Department of Radiotherapy, University Medical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The Netherlands.
Int J Radiat Oncol Biol Phys. 2010 Mar 15;76(4):1259-65. doi: 10.1016/j.ijrobp.2009.07.1685. Epub 2009 Dec 16.
The dose-response relationship of the parotid gland has been described most frequently using the Lyman-Kutcher-Burman model. However, various other normal tissue complication probability (NTCP) models exist. We evaluated in a large group of patients the value of six NTCP models that describe the parotid gland dose response 1 year after radiotherapy.
A total of 347 patients with head-and-neck tumors were included in this prospective parotid gland dose-response study. The patients were treated with either conventional radiotherapy or intensity-modulated radiotherapy. Dose-volume histograms for the parotid glands were derived from three-dimensional dose calculations using computed tomography scans. Stimulated salivary flow rates were measured before and 1 year after radiotherapy. A threshold of 25% of the pretreatment flow rate was used to define a complication. The evaluated models included the Lyman-Kutcher-Burman model, the mean dose model, the relative seriality model, the critical volume model, the parallel functional subunit model, and the dose-threshold model. The goodness of fit (GOF) was determined by the deviance and a Monte Carlo hypothesis test. Ranking of the models was based on Akaike's information criterion (AIC).
None of the models was rejected based on the evaluation of the GOF. The mean dose model was ranked as the best model based on the AIC. The TD(50) in these models was approximately 39 Gy.
The mean dose model was preferred for describing the dose-response relationship of the parotid gland.
腮腺的剂量-反应关系最常使用 Lyman-Kutcher-Burman 模型来描述。然而，还存在其他各种正常组织并发症概率（NTCP）模型。我们在一大组患者中评估了六种描述放疗后 1 年腮腺剂量反应的 NTCP 模型的价值。
共有 347 名头颈部肿瘤患者纳入本前瞻性腮腺剂量反应研究。这些患者接受常规放疗或调强放疗。使用 CT 扫描的三维剂量计算得出腮腺的剂量-体积直方图。在放疗前和放疗后 1 年测量刺激唾液流量。将预处理流量的 25%作为并发症的阈值。评估的模型包括 Lyman-Kutcher-Burman 模型、平均剂量模型、相对序列模型、临界体积模型、并行功能亚单位模型和剂量阈值模型。拟合优度（GOF）通过偏差和蒙特卡罗假设检验确定。基于 Akaike 信息准则（AIC）对模型进行排名。
没有一个模型因 GOF 的评估而被拒绝。基于 AIC，平均剂量模型被评为最佳模型。这些模型中的 TD（50）约为 39Gy。
平均剂量模型更适合描述腮腺的剂量反应关系。
