive the combination of a selective COX‐2 inhibitor (firocoxib) and palliative RT (Group 1) or RT and placebo (Group 2). Patients were regularly monitored with blood tests, urinalysis, and computed tomography. Pet owners were asked to complete monthly a quality‐of‐life questionnaire. Twenty‐four dogs were prospectively enrolled. According to Adams modified system, there were five stage 1, five stage 2, three stage 3, and 11 stage 4 tumors. Two dogs had metastases to regional lymph nodes. Median progression‐free interval and overall survival were 228 and 335 days in Group 1 (n = 12) and 234 and 244 days in Group 2 (n = 12). These differences were not statistically significant. The involvement of regional lymph nodes was significantly associated with progression‐free interval and overall survival (P = 0.004). Quality of life was significantly improved in Group 1 (P = 0.008). In particular, a significant difference was observed for activity and appetite. Although not providing a significant enhancement of progression‐free interval and overall survival, firocoxib in combination with RT is safe and improved life quality in dogs with nasal carcinomas.     

AN ACCELERATED TECHNIQUE FOR IRRADIATION OF MALIGNANT CANINE NASAL AND PARANASAL SINUS TUMORS

Tumor and normal tissue response was assessed in 21 dogs with malignant nasal tumors given 42 Gy cobalt radiation in 9 or 10 fractions over 11 to 13 days. Local tumor/clinical relapse recurred in 68% of dogs, with a median relapse free interval (RFI) of 270 days. Median survival was 428 days. One year survival for all dogs was 60%. RFI and survival times are better than, or similar to, previous reports of dogs treated with radiotherapy only. Acute radiation effects were severe in one dog. Late effects were severe in six of 15 dogs (40%) with durable tumor control. Late effects included bilateral blindness (3), osteoradionecrosis (3), and seizures (1). These six dogs had a median survival of 705 days. Loss of vision occurred in at least one eye in nine dogs (47%). Tumor staging based on CT findings were predictive for survival duration. Tumor histology was not predictive of outcome. Labrador Retrievers were significantly over-represented. Despite comparable or improved tumor control and survival times provided by this accelerated protocol, relative to other radiotherapy reports, local failure remains the major cause of death, and late radiation effects can be severe in dogs with durable tumor control.     

USE OF RADIATION AND A SLOW-RELEASE CISPLATIN FORMULATION FOR TREATMENT OF CANINE NASAL TUMORS

The purpose of this study was to evaluate the combined use of radiation and a slow-release cisplatin chemotherapy formulation for treatment of malignant nasal tumors in dogs. In this retrospective analysis, 51 dogs were evaluated with respect to treatment toxicity, tumor type, stage of disease, cribriform plate involvement, and overall survival. In general, treatment was well tolerated. Mean and median survival as assessed by the Kaplan-Meier product limit method was 570 and 474 days, respectively. No other factors, including tumor type, stage of disease, or cribriform plate invasion had a significant impact on survival. In conclusion, a combination of slow release cisplatin chemotherapy and radiation for the treatment of canine nasal tumors is well tolerated. Results of this analysis warrant further study to elucidate possible other beneficial radiation potentiating drugs and dosing schedules.     

PROGNOSTIC SIGNIFICANCE OF TUMOR HISTOLOGY AND COMPUTED TOMOGRAPHIC STAGING FOR RADIATION TREATMENT RESPONSE OF CANINE NASAL TUMORS

Prognostic significance of tumor histology and four computed tomography (CT) staging methods was tested retrospectively in dogs from three treatment centers that underwent intent-to-cure-radiotherapy for intranasal neoplasia. Disease-free and overall survival times were available for 94 dogs. A grouping of anaplastic, squamous cell, and undifferentiated carcinomas had a significantly shorter median disease-free survival (4.4 mo) than a grouping of all sarcomas (10.6 months). Disease-free survivals were not significantly different, when all carcinomas were compared with all sarcomas. The published original and modified WHO staging methods did not significantly relate to either survival endpoint. A modified human maxillary tumor staging system previously applied to canine nasal tumors was prognostically significant for both survival endpoints; a further modified version of that CT-based staging system resulted in improved significance for both survival endpoints. Dogs with unilateral intranasal involvement without bone destruction beyond the turbinates on CT, had longest median survival (23.4 months); CT evidence of cribriform plate involvement was associated with shortest median survival (6.7 months). Combining CT and histology statistically improved prognostic significance for both survival endpoints over the proposed CT staging method alone. Significance was lost when CT stages were collapsed to 相似文献   

SLOW RELEASE CISPLATIN COMBINED WITH RADIATION FOR THE TREATMENT OF CANINE NASAL TUMORS

Thirteen dogs with malignant tumors of the nasal cavity were treated with a combination of slow release cisplatin and megavoltage radiation. Radiation was delivered on a Monday through Friday schedule using a 6 MV linear accelerator. The median total dose was 49.5 Gy (range 49.5-56 Gy). Cisplatin was given using an open-cell polylactic acid polymer, impregnated with the drug and implanted intramus-cularly at a distant site, as a slow release delivery system (OPLAn-Pt [THM Biomedical, Inc]). The median dose used was 60 mg/m2 (range 60–100 mg/m2). When combined with radiation, this delivery system caused no systemic drug toxicity, and a local tissue reaction was seen in only two dogs. Acute side effects to normal tissue from radiation were not enhanced, as measured by subjective assessment. When compared to a group of historical controls that received radiation without OPLA-Pt, the dogs that received combined radiation and cisplatin had longer overall survival times, with a median of 580 days. The control group had a median survival of 325 days. Previously reported median survival times for comparable megavoltage radiation treatment range from 6 to 13 months. Some dogs in both groups also received adjubant chemotherapy but this did not influence survival time. By multivariate analysis, only the use of OPLA-Pt was found to significantly influence survival, with a p value of p = 0.023. Mega-voltage radiation and slow release cisplatin appears to be a well tolerated combination that may favorably affect survival of dogs with nasal tumors.     

DOSIMETRIC IMPACT OF DAILY SETUP VARIATIONS DURING TREATMENT OF CANINE NASAL TUMORS USING INTENSITY-MODULATED RADIATION THERAPY

Intensity-modulated radiation therapy (IMRT) can be employed to yield precise dose distributions that tightly conform to targets and reduce high doses to normal structures by generating steep dose gradients. Because of these sharp gradients, daily setup variations may have an adverse effect on clinical outcome such that an adjacent normal structure may be overdosed and/or the target may be underdosed. This study provides a detailed analysis of the impact of daily setup variations on optimized IMRT canine nasal tumor treatment plans when variations are not accounted for due to the lack of image guidance. Setup histories of ten patients with nasal tumors previously treated using helical tomotherapy were replanned retrospectively to study the impact of daily setup variations on IMRT dose distributions. Daily setup shifts were applied to IMRT plans on a fraction-by-fraction basis. Using mattress immobilization and laser alignment, mean setup error magnitude in any single dimension was at least 2.5 mm (0–10.0 mm). With inclusions of all three translational coordinates, mean composite offset vector was 5.9±3.3 mm. Due to variations, a loss of equivalent uniform dose for target volumes of up to 5.6% was noted which corresponded to a potential loss in tumor control probability of 39.5%. Overdosing of eyes and brain was noted by increases in mean normalized total dose and highest normalized dose given to 2% of the volume. Findings suggest that successful implementation of canine nasal IMRT requires daily image guidance to ensure accurate delivery of precise IMRT distributions when non-rigid immobilization techniques are utilized. Unrecognized geographical misses may result in tumor recurrence and/or radiation toxicities to the eyes and brain.     

POTENTIAL FOR INTENSITY-MODULATED RADIATION THERAPY TO PERMIT DOSE ESCALATION FOR CANINE NASAL CANCER

We evaluated the impact of inverse planned intensity-modulated radiation therapy (IMRT) on the dose-volume histograms (DVHs) and on the normal tissue complication probabilities (NTCPs) of brain and eyes in dogs with nasal tumors. Nine dogs with large, caudally located nasal tumors were planned using conventional techniques and inverse planned IMRT for a total prescribed dose of 52.5 Gy in 3.5 Gy fractions. The equivalent uniform dose for brain and eyes was calculated to estimate the normal tissue complication probability (NTCP) of these organs. The NTCP values as well as the DVHs were used to compare the treatment plans. The dose distribution in IMRT plans was more conformal than in conventional plans. The average dose delivered to one-third of the brain was 10 Gy lower with the IMRT plan compared with conventional planning. The mean partial brain volume receiving 43.6 Gy or more was reduced by 25.6% with IMRT. As a consequence, the NTCPs were also significantly lower in the IMRT plans. The mean NTCP of brain was two times lower and at least one eye could be saved in all patients planed with IMRT. Another possibility with IMRT is dose escalation in the target to improve tumor control while keeping the NTCPs at the same level as for conventional planning. Veterinary     

A COMPARISON OF NORMAL TISSUE COMPLICATION PROBABILITY OF BRAIN FOR PROTON AND PHOTON THERAPY OF CANINE NASAL TUMORS

This study compared the calculated normal tissue complication probability of brain in dogs with a nasal tumor, which had both photon and proton treatment planning. Nine dogs diagnosed with a variety of histologies, but all with large, caudally located nasal tumors were studied. Three-dimensional (3-D) photon dose distribution, and a proton dose distribution was calculated for each dog. To calculate the normal tissue complication probability (NTCP) for brain, the partial brain volume irradiated with the prescribed dose was determined, then a mathematic model relating complications to partial volume and radiation dose was used. The NTCP was always smaller for proton plans as compared to photon plans, indicating conformation of the dose to the target allows a higher dose to be given. If a 5% NTCP were accepted, the mean applicable dose for this group of dogs was 50.2 Gy for photons, but 58.3 Gy for protons. Not all dogs would benefit the same from proton irradiation. If a large partial brain volume has to be irradiated, the advantage becomes minimal. There is also a minimal advantage if the planning target volume (PTV) includes a small, superficial brain volume. However, for a complex PTV shape the degree of conformation is clearly superior for protons and results in smaller calculated NTCPs.     

REIRRADIATION OF RECURRENT CANINE NASAL TUMORS

Canine nasal tumors are typically treated with radiation therapy but most patients develop local recurrence. Our purpose was to evaluate tumor and normal tissue response to reirradiation in nine dogs. The median dose delivered with the first protocol was 50 Gy (range 44–55 Gy) and the median fraction number was 18 (range 15–20). For the second protocol, the median dose was lower intentionally, median of 36 Gy (range 23–44 Gy), without changing the median fraction number of 18 (range 14–20) to avoid late effects. The median time between protocols was 539 days (range 258–1652 days). Median survival was 927 days (95% confidence interval [CI] 423–1767 days). Median time to progression following the first and second courses was 513 days (95% CI 234–1180 days) and 282 days (95% CI 130–453 days), respectively. These were not significantly different (P=0.086). The qualitative response assessment was better for the first course compared with the second (P=0.018). Severity and timing of skin, mucous membrane, and ocular effects were similar for early side effects between the two courses (P>0.05 for all comparisons). All dogs experienced some late side effects, with two out of nine being classified as severe. These severe effects were blindness in each dog, possibly related to tumor recurrence. Reirradiation of canine nasal tumors resulted in a second clinical remission in eight of nine dogs, although the second response was less complete. Acute and late effects for seven of nine patients were not life threatening, indicating that reirradiation of canine nasal tumors may be a viable treatment option after recurrence.     

RETROSPECTIVE COMPARISON OF THREE‐DIMENSIONAL CONFORMAL RADIATION THERAPY VS. PREDNISOLONE ALONE IN 30 CASES OF CANINE INFRATENTORIAL BRAIN TUMORS

Infratentorial tumors are relatively infrequent in dogs and a lack of data makes it difficult to offer prognostic information. Untreated, dogs with these neoplasms have shorter survival times than those with supratentorial tumors. The role of radiation therapy (RT) in the management of infratentorial tumors is poorly defined and tumoral/peritumoral swelling in this site is a potential cause of serious acute side effects. The aim of this retrospective, cohort study was to describe cases of infratentorial tumors treated with fractionated three‐dimensional conformal RT (3D CRT) and glucocorticoids (GC), and compare outcomes and survival with dogs affected by tumors in the same location that received GC alone. Thirty patients with a MRI diagnosis of infratentorial tumors were recruited (15 received RT and GC and 15 GC alone). None had mentation changes at presentation. For both groups, MRI and medical records were reviewed; and factors associated with survival were evaluated with Kaplan–Meier product limit survival and Cox regression analysis. Overall median survival time (MST) was 294 days (95% CI 42–545). The MST in the RT group was 756 days (95% CI 209–1302) vs. 89 days (95% CI 34.7–143.3 days) for those dogs treated palliatively with GC alone. This difference was statistically significant (P = 0.001). No other factors (including neurological signs, MRI features, tumor volume and total RT dose) were statistically associated with survival in the RT group. This study suggests that 3D CRT offers a survival advantage for dogs with infratentorial tumors compared to GC alone, and significant complications are uncommon.     

RADIOBIOLOGICAL AND TREATMENT PLANNING STUDY OF A SIMULTANEOUSLY INTEGRATED BOOST FOR CANINE NASAL TUMORS USING HELICAL TOMOTHERAPY

Feasibility of delivering a simultaneously integrated boost to canine nasal tumors using helical tomotherapy to improve tumor control probability (TCP) via an increase in total biological equivalent uniform dose (EUD) was evaluated. Eight dogs with varying size nasal tumors (5.8-110.9 cc) were replanned to 42 Gy to the nasal cavity and integrated dose boosts to gross disease of 45.2, 48.3, and 51.3 Gy in 10 fractions. EUD values were calculated for tumors and mean normalized total doses (NTD(mean)) for organs at risk (OAR). Normal Tissue Complication Probability (NTCP) values were obtained for OARs, and estimated TCP values were computed using a logistic dose-response model and based on deliverable EUD boost doses. Significant increases in estimated TCP to 54%, 74%, and 86% can be achieved with 10%, 23%, and 37% mean relative EUD boosts to the gross disease, respectively. NTCP values for blindness of either eye and for brain necrosis were < 0.01% for all boosts. Values for cataract development were 31%, 42%, and 46% for studied boost schemas, respectively. Average NTD(mean) to eyes and brain for mean EUD boosts were 10.2, 11.3, and 12.1 Gy3, and 7.5, 7.2, and 7.9 Gy2, respectively. Using helical tomotherapy, simultaneously integrated dose boosts can be delivered to increase the estimated TCP at 1-year without significantly increasing the NTD(mean) to eyes and brain. Delivery of these treatments in a prospective trial may allow quantification of a dose-response relationship in canine nasal tumors.     

RADIATION THERAPY FOR INCOMPLETELY RESECTED CANINE MAST CELL TUMORS

The records of 56 dogs treated with megavoltage radiation for mast cell neoplasia were reviewed to determine the efficacy of this treatment modality. Total radiation dose ranged from 45 to 57 Gray (Gy), dose per fraction ranged from 3.0 to 4.0 Gy, and radiation treatment time ranged from 14–28 days. Median disease free interval (95% CI) was 32.7 (19–70) months. Median disease free interval for dogs older than 7.5 years was 15 (lower limit 7) months as compared to 62 (lower limit 20) for dogs younger than 7.5 years of age (p = 0.006). Median disease free interval for dogs with measurable disease was 12 (lower limit 5) months as compared to 54 (32–70) months for dogs with microscopic disease (p = 0.006). Radiation treatment time was also significantly related to disease free interval. Median disease free interval for dogs treated longer than 22 days was 12 (7–19) months as compared to greater than 50 (lower limit 20) months for dogs treated in 22 or fewer days (p < 0.001). This appeared to be due to more recurrences in dogs treated with 3-per-week fractionation and suggests that tumor proliferation in the interfraction interval may be important. Sex, tumor location, histologic grade, WHO clinical stage, number of radiation fractions, total radiation dose, and dose-per-fraction, as well as the following "yes/no" variables: steroids given, surgery prior to radiation, lymph nodes irradiated, and development of another mast cell tumor did not appear to influence median disease free interval or survival. Data presented herein support megavoltage radiation as an effective treatment for canine mast cell neoplasia, and suggest that disease free interval in dogs treated with daily fractions may be longer than that achieved with alternating day fractions.     

RESPONSE OF NINETEEN CATS WITH NASAL LYMPHOMA TO RADIATION THERAPY AND CHEMOTHERAPY

The records of 19 cats treated for stage I nasal lymphoma with radiation therapy and chemotherapy were reviewed to determine response to therapy, treatment outcome and possible prognostic indicators. All cats were treated with megavoltage radiation therapy to a total dose ranging from 22 to 48 Gy (median dose = 42 Gy). All cats were prescribed at least 6 months of multiagent chemotherapy. The median progression-free interval for all cats was 945 days (31 months). Two cats did not achieve clinical remission. Of 17 cats evaluable for relapse, 10 (58.8%) were progression free during the entire follow-up period. Four cats (23.5%) suffered local recurrence, while three (17.6%) experienced distant relapse. The median survival time was 955 days (31.4 months). The only variable found to have a significant negative impact on survival was destruction of the cribriform plate before therapy (P= 0.002). The long progression free and survival times reported here indicate that cats with stage I nasal lymphoma treated with aggressive local and systemic therapy can have a favorable outcome when compared with other anatomic forms of lymphoma. Despite strong clinical responses to the multimodality therapy used, the fact that three (17.6%) cats relapsed distantly supports the recommendation that treatment with radiation therapy alone is insufficient until further prospective studies can be performed.     

FACTORS INFLUENCING SURVIVAL AFTER RADIOTHERAPY OF NASAL TUMORS IN 130 DOGS

Improvements in survival of dogs with nasal tumors have been slow to develop throughout the past three decades. Despite multiple studies examining various radiation time-dose schema, the advancement of CT-based computerized treatment planning, and the evaluation of detailed staging systems, the optimal treatment regimen, and most important prognostic factors regarding survival remain unclear. In this study, data from four previous studies were combined with data from 44 additional dogs, and this population of 130 dogs was evaluated for factors which influenced survival. Twenty-one dogs were treated with orthovoltage at the University of Pennsylvania. One hundred nine dogs were treated with cobalt photons at North Carolina State University. Sixty-five of these 109 dogs had been previously described. Of the 44 dogs not previously described, 35 were treated with a shrinking field technique. Survival was determined from the medical record, or from information derived by telephone or mail survey. The univariate Cox regression model was used to examine for relationship between various patient, tumor, and treatment variables and survival. Significant relationships identified in the univariate analysis were further analyzed using the multivariate Cox regression model. Median survival of the 130 dogs was 8.9 months (95% C.I., 8-11 months). In the univariate analysis, the following variables were associated with decreased survival: 1) age >10 years old, 2) regional lymph node metastasis, 3) advanced tumor stage, 4) use of megavoltage radiation, 5) overall total dose >55 Gray, and 6) boost technique performed. In a multivariate analysis of 125 dogs with complete data for age, radiation type, and radiation dose, age (p < .001) and radiation type (p = .02) were identified as joint predictors of survival. After adjusting for age, the staging system lost prognostic significance (p = .06). In a subset of dogs that received cobalt radiation, after adjusting for age, dogs treated with a boost technique had decreased survival (p = .001). In general, local control of canine nasal tumors following aggressive radiation therapy is poor. Early diagnosis and selection of appropriate patients is warranted and palliative types of treatment should be considered in dogs with a poor chance of long term survival.     

RESPONSE OF FELINE ORAL SQUAMOUS CELL CARCINOMA TO PALLIATIVE RADIATION THERAPY

Seven cats with advanced oral squamous cell carcinoma were treated with palliative radiotherapy. Megavoltage radiation in 8 Gray (Gy) fractions was delivered on days 0, 7, and 21 for a total dose of 24 Gy. Treatment field included the mandible, oropharynx, retropharyngeal lymph nodes, and tonsils. Adjuvant treatment with chemotherapy was variable. Age ranged from 13 to 18 years old with a median age of 15 years. Three of the seven cats (43%) did not complete treatment. Six cats were euthanized due to tumor growth and/or radiation side effects with a median survival time of 60 days (range = 42 to 97 days, mean = 63 +/- 8.4 days). Radiotherapy complications or progression of disease occurred in 6 of 7 (85.7 %) cats and included adverse clinical signs, such as mucositis, serosanguinous oral discharge, pain, and dysphagia. These data suggest that coarse fractionation radiotherapy did not result in palliation in cats with inoperable oral squamous cell carcinoma.     

ORTHOVOLTAGE RADIOTHERAPY OF CANINE TRANSMISSIBLE VENEREAL TUMORS

Eighteen dogs with transmissible venereal tumors were treated with orthovoltage radiotherapy. Total radiation doses ranged from 10 to 30 Gy, given during periods of one to 34 days. Tumor cure, defined as lack of recurrence within one year after completion of radiotherapy, was observed in all 18 dogs. In seven of eight dogs, tumors were cured with a single dose of 10 Gy. The tumor that recurred after the single 10 Gy dose had been treated with chemotherapy prior to radiotherapy and was subsequently cured with additional radiotherapy. Data presented indicate that transmissible venereal tumors are radiocurable and that a single dose of 10 Gy is sufficient for cure in most dogs.     

DYNAMIC CT MEASUREMENT OF CONTRAST MEDIUM WASHIN KINETICS IN CANINE NASAL TUMORS

Tumor oxygenation affects the biologic behavior of a tumor and also its radiation response. Decreased tumor oxygenation has been associated with an aggressive phenotype and with decreased local tumor control following irradiation. Thus, measurement of oxygenation may be useful for pretreatment evaluation of a tumor. Many methods for assessing tumor oxygenation are available but most are invasive. There is a need for a non-invasive measure of oxygenation, or a surrogate for oxygenation. Measurement of perfusion has been suggested as a substitute for measurement of oxygenation. The use of washin kinetics of iodinated contrast medium to estimate perfusion has been shown to be related to radiation response of human carcinomas. We quantified the washin kinetics of iodinated contrast medium using dynamic CT in 9 dogs. All dogs had a malignant nasal tumor and perfusion was quantified at two sites in each tumor to evaluate intratumoral variation in perfusion. Dogs were given an intravenous bolus injection of contrast medium and arterial and tumor washin kinetics quantified using a helical CT scanner. Perfusion was estimated from these data using previously validated methods. Eight of the 9 dogs received definitive radiation therapy and perfusion was quantified a second time in these 8 dogs midway through irradiation. Pretreatment perfusion varied between dogs by a factor of 16.9. Between dog variation in perfusion was subjectively greater than within tumor variation based on comparison of two intratumoral regions. Changes in perfusion in individual dogs during irradiation were observed, but no identifiable pattern of perfusion alteration was detected. Measurement of perfusion in canine nasal tumors using dynamic CT is possible and further study of this parameter as it relates to radiation response is reasonable.     

COMPARISON OF THE ACCURACY OF POSITIONING DEVICES FOR RADIATION THERAPY OF CANINE AND FELINE HEAD TUMORS

The purpose of this retrospective study was to evaluate the repositioning accuracy of different positioning devices in order to determine their applicability for potential use in conformal radiation therapy for animals. Forty-four animals with spontaneous tumors of the head were included. The animals were divided into 3 groups determined according to the positioning device used. Group 1 animals were positioned using a thermoplastic mask. Group 2 animals were positioned using a head holder. Group 3 animals were positioned using the head holder and an inflatable pillow. The time of presentation determined which position device was used. Port films of the 44 patients were reviewed retrospectively, and the repositioning precision was recorded by measurements in three orthogonal planes. Groups 2 and 3 had significantly better repositioning accuracy (P < or = 0.05) compared to Group 1. The position variation was not significantly different (P < or = 0.05) between Groups 2 and 3 in the lateral and longitudinal direction. Group 3 had a median reposition variation of 0.5 to 1.0 mm, with a standard deviation of 1.0 to 1.5 mm.