The role of adjuvant radiation in women with lymph node-negative breast cancer who have had mastectomy remains controversial. Recently, two Surveillance, Epidemiology, and End Results (SEER) database analyses failed to find a solitary risk factor that identifies a subset of T3N0 patients who gain a survival benefit from postmastectomy radiotherapy (PMRT). These studies found that estrogen receptor-negative disease and high-grade tumors were associated with increased mortality, although individually these factors did not identify a subset benefiting from PMRT. We hypothesize that patients with multiple high risk characteristics may benefit from PMRT. In this analysis, we examined whether younger women (age < 50 years) with large primary tumors that are both hormone receptor-negative and high-grade would demonstrate a survival benefit from PMRT despite having lymph node-negative disease. The SEER database was searched for women age 20-49 diagnosed from 1990-2005 with N0, M0, invasive breast cancer and treated with mastectomy. Multivariate proportional hazards analysis assessed the association between PMRT and survival in high risk (tumor size >5.0cm, grade III-IV, and estrogen receptor (ER) or progesterone receptor (PR) negative), intermediate risk (having one or two of the high risk characteristics), and low risk (size ≤5cm, grade I-II, and ER and PR positive) groups. A total of 12,104 patients were identified, of which 301 (2.5%) were high risk, 7,220 (59.6%) were intermediate risk, and 4,583 (37.9%) were low risk. Receipt of PMRT was 49.2%, 11.4%, and 7.5% in the risk groups, respectively. In the high risk group, PMRT was associated with significantly improved cause specific survival (CSS) (hazard ratio [HR]=0.439; 95% confidence interval [CI]=0.223-0.865; p = 0.017) and overall survival (OS) (HR=0.527; CI=0.283-0.981; p = 0.043). PMRT was associated with worse CSS and OS in the intermediate and low risk groups (intermediate risk CSS: HR=1.526; CI=1.215-1.916; p <0.001; intermediate risk OS: HR=1.511; CI=1.225-1.862; p <0.001; low risk CSS: HR=2.023; CI=1.220-3.354; p = 0.006; low risk OS: HR=1.601; CI=1.023-2.507; p = 0.04). Young patients with large primary tumors who have hormone receptor-negative and high grade breast cancer may be an additional uncommon subset of lymph-node negative patients who derive a survival benefit from PMRT. The decrease in CSS and OS with PMRT in the low and intermediate risk groups suggests that the choice to give PMRT was associated with risk factors for death from breast cancer that are not reported in the SEER database (such as lymphovascular invasion, close or involved margin status, and use of systemic therapy).

Introduction: Cylindrical applicators for vaginal cuff irradiation are now commercially available for use with a 50 kV electronic brachytherapy (EB) source and are similar in size and shape to cylinders that have been long employed with iridium‐192 brachytherapy (IB). Although the insertion of the new applicator is relatively the same, the implementation from simulation to treatment can be quite different than conventional IB. This study describes the clinical implementation process of the first ever treatment of a patient using this applicator. Methods: The EB endometrial cylinder applicator (Axxent Vaginal Applicator) was supplied and manufactured by Xoft Inc., Sunnyvale, CA. Prior to treatment, a detailed analysis was performed to verify the position of dwell positions, dose distribution verification using gafchromic film and applicator geometry measurements to verify length and diameter. Finally, treatment plans were generated using PLATO software (Nucletron Corp., Columbia, MD) on CT scan data and verified using TLD data. Results: The results of the dwell position verification show that the dummy seed insert agree with the gafchromic film measurements. The distance from dwell position 1 and the surface of the apex of the applicator vary with cylinder size. The dose distribution can be verified at the point of prescription. Discussion: The results of this study show that the EB endometrial cylinder applicator can and has been implemented safely. It should be noted that this applicator has some differences from the traditional IB applicators and need to be accounted for.

Purpose: The purpose of this study was to evaluate the volume changes of target and critical organs using cone-beam CT (CB-CT) and to investigate their effects on the organ doses for prostate IMRT cases. Previously the feasibility of CB-CT based treatment planning was demonstrated by researchers (L. Lee et al., 2006; S. Yoo et al., 2008, Int. J. Rad. Onc. Biol. Phys.) Method and Materials: Ten prostate IMRT patients had daily onboard imaging and weekly kV CB-CT using a Trilogy system for treatment position verification. The latest CB-CT data were imported into Eclipse treatment planning system and used for drawing CB-CT based contours for prostate, rectum, and bladder. The volume of the organs in CB-CT were measured and compared to the organ volumes from the initial CT images. Organ contours were exported to the primary IMRT plan and the plan was rerun using CB-CT based volumes with the same field parameters. The volume and mean dose changes were measured. Results: The average interval between the first CT and latest CB-CT was 29 days. The mean volume changes of the prostate, rectum and bladder were −8.3%, 10.8%, and −5.6%, respectively between the first CT and latest CB-CT for 10 prostate IMRT cases. The dose coverage ratios of CB-CT based volumes to primary volumes were 99.9%, 111.7% and 100.6% for prostate, rectum and bladder, respectively. Conclusion: The prostate volume got decreased about 8% (n=10) after a month based on kV CB-CT and the dose coverage remained the same. However, in the average approximately 10% increase in rectal volume and dose were observed. The greater mean volume was measured for bladder with CB-CT, however, the mean dose was almost equal to the primary plan. Therefore, based on the observation, updated volumes with CB-CT can be used for improving the rectal dose as needed.

Purpose/Objective(s)IBTR! is a new web-based tool developed by investigators at Tufts University to predict ipsilateral breast recurrence risks after breast conserving surgery (BCS) with or without radiotherapy (RT). The tool generates IBTR estimates by assigning relative risk ratios to seven variables: age, tumor size, grade, lymphovascular invasion (LVI), margin status, chemotherapy, and hormone therapy use. This study evaluates the validity of IBTR! using population data from British Columbia.Materials/MethodsThe BC Cancer Agency Breast Cancer Outcomes Unit (BCOU) database identified 2071 women diagnosed between 1989 and 1999 with invasive breast cancer, pT1-3, all N stages, M0, treated with BCS. All subjects had complete information on the 7 prognostic variables in IBTR! and at least 10 years of follow-up. Anonymized data on demographics, pathologic characteristics, and adjuvant local and systemic treatment were abstracted and entered into the model to generate predicted 10-year IBTR estimates for each subject. Mean IBTR! predicted and BCOU observed breast recurrence rates were compared for the entire cohort and for each prognostic variable in the model.ResultsMean follow-up time was 13 years (range 10-17 years). RT was delivered in 1917 (93%) subjects. In the overall cohort, the difference between 10-year predicted (10.5%, standard error (SE) 0.3) and observed IBTR (6.2%, SE 0.5) was 4.3% (p < 0.001). Predicted and observed IBTR risks were within 2% in the following subgroups: age>50 years, tumors ≤1cm, grade I histology, and patients treated with hormone therapy (all p > 0.05). In these subjects, the predicted risks ranged from 5.4-6.8% compared to observed risks of 3.1-6.2%. The predicted and observed outcomes were within 2.1-5% for subgroups with tumors 1.1-2cm (p < 0.001), grade II histology (p < 0.001), LVI-negative disease (p = 0.001), and subgroups who received RT (p < 0.001). The model overestimated IBTR by 5-10% in subjects with age 41-50 years, tumors >2cm, grade III histology, LVI-positive disease, and subjects treated with chemotherapy (all p < 0.001). The model overestimated IBTR by >10% in subjects aged <41 years (predicted 23% vs. observed 12%), with positive margins (predicted 28% vs. observed 11%), and subjects treated without RT (predicted 25% vs. observed 14%), (all p < 0.001).ConclusionsIBTR! performed well in predicting breast recurrence risk to within 2% in women with favorable risk factors. The model performed less well in subjects with young age or intermediate to high risk pathologic characteristics, with a tendency toward overestimating risks. Modification of the relative risk ratios of these prognostic variables may improve the predictive reliability of IBTR! Purpose/Objective(s)IBTR! is a new web-based tool developed by investigators at Tufts University to predict ipsilateral breast recurrence risks after breast conserving surgery (BCS) with or without radiotherapy (RT). The tool generates IBTR estimates by assigning relative risk ratios to seven variables: age, tumor size, grade, lymphovascular invasion (LVI), margin status, chemotherapy, and hormone therapy use. This study evaluates the validity of IBTR! using population data from British Columbia. IBTR! is a new web-based tool developed by investigators at Tufts University to predict ipsilateral breast recurrence risks after breast conserving surgery (BCS) with or without radiotherapy (RT). The tool generates IBTR estimates by assigning relative risk ratios to seven variables: age, tumor size, grade, lymphovascular invasion (LVI), margin status, chemotherapy, and hormone therapy use. This study evaluates the validity of IBTR! using population data from British Columbia. Materials/MethodsThe BC Cancer Agency Breast Cancer Outcomes Unit (BCOU) database identified 2071 women diagnosed between 1989 and 1999 with invasive breast cancer, pT1-3, all N stages, M0, treated with BCS. All subjects had complete information on the 7 prognostic variables in IBTR! and at least 10 years of follow-up. Anonymized data on demographics, pathologic characteristics, and adjuvant local and systemic treatment were abstracted and entered into the model to generate predicted 10-year IBTR estimates for each subject. Mean IBTR! predicted and BCOU observed breast recurrence rates were compared for the entire cohort and for each prognostic variable in the model. The BC Cancer Agency Breast Cancer Outcomes Unit (BCOU) database identified 2071 women diagnosed between 1989 and 1999 with invasive breast cancer, pT1-3, all N stages, M0, treated with BCS. All subjects had complete information on the 7 prognostic variables in IBTR! and at least 10 years of follow-up. Anonymized data on demographics, pathologic characteristics, and adjuvant local and systemic treatment were abstracted and entered into the model to generate predicted 10-year IBTR estimates for each subject. Mean IBTR! predicted and BCOU observed breast recurrence rates were compared for the entire cohort and for each prognostic variable in the model. ResultsMean follow-up time was 13 years (range 10-17 years). RT was delivered in 1917 (93%) subjects. In the overall cohort, the difference between 10-year predicted (10.5%, standard error (SE) 0.3) and observed IBTR (6.2%, SE 0.5) was 4.3% (p < 0.001). Predicted and observed IBTR risks were within 2% in the following subgroups: age>50 years, tumors ≤1cm, grade I histology, and patients treated with hormone therapy (all p > 0.05). In these subjects, the predicted risks ranged from 5.4-6.8% compared to observed risks of 3.1-6.2%. The predicted and observed outcomes were within 2.1-5% for subgroups with tumors 1.1-2cm (p < 0.001), grade II histology (p < 0.001), LVI-negative disease (p = 0.001), and subgroups who received RT (p < 0.001). The model overestimated IBTR by 5-10% in subjects with age 41-50 years, tumors >2cm, grade III histology, LVI-positive disease, and subjects treated with chemotherapy (all p < 0.001). The model overestimated IBTR by >10% in subjects aged <41 years (predicted 23% vs. observed 12%), with positive margins (predicted 28% vs. observed 11%), and subjects treated without RT (predicted 25% vs. observed 14%), (all p < 0.001). Mean follow-up time was 13 years (range 10-17 years). RT was delivered in 1917 (93%) subjects. In the overall cohort, the difference between 10-year predicted (10.5%, standard error (SE) 0.3) and observed IBTR (6.2%, SE 0.5) was 4.3% (p < 0.001). Predicted and observed IBTR risks were within 2% in the following subgroups: age>50 years, tumors ≤1cm, grade I histology, and patients treated with hormone therapy (all p > 0.05). In these subjects, the predicted risks ranged from 5.4-6.8% compared to observed risks of 3.1-6.2%. The predicted and observed outcomes were within 2.1-5% for subgroups with tumors 1.1-2cm (p < 0.001), grade II histology (p < 0.001), LVI-negative disease (p = 0.001), and subgroups who received RT (p < 0.001). The model overestimated IBTR by 5-10% in subjects with age 41-50 years, tumors >2cm, grade III histology, LVI-positive disease, and subjects treated with chemotherapy (all p < 0.001). The model overestimated IBTR by >10% in subjects aged <41 years (predicted 23% vs. observed 12%), with positive margins (predicted 28% vs. observed 11%), and subjects treated without RT (predicted 25% vs. observed 14%), (all p < 0.001). ConclusionsIBTR! performed well in predicting breast recurrence risk to within 2% in women with favorable risk factors. The model performed less well in subjects with young age or intermediate to high risk pathologic characteristics, with a tendency toward overestimating risks. Modification of the relative risk ratios of these prognostic variables may improve the predictive reliability of IBTR! IBTR! performed well in predicting breast recurrence risk to within 2% in women with favorable risk factors. The model performed less well in subjects with young age or intermediate to high risk pathologic characteristics, with a tendency toward overestimating risks. Modification of the relative risk ratios of these prognostic variables may improve the predictive reliability of IBTR!

Purpose: This report lists the necessary steps to install, commission, and implement on-board imaging (OBI), and cone beam CT (CBCT) in a department with a LINAC and a competing-vendor information system. Method and Materials: The Varian Trilogy is a multi-use LINAC. This work will concentrate on the new treatment control module (4DTC). The 4DCT sequences the treatment fields, control of the MLC and the interface to both the OBI computer and the CBCT reconstructor. The record and verify system currently employed is the IMPAC MOSAIQ 4DT sequencer module. This unit receives the treatment parameters in DICOMRT format from the treatment planning system (TPS) and exports the information to the LINAC's 4DTC. The sequencer module then waits for the treatment to be completed and receives the final treatment data including images taken by the OBI. Results: The commissioning and implementation of the LINAC was performed in stages. The first stage was to collect all the necessary data to allow our 3 TPSs to properly calculate dose and send the correct parameters to the LINAC for treatment. The second stage included confirming that the OBI performed diagnostic quality x-rays and that remote couch movements were accurately employed. The third and final stage was to verify the CBCT images could be transferred from the TPSs, registered, and prepared by MOSAIQ. After imaging, couch shifts are made to align the patient to their simulation CT and a screen capture is used to document the shifts. Since the CBCT data resides on the 4DTC, there is currently no mechanism to export it back into the MOSAIQ system. Conclusion: The Varian Trilogy LINAC with OBI and CBCT was successfully implemented using IMPAC MOSAIQ information system in a semi-chart-less and filmless department.

Occasionally, post mastectomy breast patients are irradiated with a tissue expander (TE) in place before their permanent implant exchange. Certain TEs contain metallic ports embedded with rare earth magnets (REM). Previous investigations into the dosimetry of these TEs reported on the effects of attenuation (u) associated with the metallic port. According to these studies, certain treatment planning systems were able to account for the dose in-homogeneity given a corrected CT HU/density curve. The purpose of this study was to investigate whether the changes in dosimetry are due to attenuation (u) that can be accounted for by most 3D treatment planning systems or the associated magnetic fields (Bo) of the REM that has yet to be modeled. Film dosimetry with single 6-MV, 18-MV photon beams and 6-MEV electron beams was performed in a solid water phantom using Gafchromic EBT film. A Magna-Site disc was placed in two opposite pole orientations. Additionally, higher field strength magnets were used perpendicular and outside the radiation beam. Measurements were compared to open fields without magnets using RIT 113 dosimetry system. The results of the phantom measurements indicate that reversing the poles of the REM had a significant affect on the isodose distributions for 6-MV photon beams. For 6-MV photons, the dose between positive and negative magnetic pole orientations showed a shift of isodose lines between 3- 5 mm. For 6-MEV electrons, the changes in pole orientation shifted the isodose distribution over 5 mm in the lateral and axial direction. The results of the higher strength magnets perpendicular to the radiation beam show significant expansion of the penumbra for all energies. The results of this study indicate that the measured dose discrepancies previously reported were not only due to attenuation but the magnetic fields associated with REMs. To accurately report dose distributions in patients with TEs containing REMs, a new model must be developed to account for the magnetic fields. Until then, patients with TEs containing REMs cannot be accurately planned on current treatment planning systems.

Extended abstract of a paper presented at Microscopy and Microanalysis 2006 in Chicago, Illinois, USA, July 30 – August 3, 2006

Purpose: The Pin‐PointTM is a technology for promoting accurate access to a designate target in the CT room, especially during biopsy procedures. This study is to confirm its capability by an independent methodology for clinical implementation. Method and Materials: The Pin‐Point system (FigA) is based on CT imaging with an invisible frame system and instant 3‐D image reconstruction in the laser (needle insertion) plane as illustrated in FigB. Two (yam and daikon) phantoms were used for the studies. The center portion of the phantoms was cut into ten 4mm thick slices. Simulated tumors were created on each slice using barium paste. Organs at risk (OAR) were carved with various shape cavities. BBs were attached as reference points (FigC). Target searching and their correlation to the neighboring OARs were explored. An optimal “reference point R” was defined to serve as navigation for the laser‐guided needle for start location, direction, and the needle depth to reach the targets and avoid OARs. Results: FigD demonstrates the needle reaching the tumor as navigated by the Pin‐Point system. We oriented the guiding laser beam and pushed its path depth from point R to the tumor edge. Next the phantom slices were studied to determine needle positioning relative to the target and OARs, and compared with the foregoing CT image. The agreement is shown in FigD. FigE illustrates the optimal needle trajectory to ‘tumor A’, by‐passing the OARs. FigF illustrates pre‐warning of OARs. FigG displays the possibility of excessive needle penetration. Conclusion: Pin‐Point system was accurate for correct tumor access without jeopardizing the nearby OARs. Our study proves valuable for the search of designated tumors and avoiding critical structures. In addition, its feasibility for further application to linac based stereotactic radiosurgery as frameless setup should be studied.