Room: AAPM ePoster Library
Purpose: To investigate treatment plan quality of head and neck cancers via tumor control and normal tissue complications probability (TCP and NTCP) objectives in the plan optimization process.
Methods: Four patients with oropharyngeal squamous cell carcinoma were used to create two sets of simultaneous integrated boost (SIB) treatment plans. One plan used conventional physical objectives for target coverage and OAR sparing while the second plan used TCP and NTCP biological objectives. The Poisson TCP model was used for the targets and the relative seriality NTCP model for the OARs. The models’ parameters that were used were taken from the literature or derived from own clinical data. TCP/NTCP values and doses of the respective plans were compared between each other and against the clinical goals.
Results: All high risk and standard risk PTV’s prescription doses satisfied D95% coverage constraint while calculated TCP values were >98% for all PTVs, in both sets of plans. V60Gy, V55.8Gy, V54Gy, V50.22Gy and conformity index (CI60Gy/CI54Gy) values were very similar for the two sets of plans. The physical plans achieved significantly lower clinical goal values for brainstem and cord (D0.1cc differences of 16.3Gy and 18.7Gy) and left cochlea and larynx (Dmean differences of 7.8Gy and 21.2Gy), with larynx mean dose the only clinical goal the radiobiological plan didn’t consistently meet. However, the NTCP values for those four OARs were 0% for both sets of plans. For the rest of the OARs (brain, parotid and submandibular glands, pharyngeal constrictors, right cochlea, esophagus and oral cavity) the differences (radiobiological minus dosimetric plans) were within -1.0Gy to 7.3Gy in the dosimetric constraints, whereas the NTCP differences ranged between -3.3% to 0.5%.
Conclusion: Radiobiological optimized planning can produce nearly clinically acceptable plans but lack the mechanism to push all OAR doses as low as can be achieved.