扇形束CT引导头颈肿瘤根治性放疗摆位精准度分析

Analysis of set-up errors by fan-beam CT guided radiotherapy for head and neck cancers

目的 定量探讨体重变化对摆位误差的影响并建立预测模型,利用联影加速器的扇形束CT图像引导技术来分析误差来源。方法 针对在联影uRT-linac 506c加速器上接受根治性放射治疗(简称放疗)的头颈肿瘤患者,常规采用扇形束CT引导的放疗(FBCT-RT),频率如下:前3次均采用FBCT-RT,之后每周1次FBCT-RT。分析患者三维方向的摆位误差。分别在首次定位CT时秤取患者基线体重(w_(0))以及放疗过程中FBCT扫描当日的周次体重(w_(1)~w_(6))。计算放疗期间患者体重相对于w_(0)的每周体重变化比率(Δw|w_(0)-w_(1~6)/w_(0)×100%|),按患者的Δw分4组:A组(Δw<2%)、B组(2%≤Δw<4%)、C组(4%≤Δw<6%)和D组(Δw≥6%)。同一患者的周次摆位误差数据会根据当时的体重变化比率进入到不同组别。用方差分析比较4组间摆位误差;用单因素线性回归模型及ROC曲线法建立体重变化与三维摆位误差关联的预测模型并给出各体重变化段摆位误差的95%单侧参考值范围。结果 2021年9月~2022年6月共有50例头颈肿瘤患者进入研究。4组间垂直方向摆位误差(VRT)比较差异有统计学意义(P<0.05),但头脚方向(LNG)和左右方向(LAT)的摆位误差无统计学意义(P>0.05);对体重指数(BMI)≥24 kg/m^(2)人群,当4%≤Δw<6%,VRT有95%的概率控制在2.98 mm内;对BMI<24 kg/m^(2)人群,即便当4%≤Δw<6%,VRT也有95%概率控制在1.93 mm内。两大人群体重变化比率的cut-off值为3.3%时,预测本次VRT达2 mm以上的敏感度为93.3%,特异度为33.7%。对Δw≥6%,颈部脂肪厚度是影响VRT方向摆位误差的重要因素。结论头颈肿瘤放疗期间的体重下降会增大VRT摆位误差。BMI≥24 kg/m^(2)人群体重减轻后,头颈脂肪层明显变薄,更容易增大摆位误差。应当对BMI≥24 kg/m^(2)且体重减轻超过6%的患者追加锥形束CT扫描以评估重新定位时机。

Objective To quantitatively explore the influence of body weight change on set-up errors in order to establish a predicting model and to analyze the sources of set-up errors by using fan beam CT (FBCT) equipped on United Imaging uRT-linac 506c accelerator.Methods FBCT guided radiotherapy(FBCT-RT) was routinely used for head and neck cancers that were treated on uRT-linac 506c accelerator.FBCT scan was carried out with the following frequency:the first three fractions,then once a week.The setup errors were all analyzed in three dimensional direction.The baseline body weight (w_(0)) was measured at the first CT simulation scan.The weekly body weight (w_(1)~w_(6)) was measured on the same day of FBCT scan.The ratio of weekly weight change during the course of radiotherapy (Δw) was calculated according to the formula:(Δw|w_(0)-w_(1~6)/w_(0)×100%|).The patients were divided into four groups according to Δw,which were Group A(Δw<2%),Group B (2%≤Δw<4%),Group C (4%≤Δw<6%) and Group D (Δw≥6%).Analysis of variance was used to compare the set-up errors among the four groups.Univariate linear regression model and receiver operating characteristic curve was applied to establish predicting models for the association between body weight change and set-up errors in three dimensional.Unilateral 95%reference interval of set-up error was given for each group.Results Fifty head and neck cancer (HNC) patients were enrolled in the study from September 2021 to June 2022.The set-up errors in vertical direction (VRT) was statistically significant (P<0.05)among the four groups,while there was no statistical significance both in the longitudinal (LNG) and lateral(LAT) directions(P>0.05).For patients with body mass index (BMI)≥24 kg/m^(2) and 4%≤Δw<6%,error in VRT has 95%probability to be controlled within 2.98 mm.For people with BMI<24 kg/m^(2) and 4%≤Δw<6%,error in VRT still has 95%probability to be controlled within 1.93 mm.When the cut-off value of the weight change ratio was 3.3%,the sensitivity and specificity of predicting error in VRT of more than 2 mm were93.3%and 33.7%,respectively.ForΔw>6%,error in VRT varied according to the baseline soft tissue thickness of the neck.Conclusions Set-up error in VRT direction increased with weight loss during radiotherapy for HNC.The thickness of neck fat layer decreased with weight loss in BMI≥24 kg/m^(2) patients,which is the main source of the set-up errors in VRT direction.Additional FBCT scans should be performed to assess the timing of reposition in patients with BMI≥24 kg/m^(2) and≥6%loss of body weight.