- Email: [email protected]
Impact of Gastric Filling on Radiation Dose Delivered to Gastroesophageal Junction Tumors
Int. J. Radiation Oncology Biol. Phys., Vol. 77, No. 1, pp. 292–300, 2010 Copyright Ó 2010 Elsevier Inc. Printed in the USA. All rights reserved 0360-3016/10/$–see front matter
doi:10.1016/j.ijrobp.2009.08.026
PHYSICS CONTRIBUTION
IMPACT OF GASTRIC FILLING ON RADIATION DOSE DELIVERED TO GASTROESOPHAGEAL JUNCTION TUMORS MYRIAM BOUCHARD, M.D.,* MARY FRANCES MCALEER, M.D., PH.D.,y AND GEORGE STARKSCHALL, PH.D.* Departments of *Radiation Physics and yRadiation Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, TX Purpose: This study examined the impact of gastric filling variation on target coverage of gastroesophageal junction (GEJ) tumors in three-dimensional conformal radiation therapy (3DCRT), intensity-modulated radiation therapy (IMRT), or IMRT with simultaneous integrated boost (IMRT-SIB) plans. Materials and Methods: Eight patients previously receiving radiation therapy for esophageal cancer had computed tomography (CT) datasets acquired with full stomach (FS) and empty stomach (ES). We generated treatment plans for 3DCRT, IMRT, or IMRT-SIB for each patient on the ES-CT and on the FS-CT datasets. The 3DCRT and IMRT plans were planned to 50.4 Gy to the clinical target volume (CTV), and the same for IMRTSIB plus 63.0 Gy to the gross tumor volume (GTV). Target coverage was evaluated using dose–volume histogram data for patient treatments simulated with ES-CT sets, assuming treatment on an FS for the entire course, and vice versa. Results: FS volumes were a mean of 3.3 (range, 1.7–7.5) times greater than ES volumes. The volume of the GTV receiving $50.4 Gy (V50.4Gy) was 100% in all situations. The planning GTV V63Gy became suboptimal when gastric filling varied, regardless of whether simulation was done on the ES-CT or the FS-CT set. Conclusions: Stomach filling has a negligible impact on prescribed dose delivered to the GEJ GTV, using either 3DCRT or IMRT planning. Thus, local relapses are not likely to be related to variations in gastric filling. Dose escalation for GEJ tumors with IMRT-SIB may require gastric filling monitoring. Ó 2010 Elsevier Inc. Gastroesophageal junction cancer, Gastric filling, Radiation therapy treatment planning, Intensity-modulated radiation therapy, Dose escalation.
at the level of the GEJ. To mitigate the effects of gastric filling, our institutional guidelines ask patients to be nil per os (NPO) 3 hours before simulation and each treatment. Compliance with this instruction is currently verified only at the simulation session, assuming the initial verbal instruction will carry over to subsequent treatment sessions. However, because patients may not be NPO for 3 hours before each treatment, it is possible that gastric filling may affect positioning of the target during treatments; in such cases, the effect of variation in gastric filling on RT dose delivery for distal esophageal tumors has not yet been well defined. The purpose of this study was to analyze the impact of variation of gastric filling on target coverage of GEJ tumors irradiated using three RT techniques: three-dimensional conformal RT (3DCRT), intensity-modulated RT (IMRT), and IMRT with simultaneous integrated boost (IMRT-SIB).
INTRODUCTION Since 1975, data have shown that the rates of esophageal adenocarcinomas have grown rapidly, by 463% and 335% in American men and women, respectively (1). Most adenocarcinomas involve the distal esophagus, close to or at the gastroesophageal junction (GEJ) (2). Most patients present with locoregionally advanced disease, and combination chemotherapy and radiation therapy (RT) is an important part of the standard treatment, either in the neoadjuvant or the definitive setting (3). Current techniques of RT for esophageal tumors include image guidance based on bony landmarks to help ensure that the radiation is delivered accurately. However, image registration based on bony landmarks does not necessarily correlate with the position of the esophageal tumor because of several factors. One such factor is gastric filling, especially Reprint requests to: George Starkschall, Ph.D., Department of Radiation Physics, Unit 94, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 3-2537; Fax: (713) 563-2479; E-mail: [email protected]; [email protected] Supported by Universite´ de Sherbrooke, Sherbrooke, Qc, Canada (M.B.). M. Bouchard is presently at Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke Qc Canada.
Accepted for presentation at the 51th meeting of the American Society for Therapeutic Radiation and Oncology in Chicago, Illinois, November 2009. Conflict of interest: G.S. receives research support from Philips Medical Systems. The authors report no other conflict of interest. Received June 22, 2009, and in revised form July 28, 2009. Accepted for publication Aug 10, 2009. 292
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
Table 1. Objectives and constraints for all plans
MATERIALS AND METHODS Patient selection Four-dimensional computed tomography (4DCT) image datasets from 8 patients previously treated with RT for esophageal cancer were analyzed. The datasets were acquired in accordance with an institutional review board—approved protocol (RCR03-0400). Of the 8 patients, 5 presented with a full stomach (FS) at the time of the initial 4DCT simulation. For that reason, the simulation was repeated with patients to be NPO for 3 hours before scanning (empty stomach [ES]). The 3 other patients were found to have had an ES during the simulation scan but an FS on repeat 4DCT scans acquired during treatment.
4DCT image selection The 4DCT image sets were acquired with a helical multislice CT scanner (Discovery PET/CT; GE Medical Systems, Waukesha, WI) using a cine mode protocol; after acquisition, images were reconstructed and binned into 10 phases over the respiratory cycle (4). To simplify the planning, we selected the 50% phase (end-expiration) of each dataset and registered the datasets from both the ES and FS scans of each patient according to the bony anatomy at the level of the GEJ (defined below). For each patient, the entire stomach was scanned, so that full and empty volumes could be compared. The 50% phase total lung volume was also determined to document breathing-induced differences that could potentially influence the GEJ tumor’s position. Because of 4DCT image-sorting factors, in certain cases the 4DCT 50% phase might not exactly represent the end-expiration anatomy, which might instead be better represented by images from the 40% or 60% phase; we therefore wanted to quantify any technology-induced bias that may have occurred in the position of the GEJ tumor.
Treatment planning Target volumes for each patient consisted of a gross tumor volume (GTV), defined as the GEJ plus a 2-cm cranial and a 2-cm caudal extension at that level. In our study, we opted for a uniform GTV definition to minimize the delineation uncertainty, knowing that a majority of tumors could have a similar shape at this level. The GEJ was defined as occurring at the last CT slice where the diaphragm’s left crus was seen clearly attached to the esophagus. The clinical target volume (CTV) included a 3-cm craniocaudal and a 1-cm radial expansion of the GTV, in addition to the distal paraesophageal, celiac, and left gastric lymph nodes. Finally, a 5-mm isotropic expansion was added to the CTV to define the planning target volume (PTV), assuming daily image guidance. A planning gross target volume (PGTV) was defined by adding a margin of 4 mm to the GTV for dose escalation planning using the IMRTSIB technique (5). The PGTV is a planning entity that is used in our practice to account for intrafractional motion and setup uncertainty of the GTV. The relationship of the PGTV to the GTV is analogous to that of the PTV to the CTV. The adjacent organs at risk (OARs), including the lungs, heart, spinal cord, liver, normal esophagus, and kidneys, were also contoured. For each patient, 3DCRT, IMRT, and IMRT-SIB plans were generated using the volumes contoured and dose calculated on the 50% phase of the ES 4DCT dataset (ES-CT). The same planning process was repeated on the 50% phase of each patient’s FS 4DCT dataset (FS-CT). The final dose prescription for the 3DCRT and IMRT plans was 50.4 Gy to the PTV, and for the IMRT-SIB plans it was 50.4 Gy to the PTV and 63.0 Gy to the PGTV.
293
Target /Organ at risk Target GTV PTV PGTV Organ at risk Lung
Heart Liver Normal esophagus Spinal cord Kidney (each)
Indicators
Objectives/Constraints
V50.4 (V63) V50.4 V50.4
100% 95% 95%
Mean dose V20 V10 V5 V50 V30
<18–20 Gy 35% 40% 60% 50% 30% or 30–40% with NTCP = 0 <32 Gy 70 Gy 50% 45 Gy 30%
Mean dose Maximum dose V50 Maximum dose V20
Abbreviations: GTV = gross tumor volume; Vn = fraction of volume receiving a dose of n Gy; PTV = planning target volume; PGTV = planning gross tumor volume, defined as the GTV plus a setup margin; NTCP = normal tissue complication probability. All plans were generated using a research version of a commercially available RT planning system (Pinnacle3, Philips Healthcare, Milpitas, CA). For 3DCRT plans, we used four 18-MV beams with the aperture shaped to the PTV plus a penumbra margin of 9–10 mm (sufficient PTV-to-field border margin to ensure a coverage similar to IMRT objectives), with individualized beam angle adjustment, beam weighting, wedges, and, in approximately half the cases, the addition of field-in-field for optimizing dose distributions. Whereas 18 MV was preferred to limit hot spots in the 3DCRT plans (most patients having >20 cm separation width), the choice of beam energy for IMRT was based on the class solutions used in our current practice. Both the IMRT and IMRT-SIB plans were generated using five or six 6-MV beams. Inverse planning was performed using the Pinnacle3 direct machine parameter optimization process (6) to meet the dose objectives and constraints presented in Table 1.
Treatment plan evaluation We analyzed the dosimetric impact of gastric filling by comparing dose–volume histograms (DVHs) that reproduced the clinical situations of patients having their treatments planned on an ES but with the entire RT treatment course delivered on an FS, and vice versa. Our results then would model a worst-case scenario that could occur if a patient was not compliant to the NPO instructions. Target volumes were contoured on FS-CT and ES-CT. To investigate the clinical scenario of planning on ES and treating on FS, we first computed the dose distributions on the ES-CT dataset using beam configurations based on the ES target volumes. Then, the plans were evaluated based on the DVHs calculated from the ES-CT target volumes and the FS-CT target volumes copied onto the ES-CT already registered with bony landmarks. The other scenario of planning on FS and treating on ES was evaluated by computing the dose distributions on the FS-CT dataset using beam configurations based on the FS target volumes. The plans were evaluated similarly, based on the DVHs calculated from the FS-CT target volumes and the ES-CT target volumes copied onto the FS-CT. The following metrics were calculated and compared between clinical scenarios for each patient: the volume of PTV (as
I. J. Radiation Oncology d Biology d Physics
294
Volume 77, Number 1, 2010
Fig. 1. Digitally reconstructed radiographs of (a) Patient 4, who had the greatest variation in stomach volume; yellow [y] = empty stomach; sky-blue [s] = full stomach, and (b) Patient 1, showing stomach filling at the first simulation session (full stomach, sky-blue [s] = 841.8 cm3) and at resimulation (empty stomach, yellow [y] = 232.1 cm3). The 3 other volumes are the stomach filling for the same patient during RT with usual nil per os instructions: week 1, lavender [l] = 835.5 cm3; week 4, violet [v] = 988.5 cm3; week 5, orange [o] = 278.62 cm3. a percentage) receiving $50.4 Gy (PTV V50.4), the CTV mean dose, the dose received by 99% of the CTV (CTV D99), and GTV V50.4 for all plans; in addition, the PGTV V63 was evaluated for the IMRTSIB plans. Means and standard deviations were obtained for these metrics as well. The impact of planning on FS-CT or on ES-CT volumes on OARs was evaluated by comparison with the dose–volume points representing the limits of acceptable toxicity for each OAR. These dose–volume limits corresponded to the constraints used for IMRT planning (Table 1) and have been part of our institutional practice. Statistical comparisons were made using the Student’s t-test, with a p value of #0.05 taken to indicate significant differences.
RESULTS The FS volumes were larger than the ES volumes by a mean factor of 3.3 (range, 1.7–7.5). Figure 1a illustrates the full and empty stomach of Patient 4, who had the greatest variation in stomach volume; likewise, Fig. 1b shows gastric filling variations for Patient 1 over the course of radiation treatment. Although Patient 1 was instructed to be NPO for 3 hours before each treatment, gastric filling varied as follows: week 1, 835.5 cm3; week 2, 461.4 cm3; week 4, 988.5 cm3; week 5, 278.6 cm3. Table 2 shows the empty and full stomach volumes for all patients studied.
Table 2. Stomach volumes (single measurements) for the 8 patients
Patient 1 2 3 4 5 6 7 8
Empty stomach volume (cm3)
Full stomach volume (cm3)
Increase in volume from empty to full (%)
214.9 258.4 272.3 160.3 230.5 275.7 370.9 226.5
841.8 696.9 799.7 1 362.0 1 103.7 1 302.0 1 000.6 953.9
292 170 194 750 379 372 170 321
To ascertain whether breathing could have an impact on the position of the GEJ tumor, we measured the lung volume for each 50% phase image set. The mean ( standard deviation) variation of the total lung volume between the 50% phase ES-CT and the 50% phase FS-CT for the 8 patients was 4.2 5.1%. Tables 3 and 4 present the dose coverage of the target volumes for the patients simulated with an ES but treated with an FS during the entire RT course. Table 3 shows the GTV, CTV, and PTV coverage for ES plans delivered on FS volumes; Table 4 details the differences when comparing the coverage of ES volumes with the original ES standard-dose treatment plans (3DCRT and IMRT) and coverage of FS volumes with the same plans. Whereas the ES CTV D99 and ES CTV mean dose were equal to or greater than the prescribed dose (with higher doses falling within clinically acceptable limits) on the initial ES 3DCRT and IMRT plans, delivering a whole course of treatment on FS volumes resulted in some underdosed areas on the CTV, with mean FS CTV D99 values of 47.7 2.9 Gy and 44.9 6.6 Gy, respectively. The mean dose to the FS CTV was greater than or equal to the prescribed dose for all plans. The mean FS PTV V50.4 value for 3DCRT ES plans was 86.1 3.8%, and that for IMRT plans was 88.4 6.0%. No statistical difference was observed between these results for the 3DCRT and IMRT techniques. The DVH curves from the 3DCRT and IMRT plans for a typical patient are given in Figs. 2 and 3, respectively. No direct correlation between the absolute or relative stomach volume increase and the lack of target coverage was observed. Tables 5 and 6 present the dose coverage of the target volumes for the patients simulated with an FS but treated with an ES during the entire RT course. Table 5 shows the GTV, CTV, and PTV coverage for FS plans delivered on ES volumes; Table 6 details the differences when comparing the coverage of FS volumes with the original FS standard-dose treatment plans (3DCRT and IMRT) and coverage of ES volumes with the same plans. Target coverage was adequate, with an ES GTV V50.4 value of 100% and a minimum ES
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
295
Table 3. Dose–volume results calculated for full stomach treatments using plans generated on empty stomach computed tomography datasets GTV V50.4 (%)
CTV D99 (Gy)*
PTV V50.4 (%)y
CTV Dmean (Gy)
PGTV V63 (%)
Patient
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
SIB
1 2 3 4 5 6 7 8 Mean SD
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
48.8 49.7 49.9 43.1 43.2 47.6 49.9 49.3 47.7 2.9
38.0 50.2 48.9 38.7 34.8 50.4 50.8 47.2 44.9 6.6
41.6 51.3 52.8 41.9 36.8 51.5 51.7 51.1 47.3 6.2
51.9 52.2 52.0 51.9 51.7 52.6 52.1 52.5 52.1 0.3
52.1 51.4 51.8 50.7 51.2 51.8 51.5 51.5 51.5 0.4
58.0 57.8 60.3 56.4 57.8 59.1 59.4 58.5 58.4 1.2
84.9 86.4 90.2 81.9 79.9 88.6 90.8 86.2 86.1 3.8
82.0 90.6 94.9 79.4 86.0 91.6 96.2 86.5 88.4 6.0
90.4 95.6 98.9 86.1 86.0 96.0 97.8 98.1 93.6 5.3
82.8 71.7 92.2 71.7 85.3 76.8 84.7 80.4 80.7 7.1
Abbreviations: GTV = gross tumor volume; Vn = fraction of volume receiving a dose of n Gy; CTV = clinical target volume; D99 = dose delivered to 99% of volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; PGTV = planning gross tumor volume, defined as the GTV plus a setup margin; 3DCRT = three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SIB = IMRT + simultaneous integrated boost; SD = standard deviation. * 3DCRT vs. IMRT, p = 0.14. y 3DCRT vs. IMRT, p = 0.11.
CTV D99 of 49.6 Gy, with most patients receiving at least the prescribed dose (and with higher doses meeting clinically acceptable limits) for 3DCRT and IMRT plans. The mean ES PTV V50.4 values were 95.5 2.5% and 96.2 4.1% for the 3DCRT and IMRT plans, respectively. Table 3 presents the FS PGTV V63 results for ES IMRTSIB plans and Table 5 shows the ES PGTV V63 results for FS IMRT-SIB plans for all patients. As shown in Table 5, any gastric filling variations resulted in a loss of PGTV cov-
erage. Figure 4 illustrates ES and FS PGTV as well as ES and FS GTV DVHs for ES IMRT-SIB plans for 2 patients. No major change in dose to OARs resulted from variation of target volume with gastric filling, except in 1 patient (Patient 4) who had the largest variation in stomach volume. In this case, there was a noticeable increase in the left kidney V20 from 3.6% to 7.0% with ES plans to 15.7% to 19.9% with FS plans, whereas the dose to the left kidney was within clinically acceptable limits on both plans. Table 7 shows the
Table 4. Dose–volume results for 3DCRT and IMRT plans made using a empty-stomach datasets with treatments delivered to empty and full stomachs CTV D99 (Gy)
CTV Dmean (Gy)
PTV V50.4 (%)
Treatment type
Patient
ES CTV
FS CTV
Diff
ES CTV
FS CTV
Diff
ES PTV
FS PTV
Diff
3DCRT
1 2 3 4 5 6 7 8 Mean SD 1 2 3 4 5 6 7 8 Mean SD
50.70 51.10 50.60 50.90 50.40 51.20 50.90 51.40 50.9 0.3 51.2 50.9 51.0 50.7 50.8 51.3 50.8 51.2 51.0 0.2
48.8 49.7 49.9 43.1 43.2 47.6 49.9 49.3 47.7 2.9 38.0 50.2 48.9 38.7 34.8 50.4 50.8 47.2 44.9 6.6
1.9 1.4 0.7 7.8 7.2 3.6 1.0 2.1 3.2 2.8 13.2 0.7 2.1 12.0 16.0 0.9 0.0 4.0 6.1 6.5
52.1 52.4 51.9 52.5 51.9 52.8 52.4 52.7 52.3 0.3 52.7 51.7 51.9 51.2 51.8 51.9 51.5 51.6 51.8 0.4
51.9 52.2 52.0 51.8 51.7 52.6 52.1 52.5 52.1 0.3 52.1 51.4 51.8 50.7 51.2 51.8 51.5 51.5 51.5 0.4
0.3 0.2 (0.1) 0.6 0.3 0.2 0.2 0.2 0.2 0.2 0.6 0.3 0.1 0.5 0.6 0.1 0.0 0.1 0.3 0.3
95.9 97.5 94.1 95.6 91.7 97.5 97.9 97.3 95.9 2.1 98.8 99.9 99.9 99.7 99.9 100.0 99.9 100.0 99.8 0.4
84.9 86.4 90.2 81.9 79.9 88.6 90.8 86.2 86.6 3.8 82.0 90.6 94.9 79.4 86.0 91.6 96.2 86.5 88.4 6.0
11.0 11.0 3.9 13.6 11.8 8.9 7.1 11.1 9.8 3.1 16.8 9.3 5.1 20.3 13.9 8.4 3.7 13.6 11.4 5.7
IMRT
Abbreviations: ES = empty stomach; FS = full stomach; CTV = clinical target volume; V50.4 = fraction of volume receiving a dose of 50.4 Gy; CTV = clinical target volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; 3DCRT, three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SD = standard deviation; Diff = absolute difference between ES-FS values, with negative values between ( ).
296
I. J. Radiation Oncology d Biology d Physics
Fig. 2. Patient 1 (average stomach volume variation), dose-volume histograms for 3-dimensional conformal radiation therapy (3DCRT) plans generated on an empty stomach (ES) and delivered to ES (top) and full stomach (FS, bottom).
DVH point-dose results for the ES and FS plans relative to recognized limits for OAR radiation toxicity risk.
DISCUSSION Our results indicate that stomach filling has a negligible impact on the standard prescribed dose delivered to GTVs located at the GEJ when either 3DCRT or IMRT planning is used. However, CTV evaluation needs careful interpretation. For plan evaluation in the general practice of RT, the geographic concept of the PTV is used as an indirect measure of CTV coverage, despite the limitations related to the fact that it is a static representation of the patient created from a single CT dataset. Using the margins defined by van Herk et al. (7) and van Herk (8), the PTV reflects a 90% probability that the CTV will receiving at least 95% of the prescribed dose. For all patients examined here, treating patients with FS during the entire GEJ tumor treatment course when the treatment was planned on an ES would have had no major impact on the CTV Dmean. Our study shows that an occasional variation in gastric filling (few fractions), even if sizeable, does not seem to affect the dose delivered to the target volumes during the course of standard
Volume 77, Number 1, 2010
Fig. 3. Patient 1 (average stomach volume variation), DVH results for IMRT plans generated on ES and delivered to ES and FS.
RT of the GEJ. A sustained increase in gastric volume compared with the volume at treatment simulation during most fractions of the treatment, however, could lead to inadequate dose to the PTV and a small CTV geographic miss, as indicated by our suboptimal PTV and CTV D99 results. We estimate that our worst-case scenario result of approximately 85% PTV coverage for a whole treatment course with an increased stomach volume could become clinically significant if, for example, more than half the RT fractions would be delivered to a patient having an FS, leading to inadequate CTV coverage. For this reason, patient NPO instructions should be emphasized throughout the patient’s treatment course. The adequacy of GTV coverage despite changes in gastric filling is an important finding, because it does not support the hypothesis that gastric filling variations explain local relapses of GEJ tumors. Many studies have reported that local treatment failure remains an important issue in esophageal cancer (9, 10). Furthermore, in a study of 66 patients in whom most tumors were located in the GEJ and were treated by definitive chemoradiation therapy using either IMRT or 3DCRT to 50.4 Gy, Settle et al. (11) showed that 24 patients were found to have experienced locoregional failure. The majority (70%) of the local failures occurred within the GTV, whereas only two occurred within the CTV and one in the PTV. By contrast, Button et al.(12) analyzed the
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
297
Table 5. Dose–volume results calculated for empty stomach treatments using plans generated on full stomach computed tomography datasets GTV V50.4 (%)
CTV D99 (Gy)
CTV Dmean (Gy)
PGTV V63 (%)
PTV V50.4 (%)
Patient
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
SIB
1 2 3 4 5 6 7 8 Mean SD
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
50.7 50.7 51.0 50.8 50.3 49.8 49.8 52.4 50.7 0.8
50.4 51.2 50.6 51.6 50.9 50.9 49.6 50.9 50.8 0.6
50.8 52.8 51.5 52.5 52.4 51.6 51.6 51.5 51.8 0.7
52.2 51.4 52.3 52.2 52.4 53.0 52.5 53.2 52.4 0.6
52.6 51.8 52.9 52.5 51.5 51.5 51.4 51.5 52.0 0.6
60.6 60.3 59.4 60.5 60.8 58.1 58.2 59.1 59.6 1.1
92.7 97.4 95.7 96.7 94.9 95.5 91.6 99.6 95.5 2.5
90.7 99.8 94.8 100.0 99.6 96.4 90.0 98.4 96.2 4.1
93.9 99.7 98.3 99.5 99.9 99.5 98.7 99.8 98.7 2.0
85.0 99.7 85.6 96.4 97.1 80.8 84.6 92.2 90.2 7.1
Abbreviations: GTV = gross tumor volume; Vn = fraction of volume receiving a dose of n Gy; CTV = clinical target volume; D99 = dose delivered to 99% of volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; PGTV = planning gross tumor volume, defined as the GTV plus a setup margin; 3DCRT = three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SIB = IMRT + simultaneous integrated boost; SD = standard deviation.
choice of margins for RT of esophageal cancer. They retrospectively analyzed patterns of disease recurrence, with local relapse compared with radiation field location in 145 patients with esophageal carcinoma who had received standard definitive chemoradiation therapy by 3DCRT techniques. The authors found that 49% of relapses were within the GTV at first presentation and that 96% of locoregional failures occurred within the radiation treatment fields, even if patients received adequate chemoradiation therapy. The same authors concluded that it would be impossible to
increase field size and thereby improve tumor control without simultaneously adding important toxicity (12). Inasmuch as a dose–response relationship has been suggested for esophageal tumors (13), and given that the current results suggest that gastric filling does not have an impact on the standard dose delivered to the GTV, the answer to better local control might be in devising ways to escalate the radiation dose to the GTV itself. Another important implication of our results is that to achieve adequate dose escalation with IMRT-SIB, direct
Table 6. Dose–volume results for 3DCRT and IMRT plans made using a full-stomach datasets with treatments delivered to empty and full stomachs CTV D99 (Gy)
CTV Dmean (Gy)
PTV V50.4 (%)
Treatment type
Patient
FS CTV
ES CTV
Diff
FS CTV
ES CTV
Diff
FS PTV
ES PTV
Diff
3DCRT
1 2 3 4 5 6 7 8 Mean SD 1 2 3 4 5 6 7 8 Mean SD
51.2 50.4 51.0 50.2 50.2 51.7 50.9 51.6 50.9 0.6 51.3 51.1 51.2 51.6 50.9 50.9 51.0 50.9 51.1 0.2
50.7 50.7 51.0 50.8 50.3 49.8 49.8 52.4 50.7 0.8 50.4 51.2 50.6 51.6 50.9 50.9 49.6 50.9 50.8 0.6
0.5 (0.3) 0.0 (0.6) (0.1) 1.9 1.1 (0.8) 0.2 0.9 0.9 (0.1) 0.6 0.0 0.0 0.0 1.4 0.0 0.4 0.6
52.4 51.4 52.3 52.0 52.7 53.2 52.6 53.2 52.5 0.6 52.7 51.8 51.9 52.5 51.5 51.5 51.5 51.5 51.9 0.5
52.2 51.4 52.3 52.1 52.4 53.0 52.5 53.2 52.4 0.5 52.6 51.8 51.9 52.5 51.5 51.5 51.4 51.5 51.8 0.5
0.2 0.0 0.0 (0.1) 0.3 0.2 0.2 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0
97.8 90.4 96.1 96.6 94 99.0 96.1 98.4 96.1 2.8 97.7 99.9 99.9 99.8 100.0 99.7 100.0 100.0 99.6 0.8
92.7 97.4 95.7 96.7 94.9 95.5 91.6 99.6 95.5 2.5 90.7 99.8 94.8 100.0 99.6 96.4 89.9 98.4 96.2 4.1
5.1 (7.0) 0.4 (0.1) (0.9) 3.5 4.5 (1.2) 0.5 3.9 7.0 0.1 5.1 (0.2) 0.4 0.3 10.0 1.2 3.0 3.9
IMRT
Abbreviations: ES = empty stomach; FS = full stomach; CTV = clinical target volume; V50.4 = fraction of volume receiving a dose of 50.4 Gy; CTV = clinical target volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; 3DCRT, three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SD = standard deviation; Diff = absolute difference between ES-FS values, with negative values between ( ).
298
I. J. Radiation Oncology d Biology d Physics
Fig. 4. Patient 1 (average stomach volume variation) and Patient 4 (greatest stomach volume variation) dose-volume histograms for an intensity-modulated radiotherapy with simultaneous integrated boost (IMRT-SIB) plan generated on and empty stomach (ES) and delivered to ES and full stomach (FS). The FS DVH represents the clinical equivalent of a patient whose plan was simulated using images taken on an ES but who received all radiation treatments with an FS.
visualization of the tumor during treatments, with appropriate image guidance techniques and/or close monitoring of gastric filling, seems necessary. Current image guidance for esophageal cancers is based on matching the bony anatomy without confirming the real tumor location. Our results clearly show that with either ES or FS, the intended dose escalation to the PGTV with IMRT-SIB would be inadequate if variations in gastric filling occur during the treatment course. Of note, radiation dose escalation in esophageal cancer is still controversial, especially given that the landmark INT 0123 study suggested worse outcomes in the patients assigned to the higher radiation dose arm than those in the 50.4-Gy standard arm (9). However, the increased incidence of treatment-related death in the high-dose arm occurred mostly in patients who did not receive the planned dose, with 7 of 11 patients receiving #50.4 Gy. Moreover, the increased radiation late effects could be explained by that study’s time frame, 1995–1999, which preceded the advent of current RT techniques and therefore might have encountered more treatment-related toxicities than would be seen
Volume 77, Number 1, 2010
today. Newer, more effective, systemic treatment combinations might be part of the solution for this problem as well, but we believe there are still opportunities to escalate radiation doses by using recent advances in RT and imaging techniques and that these opportunities should be tested in clinical trials. Our results showed that planning on FS volumes allows adequate CTV and PTV coverage for 3DCRT and IMRT plans, even with a reduction in gastric filling during the course of the radiation treatments. This result indicates that it is adequate to plan RT for GEJ cancer on an FS-CT and that there is no need to repeat the scan if the patient is found to have an FS at the simulation session. No major resulting dose increase to OARs was noticed in most such cases. However, when normal tissue constraints are not met, it may be beneficial to rescan the patient by means of ES-CT and then replan the treatment to optimize OAR sparing. To our knowledge, our study is the first to quantify the dosimetric impact of variations in gastric filling on the radiation dose delivered to a GEJ tumor. Target underdosing with changes in gastric filling was similar with either 3DCRT or IMRT. However, the small sample size of our study might have precluded finding a statistically significant difference between the two techniques. The larger standard deviation on FS CTV D99 for IMRT technique and its lowest individual FS CTV D99 result (34.8 Gy for IMRT vs. 43.1 for 3DCRT) might favor 3DCRT to prevent geographic miss. Nevertheless, the same precautions regarding NPO instructions should be followed regardless of which of these techniques is used, inasmuch as both planning techniques resulted in some cases where the worst-case scenario could be clinically significant. Future investigations are needed to assess the frequency of gastric filling increases to determine the risk of CTV underdosage during treatment with current NPO instructions. Then, the need for intervention by a dietician and/or for patient instructions on methods of minimizing gastric emptying delay could be proposed in certain cases. Up to now, only Watanabe et al. (14) have described interfractional variations of stomach position in patients with gastric lymphomas, which they detected using fluoroscopic visualization. In that study, variations were important even if the patients had an overnight fast before treatment. Variations in gastric filling may differ from the one experienced by GEJ cancer patients because, in contrast to a more localized GEJ tumor, diffuse gastric lymphoma can harden tissues that may limit gastric filling and at the same time delay emptying, which could then induce variable responses to NPO instructions. Furthermore, it is important to recognize that gastric filling might vary because of different underlying diseases or other factors, such as the use of opioids or enteral nutrition techniques, which should also be evaluated individually (15, 16). Gastric filling has an unknown effect on GEJ tumor position in patients with a sliding hiatal hernia, one of the most common abnormalities involving the esophagus. The uniform GTV definition we used for our study did not allow
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
299
Table 7. Dosevolume results relative to recognized limits for radiation toxicity for empty stomach (ES) and full stomach (FS) 3DCRT, IMRT, and IMRT-SIB plans Liver V30 (%)
Heart V50 (%)
Patient 3DCRT IMRT
SIB
1 ES FS 2 ES FS 3 ES FS 4 ES FS 5 ES FS 6 ES FS 7 ES FS 8 ES FS
7.0 10.6 8.2 12.6 12.0 12.6 8.7 36.1 11.6 17.5 7.9 9.3 16.5 15.0 12.6 10.6
20.3 28.0 9.3 14.9 23.6 24.9 28.4 25.2 31.1 37.0* 15.5 20.3 22.0 17.5 22.4 30.2*
5.0 8.0 5.1 6.5 19.1 14.1 6.5 15.1 10.3 12.7 5.2 5.7 8.9 8.4 10.3 9.2
3DCRT IMRT 8.5 11.1 16.4 14.7 16.5 20.9 11.2 8.7 10.1 9.2 2.7 3.6 8.4 8.0 8.2 2.2
4.8 6.4 8.5 7.1 7.7 5.3 4.2 5.4 8.0 7.3 1.2 1.0 3.5 3.9 4.4 1.0
Lung V20 (%) SIB 6.8 8.8 10.0 7.8 12.7 11.4 5.7 6.4 9.0 9.2 2.1 2.2 5.7 5.4 5.0 2.1
3DCRT IMRT 8.4 11.8 9.2 10.7 12.5 15.0 13.2 10.6 17.2 20.3 7.3 7.4 8.7 11.1 8.8 13.9
8.6 12.0 8.7 7.9 9.7 8.8 10.0 10.6 19.8 19.0 8.4 7.6 8.6 10.2 8.6 14.9
Left kidney V20 (%)
Spinal cord Dmax (Gy)
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
9.9 13.0 9.7 8.9 10.8 11.2 11.1 12.9 20.3 20.2 9.0 8.4 10.2 12.0 9.6 16.2
1.7 2.8 0.0 0.0 0.0 0.0 7.0 19.9 10.1 1.0 –y – – – – –
0.7 1.5 0.0 0.0 0.0 0.0 3.6 15.7 10.4 0.9 – – – – – –
0.8 1.7 0.0 0.0 0.0 0.0 4.4 16.1 9.9 1.1 – – – – – –
37.5 35.7 40.1 33.7 43.1 42.4 30.0 29.7 44.4 34.6 37.3 38.8 39.7 36.8 39.4 41.1
35.2 38.3 39.2 42.5 39.0 41.5 36.6 37.6 40.0 40.5 35.3 37.3 41.8 39.8 39.2 40.1
40.9 39.4 44.9 36.2 44.6 43.8 39.4 38.9 44.1 44.9 40.4 43.5 37.2 43.2 39.6 44.2
Abbreviations: Vn = fraction of volume receiving a dose of n Gy; Dmax = maximum dose delivered to the volume; 3DCRT, three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SIB = IMRT + simultaneous integrated boost; ES = empty stomach; FS = full stomach. * NTCP = 0. y Left kidney located inferior to the imaging field and therefore not included in the CT dataset.
the analysis of this factor. Sliding hiatal hernia is probably underdiagnosed, with estimates of its incidence varying between 8% and 33% (17). Moreover, Bremner et al. (18) mentioned that hiatal hernias smaller than 2 cm are usually asymptomatic, suggesting that attempts to detect them are unnecessary. However, a 2-cm sliding displacement of the GTV could become significant in RT. By contrast, Kahrilas et al. (19) found an inverse proportional relation between the severity of sliding hiatal hernia and esophageal shortening with swallowing. Overall, the influence of gastric filling on GEJ tumor motion in the presence of a hiatal hernia is complex, and the need for individualized target visualization, in addition to a constant gastric volume, might be crucial to ensure that dose-escalated RT is administered to the appropriate volume in patients with this clinical condition. This complex issue was beyond the scope of this study, but it should be kept in mind when planning RT for a GEJ cancer. Besides the impact of gastric filling on target coverage, tumor motion related to breathing is another factor to consider (20, 21). Dieleman et al. (22) described that breathing has a more profound effect on the position of the normal distal esophagus. We purposely did not use an internal target vol-
ume (ITV) based on a complete 4DCT dataset to specifically isolate the effects of stomach filling. However, our practice for esophageal cancer is to add an ITV based on a 4DCT dataset to the previously described contour guidelines. Thus, although this study shows the isolated impact of variations in gastric filling, it is limited in that it cannot show whether, in our daily practice, the larger target volumes resulting from the addition of an ITV might prevent some geographic miss on the CTV. CONCLUSION Stomach filling has a negligible impact on the standard prescribed dose delivered to a GTV located at the GEJ, when either 3DCRT or IMRT is planned. Thus, local relapses are not likely to be related to variations in gastric filling. Only a sustained increase in gastric filling could result in CTV underdosage for RT of a GEJ cancer. However, changes in stomach filling result in boost target miss when IMRT-SIB is planned, suggesting that caution be taken when dose escalation for GEJ tumors is attempted without adequate tumor visualization and close monitoring to ensure reproducible stomach filling.
REFERENCES 1. Brown LM, Devesa SS, Chow WH. Incidence of adenocarcinoma of the esophagus among white Americans by sex, stage, and age. J Natl Cancer Inst 2008;100:1184–1187. 2. Daly JM, Fry WA, Little AG, et al. Esophageal cancer: Results of an American College of Surgeons Patient Care Evaluation Study. J Am Coll Surg 2000;190:562–572. discussion 572–563.
3. NCCN guidelines: http://www.nccn.org/professionals/ physician_gls/f_guidelines.asp; Last accessed March 24, 2009. 4. Pan T, Lee TY, Rietzel E, et al. 4D-CT imaging of a volume influenced by respiratory motion on multi-slice CT. Med Phys 2004;31:333–340. 5. O’Daniel J, Bucci MK, Balter P. Image-guided planning margin determination for concomitant boost treatment of 3D-
300
6. 7.
8. 9.
10. 11.
12.
13.
I. J. Radiation Oncology d Biology d Physics
conformal, IMRT, and stereotactic lung cancer radiotherapy. Med Phys 2008;35:2640 [Abstract]. Shepard DM, Earl MA, Li XA, et al. Direct aperture optimization: A turnkey solution for step-and-shoot IMRT. Med Phys 2002;29:1007–1018. van Herk M, Remeijer P, Rasch C, et al. The probability of correct target dosage: Dose-population histograms for deriving treatment margins in radiotherapy. Int J Radiat Oncol Biol Phys 2000;47:1121–1135. van Herk M. Errors and margins in radiotherapy. Semin Radiat Oncol 2004;14:52–64. Minsky BD, Pajak TF, Ginsberg RJ, et al. INT 0123 (Radiation Therapy Oncology Group 94-05) phase III trial of combinedmodality therapy for esophageal cancer: High-dose versus standard-dose radiation therapy. J Clin Oncol 2002;20:1167–1174. Denham JW, Steigler A, Kilmurray J, et al. Relapse patterns after chemo-radiation for carcinoma of the oesophagus. Clin Oncol (R Coll Radiol) 2003;15:98–108. Settle SH, Bucci MK, Palmer MB, et al. PET/CT fusion with treatment planning CT (TP CT) shows predominant pattern of locoregional failure in esophageal patients treated with chemoradiation (CRT) is in GTV. Int J Radiat Oncol Biol Phys 2008; 72:S72–S73. Button MR, Morgan CA, Croydon ES, et al. Study to determine adequate margins in radiotherapy planning for esophageal carcinoma by detailing patterns of recurrence after definitive chemoradiotherapy. Int J Radiat Oncol Biol Phys 2009;73:818–823. Geh JI, Bond SJ, Bentzen SM, et al. Systematic overview of preoperative (neoadjuvant) chemoradiotherapy trials in oeso-
Volume 77, Number 1, 2010
14. 15. 16. 17. 18. 19. 20.
21. 22.
phageal cancer: Evidence of a radiation and chemotherapy dose response [see comment]. Radiother Oncol 2006;78: 236–244. Watanabe M, Isobe K, Takisima H, et al. Intrafractional gastric motion and interfractional stomach deformity during radiation therapy. Radiother Oncol 2008;87:425–431. Patrick A, Epstein O. Review article: gastroparesis. Aliment Pharmacol Ther 2008;27:724–740. Wood JD, Galligan JJ. Function of opioids in the enteric nervous system. Neurogastroenterol Motil 2004;16(Suppl 2): 17–28. Payne WS, Olsen AM. The esophagus. Philadelphia: Lea & Febiger; 1974. Bremner CG, Demeester TR, Huprich JE, et al. Esophageal disease and testing. New York: Taylor & Francis; 2005. Kahrilas PJ, Wu S, Lin S, et al. Attenuation of esophageal shortening during peristalsis with hiatus hernia. Gastroenterology 1995;109:1818–1825. Yaremko BP, Guerrero TM, McAleer MF, et al. Determination of respiratory motion for distal esophagus cancer using four-dimensional computed tomography. Int J Radiat Oncol Biol Phys 2008;70:145–153. Zhao KL, Liao Z, Bucci MK, et al. Evaluation of respiratory-induced target motion for esophageal tumors at the gastroesophageal junction. Radiother Oncol 2007;84:283–289. Dieleman EM, Senan S, Vincent A, et al. Four-dimensional computed tomographic analysis of esophageal mobility during normal respiration. Int J Radiat Oncol Biol Phys 2007;67: 775–780.
doi:10.1016/j.ijrobp.2009.08.026
PHYSICS CONTRIBUTION
IMPACT OF GASTRIC FILLING ON RADIATION DOSE DELIVERED TO GASTROESOPHAGEAL JUNCTION TUMORS MYRIAM BOUCHARD, M.D.,* MARY FRANCES MCALEER, M.D., PH.D.,y AND GEORGE STARKSCHALL, PH.D.* Departments of *Radiation Physics and yRadiation Oncology, The University of Texas M. D. Anderson Cancer Center, Houston, TX Purpose: This study examined the impact of gastric filling variation on target coverage of gastroesophageal junction (GEJ) tumors in three-dimensional conformal radiation therapy (3DCRT), intensity-modulated radiation therapy (IMRT), or IMRT with simultaneous integrated boost (IMRT-SIB) plans. Materials and Methods: Eight patients previously receiving radiation therapy for esophageal cancer had computed tomography (CT) datasets acquired with full stomach (FS) and empty stomach (ES). We generated treatment plans for 3DCRT, IMRT, or IMRT-SIB for each patient on the ES-CT and on the FS-CT datasets. The 3DCRT and IMRT plans were planned to 50.4 Gy to the clinical target volume (CTV), and the same for IMRTSIB plus 63.0 Gy to the gross tumor volume (GTV). Target coverage was evaluated using dose–volume histogram data for patient treatments simulated with ES-CT sets, assuming treatment on an FS for the entire course, and vice versa. Results: FS volumes were a mean of 3.3 (range, 1.7–7.5) times greater than ES volumes. The volume of the GTV receiving $50.4 Gy (V50.4Gy) was 100% in all situations. The planning GTV V63Gy became suboptimal when gastric filling varied, regardless of whether simulation was done on the ES-CT or the FS-CT set. Conclusions: Stomach filling has a negligible impact on prescribed dose delivered to the GEJ GTV, using either 3DCRT or IMRT planning. Thus, local relapses are not likely to be related to variations in gastric filling. Dose escalation for GEJ tumors with IMRT-SIB may require gastric filling monitoring. Ó 2010 Elsevier Inc. Gastroesophageal junction cancer, Gastric filling, Radiation therapy treatment planning, Intensity-modulated radiation therapy, Dose escalation.
at the level of the GEJ. To mitigate the effects of gastric filling, our institutional guidelines ask patients to be nil per os (NPO) 3 hours before simulation and each treatment. Compliance with this instruction is currently verified only at the simulation session, assuming the initial verbal instruction will carry over to subsequent treatment sessions. However, because patients may not be NPO for 3 hours before each treatment, it is possible that gastric filling may affect positioning of the target during treatments; in such cases, the effect of variation in gastric filling on RT dose delivery for distal esophageal tumors has not yet been well defined. The purpose of this study was to analyze the impact of variation of gastric filling on target coverage of GEJ tumors irradiated using three RT techniques: three-dimensional conformal RT (3DCRT), intensity-modulated RT (IMRT), and IMRT with simultaneous integrated boost (IMRT-SIB).
INTRODUCTION Since 1975, data have shown that the rates of esophageal adenocarcinomas have grown rapidly, by 463% and 335% in American men and women, respectively (1). Most adenocarcinomas involve the distal esophagus, close to or at the gastroesophageal junction (GEJ) (2). Most patients present with locoregionally advanced disease, and combination chemotherapy and radiation therapy (RT) is an important part of the standard treatment, either in the neoadjuvant or the definitive setting (3). Current techniques of RT for esophageal tumors include image guidance based on bony landmarks to help ensure that the radiation is delivered accurately. However, image registration based on bony landmarks does not necessarily correlate with the position of the esophageal tumor because of several factors. One such factor is gastric filling, especially Reprint requests to: George Starkschall, Ph.D., Department of Radiation Physics, Unit 94, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 3-2537; Fax: (713) 563-2479; E-mail: [email protected]; [email protected] Supported by Universite´ de Sherbrooke, Sherbrooke, Qc, Canada (M.B.). M. Bouchard is presently at Centre Hospitalier Universitaire de Sherbrooke, Sherbrooke Qc Canada.
Accepted for presentation at the 51th meeting of the American Society for Therapeutic Radiation and Oncology in Chicago, Illinois, November 2009. Conflict of interest: G.S. receives research support from Philips Medical Systems. The authors report no other conflict of interest. Received June 22, 2009, and in revised form July 28, 2009. Accepted for publication Aug 10, 2009. 292
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
Table 1. Objectives and constraints for all plans
MATERIALS AND METHODS Patient selection Four-dimensional computed tomography (4DCT) image datasets from 8 patients previously treated with RT for esophageal cancer were analyzed. The datasets were acquired in accordance with an institutional review board—approved protocol (RCR03-0400). Of the 8 patients, 5 presented with a full stomach (FS) at the time of the initial 4DCT simulation. For that reason, the simulation was repeated with patients to be NPO for 3 hours before scanning (empty stomach [ES]). The 3 other patients were found to have had an ES during the simulation scan but an FS on repeat 4DCT scans acquired during treatment.
4DCT image selection The 4DCT image sets were acquired with a helical multislice CT scanner (Discovery PET/CT; GE Medical Systems, Waukesha, WI) using a cine mode protocol; after acquisition, images were reconstructed and binned into 10 phases over the respiratory cycle (4). To simplify the planning, we selected the 50% phase (end-expiration) of each dataset and registered the datasets from both the ES and FS scans of each patient according to the bony anatomy at the level of the GEJ (defined below). For each patient, the entire stomach was scanned, so that full and empty volumes could be compared. The 50% phase total lung volume was also determined to document breathing-induced differences that could potentially influence the GEJ tumor’s position. Because of 4DCT image-sorting factors, in certain cases the 4DCT 50% phase might not exactly represent the end-expiration anatomy, which might instead be better represented by images from the 40% or 60% phase; we therefore wanted to quantify any technology-induced bias that may have occurred in the position of the GEJ tumor.
Treatment planning Target volumes for each patient consisted of a gross tumor volume (GTV), defined as the GEJ plus a 2-cm cranial and a 2-cm caudal extension at that level. In our study, we opted for a uniform GTV definition to minimize the delineation uncertainty, knowing that a majority of tumors could have a similar shape at this level. The GEJ was defined as occurring at the last CT slice where the diaphragm’s left crus was seen clearly attached to the esophagus. The clinical target volume (CTV) included a 3-cm craniocaudal and a 1-cm radial expansion of the GTV, in addition to the distal paraesophageal, celiac, and left gastric lymph nodes. Finally, a 5-mm isotropic expansion was added to the CTV to define the planning target volume (PTV), assuming daily image guidance. A planning gross target volume (PGTV) was defined by adding a margin of 4 mm to the GTV for dose escalation planning using the IMRTSIB technique (5). The PGTV is a planning entity that is used in our practice to account for intrafractional motion and setup uncertainty of the GTV. The relationship of the PGTV to the GTV is analogous to that of the PTV to the CTV. The adjacent organs at risk (OARs), including the lungs, heart, spinal cord, liver, normal esophagus, and kidneys, were also contoured. For each patient, 3DCRT, IMRT, and IMRT-SIB plans were generated using the volumes contoured and dose calculated on the 50% phase of the ES 4DCT dataset (ES-CT). The same planning process was repeated on the 50% phase of each patient’s FS 4DCT dataset (FS-CT). The final dose prescription for the 3DCRT and IMRT plans was 50.4 Gy to the PTV, and for the IMRT-SIB plans it was 50.4 Gy to the PTV and 63.0 Gy to the PGTV.
293
Target /Organ at risk Target GTV PTV PGTV Organ at risk Lung
Heart Liver Normal esophagus Spinal cord Kidney (each)
Indicators
Objectives/Constraints
V50.4 (V63) V50.4 V50.4
100% 95% 95%
Mean dose V20 V10 V5 V50 V30
<18–20 Gy 35% 40% 60% 50% 30% or 30–40% with NTCP = 0 <32 Gy 70 Gy 50% 45 Gy 30%
Mean dose Maximum dose V50 Maximum dose V20
Abbreviations: GTV = gross tumor volume; Vn = fraction of volume receiving a dose of n Gy; PTV = planning target volume; PGTV = planning gross tumor volume, defined as the GTV plus a setup margin; NTCP = normal tissue complication probability. All plans were generated using a research version of a commercially available RT planning system (Pinnacle3, Philips Healthcare, Milpitas, CA). For 3DCRT plans, we used four 18-MV beams with the aperture shaped to the PTV plus a penumbra margin of 9–10 mm (sufficient PTV-to-field border margin to ensure a coverage similar to IMRT objectives), with individualized beam angle adjustment, beam weighting, wedges, and, in approximately half the cases, the addition of field-in-field for optimizing dose distributions. Whereas 18 MV was preferred to limit hot spots in the 3DCRT plans (most patients having >20 cm separation width), the choice of beam energy for IMRT was based on the class solutions used in our current practice. Both the IMRT and IMRT-SIB plans were generated using five or six 6-MV beams. Inverse planning was performed using the Pinnacle3 direct machine parameter optimization process (6) to meet the dose objectives and constraints presented in Table 1.
Treatment plan evaluation We analyzed the dosimetric impact of gastric filling by comparing dose–volume histograms (DVHs) that reproduced the clinical situations of patients having their treatments planned on an ES but with the entire RT treatment course delivered on an FS, and vice versa. Our results then would model a worst-case scenario that could occur if a patient was not compliant to the NPO instructions. Target volumes were contoured on FS-CT and ES-CT. To investigate the clinical scenario of planning on ES and treating on FS, we first computed the dose distributions on the ES-CT dataset using beam configurations based on the ES target volumes. Then, the plans were evaluated based on the DVHs calculated from the ES-CT target volumes and the FS-CT target volumes copied onto the ES-CT already registered with bony landmarks. The other scenario of planning on FS and treating on ES was evaluated by computing the dose distributions on the FS-CT dataset using beam configurations based on the FS target volumes. The plans were evaluated similarly, based on the DVHs calculated from the FS-CT target volumes and the ES-CT target volumes copied onto the FS-CT. The following metrics were calculated and compared between clinical scenarios for each patient: the volume of PTV (as
I. J. Radiation Oncology d Biology d Physics
294
Volume 77, Number 1, 2010
Fig. 1. Digitally reconstructed radiographs of (a) Patient 4, who had the greatest variation in stomach volume; yellow [y] = empty stomach; sky-blue [s] = full stomach, and (b) Patient 1, showing stomach filling at the first simulation session (full stomach, sky-blue [s] = 841.8 cm3) and at resimulation (empty stomach, yellow [y] = 232.1 cm3). The 3 other volumes are the stomach filling for the same patient during RT with usual nil per os instructions: week 1, lavender [l] = 835.5 cm3; week 4, violet [v] = 988.5 cm3; week 5, orange [o] = 278.62 cm3. a percentage) receiving $50.4 Gy (PTV V50.4), the CTV mean dose, the dose received by 99% of the CTV (CTV D99), and GTV V50.4 for all plans; in addition, the PGTV V63 was evaluated for the IMRTSIB plans. Means and standard deviations were obtained for these metrics as well. The impact of planning on FS-CT or on ES-CT volumes on OARs was evaluated by comparison with the dose–volume points representing the limits of acceptable toxicity for each OAR. These dose–volume limits corresponded to the constraints used for IMRT planning (Table 1) and have been part of our institutional practice. Statistical comparisons were made using the Student’s t-test, with a p value of #0.05 taken to indicate significant differences.
RESULTS The FS volumes were larger than the ES volumes by a mean factor of 3.3 (range, 1.7–7.5). Figure 1a illustrates the full and empty stomach of Patient 4, who had the greatest variation in stomach volume; likewise, Fig. 1b shows gastric filling variations for Patient 1 over the course of radiation treatment. Although Patient 1 was instructed to be NPO for 3 hours before each treatment, gastric filling varied as follows: week 1, 835.5 cm3; week 2, 461.4 cm3; week 4, 988.5 cm3; week 5, 278.6 cm3. Table 2 shows the empty and full stomach volumes for all patients studied.
Table 2. Stomach volumes (single measurements) for the 8 patients
Patient 1 2 3 4 5 6 7 8
Empty stomach volume (cm3)
Full stomach volume (cm3)
Increase in volume from empty to full (%)
214.9 258.4 272.3 160.3 230.5 275.7 370.9 226.5
841.8 696.9 799.7 1 362.0 1 103.7 1 302.0 1 000.6 953.9
292 170 194 750 379 372 170 321
To ascertain whether breathing could have an impact on the position of the GEJ tumor, we measured the lung volume for each 50% phase image set. The mean ( standard deviation) variation of the total lung volume between the 50% phase ES-CT and the 50% phase FS-CT for the 8 patients was 4.2 5.1%. Tables 3 and 4 present the dose coverage of the target volumes for the patients simulated with an ES but treated with an FS during the entire RT course. Table 3 shows the GTV, CTV, and PTV coverage for ES plans delivered on FS volumes; Table 4 details the differences when comparing the coverage of ES volumes with the original ES standard-dose treatment plans (3DCRT and IMRT) and coverage of FS volumes with the same plans. Whereas the ES CTV D99 and ES CTV mean dose were equal to or greater than the prescribed dose (with higher doses falling within clinically acceptable limits) on the initial ES 3DCRT and IMRT plans, delivering a whole course of treatment on FS volumes resulted in some underdosed areas on the CTV, with mean FS CTV D99 values of 47.7 2.9 Gy and 44.9 6.6 Gy, respectively. The mean dose to the FS CTV was greater than or equal to the prescribed dose for all plans. The mean FS PTV V50.4 value for 3DCRT ES plans was 86.1 3.8%, and that for IMRT plans was 88.4 6.0%. No statistical difference was observed between these results for the 3DCRT and IMRT techniques. The DVH curves from the 3DCRT and IMRT plans for a typical patient are given in Figs. 2 and 3, respectively. No direct correlation between the absolute or relative stomach volume increase and the lack of target coverage was observed. Tables 5 and 6 present the dose coverage of the target volumes for the patients simulated with an FS but treated with an ES during the entire RT course. Table 5 shows the GTV, CTV, and PTV coverage for FS plans delivered on ES volumes; Table 6 details the differences when comparing the coverage of FS volumes with the original FS standard-dose treatment plans (3DCRT and IMRT) and coverage of ES volumes with the same plans. Target coverage was adequate, with an ES GTV V50.4 value of 100% and a minimum ES
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
295
Table 3. Dose–volume results calculated for full stomach treatments using plans generated on empty stomach computed tomography datasets GTV V50.4 (%)
CTV D99 (Gy)*
PTV V50.4 (%)y
CTV Dmean (Gy)
PGTV V63 (%)
Patient
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
SIB
1 2 3 4 5 6 7 8 Mean SD
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
48.8 49.7 49.9 43.1 43.2 47.6 49.9 49.3 47.7 2.9
38.0 50.2 48.9 38.7 34.8 50.4 50.8 47.2 44.9 6.6
41.6 51.3 52.8 41.9 36.8 51.5 51.7 51.1 47.3 6.2
51.9 52.2 52.0 51.9 51.7 52.6 52.1 52.5 52.1 0.3
52.1 51.4 51.8 50.7 51.2 51.8 51.5 51.5 51.5 0.4
58.0 57.8 60.3 56.4 57.8 59.1 59.4 58.5 58.4 1.2
84.9 86.4 90.2 81.9 79.9 88.6 90.8 86.2 86.1 3.8
82.0 90.6 94.9 79.4 86.0 91.6 96.2 86.5 88.4 6.0
90.4 95.6 98.9 86.1 86.0 96.0 97.8 98.1 93.6 5.3
82.8 71.7 92.2 71.7 85.3 76.8 84.7 80.4 80.7 7.1
Abbreviations: GTV = gross tumor volume; Vn = fraction of volume receiving a dose of n Gy; CTV = clinical target volume; D99 = dose delivered to 99% of volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; PGTV = planning gross tumor volume, defined as the GTV plus a setup margin; 3DCRT = three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SIB = IMRT + simultaneous integrated boost; SD = standard deviation. * 3DCRT vs. IMRT, p = 0.14. y 3DCRT vs. IMRT, p = 0.11.
CTV D99 of 49.6 Gy, with most patients receiving at least the prescribed dose (and with higher doses meeting clinically acceptable limits) for 3DCRT and IMRT plans. The mean ES PTV V50.4 values were 95.5 2.5% and 96.2 4.1% for the 3DCRT and IMRT plans, respectively. Table 3 presents the FS PGTV V63 results for ES IMRTSIB plans and Table 5 shows the ES PGTV V63 results for FS IMRT-SIB plans for all patients. As shown in Table 5, any gastric filling variations resulted in a loss of PGTV cov-
erage. Figure 4 illustrates ES and FS PGTV as well as ES and FS GTV DVHs for ES IMRT-SIB plans for 2 patients. No major change in dose to OARs resulted from variation of target volume with gastric filling, except in 1 patient (Patient 4) who had the largest variation in stomach volume. In this case, there was a noticeable increase in the left kidney V20 from 3.6% to 7.0% with ES plans to 15.7% to 19.9% with FS plans, whereas the dose to the left kidney was within clinically acceptable limits on both plans. Table 7 shows the
Table 4. Dose–volume results for 3DCRT and IMRT plans made using a empty-stomach datasets with treatments delivered to empty and full stomachs CTV D99 (Gy)
CTV Dmean (Gy)
PTV V50.4 (%)
Treatment type
Patient
ES CTV
FS CTV
Diff
ES CTV
FS CTV
Diff
ES PTV
FS PTV
Diff
3DCRT
1 2 3 4 5 6 7 8 Mean SD 1 2 3 4 5 6 7 8 Mean SD
50.70 51.10 50.60 50.90 50.40 51.20 50.90 51.40 50.9 0.3 51.2 50.9 51.0 50.7 50.8 51.3 50.8 51.2 51.0 0.2
48.8 49.7 49.9 43.1 43.2 47.6 49.9 49.3 47.7 2.9 38.0 50.2 48.9 38.7 34.8 50.4 50.8 47.2 44.9 6.6
1.9 1.4 0.7 7.8 7.2 3.6 1.0 2.1 3.2 2.8 13.2 0.7 2.1 12.0 16.0 0.9 0.0 4.0 6.1 6.5
52.1 52.4 51.9 52.5 51.9 52.8 52.4 52.7 52.3 0.3 52.7 51.7 51.9 51.2 51.8 51.9 51.5 51.6 51.8 0.4
51.9 52.2 52.0 51.8 51.7 52.6 52.1 52.5 52.1 0.3 52.1 51.4 51.8 50.7 51.2 51.8 51.5 51.5 51.5 0.4
0.3 0.2 (0.1) 0.6 0.3 0.2 0.2 0.2 0.2 0.2 0.6 0.3 0.1 0.5 0.6 0.1 0.0 0.1 0.3 0.3
95.9 97.5 94.1 95.6 91.7 97.5 97.9 97.3 95.9 2.1 98.8 99.9 99.9 99.7 99.9 100.0 99.9 100.0 99.8 0.4
84.9 86.4 90.2 81.9 79.9 88.6 90.8 86.2 86.6 3.8 82.0 90.6 94.9 79.4 86.0 91.6 96.2 86.5 88.4 6.0
11.0 11.0 3.9 13.6 11.8 8.9 7.1 11.1 9.8 3.1 16.8 9.3 5.1 20.3 13.9 8.4 3.7 13.6 11.4 5.7
IMRT
Abbreviations: ES = empty stomach; FS = full stomach; CTV = clinical target volume; V50.4 = fraction of volume receiving a dose of 50.4 Gy; CTV = clinical target volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; 3DCRT, three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SD = standard deviation; Diff = absolute difference between ES-FS values, with negative values between ( ).
296
I. J. Radiation Oncology d Biology d Physics
Fig. 2. Patient 1 (average stomach volume variation), dose-volume histograms for 3-dimensional conformal radiation therapy (3DCRT) plans generated on an empty stomach (ES) and delivered to ES (top) and full stomach (FS, bottom).
DVH point-dose results for the ES and FS plans relative to recognized limits for OAR radiation toxicity risk.
DISCUSSION Our results indicate that stomach filling has a negligible impact on the standard prescribed dose delivered to GTVs located at the GEJ when either 3DCRT or IMRT planning is used. However, CTV evaluation needs careful interpretation. For plan evaluation in the general practice of RT, the geographic concept of the PTV is used as an indirect measure of CTV coverage, despite the limitations related to the fact that it is a static representation of the patient created from a single CT dataset. Using the margins defined by van Herk et al. (7) and van Herk (8), the PTV reflects a 90% probability that the CTV will receiving at least 95% of the prescribed dose. For all patients examined here, treating patients with FS during the entire GEJ tumor treatment course when the treatment was planned on an ES would have had no major impact on the CTV Dmean. Our study shows that an occasional variation in gastric filling (few fractions), even if sizeable, does not seem to affect the dose delivered to the target volumes during the course of standard
Volume 77, Number 1, 2010
Fig. 3. Patient 1 (average stomach volume variation), DVH results for IMRT plans generated on ES and delivered to ES and FS.
RT of the GEJ. A sustained increase in gastric volume compared with the volume at treatment simulation during most fractions of the treatment, however, could lead to inadequate dose to the PTV and a small CTV geographic miss, as indicated by our suboptimal PTV and CTV D99 results. We estimate that our worst-case scenario result of approximately 85% PTV coverage for a whole treatment course with an increased stomach volume could become clinically significant if, for example, more than half the RT fractions would be delivered to a patient having an FS, leading to inadequate CTV coverage. For this reason, patient NPO instructions should be emphasized throughout the patient’s treatment course. The adequacy of GTV coverage despite changes in gastric filling is an important finding, because it does not support the hypothesis that gastric filling variations explain local relapses of GEJ tumors. Many studies have reported that local treatment failure remains an important issue in esophageal cancer (9, 10). Furthermore, in a study of 66 patients in whom most tumors were located in the GEJ and were treated by definitive chemoradiation therapy using either IMRT or 3DCRT to 50.4 Gy, Settle et al. (11) showed that 24 patients were found to have experienced locoregional failure. The majority (70%) of the local failures occurred within the GTV, whereas only two occurred within the CTV and one in the PTV. By contrast, Button et al.(12) analyzed the
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
297
Table 5. Dose–volume results calculated for empty stomach treatments using plans generated on full stomach computed tomography datasets GTV V50.4 (%)
CTV D99 (Gy)
CTV Dmean (Gy)
PGTV V63 (%)
PTV V50.4 (%)
Patient
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
SIB
1 2 3 4 5 6 7 8 Mean SD
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
100 100 100 100 100 100 100 100 100 0
50.7 50.7 51.0 50.8 50.3 49.8 49.8 52.4 50.7 0.8
50.4 51.2 50.6 51.6 50.9 50.9 49.6 50.9 50.8 0.6
50.8 52.8 51.5 52.5 52.4 51.6 51.6 51.5 51.8 0.7
52.2 51.4 52.3 52.2 52.4 53.0 52.5 53.2 52.4 0.6
52.6 51.8 52.9 52.5 51.5 51.5 51.4 51.5 52.0 0.6
60.6 60.3 59.4 60.5 60.8 58.1 58.2 59.1 59.6 1.1
92.7 97.4 95.7 96.7 94.9 95.5 91.6 99.6 95.5 2.5
90.7 99.8 94.8 100.0 99.6 96.4 90.0 98.4 96.2 4.1
93.9 99.7 98.3 99.5 99.9 99.5 98.7 99.8 98.7 2.0
85.0 99.7 85.6 96.4 97.1 80.8 84.6 92.2 90.2 7.1
Abbreviations: GTV = gross tumor volume; Vn = fraction of volume receiving a dose of n Gy; CTV = clinical target volume; D99 = dose delivered to 99% of volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; PGTV = planning gross tumor volume, defined as the GTV plus a setup margin; 3DCRT = three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SIB = IMRT + simultaneous integrated boost; SD = standard deviation.
choice of margins for RT of esophageal cancer. They retrospectively analyzed patterns of disease recurrence, with local relapse compared with radiation field location in 145 patients with esophageal carcinoma who had received standard definitive chemoradiation therapy by 3DCRT techniques. The authors found that 49% of relapses were within the GTV at first presentation and that 96% of locoregional failures occurred within the radiation treatment fields, even if patients received adequate chemoradiation therapy. The same authors concluded that it would be impossible to
increase field size and thereby improve tumor control without simultaneously adding important toxicity (12). Inasmuch as a dose–response relationship has been suggested for esophageal tumors (13), and given that the current results suggest that gastric filling does not have an impact on the standard dose delivered to the GTV, the answer to better local control might be in devising ways to escalate the radiation dose to the GTV itself. Another important implication of our results is that to achieve adequate dose escalation with IMRT-SIB, direct
Table 6. Dose–volume results for 3DCRT and IMRT plans made using a full-stomach datasets with treatments delivered to empty and full stomachs CTV D99 (Gy)
CTV Dmean (Gy)
PTV V50.4 (%)
Treatment type
Patient
FS CTV
ES CTV
Diff
FS CTV
ES CTV
Diff
FS PTV
ES PTV
Diff
3DCRT
1 2 3 4 5 6 7 8 Mean SD 1 2 3 4 5 6 7 8 Mean SD
51.2 50.4 51.0 50.2 50.2 51.7 50.9 51.6 50.9 0.6 51.3 51.1 51.2 51.6 50.9 50.9 51.0 50.9 51.1 0.2
50.7 50.7 51.0 50.8 50.3 49.8 49.8 52.4 50.7 0.8 50.4 51.2 50.6 51.6 50.9 50.9 49.6 50.9 50.8 0.6
0.5 (0.3) 0.0 (0.6) (0.1) 1.9 1.1 (0.8) 0.2 0.9 0.9 (0.1) 0.6 0.0 0.0 0.0 1.4 0.0 0.4 0.6
52.4 51.4 52.3 52.0 52.7 53.2 52.6 53.2 52.5 0.6 52.7 51.8 51.9 52.5 51.5 51.5 51.5 51.5 51.9 0.5
52.2 51.4 52.3 52.1 52.4 53.0 52.5 53.2 52.4 0.5 52.6 51.8 51.9 52.5 51.5 51.5 51.4 51.5 51.8 0.5
0.2 0.0 0.0 (0.1) 0.3 0.2 0.2 0.0 0.1 0.1 0.1 0.0 0.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0
97.8 90.4 96.1 96.6 94 99.0 96.1 98.4 96.1 2.8 97.7 99.9 99.9 99.8 100.0 99.7 100.0 100.0 99.6 0.8
92.7 97.4 95.7 96.7 94.9 95.5 91.6 99.6 95.5 2.5 90.7 99.8 94.8 100.0 99.6 96.4 89.9 98.4 96.2 4.1
5.1 (7.0) 0.4 (0.1) (0.9) 3.5 4.5 (1.2) 0.5 3.9 7.0 0.1 5.1 (0.2) 0.4 0.3 10.0 1.2 3.0 3.9
IMRT
Abbreviations: ES = empty stomach; FS = full stomach; CTV = clinical target volume; V50.4 = fraction of volume receiving a dose of 50.4 Gy; CTV = clinical target volume; Dmean = mean dose delivered to the volume; PTV = planning target volume; 3DCRT, three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SD = standard deviation; Diff = absolute difference between ES-FS values, with negative values between ( ).
298
I. J. Radiation Oncology d Biology d Physics
Fig. 4. Patient 1 (average stomach volume variation) and Patient 4 (greatest stomach volume variation) dose-volume histograms for an intensity-modulated radiotherapy with simultaneous integrated boost (IMRT-SIB) plan generated on and empty stomach (ES) and delivered to ES and full stomach (FS). The FS DVH represents the clinical equivalent of a patient whose plan was simulated using images taken on an ES but who received all radiation treatments with an FS.
visualization of the tumor during treatments, with appropriate image guidance techniques and/or close monitoring of gastric filling, seems necessary. Current image guidance for esophageal cancers is based on matching the bony anatomy without confirming the real tumor location. Our results clearly show that with either ES or FS, the intended dose escalation to the PGTV with IMRT-SIB would be inadequate if variations in gastric filling occur during the treatment course. Of note, radiation dose escalation in esophageal cancer is still controversial, especially given that the landmark INT 0123 study suggested worse outcomes in the patients assigned to the higher radiation dose arm than those in the 50.4-Gy standard arm (9). However, the increased incidence of treatment-related death in the high-dose arm occurred mostly in patients who did not receive the planned dose, with 7 of 11 patients receiving #50.4 Gy. Moreover, the increased radiation late effects could be explained by that study’s time frame, 1995–1999, which preceded the advent of current RT techniques and therefore might have encountered more treatment-related toxicities than would be seen
Volume 77, Number 1, 2010
today. Newer, more effective, systemic treatment combinations might be part of the solution for this problem as well, but we believe there are still opportunities to escalate radiation doses by using recent advances in RT and imaging techniques and that these opportunities should be tested in clinical trials. Our results showed that planning on FS volumes allows adequate CTV and PTV coverage for 3DCRT and IMRT plans, even with a reduction in gastric filling during the course of the radiation treatments. This result indicates that it is adequate to plan RT for GEJ cancer on an FS-CT and that there is no need to repeat the scan if the patient is found to have an FS at the simulation session. No major resulting dose increase to OARs was noticed in most such cases. However, when normal tissue constraints are not met, it may be beneficial to rescan the patient by means of ES-CT and then replan the treatment to optimize OAR sparing. To our knowledge, our study is the first to quantify the dosimetric impact of variations in gastric filling on the radiation dose delivered to a GEJ tumor. Target underdosing with changes in gastric filling was similar with either 3DCRT or IMRT. However, the small sample size of our study might have precluded finding a statistically significant difference between the two techniques. The larger standard deviation on FS CTV D99 for IMRT technique and its lowest individual FS CTV D99 result (34.8 Gy for IMRT vs. 43.1 for 3DCRT) might favor 3DCRT to prevent geographic miss. Nevertheless, the same precautions regarding NPO instructions should be followed regardless of which of these techniques is used, inasmuch as both planning techniques resulted in some cases where the worst-case scenario could be clinically significant. Future investigations are needed to assess the frequency of gastric filling increases to determine the risk of CTV underdosage during treatment with current NPO instructions. Then, the need for intervention by a dietician and/or for patient instructions on methods of minimizing gastric emptying delay could be proposed in certain cases. Up to now, only Watanabe et al. (14) have described interfractional variations of stomach position in patients with gastric lymphomas, which they detected using fluoroscopic visualization. In that study, variations were important even if the patients had an overnight fast before treatment. Variations in gastric filling may differ from the one experienced by GEJ cancer patients because, in contrast to a more localized GEJ tumor, diffuse gastric lymphoma can harden tissues that may limit gastric filling and at the same time delay emptying, which could then induce variable responses to NPO instructions. Furthermore, it is important to recognize that gastric filling might vary because of different underlying diseases or other factors, such as the use of opioids or enteral nutrition techniques, which should also be evaluated individually (15, 16). Gastric filling has an unknown effect on GEJ tumor position in patients with a sliding hiatal hernia, one of the most common abnormalities involving the esophagus. The uniform GTV definition we used for our study did not allow
Gastric filling and dose to gastroesophageal junction tumors d M. BOUCHARD et al.
299
Table 7. Dosevolume results relative to recognized limits for radiation toxicity for empty stomach (ES) and full stomach (FS) 3DCRT, IMRT, and IMRT-SIB plans Liver V30 (%)
Heart V50 (%)
Patient 3DCRT IMRT
SIB
1 ES FS 2 ES FS 3 ES FS 4 ES FS 5 ES FS 6 ES FS 7 ES FS 8 ES FS
7.0 10.6 8.2 12.6 12.0 12.6 8.7 36.1 11.6 17.5 7.9 9.3 16.5 15.0 12.6 10.6
20.3 28.0 9.3 14.9 23.6 24.9 28.4 25.2 31.1 37.0* 15.5 20.3 22.0 17.5 22.4 30.2*
5.0 8.0 5.1 6.5 19.1 14.1 6.5 15.1 10.3 12.7 5.2 5.7 8.9 8.4 10.3 9.2
3DCRT IMRT 8.5 11.1 16.4 14.7 16.5 20.9 11.2 8.7 10.1 9.2 2.7 3.6 8.4 8.0 8.2 2.2
4.8 6.4 8.5 7.1 7.7 5.3 4.2 5.4 8.0 7.3 1.2 1.0 3.5 3.9 4.4 1.0
Lung V20 (%) SIB 6.8 8.8 10.0 7.8 12.7 11.4 5.7 6.4 9.0 9.2 2.1 2.2 5.7 5.4 5.0 2.1
3DCRT IMRT 8.4 11.8 9.2 10.7 12.5 15.0 13.2 10.6 17.2 20.3 7.3 7.4 8.7 11.1 8.8 13.9
8.6 12.0 8.7 7.9 9.7 8.8 10.0 10.6 19.8 19.0 8.4 7.6 8.6 10.2 8.6 14.9
Left kidney V20 (%)
Spinal cord Dmax (Gy)
SIB
3DCRT
IMRT
SIB
3DCRT
IMRT
SIB
9.9 13.0 9.7 8.9 10.8 11.2 11.1 12.9 20.3 20.2 9.0 8.4 10.2 12.0 9.6 16.2
1.7 2.8 0.0 0.0 0.0 0.0 7.0 19.9 10.1 1.0 –y – – – – –
0.7 1.5 0.0 0.0 0.0 0.0 3.6 15.7 10.4 0.9 – – – – – –
0.8 1.7 0.0 0.0 0.0 0.0 4.4 16.1 9.9 1.1 – – – – – –
37.5 35.7 40.1 33.7 43.1 42.4 30.0 29.7 44.4 34.6 37.3 38.8 39.7 36.8 39.4 41.1
35.2 38.3 39.2 42.5 39.0 41.5 36.6 37.6 40.0 40.5 35.3 37.3 41.8 39.8 39.2 40.1
40.9 39.4 44.9 36.2 44.6 43.8 39.4 38.9 44.1 44.9 40.4 43.5 37.2 43.2 39.6 44.2
Abbreviations: Vn = fraction of volume receiving a dose of n Gy; Dmax = maximum dose delivered to the volume; 3DCRT, three-dimensional conformal radiation therapy; IMRT = intensity-modulated radiation therapy; SIB = IMRT + simultaneous integrated boost; ES = empty stomach; FS = full stomach. * NTCP = 0. y Left kidney located inferior to the imaging field and therefore not included in the CT dataset.
the analysis of this factor. Sliding hiatal hernia is probably underdiagnosed, with estimates of its incidence varying between 8% and 33% (17). Moreover, Bremner et al. (18) mentioned that hiatal hernias smaller than 2 cm are usually asymptomatic, suggesting that attempts to detect them are unnecessary. However, a 2-cm sliding displacement of the GTV could become significant in RT. By contrast, Kahrilas et al. (19) found an inverse proportional relation between the severity of sliding hiatal hernia and esophageal shortening with swallowing. Overall, the influence of gastric filling on GEJ tumor motion in the presence of a hiatal hernia is complex, and the need for individualized target visualization, in addition to a constant gastric volume, might be crucial to ensure that dose-escalated RT is administered to the appropriate volume in patients with this clinical condition. This complex issue was beyond the scope of this study, but it should be kept in mind when planning RT for a GEJ cancer. Besides the impact of gastric filling on target coverage, tumor motion related to breathing is another factor to consider (20, 21). Dieleman et al. (22) described that breathing has a more profound effect on the position of the normal distal esophagus. We purposely did not use an internal target vol-
ume (ITV) based on a complete 4DCT dataset to specifically isolate the effects of stomach filling. However, our practice for esophageal cancer is to add an ITV based on a 4DCT dataset to the previously described contour guidelines. Thus, although this study shows the isolated impact of variations in gastric filling, it is limited in that it cannot show whether, in our daily practice, the larger target volumes resulting from the addition of an ITV might prevent some geographic miss on the CTV. CONCLUSION Stomach filling has a negligible impact on the standard prescribed dose delivered to a GTV located at the GEJ, when either 3DCRT or IMRT is planned. Thus, local relapses are not likely to be related to variations in gastric filling. Only a sustained increase in gastric filling could result in CTV underdosage for RT of a GEJ cancer. However, changes in stomach filling result in boost target miss when IMRT-SIB is planned, suggesting that caution be taken when dose escalation for GEJ tumors is attempted without adequate tumor visualization and close monitoring to ensure reproducible stomach filling.
REFERENCES 1. Brown LM, Devesa SS, Chow WH. Incidence of adenocarcinoma of the esophagus among white Americans by sex, stage, and age. J Natl Cancer Inst 2008;100:1184–1187. 2. Daly JM, Fry WA, Little AG, et al. Esophageal cancer: Results of an American College of Surgeons Patient Care Evaluation Study. J Am Coll Surg 2000;190:562–572. discussion 572–563.
3. NCCN guidelines: http://www.nccn.org/professionals/ physician_gls/f_guidelines.asp; Last accessed March 24, 2009. 4. Pan T, Lee TY, Rietzel E, et al. 4D-CT imaging of a volume influenced by respiratory motion on multi-slice CT. Med Phys 2004;31:333–340. 5. O’Daniel J, Bucci MK, Balter P. Image-guided planning margin determination for concomitant boost treatment of 3D-
300
6. 7.
8. 9.
10. 11.
12.
13.
I. J. Radiation Oncology d Biology d Physics
conformal, IMRT, and stereotactic lung cancer radiotherapy. Med Phys 2008;35:2640 [Abstract]. Shepard DM, Earl MA, Li XA, et al. Direct aperture optimization: A turnkey solution for step-and-shoot IMRT. Med Phys 2002;29:1007–1018. van Herk M, Remeijer P, Rasch C, et al. The probability of correct target dosage: Dose-population histograms for deriving treatment margins in radiotherapy. Int J Radiat Oncol Biol Phys 2000;47:1121–1135. van Herk M. Errors and margins in radiotherapy. Semin Radiat Oncol 2004;14:52–64. Minsky BD, Pajak TF, Ginsberg RJ, et al. INT 0123 (Radiation Therapy Oncology Group 94-05) phase III trial of combinedmodality therapy for esophageal cancer: High-dose versus standard-dose radiation therapy. J Clin Oncol 2002;20:1167–1174. Denham JW, Steigler A, Kilmurray J, et al. Relapse patterns after chemo-radiation for carcinoma of the oesophagus. Clin Oncol (R Coll Radiol) 2003;15:98–108. Settle SH, Bucci MK, Palmer MB, et al. PET/CT fusion with treatment planning CT (TP CT) shows predominant pattern of locoregional failure in esophageal patients treated with chemoradiation (CRT) is in GTV. Int J Radiat Oncol Biol Phys 2008; 72:S72–S73. Button MR, Morgan CA, Croydon ES, et al. Study to determine adequate margins in radiotherapy planning for esophageal carcinoma by detailing patterns of recurrence after definitive chemoradiotherapy. Int J Radiat Oncol Biol Phys 2009;73:818–823. Geh JI, Bond SJ, Bentzen SM, et al. Systematic overview of preoperative (neoadjuvant) chemoradiotherapy trials in oeso-
Volume 77, Number 1, 2010
14. 15. 16. 17. 18. 19. 20.
21. 22.
phageal cancer: Evidence of a radiation and chemotherapy dose response [see comment]. Radiother Oncol 2006;78: 236–244. Watanabe M, Isobe K, Takisima H, et al. Intrafractional gastric motion and interfractional stomach deformity during radiation therapy. Radiother Oncol 2008;87:425–431. Patrick A, Epstein O. Review article: gastroparesis. Aliment Pharmacol Ther 2008;27:724–740. Wood JD, Galligan JJ. Function of opioids in the enteric nervous system. Neurogastroenterol Motil 2004;16(Suppl 2): 17–28. Payne WS, Olsen AM. The esophagus. Philadelphia: Lea & Febiger; 1974. Bremner CG, Demeester TR, Huprich JE, et al. Esophageal disease and testing. New York: Taylor & Francis; 2005. Kahrilas PJ, Wu S, Lin S, et al. Attenuation of esophageal shortening during peristalsis with hiatus hernia. Gastroenterology 1995;109:1818–1825. Yaremko BP, Guerrero TM, McAleer MF, et al. Determination of respiratory motion for distal esophagus cancer using four-dimensional computed tomography. Int J Radiat Oncol Biol Phys 2008;70:145–153. Zhao KL, Liao Z, Bucci MK, et al. Evaluation of respiratory-induced target motion for esophageal tumors at the gastroesophageal junction. Radiother Oncol 2007;84:283–289. Dieleman EM, Senan S, Vincent A, et al. Four-dimensional computed tomographic analysis of esophageal mobility during normal respiration. Int J Radiat Oncol Biol Phys 2007;67: 775–780.