| BackgroundWith the increasing incidence of malignant tumors year by year,the number of lung cancer cases and deaths has risen to the first place in China.Surgical resection is the preferred treatment for early stage non-small cell lung cancer(NSCLC).The local control rate and survival rate of surgical treatment are higher than traditional non-surgery treatment.Due to the difficulty in detection and diagnosis of early stage NSCLC,about 85%of patients have already been diagnosed as staged III-IV once the obvious clinical symptoms occurred,and only part of these patients have the opportunity to receive surgery treatment.Radiation therapy is an important local treatment for patients with advanced NSCLC who cannot undergo surgery or radical resection,and patients with stage I or II lung cancer who are unable to tolerate surgery or refuse surgery due to severe medical conditions.64.3%of patients with NSCLC require radiation therapy during the treatment,and 45.4%of patients with small cell lung cancer should receive radiation therapy at the initial treatment.Unfortunately,the probability of local control rate of lung cancer is only about 50%with conventional radiotherapy.One of the major reasons for the low control rate is the geometric target miss caused by lung tumor motion during respiratory cycle.The second reason is the sensitivity of the normal tissue to radiation limits the escalation of radiation dose on tumor.Clinical dosimetry studies have shown that escalating the radiation dose on tumors was expected to improve the local control rate and to prolong the survival.The development of radiotherapy technology has brought hope for escalating the radiation dose on tumor target and reducing radiation-induced lung injury.With the development of computer technology,medical imaging,radiophysics and radiobiology,three-dimensional conformal radiotherapy(3DCRT)and intensity modulated radiotherapy(IMRT)have become the mainstream technology of radiotherapy.Lung cancer radiotherapy has entered the era of image-guided precise radiotherapy.In particular,stereotactic body radiotherapy(SBRT),which has undergone the technological innovation for several years,has developed into a highly accurate radiotherapy technology.The advantages of SBRT are that the position is well-repeated and the target dose distribution is highly conformable,which could better protect the normal lung tissue and ensure accurate irradiation through online and offline image guidance.The application of SBRT technology in solitary pulmonary tumor(SPT)is particularly noticeable.To better protect normal lung tissue,SBRT technology has gradually become the standard treatment for early NSCLC and lung oligometastases which can not tolerate surgery or refuse surgery.The treatment efficacy on early solitary lung cancer is similar to that of surgery.Simulation is the primary process in the precise radiotherapy,which directly affects the accuracy of target field and the treatment efficacy.While the lung tumor is deeply affected by the physiological activities,such as respiratory movement,the pulsation of heart and large blood vessel,etc.Tumor and normal organs have changes in position and shape during the respiratory cycle.In particular,it is the most obvious that the displacement of the lower lobe tumor in the superior-inferior(SI)direction.At present,CT simulation in SPT radiotherapy usually adopts three-dimensional computed tomography(3DCT)spiral scanning under free breathing.The collection of tumor motion information is done instantaneously during the respiratory cycle.So the 3DCT image only represents the transitory location and the shape of the tumor at a certain moment.The gross tumor volume(GTV)and clinical target volume(CTV)delineated on the 3DCT image cannot completely contain the tumor motion information during the respiratory cycle.The changes in tumor position and shape during simulation may cause errors in image acquisition,target delineation and planning design.It will result in "off target" that tumor escape from the radiation field.In order to compensate for the tumor displacement and deformation caused by respiratory movement,the radiologist defines the internal target volume(ITV)by employed a certain security margin based on CTV.Therefore,accurately determining the displacement of SPT in the three-dimensional direction is an important problem to be solved in the target delineation.Domestic and foreign scholars adopt a variety of methods to measure the tumor displacement.Such as X-ray fluoroscopy,labeling on body surface,marker implantation technology,inspiratory combined with expiratory phase techniques and four-dimensional CT(4DCT)technology.Labeling on body surface is an indirect way and it can not represent the true range of tumor movement.Marker implantation technology can precisely measure the tumor displacement through invasive and auxiliary means,which is uneasy to be accept for some patients.Inspiratory combined with expiratory phase technique has errors in the tumor position between the sessions.4DCT technology is composed of 3DCT image and the time factor.It can dynamically observe the tumor trajectory during the respiratory cycle and accurately measure the tumor displacement in three-dimensional direction.However,due to the constraints in economy and technology,4DCT technology have not been widely applied in many primary hospital and the underdeveloped regions.At present,there are two main approaches to define the security margin from CTV to ITV based on 3DCT images.The first one is the radiologists' experience combined with the literature review,but the tumor displacements was found to be different from one direction and lobe to another.The difference also exist in different patients.Therefore,the radiologists' experience cannot represent the individualized tumor motion.Another method of defining the security margin in three-dimensional direction is individualized principal based on X-ray simulator fluoroscopy.This method can reduce the probability that the margin is larger or smaller than the tumor movement.But the X-ray fluoroscopy is not applicable to the small-sized tumor and low density tumor because of its low density resolution.When the tumor overlaps with the structure of the heart and vertebral,the tumor displacement cannot be clearly displayed in one direction at least.Therefore,expanding the security margin based on clinical experience,literature review or X-ray fluoroscopy is difficult to meet the needs of accurate individualized ITV delineation in SPT radiotherapy.Therefor,the precise radiotherapy is not on the way completely.Tumor displacement and deformation caused by respiratory movement have become a bottleneck to escalate the radiation dose on tumor target and improve the radiotherapy efficacy.Since 4DCT technology has not been widely applied in the underdeveloped regions,and other imaging techniques are difficult to accurately measure SPT displacement during respiratory cycle,establishing a displacement model for SPT is beneficial to access the displacement of different lobe tumor and reduce the radiation dose on normal lung tissue.The displacement model is of great significance for escalating the radiation dose on the tumor and improve the local tumor control rate.Part 1 A Comparison of the Solitary Pulmonary Tumor Displacement Based on 4DCT and A Establishment of the Displacement ModelThe application of SBRT in radiotherapy requires more accurate target delineation.Studies have shown that there is difference in tumor displacement from one pulmonary lobe to another.The anatomic pulmonary lobe is divided into different segments.Defining the difference in SPT displacement between different segments is beneficial to improve the accuracy of target delineation.Since 4DCT has not been widely used,the establishment of SPT displacement model can provide a reference for the target delineation based on 3DCT simulation in the underdeveloped regions.ObjectiveTo compare the difference in displacements of SPTs located at different pulmonary segments based on 4DCT technology,and analyze the correlated factors of tumor displacement.Then try to establish a relevant displacement model for SPT.MethodsFrom September 2015 to May 2017,a total of 290 SPTs prepared to be treated with radiotherapy at Shandong Cancer Hospital were selected into this study.All the patients underwent both 4DCT simulation under free breathing and 3DCT spiral scan in the breath holding status.The tumor displacements in three-dimensional direction were measured based on 4DCT technology.Tumor volumes were obtained by contouring GTV on spiral 3DCT images.The diaphragm movement was measured by X-ray simulator under free breathing.The vital capacity and tidal volume were acquired by pneumatometer.The patient's information,such as gender,age,height,weight,respiratory rate,and tumor location were collected.One-Way ANOVA was used to compare the difference in displacements of tumors located at different pulmonary segments.The independent sample t-test was used to compare the tumor displacements of segment IV and V.Taking the pulmonary lobe as a unit,the correlations between the tumor displacement and patients ' gender,age,height,weight,tidal volume,vital capacity,respiratory rate,tumor volume,location,and diaphragm movement were evaluated using Partial Correlations analyses.A displacement model for SPT was subsequently established using the multivariate linear regression analyses.Results1.For SPTs located at the upper lobe,the tumor displacements in left-right(LR),SI,anterior-posterior(AP)directions were 1.03±0.95mm,2.39±1.86mm,1.1 9±0.82 mm,respectively.The tumor displacements in SI direction were significantly different among the three segments(P<0.001).The difference was also significant in AP direction(P=0.002).There was no difference in LR direction(P=0.690).There were correlations between the diaphragm movement and the tumor displacements in LR,SI,AP directions.The tumor displacements in LR and SI directions were positive correlated with the tidal volume.There were correlations between the tumor displacement in SI,AP directions and the pulmonary segments tumor located.The displacement model of tumor located at upper lobe in LR direction was established as Xupper =-0.267+0.002*TV+0.446*DM.The displacement model in SI direction was Yupper=-1.704+0.004*TV+0.725*DM+2.250*SII+1.349*S?.The model in AP direction was Zupper=0.043+0.626*DM+0.599*SII+0.519*S?.(TV:tidal volume,DM:diaphragm movement,SII:posterior segment,S?:anterior segment.the same below)2.For SPTs located at segment ? and ?,the displacements in LR,SI,AP directions were 1.51±0.97mm,4.84±3.10mm,2.14±1.51mm,respectively.In three-dimensional direction,there were no difference in the tumor displacements between segment IV and V(PLR=0.367,PSI=0.724,PAP=0.575,respectively).There were correlations between the diaphragm movement and the tumor displacements in LR,SI,AP dimensions.The tumor displacement in SI direction was also positive correlated with the tidal volume.The tumor displacement model of segment IV and V in LR direction was established as Xmiddle=0.539+0.758*DM,The displacement model in SI direction was Ymiddle=-2.316+2.707*DM+0.009*TV,and the displacement model in AP direction was Zmiddle=0.717+1.112*DM?3.For SPTs located at the lower lobe,the displacements in LR,SI,AP directions were 2.30±1.35mm,8.86±4.23mm,2.40±1.60mm,respectively.There was no difference in the tumor displacements among the five segments in LR,AP directions(PLR=0.810,PAP=0.803).The tumor displacements in SI direction was significantly different among the five segments(P=0.032).The diaphragm movement and the tidal volume were both positively correlated with the tumor displacement in LR,SI,AP directions.There was a correlation between the tumor displacement in SI direction and the segment that tumor located.There was a negative correlation between the displacement in SI direction and the respiratory rate.The displacement model of tumor located at lower lobe in LR direction was established as Xlower=-0.425+0.004*TV+0.857*DM.The displacement model in SI direction was Ylower=4.691+4.817*DM+0.005*TV-0.307*RR+3.148*SIX+2.655*SX,and the model in AP direction was Zlower=0.177+0.003*TV+0.908*DM.(RR:respiratory rate,SIX:lateral segment,SX:posterior segment)ConclusionsThe displacement models showed that the diaphragm movement,the tidal volume and the tumor location were the major factors that influenced SPT displacements in three-dimensional direction.The displacement of the lower lobe tumor in SI direction was affected by the respiratory rate at the same time.The displacements of SPTs located at different pulmonary segments belonged to the same lobe were not exactly consistent.The results suggest that the margin employed to CTV should be different according to pulmonary segment that tumor located in individualized target delineation.Part 2 A Verification of the Displacement Model for Solitary Pulmonary Tumor Based on 4DCTCurrently,4DCT scanning technology is the best way to measure tumor displacement caused by respiratory movement.The displacement model for SPT was established based on 4DCT technology in the first part of this study.It is not clear whether the prediction result of the displacement model is consistent with the displacement measured by 4DCT technology.ObjectiveTo compare the difference in SPT displacements between the methods of 4DCT and the displacement model.In order to explore the accuracy of displacement model in predicting the SPT displacement.MethodsFrom May 2017 to April 2018,108 SPTs prepared to be treated with radiotherapy at Shandong Cancer Hospital were selected as the verification sample of the displacement model.All the patients underwent 4DCT simulation under free breathing.The displacements of the tumor in three-dimensional direction were measured based on 4DCT technology.The diaphragm movement was measured by X-ray simulator under free breathing and the patient's tidal volume was measured by pneumatometer.The respiratory rate was recorded under free breathing.According to the pulmonary segment that tumor located and the information collected above,the displacement of SPTs located at different segments were calculated based on the displacement model.The Wilcoxon signed-rank test was performed to determine whether there was any significant difference in the displacements between the methods of 4DCT and the displacement model.Difference was considered significant for P<0.05.Results1.For SPTs located at the upper lobe,the tumor displacements obtained by the displacement model were 0.98±0.37mm,2.25±1.11mm,1.34±0.43mm in RL,SI,AP directions,respectively.The displacements measured by 4DCT were 1.20±0.68mm,2.62±1.60mm,1.19±0.67mm in RL,SI and AP directions,respectively.There was no statistical difference in the tumor displacements between the two approaches(PRL=0.149,PSI=0.061,PAp=0.252).2.For SPTs located at segment IV and V,the tumor displacements obtained by the displacement model were 1.75±0.49mm,4.68±1.13mm,1.98±0.47mm in RL,SI,AP directions,respectively.The displacements measured by 4DCT were 1.60±0.66mm,4.39±1.71mm,2.21±0.83mmin in RL,SI and AP directions,respectively.There was no statistical difference in the tumor displacements between the two approaches(PRL=0.073,PSI=0.229,P,p=0.104).3.For SPTs located at the lower lobe,the tumor displacements obtained by the displacement model were 2.50±1.13mm,9.899±2.95mm,2.74±0.70mm in RL,SI,AP directions,respectively.The displacements measured by 4DCT were 2.30±1.19mm,9.31±4.41mm,2.52±1.01mm in RL,SI and AP directions,respectively.There was no statistical difference in the tumor displacements between the two approaches(PRL= 0.1 57,PSI=0.194,PAP=0.176).ConclusionsThe SPT displacement obtained by the displacement model was similar to the result measured by 4DCT technology.The displacement model could accurately predict SPT displacements in three-dimensional direction under free breathing.In regions where 4DCT simulation has not yet been applied,the prediction result could provide a reference for the individualized ITV delineation on the conventional 3DCT images.Part 3 Application of the Displacement Model for Solitary Pulmonary Tumor in RadiotherapyClinical dosimetry studies have shown that there is a specific dose-efficacy relationship between the radiotherapy efficacy and the radiation dose tumor received.However,the normal lung tissue is sensitive to radiation dose,the correlation between radiation-induced lung injury and radiation dose on the normal tissue is also widely recognized.So,the escalation of radiation dose on planning target volume(PTV)is often limited by the toleration of normal lung tissue to radiation.The displacement model for SPT was established in the first part of this study.The accuracy of the displacement model in predicting SPT displacement was verified in the second part.On the basis of ensuring the radiation dose on tumor target,whether using the model for the individualized target delineation and planning design can reduce the radiation dose on normal lung tissue is an important aspect to evaluate the application value of the displacement model.ObjectiveThe aim of this study was to evaluate the difference in target volumes contoured by different methods.The dosimetry parameters in treatment planning designed by three methods were compared as well.To explore the application value of the displacement model for SPT in radiotherapy.Methods52 patients treated with radiotherapy for inoperable early stage NSCLC or pulmonary metastases at Shandong cancer hospital from April 2018 to November 201 8 were included in this retrospecitve study.All the patients sequentially underwent 3DCT and 4DCT simulation under free breathing.The methods of measuring the diaphragm movement,the respiratory rate and tidal volume were consistent with the first part of this study.PTVcon,PTVmodel and PTV4D were contoured according to the conventional experience,the displacement model and 4DCT technology,respectively.GTV was delineated on 3DCT images.GTV was expanded with an 8mm margin for adenocarcinoma,a 6mm margin for squamous cell carcimoma and 5mm margin for metastases to generate CTVcon.PTVcon.was created from CTVcon by adding the anisotropic margins for tumor displacement and setup uncertainly.The anisotropic margins were 5mm in LR and AP directions,10mm in SI direction for SPTs located at upper and middle lobe,15mm for the lower lobe SPT.IGTVmodei was created based on GTV by expanding margins,and the margins were equal to the displacements in three-dimensional direction obtained by the displacement model.IGTV4D was created by fusing each of GTVs delineated on 4DCT images of different respiratory phases.ITVmodei and ITV4D were contoured by expand the subclinical lesion areas of 8-,6-,5-mm according to the different tumor pathological type base on IGTVmodei and IGTV4D,respectively.PTVmodei and PTV4D were defined by adding 3mm setup margin based on ITVmodei and ITV4D according to our experience,respectively.In addition,the treatment PLANcon,PLANmodei and PLAN4D were designed according to PTVcon,PTVmadel and PTV4D,respectively.The Wilcoxon signed-rank test was used to compare the differences in volumes between IGTVmodei and IGTV4D,PTVs generated by different methods were compared as well.The dosimetry parameters of normal lung tissue in PLANcon,PLANmodel and PLAN4D were also compared using Wilcoxon signed-rank test.The paired sample t-test was used to compare the difference between MIcon and MImodei.The dosimetry parameters of PTV in PLANcon,PLANmodel and PLAN4D were compared using Friedman M test.Value of P<0.05 was considered significant.Results1.Comparison of IGTVmodel and IGTV4D.The volumes of IGTVmodel and IGTV4D were 29.65±16.73cm3 and 25.61±9.94cm3,respectively.There was significant difference between them(P<0.001).2.Comparison of PTVcon,PTVmodei and PTV4D.The volumes of PTVcon and PTVmodei were 59.07±21.45cm3 and 51.33±18.22cm3,respectively.There was significant difference in volume between PTVcon and PTVmodel(P<0.001).The volume of PTV4D is 43.44±17.55cm3.PTVcon and PTVmodel were both bigger than PTV4Dd(P<0.001).3.Comparison of MIcon and MImodel.The MIcon.of PTVcon to PTV4D was 0.62±0.10,and the MImodel of PTVmodel to PTV4D was 0.66±0.11.The difference was significant between MIcon and MImodel(P=0.005).4.Comparison of dosimetry parameters of lung tissue in PLANcon,PLANmodel and PLAN4D.The mean lung dose(MLD),V5Gy,V10Gy,V20Gy,V30Gy in PLANmodel were all significantly lower than those in PLANcon.There was statistical difference in the dosimetry parameters between PLANcon and PLANmodel(all P<0.001).The radiation dose on normal lung tissue in PLANmodel and PLANcon were both higher than that in PLAN4D.5.Comparison of PTV dosimetry parameters in PLANcon,PLANmodel and PLAN4D.There was no significant difference in dosimetry parameters(Dmean,D2%,D98%,CI,HI)among PTVcon and PTVmodel and PTV4D.(PDmean=0.227,PD2%=0.375,PD98%=0.2 19,PCI=0.491,PHI=0.199,respectively).ConclusionsThe volume of PTVmodel was samller than PTVcon,PTVmodel was more similar to PTV4D in volume and geometry than PTVcon.There was less normal lung tissue that exposed to radiation in PTVmodel than that in PTVcon.PLANmodel was superior to PLANcon in dosimetry parameters of lung tissue.On the basis of ensuring the radiation dose on PTV,the radiation dose on normal lung tissue in PLANmodel was lower than that in PLANcon.PTVmodel and PLANmodel based on the displacement model were beneficial to escalate the radiation dose on tumor and reduce the radiation-induced lung injury.The displacement model for SPT was feasible in target delineation and treatment planning design. |
PDF Full Text Request