Eclipse中三变量对20例直肠癌IMRT剂量学研究

Dosimetric study of 20 cases of rectal cancer with three variables in Eclipse

目的基于Varian Eclipse三维计划系统的直肠癌调强适形放射治疗(intensity-modulated radiotherapy,IMRT)中,存在计算算法、多叶准直器(multileaf collimator, MLC)运动方式和X射线能量3种条件变量。现有文献通常采用单变量法对某一条件变量的剂量学进行了研究,并未考虑变量之间的相互影响,存在一定不足。本研究结合了单变量法和多变量法进行剂量学对比分析,旨在探索Eclipse中直肠癌IMRT的最佳变量组合模式。方法采用单变量法研究3种条件变量对放疗计划评估参数的影响,这些评估参数包括靶区的Dmean.适形指数(conformal index,CI)和均匀性指数(homogeneity index, HI),以及膀胱D50、脊髓Dniax 小肠Dm^和股骨头D5 o在此基础上进行多变量研究即将3种条件变量中的优变量和劣变量各自串联组成A、B2个对照组,并统计分析A组和E组的计划评估参数。结果(1)算法变量研究中,各项异性算法(anisotropic analytical algorithm, AAA)与笔形束卷积算法(pencil beam convolution, PBC)相比,计划靶区体积(planning target volume, PTV)的 Dmean 偏大(Z = 3. 264 , F V 0. 001), CI 偏大(Z = 3. 295, P =0. 004);危及器官(organ at risk, OAR)均偏小,P V0. 05。(2) MLC运动方式变量研究中,动态调强(sliding window,SW)与静态调强(multiple static segments, MSS)相比,PTV 的 Dmean更接近处方剂量(/= 2. 479 ,P = 0. 023),OAR 中小肠Dmax、脊髓Dmax和膀胱D50也偏小,PV0.05。(3)X射线能量变量研究中,15与6 MV相比,PTV的Dnzn偏大&=15? 153,PV0. 001),CI 偏大(r=6. O65,PVO. 001),和 HI 也更接近 1(/=一4. 807, P<0. 001);OAR 中小肠 Dm“偏小(Z=-3. 92,PV0. 001),脊髓 Dm“x 偏小(Z=-4. 225, PV0. 001),股骨头 D5 偏小(Z= 2. 962.P = 0. 008),膀胱 D50 偏大,?=10. 599,P<0. 001o (4)多变量串联组合研究中,A组与E组相比,A组PTV的Dm“n偏大(Z= 3. 17, P =0.002),CI 偏大(Z=18. 118,P<0. 001),和 HI 也更接近 1(/=一 14. 899,P<0. 001);OAR 各器官均偏小,F<0. 05。结论利用Eclispe设计直肠癌IMRT计划时,采用AAA.SW以及15 MV组成的多变量模式能最大程度的增大放疗增益比,是一种最佳变量组合模式。

OBJECTIVE Based on the Varian Eclipse three-dimensional planning system, there are three kinds of condition variables: computational algorithm, multi-leaf collimator( MLC) motion mode and X-ray energy in intensity modulated radiation therapy(IMRT). The existing literature usually uses the univariate method to study the dosimetry of a certain conditional variable,and does not consider the interaction between the variables and there are certain deficiencies. This study combined univariate and multivariate methods for dosimetric comparison analysis to explore the optimal combination of variables for rectal cancer IMRT in Eclipse. METHODS The univariate method was used to study the effects of three condition variables on the evaluation parameters of radiotherapy plans. These parameters included Dmean,conformal index(CI) and uniformity index( HI) in the target area, as well as bladder D50 ?spinal cord Dniax , and small intestine Dnwx and femoral head D5. On this basis a multivariate study was conducted in which the superior and inferior variables of the three conditional variables were respectively connected in series to form A and B control groups, and the planned evaluation parameters of the two control groups A and B were statistically analyzed. RESULTS (1) In the algorithm variable research the anisotropic algorithm( AAA) was compared with the pencil beam convolution algorithm(PBC): the Dmean of the PTV was larger(Z=- 3. 264 ,PV0. 001), and the CI was larger too(f= 3. 295 ,P = 0. 004),()ARs were both smaller( P<Z 0. 05);(2) In the MLC motion mode variable study, dynamic intensity(SW) was compared with static intensity( MSS): PTV's Dmean is closer to the prescribed dose(z = 2. 479 , P = 0. 023), small intestine Dmax, spinal cord Dmax and bladder D50 were also smaller(P<C0. 05);(3)X-ray variable study, 15 MV compared with 6 MV:PTV's Dmean was larger(z = 15. 153, P<0. 001),CI was larger(r = 6. 065,P<0. 001) and HI was closer to l(r=-4. 807,PV0. 001),and small intestine Dmax is smaller in ()ARs(Z=- 3. 92 , PV0. 001), spinal cord Dmax was small(t =- 4. 225 , P V0. 001), femoral head D5 was small(Z=- 2. 962,P=0. 008),bladder D50 was too large(r= 10. 599,PV0. 001);(4) In the multivariate series combination study,group A was compared with group B:group A had a larger Dmean(Z=-3. 17,P = 0.002) and CI(Z=1& 118, PV0. 001),and HI was also closer to l(t--14. 899 , PV0. 001),and the organs of OARs were smaller( PV0. 05). CONCLUSION When using Eclispe to design the rectal cancer IMRT plan,the multivariate model consisting of AAA,SW and 15 MV can maximize the radiotherapy therapeutic gain factor( TGF), which is the best variable combination mode.