- Email: [email protected]
Optimal beam design on intensity-modulated radiation therapy with simultaneous integrated boost in nasopharyngeal cancer
Medical Dosimetry ] (2014) ]]]–]]]
Medical Dosimetry journal homepage: www.meddos.org
Technical report
Optimal beam design on intensity-modulated radiation therapy with simultaneous integrated boost in nasopharyngeal cancer Mei-Chun Cheng, M.S.,†,‡ Yu-Wen Hu, M.D.,* Ching-Sheng Liu Ph.D.,* Jeun-Shenn Lee, Ph.D.,† Pin-I Huang, M.D.,* Sang-Hue Yen, M.D.,* Yuh-Lin Lee, M.S.,* Chun-Mei Hsieh, B.S.,* and Cheng-Ying Shiau, M.D.* Division of Radiation Oncology, Department of Oncology Medicine, Taipei Veterans General Hospital, Taipei, Taiwan; †Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, Taipei, Taiwan; and ‡Department of Radiation Oncology, Taichung Veterans General Hospital, Taichung, Taiwan
*
A R T I C L E I N F O
A B S T R A C T
Article history: Received 6 January 2014 Accepted 20 March 2014
This study aims to determine the optimal beam design among various combinations of field numbers and beam trajectories for intensity-modulated radiation therapy (IMRT) with simultaneous integrated boost (SIB) technique for the treatment of nasopharyngeal cancer (NPC). We used 10 fields with gantry angles of 1551, 1301, 751, 251, 01 L, 01 R, 3351, 2851, 2301, and 2051 denoted as F10. To decrease doses in the spinal cord, the F10 technique was designed by featuring 2 pairs of split-opposed beam fields at 1551 to 3351 and 2051 to 251, as well as one pair of manually split beam fields at 01. The F10 technique was compared with 4 other common field arrangements: F7E, 7 fields with 501 equally spaced gantry angles; F7, the basis of F10 with 1551, 1301, 751, 01, 2851, 2301, and 2051; F9E, 9 fields with 401 equally spaced gantry angles; and FP, 7 posterior fields with 1801, 1501, 1201, 901, 2701, 2401, and 2101. For each individual case of 10 patients, the customized constraints derived after optimization with the standard F10 technique were applied to 4 other field arrangements. The 4 new optimized plans of each individual case were normalized to achieve the same coverage of planning target volume (PTV)63 Gy as that of the standard F10 technique. The F10 field arrangement exhibited the best coverage in PTV70 Gy and the least mean dose in the trachea-esophagus region. Furthermore, the F10 field arrangement demonstrated the highest level of conformity in the low-dose region and the least monitor unit. The F10 field arrangement performed more outstandingly than the other field arrangements in PTV70 Gy coverage and spared the central organ. This arrangement also exhibited the highest conformity and delivery efficiency. The F10 technique is recommended as the standard beam geometry for the SIB-IMRT of NPC. & 2014 American Association of Medical Dosimetrists.
Keywords: Intensity-modulated radiation therapy (IMRT) Simultaneous integrated boost (SIB) Nasopharyngeal cancer (NPC) Beam design
Introduction According to the World Health Organization, the incidence rate of nasopharyngeal cancer (NPC) is o 1 case per 100,000 cases worldwide. However, the incidence rate is 4 20 cases per 100,000 cases in Southern Asia. In Taiwan, the incidence rate is approximately 8.29 and 2.77 for men and women, respectively, based on a 2008 report
Reprint requests to: Cheng-Ying Shiau, Division of Radiation Oncology, Department of Oncology Medicine, Taipei Veterans General Hospital, No. 201, Sec. 2, Shihpai Rd, Beitou District, Taipei City 112, Taiwan. Tel.: þ88 6228 757 270, ext: 211; fax: þ88 6228 749 425. E-mail: [email protected]. http://dx.doi.org/10.1016/j.meddos.2014.03.003 0958-3947/Copyright Ó 2014 American Association of Medical Dosimetrists
of the Cancer Registry Department of Health and Welfare, Executive Yuan, Taiwan. NPC is radiosensitive, hence radiotherapy or combined radiotherapy and chemotherapy is the standard method. With the advancement of 3-dimensional conformal radiotherapy (3D-CRT) to intensity-modulated radiation therapy (IMRT), tumor coverage has been improved and critical structures have been clearly spared.1 Therefore, the IMRT technique is recommended for the treatment of head and neck cancer.2 Clark et al.3 observed that the low-dose volume in IMRT is higher than that in 3D-CRT. IMRT has been used for treatment of NPC at Taipei Veterans General Hospital (VGH) since 2002. We found that the proper arrangement of beams not only achieved the plan criteria but also improved dose conformity in high-dose and low-
2
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]] corresponding levels. Level Ib was included only when definitive lymphadenopathy was observed. PTV administered at 63 Gy (PTV63 Gy) was created with margins of 3 mm around the upper CTV63 Gy above the C2-3 junction and 5 mm around the lower CTV63 Gy. The planning organ-at-risk volume was also created around the brain stem and at 10 cm of the spinal cord with margins of 3 and 5 mm, respectively. The dose constraints also included the inner ears and whole parotid glands. With a chin-up position, the eyes and optic nerves were excluded from the constraints, except for advanced T4 cases. These structures are shown in Fig. 2. IMRT planning technique
Fig. 1. Schematic of the beam arrangement. The regions of nasal cavity and neck in the axial view are shown in left side and right side of figures, respectively. The red area indicates the PTV, the brown area indicates the brain stem or the spinal cord, and the blue area indicates the trachea-esophagus region. (A) The PTV is covered by a beam at a gantry angle of 1551. (B) The PTV is covered by a partial beam at a gantry angle of 1551 and the split-opposed beam at 3351. (Color version of figure is available online.)
dose volumes. In our study, manually split or opposing partial fields were designed based on the relative position between planning target volume (PTV) and critical structures. Critical structures receive higher doses than the PTV if they are located in front of the PTV along the beam trajectory because of the physical characteristics of a photon beam. A nonopposing field arrangement is the basic principle of beam angle selection in IMRT. However, not all critical structures of each individual field could be spared or placed at the back of the PTV. After analyzing the geometric characteristics of critical structures vs the PTV of NPC, we designed the F10 beam arrangement with the original F7 fields split and partially placed in opposing trajectories to relocate the critical structures at the back of the PTV (Fig. 1). The 01 field was also split in half in our F10 field arrangement to decrease dose leakage to the larynx, trachea, and spinal cord of the dynamic multileaf collimator (DMLC) motion. This study aims to confirm that the Taipei VGH F10 field arrangement is a better IMRT beam design for NPC than 4 other popular field arrangements.
The SIB-IMRT for NPC at Taipei VGH had 2 dose levels: 70 Gy to CTV70 Gy and 63 Gy to CTV63 Gy in 35 fractions. The field arrangement was 10 fields, denoted as F10, which was developed in the era of Varian’s CadPlan Radiotherapy Treatment Planning System version 6.32 (Table 2). In this study, the IMRT was planned by Varian's Eclipse Radiotherapy Treatment Planning System version 7.1. The sweeping step-and-shoot technique with 10 dose levels was adopted to convert the optimal fluence into segments of leaf motion. The schematic of our idea is shown in Fig. 1. In Fig. 1A, the right side of the PTV at a beam trajectory of gantry 1551 is found behind critical structures. If the right side of the PTV was covered with sufficient doses, the critical structures in front of the PTV receive a higher dose than the PTV. Considering this finding, we developed a splitopposed field technique (Fig. 1B). Half of the field that covered the right side of the PTV was opposed to the gantry angle at 3351; hence, the right side of the PTV was proximal to the source at 3351. Considering this idea, we featured 2 pairs of splitopposed fields at gantry angles of 1551 to 3351 and 2051 to 251. Furthermore, the large field size was divided into multiple subfields in Varian's system. In our study, 01 L to 01 R were manually split to prevent leakage to the central structures. The IMRT covered the nasopharynx and the upper neck, and the anterior field covered the lower neck. The center of the IMRT field was placed at the middle neck junction with the half-beam lower neck single anterior field, which was not compared in this study. The F10 technique was compared with 4 other popular field arrangements, as shown in Table 2. The 4 other field arrangements were F7, F7E, F9E, and FP. The F7 technique, the predecessor of the F10 technique, exhibits a similar gantry angle, except for split-opposed fields. In the F7E technique, 7 equally spaced gantry angles are considered.5 By comparison, the F9E technique uses 9 equally spaced gantry angles. The FP technique uses 7 posterior gantry angles.6-8 A standard template of dose-volume histogram (DVH) constraints for the inverse treatment plan of the SIB-IMRT of NPC was routinely applied for actual treatment planning using the F10 beam arrangement. The constraints of the standard template were adjusted, if desired, during the optimization process to achieve a satisfactory plan for each patient. Derived from the approved inverse optimization planning with the F10 technique for actual clinical treatment, the customized constraints were then applied to the 4 other techniques to create the respective SIB-IMRT plans for the 10 patients with NPC included in this study. The approved criteria for DVH are shown in Tables 3 and 4. Analysis The 4 new optimized plans of each individual case were normalized to obtain the same coverage of PTV63 Gy as that of the F10 technique. The DVH of targets and other critical organs were compared among the 5 techniques (Tables 5 and 6). Plans were rejected because of inadequate coverage of PTV70 Gy or excessive dose to the spinal cord or brain stem. The conformity index9 was used to evaluate plan conformity. The equation is expressed as follows: CI ¼
¼ Methods and Materials Patient selection and structure delineation During 2005, 10 consecutive patients with NPC who were treated with 10-field simultaneous integrated boost (SIB)4 IMRT at Taipei VGH were selected for this study. The patients immobilized with a thermoplastic mask were subjected to computed tomography simulation (5 mm/slice). The TNM categories based on the American Joint Committee on Cancer sixth edition cancer staging system for the 10 patients are shown in Table 1. Gross tumor volume was defined using 5 sets of magnetic resonance image fusion, including axial T2-weighted image (WI), axial T1WI, postcontrast axial T1 with fat saturation, postcontrast coronal T1WI, and sagittal T1WI. Clinical target volume (CTV)70 Gy was defined by gross tumor volume of both primary tumor and regional lymphadenopathy. PTV was administered at 70 Gy (PTV70 Gy) and created by adding margins of 3 mm for the primary tumor and margins of 5 mm for the neck lymphadenopathy. CTV63 Gy encompassed the CTV70 Gy, whole nasopharynx, skull base, bilateral tonsils, and at least level II of lymphatics. The CTV63 Gy extended to level III or level IV for patients with gross lymphadenopathy in the
RV 1 PTV in RV PTV in RV=PTV RV PTV
1 PTV in RV=PTV
ð1Þ
2 ð2Þ
where RV is the radiation volume. Overall, 4 conformity indices, namely, CI70 Gy, CI63 Gy, CI49 Gy, and CI35 Gy, were defined in our study. CI70 Gy and CI63 Gy are the conformity indices of 70 and 63 Gy, respectively and are described as the dose conformity to high-dose regions (Eqs. (3) and (4)). CI49 Gy and CI35 Gy are the Table 1 TNM stages of the 10 patients in our study T stage
N stage 0
1 2 3 4 Total
1 2 3
1
2
1 3
1
1 5
1 2
3
Total
0
1 4 1 4 10
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]]
3
Fig. 2. Definition of structures included PTV70 Gy (red segment), PTV63 Gy (light green segment), parotid gland (purple segment), brain stem (blue segment), and spinal cord (blue segment). The typical dose distributions included 7000, 6300, 5600, 4900, 4200, and 3500 cGy. (Color version of figure is available online.) conformity indices of 49 and 35 Gy, respectively, and are described as the dose conformity to low-dose regions (Eqs. (5) and (6)). CI70
Gy
¼
RV70 Gy PTV70 Gy
CI63
Gy
¼
RV63 Gy PTV63 Gy
V70
Gy
1 of PTV70
Gy
1 of PTV63
V63
CI49
Gy
RV49 Gy ¼ PTV63 Gy
CI35
Gy
¼
2 ð3Þ
Gy
2 Gy
Results A typical dose distribution is shown in Fig. 2. Dose distributions include the coverage dose of PTV70 Gy (7000 cGy), coverage dose of PTV63 Gy (6300 cGy), and sparing doses of the brain stem (5600 cGy), spinal cord (4900 cGy), and parotid gland (3500 cGy).
ð4Þ Table 3 Approved criteria (target) for DVH ð5Þ
RV35 Gy PTV63 Gy
ð6Þ
We performed repeated-measures analysis of variance by using SPSS 13.0 (SPSS Inc., Chicago, IL) to detect the statistical difference between the 5 techniques. Bonferroni-corrected post hoc t-tests were used in pairwise comparison.10 Table 2 Beam trajectory
Target
Criterion
PTV70 Gy CTV70 Gy PTV63 Gy CTV63 Gy
V70 Gy V70 Gy V63 Gy V63 Gy
Z 95% ¼ 100% Z 95% ¼ 100%
Table 4 Approved criteria (structure) for DVH
Technique
Beam design
Structure
Criterion
F10 F7 F7E F9E FP
1551, 1301, 751, 251, 01 L, 01 R, 3351, 2851, 2301, and 2051 1551, 1301, 751, 01, 2851, 2301, and 2051 1501, 1001, 501, 01, 3101, 2601, and 2101 1601, 1201, 801, 401, 01, 3201, 2801, 2401, and 2001 1801, 1501, 1201, 901, 2701, 2401, and 2101
Brain stem Spinal cord Inner ear Parotid glands
D5% o 56 Gy D5% o 49 Gy D5% o 51 Gy D50% o 31.5 Gy V35 Gy o 45%
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]]
4 Table 5 Coverage of targets
Technique (%) F10 V70 Gy V70 Gy V63 Gy V63 Gy No. of
of PTV70 Gy (%) of CTV70 Gy (%) of PTV63 Gy (%) of CTV63 Gy (%) plans approved/10 plans
95.3 99.6 94.7 99.8 9
F7 ⫾ ⫾ ⫾ ⫾
2.1 0.4 0.7 0.3
88.8 99.7 94.6 99.8 6
⫾ ⫾ ⫾ ⫾
12.6 0.4 0.5 0.2
F7E
F9
FP
91.8 ⫾ 7.9 99.7 ⫾ 0.6 94.6 ⫾ 0.5 99.8 ⫾ 0.3 6
92.1 ⫾ 8.8 99.5 ⫾ 1.2 94.5 ⫾ 0.5 99.8 ⫾ 0.2 7
88.9 99.7 94.6 99.7 6
⫾ ⫾ ⫾ ⫾
11.8 0.6 0.6 0.3
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.3 1.7 3.1* 8.3 9.1 7.7
Table 6 Sparing of critical structures Technique (%) F10 D5% of brain stem (Gy) D5% of spinal cord (Gy) D50% of parotid glands (Gy) V35 Gy of parotid glands (%) D5% of right inner ear (Gy) D5% of left inner ear (Gy) Mean dose of trachea-esophagus region (Gy) Mean dose of oral cavity (Gy) n
† ‡
50.1 44.3 30.6 41.3 49.2 49.9
F7 ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.7 1.6 3.1* 6.5 7.9 7.0
50.7 44.4 30.3 40.8 49.5 50.2
49.1 ⫾ 1.9† 36.9 ⫾ 3.0‡
1.9 2.1 3.3 6.7 8.0 7.4
50.9 44.4 31.1 42.2 50.2 49.4
F9 ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
1.9 1.8 3.4 6.8 8.9 8.0
51 44.1 30.5 41.7 48.9 49.3
56.0 ⫾ 2.3† 37.9 ⫾ 2.3‡
FP ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.1 2.0 3.7 7.0 8.0 7.5
51.1 44.3 31.7 41.9 49.0 49.5
54.9 ⫾ 2.1† 40.1 ⫾ 3.7‡
52.4 ⫾ 2.6† 31.0 ⫾ 2.3‡
F10 was statistically significant vs FP (p o 0.05). F10 was statistically significant vs F7, F7E, F9, or FP (p o 0.05). FP was statistically significant vs F10, F7, F7E, or F9 (p o 0.05).
Table 7 Conformity of the plans Conformity index
CI70 Gy CI63 Gy CI49 Gy CI35 Gy
†
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
53.3 ⫾ 2.1† 38.4 ⫾ 1.9‡
The coverage of PTV and CTV and the ratio of approved plans are shown in Table 5. All plans were normalized to the same V63 Gy of PTV63 Gy. In Table 5, the V70 Gy of CTV70 Gy in all techniques was 4 99.5%. The V70 Gy of PTV70 Gy was the highest in the F10 technique. The percentage of the approved plan was the highest in the F10 technique. The sparing of critical structures is shown in Table 6. The D5% of the spinal cord and the brain stem showed no significant difference in each technique. The D50% and V35 Gy of the parotid glands were slightly high in the F7E and FP techniques. In the D50% of the parotid glands, F10 was statistically significantly less than FP. In the F10 technique, the mean dose of the trachea-esophagus region was statistically the least value. In the FP technique, the mean dose of the oral cavity was statistically the least value. Dose conformity was evaluated using 4 conformity indices. The results are shown in Table 7. The CI70 Gy of the F10 technique was slightly better. The CI63 Gy of each technique did not exhibit any difference. The CI49 Gy of the F10 and FP techniques was better. The CI35 Gy of the F10 technique was the best. The CI35 Gy of the F10 technique was also statistically significant in F10 vs F7 and F10 vs F7E. The efficiency of the plans was expressed as monitor unit (MU)/ 2 Gy. The average MU and subfields are shown in Table 8. The average MU of the F10 technique was the least. The average
n
F7E
Technique F10
F7
F7E
F9
FP
1.90 1.31 2.62 4.56*,†
2.03 1.29 2.76 4.9*
1.93 1.32 2.69 4.78†
1.91 1.31 2.71 4.87
2.02 1.32 2.63 4.9
F10 was statistically significant vs F7 (p o 0.05). F10 was statistically significant vs F7E (p o 0.05).
subfields were 10, 8.4, 8.3, 11.3, and 8.1 in the F10, F7, F7E, F9E, and FP techniques, respectively.
Discussion All treatment plans were normalized to a similar coverage of PTV63 Gy. The results of the evaluation are shown in Table 5. In the V70 Gy of PTV70 Gy, only the F10 technique satisfied the criteria and was better than the other techniques. The plans were approved when the target coverage at the V70 Gy of PTV70 Gy was achieved and critical structures were also spared. The ratio of the approved plans in the F10 technique reached 90%. However, the ratio of approved plans in the other techniques ranged from 60% to 70% because of inadequate PTV70 Gy coverage. The coverage of PTV63 Gy was 4 95%, and the dose in the critical structures was relatively increased when renormalization was applied to achieve the coverage of PTV70 Gy in all techniques. Therefore, the appropriate dose coverage and acceptable, spared critical structures were satisfied in our technique. The F10 technique was designed to reduce the dose in the spinal cord and brain stem based on split-opposed fields. Although dose decreases in the spinal cord and the brain stem were not evident, the mean dose decreased significantly in the tracheaesophagus region. In our IMRT optimization for F10 field arrangement, DVH constraints were applied to the spinal cord and brain stem but not to the trachea-esophagus region. Although the F10 technique can reduce the dose in the central axis without the need for specifying constraints to the trachea and esophagus, the Table 8 Efficiency of the plans Technique
MU/2 Gy Total no. of subfields
F10
F7
F7E
F9
FP
547 10
554 8.4
573 8.3
581 11.3
590 8.1
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]]
constraints set was inadequate for the other 4 beam arrangements to reduce dose to the trachea-esophagus region. Conformity indices in the high-dose region and low-dose region were selected to evaluate the dose conformity of plans. For the conformity index in the high-dose region CI70 Gy, the F10 technique was slightly better than the others were. For the conformity index in the low-dose regions CI49 Gy and CI35 Gy, the F10 technique was evidently better than the others. In this study, the split-opposed technique can evidently improve dose conformity in the low-dose region CI35 Gy. A large DMLC field was divided into multiple carriages because of the characteristics of Varian's DMLC device controller. In the F10 technique, the field size was fit to PTV63 Gy with a margin of 5 mm based on our 3D-CRT experience. Three pairs of fields were split manually. As such, the subfield number was fixed in the F10 technique. The total MU in the F10 technique was the least. Hence, the beam delivery efficiency of the F10 technique was better than that of the other techniques. This study indicated that the F10 technique is applicable for the treatment of NPC because of the standard shape and nonsuperficial CTV. The split-opposed technique is useful if critical structures are located in the central axis and adjacent to the PTV.
Conclusions The split-opposed technique was suitable when partial PTV was adjacent and behind critical structures in the beam trajectory. A better coverage in PTV70 Gy and spared central organ were observed in our study. In addition, F10 exhibited the highest level
5
of conformity and delivery efficiency among the 5 field arrangements. Therefore, the F10 technique is recommended as the standard beam geometry for the SIB-IMRT of NPC. References 1. Teo, P.M.L.; Ma, B.B.Y.; Chan, A.T.C.; et al. Radiotherapy for nasopharyngeal carcinoma—Transition from two-dimensional to three-dimensional methods. Radiother. Oncol. 73:163–72; 2004. 2. Halperin, E.C.; Perez, C.A.; Brady, L.W.; et al. Perez and Brady's Principles and Practice of Radiation Oncology. 5th ed., Philadelphia, PA, 2007. 3. Clark, C.H.; Bidmead, A.M.; Mubata, C.D.; et al. Intensity-modulated radiotherapy improves target coverage, spinal cord sparing and allows dose escalation in patients with locally advanced cancer of the larynx. Radiother. Oncol. 70:189–98; 2004. 4. Orlandi, E.; Palazzi, M.; Pignoli, E.; et al. Radiobiological basis and clinical results of the simultaneous integrated boost (SIB) in intensity modulated radiotherapy (IMRT) for head and neck cancer: A review. Crit. Rev. Oncol. Hematol. 73:111–25; 2010. 5. Chen, S.W.; Yang, S.N.; Lin, F.J.; et al. Comparative dosimetric study of two strategies of intensity-modulated radiotherapy in nasopharygeal cancer. Med. Dosim. 30:219–27; 2005. 6. Hunt, M.A.; Zelefsky, M.J.; Leibel, S.A.; et al. Treatment planning and delivery of intensity-modulated radiation therapy for primary nasopharynx cancer. Int. J. Radiat. Oncol. Biol. Phys. 49:623–32; 2001. 7. Lee, N.; Mechalakos, J.; Hunt, M.A.; et al. Choosing an intensity-modulated radiation therapy technique in the treatment of head-and-neck cancer. Int. J. Radiat. Oncol. Biol. Phys. 68:1299–309; 2007. 8. Lee, N.; Xia, P.; Fu, K.K.; et al. Intensity-modulated radiotherapy in the treatment of nasopharyngeal carcinoma: An update of the UCSF experience. Int. J. Radiat. Oncol. Biol. Phys. 53:12–22; 2002. 9. Paddick, I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg. 93(suppl. 3):219–22; 2000. 10. Singla, R.; King, S.; Albuquerque, K.; et al. Simultaneous-integrated boost intensity-modulated radiation therapy (SIB-IMRT) in the treatment of earlystage left-sided breast carcinoma. Med. Dosim. 31:190–6; 2006.
Medical Dosimetry journal homepage: www.meddos.org
Technical report
Optimal beam design on intensity-modulated radiation therapy with simultaneous integrated boost in nasopharyngeal cancer Mei-Chun Cheng, M.S.,†,‡ Yu-Wen Hu, M.D.,* Ching-Sheng Liu Ph.D.,* Jeun-Shenn Lee, Ph.D.,† Pin-I Huang, M.D.,* Sang-Hue Yen, M.D.,* Yuh-Lin Lee, M.S.,* Chun-Mei Hsieh, B.S.,* and Cheng-Ying Shiau, M.D.* Division of Radiation Oncology, Department of Oncology Medicine, Taipei Veterans General Hospital, Taipei, Taiwan; †Department of Biomedical Imaging and Radiological Sciences, National Yang-Ming University, Taipei, Taiwan; and ‡Department of Radiation Oncology, Taichung Veterans General Hospital, Taichung, Taiwan
*
A R T I C L E I N F O
A B S T R A C T
Article history: Received 6 January 2014 Accepted 20 March 2014
This study aims to determine the optimal beam design among various combinations of field numbers and beam trajectories for intensity-modulated radiation therapy (IMRT) with simultaneous integrated boost (SIB) technique for the treatment of nasopharyngeal cancer (NPC). We used 10 fields with gantry angles of 1551, 1301, 751, 251, 01 L, 01 R, 3351, 2851, 2301, and 2051 denoted as F10. To decrease doses in the spinal cord, the F10 technique was designed by featuring 2 pairs of split-opposed beam fields at 1551 to 3351 and 2051 to 251, as well as one pair of manually split beam fields at 01. The F10 technique was compared with 4 other common field arrangements: F7E, 7 fields with 501 equally spaced gantry angles; F7, the basis of F10 with 1551, 1301, 751, 01, 2851, 2301, and 2051; F9E, 9 fields with 401 equally spaced gantry angles; and FP, 7 posterior fields with 1801, 1501, 1201, 901, 2701, 2401, and 2101. For each individual case of 10 patients, the customized constraints derived after optimization with the standard F10 technique were applied to 4 other field arrangements. The 4 new optimized plans of each individual case were normalized to achieve the same coverage of planning target volume (PTV)63 Gy as that of the standard F10 technique. The F10 field arrangement exhibited the best coverage in PTV70 Gy and the least mean dose in the trachea-esophagus region. Furthermore, the F10 field arrangement demonstrated the highest level of conformity in the low-dose region and the least monitor unit. The F10 field arrangement performed more outstandingly than the other field arrangements in PTV70 Gy coverage and spared the central organ. This arrangement also exhibited the highest conformity and delivery efficiency. The F10 technique is recommended as the standard beam geometry for the SIB-IMRT of NPC. & 2014 American Association of Medical Dosimetrists.
Keywords: Intensity-modulated radiation therapy (IMRT) Simultaneous integrated boost (SIB) Nasopharyngeal cancer (NPC) Beam design
Introduction According to the World Health Organization, the incidence rate of nasopharyngeal cancer (NPC) is o 1 case per 100,000 cases worldwide. However, the incidence rate is 4 20 cases per 100,000 cases in Southern Asia. In Taiwan, the incidence rate is approximately 8.29 and 2.77 for men and women, respectively, based on a 2008 report
Reprint requests to: Cheng-Ying Shiau, Division of Radiation Oncology, Department of Oncology Medicine, Taipei Veterans General Hospital, No. 201, Sec. 2, Shihpai Rd, Beitou District, Taipei City 112, Taiwan. Tel.: þ88 6228 757 270, ext: 211; fax: þ88 6228 749 425. E-mail: [email protected]. http://dx.doi.org/10.1016/j.meddos.2014.03.003 0958-3947/Copyright Ó 2014 American Association of Medical Dosimetrists
of the Cancer Registry Department of Health and Welfare, Executive Yuan, Taiwan. NPC is radiosensitive, hence radiotherapy or combined radiotherapy and chemotherapy is the standard method. With the advancement of 3-dimensional conformal radiotherapy (3D-CRT) to intensity-modulated radiation therapy (IMRT), tumor coverage has been improved and critical structures have been clearly spared.1 Therefore, the IMRT technique is recommended for the treatment of head and neck cancer.2 Clark et al.3 observed that the low-dose volume in IMRT is higher than that in 3D-CRT. IMRT has been used for treatment of NPC at Taipei Veterans General Hospital (VGH) since 2002. We found that the proper arrangement of beams not only achieved the plan criteria but also improved dose conformity in high-dose and low-
2
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]] corresponding levels. Level Ib was included only when definitive lymphadenopathy was observed. PTV administered at 63 Gy (PTV63 Gy) was created with margins of 3 mm around the upper CTV63 Gy above the C2-3 junction and 5 mm around the lower CTV63 Gy. The planning organ-at-risk volume was also created around the brain stem and at 10 cm of the spinal cord with margins of 3 and 5 mm, respectively. The dose constraints also included the inner ears and whole parotid glands. With a chin-up position, the eyes and optic nerves were excluded from the constraints, except for advanced T4 cases. These structures are shown in Fig. 2. IMRT planning technique
Fig. 1. Schematic of the beam arrangement. The regions of nasal cavity and neck in the axial view are shown in left side and right side of figures, respectively. The red area indicates the PTV, the brown area indicates the brain stem or the spinal cord, and the blue area indicates the trachea-esophagus region. (A) The PTV is covered by a beam at a gantry angle of 1551. (B) The PTV is covered by a partial beam at a gantry angle of 1551 and the split-opposed beam at 3351. (Color version of figure is available online.)
dose volumes. In our study, manually split or opposing partial fields were designed based on the relative position between planning target volume (PTV) and critical structures. Critical structures receive higher doses than the PTV if they are located in front of the PTV along the beam trajectory because of the physical characteristics of a photon beam. A nonopposing field arrangement is the basic principle of beam angle selection in IMRT. However, not all critical structures of each individual field could be spared or placed at the back of the PTV. After analyzing the geometric characteristics of critical structures vs the PTV of NPC, we designed the F10 beam arrangement with the original F7 fields split and partially placed in opposing trajectories to relocate the critical structures at the back of the PTV (Fig. 1). The 01 field was also split in half in our F10 field arrangement to decrease dose leakage to the larynx, trachea, and spinal cord of the dynamic multileaf collimator (DMLC) motion. This study aims to confirm that the Taipei VGH F10 field arrangement is a better IMRT beam design for NPC than 4 other popular field arrangements.
The SIB-IMRT for NPC at Taipei VGH had 2 dose levels: 70 Gy to CTV70 Gy and 63 Gy to CTV63 Gy in 35 fractions. The field arrangement was 10 fields, denoted as F10, which was developed in the era of Varian’s CadPlan Radiotherapy Treatment Planning System version 6.32 (Table 2). In this study, the IMRT was planned by Varian's Eclipse Radiotherapy Treatment Planning System version 7.1. The sweeping step-and-shoot technique with 10 dose levels was adopted to convert the optimal fluence into segments of leaf motion. The schematic of our idea is shown in Fig. 1. In Fig. 1A, the right side of the PTV at a beam trajectory of gantry 1551 is found behind critical structures. If the right side of the PTV was covered with sufficient doses, the critical structures in front of the PTV receive a higher dose than the PTV. Considering this finding, we developed a splitopposed field technique (Fig. 1B). Half of the field that covered the right side of the PTV was opposed to the gantry angle at 3351; hence, the right side of the PTV was proximal to the source at 3351. Considering this idea, we featured 2 pairs of splitopposed fields at gantry angles of 1551 to 3351 and 2051 to 251. Furthermore, the large field size was divided into multiple subfields in Varian's system. In our study, 01 L to 01 R were manually split to prevent leakage to the central structures. The IMRT covered the nasopharynx and the upper neck, and the anterior field covered the lower neck. The center of the IMRT field was placed at the middle neck junction with the half-beam lower neck single anterior field, which was not compared in this study. The F10 technique was compared with 4 other popular field arrangements, as shown in Table 2. The 4 other field arrangements were F7, F7E, F9E, and FP. The F7 technique, the predecessor of the F10 technique, exhibits a similar gantry angle, except for split-opposed fields. In the F7E technique, 7 equally spaced gantry angles are considered.5 By comparison, the F9E technique uses 9 equally spaced gantry angles. The FP technique uses 7 posterior gantry angles.6-8 A standard template of dose-volume histogram (DVH) constraints for the inverse treatment plan of the SIB-IMRT of NPC was routinely applied for actual treatment planning using the F10 beam arrangement. The constraints of the standard template were adjusted, if desired, during the optimization process to achieve a satisfactory plan for each patient. Derived from the approved inverse optimization planning with the F10 technique for actual clinical treatment, the customized constraints were then applied to the 4 other techniques to create the respective SIB-IMRT plans for the 10 patients with NPC included in this study. The approved criteria for DVH are shown in Tables 3 and 4. Analysis The 4 new optimized plans of each individual case were normalized to obtain the same coverage of PTV63 Gy as that of the F10 technique. The DVH of targets and other critical organs were compared among the 5 techniques (Tables 5 and 6). Plans were rejected because of inadequate coverage of PTV70 Gy or excessive dose to the spinal cord or brain stem. The conformity index9 was used to evaluate plan conformity. The equation is expressed as follows: CI ¼
¼ Methods and Materials Patient selection and structure delineation During 2005, 10 consecutive patients with NPC who were treated with 10-field simultaneous integrated boost (SIB)4 IMRT at Taipei VGH were selected for this study. The patients immobilized with a thermoplastic mask were subjected to computed tomography simulation (5 mm/slice). The TNM categories based on the American Joint Committee on Cancer sixth edition cancer staging system for the 10 patients are shown in Table 1. Gross tumor volume was defined using 5 sets of magnetic resonance image fusion, including axial T2-weighted image (WI), axial T1WI, postcontrast axial T1 with fat saturation, postcontrast coronal T1WI, and sagittal T1WI. Clinical target volume (CTV)70 Gy was defined by gross tumor volume of both primary tumor and regional lymphadenopathy. PTV was administered at 70 Gy (PTV70 Gy) and created by adding margins of 3 mm for the primary tumor and margins of 5 mm for the neck lymphadenopathy. CTV63 Gy encompassed the CTV70 Gy, whole nasopharynx, skull base, bilateral tonsils, and at least level II of lymphatics. The CTV63 Gy extended to level III or level IV for patients with gross lymphadenopathy in the
RV 1 PTV in RV PTV in RV=PTV RV PTV
1 PTV in RV=PTV
ð1Þ
2 ð2Þ
where RV is the radiation volume. Overall, 4 conformity indices, namely, CI70 Gy, CI63 Gy, CI49 Gy, and CI35 Gy, were defined in our study. CI70 Gy and CI63 Gy are the conformity indices of 70 and 63 Gy, respectively and are described as the dose conformity to high-dose regions (Eqs. (3) and (4)). CI49 Gy and CI35 Gy are the Table 1 TNM stages of the 10 patients in our study T stage
N stage 0
1 2 3 4 Total
1 2 3
1
2
1 3
1
1 5
1 2
3
Total
0
1 4 1 4 10
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]]
3
Fig. 2. Definition of structures included PTV70 Gy (red segment), PTV63 Gy (light green segment), parotid gland (purple segment), brain stem (blue segment), and spinal cord (blue segment). The typical dose distributions included 7000, 6300, 5600, 4900, 4200, and 3500 cGy. (Color version of figure is available online.) conformity indices of 49 and 35 Gy, respectively, and are described as the dose conformity to low-dose regions (Eqs. (5) and (6)). CI70
Gy
¼
RV70 Gy PTV70 Gy
CI63
Gy
¼
RV63 Gy PTV63 Gy
V70
Gy
1 of PTV70
Gy
1 of PTV63
V63
CI49
Gy
RV49 Gy ¼ PTV63 Gy
CI35
Gy
¼
2 ð3Þ
Gy
2 Gy
Results A typical dose distribution is shown in Fig. 2. Dose distributions include the coverage dose of PTV70 Gy (7000 cGy), coverage dose of PTV63 Gy (6300 cGy), and sparing doses of the brain stem (5600 cGy), spinal cord (4900 cGy), and parotid gland (3500 cGy).
ð4Þ Table 3 Approved criteria (target) for DVH ð5Þ
RV35 Gy PTV63 Gy
ð6Þ
We performed repeated-measures analysis of variance by using SPSS 13.0 (SPSS Inc., Chicago, IL) to detect the statistical difference between the 5 techniques. Bonferroni-corrected post hoc t-tests were used in pairwise comparison.10 Table 2 Beam trajectory
Target
Criterion
PTV70 Gy CTV70 Gy PTV63 Gy CTV63 Gy
V70 Gy V70 Gy V63 Gy V63 Gy
Z 95% ¼ 100% Z 95% ¼ 100%
Table 4 Approved criteria (structure) for DVH
Technique
Beam design
Structure
Criterion
F10 F7 F7E F9E FP
1551, 1301, 751, 251, 01 L, 01 R, 3351, 2851, 2301, and 2051 1551, 1301, 751, 01, 2851, 2301, and 2051 1501, 1001, 501, 01, 3101, 2601, and 2101 1601, 1201, 801, 401, 01, 3201, 2801, 2401, and 2001 1801, 1501, 1201, 901, 2701, 2401, and 2101
Brain stem Spinal cord Inner ear Parotid glands
D5% o 56 Gy D5% o 49 Gy D5% o 51 Gy D50% o 31.5 Gy V35 Gy o 45%
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]]
4 Table 5 Coverage of targets
Technique (%) F10 V70 Gy V70 Gy V63 Gy V63 Gy No. of
of PTV70 Gy (%) of CTV70 Gy (%) of PTV63 Gy (%) of CTV63 Gy (%) plans approved/10 plans
95.3 99.6 94.7 99.8 9
F7 ⫾ ⫾ ⫾ ⫾
2.1 0.4 0.7 0.3
88.8 99.7 94.6 99.8 6
⫾ ⫾ ⫾ ⫾
12.6 0.4 0.5 0.2
F7E
F9
FP
91.8 ⫾ 7.9 99.7 ⫾ 0.6 94.6 ⫾ 0.5 99.8 ⫾ 0.3 6
92.1 ⫾ 8.8 99.5 ⫾ 1.2 94.5 ⫾ 0.5 99.8 ⫾ 0.2 7
88.9 99.7 94.6 99.7 6
⫾ ⫾ ⫾ ⫾
11.8 0.6 0.6 0.3
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.3 1.7 3.1* 8.3 9.1 7.7
Table 6 Sparing of critical structures Technique (%) F10 D5% of brain stem (Gy) D5% of spinal cord (Gy) D50% of parotid glands (Gy) V35 Gy of parotid glands (%) D5% of right inner ear (Gy) D5% of left inner ear (Gy) Mean dose of trachea-esophagus region (Gy) Mean dose of oral cavity (Gy) n
† ‡
50.1 44.3 30.6 41.3 49.2 49.9
F7 ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.7 1.6 3.1* 6.5 7.9 7.0
50.7 44.4 30.3 40.8 49.5 50.2
49.1 ⫾ 1.9† 36.9 ⫾ 3.0‡
1.9 2.1 3.3 6.7 8.0 7.4
50.9 44.4 31.1 42.2 50.2 49.4
F9 ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
1.9 1.8 3.4 6.8 8.9 8.0
51 44.1 30.5 41.7 48.9 49.3
56.0 ⫾ 2.3† 37.9 ⫾ 2.3‡
FP ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
2.1 2.0 3.7 7.0 8.0 7.5
51.1 44.3 31.7 41.9 49.0 49.5
54.9 ⫾ 2.1† 40.1 ⫾ 3.7‡
52.4 ⫾ 2.6† 31.0 ⫾ 2.3‡
F10 was statistically significant vs FP (p o 0.05). F10 was statistically significant vs F7, F7E, F9, or FP (p o 0.05). FP was statistically significant vs F10, F7, F7E, or F9 (p o 0.05).
Table 7 Conformity of the plans Conformity index
CI70 Gy CI63 Gy CI49 Gy CI35 Gy
†
⫾ ⫾ ⫾ ⫾ ⫾ ⫾
53.3 ⫾ 2.1† 38.4 ⫾ 1.9‡
The coverage of PTV and CTV and the ratio of approved plans are shown in Table 5. All plans were normalized to the same V63 Gy of PTV63 Gy. In Table 5, the V70 Gy of CTV70 Gy in all techniques was 4 99.5%. The V70 Gy of PTV70 Gy was the highest in the F10 technique. The percentage of the approved plan was the highest in the F10 technique. The sparing of critical structures is shown in Table 6. The D5% of the spinal cord and the brain stem showed no significant difference in each technique. The D50% and V35 Gy of the parotid glands were slightly high in the F7E and FP techniques. In the D50% of the parotid glands, F10 was statistically significantly less than FP. In the F10 technique, the mean dose of the trachea-esophagus region was statistically the least value. In the FP technique, the mean dose of the oral cavity was statistically the least value. Dose conformity was evaluated using 4 conformity indices. The results are shown in Table 7. The CI70 Gy of the F10 technique was slightly better. The CI63 Gy of each technique did not exhibit any difference. The CI49 Gy of the F10 and FP techniques was better. The CI35 Gy of the F10 technique was the best. The CI35 Gy of the F10 technique was also statistically significant in F10 vs F7 and F10 vs F7E. The efficiency of the plans was expressed as monitor unit (MU)/ 2 Gy. The average MU and subfields are shown in Table 8. The average MU of the F10 technique was the least. The average
n
F7E
Technique F10
F7
F7E
F9
FP
1.90 1.31 2.62 4.56*,†
2.03 1.29 2.76 4.9*
1.93 1.32 2.69 4.78†
1.91 1.31 2.71 4.87
2.02 1.32 2.63 4.9
F10 was statistically significant vs F7 (p o 0.05). F10 was statistically significant vs F7E (p o 0.05).
subfields were 10, 8.4, 8.3, 11.3, and 8.1 in the F10, F7, F7E, F9E, and FP techniques, respectively.
Discussion All treatment plans were normalized to a similar coverage of PTV63 Gy. The results of the evaluation are shown in Table 5. In the V70 Gy of PTV70 Gy, only the F10 technique satisfied the criteria and was better than the other techniques. The plans were approved when the target coverage at the V70 Gy of PTV70 Gy was achieved and critical structures were also spared. The ratio of the approved plans in the F10 technique reached 90%. However, the ratio of approved plans in the other techniques ranged from 60% to 70% because of inadequate PTV70 Gy coverage. The coverage of PTV63 Gy was 4 95%, and the dose in the critical structures was relatively increased when renormalization was applied to achieve the coverage of PTV70 Gy in all techniques. Therefore, the appropriate dose coverage and acceptable, spared critical structures were satisfied in our technique. The F10 technique was designed to reduce the dose in the spinal cord and brain stem based on split-opposed fields. Although dose decreases in the spinal cord and the brain stem were not evident, the mean dose decreased significantly in the tracheaesophagus region. In our IMRT optimization for F10 field arrangement, DVH constraints were applied to the spinal cord and brain stem but not to the trachea-esophagus region. Although the F10 technique can reduce the dose in the central axis without the need for specifying constraints to the trachea and esophagus, the Table 8 Efficiency of the plans Technique
MU/2 Gy Total no. of subfields
F10
F7
F7E
F9
FP
547 10
554 8.4
573 8.3
581 11.3
590 8.1
M. Cheng et al. / Medical Dosimetry ] (2014) ]]]–]]]
constraints set was inadequate for the other 4 beam arrangements to reduce dose to the trachea-esophagus region. Conformity indices in the high-dose region and low-dose region were selected to evaluate the dose conformity of plans. For the conformity index in the high-dose region CI70 Gy, the F10 technique was slightly better than the others were. For the conformity index in the low-dose regions CI49 Gy and CI35 Gy, the F10 technique was evidently better than the others. In this study, the split-opposed technique can evidently improve dose conformity in the low-dose region CI35 Gy. A large DMLC field was divided into multiple carriages because of the characteristics of Varian's DMLC device controller. In the F10 technique, the field size was fit to PTV63 Gy with a margin of 5 mm based on our 3D-CRT experience. Three pairs of fields were split manually. As such, the subfield number was fixed in the F10 technique. The total MU in the F10 technique was the least. Hence, the beam delivery efficiency of the F10 technique was better than that of the other techniques. This study indicated that the F10 technique is applicable for the treatment of NPC because of the standard shape and nonsuperficial CTV. The split-opposed technique is useful if critical structures are located in the central axis and adjacent to the PTV.
Conclusions The split-opposed technique was suitable when partial PTV was adjacent and behind critical structures in the beam trajectory. A better coverage in PTV70 Gy and spared central organ were observed in our study. In addition, F10 exhibited the highest level
5
of conformity and delivery efficiency among the 5 field arrangements. Therefore, the F10 technique is recommended as the standard beam geometry for the SIB-IMRT of NPC. References 1. Teo, P.M.L.; Ma, B.B.Y.; Chan, A.T.C.; et al. Radiotherapy for nasopharyngeal carcinoma—Transition from two-dimensional to three-dimensional methods. Radiother. Oncol. 73:163–72; 2004. 2. Halperin, E.C.; Perez, C.A.; Brady, L.W.; et al. Perez and Brady's Principles and Practice of Radiation Oncology. 5th ed., Philadelphia, PA, 2007. 3. Clark, C.H.; Bidmead, A.M.; Mubata, C.D.; et al. Intensity-modulated radiotherapy improves target coverage, spinal cord sparing and allows dose escalation in patients with locally advanced cancer of the larynx. Radiother. Oncol. 70:189–98; 2004. 4. Orlandi, E.; Palazzi, M.; Pignoli, E.; et al. Radiobiological basis and clinical results of the simultaneous integrated boost (SIB) in intensity modulated radiotherapy (IMRT) for head and neck cancer: A review. Crit. Rev. Oncol. Hematol. 73:111–25; 2010. 5. Chen, S.W.; Yang, S.N.; Lin, F.J.; et al. Comparative dosimetric study of two strategies of intensity-modulated radiotherapy in nasopharygeal cancer. Med. Dosim. 30:219–27; 2005. 6. Hunt, M.A.; Zelefsky, M.J.; Leibel, S.A.; et al. Treatment planning and delivery of intensity-modulated radiation therapy for primary nasopharynx cancer. Int. J. Radiat. Oncol. Biol. Phys. 49:623–32; 2001. 7. Lee, N.; Mechalakos, J.; Hunt, M.A.; et al. Choosing an intensity-modulated radiation therapy technique in the treatment of head-and-neck cancer. Int. J. Radiat. Oncol. Biol. Phys. 68:1299–309; 2007. 8. Lee, N.; Xia, P.; Fu, K.K.; et al. Intensity-modulated radiotherapy in the treatment of nasopharyngeal carcinoma: An update of the UCSF experience. Int. J. Radiat. Oncol. Biol. Phys. 53:12–22; 2002. 9. Paddick, I. A simple scoring ratio to index the conformity of radiosurgical treatment plans. J Neurosurg. 93(suppl. 3):219–22; 2000. 10. Singla, R.; King, S.; Albuquerque, K.; et al. Simultaneous-integrated boost intensity-modulated radiation therapy (SIB-IMRT) in the treatment of earlystage left-sided breast carcinoma. Med. Dosim. 31:190–6; 2006.