patient was 45 Gy in 1.8 Gy daily fractions using 15 MV photon beams, prescribed to encompass 95% of the PTV. All patients underwent daily image-guided setup using a Varian Trilogy on-board imager. Daily shifts obtained after matching to pelvic bony anatomy were used to retrospectively analyze the effect of the setup shifts on BM dose-volume parameters. Dose calculations were performed for each fraction by translating the isocenter in the direction of the measured shifts (negating their effect) and recalculat- ing the doses using the originally planned monitor units. The doses from each fraction were summed to obtain the cumulative dose delivered during treatment. The percentage volumes of BM receiving doses greater than 10 (V 10 ), 20 (V 20 ), 30 (V 30 ) and 40 (V 40 ) Gy were calculated and compared with the original planned values. Results: The average shifts observed were 0.03 cm (range, -0.7 to 0.8), -0.14 cm (range, -0.9 to 0.5) and 0.08 cm (range, -0.5 to 0.5) in the AP, LR and SI directions respectively. The average planned V 10 ,V 20 ,V 30 and V 40 was 88.5% (range, 63.0 to 94.0), 71.4% (range, 52.6 to 81.2), 42.5% (range, 38.2 to 49) and 16.2% (range, 12.5 - 18.6) respectively. The average V 10 ,V 20 ,V 30 and V 40 without image guidance would have been 87.4% (range, 63.0 to 96.0), 70.1% (range, 52.3 to 82.2), 41.0% (range, 30.5 to 53.4) and 14.25% (range, 12.5 - 19.6) respectively. Conclusions: Daily patient setup using image guidance did not improve the sparing of pelvic BM in the cervical cancer patients studied using our protocol with a 0.7 cm PTV expansion. Future studies will investigate the impact of image guidance on BM dose- volume parameters using smaller expansions. Author Disclosure: J.H. Lewis, None; N. Tyagi, None; C.M. Yashar, None; D. Vo, None; L.K. Mell, None; S.B. Jiang, None; A.J. Mundt, None. 2924 The use of Cone Beam CT-based Three-dimensional Planning in Intracavitary Brachytherapy for Cervical Cancer H. Al-Halabi 1 , L. Portelance 1 , M. Duclos 1 , B. Reniers 1 , M. Bloemers 1 , B. Bahoric 2 , T. Niazi 2 , L. Souhami 1 1 McGill University Health Center, Montreal, QC, Canada, 2 Jewish General Hospital, Montreal, QC, Canada Purpose/Objective(s): To evaluate the feasibility of using cone beam CT (CBCT) based three-dimensional (3-D) treatment plan- ning to calculate dose-volume histograms (DVHs) of the bladder and rectum, and to compare these values with International Com- mission on Radiation Units and Measurements (ICRU) reference points in intracavitary brachytherapy (ICBT) for cervical cancer patients. Materials/Methods: Between July and December of 2007, 10 cervical cancer patients underwent an average of 3 ICBT insertions. Using CT compatible applicators intraoperative CBCT images were obtained in 21 of these procedures. A rectal tube was inserted and bladder contrast was injected before scanning to facilitate contouring the rectum and bladder respectively. All patients under- went intraoperative orthogonal film and treatments were prescribed using standard two-dimensional planning and dosimetry. ICRU reference point doses to the bladder (B ICRU ) and rectum (R ICRU ) were calculated for each treatment. Using the Varian Eclipse treat- ment planning system, contours of the bladder and rectum were delineated onto the CBCT images. The rectum was contoured from the sigmoid flexure to the inferior border of the ischial tuberosities. The whole bladder was contoured. Using the PLATO planning system, DVHs for the bladder and rectum were obtained for each treatment by reconstructing the same dose distribution planned conventionally. For dose comparison, the minimum dose in 2.0 cm 3 of bladder (B D2V ) and rectum (R D2V ) receiving the highest dose were determined from DVHs. B D2V and R D2V were compared with ICRU point estimates of bladder and rectal doses. Results: A total of 21 CBCT based plans were obtained. CBCT image quality improved with experience in performing the scan. The use of bladder contrast and rectal tube allowed more accurate, reproducible delineation of these organ contours. Performing intraoperative scans as opposed to moving the patient to a CT scanner limits patient movement, this minimizes post- insertion ap- plicator motion. The mean B ICRU dose (348 cGy) was significantly lower than the mean B D2V (570 cGy), (p \ 0.001). The mean R ICRU dose was 406 cGy, whereas the mean R D2V was 474 cGy (p = 0.044). These results are comparable to data derived from studies comparing CT-based 3-D planning in ICBT to conventional planning with orthogonal films. Conclusions: CBCT based 3-D planning can be utilized in image guided brachytherapy for cervical cancer and it is a convenient alternative to CT based planning with the advantage of minimizing applicator motion. Our data suggests that the ICRU bladder and rectal point doses significantly underestimate the maximum dose to both organs calculated using CBCT based 3-D planning. Cor- relation with late effects will further evaluate the role of CBCT based 3-D dosimetry in ICBT planning. Author Disclosure: H. Al-Halabi, None; L. Portelance, None; M. Duclos, None; B. Reniers, None; M. Bloemers, None; B. Bahoric, None; T. Niazi, None; L. Souhami, None. 2925 Interfractional Overlap of Bladder and Rectal Hotspots during 3D Image-based HDR Brachytherapy of the Cervix F. S. Vali, W. Hall, M. Gao, R. Hong, A. Chi, D. Milling, J. Sinacore, K. Albuquerque Loyola University Medical Center, Maywood, IL Purpose/Objective(s): The GEC-ESTRO working group prioritizes the 0.1cc hotspot as an indicator of toxicity to the bladder and rectum during 3D cervical cancer brachytherapy. Using vector math, we previously demonstrated that these hotspots may move considerably. Since it is possible that hotspots within an organ may retrace their positions as they shift from fraction to fraction, we aim to study the percent of those fractions in which there is no overlap at all between the 0.1cc hotspots during the course of treatment. Materials/Methods: Fifty 3D CT based plans from 10 cervical cancer patients treated with HDR brachytherapy were studied. For each of the 50 CT scans, the rectal and bladder volumes were contoured, dose-volume histograms (DVH) were generated and the 0.1cc hotspots were identified. To study the motion of the hotspots we assessed the movement of their respective geometric centers from fraction to fraction. For each fraction, the hotspot location in 3D space was represented by a vector between the center of the hotspot and a fixed reproducible point within the organ. The 3D shift of this vector from fraction to fraction was recorded. We then identified the number of fractions during which the displacement of the hotspot from each of its previous or subsequent locations in Proceedings of the 50th Annual ASTRO Meeting S583