Radiotherapy and Oncology 112 (2014) 308–313

Contents lists available at ScienceDirect

Radiotherapy and Oncology journal homepage: www.thegreenjournal.com

Retroperitoneal sarcoma

Spatial and volumetric changes of retroperitoneal sarcomas during pre-operative radiotherapy Philip Wong a,b, Colleen Dickie a,b, David Lee a,b, Peter Chung a,b, Brian O’Sullivan a,b, Daniel Letourneau a,b, Wei Xu c, Carol Swallow d, Rebecca Gladdy d, Charles Catton a,b,⇑ a d

Radiation Medicine Program, Princess Margaret Cancer Centre; b Department of Radiation Oncology; c Department of Biostatistics, Princess Margaret Cancer Centre; and Department of Surgical Oncology, Mount Sinai and Princess Margaret Cancer Centre, University of Toronto, Canada

a r t i c l e

i n f o

Article history: Received 7 March 2014 Received in revised form 8 July 2014 Accepted 2 August 2014 Available online 20 August 2014 Keywords: Retroperitoneal Sarcoma Pre-operative Radiotherapy Volume Displacement

a b s t r a c t Purpose: To determine the positional and volumetric changes of retroperitoneal sarcomas (RPS) during pre-operative external beam radiotherapy (PreRT). Material and methods: After excluding 2 patients who received chemotherapy prior to PreRT and 15 RPS that were larger than the field-of-view of cone-beam CT (CBCT), the positional and volumetric changes of RPS throughout PreRT were characterized in 19 patients treated with IMRT using CBCT image guidance. Analysis was performed on 118 CBCT images representing one image per week of those acquired daily during treatment. Intra-fraction breathing motions of the gross tumor volume (GTV) and kidneys were measured in 22 RPS patients simulated using 4D-CT. Fifteen other patients were excluded whose tumors were incompletely imaged on CBCT or who received pre-RT chemotherapy. Results: A GTV volumetric increase (mean: 6.6%, p = 0.035) during the first 2 weeks (CBCT1 vs. CBCT2) of treatment was followed by GTV volumetric decrease (mean: 4%, p = 0.009) by completion of radiotherapy (CBCT1 vs. CBCT6). Internal margins of 8.6, 15 and 15 mm in the lateral, anterior/posterior and superior/ inferior directions would be required to account for inter-fraction displacements. The extent of GTV respiratory motion was significantly (p < 0.0001) correlated with more superiorly positioned tumors. Conclusion: Inter-fraction CBCT provides important volumetric and positional information of RPS which may improve PreRT quality and prompt re-planning. Planning target volume may be reduced using online soft-tissue matching to account for interfractional displacements of GTVs. Important breathing motion occurred in superiorly placed RPS supporting the utility of 4D-CT planning. Ó 2014 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 112 (2014) 308–313

Retroperitoneal sarcomas (RPS) are rare cancers that comprise 10–15% of soft-tissue sarcomas (STS) [1]. Unlike STS from the extremities in which approximately 90% local control is achieved largely through wide excision ± adjuvant radiotherapy (RT), the 5-year local control rate in RPS from surgical series ranges from 30 to 50% [2] and locoregional abdominal failure is the principle cause of death [3–11]. Efforts to improve local control in RPS have included the use of pre-operative RT (PreRT) [12–14]. This strategy produced 5 and 10-year relapse free survival rates of 69% and 63% respectively, in a Princess Margaret Cancer Centre (PMH)/Mt-Sinai Hospital Phase I/II study and has been adopted as standard practice at the PMH [15]. The European Organization for Research and Treatment of Cancer (EORTC) is currently leading an ongoing phase III randomized control trial (RCT) (EORTC 66092-22092) that stud-

⇑ Corresponding author at: Radiation Medicine Program, Princess Margaret Cancer Centre, 5th Floor, 610 University Ave., Toronto, ON M5G 2M9, Canada. E-mail address: [email protected] (C. Catton). http://dx.doi.org/10.1016/j.radonc.2014.08.004 0167-8140/Ó 2014 Elsevier Ireland Ltd. All rights reserved.

ies the addition of PreRT to surgery in the management of primary resectable RPS. The advantages of delivering PreRT for RPS include displacement of normal tissues and organs outside the high-dose volume. Post-operatively, radiosensitive normal organs fill the surgical cavity and render RT difficult to tolerate. Moreover, delineating the clinical target volume (CTV) is more accurate in the presence of the gross tumor volume (GTV) in the pre-operative setting than post-operatively where lack of a GTV and surgical disruption add uncertainty. A potential criticism of PreRT is the risk that RPS continues to grow during PreRT, which may render the tumor unresectable. While prior studies have described the range of interfractional and intrafractional motions of abdominal organs and structures [16–18], there are no prior publications to describe RPS motions, which may differ from other abdominal tumors and structures due to RPS’s large sizes and lack of organ confinement. Since 2007, PMH implemented the use of daily cone-beam CT (CBCT) to guide the delivery of intensity modulated radiotherapy (IMRT) for PreRT of RPS. Since 2010, 4D-CT simulation of RPS has

309

P. Wong et al. / Radiotherapy and Oncology 112 (2014) 308–313

been used to incorporate potential breathing motion of the target into the planning target volumes. The current study aims to characterize the volumetric and spatial changes of RPS during the course of radiotherapy by using the information gathered through CBCT imaging and 4D-CT simulations of RPS.

right–left) and volumes): 23.5  13.3 cm (2883 cc), 27.9  21.0 cm (5484 cc), 31.0  15.7 cm (6809 cc), AND 33.7  26.0 cm (11,456 cc) that were incompletely encompassed by the field-ofview of the 4D-CTs, which required the imaging of the diaphragmatic dome, and were excluded from analysis; these were also excluded from the IGRT cohort. The remaining 22 patients formed the ‘‘4D-CT cohort’’ and were used to characterize the intra-fractional breathing motion of RPS and kidneys. Table 1 summarizes the characteristics of the patients.

Materials and methods Patients Approval from the PMH Research Ethics Board was obtained. From 2007 to 2012, 56 patients with RPS were treated with PreRT using IMRT and daily CBCT (Elekta XVI, Stockholm, Sweden) softtissue image guided radiotherapy (IGRT). Patients with primary (n = 52) or recurrent (n = 4) RPS of any histology with no prior exposure to RT were included. Patients initially judged by the consulting surgical oncologists to have resectable, borderline resectable, or unresectable disease were included. There were 2 cohorts of patients included in this study. From 2007 to 2010, 36 patients underwent PreRT and daily image guidance with CBCT. Seventeen patients were excluded from analysis due to chemotherapy prior to RT (n = 2) or tumor larger than CBCT field-of-view (n = 15) which represent a conical volume with a maximum diameter of 41 cm (axial direction) and 26 cm width (superior–inferior direction). The remaining 19 patients formed the ‘‘IGRT cohort’’ in which GTV expansion, contraction and inter-fractional motion during RT were characterized. Of these 19 patients, 3 had been deemed unresectable prior to RT, 15 resectable, and 1 borderline. From 2010 to 2012, 26 patients were simulated using 4D-CT simulation (Philips Brilliance Large Bore, Amsterdam, Netherlands). Four patients had GTVs (maximum dimensions (sup-inf;

Patient positioning and simulation Patients were positioned supine with their arms above the head and immobilized using an evacuated cushion and leg immobilizers. Helical CT scans were used for planning from 2007 to 2010. From 2010, CT simulation was acquired with 4D-bellows around the abdomen. Following the helical scan acquisition, 4D-CT free breathing acquisition was performed. The 4D-CT dataset at maximum inhale and exhale phases was used for planning. The estimated additional irradiation dose from the 4D-CT simulation is estimated to be 4 times that of a helical CT scan [19].

Planning and treatment All patients treated for RPS pre-operatively were planned using Pinnacle3 (Philips Healthcare, Amsterdam, Netherlands). Prior to 2010, planning was done using the following targets and structures contoured by radiation oncologists: GTV, CTV, individual kidneys, and specific organs at risk relative to the local anatomy. Since 2010, patients were simulated using 4D-CT scans in which targets and critical structures were contoured at maximum inspiration and expiration. The combined contours from inspiration and

Table 1 Characteristics of patients in the ‘‘IGRT cohort’’ and ‘‘4D-CT cohort’’. 2010

2007

2012

IMRT + CBCT

4D-CT simulaon

IGRT cohort N=19

Excluded paents N=17 (2: chemotherapy before RT) (15: GTV beyond field of view)

4D-CT cohort N=22

2 overlap

Variables

IGRT cohort (n = 19)

4D-CT cohort (n = 22)

Included patients ‘‘IGRT’’ + ‘‘4D-CT’’ cohorts (n = 39) 2 overlaps

Excluded patients (n = 17)

All patients (n = 56)

Median age (range)

61 (31–88)

61 (38–88)

60 (31–88)

55 (34–81)

59 (31–88)

Gender F M

14 (74%) 5 (16%)

8 (36%) 14 (64%)

21 (54%) 18 (46%)

6 (35%) 11 (65%)

27 (48%) 29 (52%)

Tumor grade Low High N/A

2 (10%) 14 (74%) 3(16%)

4 (18%) 17 (77%) 1 (4%)

6 (15%) 29 (74%) 4 (19%)

10 (59%) 7 (41%) 0

16 (28%) 36 (64%) 4 (7%)

Histology Liposarcoma Leiomyosarcoma Sarcoma NOS/UPS Synovial sarcoma Malignant solitary fibrous tumor Malignant peripheral nerve sheath tumor

8 4 5 0 1 1

7 6 5 2 1 1

14 (36%) 10 (26%) 9 (23%) 2 (5%) 2 (5%) 2 (5%)

14 (82%) 3 (18%) 0 0 0 0

28 (50%) 13 (23%) 9 (16%) 2 (4%) 2 (4%) 2 (4%)

Median GTV (range) cc

417 (61–3516)

836 (57–7639)

506 (57–7639)

4828 (1088–11,456)

1196 (57–11,456)

Median RT dose (range)

50 Gy (14.4–50.4)

50.4 Gy (45–50.4)

50 Gy (14.4–50.4)

45 Gy (34.4–50)

50 Gy (14.4–50.4)

(42%) (21%) (26%) (0%) (5%) (5%)

(32%) (27%) (23%) (9%) (4%) (4%)

310

Radiotherapy in retroperitoneal sarcomas

expiration formed the final target volumes for treatment planning and delivery. Specifically, an internal target volume (ITV) was delineated consisting of the combination of the CTVs at inhalation and expiration. CTVs and ITVs were expanded by 5 mm to form the PTV. All radiotherapy was delivered with the primary tumor (GTV) in place and was given using 3–7 beam IMRT arrangements. Patients were prescribed 45–50.4 Gy to the ICRU point, given in 1.8–2 Gy daily fractions. Patient position was corrected daily for translational setup errors based on 3D-CBCT image guidance using co-registrations of bony and soft-tissue anatomy. An internal laboratory phantom study (unpublished data) estimates that the cumulative radiation dose from daily CBCT would not exceed 0.5 Gy over the course of sarcoma radiotherapy (45–50.4 Gy).

Volumetric changes and interfraction displacements Median time from CT-simulation to RT start was 10 days (range: 6–31), with the aim of complying with the Cancer Care Ontario guideline requiring that 80% of cases commence their treatments within 2 weeks from CT-simulation [20]. One CBCT per week was contoured and analyzed to quantify the inter-fractional volumetric and positional changes of the GTVs and kidneys. Volumes were contoured for each CBCT by 2 observers (PW, DL). CBCTs acquired from the same patient were contoured using the same window levels and width. The volumes and centroid locations of the volumes were calculated using the Pinnacle3 software and compared between the 2 observers to determine the inter-observer differences. There were no significant inter-observer differences in GTV and kidney volumes and locations (intraclass correlation coefficient P 0.992); thus, the contours from one observer (PW) were used for all further analysis. A total of 118 GTVs and 177 kidney volumes were analyzed on 118 CBCTs. Volumes of GTV and kidneys at each CBCT were normalized to their sizes at the time of simulation. Comparisons were made between the volumes of GTV and kidneys at the initial CBCT (CBCT1) with volumes from subsequent CBCTs (CBCT2–6) to maintain consistent imaging methodology. CBCT1 was taken from the first day of treatment. Student paired t-test analysis was performed with SAS version 9.3 (SAS institute, Cary, USA) to determine the significance of the change in volume over the course of RT. Resection margins were defined as R0 when all gross disease was resected en-bloc with negative histological margins, and R1 when all gross disease was resected en-bloc but with tumor cells present at or within 1 mm of the resection margin. Contours from the volumetric analysis were used to determine the centroid of the volumes at simulation and at each CBCT. The location of the centroids at CT-simulation and at subsequent CBCTs was compared to determine inter-fractional displacements in the right/left (R/L), anterior–posterior (A/P) and superior–inferior (S/ I) directions. The Van Herk formula (Margin = 2.5R + 0.7r) [21] was used to estimate the internal margin necessary to ensure a minimum dose to the CTV of 95% for 90% of the patients.

Results Patient characteristics Between 2007 and 2012, 56 adult patients with RPS were treated with PreRT. The characteristics of these patients are listed in Table 1. Exclusion from the present study was most frequently due to large tumor size that was beyond the CBCT or 4D-CT field of views: the average volume of RPS of excluded cases (n = 17, median volume 4828 cc) was significantly greater than the ones analyzed (n = 19, median volume 417 cc, p < 0.0001 vs. excluded cases). In addition, excluded patients more often had low-grade tumors (59% vs. 10%, v2 test p = 0.003) and liposarcoma as the histology (82% vs. 42%, v2 test p < 0.0001). Patient age, gender and radiotherapy dose were not significantly different between analyzed and excluded patients (Table 1).

Volumetric change Initial GTV expansion (mean volume increase: 6.6%, Standard deviation (SD): 9%, p = 0.035) during the first 2 weeks of treatment (CBCT2 vs. CBCT1) was followed by GTV regression (mean volume decrease: 4%, SD: 7%, p = 0.009) by treatment completion (CBCT6 vs. CBCT1) (Fig. 1). Change in GTV volume at any timepoint (CBCT2–6 compared to CBCT1) during RT ranged from +76% to 21%, with 1 patient with GTV expansion re-CT simulated and re-planned to improve coverage. Kidney volumes from the ipsilateral and contralateral side of the GTV were analyzed in the same fashion as the GTVs. There were no significant volumetric changes over the course of radiotherapy for the kidneys. By the end of radiotherapy, 14 of 19 GTVs were smaller than the initial PreRT volume (Table 2). All 5 patients whose GTVs enlarged during PreRT underwent resection (3 R0 and 2 R1 resections). Of the 14 patients whose GTVs contracted during radiotherapy, 10 underwent resection (5 R0 and 5 R1). Four of the 14 patients did not undergo surgery due to poor performance status (n = 1), development of distant metastasis during PreRT (n = 1) or initially unresectable disease that did not respond sufficiently to permit resection following RT (n = 2).

Interfractional displacement Following bone co-registration, RPS centroids were displaced by a mean of 1.2 (SD: 4.8) R/L, 2.7 (SD: 5.5) A/P and 2.2 mm (SD: 7.1)

Intra-fractional breathing motion analysis Breathing motion of RPS was measured from 4D-CT simulation images acquired in the ‘‘4DCT cohort’’ of 22 patients. Centroid positions from contours made at the time of treatment planning were used to determine the position of the targets and normal organs at maximum inspiration and expiration. All centroid positions were normalized to the position of the most anterior and superior point of the L3 vertebral body at the L2/3 junction. The amount of GTV respiratory motion in each direction was analyzed to determine whether correlations (Pearson’s) exist between breathing motion and tumor size, tumor location, patient height and weight.

Fig. 1. Mean tumor (GTV) volumes over the course of radiotherapy (RT). GTVs volumes were normally distributed so Student t-test was used for analysis. Volumes at each week of RT (CBCT1–6) were normalized to the volume at the time of CTsimulation. Volumetric comparisons were made between the volumes at CBCT1 vs. subsequent CBCTs (CBCT2–6) for consistency in imaging modality.

311

P. Wong et al. / Radiotherapy and Oncology 112 (2014) 308–313 Table 2 Characteristics of the 19 patients in the ‘‘IGRT cohort’’ treated for which volumetric analysis were made. Variables

Tumors (n = 14) with decrease in volume

Tumors (n = 5) with increase in volume

Age (median) Gender (F:M)

66.5 10:4

56 5:0

Tumor grade Low High N/A

2 9 3

0 5 0

Histologya LPS LMS UPS/NOS M. SFT MPNST

6 2 4 1 1

2 2 1 0 0

Median GTV (range) at simulation

909 (61–3516) cc

267 (70–3097) cc

Median RT dose

50 Gy

50 Gy

Resection Yes No

10 4

5 0

R0 (Clear) R1 (Micro. Pos. Margins) No resection

5 5 4

3 2 0

tion of the GTV relative to maximum inspiration and expiration was 0.060 mm R/L, 0.46 mm A/P and 0.069 mm S/I from the GTV position on helical scan suggesting that the motion was symmetrical in all directions. The extent of CTV and kidney intrafractional breathing motion was similar to the GTV (Table 3). Exploratory analyses revealed only the S/I location of the tumor to be associated with the magnitude of RPS S/I, A/P and radial breathing motion. Tumors located superior to the L2/3 junction moved significantly more than those located inferior to that point (t-test p = 0.0014). Overall, as the tumor centroid location approached the superior limits of the abdomen, breathing motion effects on RPS increased (R2 = 0.66, p < 0.0001) (Fig. 3). Compared to volumes contoured in one breathing phase, combining the GTVs and CTVs at maximum inspiration and expiration for treatment planning enlarged the GTV and CTV by 1.1 times each. This translates into mean volumetric increase in GTV and CTV of 85 cc and 163 cc respectively.

Discussion

a LPS – liposarcoma, LMS – leiomyosarcoma, UPS/NOS – undifferentiated pleomorphic sarcoma/Not otherwise specified, M. SFT – malignant solitary fibrous tumor, MPNST – malignant peripheral nerve sheath tumor.

Fig. 2. Inter-fraction displacements of GTV and kidneys from their position at simulation in the right/left (R/L), anterior/posterior (A/P) and superior/inferior (S/I) directions.

S/I from the simulated position (Fig. 2). The mean displacements of kidneys were similar of 0.4 (SD: 2.7) R/L, 1.6 (SD: 3.8) A/P and 2.9 mm (SD:7.6) S/I. Using the measured displacements and the Van Herk’s formula, internal margins of 8.6, 15 and 15 mm are needed in the R/L, A/P and S/I directions to ensure 95% coverage of the CTV in 90% of patients.

Intrafraction breathing motion The mean extent of GTV respiratory excursions was 0.66 mm R/ L, 1.7 mm A/P and 4.6 mm S/I (Table 3). The mean resultant posi-

The role of PreRT in RPS is being evaluated in EORTC 6609222092. Evidence from PreRT studies for extremity sarcomas shows that the STS could progress or regress on therapy [22–25]. Two studies of 51 and 25 patients respectively demonstrated volumetric changes in tumor size ranging from 85% to +285% when comparing cross-sectional images obtained prior to and after RT (50 Gy/25fx) [23,24]. However, radiographic tumor regression was not consistently correlated with increased histopathological signs of response [23–25] or improved clinical outcome [25]. To our knowledge there have been no publications characterizing volumetric changes of RPS during the course of PreRT. GTV changes during treatment may lead to over-treatment of the PTV, or under-coverage of the CTV, depending on whether the volume contracted or expanded during the course of radiotherapy. In either situation, a RT treatment re-planning would be required. In the current study, important RPS volumetric changes (maximum growth: 76%; maximum reduction: 21%) occurred during the course of pre-operative RT. This range in tumor contraction and expansion is narrower than the range described in the above-mentioned studies [23–25]. In the current study, RPS typically grew during the first 2 weeks of treatments before stabilizing and regressing by the end of the treatment (CBCT1 vs. CBCT6) (Fig. 1). The results advocate for the use of IGRT such as CBCT in which soft-tissues are imaged to monitor changes in GTV volume and permit re-planning as necessary to ensure proper target coverage. While the long-term clinical impact of RPS volumetric growth or reduction during RT is unclear, these events had no apparent effect on the feasibility or quality of surgery (Table 2). It should also be noted that of the 3 tumors initially deemed technically unresectable, none became resectable by the completion of RT. Taken together, these observations are consonant with the intent of RT to sterilize the microscopic disease at the periphery of the tumor without necessarily influencing tumor size or resectability. Patient age, gender, RT dose, RPS histology, grade and size were not correlated with volumetric changes observed during RT.

Table 3 Intra-fraction breathing motion of the tumor (GTV), clinical target volume (CTV) and kidneys in the ‘‘4D-CT cohort’’. R/L

A/P

S/I

Average GTV position from helical scan-simulation (mm) GTV motion extent in mm (SD) CTV motion extent in mm (SD)

0.060 0.71 (0.99) 0.46 (0.60)

0.46 1.9 (3.7) 1.6 (2.2)

0.069 4.9 (4.6) 3.4 (4.0)

Left kidney in mm (SD) Right kidney in mm (SD) Both kidneys in mm (SD)

0.59 (1.1) 0.3 (0.86) 0.3 (1.3)

2.2 (2.7) 2.4 (3.3) 2.4 (2.9)

7.2 (5.0) 7.0 (6.2) 7.1 (5.4)

312

Radiotherapy in retroperitoneal sarcomas

Fig. 3. Superior/inferior (S/I) breathing motion of the tumors (GTV) in relation to the S/I location of the GTV centroid. Y-axis plotted in log-scale. Regression analysis: R2 = 0.65, p < 0.0001 (Spearman’s and Pearson’s correlations). p < 0.02 correlations were also found between GTV S/I location and A/P and radial tumor breathing motion. No other correlations were found between tumor breathing motion in any direction with GTV R/L location, GTV A/P location, patient height and weight.

In addition to volumetric changes, inter- and intra-fractional breathing motion may affect the coverage of the target volumes during radiotherapy. Although IMRT in RPS may permit for more conformal treatment and improved sparing of normal tissues from high doses of irradiation [26], little information is available describing RPS motion during radiotherapy. Two prior studies on liver and gastric tumors exemplified the amount of inter-fractional displacement of abdominal structures relative to the skeleton. Dawson et al. used repeated CT scans of 8 patients on active breathing control to measure the remaining inter-fractional displacements of hepatic microcoils relative to the skeleton. The average residual inter-fraction displacement of the microcoils was 3.3, 3.2 and 6.6 mm in the R/L, A/P and S/I directions, respectively [16]. Using CT scans acquired during the radiotherapy course of 22 resected gastric cancer patients, Wysocka et al. observed that following bone co-registration, different retroperitoneal structures and organs had differing amount of residual inter-fractional displacements [17]. Median left kidney inter-fractional displacements were 1.3, 2.7 and 6.1 mm in the R/L, A/P and S/I directions. Pancreas (median displacements: 4.6, 3 and 5.7 mm) and celiac axis (median displacements: 1.4, 1.2 and 1.7 mm) moved less between fractions. Recently, Hallman et al. reviewed the intrafractional respiratory motion of pancreas and livers from 7 studies which included a total of 72 patients [18]. During breathing, mean pancreas and liver excursion ranged from 5 to 25 and 9.8 to 25 mm respectively. In the current study, the largest extent of inter-fractional displacements secondary to motion of RPS within the abdominal cavity beyond skeletal co-registration occurred in the S/I and A/P directions (Fig. 2). Using the Van Herk formula to estimate the internal margin, 8.6, 15 and 15 mm margins in the R/L, A/P and S/I directions would be needed in the absence of soft-tissue matching IGRT techniques in the treatment of RPS. The largest intra-fractional breathing motion of RPS occurred along the S/I axis (4.6 mm (SD: 4.4)). Implementation of 4D-planning increased the GTVs and CTVs by a mean of 85 and 163 cc (Table 3), which represented the mean target volume that would be under-covered in the absence of 4D-planning or normal tissue included in the high-dose region using 4D-planning. GTV centroid S/I location was significantly

(Pearson’s p < 0.0001) correlated with increased breathing motion (Fig. 3). Specifically, none of the RPS with centroids located below the L2/3 junction were displaced by more than 4 mm S/I during breathing. In contrast, RPS located more superiorly may move by up to 14 mm along the S/I axis with 1 patient treated using active breathing control due to large observed respiratory excursions. The average and range of inter- and intra-fractional respiratory motion of kidneys, GTVs and CTVs were within the lower range of the organ motions described in prior studies [16–18]. There are several possible explanations for the reduced inter- and intra-fractional displacements observed in our patient cohort. Firstly, the current study was based on patient treated pre-operatively for RPS, which may arise from structures and organs with different mobility within the retroperitoneum. Secondly, patients generally had large tumors (median size: 1207 cc) and were treated with the RPS in-situ which may obstruct organ inter- and intra-fractional motion. In the previously quoted publications, the tumors are either resected (gastric) or are smaller organ-confined tumors (liver and pancreas) and may be less likely to interfere with organ motions secondary to breathing. Finally, respiratory excursions of organs and targets were assessed in free-breathing, which may yield smaller excursions than measured by Wysocka et al. in which patients were assessed using breath hold techniques [17]. Our current study is limited by the sample size, which was inherent to the rarity of the disease and the exclusion of incompletely imaged tumors on CBCT and 4D-CT. Patients excluded often had low-grade liposarcomas, which may grow to a large size before becoming symptomatic. Correspondingly, these large low-grade liposarcomas excluded from the analysis likely grow and respond slower and to a lesser extent during radiotherapy than the studied RPS or the excluded high-grade RPS. Nevertheless, no correlation between tumor sizes and the amount of volumetric and spatial changes was found within the range of RPS analyzed (range: 57– 7639 cc). Thus, our observation from the analyzed patients may represent the worst-case scenarios. This study is the largest series to characterize RPS motion and volumetric changes during PreRT. A second observer and control structures (kidneys) were used to assess potential inter-observer differences in CBCT segmentation and demonstrated that the contouring was consistent between

P. Wong et al. / Radiotherapy and Oncology 112 (2014) 308–313

observers and the measurements are reproducible in RPS and normal structures. The current study aims to provide margin estimates for use in the PreRT of RPS and to describe the roles of CBCT IGRT and 4Dsimulation to optimize RT planning and treatments for individual patients. Although intrafraction motion of GTVs may be underestimated when measured using inspiration and expiration phases only [27], our study demonstrated important respiratory motion, which is seldom accounted for in the treatment of RPS. The importance of radiotherapy quality in improving patient outcome in head and neck cancer is known [28]. In an effort to determine the utility of RT in the management of RPS, high quality RT will be needed to ensure optimal target coverage to maximize the impact of PreRT on patient outcome. Firstly, our results suggest that important volumetric changes occur during PreRT and soft-tissue imaging may be required to determine the need for RT re-planning. The role and benefit in applying adaptive radiotherapy (ART) for RPS Pre-RT should be examined by establishing the optimal time points and threshold GTV changes for ART. The use of PreRT did not seem to interfere with surgical resectability or margin quality in this series of patients. Secondly, IGRT using soft-tissue matching may reduce the internal margin needed. Finally, implementation of 4D-CT simulation and planning is indicated with superiorly placed tumors to account for respiratory motion. In parallel with the ongoing EORTC 66092-22092 trial to define the role of PreRT in the management of RPS, this study describes findings that should improve the quality of RT when used in the treatment of RPS. Research funding This work was supported by funds from the CARO-Elekta Research Fellowship. Previous presentations Some of these data were presented at the 54th Annual Meeting of the American Society for Radiation Oncology, and 17th Annual Connective Tissue Oncology Society meetings. Conflicts of interest Philip Wong was the recipient of the CARO-Elekta Research Fellowship and the CIHR Excellence in Radiation Research for the 21st century scholarship. There are no other conflicts of interests from the other authors. References [1] Porter GA, Baxter NN, Pisters PW. Retroperitoneal sarcoma: a populationbased analysis of epidemiology, surgery, and radiotherapy. Cancer 2006;106:1610–6. [2] Cheifetz R, Catton CN, Kandel R, O’Sullivan B, Couture J, Swallow CJ. Recent progress in the management of retroperitoneal sarcoma. Sarcoma 2001;5:17–26. [3] Gronchi A, Lo Vullo S, Fiore M, Mussi C, Stacchiotti S, Collini P, et al. Aggressive surgical policies in a retrospectively reviewed single-institution case series of retroperitoneal soft tissue sarcoma patients. J Clin Oncol 2009;27:24–30. [4] Bonvalot S, Rivoire M, Castaing M, Stoeckle E, Le Cesne A, Blay JY, et al. Primary retroperitoneal sarcomas: a multivariate analysis of surgical factors associated with local control. J Clin Oncol 2009;27:31–7.

313

[5] Erzen D, Sencar M, Novak J. Retroperitoneal sarcoma: 25 years of experience with aggressive surgical treatment at the Institute of Oncology, Ljubljana. J Surg Oncol 2005;91:1–9. [6] Hassan I, Park SZ, Donohue JH, Nagorney DM, Kay PA, Nasciemento AG, et al. Operative management of primary retroperitoneal sarcomas: a reappraisal of an institutional experience. Ann Surg 2004;239:244–50. [7] Stojadinovic A, Yeh A, Brennan MF. Completely resected recurrent soft tissue sarcoma: primary anatomic site governs outcomes. J Am Coll Surg 2002;194:436–47. [8] Lewis JJ, Leung D, Woodruff JM, Brennan MF. Retroperitoneal soft-tissue sarcoma: analysis of 500 patients treated and followed at a single institution. Ann Surg 1998;228:355–65. [9] Singer S, Corson JM, Demetri GD, Healey EA, Marcus K, Eberlein TJ. Prognostic factors predictive of survival for truncal and retroperitoneal soft-tissue sarcoma. Ann Surg 1995;221:185–95. [10] Rosenberg SA, Tepper J, Glatstein E, Costa J, Baker A, Brennan M, et al. The treatment of soft-tissue sarcomas of the extremities: prospective randomized evaluations of (1) limb-sparing surgery plus radiation therapy compared with amputation and (2) the role of adjuvant chemotherapy. Ann Surg 1982;196:305–15. [11] O’Sullivan B, Davis AM, Turcotte R, Bell R, Catton C, Chabot P, et al. Preoperative versus postoperative radiotherapy in soft-tissue sarcoma of the limbs: a randomised trial. Lancet 2002;359:2235–41. [12] Sampath S, Hitchcock YJ, Shrieve DC, Randall RL, Schultheiss TE, Wong JY. Radiotherapy and extent of surgical resection in retroperitoneal soft-tissue sarcoma: multi-institutional analysis of 261 patients. J Surg Oncol 2010;101:345–50. [13] Paryani NN, Zlotecki RA, Swanson EL, Morris CG, Grobmyer SR, Hochwald SN, et al. Multimodality local therapy for retroperitoneal sarcoma. Int J Radiat Oncol Biol Phys 2012;82:1128–34. [14] Pawlik TM, Pisters PW, Mikula L, Feig BW, Hunt KK, Cormier JN, et al. Longterm results of two prospective trials of preoperative external beam radiotherapy for localized intermediate- or high-grade retroperitoneal soft tissue sarcoma. Ann Surg Oncol 2006;13:508–17. [15] Smith MJ, Ridgway PF, Catton CN, Cannell AJ, O’Sullivan B, Mikula LA, et al. Combined management of retroperitoneal sarcoma with dose intensification radiotherapy and resection: long-term results of a prospective trial. Radiother Oncol 2014. [16] Dawson LA, Brock KK, Kazanjian S, Fitch D, McGinn CJ, Lawrence TS, et al. The reproducibility of organ position using active breathing control (ABC) during liver radiotherapy. Int J Radiat Oncol Biol Phys 2001;51:1410–21. [17] Wysocka B, Kassam Z, Lockwood G, Brierley J, Dawson LA, Buckley CA, et al. Interfraction and respiratory organ motion during conformal radiotherapy in gastric cancer. Int J Radiat Oncol Biol Phys 2010;77:53–9. [18] Hallman JL, Mori S, Sharp GC, Lu HM, Hong TS, Chen GT. A four-dimensional computed tomography analysis of multiorgan abdominal motion. Int J Radiat Oncol Biol Phys 2012;83:435–41. [19] Mori S, Ko S, Ishii T, Nishizawa K. Effective doses in four-dimensional computed tomography for lung radiotherapy planning. Med Dosim 2009;34:87–90. [20] Ontario CC. Radiation Treatment Wait Times 2013. [21] van Herk M, Remeijer P, Rasch C, Lebesque JV. The probability of correct target dosage: dose-population histograms for deriving treatment margins in radiotherapy. Int J Radiat Oncol Biol Phys 2000;47:1121–35. [22] Betgen AHR, Sonke J. One-beam CT guidance for set-up verification in extremity soft tissue sarcoma patients. Miami: CTOS; 2009. [23] Canter RJ, Martinez SR, Tamurian RM, Wilton M, Li CS, Ryu J, et al. Radiographic and histologic response to neoadjuvant radiotherapy in patients with soft tissue sarcoma. Ann Surg Oncol 2010;17:2578–84. [24] Roberge D, Skamene T, Nahal A, Turcotte RE, Powell T, Freeman C. Radiological and pathological response following pre-operative radiotherapy for soft-tissue sarcoma. Radiother Oncol 2010;97:404–7. [25] Miki Y, Ngan S, Clark JC, Akiyama T, Choong PF. The significance of size change of soft tissue sarcoma during preoperative radiotherapy. Eur J Surg Oncol 2010;36:678–83. [26] Paumier A, Le Pechoux C, Beaudre A, Negretti L, Ferreira I, Roberti E, et al. IMRT or conformal radiotherapy for adjuvant treatment of retroperitoneal sarcoma? Radiother Oncol 2011;99:73–8. [27] Ezhil M, Vedam S, Balter P, Choi B, Mirkovic D, Starkschall G, et al. Determination of patient-specific internal gross tumor volumes for lung cancer using four-dimensional computed tomography. Radiat oncol 2009;4:4. [28] Peters LJ, O’Sullivan B, Giralt J, Fitzgerald TJ, Trotti A, Bernier J, et al. Critical impact of radiotherapy protocol compliance and quality in the treatment of advanced head and neck cancer: results from TROG 02.02. J Clin Oncol 2010;28:2996–3001.