Purpose/Objective(s)Increasingly, MR-based planning is used in brachytherapy (BT) for cervix cancer, allowing for optimal therapeutic ratio. Daily MRI during pulsed-dose-rate BT (PDR BT) was used to evaluate changes in high risk CTV (HRCTV), organ at risk volumes and dosimetry.Materials/Methods50 patients (pts) were accrued to an MRI-guided PDR BT protocol. An MRI was done after insertion of an intrauterine tandem, followed by MRIs on Day 2 (d2) and Day 3 (d3). A BT plan was optimized along GEC ESTRO guidelines on Day 1 (d1). D2 and d3 MRIs were compared to d1 images and plan to determine volumetric and dosimetric differences. Deformable soft tissue modeling with dose accumulation over time was used to validate the dosimetric changes in a subset of 8 pts with 3 MRIs.ResultsAll pts had volumetric and dosimetric data for d1, 43 for d2 and 27 for d3. Mean d1 HRCTV was 30.4cm3 (10.6cm3 - 74.6cm3). HRCTV increased over time (p<0.0001): mean volume 33.5 cm3 on d2 and 34.8cm3 on d3. Mean d1 rectal volume was 54.6 cm3 with an increase in d2 mean volume to 67.6cm3 (p = 0.002). HRCTV D90 decreased over time (p<0.0001). Optimal dosimetry (D90≥87Gy) was achieved in 92%, 67% and 52% of all cases on d1, d2 and d3 respectively; with mean D90 on d1, d2, and d3 of 96, 89, and 88 Gy. Four pts with suboptimal d1 dosimetry had poor dosimetry throughout treatment. Of the 24 pts with optimal d1 dosimetry and d2, d3 scans, 10 had acceptable dosimetry throughout and 14 were unacceptable on at least one of the subsequent days. However, the average D90 over 3 days was ≥87Gy in 92% of pts. The only factor predictive of a reduction in HRCTV D90 during treatment was initial target volume (p = 0.009); the smaller the d1 HRCTV, the larger the decrease in HRCTV dose on d2 or d3. There was a trend of increasing rectal 2cc maximum dose (D2cc) over time (p = 0.06) with mean D2cc on d1, d2, d3 of 70, 73 and 70 Gy. High rectal D2cc on d1 predicted for high doses on subsequent days (p = 0.01) but after adjusting for baseline rectal dosimetry, a change in rectal volume did not predict for poor rectal dosimetry (p = 0.11). Bladder volumes and dosimetry did not change significantly over time. Deformable registration confirmed cumulative doses for HRCTV D90 were lower than predicted by the d1 plan alone in 6 of 8 pts. Cumulative rectal doses were higher than as planned on d1 for 4 of 8 pts.ConclusionsDaily MRI shows significant HRCTV and rectal volumetric changes over time, likely due to edema. Despite this, the overall D90 in 92% of pts with d1 - 3 MRIs and good d1 dosimetry met GEC ESTRO criteria. Longer pt follow-up is required to understand the clinical significance of these volume changes and the dosimetric consequences. Pts with small volume HRCTV and borderline dosimetry, particularly with regard to rectal dose, may benefit from daily MRI planning. Purpose/Objective(s)Increasingly, MR-based planning is used in brachytherapy (BT) for cervix cancer, allowing for optimal therapeutic ratio. Daily MRI during pulsed-dose-rate BT (PDR BT) was used to evaluate changes in high risk CTV (HRCTV), organ at risk volumes and dosimetry. Increasingly, MR-based planning is used in brachytherapy (BT) for cervix cancer, allowing for optimal therapeutic ratio. Daily MRI during pulsed-dose-rate BT (PDR BT) was used to evaluate changes in high risk CTV (HRCTV), organ at risk volumes and dosimetry. Materials/Methods50 patients (pts) were accrued to an MRI-guided PDR BT protocol. An MRI was done after insertion of an intrauterine tandem, followed by MRIs on Day 2 (d2) and Day 3 (d3). A BT plan was optimized along GEC ESTRO guidelines on Day 1 (d1). D2 and d3 MRIs were compared to d1 images and plan to determine volumetric and dosimetric differences. Deformable soft tissue modeling with dose accumulation over time was used to validate the dosimetric changes in a subset of 8 pts with 3 MRIs. 50 patients (pts) were accrued to an MRI-guided PDR BT protocol. An MRI was done after insertion of an intrauterine tandem, followed by MRIs on Day 2 (d2) and Day 3 (d3). A BT plan was optimized along GEC ESTRO guidelines on Day 1 (d1). D2 and d3 MRIs were compared to d1 images and plan to determine volumetric and dosimetric differences. Deformable soft tissue modeling with dose accumulation over time was used to validate the dosimetric changes in a subset of 8 pts with 3 MRIs. ResultsAll pts had volumetric and dosimetric data for d1, 43 for d2 and 27 for d3. Mean d1 HRCTV was 30.4cm3 (10.6cm3 - 74.6cm3). HRCTV increased over time (p<0.0001): mean volume 33.5 cm3 on d2 and 34.8cm3 on d3. Mean d1 rectal volume was 54.6 cm3 with an increase in d2 mean volume to 67.6cm3 (p = 0.002). HRCTV D90 decreased over time (p<0.0001). Optimal dosimetry (D90≥87Gy) was achieved in 92%, 67% and 52% of all cases on d1, d2 and d3 respectively; with mean D90 on d1, d2, and d3 of 96, 89, and 88 Gy. Four pts with suboptimal d1 dosimetry had poor dosimetry throughout treatment. Of the 24 pts with optimal d1 dosimetry and d2, d3 scans, 10 had acceptable dosimetry throughout and 14 were unacceptable on at least one of the subsequent days. However, the average D90 over 3 days was ≥87Gy in 92% of pts. The only factor predictive of a reduction in HRCTV D90 during treatment was initial target volume (p = 0.009); the smaller the d1 HRCTV, the larger the decrease in HRCTV dose on d2 or d3. There was a trend of increasing rectal 2cc maximum dose (D2cc) over time (p = 0.06) with mean D2cc on d1, d2, d3 of 70, 73 and 70 Gy. High rectal D2cc on d1 predicted for high doses on subsequent days (p = 0.01) but after adjusting for baseline rectal dosimetry, a change in rectal volume did not predict for poor rectal dosimetry (p = 0.11). Bladder volumes and dosimetry did not change significantly over time. Deformable registration confirmed cumulative doses for HRCTV D90 were lower than predicted by the d1 plan alone in 6 of 8 pts. Cumulative rectal doses were higher than as planned on d1 for 4 of 8 pts. All pts had volumetric and dosimetric data for d1, 43 for d2 and 27 for d3. Mean d1 HRCTV was 30.4cm3 (10.6cm3 - 74.6cm3). HRCTV increased over time (p<0.0001): mean volume 33.5 cm3 on d2 and 34.8cm3 on d3. Mean d1 rectal volume was 54.6 cm3 with an increase in d2 mean volume to 67.6cm3 (p = 0.002). HRCTV D90 decreased over time (p<0.0001). Optimal dosimetry (D90≥87Gy) was achieved in 92%, 67% and 52% of all cases on d1, d2 and d3 respectively; with mean D90 on d1, d2, and d3 of 96, 89, and 88 Gy. Four pts with suboptimal d1 dosimetry had poor dosimetry throughout treatment. Of the 24 pts with optimal d1 dosimetry and d2, d3 scans, 10 had acceptable dosimetry throughout and 14 were unacceptable on at least one of the subsequent days. However, the average D90 over 3 days was ≥87Gy in 92% of pts. The only factor predictive of a reduction in HRCTV D90 during treatment was initial target volume (p = 0.009); the smaller the d1 HRCTV, the larger the decrease in HRCTV dose on d2 or d3. There was a trend of increasing rectal 2cc maximum dose (D2cc) over time (p = 0.06) with mean D2cc on d1, d2, d3 of 70, 73 and 70 Gy. High rectal D2cc on d1 predicted for high doses on subsequent days (p = 0.01) but after adjusting for baseline rectal dosimetry, a change in rectal volume did not predict for poor rectal dosimetry (p = 0.11). Bladder volumes and dosimetry did not change significantly over time. Deformable registration confirmed cumulative doses for HRCTV D90 were lower than predicted by the d1 plan alone in 6 of 8 pts. Cumulative rectal doses were higher than as planned on d1 for 4 of 8 pts. ConclusionsDaily MRI shows significant HRCTV and rectal volumetric changes over time, likely due to edema. Despite this, the overall D90 in 92% of pts with d1 - 3 MRIs and good d1 dosimetry met GEC ESTRO criteria. Longer pt follow-up is required to understand the clinical significance of these volume changes and the dosimetric consequences. Pts with small volume HRCTV and borderline dosimetry, particularly with regard to rectal dose, may benefit from daily MRI planning. Daily MRI shows significant HRCTV and rectal volumetric changes over time, likely due to edema. Despite this, the overall D90 in 92% of pts with d1 - 3 MRIs and good d1 dosimetry met GEC ESTRO criteria. Longer pt follow-up is required to understand the clinical significance of these volume changes and the dosimetric consequences. Pts with small volume HRCTV and borderline dosimetry, particularly with regard to rectal dose, may benefit from daily MRI planning.