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

Learn More →

A dosimetric comparison of four treatment planning methods for high grade glioma

A dosimetric comparison of four treatment planning methods for high grade glioma Background: High grade gliomas (HGG) are typically treated with a combination of surgery, radiotherapy and chemotherapy. Three dimensional (3D) conformal radiotherapy treatment planning is still the main stay of treatment for these patients. New treatment planning methods suggest better dose distributions and organ sparing but their clinical benefit is unclear. The purpose of the current study was to compare normal tissue sparing and tumor coverage using four different radiotherapy planning methods in patients with high grade glioma. Methods: Three dimensional conformal (3D), sequential boost IMRT, integrated boost (IB) IMRT and Tomotherapy (TOMO) treatment plans were generated for 20 high grade glioma patients. T1 and T2 MRI abnormalities were used to define GTV and CTV with 2 and 2.5 cm margins to define PTV1 and PTV2 respectively. Results: The mean dose to PTV2 but not to PTV1 was less then 95% of the prescribed dose with IB and IMRT plans. The mean doses to the optic chiasm and the ipsilateral globe were highest with 3D plans and least with IB plans. The mean dose to the contralateral globe was highest with TOMO plans. The mean of the integral dose (ID) to the brain was least with the IB plan and was lower with IMRT compared to 3D plans. The TOMO plans had the least mean D10 to the normal brain but higher mean D50 and D90 compared to IB and IMRT plans. The mean D10 and D50 but not D90 were significantly lower with the IMRT plans compared to the 3D plans. Conclusion: No single treatment planning method was found to be superior to all others and a personalized approach is advised for planning and treating high-grade glioma patients with radiotherapy. Integral dose did not reflect accurately the dose volume histogram (DVH) of the normal brain and may not be a good indicator of delayed radiation toxicity. Page 1 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 ian Medical Systems, Palo Alto, CA). For each patient, a Background High grade gliomas (HGG) are the most prevalent pri- Gross Tumor Volume (GTV) and a Clinical Target Volume mary malignant brain tumors in adults. These malignan- (CTV) [9] were contoured using the contrast enhanced T1 cies are typically treated with a combination of surgery, and the T2 MRI abnormalities, respectively. A 2-cm mar- radiotherapy and chemotherapy. Three dimensional (3D) gin to the CTV was used to define the Planning Target Vol- conformal radiotherapy treatment planning is still the ume 1 (PTV1 [9]), and a 2.5-cm margin to the GTV was main stay of treatment for these patients with treatment used to define PTV2. Areas of the PTV1 and PTV2 that volume delineation based on Magnetic Resonance Images were outside the skull were trimmed with 0.5 cm inner (MRI) fused to the patient's simulation computed tomog- margin to the body contour. The globes and the optic chi- raphy. New technologies for radiotherapy planning and asm were contoured and were designated as organs at risk treatment such as Intensity Modulated Radiotherapy during the treatment planning. The brain stem, subven- (IMRT) and new treatment instruments such as Tomo- tricular zones (SVZ) and the normal brain (the volume of therapy are becoming widely used. These provide better brain that was left after excluding the PTV1 and PTV2 vol- dose conformality, add certainty to dose delivery to the umes using the software Boolean operators) were also target volumes, and allow sparing of sensitive organs adja- contoured for toxicity evaluation, but were not taken into cent to the treatment field and/or escalation of the dose to consideration during treatment planning as organs at risk. the target volumes [1-7]. Although technically feasible, The SVZs, which are believed to harbor the brain progen- the clinical benefit of the use of these technologies in the itor cells [10], were contoured as previously described treatment of HGG patients is unclear. Dose escalation in [11]. Briefly, the lateral ventricles were contoured in both patients with HGG has thus far yielded disappointing sides of the brain. The lateral edges of the ventricles were results [1,2] and the use of advanced planning techniques marked using a brush tool with the width of 0.5 cm. Treat- to spare a presumably healthy tissue surrounding the pri- ment volumes and normal structures were contoured by a mary lesions to reduce toxicity is of uncertain benefit [8]. single physician and verified by a second physician. Furthermore, the problematic quality assurance and reproducibility of some of these advanced treatment plan- For each patient, four treatment plans were generated. ning methods may compromise the ability to test them in Three LINAC based treatment plans included a 3D plan, a controlled randomized trial [3]. In the current study, we an IMRT plan and an IB IMRT plan. These were done for aimed to compare the dose distribution in target volumes Varian Clinac-21EX beams. This machine is equipped as well as normal tissues with four treatment planning with the Millennium120 multi leaf collimator (MLC). The methods, done for the same patients with HGG. Our pur- leaf width for the central 40 pairs is 5 mm and for the pose for conducting this dosimetric comparison was to outer 20 pairs is 1 cm. The simulation CT images and asso- discover the benefits and drawbacks of each planning ciated contours were then transferred from the Eclipse method. We also aimed to evaluate the ability of the cal- treatment planning software to the Tomotherapy treat- culated integral dose (ID) to reflect the actual dose distri- ment planning software (TomoTherapy Inc., Madison, bution in the normal brain. A conformal three WI) using the DICOM-RT protocol to generate the fourth dimensional (3D) plan as well as a Linear Accelerator treatment plan with the Tomotherapy treatment planning (LINAC, Varian Clinac-21EX equipped with the station. The optimal beam arrangement that delivered Millennium120 Multi Leaf Collimator) based sequential optimal tumor coverage and normal tissue sparing was boost IMRT plan were generated. We generated a third selected after comparisons of various beam arrangements. LINAC based plan that was also an IMRT plan but was pre- Dose constrains and priorities were modified as needed in scribed as an Integrated Boost (IB) plan. Tomotherapy the IMRT, IB and Tomotherapy algorithms during the (TOMO) plans were also generated for each patient using optimization process. The 3D LINAC plans typically IB dose prescription. included 3-5 treatment fields to conform the dose for each target volume and the IMRT plans included typically 4-5 non co-planar treatment fields. If possible the dose to con- Methods Twenty adult patients with high grade glioma, previously tra-lateral brain was limited, when this did not compro- treated with conventional 3D conformal radiotherapy at mise the dose to critical structures. The beams were the Radiation Oncology Branch of the National Cancer chosen accordingly. The IMRT and IB plans were identical Institute during the period 2004-2008, were included in in field arrangement and differed only by the dose pre- this study. Available treatment planning simulation CT scription parameters. The helical Tomotherapy parame- images and diagnostic contrast enhanced pre-operative T1 ters definitions were 1 cm for the field size (slice and T2 MR Images were mandatory for the patients to be thickness) and 0.2-0.3 for pitch (the ratio of the distance included. the couch travels to the field width per one full rotation of the gantry). The Planning Modulation Factor (the ratio The contrast enhanced MR images were fused to the sim- between the longest time a leaf is opened to the mean leaf ulation CT images using the Eclipse planning system (Var- opening time) was 2.00 and the mean actual modulation Page 2 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 factor was 1.71 (range 1.22-1.96). The Plan Calculation globes below 5 Gy. Chiasm but not globes constrains had Grid (image resolution for dose calculations) was typi- higher priority than the target volume inhomogeneities in cally 0.274*0.274 cm (range 0.196*0.196 cm to the IMRT, IB and TOMO optimization algorithms when 0.424*0.424 cm). there was an overlap of the structures. The brain stem, the subventricular zones and the normal brain did not have Dose calculations for all plans were based on photon dose constrains during treatment planning. Bioequivalent beams with maximal energy of 6-15 MV. The 3D and dose calculations were used to allow the comparison of IMRT plans included two sequential plans each, using the the IB and TOMO plans to the 3D and IMRT plans. PTV1 and PTV2 (boost) as the target volumes, with 46 Gy in 23 fractions and 14 Gy in 7 fractions, respectively, as Two methods were used to compare the dose distribution the prescribed doses. Both the LINAC based IB plan and in the normal brain. First, the Integral Dose (ID) was cal- the TOMO plan were prescribed as integrated boost plans culated as previously described. Briefly, the volume of the as previously described [4]. Briefly, the integrated boost normal brain was multiplied by the mean dose to the included 23 fractions in a single plan, with a differential brain [12,13]. Since different dose volume histogram dose prescription to the target volumes. A total dose of 46 (DVH) curves can generate the same ID value (figure 2) Gy in 2 Gy fractions was prescribed to PTV1 and a total we decided to compare three points on the DVH of the dose of 53.8 Gy in 2.34 Gy fractions was prescribed for normal brain in each plan. D10, D50 and D90 represent PTV2. The PTV2 total dose was calculated as the bioequiv- the dose received by 10%, 50% and 90% of the normal alent dose of 30 fractions of 2 Gy given in 23 fractions brain volume, respectively according to the linear quadratic model with a α/β ratio of 3. The integrated boost concept is illustrated in figure 1. The mean percent volume coverage of the target volumes Acceptable inhomogeneity was defined as 5% above and as well as the mean of the maximal dose to normal organs 7% below the prescribed dose inside the target volumes. were calculated for each plan. The means of the IC, the An inhomogeneity coefficient (IC) of the dose in the tar- normal brain ID and D10, D50 and D90 to the brain were get volumes was calculated using the formula (Dmax- also calculated. Since all these values relate to the same Dmin)/Dmean as previously described [12]. The closer volumes, there was no need to normalize them for the the IC to zero, the more homogenous the plan was con- purpose of this comparison. sidered. An Excel based (Microsoft Office) two-tailed paired stu- The maximal dose allowed to the optic chiasm was 54 Gy, dent T test was used to determine if there was a statistically in the 3D and IMRT plans and 51.5 Gy in the IB and TOMO plans. An effort was made to keep the dose to the PTV 2, 53.8 Gy 2.34Gy/fraction PTV 1, 46Gy 2Gy/fraction A sample Dose Volume Histog dose with two differ Figure 2 ent plans rams (DVH) of normal brain A sample Dose Volume Histograms (DVH) of normal brain dose with two different plans. Although the Inte- Schematic and dose prescription Figure 1 illustration of the Integrated Boost target volumes gral Dose to the brain according to these to DVH's is the Schematic illustration of the Integrated Boost target same (15.4 Gyxcm × 1000), it is obvious that these histo- volumes and dose prescription. Abbreviation: PTV1 - grams are different in both high and low dose areas. The Planning Treatment Volume 1 (corresponds with T2 MR dose received by 10% of the brain volume (D10) and 90% of Image abnormality with 2 cm margins), PTV2 - Planning the brain volume (D90) can describe more accurately such a Treatment Volume 2 (corresponds with T1+ contrast MR difference. Abbreviation: NA-not applicable. SD-Standard Image abnormality with 2.5 cm margins). Deviation. Page 3 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 significant difference between the means of the above val- 3D plans regarding PTV 1 (p < 0.02) and higher (worse) ues accomplished by each treatment planning method. A then the mean IC of the IB plans regarding PTV2 (p < statistically significant difference was defined when the T 0.03) (figure 3). No significant difference was found test resulted in a p value of < 0.05. between the means of the IC of the 3D and IB plans. Results Normal tissue sparing Patients' characteristics The mean of the maximal dose to the normal structures A total of 20 patients were included in our study, 11 males with the various treatment plans was used as a surrogate and 9 females with a mean age of 54 y (range 37-71). to normal tissue sparing [14] (figure 4). The optic chiasm, Nineteen patients had a pathological diagnosis of Gliob- which was designated as organ at risk during treatment lastoma Multiforme (GBM, World Health Organization planning, was better spared with all other planning meth- grade IV) and one had Anaplastic Astrocytoma (AA, World ods compared to the 3D plan (p < 0.05 for all compari- Health Organization grade III). sons). The IB plan (p < 0.00002) but not the TOMO plan (p > 0.5) spared the optic chiasm significantly better then Target volumes' coverage the IMRT plan. All plans met the pre defined dose con- The mean PTV1 and PTV2 volumes were 452 cm (range strains for the optic chiasm. 3 3 3 276-1074 cm ) and 300 cm (range 137-567 cm ) respec- tively. A mean of >98% (range 92-100%) of the PTV1 The TOMO and IB plans spared the ipsilateral globe sig- received 100% of the prescribed dose in all planning nificantly better then the 3D plans (p < 0.002, p < 0.0008 methods. A mean of 95.5% and 95.7% of the PTV2 respectively) but only the IB plan spared that globe signif- received 100% of the prescribed dose with TOMO and icantly better then the IMRT plan (p < 0.005). No signifi- IMRT plans respectively. A mean of 94% and 92% of the cant difference was found between the IB and the TOMO PTV2 received 100% of the prescribed dose with IB and sparing of these organs (p > 0.3). 3D treatment plans, respectively. The mean IB and 3D plans PTV2 coverage was significantly inferior then the The contralateral globe in contrast, was not spared with IMRT and TOMO plans (p < 0.02, for all comparisons). the TOMO plan which had the highest mean maximal dose compared to all other plans (p value < 0.0001 com- The mean Inhomogeneity Coefficient (IC) was signifi- pared to the IMRT and IB plans and p > 0.05 compared to cantly lower (better) with the TOMO plans, compared to the 3D plan). No other significant difference was found all other plans for both PTV1 and PTV2 (p < 0.0003 for all between the plans in respect to the ability to spare the con- comparisons) (figure 3). The mean IC of the IMRT plans was significantly higher (worse) than the mean IC of the 0.50 0.41 0.38 0.35 0.40 0.30 0.23 0.26 0.28 Ipsi Contra Ipsi Contra 0.20 0.25 Brain Optic lateral lateral Stem Chiasm SVZ lateral globe lateral 0.10 3D 6176 5733 5319 3737 2758 756 0.12 0.00 PTV 1 6380 5675 4999 3383 2024 480 IMRT PTV 2 3D IMRT 6593 5887 5269 3036 1706 443 IB IB TOMO TOMO 6425 5842 4750 3117 1837 796 Th different pla Figure 3 e mean Inh nning methods omogeneity Coefficient (IC) achieved by the The mean ma each treatm Figure 4 ent xima plan l dose (cGy) in ning method normal tissues found with The mean Inhomogeneity Coefficient (IC) achieved The mean maximal dose (cGy) in normal tissues by the different planning methods. The mean of the found with each treatment planning method. Abbrevi- Inhomogeneity Coefficient is a measure of dose inhomogene- ations: SVZ- sub ventricular zone, 3D- conformal three ity in the target volumes. The closer the IC to zero, the dimensional, IMRT-Intensity Modulated Radio Therapy, IB- more homogenous the dose is. Integrated Boost, TOMO - Tomo Therapy. Page 4 of 7 (page number not for citation purposes) Inhomogeneity Co ef f icien t (IC) Dose (cGy) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 tralateral globe. Although the mean dose of the IMRT and IB plans to the contralateral globe was lower then the 3D plan, a careful evaluation of the dose to that globe in some individual cases was lowest with the 3D plan compared to all others (data not shown). 3000 The mean maximal dose to the ipsilateral subventricular zone (SVZ) was the highest with the IB treatment plan- ning followed by TOMO, IMRT and 3D (p < 0.05 for all comparisons except IMRT vs. TOMO plans). The contral- ateral SVZ was best spared using the IMRT (p < 0.05 com- D10 D50 D90 pared to every other planning method). No significant 4167 1676 375 IB difference was found between the IB, 3D and TOMO 3931 2152 583 TOMO plans in this respect. IMRT 4762 1791 435 5467 2008 425 3D The mean maximal dose to the brain stem with IMRT plan The mean D10, D5 pla Figure 5 nning method 0, and D90 found with each treatment was significantly lower than with IB and TOMO (p < The mean D10, D50, and D90 found with each treat- 0.008 for both comparisons) but not the 3D plan (p > ment planning method. The mean dose to 10% (D10), 0.6). 50% (D50) and 90% (D90) of the normal brain volume (after the PTV1 and PTV2 volumes were excluded using the soft- The mean of the integral dose (ID) to the brain was signif- ware Boolean Operators) with each treatment planning icantly lower (range 10-17.5%) with the IB plan com- method. pared to all other plans (p < 0.006) (table 1). No significant difference in the ID to the brain was found among the other plans. A different pattern was noted treatment planning method that is superior to all others in when the D10, 50 and 90 were extracted out of the DVHs all aspects compared. of the various plans and compared. The mean dose to 10 percent of the normal brain was consistently lowest with Target volumes' coverage the TOMO plans followed by the IB, IMRT and 3D plans. Sequential boost plans assume 100% dose coverage to the Conversely, the D50 and D90 values were significantly boost volume by the initial part of the treatment. In gli- higher with the TOMO plans compared to both IB and oma patients, were this is not always the case (PTV2 is not IMRT plans and the D50 but not the D90 was significantly a geometrical cone down of PTV1); cold spots might be lower in the IB plan compared to the IMRT and 3D plans. noticed within the boost volume. Non standard target vol- Interestingly, IMRT D10 and D50 were significantly lower umes [4,5] can overcome this problem but these were not than the 3D doses, but no significant difference was found tested prospectively [15]. between these plans regarding D90 (figure 5). More treatment fields may suggest some advantage in the Discussion PTV2 dose coverage (IMRT plan was better then the 3D In this study, we compared the delivery of radiation dose plan). to the target volumes and the adjacent normal structures in high-grade glioma patients by using four treatment Surprisingly, the PTV2 dose coverage with the IB plan was planning methods. This kind of comparison harbors worse then the IMRT plan despite the use of the same numerous biases due to the use of different planning soft- beam arrangement and dose constrains. As previously ware, different optimization algorithms and different reported [4], there is a trade off between the target volume dose prescriptions. The use of mean values to compare coverage and the homogeneity of the dose (which is sup- these plans harbors another potential bias since it fails to ported by our IC results). Prescribing the IB plans to a reflect a better dose profile offered for individual patients lower isodose line improved the coverage but compro- by a specific planning method (e.g. target volume dose mised the homogeneity of the dose in the target volume goals). Furthermore, a lower dose to a normal tissue is an (data not shown). The beam weighing algorithm used to important goal in treatment planning (e.g. SVZ or normal produce a plan sum in a sequential boost (IMRT) com- brain), but does not necessarily give an advantage if the pared to an integrated boost (IB) gives another explana- tolerance of that tissue is not met (e.g. the optic chiasm). tion to the different PTV2 coverage [4]. A qualitative comparison of the various plans is summa- The TOMO plan did not result in an inferior PTV2 cover- rized in table 2. According to our results, there is no single age and achieved the best IC as well since it uses infinite number of fields by definition. Page 5 of 7 (page number not for citation purposes) D10 D50 D90 Dose (cGy) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 Table 1: The mean Integral Dose (ID) to the brain with each treatment planning method. % Difference (p value) Integral Dose ± SD (Gy × cm × 1000) Plan 3D IMRT IB 22.8 ± 7.2 3D NA 21.1 ± 3.7 IMRT 7.5 (0.2) NA 18.8 ± 3.1 IB 17.5 (0.006)* 10.9 (8.7E-09)* NA 21 ± 2.3 TOMO 7.8 (0.3) 0.3 (0.6) 10.6 (0.003)* * Statistically significant Normal tissue sparing tumors, associated with the tumor ability to resist radia- The TOMO plans were able to spare best small organs tion treatment and recur[19-21]. Higher radiosensitivity which usually lie close to the target volumes and were of the brain stem compared to other parts of the brain is highly prioritized (e.g. the chiasm and ipsilateral globe) not reported and its tolerance to fractionated radiotherapy but the advantage was relatively small compared to the appears to be a function of the volume receiving high dose IMRT and IB plans. The ability of the TOMO to spare these rather than the maximum point dose [22]. organs in patients with tumors located distant from them is questionable since blocking both the entering and exit Integral Dose (ID) is the value usually used to compare doses to a distant volume without compromising the tar- the dose received by healthy tissues outside the target vol- get volume coverage, is unlikely with the helical beam umes [12-14]. We added an analysis of three points in the arrangement. DVH curve (D10, D50 and D90) of the normal brain to evaluate the validity of the ID as a tool for this kind of No benefit and even a potential disadvantage was found comparisons since different DVH curves may result in the with the IB and TOMO plans for structures with larger vol- same ID. The TOMO plans, with the high conformality umes that often overlap with the target volumes, such as and a rapid fall off of the dose around the target volumes the SVZ and the brain stem (due to a higher dose per frac- had the lowest D10 to the normal brain (only a small vol- tion translated into higher total BED in these areas). We ume around the target volume received a high dose). At chose not to give dose constrains to these structures and the same time, larger volumes of normal brain received compared their dose distribution after the plans were gen- irradiation at all (high D50 and D90). This significant erated. The SVZs (believed to harbor the normal brain dose of irradiation received by the normal brain was not progenitor cells [10]) viability is correlated with late radi- demonstrated when the ID values were compared. Along ation toxicity [11,16,17] on one hand justifying an with the expected longer survival of HGG patients in the attempt to lower their dose [11,18] but are also suspected future these differences may translate to toxicity. Lower ID as the source of cancerous stem cells in primary brain values with IMRT compared to 3D plans were used to jus- tify its use for HGG patients [14]. Our mean ID failed to Table 2: A qualitative comparison of the four treatment planning indicate such a difference but it is in line with the lower methods. D10 and D50 for the IMRT plans we found. We suggest caution in the use and interpretation of ID for treatment 3D IMRT IB TOMO plans comparisons due to its failure to predict differences in DVH curves which might have significant implications Target Volumes in the future. PTV1 coverage ++ + + PTV2 coverage -+ - + Inhomogeneity Coefficient -+ Conclusion Normal Tissues Sparing Our data suggest a distinctive approach in the use of new Optic Chiasm -+ treatment planning tools. Further investigation is indi- Ipsilateral Globe -+ + cated to better choose the correct tool for each patient. Contralateral Globe +- Larger series might suggest a decision algorithm according Ipsilateral SVZ +- to the patient's tumor size, location and prognosis. Long Contralateral SVZ + Brainstem +- - follow up periods with HGG patients are becoming a Normal Brain ID + widespread phenomenon, and may allow better under- Normal Brain D10 -+ standing of the effect of the different DVH curves of the Normal Brain D50 -+ - normal brain and the SVZ. A close follow up on patient's Normal Brain D90 - toxicity profiles and correlations with a specific treatment planning method are indicated. The use of imprecise and + Significant advantage - Significant disadvantage Page 6 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 10. Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM, Alvarez-Buylla A: insensitive tools like ID to compare potential toxicity due Subventricular zone astrocytes are neural stem cells in the to large irradiated volumes should be discouraged and adult mammalian brain. Cell 1999, 97(6):703-16. better tools should be developed. 11. Barani IJ, Cuttino LW, Benedict SH, Todor D, Bump E, Wu Y, Chung T, Broaddus W, Lin P: Neural stem cell-preserving external- beam radiotherapy of central nervous system malignancies. Competing interests IJROBP 2007, 68(4):978-85. 12. Shi C, Peñagarìcano J, Papanikolaou N: Comparison of IMRT The authors declare that they have no competing interests. treatment plans between linac and helical tomotherapy based on integral dose and inhomogeneity index. Med Dosim Authors' contributions 2008, 33(3):215-21. 13. Aoyama H, Westerly DC, MacKie TR, Olivera G, Bentzen S, Patel R, LZ carried out the contouring, and participated in the Jaradat H, Tome W, Ritter M, Mehta M: Integral radiation dose to study design, coordination, treatment planning and writ- normal structures with conformal external beam radiation. ing of the manuscript. BS participated in contouring and IJROBP 2006, 64(3):962-7. 14. Hermanto U, Frija EK, Lii MJ, Chang E, Mhajan A, Woo S: Intensity- helped revising the draft manuscript. HN carried out the modulated radiotherapy (IMRT) and conventional three- treatment planning. JO carried out MRI fusions and par- dimensional conformal radiotherapy for high-grade gliomas: Does IMRT increase the integral dose to normal brain? ticipated in treatment planning. BA participated in treat- IJROBP 2007, 67(4):1135-44. ment planning. US participated in the statistical analysis 15. Fiveash JB, Nordal JM, Markert RS, Ahmed RS, Nabors LB: High of the results and helped revising the draft manuscript. Grade Gliomas. In Clinical Radiation Oncology 2nd edition. Edited by: Gunderson L, Tepper J. Churchill Livingstone, Philadelphia, (PA); RWM participated in treatment planning. DC participated 2006:515-537. in the data analysis and helped revising the draft manu- 16. Barani IJ, Benedict SH, Lin P: Neural stem cells: Implications for the conventional radiotherapy of central nervous system script. KC conceived of the study, and participated in its malignancies. IJROBP 2007, 68(2):324-33. design and coordination and helped to draft the manu- 17. Mizumatsu S, Monje ML, Morhardt DR, Rola R, Palmer T, Fike J: script. All authors read and approved the final manu- Extreme sensitivity of adult neurogenesis to low doses of X- irradiation. Cancer Res 2003, 63(14):4021-7. script. 18. Gutieìrrez AN, Westerly DC, Tomeì WA, Jaradat H, Mackie T, Bentzen S, Khuntia D, Mehta M: Whole brain radiotherapy with References hippocampal avoidance and simultaneously integrated brain metastases boost: A planning study. IJROBP 2007, 69(2):589-97. 1. Souhami L, Seiferheld W, Brachman D, Podgorsak P, Werner-Wasik 19. Quiñones-Hinojosa A, Chaichana K: The human subventricular M, Lustig R, Schultz C, Sause W, Okunieff P, Buckner J: Randomized zone: A source of new cells and a potential source of brain comparison of stereotactic radiosurgery followed by conven- tumors. Exp Neurol 2007, 205(2):313-24. tional radiotherapy with carmustine to conventional radio- 20. Lim DA, Cha S, Mayo MC, Chen MH, Keles E, VandenBerg S, Berger therapy with carmustine for patients with glioblastoma M: Relationship of glioblastoma multiforme to neural stem multiforme: Report of radiation therapy oncology group 93- cell regions predicts invasive and multifocal tumor pheno- 05 protocol. IJROBP 2004, 60(3):853-60. type. Neuro Oncol 2007, 9(4):424-9. 2. Cardinale R, Won M, Choucair A, Gillin M, Chakravarti A, Schultz C, 21. Sanai N, Alvarez-Buylla A, Berger MS: Mechanisms of disease: Souhami L, Chen A, Pham H, Mehta M: A phase II trial of acceler- Neural stem cells and the origin of gliomas. N Engl J Med 2005, ated radiotherapy using weekly stereotactic conformal 353(8):811-22. boost for supratentorial glioblastoma multiforme: RTOG 22. Debus J, Hug EB, Liebsch NJ, O'farrel D, Finkelstein D, Efird J, 0023. IJROBP 2006, 65(5):1422-8. Munzenrider J: Brainstem tolerance to conformal radiother- 3. Hartford AC, Palisca MG, Eichler TJ, Beyer D, Devineni V, Ibbott G, apy of skull base tumors. IJROBP 1997, 39(5):967-75. Kavanagh B, Kent J, Rosenthal S, Schultz C, Tripuraneni P, Gaspar L: American society for therapeutic radiology and oncology (ASTRO) and american college of radiology (ACR) practice guidelines for intensity-modulated radiation therapy (IMRT). IJROBP 2009, 73(1):9-14. 4. Thilmann C, Zabel A, Grosser KH, Hoess A, Wannenmacher M, Debus J: Intensity-modulated radiotherapy with an integrated boost to the macroscopic tumor volume in the treatment of high-grade gliomas. Int J Cancer 2001, 96(6):341-9. 5. Suzuki M, Nakamatsu K, Kanamori S, Okumra M, Uchiyama T, Akai F, Nishimura Y: Feasibility study of the simultaneous integrated boost (SIB) method for malignant gliomas using intensity- modulated radiotherapy (IMRT). Jpn J Clin Oncol 2003, 33(6):271-7. 6. Suzuki M, Nakamatsu K, Kanamori S, Okajima K, Okumura M, Publish with Bio Med Central and every Nishimura Y: Comparison of outcomes between overlapping scientist can read your work free of charge structure-based and non-overlapping structure-based opti- mization for simultaneous integrated boost IMRT for malig- "BioMed Central will be the most significant development for nant gliomas. Int J Clin Oncol 2004, 9(6):491-7. disseminating the results of biomedical researc h in our lifetime." 7. Iuchi T, Hatano K, Narita Y, Kodama T, Yamaki T, Osato K: Hypof- Sir Paul Nurse, Cancer Research UK ractionated high-dose irradiation for the treatment of malig- nant astrocytomas using simultaneous integrated boost Your research papers will be: technique by IMRT. IJROBP 2006, 64(5):1317-24. available free of charge to the entire biomedical community 8. Chan JL, Lee SW, Fraass BA, Normolle DP, Greenberg HS, Junck LR, Gebarski SS, Sandler HM: Survival and failure patterns of high- peer reviewed and published immediately upon acceptance grade gliomas after three-dimensional conformal radiother- cited in PubMed and archived on PubMed Central apy. JCO 2002, 20(6):1635-42. 9. ICRU report 50, prescribing, recording, and reporting pho- yours — you keep the copyright ton beam therapy. International commission on Radiatrion BioMedcentral Submit your manuscript here: units and Measurments. Bethesda, MD 1993. http://www.biomedcentral.com/info/publishing_adv.asp Page 7 of 7 (page number not for citation purposes) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Radiation Oncology Springer Journals

A dosimetric comparison of four treatment planning methods for high grade glioma

Loading next page...
 
/lp/springer-journals/a-dosimetric-comparison-of-four-treatment-planning-methods-for-high-cHvZVXJzPI
Publisher
Springer Journals
Copyright
Copyright © 2009 by Zach et al; licensee BioMed Central Ltd.
Subject
Medicine & Public Health; Oncology; Radiotherapy
eISSN
1748-717X
DOI
10.1186/1748-717X-4-45
pmid
19845946
Publisher site
See Article on Publisher Site

Abstract

Background: High grade gliomas (HGG) are typically treated with a combination of surgery, radiotherapy and chemotherapy. Three dimensional (3D) conformal radiotherapy treatment planning is still the main stay of treatment for these patients. New treatment planning methods suggest better dose distributions and organ sparing but their clinical benefit is unclear. The purpose of the current study was to compare normal tissue sparing and tumor coverage using four different radiotherapy planning methods in patients with high grade glioma. Methods: Three dimensional conformal (3D), sequential boost IMRT, integrated boost (IB) IMRT and Tomotherapy (TOMO) treatment plans were generated for 20 high grade glioma patients. T1 and T2 MRI abnormalities were used to define GTV and CTV with 2 and 2.5 cm margins to define PTV1 and PTV2 respectively. Results: The mean dose to PTV2 but not to PTV1 was less then 95% of the prescribed dose with IB and IMRT plans. The mean doses to the optic chiasm and the ipsilateral globe were highest with 3D plans and least with IB plans. The mean dose to the contralateral globe was highest with TOMO plans. The mean of the integral dose (ID) to the brain was least with the IB plan and was lower with IMRT compared to 3D plans. The TOMO plans had the least mean D10 to the normal brain but higher mean D50 and D90 compared to IB and IMRT plans. The mean D10 and D50 but not D90 were significantly lower with the IMRT plans compared to the 3D plans. Conclusion: No single treatment planning method was found to be superior to all others and a personalized approach is advised for planning and treating high-grade glioma patients with radiotherapy. Integral dose did not reflect accurately the dose volume histogram (DVH) of the normal brain and may not be a good indicator of delayed radiation toxicity. Page 1 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 ian Medical Systems, Palo Alto, CA). For each patient, a Background High grade gliomas (HGG) are the most prevalent pri- Gross Tumor Volume (GTV) and a Clinical Target Volume mary malignant brain tumors in adults. These malignan- (CTV) [9] were contoured using the contrast enhanced T1 cies are typically treated with a combination of surgery, and the T2 MRI abnormalities, respectively. A 2-cm mar- radiotherapy and chemotherapy. Three dimensional (3D) gin to the CTV was used to define the Planning Target Vol- conformal radiotherapy treatment planning is still the ume 1 (PTV1 [9]), and a 2.5-cm margin to the GTV was main stay of treatment for these patients with treatment used to define PTV2. Areas of the PTV1 and PTV2 that volume delineation based on Magnetic Resonance Images were outside the skull were trimmed with 0.5 cm inner (MRI) fused to the patient's simulation computed tomog- margin to the body contour. The globes and the optic chi- raphy. New technologies for radiotherapy planning and asm were contoured and were designated as organs at risk treatment such as Intensity Modulated Radiotherapy during the treatment planning. The brain stem, subven- (IMRT) and new treatment instruments such as Tomo- tricular zones (SVZ) and the normal brain (the volume of therapy are becoming widely used. These provide better brain that was left after excluding the PTV1 and PTV2 vol- dose conformality, add certainty to dose delivery to the umes using the software Boolean operators) were also target volumes, and allow sparing of sensitive organs adja- contoured for toxicity evaluation, but were not taken into cent to the treatment field and/or escalation of the dose to consideration during treatment planning as organs at risk. the target volumes [1-7]. Although technically feasible, The SVZs, which are believed to harbor the brain progen- the clinical benefit of the use of these technologies in the itor cells [10], were contoured as previously described treatment of HGG patients is unclear. Dose escalation in [11]. Briefly, the lateral ventricles were contoured in both patients with HGG has thus far yielded disappointing sides of the brain. The lateral edges of the ventricles were results [1,2] and the use of advanced planning techniques marked using a brush tool with the width of 0.5 cm. Treat- to spare a presumably healthy tissue surrounding the pri- ment volumes and normal structures were contoured by a mary lesions to reduce toxicity is of uncertain benefit [8]. single physician and verified by a second physician. Furthermore, the problematic quality assurance and reproducibility of some of these advanced treatment plan- For each patient, four treatment plans were generated. ning methods may compromise the ability to test them in Three LINAC based treatment plans included a 3D plan, a controlled randomized trial [3]. In the current study, we an IMRT plan and an IB IMRT plan. These were done for aimed to compare the dose distribution in target volumes Varian Clinac-21EX beams. This machine is equipped as well as normal tissues with four treatment planning with the Millennium120 multi leaf collimator (MLC). The methods, done for the same patients with HGG. Our pur- leaf width for the central 40 pairs is 5 mm and for the pose for conducting this dosimetric comparison was to outer 20 pairs is 1 cm. The simulation CT images and asso- discover the benefits and drawbacks of each planning ciated contours were then transferred from the Eclipse method. We also aimed to evaluate the ability of the cal- treatment planning software to the Tomotherapy treat- culated integral dose (ID) to reflect the actual dose distri- ment planning software (TomoTherapy Inc., Madison, bution in the normal brain. A conformal three WI) using the DICOM-RT protocol to generate the fourth dimensional (3D) plan as well as a Linear Accelerator treatment plan with the Tomotherapy treatment planning (LINAC, Varian Clinac-21EX equipped with the station. The optimal beam arrangement that delivered Millennium120 Multi Leaf Collimator) based sequential optimal tumor coverage and normal tissue sparing was boost IMRT plan were generated. We generated a third selected after comparisons of various beam arrangements. LINAC based plan that was also an IMRT plan but was pre- Dose constrains and priorities were modified as needed in scribed as an Integrated Boost (IB) plan. Tomotherapy the IMRT, IB and Tomotherapy algorithms during the (TOMO) plans were also generated for each patient using optimization process. The 3D LINAC plans typically IB dose prescription. included 3-5 treatment fields to conform the dose for each target volume and the IMRT plans included typically 4-5 non co-planar treatment fields. If possible the dose to con- Methods Twenty adult patients with high grade glioma, previously tra-lateral brain was limited, when this did not compro- treated with conventional 3D conformal radiotherapy at mise the dose to critical structures. The beams were the Radiation Oncology Branch of the National Cancer chosen accordingly. The IMRT and IB plans were identical Institute during the period 2004-2008, were included in in field arrangement and differed only by the dose pre- this study. Available treatment planning simulation CT scription parameters. The helical Tomotherapy parame- images and diagnostic contrast enhanced pre-operative T1 ters definitions were 1 cm for the field size (slice and T2 MR Images were mandatory for the patients to be thickness) and 0.2-0.3 for pitch (the ratio of the distance included. the couch travels to the field width per one full rotation of the gantry). The Planning Modulation Factor (the ratio The contrast enhanced MR images were fused to the sim- between the longest time a leaf is opened to the mean leaf ulation CT images using the Eclipse planning system (Var- opening time) was 2.00 and the mean actual modulation Page 2 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 factor was 1.71 (range 1.22-1.96). The Plan Calculation globes below 5 Gy. Chiasm but not globes constrains had Grid (image resolution for dose calculations) was typi- higher priority than the target volume inhomogeneities in cally 0.274*0.274 cm (range 0.196*0.196 cm to the IMRT, IB and TOMO optimization algorithms when 0.424*0.424 cm). there was an overlap of the structures. The brain stem, the subventricular zones and the normal brain did not have Dose calculations for all plans were based on photon dose constrains during treatment planning. Bioequivalent beams with maximal energy of 6-15 MV. The 3D and dose calculations were used to allow the comparison of IMRT plans included two sequential plans each, using the the IB and TOMO plans to the 3D and IMRT plans. PTV1 and PTV2 (boost) as the target volumes, with 46 Gy in 23 fractions and 14 Gy in 7 fractions, respectively, as Two methods were used to compare the dose distribution the prescribed doses. Both the LINAC based IB plan and in the normal brain. First, the Integral Dose (ID) was cal- the TOMO plan were prescribed as integrated boost plans culated as previously described. Briefly, the volume of the as previously described [4]. Briefly, the integrated boost normal brain was multiplied by the mean dose to the included 23 fractions in a single plan, with a differential brain [12,13]. Since different dose volume histogram dose prescription to the target volumes. A total dose of 46 (DVH) curves can generate the same ID value (figure 2) Gy in 2 Gy fractions was prescribed to PTV1 and a total we decided to compare three points on the DVH of the dose of 53.8 Gy in 2.34 Gy fractions was prescribed for normal brain in each plan. D10, D50 and D90 represent PTV2. The PTV2 total dose was calculated as the bioequiv- the dose received by 10%, 50% and 90% of the normal alent dose of 30 fractions of 2 Gy given in 23 fractions brain volume, respectively according to the linear quadratic model with a α/β ratio of 3. The integrated boost concept is illustrated in figure 1. The mean percent volume coverage of the target volumes Acceptable inhomogeneity was defined as 5% above and as well as the mean of the maximal dose to normal organs 7% below the prescribed dose inside the target volumes. were calculated for each plan. The means of the IC, the An inhomogeneity coefficient (IC) of the dose in the tar- normal brain ID and D10, D50 and D90 to the brain were get volumes was calculated using the formula (Dmax- also calculated. Since all these values relate to the same Dmin)/Dmean as previously described [12]. The closer volumes, there was no need to normalize them for the the IC to zero, the more homogenous the plan was con- purpose of this comparison. sidered. An Excel based (Microsoft Office) two-tailed paired stu- The maximal dose allowed to the optic chiasm was 54 Gy, dent T test was used to determine if there was a statistically in the 3D and IMRT plans and 51.5 Gy in the IB and TOMO plans. An effort was made to keep the dose to the PTV 2, 53.8 Gy 2.34Gy/fraction PTV 1, 46Gy 2Gy/fraction A sample Dose Volume Histog dose with two differ Figure 2 ent plans rams (DVH) of normal brain A sample Dose Volume Histograms (DVH) of normal brain dose with two different plans. Although the Inte- Schematic and dose prescription Figure 1 illustration of the Integrated Boost target volumes gral Dose to the brain according to these to DVH's is the Schematic illustration of the Integrated Boost target same (15.4 Gyxcm × 1000), it is obvious that these histo- volumes and dose prescription. Abbreviation: PTV1 - grams are different in both high and low dose areas. The Planning Treatment Volume 1 (corresponds with T2 MR dose received by 10% of the brain volume (D10) and 90% of Image abnormality with 2 cm margins), PTV2 - Planning the brain volume (D90) can describe more accurately such a Treatment Volume 2 (corresponds with T1+ contrast MR difference. Abbreviation: NA-not applicable. SD-Standard Image abnormality with 2.5 cm margins). Deviation. Page 3 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 significant difference between the means of the above val- 3D plans regarding PTV 1 (p < 0.02) and higher (worse) ues accomplished by each treatment planning method. A then the mean IC of the IB plans regarding PTV2 (p < statistically significant difference was defined when the T 0.03) (figure 3). No significant difference was found test resulted in a p value of < 0.05. between the means of the IC of the 3D and IB plans. Results Normal tissue sparing Patients' characteristics The mean of the maximal dose to the normal structures A total of 20 patients were included in our study, 11 males with the various treatment plans was used as a surrogate and 9 females with a mean age of 54 y (range 37-71). to normal tissue sparing [14] (figure 4). The optic chiasm, Nineteen patients had a pathological diagnosis of Gliob- which was designated as organ at risk during treatment lastoma Multiforme (GBM, World Health Organization planning, was better spared with all other planning meth- grade IV) and one had Anaplastic Astrocytoma (AA, World ods compared to the 3D plan (p < 0.05 for all compari- Health Organization grade III). sons). The IB plan (p < 0.00002) but not the TOMO plan (p > 0.5) spared the optic chiasm significantly better then Target volumes' coverage the IMRT plan. All plans met the pre defined dose con- The mean PTV1 and PTV2 volumes were 452 cm (range strains for the optic chiasm. 3 3 3 276-1074 cm ) and 300 cm (range 137-567 cm ) respec- tively. A mean of >98% (range 92-100%) of the PTV1 The TOMO and IB plans spared the ipsilateral globe sig- received 100% of the prescribed dose in all planning nificantly better then the 3D plans (p < 0.002, p < 0.0008 methods. A mean of 95.5% and 95.7% of the PTV2 respectively) but only the IB plan spared that globe signif- received 100% of the prescribed dose with TOMO and icantly better then the IMRT plan (p < 0.005). No signifi- IMRT plans respectively. A mean of 94% and 92% of the cant difference was found between the IB and the TOMO PTV2 received 100% of the prescribed dose with IB and sparing of these organs (p > 0.3). 3D treatment plans, respectively. The mean IB and 3D plans PTV2 coverage was significantly inferior then the The contralateral globe in contrast, was not spared with IMRT and TOMO plans (p < 0.02, for all comparisons). the TOMO plan which had the highest mean maximal dose compared to all other plans (p value < 0.0001 com- The mean Inhomogeneity Coefficient (IC) was signifi- pared to the IMRT and IB plans and p > 0.05 compared to cantly lower (better) with the TOMO plans, compared to the 3D plan). No other significant difference was found all other plans for both PTV1 and PTV2 (p < 0.0003 for all between the plans in respect to the ability to spare the con- comparisons) (figure 3). The mean IC of the IMRT plans was significantly higher (worse) than the mean IC of the 0.50 0.41 0.38 0.35 0.40 0.30 0.23 0.26 0.28 Ipsi Contra Ipsi Contra 0.20 0.25 Brain Optic lateral lateral Stem Chiasm SVZ lateral globe lateral 0.10 3D 6176 5733 5319 3737 2758 756 0.12 0.00 PTV 1 6380 5675 4999 3383 2024 480 IMRT PTV 2 3D IMRT 6593 5887 5269 3036 1706 443 IB IB TOMO TOMO 6425 5842 4750 3117 1837 796 Th different pla Figure 3 e mean Inh nning methods omogeneity Coefficient (IC) achieved by the The mean ma each treatm Figure 4 ent xima plan l dose (cGy) in ning method normal tissues found with The mean Inhomogeneity Coefficient (IC) achieved The mean maximal dose (cGy) in normal tissues by the different planning methods. The mean of the found with each treatment planning method. Abbrevi- Inhomogeneity Coefficient is a measure of dose inhomogene- ations: SVZ- sub ventricular zone, 3D- conformal three ity in the target volumes. The closer the IC to zero, the dimensional, IMRT-Intensity Modulated Radio Therapy, IB- more homogenous the dose is. Integrated Boost, TOMO - Tomo Therapy. Page 4 of 7 (page number not for citation purposes) Inhomogeneity Co ef f icien t (IC) Dose (cGy) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 tralateral globe. Although the mean dose of the IMRT and IB plans to the contralateral globe was lower then the 3D plan, a careful evaluation of the dose to that globe in some individual cases was lowest with the 3D plan compared to all others (data not shown). 3000 The mean maximal dose to the ipsilateral subventricular zone (SVZ) was the highest with the IB treatment plan- ning followed by TOMO, IMRT and 3D (p < 0.05 for all comparisons except IMRT vs. TOMO plans). The contral- ateral SVZ was best spared using the IMRT (p < 0.05 com- D10 D50 D90 pared to every other planning method). No significant 4167 1676 375 IB difference was found between the IB, 3D and TOMO 3931 2152 583 TOMO plans in this respect. IMRT 4762 1791 435 5467 2008 425 3D The mean maximal dose to the brain stem with IMRT plan The mean D10, D5 pla Figure 5 nning method 0, and D90 found with each treatment was significantly lower than with IB and TOMO (p < The mean D10, D50, and D90 found with each treat- 0.008 for both comparisons) but not the 3D plan (p > ment planning method. The mean dose to 10% (D10), 0.6). 50% (D50) and 90% (D90) of the normal brain volume (after the PTV1 and PTV2 volumes were excluded using the soft- The mean of the integral dose (ID) to the brain was signif- ware Boolean Operators) with each treatment planning icantly lower (range 10-17.5%) with the IB plan com- method. pared to all other plans (p < 0.006) (table 1). No significant difference in the ID to the brain was found among the other plans. A different pattern was noted treatment planning method that is superior to all others in when the D10, 50 and 90 were extracted out of the DVHs all aspects compared. of the various plans and compared. The mean dose to 10 percent of the normal brain was consistently lowest with Target volumes' coverage the TOMO plans followed by the IB, IMRT and 3D plans. Sequential boost plans assume 100% dose coverage to the Conversely, the D50 and D90 values were significantly boost volume by the initial part of the treatment. In gli- higher with the TOMO plans compared to both IB and oma patients, were this is not always the case (PTV2 is not IMRT plans and the D50 but not the D90 was significantly a geometrical cone down of PTV1); cold spots might be lower in the IB plan compared to the IMRT and 3D plans. noticed within the boost volume. Non standard target vol- Interestingly, IMRT D10 and D50 were significantly lower umes [4,5] can overcome this problem but these were not than the 3D doses, but no significant difference was found tested prospectively [15]. between these plans regarding D90 (figure 5). More treatment fields may suggest some advantage in the Discussion PTV2 dose coverage (IMRT plan was better then the 3D In this study, we compared the delivery of radiation dose plan). to the target volumes and the adjacent normal structures in high-grade glioma patients by using four treatment Surprisingly, the PTV2 dose coverage with the IB plan was planning methods. This kind of comparison harbors worse then the IMRT plan despite the use of the same numerous biases due to the use of different planning soft- beam arrangement and dose constrains. As previously ware, different optimization algorithms and different reported [4], there is a trade off between the target volume dose prescriptions. The use of mean values to compare coverage and the homogeneity of the dose (which is sup- these plans harbors another potential bias since it fails to ported by our IC results). Prescribing the IB plans to a reflect a better dose profile offered for individual patients lower isodose line improved the coverage but compro- by a specific planning method (e.g. target volume dose mised the homogeneity of the dose in the target volume goals). Furthermore, a lower dose to a normal tissue is an (data not shown). The beam weighing algorithm used to important goal in treatment planning (e.g. SVZ or normal produce a plan sum in a sequential boost (IMRT) com- brain), but does not necessarily give an advantage if the pared to an integrated boost (IB) gives another explana- tolerance of that tissue is not met (e.g. the optic chiasm). tion to the different PTV2 coverage [4]. A qualitative comparison of the various plans is summa- The TOMO plan did not result in an inferior PTV2 cover- rized in table 2. According to our results, there is no single age and achieved the best IC as well since it uses infinite number of fields by definition. Page 5 of 7 (page number not for citation purposes) D10 D50 D90 Dose (cGy) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 Table 1: The mean Integral Dose (ID) to the brain with each treatment planning method. % Difference (p value) Integral Dose ± SD (Gy × cm × 1000) Plan 3D IMRT IB 22.8 ± 7.2 3D NA 21.1 ± 3.7 IMRT 7.5 (0.2) NA 18.8 ± 3.1 IB 17.5 (0.006)* 10.9 (8.7E-09)* NA 21 ± 2.3 TOMO 7.8 (0.3) 0.3 (0.6) 10.6 (0.003)* * Statistically significant Normal tissue sparing tumors, associated with the tumor ability to resist radia- The TOMO plans were able to spare best small organs tion treatment and recur[19-21]. Higher radiosensitivity which usually lie close to the target volumes and were of the brain stem compared to other parts of the brain is highly prioritized (e.g. the chiasm and ipsilateral globe) not reported and its tolerance to fractionated radiotherapy but the advantage was relatively small compared to the appears to be a function of the volume receiving high dose IMRT and IB plans. The ability of the TOMO to spare these rather than the maximum point dose [22]. organs in patients with tumors located distant from them is questionable since blocking both the entering and exit Integral Dose (ID) is the value usually used to compare doses to a distant volume without compromising the tar- the dose received by healthy tissues outside the target vol- get volume coverage, is unlikely with the helical beam umes [12-14]. We added an analysis of three points in the arrangement. DVH curve (D10, D50 and D90) of the normal brain to evaluate the validity of the ID as a tool for this kind of No benefit and even a potential disadvantage was found comparisons since different DVH curves may result in the with the IB and TOMO plans for structures with larger vol- same ID. The TOMO plans, with the high conformality umes that often overlap with the target volumes, such as and a rapid fall off of the dose around the target volumes the SVZ and the brain stem (due to a higher dose per frac- had the lowest D10 to the normal brain (only a small vol- tion translated into higher total BED in these areas). We ume around the target volume received a high dose). At chose not to give dose constrains to these structures and the same time, larger volumes of normal brain received compared their dose distribution after the plans were gen- irradiation at all (high D50 and D90). This significant erated. The SVZs (believed to harbor the normal brain dose of irradiation received by the normal brain was not progenitor cells [10]) viability is correlated with late radi- demonstrated when the ID values were compared. Along ation toxicity [11,16,17] on one hand justifying an with the expected longer survival of HGG patients in the attempt to lower their dose [11,18] but are also suspected future these differences may translate to toxicity. Lower ID as the source of cancerous stem cells in primary brain values with IMRT compared to 3D plans were used to jus- tify its use for HGG patients [14]. Our mean ID failed to Table 2: A qualitative comparison of the four treatment planning indicate such a difference but it is in line with the lower methods. D10 and D50 for the IMRT plans we found. We suggest caution in the use and interpretation of ID for treatment 3D IMRT IB TOMO plans comparisons due to its failure to predict differences in DVH curves which might have significant implications Target Volumes in the future. PTV1 coverage ++ + + PTV2 coverage -+ - + Inhomogeneity Coefficient -+ Conclusion Normal Tissues Sparing Our data suggest a distinctive approach in the use of new Optic Chiasm -+ treatment planning tools. Further investigation is indi- Ipsilateral Globe -+ + cated to better choose the correct tool for each patient. Contralateral Globe +- Larger series might suggest a decision algorithm according Ipsilateral SVZ +- to the patient's tumor size, location and prognosis. Long Contralateral SVZ + Brainstem +- - follow up periods with HGG patients are becoming a Normal Brain ID + widespread phenomenon, and may allow better under- Normal Brain D10 -+ standing of the effect of the different DVH curves of the Normal Brain D50 -+ - normal brain and the SVZ. A close follow up on patient's Normal Brain D90 - toxicity profiles and correlations with a specific treatment planning method are indicated. The use of imprecise and + Significant advantage - Significant disadvantage Page 6 of 7 (page number not for citation purposes) Radiation Oncology 2009, 4:45 http://www.ro-journal.com/content/4/1/45 10. Doetsch F, Caille I, Lim DA, Garcia-Verdugo JM, Alvarez-Buylla A: insensitive tools like ID to compare potential toxicity due Subventricular zone astrocytes are neural stem cells in the to large irradiated volumes should be discouraged and adult mammalian brain. Cell 1999, 97(6):703-16. better tools should be developed. 11. Barani IJ, Cuttino LW, Benedict SH, Todor D, Bump E, Wu Y, Chung T, Broaddus W, Lin P: Neural stem cell-preserving external- beam radiotherapy of central nervous system malignancies. Competing interests IJROBP 2007, 68(4):978-85. 12. Shi C, Peñagarìcano J, Papanikolaou N: Comparison of IMRT The authors declare that they have no competing interests. treatment plans between linac and helical tomotherapy based on integral dose and inhomogeneity index. Med Dosim Authors' contributions 2008, 33(3):215-21. 13. Aoyama H, Westerly DC, MacKie TR, Olivera G, Bentzen S, Patel R, LZ carried out the contouring, and participated in the Jaradat H, Tome W, Ritter M, Mehta M: Integral radiation dose to study design, coordination, treatment planning and writ- normal structures with conformal external beam radiation. ing of the manuscript. BS participated in contouring and IJROBP 2006, 64(3):962-7. 14. Hermanto U, Frija EK, Lii MJ, Chang E, Mhajan A, Woo S: Intensity- helped revising the draft manuscript. HN carried out the modulated radiotherapy (IMRT) and conventional three- treatment planning. JO carried out MRI fusions and par- dimensional conformal radiotherapy for high-grade gliomas: Does IMRT increase the integral dose to normal brain? ticipated in treatment planning. BA participated in treat- IJROBP 2007, 67(4):1135-44. ment planning. US participated in the statistical analysis 15. Fiveash JB, Nordal JM, Markert RS, Ahmed RS, Nabors LB: High of the results and helped revising the draft manuscript. Grade Gliomas. In Clinical Radiation Oncology 2nd edition. Edited by: Gunderson L, Tepper J. Churchill Livingstone, Philadelphia, (PA); RWM participated in treatment planning. DC participated 2006:515-537. in the data analysis and helped revising the draft manu- 16. Barani IJ, Benedict SH, Lin P: Neural stem cells: Implications for the conventional radiotherapy of central nervous system script. KC conceived of the study, and participated in its malignancies. IJROBP 2007, 68(2):324-33. design and coordination and helped to draft the manu- 17. Mizumatsu S, Monje ML, Morhardt DR, Rola R, Palmer T, Fike J: script. All authors read and approved the final manu- Extreme sensitivity of adult neurogenesis to low doses of X- irradiation. Cancer Res 2003, 63(14):4021-7. script. 18. Gutieìrrez AN, Westerly DC, Tomeì WA, Jaradat H, Mackie T, Bentzen S, Khuntia D, Mehta M: Whole brain radiotherapy with References hippocampal avoidance and simultaneously integrated brain metastases boost: A planning study. IJROBP 2007, 69(2):589-97. 1. Souhami L, Seiferheld W, Brachman D, Podgorsak P, Werner-Wasik 19. Quiñones-Hinojosa A, Chaichana K: The human subventricular M, Lustig R, Schultz C, Sause W, Okunieff P, Buckner J: Randomized zone: A source of new cells and a potential source of brain comparison of stereotactic radiosurgery followed by conven- tumors. Exp Neurol 2007, 205(2):313-24. tional radiotherapy with carmustine to conventional radio- 20. Lim DA, Cha S, Mayo MC, Chen MH, Keles E, VandenBerg S, Berger therapy with carmustine for patients with glioblastoma M: Relationship of glioblastoma multiforme to neural stem multiforme: Report of radiation therapy oncology group 93- cell regions predicts invasive and multifocal tumor pheno- 05 protocol. IJROBP 2004, 60(3):853-60. type. Neuro Oncol 2007, 9(4):424-9. 2. Cardinale R, Won M, Choucair A, Gillin M, Chakravarti A, Schultz C, 21. Sanai N, Alvarez-Buylla A, Berger MS: Mechanisms of disease: Souhami L, Chen A, Pham H, Mehta M: A phase II trial of acceler- Neural stem cells and the origin of gliomas. N Engl J Med 2005, ated radiotherapy using weekly stereotactic conformal 353(8):811-22. boost for supratentorial glioblastoma multiforme: RTOG 22. Debus J, Hug EB, Liebsch NJ, O'farrel D, Finkelstein D, Efird J, 0023. IJROBP 2006, 65(5):1422-8. Munzenrider J: Brainstem tolerance to conformal radiother- 3. Hartford AC, Palisca MG, Eichler TJ, Beyer D, Devineni V, Ibbott G, apy of skull base tumors. IJROBP 1997, 39(5):967-75. Kavanagh B, Kent J, Rosenthal S, Schultz C, Tripuraneni P, Gaspar L: American society for therapeutic radiology and oncology (ASTRO) and american college of radiology (ACR) practice guidelines for intensity-modulated radiation therapy (IMRT). IJROBP 2009, 73(1):9-14. 4. Thilmann C, Zabel A, Grosser KH, Hoess A, Wannenmacher M, Debus J: Intensity-modulated radiotherapy with an integrated boost to the macroscopic tumor volume in the treatment of high-grade gliomas. Int J Cancer 2001, 96(6):341-9. 5. Suzuki M, Nakamatsu K, Kanamori S, Okumra M, Uchiyama T, Akai F, Nishimura Y: Feasibility study of the simultaneous integrated boost (SIB) method for malignant gliomas using intensity- modulated radiotherapy (IMRT). Jpn J Clin Oncol 2003, 33(6):271-7. 6. Suzuki M, Nakamatsu K, Kanamori S, Okajima K, Okumura M, Publish with Bio Med Central and every Nishimura Y: Comparison of outcomes between overlapping scientist can read your work free of charge structure-based and non-overlapping structure-based opti- mization for simultaneous integrated boost IMRT for malig- "BioMed Central will be the most significant development for nant gliomas. Int J Clin Oncol 2004, 9(6):491-7. disseminating the results of biomedical researc h in our lifetime." 7. Iuchi T, Hatano K, Narita Y, Kodama T, Yamaki T, Osato K: Hypof- Sir Paul Nurse, Cancer Research UK ractionated high-dose irradiation for the treatment of malig- nant astrocytomas using simultaneous integrated boost Your research papers will be: technique by IMRT. IJROBP 2006, 64(5):1317-24. available free of charge to the entire biomedical community 8. Chan JL, Lee SW, Fraass BA, Normolle DP, Greenberg HS, Junck LR, Gebarski SS, Sandler HM: Survival and failure patterns of high- peer reviewed and published immediately upon acceptance grade gliomas after three-dimensional conformal radiother- cited in PubMed and archived on PubMed Central apy. JCO 2002, 20(6):1635-42. 9. ICRU report 50, prescribing, recording, and reporting pho- yours — you keep the copyright ton beam therapy. International commission on Radiatrion BioMedcentral Submit your manuscript here: units and Measurments. Bethesda, MD 1993. http://www.biomedcentral.com/info/publishing_adv.asp Page 7 of 7 (page number not for citation purposes)

Journal

Radiation OncologySpringer Journals

Published: Oct 21, 2009

References