- Email: [email protected]
Comparison of a new noncoplanar intensity-modulated radiation therapy technique for craniospinal irradiation with 3 coplanar techniques
Medical Dosimetry 40 (2015) 296–303
Medical Dosimetry journal homepage: www.meddos.org
Comparison of a new noncoplanar intensity-modulated radiation therapy technique for craniospinal irradiation with 3 coplanar techniques Anders T. Hansen, M.S.,* Slavka Lukacova, M.D., Ph.D.,† Yasmin Lassen-Ramshad, M.D., Ph.D.,† and Jørgen B. Petersen, M.S.* Department of Medical Physics, Aarhus University Hospital, Nørrebrogade 44, Building 5, DK-8000 Aarhus C, Denmark; and †Department of Oncology, Aarhus University Hospital, Nørrebrogade 44, Building 5, DK-8000 Aarhus C, Denmark
*
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 September 2014 Received in revised form 15 February 2015 Accepted 19 March 2015
When standard conformal x-ray technique for craniospinal irradiation is used, it is a challenge to achieve satisfactory dose coverage of the target including the area of the cribriform plate, while sparing organs at risk. We present a new intensity-modulated radiation therapy (IMRT), noncoplanar technique, for delivering irradiation to the cranial part and compare it with 3 other techniques and previously published results. A total of 13 patients who had previously received craniospinal irradiation with standard conformal x-ray technique were reviewed. New treatment plans were generated for each patient using the noncoplanar IMRT-based technique, a coplanar IMRT-based technique, and a coplanar volumetric-modulated arch therapy (VMAT) technique. Dosimetry data for all patients were compared with the corresponding data from the conventional treatment plans. The new noncoplanar IMRT technique substantially reduced the mean dose to organs at risk compared with the standard radiation technique. The 2 other coplanar techniques also reduced the mean dose to some of the critical organs. However, this reduction was not as substantial as the reduction obtained by the noncoplanar technique. Furthermore, compared with the standard technique, the IMRT techniques reduced the total calculated radiation dose that was delivered to the normal tissue, whereas the VMAT technique increased this dose. Additionally, the coverage of the target was significantly improved by the noncoplanar IMRT technique. Compared with the standard technique, the coplanar IMRT and the VMAT technique did not improve the coverage of the target significantly. All the new planning techniques increased the number of monitor units (MU) used—the noncoplanar IMRT technique by 99%, the coplanar IMRT technique by 122%, and the VMAT technique by 26%—causing concern for leak radiation. The noncoplanar IMRT technique covered the target better and decreased doses to organs at risk compared with the other techniques. All the new techniques increased the number of MU compared with the standard technique. & 2015 American Association of Medical Dosimetrists.
Keywords: Craniospinal irradiation IMRT Medulloblastoma Pediatric radiotherapy
Introduction Medulloblastoma represents approximately 20% of intracranial malignant tumors in children.1-3 This disease typically spreads through the cerebrospinal fluid within the central nervous system. The standard treatment consists of maximal surgical resection followed by radiation therapy and chemotherapy.1,3 The whole craniospinal axis is treated with radiation followed by a boost in the posterior fossa. The standard radiotherapy technique used for craniospinal irradiation consists of 2 almost opposing lateral
Reprint requests to: Anders Traberg Hansen, Department of Medical Physics, Aarhus University Hospital, Nørrebrogade 44, 8000 Arhus C, Denmark. E-mail: [email protected] http://dx.doi.org/10.1016/j.meddos.2015.03.007 0958-3947/Copyright Ó 2015 American Association of Medical Dosimetrists
photon fields, rotated slightly in the anterior direction, to place the anterior field edges just behind both lenses, as seen in Fig. 1B. To these 2 lateral fields, several posterior fields are typically added end to end to irradiate the spine from the caudal field edge of the 2 cranial fields to the sacrum.4-8 It has previously been shown that sufficient dose coverage of the whole central nervous system, including the cribriform plate, is essential in preventing recurrence of the disease.9,10 The disadvantage of the standard conformal xray technique is inadequate target coverage, mainly of the cribriform plate, when certain organs at risk like the parotid glands, the inner ears, or the lenses are to be spared.11 The risk of xerostomia increases when the parotid glands receive doses more than 25 Gy.12,13 The risk of hearing loss has been reported to increase when doses to the inner ears are more than 35 Gy.14,15 Especially for
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
297
Fig. 1. The field arrangements for the 4 treatment techniques are displayed. In the top left illustration is the noncoplanar IMRT treatment technique (A) and in the top right the standard technique (B). In the bottom left is the coplanar IMRT technique (C) and in the bottom right the VMAT technique (D).
young children, the cribriform plate tends to be more anteriorly located, which is why a full irradiation of the lenses often is unavoidable. If doses more than 0.5 Gy are given to the lenses, there is an increased risk for development of radiation-induced cataract.16 Additionally, radiotherapy for medulloblastoma may lead to a series of other late side effects.1,2 Therefore, a radiation technique that is able to cover the target with a sufficient dose and at the same time limits the doses given to the described organs at risk is needed.17 This can be achieved using proton therapy.18 However, the use of proton therapy is associated with large economic cost and limited capacity. Therefore, we propose a new noncoplanar intensitymodulated radiation therapy (IMRT)–based technique for craniospinal irradiation, which improves the dose coverage of the target and reduces the dose to the described critical organs.
organs at risk. The CTV consisted of the whole brain including the cribriform plate and the spinal canal. The planning target volume (PTV) was generated by adding a margin of 5 mm to the CTV; however, a margin of 6 mm in the cranial and caudal direction was used because of the uncertainty introduced by the spacing between the CT slices. The critical organs defined for this study were as follows: the parotid glands, the inner ears, the lenses, and the frontal and posterior parts of the eyes. To
Methods and Materials Treatment plans for 13 patients who had been treated for medulloblastoma, at our hospital, in the time period 2007 to 2013 with the standard conformal x-ray technique were collected for this study. In total, 6 patients were children (2 to 17 years), 2 patients were young adults (18 to 30 years), and 5 were adults (30 to 60 years). In the treatment plans, the cranial part as well as the spinal part of the craniospinal treatment was point normalized for all patients. The cranial normalization points were in all cases located centrally in the brain. The spinal normalization points were in all cases placed at representative positions in the medulla spinalis. The original spinal part of the treatment plans was left almost unchanged, but the cranial part of the treatment plans was replanned using 3 planning techniques: our novel noncoplanar IMRT, a coplanar IMRT, and a coplanar volumetric-modulated arch therapy (VMAT) technique.
Delineation The original computed tomography (CT) scans were in 1 case performed with a slice thickness of 4.5 mm and in 12 cases with a slice thickness of 3 mm. They encompassed the whole cranium, the thorax, and the abdominal and pelvic regions. Typically, parts of the arms and legs were not included in the CT scans. The CT scans were used for delineation of the clinical target volume (CTV) and
Fig. 2. An example of the spinal part of the original conventional treatment plan. Notice the 6 segments in the cervical junction zone.
298
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
Fig. 3. Digitally reconstructed radiographs for the 5 fields constituting the noncoplanar technique. The 2 images in the top of the figure are the coplanar fields and the lower 3 are the noncoplanar. The eyes have been indicated.
increase the clarity and simplicity of the analysis, organs of the same type on the left and right side were unified into a single volume. The volume, referred to as normal tissue, was defined as the total body covered by the CT scan, but outside of the PTV. In addition, the structure of foramen magnum was introduced as an anatomical landmark for the orientation of the noncoplanar fields. The part of the PTV extending cranially from the foramen magnum was also defined as an independent structure called the cranial PTV, to evaluate the capability of the different techniques to cover this part.
Dose calculation and treatment planning The commercial treatment planning system Eclipse version 11.3.31 (Varian Medical Systems Inc.) and the convolution-superposition dose calculation algorithm AAA were used for all treatment plans. Originally, the spinal axis was treated for all patients according to the standard technique with several posterior fields. Overall, 2 to 3 fields, consisting of 3 segments each, were used to cover the whole length of the spine. The field junctions moved over 3 positions, separated with the gap of 1.5 to 2.5 cm, to smear out the dose heterogeneities in the target and to increase the plan robustness for
errors of field abutment. Several low-weight segments were used to account for the depth variation of the spine. The spinal part of the original conventional treatment plan was used in all new treatment plans. However, the number of segments in the most cranial spinal field was doubled to smooth the longitudinal dose gradient in the cervical junction zone, as displayed in Fig. 2. The dose given to the target by the spinal part of the treatment plan served as the basis for the optimization of the new plans. This approach of using the optimization algorithm to produce a smooth dose transition in the junction zone has previously been demonstrated to be both safe and efficient.1922 All treatment planning techniques for the cranial part were symmetrical in the sagittal plane, and 6-MV photons were used for all fields, as shown in Fig. 1. Noncoplanar IMRT technique: In total, 5 IMRT fields including 3 noncoplanar fields were used as shown in Fig. 1A. The central axis of the treatment fields was set to be in a plane defined by the cribriform plate and the foramen magnum. The treatment fields were placed within this plane to obtain a sufficient coverage of the cribriform plate. This approach allowed keeping the volume of irradiated normal tissue outside the brain low. Moreover, 2 of the fields were coplanar and defined by gantry position of 751 and 2851. The angles for the gantry and table rotation of the noncoplanar fields depended strongly on the anatomy and fixation of the individual patient. The jaws of the noncoplanar fields were fixed to avoid
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
299
Fig. 4. An example of dose distributions near the eyes and the cribriform plate for the noncoplanar (A), the standard (B), the coplanar (C), and the VMAT (D) techniques. The PTV is drawn in solid black. The 50% isodose curve is drawn in white, the 95% isodose curve is drawn in dark gray, and the 105% isodose curve is if present drawn in light gray. unnecessary irradiation to the normal tissue in the front of the head. The 2 coplanar fields were used for the treatment of the upper part of the spinal cord. A set of beams-eye views from the fields of a noncoplanar plan is displayed in Fig. 3. Because all the noncoplanar fields were planned from an anterior and superior direction, it is considered unlikely, that collisions between the gantry and the table or the patient would occur, using this technique. Coplanar IMRT technique: This technique consisted of 5 fields produced by projecting the 3 noncoplanar fields from the noncoplanar plan onto the transversal plane of the patient (Fig. 1C). The jaws of the 3 frontal fields were fixed just caudally of the eyes to limit the unnecessary irradiation of the frontal part of the head. Coplanar VMAT technique: This technique consisted of 2 intensity-modulated arcs going clockwise from 1821 to 01 and then to 1781 and the same way counterclockwise as previously described by Studenski et al.23 The collimator was set in 451 and 3151 for the 2 arcs (Fig. 1D). Planning optimization: The same set of optimization objectives was used for all treatment plans in the optimization process, with the exception of virtual volumes. Virtual volumes were defined as the volumes of either an overdose or underdose in
the initiation of the optimization process. The fluences of the IMRT fields, which had fixed jaws, were edited by hand, before the start of a subsequent optimization to avoid steep gradients and an inhomogeneous dose distribution. At least 2 subsequent optimizations were performed for all plans. The new treatment plans irradiating the brain and the cranial part of the spinal cord were all normalized to the mean dose in the part of the CTV, which was located cranially off the foramen magnum, which was equal to the brain. A final treatment plan, consisting of the sum of the new treatment plan and the modified spinal part of the original plan, was produced for each patient. An illustration of calculated doses near the eyes and the cribriform plate can be seen in Fig. 4. The prescribed dose was set to be 36 Gy in 20 fractions, which is one of the prescriptions used for treating medulloblastoma.2,22,23
Plan evaluation The mean doses to the parotid glands, the eyes, the lenses, the inner ears, the frontal and posterior parts of the eyes, and the normal tissues and the number of
Table 1 The mean dose to the critical organs and the number of monitor units are shown for each treatment planning technique Standard Parotid gland Lens Eye Frontal eye Posterior eye Ear Body-PTV Monitor units
27.7 (4.9) 11.3 (9.8) 21.8 (6.1) 15.8 (6.1) 27.7 (7.4) 36.5 (0.5) 8.5 (1.4)
Noncoplanar [17.8 to 35.7] [4.1 to 36.6] [12.6 to 32.9] [9.2 to 31.4] [15.0 to 35.9] [35.9 to 37.3] [5.8 to 11.2]
726 (94) [522 to 829]
20.5 (1.5) 5.9 (1.2) 13.7 (2.1) 9.3 (1.6) 18.0 (2.9) 35.0 (0.7) 8.3 (1.4)
[17.7 to 22.6] [3.9 to 7.7] [11.0 to 17.0] [6.8 to 11.9] [14.4 to 23.0] [33.8 to 36.4] [5.8 to 10.8]
1445 (133) [1264 to 1696]
Standard deviations are specified in brackets, and ranges are specified in square brackets.
Coplanar 26.6 (1.5) 9.7 (3.0) 18.9 (2.7) 14.7 (2.7) 23.0 (3.0) 35.8 (0.7) 8.4 (1.4)
VMAT [23.2 to 29.8] [6.0 to 16.2] [13.8 to 23.0] [10.0 to 18.6] [17.6 to 27.4] [34.1 to 36.7] [5.8 to 10.9]
1617 (120) [1449 to 1836]
20.0 (1.4) 10.6 (1.6) 17.6 (2.1) 14.2 (2.0) 20.9 (2.4) 35.7 (0.6) 8.9 (1.5)
[18.5 to 23.3] [7.6 to 13.3] [14.7 to 21.3] [11.1 to 17.4] [17.7 to 24.6] [34.7 to 36.5] [6.0 to 11.5]
916 (77) [751 to 2005]
300
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
Table 2 The values for p from a statistical analysis using Student paired t-test of the data presented in Table 1 Standard Parotid gland Noncoplanar Coplanar VMAT
o0.001* 0.344 o0.001*
Eye Noncoplanar Coplanar VMAT
o0.001* 0.044* 0.022*
Posterior eye Noncoplanar Coplanar VMAT
o0.001* 0.014* 0.003*
Body-PTV Noncoplanar Coplanar VMAT
0.014* 0.009* o0.001*
Noncoplanar
Coplanar
o0.001* 0.212
o0.001* o0.001*
o0.001* o0.001*
0.090 o0.001*
Standard
Noncoplanar
Coplanar
o0.001*
Lens Noncoplanar Coplanar VMAT
0.052 0.473 0.795
o0.001* o0.001*
0.235
0.029*
Frontal eye Noncoplanar Coplanar VMAT
0.002* 0.456 0.367
o0.001* o0.001*
0.389
0.001*
Ear Noncoplanar Coplanar VMAT
o0.001* 0.005* 0.001*
o0.001* 0.001*
0.577
o0.001*
No. of MU Noncoplanar Coplanar VMAT
o0.001* o0.001* o0.001*
o0.001* o0.001*
o0.001*
Significance is marked (*).
monitor units (MU) were calculated for all total treatment plans and compared as illustrated in Table 1. Also, the mean, maximum, and minimum dose to the cranial PTV and the relative volume of it receiving more than 95% and more than 107% of the prescribed dose was also calculated and compared as illustrated in Table 3. Paired 2-tailed Student t-tests were used to compare results between treatment techniques. A confidence level of 5% was regarded statistically significant. The values for p for the correlations of the data in Table 1 can be seen in Table 2. Moreover, p values for the data displayed in Table 3 are in the same manner displayed in Table 4.
Results The noncoplanar IMRT technique significantly reduced the mean dose to the parotid gland, the frontal and posterior eye, the inner ear, and the normal tissue compared with the standard technique as illustrated in Tables 1 and 2. Moreover, this technique was significantly better to spare all critical organs than the coplanar IMRT technique. The technique was also superior to the VMAT technique regarding the mean dose to the lens, the frontal and posterior eye, the inner ear, and the normal tissue, but not regarding the parotid gland, where the results were similar. We observed that the VMAT technique delivered lower mean doses to the parotid gland and posterior eye, but higher doses to the normal tissue, compared with the coplanar IMRT-based technique. Especially regarding the lens, both the noncoplanar IMRT-based technique and VMAT technique reduced the mean dose compared with the standard technique, but this reductions was not found to be statistically significant. However, a major reduction in mean dose to the lens from 22.7 to 7.2 Gy was found for the 4 patients younger than 15 years using the noncoplanar IMRT-based technique relative to the standard technique. The reduction using the coplanar technique was to 13.3 Gy and to 11.3 Gy using VMAT. For adult patients, there was almost no difference in mean dose to the
lens between the noncoplanar IMRT-based technique and the standard technique. The coplanar and the noncoplanar IMRTbased techniques significantly reduced the dose to the normal tissue compared with the standard technique. The dose to the normal tissue was found to be highest using VMAT technique. The dose coverage of the cranial PTV was found to be significantly improved using the noncoplanar IMRT-based technique relative to the standard and coplanar IMRT-based techniques. The relative volume covered by 95% of the prescribed dose is displayed in Table 3. The dose-volume relationship regarding the eyes and the cranial PTV for a typical patient is displayed in Figs. 5 and 6. The number of MU for a whole treatment was found to be lowest for the standard technique, 26% higher for the VMAT, 99% higher for the noncoplanar IMRT-based technique, and 123% higher for the coplanar IMRT-based technique as illustrated in Table 1. Discussion The results of the present study indicate that the new noncoplanar IMRT-based technique for craniospinal irradiation is superior regarding both the mean dose to the critical organs and improvement of the target dose coverage. The dose reduction to the organs at risk may result in the reduction of the incidence and severity of potential side effects, such as xerostomia, hearing loss, and lens cataract. Furthermore, the improvement of the dose coverage of the target may decrease the risk of recurrence and thus improve the survival of patients with medulloblastoma. The fact that the noncoplanar IMRT-based technique was identical to the coplanar IMRT-based technique in all aspects except for the use of noncoplanar fields, but produced significantly superior results, demonstrates the benefit of the noncoplanar treatment
Table 3 Mean values for the doses to the cranial PTV are shown for each treatment planning technique
V95%, % V107%, % Max, Gy Mean, Gy Min, Gy
Standard
Noncoplanar
Coplanar
VMAT
99.56 1.3 39.0 36.70 20.9
99.94 0.002 38.2 36.04 30.2
99.77 0.12 40.1 36.06 28.4
99.80 0.019 39.0 36.03 32.2
(0.33) [99.13 to 99.99] (3.0) [0 to 11.0] (1.1) [37.8 to 41.3] (0.47) [35.81 to 37.42] (9.7) [5.8 to 31.8]
(0.06) [99.75 to 99.99] (0.005) [0 to 0.013] (1.1) [36.8 to 40.1] (0.04) [35.97 to 36.13] (1.8) [27.8 to 32.6]
Standard deviations are specified in brackets, ranges are specified in square brackets.
(0.20) [99.36 to 99.97] (1.20) [0 to 0.70] (1.1) [38.4 to 42.0] (0.04) [36.01 to 36.13] (3.3) [20.9 to 32.4]
(0.31) [98.89 to 99.99] (0.030) [0 to 0.096] (0.7) [38.1 to 40.5] (0.06) [35.86 to 36.09] (1.0) [30.2 to 33.8]
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
301
Table 4 The values for p from a statistical analysis using Student paired t-test of the data presented in Table 3 Standard
Noncoplanar
Coplanar
Standard
Noncoplanar
Coplanar
0.775
V107% Noncoplanar Coplanar VMAT
0.154 0.192 0.158
0.071 0.039*
0.108
Mean Noncoplanar Coplanar VMAT
0.000* 0.000* 0.000*
0.004* 0.881
0.103
V95% Noncoplanar Coplanar VMAT
0.002* 0.074 0.082
Max Noncoplanar Coplanar VMAT
0.020* 0.030* 0.925
0.000* 0.000*
0.003*
Min Noncoplanar Coplanar VMAT
0.006* 0.028* 0.001*
0.071 0.001*
0.002*
0.007* 0.152
Significance is marked (*).
approach for craniospinal irradiation. It should be pointed out that especially the mean dose to the lens was sharply reduced in young patients by all the new techniques used, but mostly by the noncoplanar IMRT-based technique. Regarding dose results from the conventional, the VMAT, and the coplanar IMRT techniques presented in this study, they are generally found to be in good agreement with similar results from other studies. The noncoplanar IMRT technique seems to deliver better results when compared with other photon techniques. It seems to deliver inferior results regarding the eyes and parotid glands when compared with proton therapy and TomoTherapy. Still, the noncoplanar technique seems to be able to deliver a better target coverage than proton therapy and TomoTherapy. Previously, several treatment planning studies of craniospinal irradiation have been performed. When interpreting results of various approaches IMRT, VMAT, or TomoTherapy, it is important to note that different optimization criteria have been used, prioritizing one objective on behalf of another. Therefore, any comparison will be influenced by a choice of optimization criteria. An overview of previously published results regarding the eyes, the parotid gland, and the coverage of the cranial PTV are displayed in Table 5.
Harron and Lewis24 studied among other issues the use of ThomoTherapy compared with a standard photon method for craniospinal irradiation of a single patient to a total prescribed dose of 23.4 Gy. Scaled to 36 Gy, they found mean doses to the left and right eyes by the use of the standard technique to be 21.4 Gy and 26.0 Gy, respectively, which is in agreement with this study. They also found that TomoTherapy reduced the mean doses to the eyes to approximately 10 Gy, which is a lower dose than any of our plans. However, the use of TomoTherapy increased the nontumor integral dose by as much as 30% to 40% compared with the standard treatment. The use of TomoTherapy was also investigated by Sharma et al.25 TomoTherapy was found to be superior especially regarding the mean doses to the critical organs. However, the authors warn about the consequences of the increased integral dose to the patient. It is found that TomoTherapy is able to deliver a dose to the eyes lower than any other technique. Yoon et al.26 found that TomoTherapy could deliver a very low dose of 3.8 Gy to the parotid gland, which is lower than any other photon technique. Studenski et al.23 compared the conventional technique with a coplanar 5-field IMRT-based technique and a coplanar VMAT technique. The geometry of the VMAT technique in the present study is the same as the one used by Studenski. The VMAT technique significantly reduced the maximum dose to the parotid
Fig. 5. The dose given to the total eye volume by each of the 4 planning techniques for the patient seen in Fig. 4.
Fig. 6. The dose-volume histograms of the cranial PTV for the patient in Fig. 4 for each of the 4 planning techniques are displayed.
302
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
Table 5 Results from other investigations compared with the findings in this study for the eye, parotid gland, and the cranial PTV Publication
Technique 30
Stoker et al. Yoon et al.26 Harron and Lewis24 Harron and Lewis24 Sharma et al.25 Yoon et al.26 Present study
IMPT IMPT TomoTherapy TomoTherapy TomoTherapy TomoTherapy Noncoplanar IMRT
Studenski et al.23 Cao et al.22 Sharma et al.25 Present study
Coplanar Coplanar Coplanar Coplanar
Studenski et al.23 Fogliata et al.27 Lee et al.28 Present study
VMAT VMAT VMAT VMAT
26
Yoon et al. Harron and Lewis24 Studenski et al.23 Cao et al.22 Lee et al.28 Sharma et al.25 Present study
IMRT IMRT IMRT IMRT
Conventional Conventional Conventional Conventional Conventional Conventional Conventional
Eye (Gy)
Parotid gland (Gy)
Target coverage, V95%
Mean dose: 14.8* Mean dose: 3.9*
100 ⫾ 1
Mean dose (L, R): 10.0,9.5* Mean dose (L and R): 10.9 and 11.2* Mean dose (L and R): 8.1 and 8.2 * Mean dose: 13.7 Median dose (L and R): 18.6 and 18.5 Mean dose (L and R): 21.3 and 19.4 * Mean dose: 18.9 Mean dose: 13.5* Median dose (L and R): 35.7 and 27.8 * Mean dose: 17.6
99.1 99.3 Mean dose: 3.8 Mean dose: 20.5
99.9
Max dose: 32.30
97.9
Mean dose: 26.6
98.3 99.8
Max dose: 24.80 Mean dose: 14.8*
97.9
Mean dose: 20.0
99.8
Mean dose: 26.6 Mean dose (L and R): 21.4 and 26.0* Median dose (L and R): 21.8 and 19.0 Median dose (L and R): 37.8 and 31.2* Mean dose (L and R): 21.3 and 19.4* Mean dose: 21.8
Max dose: 38.95
99.4 97.9
Mean dose: 27.7
98.2 99.6
L-left, R-right. Results that have been rescaled to the prescribed dose of 36 Gy, recalculated or read from a graph are indicated (*).
glands compared with the standard technique from 38.95 to 24.80 Gy, but no further dose reduction was found using the coplanar IMRT approach. This finding is consistent with our results. Interestingly, Studenski et al. reported increased mean doses to the eyes and lenses by both coplanar IMRT and VMAT, which is only partially found in the present study. Fogliata et al.27 reported results of craniospinal irradiation using VMAT in 5 patients. Of them, 2 patients were treated with the dose of 36 Gy. The approximate mean dose to their eyes was 13.5 Gy, to the lenses 7.5 Gy, and to the parotid glands 14.8 Gy. However, these low doses resulted in the reduction of the target coverage. Lee et al.28 compared a VMAT planning technique with only one arc with the standard craniospinal irradiation technique. The median dose to the eyes and lenses was found to be reduced by using VMAT, normalized to 36 Gy; this reduction was approximately from 34.5 to 31.7 Gy for the eyes and 29.9 to 23.7 Gy for the lenses. However, the dose to the normal tissue increased. All these findings are in line with the findings of the present study. Coplanar IMRT was investigated by Cao et al.22 Conventional treatment plans for craniospinal irradiation of 3 patients were compared with treatment plans based on coplanar IMRT technique, consisting of 7 fields for the cranial part and 3 for the spinal part. The coplanar IMRT technique improved the dose coverage of the target, compared with the conventional planning technique. The V95% of the whole PTV was improved from 98.1% to 99.5%, which is consistent with the improvement of cranial PTV coverage using coplanar IMRT in our study. Additionally, Cao et al. found a reduction of the median dose to the eyes from approximately 20.4 to 18.6 Gy, which is in agreement with the reduction found in the present study. Sharma et al.25 also performed a comparison between the standard photon technique and a coplanar IMRT technique. Mean doses to the eyes of 21.3 Gy and 19.4 Gy were found to be in agreement with the findings in the present study. Treatment based on protons has the potential to spare the normal tissue even more than photon-based therapy.26,29 This is demonstrated by Stoker et al.30 who found good sparing of the parotid gland, which received only 14.8 Gy, and also sparing of other critical organs. Additionally they found good coverage of the target, especially the cribriform plate. However, the availability of
proton therapy is often limited by economy or capacity; therefore, photon techniques are mostly used.18 Craniospinal irradiation is typically used as a part of the curative treatment of children and young adults with medulloblastoma. Fortunately, the prognosis of the most patients is very good, and they are expected to survive for a long time. Therefore, the risk of radiation-induced secondary cancer is important to take into account during treatment planning.31-33 In our study, mean dose to the surrounding normal tissues was found to be reduced using both the coplanar and noncoplanar IMRT-based techniques, giving the expectation that the risk of secondary cancer by primary irradiation is not increased. However, caution is mandatory in the interpretation of the reported results, because actual radiation exposure to the healthy tissue may differ from that predicted by the treatment planning system. The treatment planning systems often predict the radiation doses to the distant tissues with a high degree of uncertainty.34 In addition, the treatment planning systems normally ignore other sources of radiation, such as leak radiation from the head of the accelerator and other scattered particles.35 Moreover, the secondary irradiation increase linearly with the number of MU used, which may indicate that any of the new methods might still increase the risk of secondary cancer. A calculated estimate of the risk of secondary cancer using the new techniques was not addressed in this study. One advantage of proton radiotherapy is the reduction of the risk of secondary radiation-induced cancer.26,36
Conclusion In conclusion, we found that the new noncoplanar IMRT technique for the cranial part of the craniospinal irradiation significantly reduced the mean doses to the eyes, the inner ears, the parotid glands, and the normal tissue compared with the standard technique for craniospinal irradiation. In addition, the noncoplanar IMRT technique significantly improved dose coverage of the target. Apparent reduced dose to the normal tissue may decrease the risk for secondary cancer. However, the increased number of MU and the uncertainty in the estimation of dose given
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
to the distant tissue need to be taken into account in the interpretation of the findings. Acknowledgments The authors thank Steffen L. Hokland for much appreciated help during the preparation of this article. References 1. Massimino, M.; Giangaspero, F.; Garrè, M.L.; et al. Childhood medulloblastoma. Crit. Rev. Oncol. Hematol. 79:65–83; 2011. 2. Gerber, N.; Mynarek, M.; Hoff, K.; et al. Recent developments and current concepts in medulloblastoma. Cancer Treat. Rev. 40:356–65; 2014. 3. Bartlett, F.; Kortmann, R.; Saran, F. Medulloblastoma. Clin. Oncol. 25:36–45; 2013. 4. Liu, X.; Yu, J.; Yu, Y.; et al. Conventional craniospinal irradiation with patient supine and source-skin distance (SSD) 100 cm for spinal field. Med. Dosim. 36:373–6; 2011. 5. Munshi, A.; Jalali, R. A simple technique of supine craniospinal irradiation. Med. Dosim. 33:1–5; 2008. 6. Hideghét, K.; Cserháti, A.; Nagy, Z.; et al. A prospective study of supine versus prone positioning and whole-body thermoplastic mask fixation for craniospinal radiotherapy in adult patients. Radiother. Oncol. 102:214–8; 2012. 7. Wilkinson, J.M.; Lewis, J.; Lawrence, G.P.; et al. Craniospinal irradiation using a forward planned segmented field technique. Br. J. Radiol. 80:209–15; 2007. 8. Chojnacka, M.; Skowro´nska-Gardas, A.; Morawska-Kaczy´nska, M.; et al. Craniospinal radiotherapy in children: electron- or photon-based technique of spinal irradiation. Rep. Pract. Oncol. Radiother 15:21–4; 2010. 9. Sun, L.M.; Yeh, S.A.; Wang, C.J.; et al. Postoperative radiation therapy for medulloblastoma—high recurrence rate in the subfrontal region. J. Neurooncol. 58:77–85; 2001. 10. Skowrońska-Gardas, A.; Chojnacka, M.; Morawska-Kaczyńska, M.; et al. Patterns of failure in children with medulloblastoma treated with 3D conformal radiotherapy. Radiother. Oncol. 84:26–33; 2007. 11. Weiss, E.; Krebeck, M.; Köhler, B.; et al. Does the standardized helmet technique lead to adequate coverage of the cribriform plate? An analysis of current practice with respect to the ICRU 50 report Int. J. Radiat. Oncol. Biol. Phys. 49:1475–80; 2001. 12. Deasy, J.; Moiseenko, V.; Marks, L.; et al. Radiotherapy dose-volume effects on salivary gland function. Int. J. Radiat. Oncol. Biol. Phys. 76:S58–63; 2010. 13. King, M.; Modlin, L.; Million, L.; et al. Parotid gland as an organ-at-risk for craniospinal irradiation. Int. J. Radiat. Oncol. Biol. Phys. 84:S634; 2012. 14. Bhandare, N.; Jackson, A.; Eisbruch, A.; et al. Radiation therapy and hearing loss. Int. J. Radiat. Oncol. Biol. Phys. 76:S50–7; 2010. 15. Paulino, A.; Lobo, M.; Teh, B.; et al. Ototoxicity after intensity-modulated radiation therapy and cisplatin-based chemotherapy in children with medulloblastoma. Int. J. Radiat. Oncol. Biol. Phys. 78:1445–50; 2010. 16. Kleiman, N. Radiation cataract. Ann. ICRP 41:80–97; 2012. 17. Christopherson, K.; Rotondo, R.; Bradley, J.; et al. Late toxicity following craniospinal radiation for early-stage medulloblastoma. Acta. Oncol. 53: 471–480; 2014.
303
18. Wolden, S.L. Protons for craniospinal radiation: Are clinical data important? Int. J. Radiat. Oncol. Biol. Phys 78:231–2; 2013. 19. Hadley, A.; Ding, G. A single-gradient junction technique to replace multiplejunction shifts for craniospinal irradiation treatment. Med. Dosim. 39:314–9; 2014. 20. Seppälä, J.; Kulmala, J.; Lindholm, P.; et al. A method to improve target dose homogeneity of craniospinal irradiation using dynamic split fields IMRT. Radiother. Oncol. 96:209–15; 2010. 21. Kusters, J.; Louwe, R.; Kollenburg, P.; et al. Optimal normal tissue sparing in craniospinal axis irradiation using IMRT with daily intrafractionally modulated junction(s). Int. J. Radiat. Oncol. Biol. Phys. 81:1405–14; 2011. 22. Cao, F.; Ramaseshan, R.; Corns, R.; et al. A three-isocenter jagged-junction IMRT approach for craniospinal irradiation without beam edge matching for field junctions. Int. J. Radiat. Oncol. Biol. Phys. 84:648–54; 2012. 23. Studenski, M.; Shen, X.; Yu, Y.; et al. Intensity-modulated radiation therapy and volumetric-modulated arc therapy for adult craniospinal irradiation—A comparison with traditional techniques. Med. Dosim. 38:48–54; 2013. 24. Harron, E.; Lewis, J. Bowel sparing in pediatric cranio-spinal radiotherapy: a comparison of combined electron and photon and helical TomoTherapy techniques to a standard photon method. Med. Dosim. 37:140–4; 2012. 25. Sharma, D.S.; Gupta, T.; Jalali, R.; et al. High-precision radiotherapy for craniospinal irradiation: evaluation of three-dimensional conformal radiotherapy, intensity modulated radiation therapy and helical TomoTherapy. Br. J. Radiol. 82:1000–9; 2009. 26. Yoon, M.; Shin, D.H.; Kim, J.; et al. Craniospinal irradiation techniques: A dosimetric comparison of proton beams with standard and advanced photon radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 81:637–46; 2011. 27. Fogliata, A.; Bergström, S.; Cafaro, I. Cranio-spinal irradiation with volumetric modulated arc therapy: A multi-institutional treatment experience. Radiother. Oncol. 99:79–85; 2011. 28. Lee, Y.; Brooks, C.; Bedford, J.; et al. Development and evaluation of multiple isocentric volumetric modulated arc therapy technique for craniospinal axis radiotherapy planning. Int. J. Radiat. Oncol. Biol. Phys. 82:1006–12; 2012. 29. Fossati, P.; Ricardi, U.; Orecchia, R. Pediatric medulloblastoma: toxicity of current treatment and potential role of proton therapy. Cancer Treat. Rev. 35:79–96; 2009. 30. Stoker, J.B.; Grant, J.; Zhu, R.; et al. Intensity modulated proton therapy for craniospinal irradiation: organ-at-risk exposure and a low-gradient junctioning technique. Int. J. Radiat. Oncol. Biol. Phys. 90:637–44; 2014. 31. Ruben, J.; Davis, S.; Evans, C.; et al. The effect of intensity-modulated radiotherapy on radiation-induced malignancies. Int. J. Radiat. Oncol. Biol. Phys. 70:1530–6; 2008. 32. Schneider, U.; Lomax, A.; Timmermann, B. Second cancers in children treated with modern radiotherapy techniques. Radiother. Oncol. 89:135–40; 2008. 33. Galloway, T.; Indelicato, D.; Amdur, R.; et al. Analysis of dose at the site of second tumor formation after radiotherapy to the central nervous system. Int. J. Radiat. Oncol. Biol. Phys. 82:90–4; 2012. 34. Taddei, P.; Jalbout, W.; Howell, R.; et al. Analytical model for out-of-field dose in photon craniospinal irradiation. Phys. Med. Biol. 58:7463–79; 2013. 35. Zaghloul, M.S. Intensity modulated radiotherapy (IMRT) for pediatric cancer patients: the advantage and fear of second malignant neoplasm. J. Egypt Nat. Cancer Inst 25:1–3; 2013. 36. Zhang, R.; Howell, R.M.; Taddei, P.J.; et al. A comparative study on the risk of radiogenic second cancers and cardiac mortality in a set of pediatric medulloblastoma patients treated with photon or proton craniospinal irradiation. Radiother. Oncol. 113:84–8; 2014.
Medical Dosimetry journal homepage: www.meddos.org
Comparison of a new noncoplanar intensity-modulated radiation therapy technique for craniospinal irradiation with 3 coplanar techniques Anders T. Hansen, M.S.,* Slavka Lukacova, M.D., Ph.D.,† Yasmin Lassen-Ramshad, M.D., Ph.D.,† and Jørgen B. Petersen, M.S.* Department of Medical Physics, Aarhus University Hospital, Nørrebrogade 44, Building 5, DK-8000 Aarhus C, Denmark; and †Department of Oncology, Aarhus University Hospital, Nørrebrogade 44, Building 5, DK-8000 Aarhus C, Denmark
*
A R T I C L E I N F O
A B S T R A C T
Article history: Received 29 September 2014 Received in revised form 15 February 2015 Accepted 19 March 2015
When standard conformal x-ray technique for craniospinal irradiation is used, it is a challenge to achieve satisfactory dose coverage of the target including the area of the cribriform plate, while sparing organs at risk. We present a new intensity-modulated radiation therapy (IMRT), noncoplanar technique, for delivering irradiation to the cranial part and compare it with 3 other techniques and previously published results. A total of 13 patients who had previously received craniospinal irradiation with standard conformal x-ray technique were reviewed. New treatment plans were generated for each patient using the noncoplanar IMRT-based technique, a coplanar IMRT-based technique, and a coplanar volumetric-modulated arch therapy (VMAT) technique. Dosimetry data for all patients were compared with the corresponding data from the conventional treatment plans. The new noncoplanar IMRT technique substantially reduced the mean dose to organs at risk compared with the standard radiation technique. The 2 other coplanar techniques also reduced the mean dose to some of the critical organs. However, this reduction was not as substantial as the reduction obtained by the noncoplanar technique. Furthermore, compared with the standard technique, the IMRT techniques reduced the total calculated radiation dose that was delivered to the normal tissue, whereas the VMAT technique increased this dose. Additionally, the coverage of the target was significantly improved by the noncoplanar IMRT technique. Compared with the standard technique, the coplanar IMRT and the VMAT technique did not improve the coverage of the target significantly. All the new planning techniques increased the number of monitor units (MU) used—the noncoplanar IMRT technique by 99%, the coplanar IMRT technique by 122%, and the VMAT technique by 26%—causing concern for leak radiation. The noncoplanar IMRT technique covered the target better and decreased doses to organs at risk compared with the other techniques. All the new techniques increased the number of MU compared with the standard technique. & 2015 American Association of Medical Dosimetrists.
Keywords: Craniospinal irradiation IMRT Medulloblastoma Pediatric radiotherapy
Introduction Medulloblastoma represents approximately 20% of intracranial malignant tumors in children.1-3 This disease typically spreads through the cerebrospinal fluid within the central nervous system. The standard treatment consists of maximal surgical resection followed by radiation therapy and chemotherapy.1,3 The whole craniospinal axis is treated with radiation followed by a boost in the posterior fossa. The standard radiotherapy technique used for craniospinal irradiation consists of 2 almost opposing lateral
Reprint requests to: Anders Traberg Hansen, Department of Medical Physics, Aarhus University Hospital, Nørrebrogade 44, 8000 Arhus C, Denmark. E-mail: [email protected] http://dx.doi.org/10.1016/j.meddos.2015.03.007 0958-3947/Copyright Ó 2015 American Association of Medical Dosimetrists
photon fields, rotated slightly in the anterior direction, to place the anterior field edges just behind both lenses, as seen in Fig. 1B. To these 2 lateral fields, several posterior fields are typically added end to end to irradiate the spine from the caudal field edge of the 2 cranial fields to the sacrum.4-8 It has previously been shown that sufficient dose coverage of the whole central nervous system, including the cribriform plate, is essential in preventing recurrence of the disease.9,10 The disadvantage of the standard conformal xray technique is inadequate target coverage, mainly of the cribriform plate, when certain organs at risk like the parotid glands, the inner ears, or the lenses are to be spared.11 The risk of xerostomia increases when the parotid glands receive doses more than 25 Gy.12,13 The risk of hearing loss has been reported to increase when doses to the inner ears are more than 35 Gy.14,15 Especially for
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
297
Fig. 1. The field arrangements for the 4 treatment techniques are displayed. In the top left illustration is the noncoplanar IMRT treatment technique (A) and in the top right the standard technique (B). In the bottom left is the coplanar IMRT technique (C) and in the bottom right the VMAT technique (D).
young children, the cribriform plate tends to be more anteriorly located, which is why a full irradiation of the lenses often is unavoidable. If doses more than 0.5 Gy are given to the lenses, there is an increased risk for development of radiation-induced cataract.16 Additionally, radiotherapy for medulloblastoma may lead to a series of other late side effects.1,2 Therefore, a radiation technique that is able to cover the target with a sufficient dose and at the same time limits the doses given to the described organs at risk is needed.17 This can be achieved using proton therapy.18 However, the use of proton therapy is associated with large economic cost and limited capacity. Therefore, we propose a new noncoplanar intensitymodulated radiation therapy (IMRT)–based technique for craniospinal irradiation, which improves the dose coverage of the target and reduces the dose to the described critical organs.
organs at risk. The CTV consisted of the whole brain including the cribriform plate and the spinal canal. The planning target volume (PTV) was generated by adding a margin of 5 mm to the CTV; however, a margin of 6 mm in the cranial and caudal direction was used because of the uncertainty introduced by the spacing between the CT slices. The critical organs defined for this study were as follows: the parotid glands, the inner ears, the lenses, and the frontal and posterior parts of the eyes. To
Methods and Materials Treatment plans for 13 patients who had been treated for medulloblastoma, at our hospital, in the time period 2007 to 2013 with the standard conformal x-ray technique were collected for this study. In total, 6 patients were children (2 to 17 years), 2 patients were young adults (18 to 30 years), and 5 were adults (30 to 60 years). In the treatment plans, the cranial part as well as the spinal part of the craniospinal treatment was point normalized for all patients. The cranial normalization points were in all cases located centrally in the brain. The spinal normalization points were in all cases placed at representative positions in the medulla spinalis. The original spinal part of the treatment plans was left almost unchanged, but the cranial part of the treatment plans was replanned using 3 planning techniques: our novel noncoplanar IMRT, a coplanar IMRT, and a coplanar volumetric-modulated arch therapy (VMAT) technique.
Delineation The original computed tomography (CT) scans were in 1 case performed with a slice thickness of 4.5 mm and in 12 cases with a slice thickness of 3 mm. They encompassed the whole cranium, the thorax, and the abdominal and pelvic regions. Typically, parts of the arms and legs were not included in the CT scans. The CT scans were used for delineation of the clinical target volume (CTV) and
Fig. 2. An example of the spinal part of the original conventional treatment plan. Notice the 6 segments in the cervical junction zone.
298
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
Fig. 3. Digitally reconstructed radiographs for the 5 fields constituting the noncoplanar technique. The 2 images in the top of the figure are the coplanar fields and the lower 3 are the noncoplanar. The eyes have been indicated.
increase the clarity and simplicity of the analysis, organs of the same type on the left and right side were unified into a single volume. The volume, referred to as normal tissue, was defined as the total body covered by the CT scan, but outside of the PTV. In addition, the structure of foramen magnum was introduced as an anatomical landmark for the orientation of the noncoplanar fields. The part of the PTV extending cranially from the foramen magnum was also defined as an independent structure called the cranial PTV, to evaluate the capability of the different techniques to cover this part.
Dose calculation and treatment planning The commercial treatment planning system Eclipse version 11.3.31 (Varian Medical Systems Inc.) and the convolution-superposition dose calculation algorithm AAA were used for all treatment plans. Originally, the spinal axis was treated for all patients according to the standard technique with several posterior fields. Overall, 2 to 3 fields, consisting of 3 segments each, were used to cover the whole length of the spine. The field junctions moved over 3 positions, separated with the gap of 1.5 to 2.5 cm, to smear out the dose heterogeneities in the target and to increase the plan robustness for
errors of field abutment. Several low-weight segments were used to account for the depth variation of the spine. The spinal part of the original conventional treatment plan was used in all new treatment plans. However, the number of segments in the most cranial spinal field was doubled to smooth the longitudinal dose gradient in the cervical junction zone, as displayed in Fig. 2. The dose given to the target by the spinal part of the treatment plan served as the basis for the optimization of the new plans. This approach of using the optimization algorithm to produce a smooth dose transition in the junction zone has previously been demonstrated to be both safe and efficient.1922 All treatment planning techniques for the cranial part were symmetrical in the sagittal plane, and 6-MV photons were used for all fields, as shown in Fig. 1. Noncoplanar IMRT technique: In total, 5 IMRT fields including 3 noncoplanar fields were used as shown in Fig. 1A. The central axis of the treatment fields was set to be in a plane defined by the cribriform plate and the foramen magnum. The treatment fields were placed within this plane to obtain a sufficient coverage of the cribriform plate. This approach allowed keeping the volume of irradiated normal tissue outside the brain low. Moreover, 2 of the fields were coplanar and defined by gantry position of 751 and 2851. The angles for the gantry and table rotation of the noncoplanar fields depended strongly on the anatomy and fixation of the individual patient. The jaws of the noncoplanar fields were fixed to avoid
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
299
Fig. 4. An example of dose distributions near the eyes and the cribriform plate for the noncoplanar (A), the standard (B), the coplanar (C), and the VMAT (D) techniques. The PTV is drawn in solid black. The 50% isodose curve is drawn in white, the 95% isodose curve is drawn in dark gray, and the 105% isodose curve is if present drawn in light gray. unnecessary irradiation to the normal tissue in the front of the head. The 2 coplanar fields were used for the treatment of the upper part of the spinal cord. A set of beams-eye views from the fields of a noncoplanar plan is displayed in Fig. 3. Because all the noncoplanar fields were planned from an anterior and superior direction, it is considered unlikely, that collisions between the gantry and the table or the patient would occur, using this technique. Coplanar IMRT technique: This technique consisted of 5 fields produced by projecting the 3 noncoplanar fields from the noncoplanar plan onto the transversal plane of the patient (Fig. 1C). The jaws of the 3 frontal fields were fixed just caudally of the eyes to limit the unnecessary irradiation of the frontal part of the head. Coplanar VMAT technique: This technique consisted of 2 intensity-modulated arcs going clockwise from 1821 to 01 and then to 1781 and the same way counterclockwise as previously described by Studenski et al.23 The collimator was set in 451 and 3151 for the 2 arcs (Fig. 1D). Planning optimization: The same set of optimization objectives was used for all treatment plans in the optimization process, with the exception of virtual volumes. Virtual volumes were defined as the volumes of either an overdose or underdose in
the initiation of the optimization process. The fluences of the IMRT fields, which had fixed jaws, were edited by hand, before the start of a subsequent optimization to avoid steep gradients and an inhomogeneous dose distribution. At least 2 subsequent optimizations were performed for all plans. The new treatment plans irradiating the brain and the cranial part of the spinal cord were all normalized to the mean dose in the part of the CTV, which was located cranially off the foramen magnum, which was equal to the brain. A final treatment plan, consisting of the sum of the new treatment plan and the modified spinal part of the original plan, was produced for each patient. An illustration of calculated doses near the eyes and the cribriform plate can be seen in Fig. 4. The prescribed dose was set to be 36 Gy in 20 fractions, which is one of the prescriptions used for treating medulloblastoma.2,22,23
Plan evaluation The mean doses to the parotid glands, the eyes, the lenses, the inner ears, the frontal and posterior parts of the eyes, and the normal tissues and the number of
Table 1 The mean dose to the critical organs and the number of monitor units are shown for each treatment planning technique Standard Parotid gland Lens Eye Frontal eye Posterior eye Ear Body-PTV Monitor units
27.7 (4.9) 11.3 (9.8) 21.8 (6.1) 15.8 (6.1) 27.7 (7.4) 36.5 (0.5) 8.5 (1.4)
Noncoplanar [17.8 to 35.7] [4.1 to 36.6] [12.6 to 32.9] [9.2 to 31.4] [15.0 to 35.9] [35.9 to 37.3] [5.8 to 11.2]
726 (94) [522 to 829]
20.5 (1.5) 5.9 (1.2) 13.7 (2.1) 9.3 (1.6) 18.0 (2.9) 35.0 (0.7) 8.3 (1.4)
[17.7 to 22.6] [3.9 to 7.7] [11.0 to 17.0] [6.8 to 11.9] [14.4 to 23.0] [33.8 to 36.4] [5.8 to 10.8]
1445 (133) [1264 to 1696]
Standard deviations are specified in brackets, and ranges are specified in square brackets.
Coplanar 26.6 (1.5) 9.7 (3.0) 18.9 (2.7) 14.7 (2.7) 23.0 (3.0) 35.8 (0.7) 8.4 (1.4)
VMAT [23.2 to 29.8] [6.0 to 16.2] [13.8 to 23.0] [10.0 to 18.6] [17.6 to 27.4] [34.1 to 36.7] [5.8 to 10.9]
1617 (120) [1449 to 1836]
20.0 (1.4) 10.6 (1.6) 17.6 (2.1) 14.2 (2.0) 20.9 (2.4) 35.7 (0.6) 8.9 (1.5)
[18.5 to 23.3] [7.6 to 13.3] [14.7 to 21.3] [11.1 to 17.4] [17.7 to 24.6] [34.7 to 36.5] [6.0 to 11.5]
916 (77) [751 to 2005]
300
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
Table 2 The values for p from a statistical analysis using Student paired t-test of the data presented in Table 1 Standard Parotid gland Noncoplanar Coplanar VMAT
o0.001* 0.344 o0.001*
Eye Noncoplanar Coplanar VMAT
o0.001* 0.044* 0.022*
Posterior eye Noncoplanar Coplanar VMAT
o0.001* 0.014* 0.003*
Body-PTV Noncoplanar Coplanar VMAT
0.014* 0.009* o0.001*
Noncoplanar
Coplanar
o0.001* 0.212
o0.001* o0.001*
o0.001* o0.001*
0.090 o0.001*
Standard
Noncoplanar
Coplanar
o0.001*
Lens Noncoplanar Coplanar VMAT
0.052 0.473 0.795
o0.001* o0.001*
0.235
0.029*
Frontal eye Noncoplanar Coplanar VMAT
0.002* 0.456 0.367
o0.001* o0.001*
0.389
0.001*
Ear Noncoplanar Coplanar VMAT
o0.001* 0.005* 0.001*
o0.001* 0.001*
0.577
o0.001*
No. of MU Noncoplanar Coplanar VMAT
o0.001* o0.001* o0.001*
o0.001* o0.001*
o0.001*
Significance is marked (*).
monitor units (MU) were calculated for all total treatment plans and compared as illustrated in Table 1. Also, the mean, maximum, and minimum dose to the cranial PTV and the relative volume of it receiving more than 95% and more than 107% of the prescribed dose was also calculated and compared as illustrated in Table 3. Paired 2-tailed Student t-tests were used to compare results between treatment techniques. A confidence level of 5% was regarded statistically significant. The values for p for the correlations of the data in Table 1 can be seen in Table 2. Moreover, p values for the data displayed in Table 3 are in the same manner displayed in Table 4.
Results The noncoplanar IMRT technique significantly reduced the mean dose to the parotid gland, the frontal and posterior eye, the inner ear, and the normal tissue compared with the standard technique as illustrated in Tables 1 and 2. Moreover, this technique was significantly better to spare all critical organs than the coplanar IMRT technique. The technique was also superior to the VMAT technique regarding the mean dose to the lens, the frontal and posterior eye, the inner ear, and the normal tissue, but not regarding the parotid gland, where the results were similar. We observed that the VMAT technique delivered lower mean doses to the parotid gland and posterior eye, but higher doses to the normal tissue, compared with the coplanar IMRT-based technique. Especially regarding the lens, both the noncoplanar IMRT-based technique and VMAT technique reduced the mean dose compared with the standard technique, but this reductions was not found to be statistically significant. However, a major reduction in mean dose to the lens from 22.7 to 7.2 Gy was found for the 4 patients younger than 15 years using the noncoplanar IMRT-based technique relative to the standard technique. The reduction using the coplanar technique was to 13.3 Gy and to 11.3 Gy using VMAT. For adult patients, there was almost no difference in mean dose to the
lens between the noncoplanar IMRT-based technique and the standard technique. The coplanar and the noncoplanar IMRTbased techniques significantly reduced the dose to the normal tissue compared with the standard technique. The dose to the normal tissue was found to be highest using VMAT technique. The dose coverage of the cranial PTV was found to be significantly improved using the noncoplanar IMRT-based technique relative to the standard and coplanar IMRT-based techniques. The relative volume covered by 95% of the prescribed dose is displayed in Table 3. The dose-volume relationship regarding the eyes and the cranial PTV for a typical patient is displayed in Figs. 5 and 6. The number of MU for a whole treatment was found to be lowest for the standard technique, 26% higher for the VMAT, 99% higher for the noncoplanar IMRT-based technique, and 123% higher for the coplanar IMRT-based technique as illustrated in Table 1. Discussion The results of the present study indicate that the new noncoplanar IMRT-based technique for craniospinal irradiation is superior regarding both the mean dose to the critical organs and improvement of the target dose coverage. The dose reduction to the organs at risk may result in the reduction of the incidence and severity of potential side effects, such as xerostomia, hearing loss, and lens cataract. Furthermore, the improvement of the dose coverage of the target may decrease the risk of recurrence and thus improve the survival of patients with medulloblastoma. The fact that the noncoplanar IMRT-based technique was identical to the coplanar IMRT-based technique in all aspects except for the use of noncoplanar fields, but produced significantly superior results, demonstrates the benefit of the noncoplanar treatment
Table 3 Mean values for the doses to the cranial PTV are shown for each treatment planning technique
V95%, % V107%, % Max, Gy Mean, Gy Min, Gy
Standard
Noncoplanar
Coplanar
VMAT
99.56 1.3 39.0 36.70 20.9
99.94 0.002 38.2 36.04 30.2
99.77 0.12 40.1 36.06 28.4
99.80 0.019 39.0 36.03 32.2
(0.33) [99.13 to 99.99] (3.0) [0 to 11.0] (1.1) [37.8 to 41.3] (0.47) [35.81 to 37.42] (9.7) [5.8 to 31.8]
(0.06) [99.75 to 99.99] (0.005) [0 to 0.013] (1.1) [36.8 to 40.1] (0.04) [35.97 to 36.13] (1.8) [27.8 to 32.6]
Standard deviations are specified in brackets, ranges are specified in square brackets.
(0.20) [99.36 to 99.97] (1.20) [0 to 0.70] (1.1) [38.4 to 42.0] (0.04) [36.01 to 36.13] (3.3) [20.9 to 32.4]
(0.31) [98.89 to 99.99] (0.030) [0 to 0.096] (0.7) [38.1 to 40.5] (0.06) [35.86 to 36.09] (1.0) [30.2 to 33.8]
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
301
Table 4 The values for p from a statistical analysis using Student paired t-test of the data presented in Table 3 Standard
Noncoplanar
Coplanar
Standard
Noncoplanar
Coplanar
0.775
V107% Noncoplanar Coplanar VMAT
0.154 0.192 0.158
0.071 0.039*
0.108
Mean Noncoplanar Coplanar VMAT
0.000* 0.000* 0.000*
0.004* 0.881
0.103
V95% Noncoplanar Coplanar VMAT
0.002* 0.074 0.082
Max Noncoplanar Coplanar VMAT
0.020* 0.030* 0.925
0.000* 0.000*
0.003*
Min Noncoplanar Coplanar VMAT
0.006* 0.028* 0.001*
0.071 0.001*
0.002*
0.007* 0.152
Significance is marked (*).
approach for craniospinal irradiation. It should be pointed out that especially the mean dose to the lens was sharply reduced in young patients by all the new techniques used, but mostly by the noncoplanar IMRT-based technique. Regarding dose results from the conventional, the VMAT, and the coplanar IMRT techniques presented in this study, they are generally found to be in good agreement with similar results from other studies. The noncoplanar IMRT technique seems to deliver better results when compared with other photon techniques. It seems to deliver inferior results regarding the eyes and parotid glands when compared with proton therapy and TomoTherapy. Still, the noncoplanar technique seems to be able to deliver a better target coverage than proton therapy and TomoTherapy. Previously, several treatment planning studies of craniospinal irradiation have been performed. When interpreting results of various approaches IMRT, VMAT, or TomoTherapy, it is important to note that different optimization criteria have been used, prioritizing one objective on behalf of another. Therefore, any comparison will be influenced by a choice of optimization criteria. An overview of previously published results regarding the eyes, the parotid gland, and the coverage of the cranial PTV are displayed in Table 5.
Harron and Lewis24 studied among other issues the use of ThomoTherapy compared with a standard photon method for craniospinal irradiation of a single patient to a total prescribed dose of 23.4 Gy. Scaled to 36 Gy, they found mean doses to the left and right eyes by the use of the standard technique to be 21.4 Gy and 26.0 Gy, respectively, which is in agreement with this study. They also found that TomoTherapy reduced the mean doses to the eyes to approximately 10 Gy, which is a lower dose than any of our plans. However, the use of TomoTherapy increased the nontumor integral dose by as much as 30% to 40% compared with the standard treatment. The use of TomoTherapy was also investigated by Sharma et al.25 TomoTherapy was found to be superior especially regarding the mean doses to the critical organs. However, the authors warn about the consequences of the increased integral dose to the patient. It is found that TomoTherapy is able to deliver a dose to the eyes lower than any other technique. Yoon et al.26 found that TomoTherapy could deliver a very low dose of 3.8 Gy to the parotid gland, which is lower than any other photon technique. Studenski et al.23 compared the conventional technique with a coplanar 5-field IMRT-based technique and a coplanar VMAT technique. The geometry of the VMAT technique in the present study is the same as the one used by Studenski. The VMAT technique significantly reduced the maximum dose to the parotid
Fig. 5. The dose given to the total eye volume by each of the 4 planning techniques for the patient seen in Fig. 4.
Fig. 6. The dose-volume histograms of the cranial PTV for the patient in Fig. 4 for each of the 4 planning techniques are displayed.
302
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
Table 5 Results from other investigations compared with the findings in this study for the eye, parotid gland, and the cranial PTV Publication
Technique 30
Stoker et al. Yoon et al.26 Harron and Lewis24 Harron and Lewis24 Sharma et al.25 Yoon et al.26 Present study
IMPT IMPT TomoTherapy TomoTherapy TomoTherapy TomoTherapy Noncoplanar IMRT
Studenski et al.23 Cao et al.22 Sharma et al.25 Present study
Coplanar Coplanar Coplanar Coplanar
Studenski et al.23 Fogliata et al.27 Lee et al.28 Present study
VMAT VMAT VMAT VMAT
26
Yoon et al. Harron and Lewis24 Studenski et al.23 Cao et al.22 Lee et al.28 Sharma et al.25 Present study
IMRT IMRT IMRT IMRT
Conventional Conventional Conventional Conventional Conventional Conventional Conventional
Eye (Gy)
Parotid gland (Gy)
Target coverage, V95%
Mean dose: 14.8* Mean dose: 3.9*
100 ⫾ 1
Mean dose (L, R): 10.0,9.5* Mean dose (L and R): 10.9 and 11.2* Mean dose (L and R): 8.1 and 8.2 * Mean dose: 13.7 Median dose (L and R): 18.6 and 18.5 Mean dose (L and R): 21.3 and 19.4 * Mean dose: 18.9 Mean dose: 13.5* Median dose (L and R): 35.7 and 27.8 * Mean dose: 17.6
99.1 99.3 Mean dose: 3.8 Mean dose: 20.5
99.9
Max dose: 32.30
97.9
Mean dose: 26.6
98.3 99.8
Max dose: 24.80 Mean dose: 14.8*
97.9
Mean dose: 20.0
99.8
Mean dose: 26.6 Mean dose (L and R): 21.4 and 26.0* Median dose (L and R): 21.8 and 19.0 Median dose (L and R): 37.8 and 31.2* Mean dose (L and R): 21.3 and 19.4* Mean dose: 21.8
Max dose: 38.95
99.4 97.9
Mean dose: 27.7
98.2 99.6
L-left, R-right. Results that have been rescaled to the prescribed dose of 36 Gy, recalculated or read from a graph are indicated (*).
glands compared with the standard technique from 38.95 to 24.80 Gy, but no further dose reduction was found using the coplanar IMRT approach. This finding is consistent with our results. Interestingly, Studenski et al. reported increased mean doses to the eyes and lenses by both coplanar IMRT and VMAT, which is only partially found in the present study. Fogliata et al.27 reported results of craniospinal irradiation using VMAT in 5 patients. Of them, 2 patients were treated with the dose of 36 Gy. The approximate mean dose to their eyes was 13.5 Gy, to the lenses 7.5 Gy, and to the parotid glands 14.8 Gy. However, these low doses resulted in the reduction of the target coverage. Lee et al.28 compared a VMAT planning technique with only one arc with the standard craniospinal irradiation technique. The median dose to the eyes and lenses was found to be reduced by using VMAT, normalized to 36 Gy; this reduction was approximately from 34.5 to 31.7 Gy for the eyes and 29.9 to 23.7 Gy for the lenses. However, the dose to the normal tissue increased. All these findings are in line with the findings of the present study. Coplanar IMRT was investigated by Cao et al.22 Conventional treatment plans for craniospinal irradiation of 3 patients were compared with treatment plans based on coplanar IMRT technique, consisting of 7 fields for the cranial part and 3 for the spinal part. The coplanar IMRT technique improved the dose coverage of the target, compared with the conventional planning technique. The V95% of the whole PTV was improved from 98.1% to 99.5%, which is consistent with the improvement of cranial PTV coverage using coplanar IMRT in our study. Additionally, Cao et al. found a reduction of the median dose to the eyes from approximately 20.4 to 18.6 Gy, which is in agreement with the reduction found in the present study. Sharma et al.25 also performed a comparison between the standard photon technique and a coplanar IMRT technique. Mean doses to the eyes of 21.3 Gy and 19.4 Gy were found to be in agreement with the findings in the present study. Treatment based on protons has the potential to spare the normal tissue even more than photon-based therapy.26,29 This is demonstrated by Stoker et al.30 who found good sparing of the parotid gland, which received only 14.8 Gy, and also sparing of other critical organs. Additionally they found good coverage of the target, especially the cribriform plate. However, the availability of
proton therapy is often limited by economy or capacity; therefore, photon techniques are mostly used.18 Craniospinal irradiation is typically used as a part of the curative treatment of children and young adults with medulloblastoma. Fortunately, the prognosis of the most patients is very good, and they are expected to survive for a long time. Therefore, the risk of radiation-induced secondary cancer is important to take into account during treatment planning.31-33 In our study, mean dose to the surrounding normal tissues was found to be reduced using both the coplanar and noncoplanar IMRT-based techniques, giving the expectation that the risk of secondary cancer by primary irradiation is not increased. However, caution is mandatory in the interpretation of the reported results, because actual radiation exposure to the healthy tissue may differ from that predicted by the treatment planning system. The treatment planning systems often predict the radiation doses to the distant tissues with a high degree of uncertainty.34 In addition, the treatment planning systems normally ignore other sources of radiation, such as leak radiation from the head of the accelerator and other scattered particles.35 Moreover, the secondary irradiation increase linearly with the number of MU used, which may indicate that any of the new methods might still increase the risk of secondary cancer. A calculated estimate of the risk of secondary cancer using the new techniques was not addressed in this study. One advantage of proton radiotherapy is the reduction of the risk of secondary radiation-induced cancer.26,36
Conclusion In conclusion, we found that the new noncoplanar IMRT technique for the cranial part of the craniospinal irradiation significantly reduced the mean doses to the eyes, the inner ears, the parotid glands, and the normal tissue compared with the standard technique for craniospinal irradiation. In addition, the noncoplanar IMRT technique significantly improved dose coverage of the target. Apparent reduced dose to the normal tissue may decrease the risk for secondary cancer. However, the increased number of MU and the uncertainty in the estimation of dose given
A. T. Hansen et al / Medical Dosimetry 40 (2015) 296–303
to the distant tissue need to be taken into account in the interpretation of the findings. Acknowledgments The authors thank Steffen L. Hokland for much appreciated help during the preparation of this article. References 1. Massimino, M.; Giangaspero, F.; Garrè, M.L.; et al. Childhood medulloblastoma. Crit. Rev. Oncol. Hematol. 79:65–83; 2011. 2. Gerber, N.; Mynarek, M.; Hoff, K.; et al. Recent developments and current concepts in medulloblastoma. Cancer Treat. Rev. 40:356–65; 2014. 3. Bartlett, F.; Kortmann, R.; Saran, F. Medulloblastoma. Clin. Oncol. 25:36–45; 2013. 4. Liu, X.; Yu, J.; Yu, Y.; et al. Conventional craniospinal irradiation with patient supine and source-skin distance (SSD) 100 cm for spinal field. Med. Dosim. 36:373–6; 2011. 5. Munshi, A.; Jalali, R. A simple technique of supine craniospinal irradiation. Med. Dosim. 33:1–5; 2008. 6. Hideghét, K.; Cserháti, A.; Nagy, Z.; et al. A prospective study of supine versus prone positioning and whole-body thermoplastic mask fixation for craniospinal radiotherapy in adult patients. Radiother. Oncol. 102:214–8; 2012. 7. Wilkinson, J.M.; Lewis, J.; Lawrence, G.P.; et al. Craniospinal irradiation using a forward planned segmented field technique. Br. J. Radiol. 80:209–15; 2007. 8. Chojnacka, M.; Skowro´nska-Gardas, A.; Morawska-Kaczy´nska, M.; et al. Craniospinal radiotherapy in children: electron- or photon-based technique of spinal irradiation. Rep. Pract. Oncol. Radiother 15:21–4; 2010. 9. Sun, L.M.; Yeh, S.A.; Wang, C.J.; et al. Postoperative radiation therapy for medulloblastoma—high recurrence rate in the subfrontal region. J. Neurooncol. 58:77–85; 2001. 10. Skowrońska-Gardas, A.; Chojnacka, M.; Morawska-Kaczyńska, M.; et al. Patterns of failure in children with medulloblastoma treated with 3D conformal radiotherapy. Radiother. Oncol. 84:26–33; 2007. 11. Weiss, E.; Krebeck, M.; Köhler, B.; et al. Does the standardized helmet technique lead to adequate coverage of the cribriform plate? An analysis of current practice with respect to the ICRU 50 report Int. J. Radiat. Oncol. Biol. Phys. 49:1475–80; 2001. 12. Deasy, J.; Moiseenko, V.; Marks, L.; et al. Radiotherapy dose-volume effects on salivary gland function. Int. J. Radiat. Oncol. Biol. Phys. 76:S58–63; 2010. 13. King, M.; Modlin, L.; Million, L.; et al. Parotid gland as an organ-at-risk for craniospinal irradiation. Int. J. Radiat. Oncol. Biol. Phys. 84:S634; 2012. 14. Bhandare, N.; Jackson, A.; Eisbruch, A.; et al. Radiation therapy and hearing loss. Int. J. Radiat. Oncol. Biol. Phys. 76:S50–7; 2010. 15. Paulino, A.; Lobo, M.; Teh, B.; et al. Ototoxicity after intensity-modulated radiation therapy and cisplatin-based chemotherapy in children with medulloblastoma. Int. J. Radiat. Oncol. Biol. Phys. 78:1445–50; 2010. 16. Kleiman, N. Radiation cataract. Ann. ICRP 41:80–97; 2012. 17. Christopherson, K.; Rotondo, R.; Bradley, J.; et al. Late toxicity following craniospinal radiation for early-stage medulloblastoma. Acta. Oncol. 53: 471–480; 2014.
303
18. Wolden, S.L. Protons for craniospinal radiation: Are clinical data important? Int. J. Radiat. Oncol. Biol. Phys 78:231–2; 2013. 19. Hadley, A.; Ding, G. A single-gradient junction technique to replace multiplejunction shifts for craniospinal irradiation treatment. Med. Dosim. 39:314–9; 2014. 20. Seppälä, J.; Kulmala, J.; Lindholm, P.; et al. A method to improve target dose homogeneity of craniospinal irradiation using dynamic split fields IMRT. Radiother. Oncol. 96:209–15; 2010. 21. Kusters, J.; Louwe, R.; Kollenburg, P.; et al. Optimal normal tissue sparing in craniospinal axis irradiation using IMRT with daily intrafractionally modulated junction(s). Int. J. Radiat. Oncol. Biol. Phys. 81:1405–14; 2011. 22. Cao, F.; Ramaseshan, R.; Corns, R.; et al. A three-isocenter jagged-junction IMRT approach for craniospinal irradiation without beam edge matching for field junctions. Int. J. Radiat. Oncol. Biol. Phys. 84:648–54; 2012. 23. Studenski, M.; Shen, X.; Yu, Y.; et al. Intensity-modulated radiation therapy and volumetric-modulated arc therapy for adult craniospinal irradiation—A comparison with traditional techniques. Med. Dosim. 38:48–54; 2013. 24. Harron, E.; Lewis, J. Bowel sparing in pediatric cranio-spinal radiotherapy: a comparison of combined electron and photon and helical TomoTherapy techniques to a standard photon method. Med. Dosim. 37:140–4; 2012. 25. Sharma, D.S.; Gupta, T.; Jalali, R.; et al. High-precision radiotherapy for craniospinal irradiation: evaluation of three-dimensional conformal radiotherapy, intensity modulated radiation therapy and helical TomoTherapy. Br. J. Radiol. 82:1000–9; 2009. 26. Yoon, M.; Shin, D.H.; Kim, J.; et al. Craniospinal irradiation techniques: A dosimetric comparison of proton beams with standard and advanced photon radiotherapy. Int. J. Radiat. Oncol. Biol. Phys. 81:637–46; 2011. 27. Fogliata, A.; Bergström, S.; Cafaro, I. Cranio-spinal irradiation with volumetric modulated arc therapy: A multi-institutional treatment experience. Radiother. Oncol. 99:79–85; 2011. 28. Lee, Y.; Brooks, C.; Bedford, J.; et al. Development and evaluation of multiple isocentric volumetric modulated arc therapy technique for craniospinal axis radiotherapy planning. Int. J. Radiat. Oncol. Biol. Phys. 82:1006–12; 2012. 29. Fossati, P.; Ricardi, U.; Orecchia, R. Pediatric medulloblastoma: toxicity of current treatment and potential role of proton therapy. Cancer Treat. Rev. 35:79–96; 2009. 30. Stoker, J.B.; Grant, J.; Zhu, R.; et al. Intensity modulated proton therapy for craniospinal irradiation: organ-at-risk exposure and a low-gradient junctioning technique. Int. J. Radiat. Oncol. Biol. Phys. 90:637–44; 2014. 31. Ruben, J.; Davis, S.; Evans, C.; et al. The effect of intensity-modulated radiotherapy on radiation-induced malignancies. Int. J. Radiat. Oncol. Biol. Phys. 70:1530–6; 2008. 32. Schneider, U.; Lomax, A.; Timmermann, B. Second cancers in children treated with modern radiotherapy techniques. Radiother. Oncol. 89:135–40; 2008. 33. Galloway, T.; Indelicato, D.; Amdur, R.; et al. Analysis of dose at the site of second tumor formation after radiotherapy to the central nervous system. Int. J. Radiat. Oncol. Biol. Phys. 82:90–4; 2012. 34. Taddei, P.; Jalbout, W.; Howell, R.; et al. Analytical model for out-of-field dose in photon craniospinal irradiation. Phys. Med. Biol. 58:7463–79; 2013. 35. Zaghloul, M.S. Intensity modulated radiotherapy (IMRT) for pediatric cancer patients: the advantage and fear of second malignant neoplasm. J. Egypt Nat. Cancer Inst 25:1–3; 2013. 36. Zhang, R.; Howell, R.M.; Taddei, P.J.; et al. A comparative study on the risk of radiogenic second cancers and cardiac mortality in a set of pediatric medulloblastoma patients treated with photon or proton craniospinal irradiation. Radiother. Oncol. 113:84–8; 2014.