PV QA 1 - Poster Viewing Q&A 1
|Analysis of OAR constraint violations for non-adaptive (NA) and adaptive (A) plans|
|Both NA and A meet OAR constraints, n (%)||Only NA meets constraints, n (%)||Only A meets constraints, n (%)||Neither NA nor A meets constraints, n (%)||p-value|
|Dmax < 38 Gy||28 (56)||4 (8)||15 (30)||3 (6)||0.019|
|V33 < 1 cc||32 (64)||4 (8)||12 (24)||2 (4)||0.077|
|Dmax < 38 Gy||40 (80)||3 (6)||4 (8)||3 (6)||1|
|V33 < 1 cc||37 (74)||3 (6)||5 (10)||5 (10)||0.727|
|Dmax < 38 Gy||44 (88)||5 (10)||1 (2)||0 (0)||0.219|
|All OAR constraints met||14 (28)||9 (18)||15 (30)||12 (24)||0.221|
Purpose/Objective(s): Magnetic Resonance Image guided (MRI-g) radiation therapy provides better soft tissue visualization than cone beam computed tomography (CBCT) and allows for daily online adaptive radiotherapy (OART) in the treatment of locally advanced pancreatic cancer (LAPC). Using MRI-g stereotactic body radiation therapy (SBRT) to treat LAPC, we hypothesized that stomach distention would correlate with dose increases to organs-at-risk (OARs) when online adaptation was not utilized. Meanwhile, we hypothesized that OART would protect against dose constraint violations to OARs.
Materials/Methods: We identified ten patients who received pancreas SBRT to a dose of 33-40 Gy in 5 fractions. The patients were instructed to avoid oral intake other than sips of water for at least 3 hours prior to treatment. Dose was prescribed to achieve 90% coverage of planning tumor volume at 100% isodose (PTV100). After each fraction’s setup MRI, the target position was aligned by 3-dimensional shifts, the normal anatomy was re-contoured, and the original radiotherapy plan was recalculated to make a non-adaptive plan. A re-optimized (adaptive) plan was then generated for each fraction and renormalized to 90% coverage of PTV100. Target and OAR doses between non-adaptive and adaptive plans were compared by paired t-test and McNemar tests to assess the dosimetric impact of daily adaptation. A linear random-coefficient model correlated changes in stomach volumes to changes in OAR objective doses.
Results: Sixty-eight percent of non-adaptive fractions met both point maxima and V33 dose constraints to the duodenum compared with 87% of adaptive fractions (p=0.0013). The original stomach volume (OSV) mean was 286cc, with a range from 126cc-856cc. The mean interfraction change from the OSV was +147cc, with a range from -404cc to +1271cc. The mean percentage change from the OSV was 53%, with a range from -47% to 334%. For non-adaptive plans, interfraction increase in stomach volume correlated with higher duodenum V33 (p=0.021), duodenum maximum dose (p=0.105), stomach V33 (p=0.004), and stomach maximum dose (p=0.009). No correlation was observed between stomach volume and dose to visceral OARs for the adaptive plans.
Conclusion: In SBRT for LAPC, interfraction variations in stomach volume occur despite precautions and correlate with increased dose to the duodenum in non-adaptive plans, leading to significant dose constraint violations. OART limits dose to the duodenum and contributes to safe SBRT dose escalation.
The asset you are trying to access is locked. Please enter your access key to unlock.