Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Determination of patient-specific internal gross tumor volumes for lung cancer using four-dimensional computed tomography

Determination of patient-specific internal gross tumor volumes for lung cancer using... Background: To determine the optimal approach to delineating patient-specific internal gross target volumes (IGTV) from four-dimensional (4-D) computed tomography (CT) image data sets used in the planning of radiation treatment for lung cancers. Methods: We analyzed 4D-CT image data sets of 27 consecutive patients with non-small-cell lung cancer (stage I: 17, stage III: 10). The IGTV, defined to be the envelope of respiratory motion of the gross tumor volume in each 4D-CT data set was delineated manually using four techniques: (1) combining the gross tumor volume (GTV) contours from ten respiratory phases (IGTV ); (2) AllPhases combining the GTV contours from two extreme respiratory phases (0% and 50%) (IGTV ); (3) 2Phases defining the GTV contour using the maximum intensity projection (MIP) (IGTV ); and (4) defining MIP the GTV contour using the MIP with modification based on visual verification of contours in individual respiratory phase (IGTV ). Using the IGTV as the optimum IGTV, we MIP-Modified AllPhases compared volumes, matching indices, and extent of target missing using the IGTVs based on the other three approaches. Results: The IGTV and IGTV were significantly smaller than the IGTV (p < 0.006 for MIP 2Phases AllPhases stage I and p < 0.002 for stage III). However, the values of the IGTV were close to those MIP-Modified determined from IGTV (p = 0.08). IGTV also matched the best with IGTV . AllPhases MIP-Modified AllPhases Conclusion: IGTV and IGTV underestimate IGTVs. IGTV is recommended to MIP 2Phases MIP-Modified improve IGTV delineation in lung cancer. Background been considered one of the main reasons for local failure Lung cancer remains the leading cause of cancer-related [2]. Researchers have reported that ~40% of lung tumors mortality. Conventional photon radiotherapy for lung move > 5 mm and that 10–12% move > 1 cm [3,4]. Sev- cancer is associated with about 50% local tumor control eral strategies have recently been developed to address the [1]. Missing the target as a result of tumor motion has issue of tumor motion and improve local control [2]. For Page 1 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 example, the development of image-guided radiotherapy touring in only one dataset. In these instances, post- (IGRT) has allowed for more accurate tumor targeting, so processing tools, such as the maximum intensity projec- it is rapidly replacing conventional radiotherapy for lung tion (MIP), have been shown to improve radiotherapy cancer [2]. In order to account for tumor motion, the planning efficiency [7]. The MIP of a 4D-CT data set International Commission on Radiation Units and Meas- reduces the multiple 3-D CT data available from a 4-D CT urements (ICRU) report 62 introduced the concept of an data set into a single 3-D CT data set, where each voxel in internal target volume (ITV), defined as the clinical target the MIP represents the maximum intensity encountered volume (CTV) plus an additional margin to account for by corresponding voxels in all individual 3-D phase image geometric uncertainties due to internal variations in sets of the 4-D CT data set. The IGTV is then determined tumor position, size, and shape. Using current imaging based on the GTV delineation on the single 3-D CT data techniques, the CTV cannot be visualized. Consequently, set. Alternatively, some cancer centers have used breath- generation of the ITV requires delineation of the gross hold spiral CT imaging to acquire images at the two tumor volume (GTV) on each of the phases that constitute extremes of the respiratory cycle [2,7]; contouring the GTV the four-dimensional (4-D) computed tomography (CT) at these extremes (the end-expiration and the end-inspira- image data set, followed by expansion of each GTV to tion phases) and then combining these two 3-D volumes account for microscopic disease. The ITV is then deter- yields the IGTV. A limited number of studies have ana- mined to be the envelope of motion of the CTV. In order lyzed the accuracy of the MIP and two-phase IGTV delin- to make the determination of the ITV more efficient, we eation techniques relative to full ten-phase method for have proposed the concept of the internal gross tumor determining IGTV [8-11]. volume (IGTV), which explicitly accounts for internal var- iations in tumor position, size, and shape but can be The aim of this study, therefore, was to evaluate the accu- derived directly from imaging studies [2]. The ITV is then racy of 4-D CT MIP-based IGTV delineation and two- determined to be the IGTV plus a margin that accounts for phase-based IGTV delineation compared to ten-phase microscopic disease. IGTV delineation as a reference. We also examined the accuracy of the MIP-based IGTV delineation after applying Traditionally, the margin necessary to account for internal a modification through visual verification of GTV cover- motion of tumors in the thorax has been determined age in individual respiratory phases. using an isotropic expansion determined by population- based estimates of respiratory motion. However, because Methods breathing characteristics vary greatly among individual Data acquisition patients, such population-based estimates may overesti- As a retrospective review of radiation treatment planning, mate or underestimate the margin needed for a given this study was included under an Institutional Review patient. Moreover, respiratory-induced tumor motion is Board-approved retrospective chart review protocol. We known not to be anisotropic; typical tumor paths are studied 27 consecutive patients with non-small-lung can- those of elongated and possible curved ellipses. The cer (NSCLC) who underwent 4-D CT simulation for treat- advent of the multislice helical CT scanner combined with ment planning and received definitive radiotherapy at our the establishment of temporal correlation between respi- institution between 2005 and 2006. Of these 27 patients, ratory motion and the CT acquisition process have 17 had stage I disease and received stereotactic body radi- allowed tumor size, shape, and position to be observed at otherapy (SBRT), and 10 had stage III disease and received multiple times during a patient's respiratory cycle [5,6]. intensity-modulated radiotherapy (IMRT). 4-D CT image The resultant CT data set, called the 4-D CT or respiration- data sets each consisting of 10 respiratory phases, were correlated CT data set, provides patient-specific informa- acquired on a multislice CT scanner (Discovery ST, GE tion about tumor position, shape, and size at different Medical Systems, Madison, WI) by sorting CT images phases of the respiratory cycle. based on the phase of an external respiratory monitor (Real-time Position Management System; Varian Medical Although using 4-D CT data provides a reliable estimate Systems, Inc., Palo Alto, CA) [12]. MIPs of the 4D-CT data of the extent of tumor motion due to respiration in three sets were then generated from the individual phase images dimensions, its clinical implementation poses some chal- as described elsewhere [5,6]. lenges. Ideally, the IGTV should be determined by con- touring the GTV on each of the ten phase image sets. The Patient-specific IGTV determination combination of these individual three dimensional (3-D) We determined patient-specific IGTVs using the demon- volumes into a single 3-D volume represents the IGTV, strable extent of tumor motion shown in the 4-D CT which accounts for respiratory motion. However, con- images. We used four approaches to determine these touring the tumor volume on ten different data sets for IGTVs: (1) contouring the GTV on each of the ten respira- each patient increases the workload compared with con- tory phases of the 4D-CT data set and combining these Page 2 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 GTVs to produce IGTV ; (2) contouring the GTV on sistency in contouring, all GTV contours in each respira- AllPhases the MIP of the 4-D CT data set to produce IGTV ; (3) tory phase of the 4-D CT and MIP data sets were drawn by MIP contouring the GTV on the extreme respiratory phases a single radiation oncologist (ME) and verified by another (0% phase = peak inhalation, 50% phase = peak exhala- radiation oncologist (JYC). We used a lung window on the tion) and combining these GTVs to produce IGTV ; CT data set to contour the primary tumor and a mediasti- 2Phases and (4) contouring the GTV on the MIP of the 4-D CT data num window to contour any involved lymph nodes. set and then modifying these contours using visual verifi- Diagnostic CT of chest with intravenous contrast and PET/ cation of coverage in each phase of the 4-D CT data set to CT were used to guide our involved lymph nodes contour- produce IGTV . Visual verification of coverage in ing as described by our previous publication (2). A total MIP-Modified each phase was achieved by overlaying the MIP based GTV of 324 GTVs were delineated with 12 GTVs delineated for contour onto each phase of the 4-D CT data set. Thus, each patient (GTV in each of 10 respiratory phases, each of these 3D volumes (IGTV , IGTV , IGTV , and IGTV ). For stage III disease, AllPhases MIP MIP MIP-Modified IGTV , and IGTV ) represented the demon- involved hilar or mediastinal lymph nodes were con- 2Phases MIP-Modified strable respiratory tumor motion volumes, or IGTVs. Fig- toured and analyzed independently. ures 1 and 2 show the results obtained using these different approaches in the determination of IGTV for cases of stage I and stage III disease, respectively. For con- 4- Figure 1 Delineation of IGTV for sta D CT data set ge I lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) IGTV of a MIP MIP-Modified 2Phases AllPhases Delineation of IGTV for stage I lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) MIP MIP-Modified 2Phases IGTV of a 4-D CT data set. MIP-based contours, as shown in panels (a) and (b), are as they appear on the MIP data AllPhases set. Phase-based contours, as shown in panels (c) and (d), are registered to the peak exhalation phase of the 4-D CT data set. Page 3 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 D 4- Figure 2 e D CT lineati da on o ta se f IGTV f t or stage III lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) IGTV of a MIP MIP-Modified 2Phases AllPhases Delineation of IGTV for stage III lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) MIP MIP-Modified 2Phases IGTV of a 4-D CT data set. MIP-based contours, as shown in panels (a) and (b), are as they appear on the MIP data AllPhases set. Phase-based contours, as shown in panels (c) and (d), are registered to the peak exhalation phase of the 4D-CT data set. Data analysis the minimum value is 0 if the volumes are completely We evaluated the IGTVs determined using each of the non-overlapping. three contouring approaches against an all phases IGTV determined by contouring all ten respiratory phases of the Volume difference calculation 4-D CT data set (IGTV ). Specifically, we compared While the matching index is a good measure of how well AllPhases the following metrics for each 3D volume: matching the shape of any two volumes match each other, it cannot index, total GTV volume and under or over-estimated vol- discriminate between overestimation and underestima- ume. tion. To gain better insight into any over/underestimation of the IGTV, we computed the differences in IGTV Matching index calculation between the all phases volume (IGTV ) and the AllPhases The matching index (MI) of any two 3D volumes A and B three test volumes (IGTV , IGTV , and IGTV MIP 2Phases MIP-Mod- is defined as the ratio of the intersection of A with B to the ). For each pair of volumes, we computed the underes- ified union of A and B, that is, timation and overestimation volumes (V and V ) Under Over using the following equations: AB ∩ MI = . VV = \V AB ∪ Under AllPhases Test VV = \, V Over Test AllPhases As can be deduced from this equation, the maximum value of the MI is 1 if the two volumes are identical, and Page 4 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 where V is the volume in ten respiratory phases, V sistently smaller than the IGTV (mean ± SD: 16.60 AllPhases test AllPhases is the test volume, and "\" denotes the set difference. The ± 17.05 cm ), whereas the IGTV (mean ± SD: MIP-Modified ) were similar to the reference IGTV. A underestimation and overestimation volumes were com- 16.33 ± 16.67 cm puted as integrals over the z coordinate of the correspond- paired sample t-test revealed that the IGTV and MIP ing transverse areas as follows: IGTV differed significantly from the IGTV (p < 2Phases AllPhases 0.001), while the IGTV did not differ signifi- MIP-Modified cantly from the reference IGTV (p = 0.08). VA = ()z \A ()zdz, Under AllPhases Test Table 2 shows the MI values for each of the three test VA = ()z \A (() zdz, Over Test AllPhases IGTVs. As shown, the IGTV (mean ± SD: 0.90 ± MIP-Modified 0.02) most closely matched the IGTV , with AllPhases IGTV (mean ± SD: 0.81 ± 0.06) and IGTV (mean where A is the area in ten respiratory phases and 2Phases MIP AllPhases ± SD: 0.80 ± 0.05) following. There were no significant A is the test area. The underestimation area (A ) and Test Under differences between IGTV and IGTV (p = 0.728), the overestimation area (A ) defined as 2Phases MIP Over and IGTV but the differences in MI between IGTV MIP MIP- and those between IGTV IGTV were Modified 2Phases MIP-Modified AA = ()z \A ()z , Under AllPhases Test significant (p < 0.001, respectively) AA = ()z \A ()z ,, Over Test AllPhases We performed a comparative analysis of the MI values of were computed for each axial level by performing the the two patient groups (patients with SI motion ≤1 cm Delaunay triangulation for the union of the all phases and and those with SI motion > 1 cm) with stage I disease. test contour points and computing the areas as a sum of There was no strong correlation between the MI and the the corresponding triangular areas (see Figure 3). Given a magnitude of SI motion, although the MI of IGTV in set of data points in the plane, the Delaunay triangulation 2Phases some patients with SI motion ≤1 cm was lower than the is a set of triangles such that no data points are contained general trend in patients with SI motion > 1 cm. Although in any triangle's circumscribed circle. Delaunay triangula- the magnitude of SI motion did not significantly impact tions maximize the minimum angle of all the triangles in the accuracy of the IGTV contouring approaches, we the triangulation and they tend to avoid skinny (or close- found that the location of the primary tumor impacted to-degenerate) triangles. We used the Delaunay triangula- IGTV contouring accuracy (Table 2). For example, we tion implemented in a high-level graphical analysis and found that tumors located near the diaphragm (cases 1, 2, programming package, MATLAB (The Mathworks, Inc.: 3, and 15), mediastinum (case 8), and chest wall (cases 4, http://www.mathworks.com), which is based on the 6, 9, 10, and 12) appeared to have worse MI values than Quickhull algorithm [13]. tumors located in the peripheral lung parenchyma (cases 5, 7, 11, 13, 14, 16, and 17) although it didn't reach sta- Statistical analysis tistical significance. To estimate any statistically significant differences between the IGTVs determined using each test volume Table 3 shows the SI motion and the IGTVs based on the (IGTV , IGTV , and IGTV ) and the IGTV MIP 2Phases MIP-Modified test and all phases volumes for the 10 stage III lung determined using the all phases volume (IGTV ), we AllPhases tumors. As shown, the majority of these tumors (9/10) used a paired sample t-test in each case to determine p, exhibited SI motion < 1 cm, so it was not meaningful to with p < 0.05 considered significant. All statistical analyses group these patients according to the 1-cm-SI motion were performed using the SPSS software package (v.10; threshold. SPSS Inc., Chicago, IL). As with stage I lung tumors, we found that, regardless of Results the magnitude of SI motion, the IGTV and IGTV MIP 2Phases Table 1 shows the superior-inferior (SI) motion and the (mean ± SD: 193.27 ± 135.09 cm and 194.81 ± 133.86 IGTVs based on the test and all phases volumes for the cm , respectively) were consistently smaller than the stage I lung tumors. SI motion ranged from 0 cm to 2.17 IGTV (mean ± SD: 209.96 ± 139.95 cm ), whereas cm, with almost half (8/17) of the tumors exhibiting SI AllPhases the IGTV (mean ± SD: 206.00 ± 137.34 cm ) was MIP-Modified motion > 1 cm. To study the influence of magnitude of SI similar to the all phases IGTV. A paired sample t-test motion on the accuracy of IGTV delineation, we grouped revealed that the IGTV and IGTV differed signifi- the 17 patients into two groups: those with tumor motion MIP 2Phases cantly from the IGTV (p < 0.001), while the > 1.00 cm and those with tumor motion ≤1.00 cm. In gen- AllPhases IGTV differed less (p = 0.01). MIP-Modified eral, we found that, regardless of the magnitude of SI motion, the IGTV and IGTV (mean ± SD: 14.14 ± MIP 2Phases 3 3 14.89 cm and 13.93 ± 15.69 cm , respectively) were con- Page 5 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 IGTV in stage III tumors. In sum, we found that MIP-Modified the volumetric underestimation for IGTV was MIP-Modified consistently lower than the underestimation for IGTVs based on the other test volumes. We also observed that the volumetric underestimation percentages in stage III dis- ease were lower than those in stage I disease. However, because GTVs are by definition larger in stage III than in stage I disease, the absolute volume underestimation was generally higher in stage III disease. Volumetric overesti- mation occurred in both stage I and stage III disease for both IGTV and IGTV . Overestimation for MIP MIP-Modified IGTV was slightly higher than that for IGTV , MIP-Modified MIP but both percentages were lower than 5.0% for the aver- age volume overestimation and 10.10% for the maximum is a subset of volume overestimation. Because IGTV 2Phases IGTV , the volumetric overestimation for IGTV AllPhases 2Phases compared to the reference IGTV was always equal to zero. Figure 4 illustrates the proportional volumetric underesti- mations (Fig. 4a) and overestimations (Fig. 4b) in the 17 individual patients with stage I disease. We found that vol- umetric underestimation was > 10% using either IGTV MIP or IGTV in 15 patients, but in no patients when 2Phases Computa overestimatio the the Figure 3 dashed l solid li tion of ne ine) compared wi ) n area the under (light gray) of the test estim th are tion a ference area (area rea (da arr ek gr a (ar ay) ea i and the in nsi side de IGTV was used. Volumetric underestimation > MIP-Modified Computation of the underestimation area (dark 20% occurred in 5 patients using the IGTV and in 7 MIP gray) and the overestimation area (light gray) of the patients using the IGTV . Of the 5 patients in whom 2Phases test area (area inside the dashed line) compared with volumetric underestimation was > 20% using IGTV , 2 MIP reference area (area inside the solid line). The areas had lesions near or attached to the diaphragm, 1 had a were computed using the Delaunay triangulation which is lesion near or attached to the chest wall, and another had shown in the regions of interest. a lesion near or attached to the mediastinum. Figure 5 illustrates the volumetric underestimations (Fig. 5a) and Table 4 shows the MI values for each IGTV based on the overestimations (Fig. 5b) in the 10 patients with stage III test volumes and on the all phases volume for patients disease. We found that volumetric underestimation was > with stage III disease. In general, we found that the 5% in 9 patients using IGTV , 8 patients using MIP GTV -based IGTV (mean ± SD: 0.93 ± 0.20) IGTV , and 2 patients using IGTV . Volu- MIP-Modified 2Phases MIP-Modified matched the GTV -based IGTV the closest, followed metric underestimation > 10% occurred in 6 patients AllPhases by the IGTVs based on GTV (mean ± SD: 0.91 ± using IGTV , 1 patient using IGTV , but no patients 2Phases MIP 2Phases 0.05) and GTV (mean ± SD: 0.86± 0.07). There was a using IGTV . In general, we found that the lowest MIP MIP-Modified significant difference between GTV -based and volumetric underestimation was achieved consistently 2Phases GTV -based IGTVs (p = 0.05) and between GTV - using the modified MIP approach to delineate the IGTV. MIP MIP based and GTV -based IGTVs (p = 0.03). MIP-Modified To analyze the accuracy of these contouring approaches in The volumetric underestimation and overestimation involved lymph nodes, we conducted the second analysis between the all phases volume and the test volumes for of involved lymph nodes in above stage III disease. Our patients with stage I and III disease are shown in Table 5. data showed that IGTV volume of lymph nodes MIP-Modified For stage I disease, the maximum volumetric underesti- (mean ± SD: 32.95 ± 40.86 cm ) matched most closely mations for IGTV , IGTV , and IGTV with IGTV volumes of lymph nodes (mean ± SD: MIP 2Phases MIP-Modified AllPhases compared to IGTV were 30.86%, 21.2%, and 34.26 ± 42.56 cm , p = 0.24), while IGTV and AllPhases 2Phases 8.53%, respectively. For stage III disease, the maximum IGTV lymph node volumes (mean ± SD: 29.15 ± 38.14 MIP volumetric underestimations for IGTV , IGTV , and and 25.63 ± 34.55 cm respectively) differed significantly MIP 2Phases IGTV compared to IGTV were 23.85%, with IGTV lymph node volume (p = 0.04 and 0.05 MIP-Modified AllPhases AllPhases 22.25%, and 6.66%, respectively. The average volumetric respectively, volume underestimation in all cases). In underestimation was 17.3% for IGTV , 19.3% for addition, the match index of lymph node IGTV MIP MIP-Modified IGTV , and 5.3% for IGTV in stage I tumors was not significantly different from IGTV (p = 0.14) 2Phases MIP-Modified 2Phases and 12.1% for IGTV , 8.9% for IGTV , and 4.2% for but was significantly different from IGTV values (p = MIP 2Phases MIP Page 6 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 1: SI motion and IGTVs based on the test volumes (IGTV , IGTV , and IGTV ) and the reference volume MIP 2Phases MIP-Modified (IGTV ) for stage I tumors AllPhases Patient No SI Motion (cm) IGTV IGTV IGTV IGTV MIP MIP-Modified AllPhases 2Phases 3 3 3 3 (cm ) (cm ) (cm ) (cm ) 1 1.06 5.82 7.98 8.32 6.56 2 1.37 8.53 10.51 9.98 7.49 3 1.70 12.88 15.85 16.54 13.23 4 1.08 4.92 5.49 5.44 3.72 5 0.15 1.64 1.79 1.80 1.46 6 2.17 17.64 22.16 23.39 18.98 7 1.27 23.08 26.06 26.28 21.77 8 0.54 12.45 15.73 15.76 12.82 9 0.18 21.80 24.50 24.96 20.92 10 0.00 60.04 66.90 67.75 63.67 11 0.41 2.46 2.80 2.85 2.27 12 1.77 32.90 37.65 39.39 33.14 13 0.14 2.08 2.27 2.23 1.84 14 0.10 1.53 1.90 1.93 1.74 15 1.62 18.59 21.26 21.69 16.66 16 0.66 10.70 11.31 10.60 7.77 17 0.09 3.35 3.45 3.33 2.70 0.001 for both cases). IGTV and IGTV not been compared with that of ten-phase contouring MIP-Modified 2Phases matched better with IGTV (match index mean ± approach particularly in stage III disease. Jin et al, in a AllPhases SD: 0.81 ± 0.08, range: 0.75–0.91 for IGTV , and phantom study, examined the feasibility of a method to MIP-Modified 0.77 ± 0.08, range: 0.65 to 0.88 for IGTV ) compared determine ITV based on motion information obtained 2Phases with IGTV (mean ± SD: 0.62 ± 0.11; Range: 0.46 to from select phases of a respiratory cycle [15]. They MIP 0.76). reported that adequate estimation of IGTV could in gen- eral be achieved by combining motion information from Discussion the extremes of motion in most cases and in some cases by Real-time tumor motion tracking provides most compre- the addition of motion information from an intermediate hensive data for respiratory tumor motion management. phase. Underberg et al. [8] reported that MIP-based con- However, it is a challenging technique to implement in touring could provide reliable margins for determining the clinical setting and more research is needed to make the IGTV for stage I lung tumors treated with SBRT. How- its clinical implementation more practical [14]. Although ever, their method did not include visual verification of both MIP-based and two-phase-based approaches have the MIP-defined GTV contour through each individual been shown to more accurately delineate the GTV than phase of the 4D CT (IGTV ). Bradley et al. [9] MIP-Modified conventional 3D CT-based planning, their accuracy has compared helical-, MIP-, and average-intensity (AI)-based Page 7 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 2: Matching index values for each IGTV based on IGTV , IGTV , and IGTV relative to the reference IGTV MIP 2Phases MIP-Modified AllPhases in stage I disease Patient No Location (Adjacent) IGTV IGTV IGTV MIP 2Phases MIP-Modified 1 Diaphragm 0.69 0.79 0.88 2 Diaphragm 0.79 0.75 0.88 3 Diaphragm 0.77 0.80 0.92 4 Chest wall 0.82 0.68 0.90 5 Lung parenchyma 0.83 0.81 0.88 6 Chest wall 0.72 0.80 0.89 7 Lung parenchyma 0.84 0.83 0.91 8 Mediastinum 0.74 0.81 0.92 9 Chest wall 0.84 0.84 0.91 10 Chest wall 0.87 0.94 0.95 11 Lung parenchyma 0.79 0.80 0.90 12 Chest wall 0.83 0.84 0.93 13 Lung parenchyma 0.82 0.83 0.90 14 Lung parenchyma 0.75 0.90 0.91 15 Diaphragm 0.79 0.77 0.89 16 Lung parenchyma 0.85 0.73 0.88 17 Lung parenchyma 0.88 0.81 0.91 4-D CT imaging to find the optimal approach for deter- over all individual phases of the 4D CT yielded the best mining the patient-specific IGTV for SBRT for stage I lung estimate of IGTV. However, there the performance of this cancer. They found that the MIP-defined GTV was signifi- approach in the delineation of involved lymph nodes was cantly larger than the helical-defined and average CT- not separately addressed [11]. In daily clinical practice, defined GTVs. However, in their study, Bradley et al. did tumor contouring in stage III disease is more challenging not compare the GTV based on GTV with that based on than in stage I disease because of the larger tumor volume, MIP GTV , the optimal reference volume. Bradley et al. more complicated tumor shape, involvement of critical AllPhases [9] did not discuss their results in the context of tumor structures, and potential involvement of multiple lymph location in their study. In another study, Cai et al. [10] nodes in which tissue density is similar to that of the determined the IGTVs for six lung tumors using a simula- tumor. In addition, although the two-phase-based tion method based on dynamic magnetic resonance imag- approach has been used to delineate IGTVs in the clinical ing (dMRI) and MIPs. They found that MIP-based IGTVs setting, there is scant data on the accuracy of such two- were smaller than dMRI-based IGTVs. They concluded phase-based IGTVs in either stage I or stage III disease that because of the low temporal resolution and retrospec- [16]. Our study showed that both MIP-based and two- tive re-sorting, 4-D CT might not accurately depict the phase-based IGTVs underestimate the 10-phase-based excursion of a moving tumor. Recent data by Rietzel et al IGTV in both stage I and III disease including involved also support our observation that tumor delineation on lymph nodes, which can potentially result in marginal the MIP with subsequent visual verification of contours under-dosing, and that the IGTV consistently MIP-Modified Page 8 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 3: SI motion and IGTVs based on the test volumes (IGTV , IGTV , and IGTV ) and the reference volume MIP 2Phases MIP-Modified (IGTV ) for stage III tumors AllPhases Patient No SI Motion (cm) IGTV IGTV IGTV IGTV MIP MIP-Modified AllPhases 2Phases 3 3 3 3 (cm ) (cm ) (cm ) (cm ) 1 0.09 64.91 77.33 79.94 74.42 2 0.12 202.85 228.55 238.40 216.29 3 0.21 135.59 146.62 151.40 138.72 4 0.18 221.50 230.85 233.74 222.22 5 0.62 23.87 30.59 29.98 23.33 6 0.11 446.14 450.31 458.21 439.86 7 0.96 242.38 265.47 268.58 244.76 8 0.14 347.06 368.97 373.61 351.46 9 0.18 36.69 39.40 36.87 34.03 10 1.77 211.70 221.96 228.89 203.01 had the lowest percentages of volumetric underestima- display mobile structures if the adjacent structures have tion, which indicates that the IGTV approach is similar (or higher) densities, which is the case for lesions MIP-Modified the most accurate in delineating the IGTV. located near the mediastinum, diaphragm, liver, or chest wall; and (2) the physician may misinterpret the MIP For the MIP-based approach, several potential sources of images because of tumor border smearing. (3) The tumor uncertainty/error exist: (1) the MIP image may not fully spicula can not be visualized on the MIP projections due to smearing of the tumor edge. Indeed, our data show that Table 4: Matching index values for each IGTV based on IGTV , MIP the MI was poor and volumetric underestimation was IGTV , and IGTV relative to the reference 2Phases MIP-Modified high using the MIP-based approach to delineate IGTVs in IGTV in stage III disease AllPhases most of lesions near the mediastinum, diaphragm, liver, Patient No GTV GTV GTV and chest wall. Of these lesions, those closer to the dia- MIP 2Phases MIP-Modified phragm and liver had the lowest MI values, which could 1 0.76 0.93 0.92 have been due to the significant motion of the diaphragm and liver and the MIP image's inability to record differ- 2 0.83 0.91 0.92 ences between the lesion and the diaphragm and liver. We are currently developing software that excludes dia- 3 0.88 0.92 0.94 phragm and liver images in some breathing phases using cine CT images so that better tumor MIP images will be 4 0.92 0.95 0.95 preserved (data to be published). We should note that MIP images do not reflect the densities of tumors, lungs, 5 0.74 0.78 0.90 and other normal tissues accurately enough for dose cal- culation in treatment planning [17]. Thus, a free-breath- 6 0.95 0.96 0.95 ing CT image set, a 4-D scan of a single respiratory phase, or an average CT image set extracted from a 4-D CT data 7 0.87 0.91 0.94 set should be used for treatment planning and dose calcu- lation. This would be especially important in proton ther- 8 0.90 0.94 0.94 apy, which is more sensitive to tumor motion and changes 9 0.87 0.92 0.91 in tissue density. In a previous study on 4-D CT in proton therapy planning, we found that a MIP density override 10 0.86 0.89 0.90 Page 9 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 5: Summary of the volumetric percentage underestimation and overestimation for each IGTV based on IGTV , IGTV , MIP 2Phases and IGTV relative to the reference IGTV . MIP-Modified AllPhases Underestimation (%) IGTV IGTV IGTV MIP 2Phases MIP-Modified Stage I patients Avg. ± SD 17.33 ± 6.56 19.32 ± 5.93 5.36 ± 1.71 Range 6.32–30.86 6.03–31.76 2.30–8.53 Stage III patients Avg. ± SD 12.11 ± 6.23 8.95 ± 5.15 4.21 ± 1.66 Range 4.04–23.85 4.01–22.25 1.20–6.66 Overestimation (%) IGTV IGTV IGTV MIP 2Phases MIP-Modified Stage I patients Avg. ± SD 3.23 ± 2.35 0 4.80 ± 2.39 Range 0.34–8.64 0 1.30–10.07 Stage III patients Avg. ± SD 2.36 ± 1.79 0 3.21 ± 2.22 Range 1.06–6.92 0 1.43–8.09 Average ± standard deviation and range are reported for stage I and stage III tumors. for tumor contouring in an average CT data set was the In clinical setting, it is common to prescribe the dose to optimal approach [18]. PTV which takes additionally clinical target volume (CTV) and set-up uncertainty into consideration. The volume- For the two-phase-based approach, tumor deformation underestimation will be reduced if PTV was used to com- between the two extreme phases of breathing and the pare above mentioned four approaches. We evaluated the curved motion pathway during each breathing cycle may effect of this underestimation on the PTV in a case with introduce uncertainty. In most cases, however, we found maximal underestimation of the IGTV in stage I disease. that the MI of the two-phase-based IGTV was slightly IGTV was expanded by 1.6 cm (0.8 cm for CTV, 0.3 cm to higher than that of MIP-based IGTV, which indicates that account for variability in the determination of motion most tumors moved in a generally straightforward SI extent and 0.5 cm for image guided patient setup). Analy- direction and that tumor deformation during breathing sis of volumetric underestimation of the PTV was carried was minimal. Particularly in stage III disease, we found out in the same manner as described for IGTV. Our results that the volumetric underestimation was generally lower showed that the volume underestimation reduced from for the two-phase-based IGTV than for the MIP-based 30.86%, 21.2%,8.53% in IGTV to 13.3%, 5.18% and IGTV. Therefore, if 4-D CT based IGTV is not 3.36% in PTV for IGTV , IGTV , IGTV MIP-Modified MIP 2Phases MIP-Modified available, the two-phase-based IGTV is a reasonable alter- respectively. In general, this improvement is more dra- native approach to take tumor motion into consideration matic in the lesions with the smaller size such as stage I although it is not optimal one. disease. However, when ablative dose is attempted in clin- ical setting but sparing critical structures is concerning Page 10 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 a. b. S Figure 4 tage I tumors Stage I tumors. (a) Volumetric underestimation for each IGTV based on IGTV , IGTV , and IGTV relative to MIP 2Phases MIP-Modified the reference IGTV . (b) Volumetric overestimation for each IGTV based on IGTV , IGTV , and IGTV AllPhases MIP 2Phases MIP-Modified relative to the IGTV . (Note: IGTV is a subset of IGTV , hence the volumetric overestimation for IGTV AllPhases 2Phases AllPhases 2Phases is always equal to zero.) Page 11 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 a. b. Stage II Figure 5 I tumors Stage III tumors. (a) Volumetric underestimation for each IGTV based on IGTV , IGTV , and IGTV relative MIP 2Phases MIP-Modified to the IGTV . (b) Volumetric overestimation for each IGTV based on IGTV , IGTV , and IGTV relative to AllPhases MIP 2Phases MIP-Modified the IGTV . (Note: IGTV is a subset of IGTV , hence the volumetric overestimation for IGTV is always AllPhases 2Phases AllPhases 2Phases equal to zero.) Page 12 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 port. We also thank the Department of Scientific Publications for its such as SBRT in stage I disease, we would accept compro- assistance in preparing this article. mised coverage for PTV but not for IGTV. Therefore, IGTV delineation accuracy is still crucial clinically. References 1. Curran W, Scott C, Langer C: Long term benefit is observed in As with other such comparative studies mentioned above, a phase III comparison of sequential vs concurrent chemo- inter or intra observer variability in the delineation of the radiation for patients with unresectable NSCLC: RTOG 9410 [abstract]. Proc Am Soc Clin Oncol 2003:S621a. GTV was not considered. The uncertainties introduced as 2. Chang JY, Dong L, Liu H, Starkschall G, Balter P, Mohan R, Liao Z, a result of the above could however be thought to be dif- Cox JD, Komaki R: Image-guided radiation therapy for non- small cell lung cancer. Journal of Thoracic Oncology 2008, ferent from those analyzed in this study, thereby requiring 3:177-186. a separate analysis that is beyond the scope of the current 3. Liu HH, Balter P, Tutt T, Choi B, Zhang J, Wang C, Chi M, Luo D, Pan report. T, Hunjan S, Starkschall G, Rosen I, Prado K, Liao Z, Komaki R, Cox JD, Mohan R, Dong L: Assessing respiration-induced tumor motion and internal target volume using 4DCT for radiation Conclusion therapy of lung cancer. Int J Radiat Oncol Biol Phys 2007, 68:531-540. We found that the MIP-based and two-phase-based 4. Donnelly ED, Parikh PJ, Lu W, Zhao T, Lechleiter K, Nystrom M, approaches to IGTV delineation significantly underesti- Hubenschmidt JP, Low DA, Bradley JD: Assessment of intrafrac- mated the IGTV in patients with stage I and stage III tion mediastinal and hilar lymph node movement and com- parison to lung tumor motion using four-dimensional CT. Int NSCLC. Due to the limitations of each approach, a signif- J Radiat Oncol Biol Phys 2007, 69:580-588. icant amount of the tumor volume could be missed in 5. Vedam SS, Keall PJ, Kini VR, Mostafavi H, Shukla HP, Mohan R: individual patient so precautions should be taken when Acquiring a four-dimensional computed tomography data- set using an external respiratory signal. Phys Med Biol 2003, these techniques are used to treat patients. We also found 48:45-62. that the IGTV approach, which requires visual MIP-Modified 6. Keall PJ, Starkschall G, Shukla H, Forster KM, Ortiz V, Stevens CW, Vedam SS, George R, Guerrero T, Mohan R: Acquiring 4D tho- verification of tumor coverage after each phase of the racic CT scans using a multislice helical method. Phys Med Biol breathing cycle, improved IGTV delineation in both cases. 2004, 49:2053-2067. 7. Keall PJ, Mageras GS, Balter JM, Emery RS, Forster KM, Juang SB, Kap- atoes JM, Low DA, Murphy MJ, Murray BR, Ramsey CR, Van Herk MB, Abbreviations Vedam SS, Wong JW, York E: The management of respiratory GTV: gross tumor volume; IGTV: internal gross tumor vol- motion in radiation oncology report of AAPM Task Group ume; CTV: Clinical target volume; PTV: Planning target 76. Med Phys 2006, 33:3874-3900. 8. Underberg RW, Lagerwaard FJ, Slotman BJ, Cuijpers JP, Senan S: Use volume; IGTV : the gross tumor volume (GTV) con- AllPhases of maximum intensity projections (MIP) for target volume tours from ten respiratory phases; IGTV : the GTV 2Phases generation in 4DCT scans for lung cancer. Int J Radiat Oncol Biol Phys 2005, 63:253-260. contours from two extreme respiratory phases (0% and 9. Bradley JD, Nofal AN, El Naqa IM, Lu W, Liu J, Hubenschmidt J, Low 50%); IGTV : the GTV contour using the maximum MIP DA, Drzymala RE, Khullar D: Comparison of helical, maximum intensity projection (MIP); IGTV : the GTV con- intensity projection (MIP), and averaged intensity (AI) 4D MIP-Modified CT imaging for stereotactic body radiation therapy (SBRT) tour using the MIP with modification based on visual ver- planning in lung cancer. Radiother Oncol 2006, 81:264-268. ification of contours in individual respiratory phase. 10. Cai J, Read PW, Baisden JM, Larner JM, Benedict SH, Sheng K: Esti- mation of error in maximal intensity projection-based inter- nal target volume of lung tumors: A simulation and Competing interests comparison study using dynamic magnetic resonance imag- The authors declare that they have no competing interests. ing. Int J Radiat Oncol Biol Phys 2007, 69:895-902. 11. Reitzel E, Liu A, Chen G, Choi N: Maximum-intensity volumes for fast contouring of lung tumors including respiratory Authors' contributions motion in 4DCT planning. Int J Radiat Oncol Biol Phys 2008, ME, SV, and JYC designed/conducted analysis and wrote 71(4):1245-1252. 12. Pan T, Lee TY, Rietzel E, Chen GT: 4-D CT imaging of a volume the manuscript. JYC was responsible for manuscript revi- influenced by respiratory motion on multi-slice CT. Med Phys sion and submission. PB, BC and GS were involved in 4- 2004, 31:333-340. 13. Barber CB, Dobkin DP, Huhdanpaa HT: The quickhull algorithm D CT simulation and treatment designed. DM was for convex hulls. ACM Transactions on Mathematical Software 1996, involved in data analysis. 22(4):469-483. 14. Keall PJ, Cattell H, Pokhrel D, Dieterich S, Wong KH, Murphy MJ, Vedam SS, Wijesooriya K, Mohan R: Geometric accuracy of a Authors' information real-time target tracking system with dynamic multileaf col- Dr. Chang is a recipient of the Research Scholar Award limator tracking system. Int J Radiat Oncol Biol Phys 2006, from the Radiological Society of North America and a 65:1579-1584. 15. Jin J, Ajlouni M, Chen Q, Yin FF, Movsas B: Lung radiotherapy: A Development Award from The University of Texas M. D. technique of using gated-CT images to determine internal Anderson Cancer Center NIH Lung Cancer SPORE (P50 target volume (ITV) for fractionated stereotactic lung radi- otherapy. Radiotherapy and Oncology 2006, 78:177-184. CA70907). 16. Sarrut D, Boldea V, Miguet S, Ginestet C: Simulation of four- dimensional CT images from deformable registration between inhale and exhale breath-hold CT scans. Med Phys Acknowledgements 2006, 33:605-617. We thank all members of the Thoracic Radiation Oncology team, the 17. Admiraal MA, Schuring D, Hurkmans CW: Dose calculations attending physicians who enrolled patients in this study, and the dosime- accounting for breathing motion in stereotactic lung radio- trists who were involved in the patients' treatment for their help and sup- Page 13 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 therapy based on 4D-CT and the internal target volume. Radiother Oncol 2008, 86:55-60. 18. Kang Y, Zhang X, Chang JY, Wang H, Wei X, Liao Z, Komaki R, Cox JD, Balter PA, Liu H, Zhu XR, Mohan R, Dong L: 4D proton treat- ment planning strategy for mobile lung tumors. Int J Radiat Oncol Biol Phys 2007, 67:906-914. Publish with Bio Med Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical researc h in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright BioMedcentral Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 14 of 14 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiation Oncology Springer Journals

Determination of patient-specific internal gross tumor volumes for lung cancer using four-dimensional computed tomography

Loading next page...
 
/lp/springer-journals/determination-of-patient-specific-internal-gross-tumor-volumes-for-zZQX8lUjG4
Publisher
Springer Journals
Copyright
Copyright © 2009 by Ezhil et al; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Oncology; Radiotherapy
eISSN
1748-717X
DOI
10.1186/1748-717X-4-4
pmid
19173738
Publisher site
See Article on Publisher Site

Abstract

Background: To determine the optimal approach to delineating patient-specific internal gross target volumes (IGTV) from four-dimensional (4-D) computed tomography (CT) image data sets used in the planning of radiation treatment for lung cancers. Methods: We analyzed 4D-CT image data sets of 27 consecutive patients with non-small-cell lung cancer (stage I: 17, stage III: 10). The IGTV, defined to be the envelope of respiratory motion of the gross tumor volume in each 4D-CT data set was delineated manually using four techniques: (1) combining the gross tumor volume (GTV) contours from ten respiratory phases (IGTV ); (2) AllPhases combining the GTV contours from two extreme respiratory phases (0% and 50%) (IGTV ); (3) 2Phases defining the GTV contour using the maximum intensity projection (MIP) (IGTV ); and (4) defining MIP the GTV contour using the MIP with modification based on visual verification of contours in individual respiratory phase (IGTV ). Using the IGTV as the optimum IGTV, we MIP-Modified AllPhases compared volumes, matching indices, and extent of target missing using the IGTVs based on the other three approaches. Results: The IGTV and IGTV were significantly smaller than the IGTV (p < 0.006 for MIP 2Phases AllPhases stage I and p < 0.002 for stage III). However, the values of the IGTV were close to those MIP-Modified determined from IGTV (p = 0.08). IGTV also matched the best with IGTV . AllPhases MIP-Modified AllPhases Conclusion: IGTV and IGTV underestimate IGTVs. IGTV is recommended to MIP 2Phases MIP-Modified improve IGTV delineation in lung cancer. Background been considered one of the main reasons for local failure Lung cancer remains the leading cause of cancer-related [2]. Researchers have reported that ~40% of lung tumors mortality. Conventional photon radiotherapy for lung move > 5 mm and that 10–12% move > 1 cm [3,4]. Sev- cancer is associated with about 50% local tumor control eral strategies have recently been developed to address the [1]. Missing the target as a result of tumor motion has issue of tumor motion and improve local control [2]. For Page 1 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 example, the development of image-guided radiotherapy touring in only one dataset. In these instances, post- (IGRT) has allowed for more accurate tumor targeting, so processing tools, such as the maximum intensity projec- it is rapidly replacing conventional radiotherapy for lung tion (MIP), have been shown to improve radiotherapy cancer [2]. In order to account for tumor motion, the planning efficiency [7]. The MIP of a 4D-CT data set International Commission on Radiation Units and Meas- reduces the multiple 3-D CT data available from a 4-D CT urements (ICRU) report 62 introduced the concept of an data set into a single 3-D CT data set, where each voxel in internal target volume (ITV), defined as the clinical target the MIP represents the maximum intensity encountered volume (CTV) plus an additional margin to account for by corresponding voxels in all individual 3-D phase image geometric uncertainties due to internal variations in sets of the 4-D CT data set. The IGTV is then determined tumor position, size, and shape. Using current imaging based on the GTV delineation on the single 3-D CT data techniques, the CTV cannot be visualized. Consequently, set. Alternatively, some cancer centers have used breath- generation of the ITV requires delineation of the gross hold spiral CT imaging to acquire images at the two tumor volume (GTV) on each of the phases that constitute extremes of the respiratory cycle [2,7]; contouring the GTV the four-dimensional (4-D) computed tomography (CT) at these extremes (the end-expiration and the end-inspira- image data set, followed by expansion of each GTV to tion phases) and then combining these two 3-D volumes account for microscopic disease. The ITV is then deter- yields the IGTV. A limited number of studies have ana- mined to be the envelope of motion of the CTV. In order lyzed the accuracy of the MIP and two-phase IGTV delin- to make the determination of the ITV more efficient, we eation techniques relative to full ten-phase method for have proposed the concept of the internal gross tumor determining IGTV [8-11]. volume (IGTV), which explicitly accounts for internal var- iations in tumor position, size, and shape but can be The aim of this study, therefore, was to evaluate the accu- derived directly from imaging studies [2]. The ITV is then racy of 4-D CT MIP-based IGTV delineation and two- determined to be the IGTV plus a margin that accounts for phase-based IGTV delineation compared to ten-phase microscopic disease. IGTV delineation as a reference. We also examined the accuracy of the MIP-based IGTV delineation after applying Traditionally, the margin necessary to account for internal a modification through visual verification of GTV cover- motion of tumors in the thorax has been determined age in individual respiratory phases. using an isotropic expansion determined by population- based estimates of respiratory motion. However, because Methods breathing characteristics vary greatly among individual Data acquisition patients, such population-based estimates may overesti- As a retrospective review of radiation treatment planning, mate or underestimate the margin needed for a given this study was included under an Institutional Review patient. Moreover, respiratory-induced tumor motion is Board-approved retrospective chart review protocol. We known not to be anisotropic; typical tumor paths are studied 27 consecutive patients with non-small-lung can- those of elongated and possible curved ellipses. The cer (NSCLC) who underwent 4-D CT simulation for treat- advent of the multislice helical CT scanner combined with ment planning and received definitive radiotherapy at our the establishment of temporal correlation between respi- institution between 2005 and 2006. Of these 27 patients, ratory motion and the CT acquisition process have 17 had stage I disease and received stereotactic body radi- allowed tumor size, shape, and position to be observed at otherapy (SBRT), and 10 had stage III disease and received multiple times during a patient's respiratory cycle [5,6]. intensity-modulated radiotherapy (IMRT). 4-D CT image The resultant CT data set, called the 4-D CT or respiration- data sets each consisting of 10 respiratory phases, were correlated CT data set, provides patient-specific informa- acquired on a multislice CT scanner (Discovery ST, GE tion about tumor position, shape, and size at different Medical Systems, Madison, WI) by sorting CT images phases of the respiratory cycle. based on the phase of an external respiratory monitor (Real-time Position Management System; Varian Medical Although using 4-D CT data provides a reliable estimate Systems, Inc., Palo Alto, CA) [12]. MIPs of the 4D-CT data of the extent of tumor motion due to respiration in three sets were then generated from the individual phase images dimensions, its clinical implementation poses some chal- as described elsewhere [5,6]. lenges. Ideally, the IGTV should be determined by con- touring the GTV on each of the ten phase image sets. The Patient-specific IGTV determination combination of these individual three dimensional (3-D) We determined patient-specific IGTVs using the demon- volumes into a single 3-D volume represents the IGTV, strable extent of tumor motion shown in the 4-D CT which accounts for respiratory motion. However, con- images. We used four approaches to determine these touring the tumor volume on ten different data sets for IGTVs: (1) contouring the GTV on each of the ten respira- each patient increases the workload compared with con- tory phases of the 4D-CT data set and combining these Page 2 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 GTVs to produce IGTV ; (2) contouring the GTV on sistency in contouring, all GTV contours in each respira- AllPhases the MIP of the 4-D CT data set to produce IGTV ; (3) tory phase of the 4-D CT and MIP data sets were drawn by MIP contouring the GTV on the extreme respiratory phases a single radiation oncologist (ME) and verified by another (0% phase = peak inhalation, 50% phase = peak exhala- radiation oncologist (JYC). We used a lung window on the tion) and combining these GTVs to produce IGTV ; CT data set to contour the primary tumor and a mediasti- 2Phases and (4) contouring the GTV on the MIP of the 4-D CT data num window to contour any involved lymph nodes. set and then modifying these contours using visual verifi- Diagnostic CT of chest with intravenous contrast and PET/ cation of coverage in each phase of the 4-D CT data set to CT were used to guide our involved lymph nodes contour- produce IGTV . Visual verification of coverage in ing as described by our previous publication (2). A total MIP-Modified each phase was achieved by overlaying the MIP based GTV of 324 GTVs were delineated with 12 GTVs delineated for contour onto each phase of the 4-D CT data set. Thus, each patient (GTV in each of 10 respiratory phases, each of these 3D volumes (IGTV , IGTV , IGTV , and IGTV ). For stage III disease, AllPhases MIP MIP MIP-Modified IGTV , and IGTV ) represented the demon- involved hilar or mediastinal lymph nodes were con- 2Phases MIP-Modified strable respiratory tumor motion volumes, or IGTVs. Fig- toured and analyzed independently. ures 1 and 2 show the results obtained using these different approaches in the determination of IGTV for cases of stage I and stage III disease, respectively. For con- 4- Figure 1 Delineation of IGTV for sta D CT data set ge I lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) IGTV of a MIP MIP-Modified 2Phases AllPhases Delineation of IGTV for stage I lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) MIP MIP-Modified 2Phases IGTV of a 4-D CT data set. MIP-based contours, as shown in panels (a) and (b), are as they appear on the MIP data AllPhases set. Phase-based contours, as shown in panels (c) and (d), are registered to the peak exhalation phase of the 4-D CT data set. Page 3 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 D 4- Figure 2 e D CT lineati da on o ta se f IGTV f t or stage III lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) IGTV of a MIP MIP-Modified 2Phases AllPhases Delineation of IGTV for stage III lung tumors based on (a) IGTV , (b) IGTV , (c) IGTV , and (d) MIP MIP-Modified 2Phases IGTV of a 4-D CT data set. MIP-based contours, as shown in panels (a) and (b), are as they appear on the MIP data AllPhases set. Phase-based contours, as shown in panels (c) and (d), are registered to the peak exhalation phase of the 4D-CT data set. Data analysis the minimum value is 0 if the volumes are completely We evaluated the IGTVs determined using each of the non-overlapping. three contouring approaches against an all phases IGTV determined by contouring all ten respiratory phases of the Volume difference calculation 4-D CT data set (IGTV ). Specifically, we compared While the matching index is a good measure of how well AllPhases the following metrics for each 3D volume: matching the shape of any two volumes match each other, it cannot index, total GTV volume and under or over-estimated vol- discriminate between overestimation and underestima- ume. tion. To gain better insight into any over/underestimation of the IGTV, we computed the differences in IGTV Matching index calculation between the all phases volume (IGTV ) and the AllPhases The matching index (MI) of any two 3D volumes A and B three test volumes (IGTV , IGTV , and IGTV MIP 2Phases MIP-Mod- is defined as the ratio of the intersection of A with B to the ). For each pair of volumes, we computed the underes- ified union of A and B, that is, timation and overestimation volumes (V and V ) Under Over using the following equations: AB ∩ MI = . VV = \V AB ∪ Under AllPhases Test VV = \, V Over Test AllPhases As can be deduced from this equation, the maximum value of the MI is 1 if the two volumes are identical, and Page 4 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 where V is the volume in ten respiratory phases, V sistently smaller than the IGTV (mean ± SD: 16.60 AllPhases test AllPhases is the test volume, and "\" denotes the set difference. The ± 17.05 cm ), whereas the IGTV (mean ± SD: MIP-Modified ) were similar to the reference IGTV. A underestimation and overestimation volumes were com- 16.33 ± 16.67 cm puted as integrals over the z coordinate of the correspond- paired sample t-test revealed that the IGTV and MIP ing transverse areas as follows: IGTV differed significantly from the IGTV (p < 2Phases AllPhases 0.001), while the IGTV did not differ signifi- MIP-Modified cantly from the reference IGTV (p = 0.08). VA = ()z \A ()zdz, Under AllPhases Test Table 2 shows the MI values for each of the three test VA = ()z \A (() zdz, Over Test AllPhases IGTVs. As shown, the IGTV (mean ± SD: 0.90 ± MIP-Modified 0.02) most closely matched the IGTV , with AllPhases IGTV (mean ± SD: 0.81 ± 0.06) and IGTV (mean where A is the area in ten respiratory phases and 2Phases MIP AllPhases ± SD: 0.80 ± 0.05) following. There were no significant A is the test area. The underestimation area (A ) and Test Under differences between IGTV and IGTV (p = 0.728), the overestimation area (A ) defined as 2Phases MIP Over and IGTV but the differences in MI between IGTV MIP MIP- and those between IGTV IGTV were Modified 2Phases MIP-Modified AA = ()z \A ()z , Under AllPhases Test significant (p < 0.001, respectively) AA = ()z \A ()z ,, Over Test AllPhases We performed a comparative analysis of the MI values of were computed for each axial level by performing the the two patient groups (patients with SI motion ≤1 cm Delaunay triangulation for the union of the all phases and and those with SI motion > 1 cm) with stage I disease. test contour points and computing the areas as a sum of There was no strong correlation between the MI and the the corresponding triangular areas (see Figure 3). Given a magnitude of SI motion, although the MI of IGTV in set of data points in the plane, the Delaunay triangulation 2Phases some patients with SI motion ≤1 cm was lower than the is a set of triangles such that no data points are contained general trend in patients with SI motion > 1 cm. Although in any triangle's circumscribed circle. Delaunay triangula- the magnitude of SI motion did not significantly impact tions maximize the minimum angle of all the triangles in the accuracy of the IGTV contouring approaches, we the triangulation and they tend to avoid skinny (or close- found that the location of the primary tumor impacted to-degenerate) triangles. We used the Delaunay triangula- IGTV contouring accuracy (Table 2). For example, we tion implemented in a high-level graphical analysis and found that tumors located near the diaphragm (cases 1, 2, programming package, MATLAB (The Mathworks, Inc.: 3, and 15), mediastinum (case 8), and chest wall (cases 4, http://www.mathworks.com), which is based on the 6, 9, 10, and 12) appeared to have worse MI values than Quickhull algorithm [13]. tumors located in the peripheral lung parenchyma (cases 5, 7, 11, 13, 14, 16, and 17) although it didn't reach sta- Statistical analysis tistical significance. To estimate any statistically significant differences between the IGTVs determined using each test volume Table 3 shows the SI motion and the IGTVs based on the (IGTV , IGTV , and IGTV ) and the IGTV MIP 2Phases MIP-Modified test and all phases volumes for the 10 stage III lung determined using the all phases volume (IGTV ), we AllPhases tumors. As shown, the majority of these tumors (9/10) used a paired sample t-test in each case to determine p, exhibited SI motion < 1 cm, so it was not meaningful to with p < 0.05 considered significant. All statistical analyses group these patients according to the 1-cm-SI motion were performed using the SPSS software package (v.10; threshold. SPSS Inc., Chicago, IL). As with stage I lung tumors, we found that, regardless of Results the magnitude of SI motion, the IGTV and IGTV MIP 2Phases Table 1 shows the superior-inferior (SI) motion and the (mean ± SD: 193.27 ± 135.09 cm and 194.81 ± 133.86 IGTVs based on the test and all phases volumes for the cm , respectively) were consistently smaller than the stage I lung tumors. SI motion ranged from 0 cm to 2.17 IGTV (mean ± SD: 209.96 ± 139.95 cm ), whereas cm, with almost half (8/17) of the tumors exhibiting SI AllPhases the IGTV (mean ± SD: 206.00 ± 137.34 cm ) was MIP-Modified motion > 1 cm. To study the influence of magnitude of SI similar to the all phases IGTV. A paired sample t-test motion on the accuracy of IGTV delineation, we grouped revealed that the IGTV and IGTV differed signifi- the 17 patients into two groups: those with tumor motion MIP 2Phases cantly from the IGTV (p < 0.001), while the > 1.00 cm and those with tumor motion ≤1.00 cm. In gen- AllPhases IGTV differed less (p = 0.01). MIP-Modified eral, we found that, regardless of the magnitude of SI motion, the IGTV and IGTV (mean ± SD: 14.14 ± MIP 2Phases 3 3 14.89 cm and 13.93 ± 15.69 cm , respectively) were con- Page 5 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 IGTV in stage III tumors. In sum, we found that MIP-Modified the volumetric underestimation for IGTV was MIP-Modified consistently lower than the underestimation for IGTVs based on the other test volumes. We also observed that the volumetric underestimation percentages in stage III dis- ease were lower than those in stage I disease. However, because GTVs are by definition larger in stage III than in stage I disease, the absolute volume underestimation was generally higher in stage III disease. Volumetric overesti- mation occurred in both stage I and stage III disease for both IGTV and IGTV . Overestimation for MIP MIP-Modified IGTV was slightly higher than that for IGTV , MIP-Modified MIP but both percentages were lower than 5.0% for the aver- age volume overestimation and 10.10% for the maximum is a subset of volume overestimation. Because IGTV 2Phases IGTV , the volumetric overestimation for IGTV AllPhases 2Phases compared to the reference IGTV was always equal to zero. Figure 4 illustrates the proportional volumetric underesti- mations (Fig. 4a) and overestimations (Fig. 4b) in the 17 individual patients with stage I disease. We found that vol- umetric underestimation was > 10% using either IGTV MIP or IGTV in 15 patients, but in no patients when 2Phases Computa overestimatio the the Figure 3 dashed l solid li tion of ne ine) compared wi ) n area the under (light gray) of the test estim th are tion a ference area (area rea (da arr ek gr a (ar ay) ea i and the in nsi side de IGTV was used. Volumetric underestimation > MIP-Modified Computation of the underestimation area (dark 20% occurred in 5 patients using the IGTV and in 7 MIP gray) and the overestimation area (light gray) of the patients using the IGTV . Of the 5 patients in whom 2Phases test area (area inside the dashed line) compared with volumetric underestimation was > 20% using IGTV , 2 MIP reference area (area inside the solid line). The areas had lesions near or attached to the diaphragm, 1 had a were computed using the Delaunay triangulation which is lesion near or attached to the chest wall, and another had shown in the regions of interest. a lesion near or attached to the mediastinum. Figure 5 illustrates the volumetric underestimations (Fig. 5a) and Table 4 shows the MI values for each IGTV based on the overestimations (Fig. 5b) in the 10 patients with stage III test volumes and on the all phases volume for patients disease. We found that volumetric underestimation was > with stage III disease. In general, we found that the 5% in 9 patients using IGTV , 8 patients using MIP GTV -based IGTV (mean ± SD: 0.93 ± 0.20) IGTV , and 2 patients using IGTV . Volu- MIP-Modified 2Phases MIP-Modified matched the GTV -based IGTV the closest, followed metric underestimation > 10% occurred in 6 patients AllPhases by the IGTVs based on GTV (mean ± SD: 0.91 ± using IGTV , 1 patient using IGTV , but no patients 2Phases MIP 2Phases 0.05) and GTV (mean ± SD: 0.86± 0.07). There was a using IGTV . In general, we found that the lowest MIP MIP-Modified significant difference between GTV -based and volumetric underestimation was achieved consistently 2Phases GTV -based IGTVs (p = 0.05) and between GTV - using the modified MIP approach to delineate the IGTV. MIP MIP based and GTV -based IGTVs (p = 0.03). MIP-Modified To analyze the accuracy of these contouring approaches in The volumetric underestimation and overestimation involved lymph nodes, we conducted the second analysis between the all phases volume and the test volumes for of involved lymph nodes in above stage III disease. Our patients with stage I and III disease are shown in Table 5. data showed that IGTV volume of lymph nodes MIP-Modified For stage I disease, the maximum volumetric underesti- (mean ± SD: 32.95 ± 40.86 cm ) matched most closely mations for IGTV , IGTV , and IGTV with IGTV volumes of lymph nodes (mean ± SD: MIP 2Phases MIP-Modified AllPhases compared to IGTV were 30.86%, 21.2%, and 34.26 ± 42.56 cm , p = 0.24), while IGTV and AllPhases 2Phases 8.53%, respectively. For stage III disease, the maximum IGTV lymph node volumes (mean ± SD: 29.15 ± 38.14 MIP volumetric underestimations for IGTV , IGTV , and and 25.63 ± 34.55 cm respectively) differed significantly MIP 2Phases IGTV compared to IGTV were 23.85%, with IGTV lymph node volume (p = 0.04 and 0.05 MIP-Modified AllPhases AllPhases 22.25%, and 6.66%, respectively. The average volumetric respectively, volume underestimation in all cases). In underestimation was 17.3% for IGTV , 19.3% for addition, the match index of lymph node IGTV MIP MIP-Modified IGTV , and 5.3% for IGTV in stage I tumors was not significantly different from IGTV (p = 0.14) 2Phases MIP-Modified 2Phases and 12.1% for IGTV , 8.9% for IGTV , and 4.2% for but was significantly different from IGTV values (p = MIP 2Phases MIP Page 6 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 1: SI motion and IGTVs based on the test volumes (IGTV , IGTV , and IGTV ) and the reference volume MIP 2Phases MIP-Modified (IGTV ) for stage I tumors AllPhases Patient No SI Motion (cm) IGTV IGTV IGTV IGTV MIP MIP-Modified AllPhases 2Phases 3 3 3 3 (cm ) (cm ) (cm ) (cm ) 1 1.06 5.82 7.98 8.32 6.56 2 1.37 8.53 10.51 9.98 7.49 3 1.70 12.88 15.85 16.54 13.23 4 1.08 4.92 5.49 5.44 3.72 5 0.15 1.64 1.79 1.80 1.46 6 2.17 17.64 22.16 23.39 18.98 7 1.27 23.08 26.06 26.28 21.77 8 0.54 12.45 15.73 15.76 12.82 9 0.18 21.80 24.50 24.96 20.92 10 0.00 60.04 66.90 67.75 63.67 11 0.41 2.46 2.80 2.85 2.27 12 1.77 32.90 37.65 39.39 33.14 13 0.14 2.08 2.27 2.23 1.84 14 0.10 1.53 1.90 1.93 1.74 15 1.62 18.59 21.26 21.69 16.66 16 0.66 10.70 11.31 10.60 7.77 17 0.09 3.35 3.45 3.33 2.70 0.001 for both cases). IGTV and IGTV not been compared with that of ten-phase contouring MIP-Modified 2Phases matched better with IGTV (match index mean ± approach particularly in stage III disease. Jin et al, in a AllPhases SD: 0.81 ± 0.08, range: 0.75–0.91 for IGTV , and phantom study, examined the feasibility of a method to MIP-Modified 0.77 ± 0.08, range: 0.65 to 0.88 for IGTV ) compared determine ITV based on motion information obtained 2Phases with IGTV (mean ± SD: 0.62 ± 0.11; Range: 0.46 to from select phases of a respiratory cycle [15]. They MIP 0.76). reported that adequate estimation of IGTV could in gen- eral be achieved by combining motion information from Discussion the extremes of motion in most cases and in some cases by Real-time tumor motion tracking provides most compre- the addition of motion information from an intermediate hensive data for respiratory tumor motion management. phase. Underberg et al. [8] reported that MIP-based con- However, it is a challenging technique to implement in touring could provide reliable margins for determining the clinical setting and more research is needed to make the IGTV for stage I lung tumors treated with SBRT. How- its clinical implementation more practical [14]. Although ever, their method did not include visual verification of both MIP-based and two-phase-based approaches have the MIP-defined GTV contour through each individual been shown to more accurately delineate the GTV than phase of the 4D CT (IGTV ). Bradley et al. [9] MIP-Modified conventional 3D CT-based planning, their accuracy has compared helical-, MIP-, and average-intensity (AI)-based Page 7 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 2: Matching index values for each IGTV based on IGTV , IGTV , and IGTV relative to the reference IGTV MIP 2Phases MIP-Modified AllPhases in stage I disease Patient No Location (Adjacent) IGTV IGTV IGTV MIP 2Phases MIP-Modified 1 Diaphragm 0.69 0.79 0.88 2 Diaphragm 0.79 0.75 0.88 3 Diaphragm 0.77 0.80 0.92 4 Chest wall 0.82 0.68 0.90 5 Lung parenchyma 0.83 0.81 0.88 6 Chest wall 0.72 0.80 0.89 7 Lung parenchyma 0.84 0.83 0.91 8 Mediastinum 0.74 0.81 0.92 9 Chest wall 0.84 0.84 0.91 10 Chest wall 0.87 0.94 0.95 11 Lung parenchyma 0.79 0.80 0.90 12 Chest wall 0.83 0.84 0.93 13 Lung parenchyma 0.82 0.83 0.90 14 Lung parenchyma 0.75 0.90 0.91 15 Diaphragm 0.79 0.77 0.89 16 Lung parenchyma 0.85 0.73 0.88 17 Lung parenchyma 0.88 0.81 0.91 4-D CT imaging to find the optimal approach for deter- over all individual phases of the 4D CT yielded the best mining the patient-specific IGTV for SBRT for stage I lung estimate of IGTV. However, there the performance of this cancer. They found that the MIP-defined GTV was signifi- approach in the delineation of involved lymph nodes was cantly larger than the helical-defined and average CT- not separately addressed [11]. In daily clinical practice, defined GTVs. However, in their study, Bradley et al. did tumor contouring in stage III disease is more challenging not compare the GTV based on GTV with that based on than in stage I disease because of the larger tumor volume, MIP GTV , the optimal reference volume. Bradley et al. more complicated tumor shape, involvement of critical AllPhases [9] did not discuss their results in the context of tumor structures, and potential involvement of multiple lymph location in their study. In another study, Cai et al. [10] nodes in which tissue density is similar to that of the determined the IGTVs for six lung tumors using a simula- tumor. In addition, although the two-phase-based tion method based on dynamic magnetic resonance imag- approach has been used to delineate IGTVs in the clinical ing (dMRI) and MIPs. They found that MIP-based IGTVs setting, there is scant data on the accuracy of such two- were smaller than dMRI-based IGTVs. They concluded phase-based IGTVs in either stage I or stage III disease that because of the low temporal resolution and retrospec- [16]. Our study showed that both MIP-based and two- tive re-sorting, 4-D CT might not accurately depict the phase-based IGTVs underestimate the 10-phase-based excursion of a moving tumor. Recent data by Rietzel et al IGTV in both stage I and III disease including involved also support our observation that tumor delineation on lymph nodes, which can potentially result in marginal the MIP with subsequent visual verification of contours under-dosing, and that the IGTV consistently MIP-Modified Page 8 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 3: SI motion and IGTVs based on the test volumes (IGTV , IGTV , and IGTV ) and the reference volume MIP 2Phases MIP-Modified (IGTV ) for stage III tumors AllPhases Patient No SI Motion (cm) IGTV IGTV IGTV IGTV MIP MIP-Modified AllPhases 2Phases 3 3 3 3 (cm ) (cm ) (cm ) (cm ) 1 0.09 64.91 77.33 79.94 74.42 2 0.12 202.85 228.55 238.40 216.29 3 0.21 135.59 146.62 151.40 138.72 4 0.18 221.50 230.85 233.74 222.22 5 0.62 23.87 30.59 29.98 23.33 6 0.11 446.14 450.31 458.21 439.86 7 0.96 242.38 265.47 268.58 244.76 8 0.14 347.06 368.97 373.61 351.46 9 0.18 36.69 39.40 36.87 34.03 10 1.77 211.70 221.96 228.89 203.01 had the lowest percentages of volumetric underestima- display mobile structures if the adjacent structures have tion, which indicates that the IGTV approach is similar (or higher) densities, which is the case for lesions MIP-Modified the most accurate in delineating the IGTV. located near the mediastinum, diaphragm, liver, or chest wall; and (2) the physician may misinterpret the MIP For the MIP-based approach, several potential sources of images because of tumor border smearing. (3) The tumor uncertainty/error exist: (1) the MIP image may not fully spicula can not be visualized on the MIP projections due to smearing of the tumor edge. Indeed, our data show that Table 4: Matching index values for each IGTV based on IGTV , MIP the MI was poor and volumetric underestimation was IGTV , and IGTV relative to the reference 2Phases MIP-Modified high using the MIP-based approach to delineate IGTVs in IGTV in stage III disease AllPhases most of lesions near the mediastinum, diaphragm, liver, Patient No GTV GTV GTV and chest wall. Of these lesions, those closer to the dia- MIP 2Phases MIP-Modified phragm and liver had the lowest MI values, which could 1 0.76 0.93 0.92 have been due to the significant motion of the diaphragm and liver and the MIP image's inability to record differ- 2 0.83 0.91 0.92 ences between the lesion and the diaphragm and liver. We are currently developing software that excludes dia- 3 0.88 0.92 0.94 phragm and liver images in some breathing phases using cine CT images so that better tumor MIP images will be 4 0.92 0.95 0.95 preserved (data to be published). We should note that MIP images do not reflect the densities of tumors, lungs, 5 0.74 0.78 0.90 and other normal tissues accurately enough for dose cal- culation in treatment planning [17]. Thus, a free-breath- 6 0.95 0.96 0.95 ing CT image set, a 4-D scan of a single respiratory phase, or an average CT image set extracted from a 4-D CT data 7 0.87 0.91 0.94 set should be used for treatment planning and dose calcu- lation. This would be especially important in proton ther- 8 0.90 0.94 0.94 apy, which is more sensitive to tumor motion and changes 9 0.87 0.92 0.91 in tissue density. In a previous study on 4-D CT in proton therapy planning, we found that a MIP density override 10 0.86 0.89 0.90 Page 9 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 Table 5: Summary of the volumetric percentage underestimation and overestimation for each IGTV based on IGTV , IGTV , MIP 2Phases and IGTV relative to the reference IGTV . MIP-Modified AllPhases Underestimation (%) IGTV IGTV IGTV MIP 2Phases MIP-Modified Stage I patients Avg. ± SD 17.33 ± 6.56 19.32 ± 5.93 5.36 ± 1.71 Range 6.32–30.86 6.03–31.76 2.30–8.53 Stage III patients Avg. ± SD 12.11 ± 6.23 8.95 ± 5.15 4.21 ± 1.66 Range 4.04–23.85 4.01–22.25 1.20–6.66 Overestimation (%) IGTV IGTV IGTV MIP 2Phases MIP-Modified Stage I patients Avg. ± SD 3.23 ± 2.35 0 4.80 ± 2.39 Range 0.34–8.64 0 1.30–10.07 Stage III patients Avg. ± SD 2.36 ± 1.79 0 3.21 ± 2.22 Range 1.06–6.92 0 1.43–8.09 Average ± standard deviation and range are reported for stage I and stage III tumors. for tumor contouring in an average CT data set was the In clinical setting, it is common to prescribe the dose to optimal approach [18]. PTV which takes additionally clinical target volume (CTV) and set-up uncertainty into consideration. The volume- For the two-phase-based approach, tumor deformation underestimation will be reduced if PTV was used to com- between the two extreme phases of breathing and the pare above mentioned four approaches. We evaluated the curved motion pathway during each breathing cycle may effect of this underestimation on the PTV in a case with introduce uncertainty. In most cases, however, we found maximal underestimation of the IGTV in stage I disease. that the MI of the two-phase-based IGTV was slightly IGTV was expanded by 1.6 cm (0.8 cm for CTV, 0.3 cm to higher than that of MIP-based IGTV, which indicates that account for variability in the determination of motion most tumors moved in a generally straightforward SI extent and 0.5 cm for image guided patient setup). Analy- direction and that tumor deformation during breathing sis of volumetric underestimation of the PTV was carried was minimal. Particularly in stage III disease, we found out in the same manner as described for IGTV. Our results that the volumetric underestimation was generally lower showed that the volume underestimation reduced from for the two-phase-based IGTV than for the MIP-based 30.86%, 21.2%,8.53% in IGTV to 13.3%, 5.18% and IGTV. Therefore, if 4-D CT based IGTV is not 3.36% in PTV for IGTV , IGTV , IGTV MIP-Modified MIP 2Phases MIP-Modified available, the two-phase-based IGTV is a reasonable alter- respectively. In general, this improvement is more dra- native approach to take tumor motion into consideration matic in the lesions with the smaller size such as stage I although it is not optimal one. disease. However, when ablative dose is attempted in clin- ical setting but sparing critical structures is concerning Page 10 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 a. b. S Figure 4 tage I tumors Stage I tumors. (a) Volumetric underestimation for each IGTV based on IGTV , IGTV , and IGTV relative to MIP 2Phases MIP-Modified the reference IGTV . (b) Volumetric overestimation for each IGTV based on IGTV , IGTV , and IGTV AllPhases MIP 2Phases MIP-Modified relative to the IGTV . (Note: IGTV is a subset of IGTV , hence the volumetric overestimation for IGTV AllPhases 2Phases AllPhases 2Phases is always equal to zero.) Page 11 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 a. b. Stage II Figure 5 I tumors Stage III tumors. (a) Volumetric underestimation for each IGTV based on IGTV , IGTV , and IGTV relative MIP 2Phases MIP-Modified to the IGTV . (b) Volumetric overestimation for each IGTV based on IGTV , IGTV , and IGTV relative to AllPhases MIP 2Phases MIP-Modified the IGTV . (Note: IGTV is a subset of IGTV , hence the volumetric overestimation for IGTV is always AllPhases 2Phases AllPhases 2Phases equal to zero.) Page 12 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 port. We also thank the Department of Scientific Publications for its such as SBRT in stage I disease, we would accept compro- assistance in preparing this article. mised coverage for PTV but not for IGTV. Therefore, IGTV delineation accuracy is still crucial clinically. References 1. Curran W, Scott C, Langer C: Long term benefit is observed in As with other such comparative studies mentioned above, a phase III comparison of sequential vs concurrent chemo- inter or intra observer variability in the delineation of the radiation for patients with unresectable NSCLC: RTOG 9410 [abstract]. Proc Am Soc Clin Oncol 2003:S621a. GTV was not considered. The uncertainties introduced as 2. Chang JY, Dong L, Liu H, Starkschall G, Balter P, Mohan R, Liao Z, a result of the above could however be thought to be dif- Cox JD, Komaki R: Image-guided radiation therapy for non- small cell lung cancer. Journal of Thoracic Oncology 2008, ferent from those analyzed in this study, thereby requiring 3:177-186. a separate analysis that is beyond the scope of the current 3. Liu HH, Balter P, Tutt T, Choi B, Zhang J, Wang C, Chi M, Luo D, Pan report. T, Hunjan S, Starkschall G, Rosen I, Prado K, Liao Z, Komaki R, Cox JD, Mohan R, Dong L: Assessing respiration-induced tumor motion and internal target volume using 4DCT for radiation Conclusion therapy of lung cancer. Int J Radiat Oncol Biol Phys 2007, 68:531-540. We found that the MIP-based and two-phase-based 4. Donnelly ED, Parikh PJ, Lu W, Zhao T, Lechleiter K, Nystrom M, approaches to IGTV delineation significantly underesti- Hubenschmidt JP, Low DA, Bradley JD: Assessment of intrafrac- mated the IGTV in patients with stage I and stage III tion mediastinal and hilar lymph node movement and com- parison to lung tumor motion using four-dimensional CT. Int NSCLC. Due to the limitations of each approach, a signif- J Radiat Oncol Biol Phys 2007, 69:580-588. icant amount of the tumor volume could be missed in 5. Vedam SS, Keall PJ, Kini VR, Mostafavi H, Shukla HP, Mohan R: individual patient so precautions should be taken when Acquiring a four-dimensional computed tomography data- set using an external respiratory signal. Phys Med Biol 2003, these techniques are used to treat patients. We also found 48:45-62. that the IGTV approach, which requires visual MIP-Modified 6. Keall PJ, Starkschall G, Shukla H, Forster KM, Ortiz V, Stevens CW, Vedam SS, George R, Guerrero T, Mohan R: Acquiring 4D tho- verification of tumor coverage after each phase of the racic CT scans using a multislice helical method. Phys Med Biol breathing cycle, improved IGTV delineation in both cases. 2004, 49:2053-2067. 7. Keall PJ, Mageras GS, Balter JM, Emery RS, Forster KM, Juang SB, Kap- atoes JM, Low DA, Murphy MJ, Murray BR, Ramsey CR, Van Herk MB, Abbreviations Vedam SS, Wong JW, York E: The management of respiratory GTV: gross tumor volume; IGTV: internal gross tumor vol- motion in radiation oncology report of AAPM Task Group ume; CTV: Clinical target volume; PTV: Planning target 76. Med Phys 2006, 33:3874-3900. 8. Underberg RW, Lagerwaard FJ, Slotman BJ, Cuijpers JP, Senan S: Use volume; IGTV : the gross tumor volume (GTV) con- AllPhases of maximum intensity projections (MIP) for target volume tours from ten respiratory phases; IGTV : the GTV 2Phases generation in 4DCT scans for lung cancer. Int J Radiat Oncol Biol Phys 2005, 63:253-260. contours from two extreme respiratory phases (0% and 9. Bradley JD, Nofal AN, El Naqa IM, Lu W, Liu J, Hubenschmidt J, Low 50%); IGTV : the GTV contour using the maximum MIP DA, Drzymala RE, Khullar D: Comparison of helical, maximum intensity projection (MIP); IGTV : the GTV con- intensity projection (MIP), and averaged intensity (AI) 4D MIP-Modified CT imaging for stereotactic body radiation therapy (SBRT) tour using the MIP with modification based on visual ver- planning in lung cancer. Radiother Oncol 2006, 81:264-268. ification of contours in individual respiratory phase. 10. Cai J, Read PW, Baisden JM, Larner JM, Benedict SH, Sheng K: Esti- mation of error in maximal intensity projection-based inter- nal target volume of lung tumors: A simulation and Competing interests comparison study using dynamic magnetic resonance imag- The authors declare that they have no competing interests. ing. Int J Radiat Oncol Biol Phys 2007, 69:895-902. 11. Reitzel E, Liu A, Chen G, Choi N: Maximum-intensity volumes for fast contouring of lung tumors including respiratory Authors' contributions motion in 4DCT planning. Int J Radiat Oncol Biol Phys 2008, ME, SV, and JYC designed/conducted analysis and wrote 71(4):1245-1252. 12. Pan T, Lee TY, Rietzel E, Chen GT: 4-D CT imaging of a volume the manuscript. JYC was responsible for manuscript revi- influenced by respiratory motion on multi-slice CT. Med Phys sion and submission. PB, BC and GS were involved in 4- 2004, 31:333-340. 13. Barber CB, Dobkin DP, Huhdanpaa HT: The quickhull algorithm D CT simulation and treatment designed. DM was for convex hulls. ACM Transactions on Mathematical Software 1996, involved in data analysis. 22(4):469-483. 14. Keall PJ, Cattell H, Pokhrel D, Dieterich S, Wong KH, Murphy MJ, Vedam SS, Wijesooriya K, Mohan R: Geometric accuracy of a Authors' information real-time target tracking system with dynamic multileaf col- Dr. Chang is a recipient of the Research Scholar Award limator tracking system. Int J Radiat Oncol Biol Phys 2006, from the Radiological Society of North America and a 65:1579-1584. 15. Jin J, Ajlouni M, Chen Q, Yin FF, Movsas B: Lung radiotherapy: A Development Award from The University of Texas M. D. technique of using gated-CT images to determine internal Anderson Cancer Center NIH Lung Cancer SPORE (P50 target volume (ITV) for fractionated stereotactic lung radi- otherapy. Radiotherapy and Oncology 2006, 78:177-184. CA70907). 16. Sarrut D, Boldea V, Miguet S, Ginestet C: Simulation of four- dimensional CT images from deformable registration between inhale and exhale breath-hold CT scans. Med Phys Acknowledgements 2006, 33:605-617. We thank all members of the Thoracic Radiation Oncology team, the 17. Admiraal MA, Schuring D, Hurkmans CW: Dose calculations attending physicians who enrolled patients in this study, and the dosime- accounting for breathing motion in stereotactic lung radio- trists who were involved in the patients' treatment for their help and sup- Page 13 of 14 (page number not for citation purposes) Radiation Oncology 2009, 4:4 http://www.ro-journal.com/content/4/1/4 therapy based on 4D-CT and the internal target volume. Radiother Oncol 2008, 86:55-60. 18. Kang Y, Zhang X, Chang JY, Wang H, Wei X, Liao Z, Komaki R, Cox JD, Balter PA, Liu H, Zhu XR, Mohan R, Dong L: 4D proton treat- ment planning strategy for mobile lung tumors. Int J Radiat Oncol Biol Phys 2007, 67:906-914. Publish with Bio Med Central and every scientist can read your work free of charge "BioMed Central will be the most significant development for disseminating the results of biomedical researc h in our lifetime." Sir Paul Nurse, Cancer Research UK Your research papers will be: available free of charge to the entire biomedical community peer reviewed and published immediately upon acceptance cited in PubMed and archived on PubMed Central yours — you keep the copyright BioMedcentral Submit your manuscript here: http://www.biomedcentral.com/info/publishing_adv.asp Page 14 of 14 (page number not for citation purposes)

Journal

Radiation OncologySpringer Journals

Published: Jan 27, 2009

References