|Ahead of print publication
Hippocampal sparing for brain tumor radiotherapy: A retrospective study comparing intensity-modulated radiotherapy and volumetric-modulated arc therapy
Gulsen Pinar Soydemir, Nazli Bilici, Elif Eda Tiken, Ayben Yentek Balkanay, Ali Firat Sisman, Didem Karacetin
Department of Radiation Oncology, Health Sciences University, Bakirköy Dr. Sadi Konuk Education and Research Hospital, Istanbul, Turkey
|Date of Submission||14-Jan-2019|
|Date of Decision||23-May-2019|
|Date of Acceptance||14-Oct-2019|
|Date of Web Publication||16-May-2020|
Gulsen Pinar Soydemir,
Department of Radiation Oncology, Health Sciences University, Bakirköy Dr. Sadi Konuk Education and Research Hospital, Bakirköy, Istanbul
Source of Support: None, Conflict of Interest: None
Context: Radiotherapy may have side effects on the brain, such as radiation necrosis, cognitive impairment, and a high chance of tumor recurrence, which has been considered the most common cause of treatment failure.
Aims: Using intensity-modulated radiotherapy (IMRT) and volumetric-modulated arc therapy (VMAT) techniques, we aimed to test the potential outcome of sparing the contralateral hippocampus (CLH) in radiotherapy for brain tumors by comparing dosimetric parameters.
Settings and Design: A prospective clinical comparative study.
Subjects and Methods: Using IMRT and VMAT, sparing CLH in radiotherapy of brain tumors was tested in ten patients, and various dosimetric parameters were compared. The treatment plans were accepted only if they met the set of planning objectives defined in the protocol.
Results: The dose delivered to 95% of the CLH volume (CLH D95), and the mean (CLH Dmean) and max (CLH Dmax) doses were found to be significantly highest in the standard IMRT (P = 0.002, <0.001, and < 0.001, respectively). The lowest CLH D95, CLH Dmean and CLH Dmax for the hippocampus were detected in sparing VMAT planning than in the other plans (P < 0.05). None of the post hoc comparisons for CLH D95 was different among any of the plans, whereas the mean dose to CLH was statistically different among all paired comparisons (P < 0.008). The maximum dose to CLH was also statistically different among all paired plans (P < 0.008), except the dose difference between standard VMAT and IMRT plans.
Conclusions: Although VMAT planning is troublesome and time-consuming, the advantage of sparing the hippocampus is beneficial, preserving the hippocampus and cognitive functions during radiotherapy.
Keywords: Brain tumor, hippocampus, intensity-modulated radiotherapy, radiotherapy, volumetric-modulated arc therapy
|How to cite this URL:|
Soydemir GP, Bilici N, Tiken EE, Balkanay AY, Sisman AF, Karacetin D. Hippocampal sparing for brain tumor radiotherapy: A retrospective study comparing intensity-modulated radiotherapy and volumetric-modulated arc therapy. J Can Res Ther [Epub ahead of print] [cited 2020 Aug 10]. Available from: http://www.cancerjournal.net/preprintarticle.asp?id=284488
| > Introduction|| |
The standard treatment approach to brain tumors is maximal surgical resection, followed by adjuvant radiotherapy. Combining radiotherapy with chemotherapy (concomitant and adjuvant temozolomide) has shown a higher survival benefit and lower additional toxicity in patients with high-grade brain tumors. However, radiotherapy may have side effects on the brain, including radiation necrosis, cognitive impairment, and a high chance of tumor recurrence, which has been considered the most common cause of treatment failure.,,, Moreover, radiation-induced memory impairment is related to the dose to the hippocampus. Thus, different strategies of dose prescription and modern planning techniques have been tested to achieve better tumor control and better sparing of the organs at risk (OARs)., Two of these radiotherapy techniques are intensity-modulated radiotherapy (IMRT) and volumetric-modulated arc therapy (VMAT), allowing the delivery of highly conformal dose distributions. The idea of these techniques has arisen to selectively spare the hippocampus during brain radiotherapy.
IMRT technique delivers variable intensity radiation with multiple radiation beams; thus, target volume conformity and sparing of normal tissues and OARs have been greatly improved. Furthermore, IMRT is able to produce inhomogeneous dose distributions, resulting in the simultaneous delivery of different doses per fraction to separate areas within the target volume. On the other hand, VMAT allows simultaneous variation of three parameters during the delivery of radiation therapy, i.e., gantry rotation speed, treatment aperture shape through movement of multileaf collimator (MLC) leaves, and dose rate. This causes further improvement in target volume conformity and OARs sparing. In this study, we aimed to test the potential outcomes of sparing the contralateral hippocampus (CLH) in radiotherapy to brain tumors in ten patients using VMAT and IMRT techniques and set various dosimetric objectives to compare dosimetric parameters between VMAT and IMRT.
| > Subjects and Methods|| |
The present study was conducted in accordance with the 2017/28 Ethics Committee Approval for VMAT/IMRT and the selection criteria. Patient consent was not obligatory as stated by the Ethics Committee of our institute.
Selection and description of the participants
In total, ten patients diagnosed with pathological grade IV brain tumor were prospectively enrolled in this prospective comparative clinical study. The median age of the patients was 54.5 years (range, 34–61 years). All patients underwent preoperative magnetic resonance imaging (MRI) or computed tomography (CT).
Datasets of ten patients who had received either IMRT or VMAT for brain tumor at our institute were included in this planning study. For all patients, the dose prescribed to the planning target volume (PTV) was 46 Gy followed by a boost of 60 Gy in 30 fractions. Patients were positioned in the supine position and immobilized with a thermoplastic mask. Normal structures such as the eyes, lenses, optic nerves, chiasm, and brain stem were contoured and designated as OARs. Expanded contours were created with safety margins of 3 mm around the brain stem and chiasm and 5 mm around the optic nerve. The hippocampus was contoured on the T1 MRI sequence, as described by Hofmaier et al.
Volumetric-modulated arc therapy planning
The VMAT plans were optimized in the research treatment planning system (TPS) of Monaco® 5.1 (Elekta AB Publ, Stockholm, Sweden), which relies on the X-ray voxel Monte Carlo (XVMC) algorithm for dose calculation. All VMAT plans were generated for a 6 MV Elekta Synergy linear accelerator (LINAC) equipped with an agility MLC. Three fields were used with a gantry angle beginning from 180°. The first field was given with both clockwise and counter clockwise gantry angle of 160° on the basis of tumor location and as the eye balls and lenses of the patients did not take the initial dose. Moreover, the collimator angles were given in between 50° and 70° on the basis of tumor location. The second field was given in between 40° and 50° as all the PTVs were caught, the lens and eyes were excluded from the exit dose, and the collimator angle was given in between 50° and 70°. The third field was given as half arc (180°) with a table angle of 55° and 305° on the basis of tumor location to protect the healthy tissues of the patients.
For Monte Carlo-based VMAT planning for the patients, two arcs for each field, 180 control points for each arc, statistic as one, and grid size as 0.3 were chosen [Figure 1].
|Figure 1: Volumetric-modulated arc therapy planning design accompanied with table angles|
Click here to view
Intensity-modulated radiotherapy planning
For generating IMRT plans for the patients, ELECTA TPS of Elekta Monaco 5.1, which relies on the XVMC algorithm, was used for dose calculation. All IMRT plans were generated in a 6 MV Elekta Synergy LINAC, equipped with an Agility MLC. IMRT planning was performed with six fields using MLC delivery method in dynamic mode. The initial gantry angles of these six fields were chosen as 260°, 230°, 170°, 140°, 110°, and 70°. Collimator or table angle was not used for any of these fields [Figure 2].
Evaluation of treatment plans
Monte Carlo algorithm was used for both plans. After contour radiation was entered by the oncologist, sonar VMAT and IMRT plans were performed by medical physicist. VMAT plans were performed twice to spare and not to spare (standard) the hippocampus contour in optimized condition. For both VMAT plans, same angles and cost functions were used. The maximum control point for each arc used in the VMAT plan was chosen as 180.
The IMRT plans were performed twice to spare and not to spare (standard) the hippocampus contour in optimized condition. For both IMRT plans, same angles and cost functions were used. The planning of these IMRT plans was adjusted at 15 segments of each field.
For both techniques, hippocampus sparing was targeted as PTV covering 100% of the volume at 95% of target volume dose. In plans where 95% of PTV was covered by 95% of target volume dose, the conformity index enhancer, planned template additional field, or cost functions were not used.
All dosimetric values were reported in descriptive statistics as mean ± standard deviation and median (minimum–maximum) values. Dosimetric parameters were tested for normal distribution using Kolmogorov–Smirnov test. Because two different plans were generated in the CT image set of every patient, the data were considered matched pair and paired tests were used to compare the two plans. For two nonnormally distributed dependent variables, a corresponding nonparametric test Wilcoxon test was used, and for more than two nonnormally distributed dependent variables, Friedman test was used. All statistical analyses were performed with 5% level of significance, and P < 0.05 was considered statistically significant. Analysis was performed using MedCalc Statistical Software version 12.7.7 program (MedCalc Software bvba, Ostend, Belgium; http://www.medcalc.org; 2013).
| > Results|| |
The VMAT and IMRT plans met the planning objectives for 10 patients, as mentioned in the protocol. The characteristics and total dosimetric results of the patients are given in [Table 1]. The mean age of the patients who underwent total (n = 9) or subtotal (n = 1) brain surgery was 52.1 ± 8.9 years, and the tumors were located in the temporal (n = 3), parietal (n = 5), and frontal (n = 2) regions of the brain. The mean clinical target volume (CTV) was 172.47 ± 108.35 cm3 (41.95–320 cm3), and the mean CLH volume was 4.16 ± 0.56 cm3 (3.18–4.8 cm3). The mean total hippocampus volume was calculated as 8.39 ± 1.85 cm3 (5.6–11.57 cm3).
The dosimetric data were obtained from dose volume histograms generated using ELECTA TPS, as shown in [Table 2]. When the dosimetric parameters of PTV were compared, we found that the coverage of the treatment target dose at PTV 46 was statistically higher in standard and sparing VMAT plans than in standard IMRT and sparing IMRT plans (P = 0.009), with the whole target receiving at least 95% of the prescription dose (D95). Post hoc comparisons of the target doses at PTV 46 showed a significant difference only between the standard VMAT and sparing IMRT plans [Table 3], P = 0.007]. The minimum dose at PTV 46 was also significantly different between the standard VMAT and IMRT plans (P = 0.005). However, the dose to receive at least 95% of 60 Gy was comparable between VMAT and IMRT plans [Table 2]. The maximum dose to reach PTV 46 was higher and the minimum dose was lower in IMRT plans than in VMAT plans (P = 0.048 and P = 0.017, respectively), whereas the maximum and minimum doses at PTV 60 in all plans were statistically similar. The volumes at PTV 46 and PTV 60 were essentially comparable among all plans, i.e., 412.77 ± 74.03 cm3( 343.45–593.22 cm3) for PTV 46 and 294.89 ± 81.2 cm3( 183.16–431.38 cm3) for PTV 60. Post hoc comparison of the mean dose to give PTV 60 was significantly different among the standard VMAT and sparing IMRT plans (P = 0.007).
|Table 2: Dosimetric comparison of standard and sparing volumetric modulated arc therapy and intensity-modulated radiotherapy for planning target volume 46 and 60 plans|
Click here to view
|Table 3: Post hoc paired comparisons of standard and sparing volumetric modulated arc therapy and standard and sparing intensity-modulated radiotherapy for planning target volume 46 and 60 plans|
Click here to view
CLH D95, CLH Dmean, and CLH Dmax were found to be significantly highest in the standard IMRT plans than in the other plans (P = 0.002, <0.001, and <0.001, respectively). On the other hand, CLH D95, CLH Dmean, and CLH Dmax were found to be significantly lowest in the sparing VMAT plans than in the other plans (P < 0.05) [Table 2]. None of the post hoc comparisons for CLH D95 was significant among any plans, whereas CLH Dmean was statistically different among all plans (P < 0.008). Moreover, the CLH Dmax also varied among all plans (P < 0.008), except a nonsignificant paired comparison between the standard VMAT and IMRT plans.
All VMAT and IMRT plans were able to meet the constraints placed on OAR and PTV. Considering PTV 60, the doses to the left eye, right lens, and left optic nerve were found to increase in both standard and sparing IMRT plans (P = 0.009, 0.044, and < 0.001, respectively), but the dose to the brain stem was increased only in the sparing VMAT plan (P = 0.041). Post hoc pairwise comparisons between the plans showed that there was a significant difference between the sparing VMAT and standard IMRT plans to give PTV 60 for the left eye (P = 0.007) and between the standard VMAT and sparing IMRT plans to give PTV 60 for the left optic nerve (P = 0.005). Other doses to the right eye, left lens, right optic nerve, and optic chiasm were statistically comparable for all plans.
All plans were optimized to keep the maximum dose within the target value <110% of the prescription dose (Dmax <110%), and all plans were able to meet this objective. The doses to give 2% and 5% of PTV were significantly higher in sparing IMRT plan than in the other three plans (P = 0.014 and P = 0.024). However, post hoc comparisons for these doses showed no significant difference for pairwise plans. 98% of PTV was similar among the plans, but VMAT did not give 107% of PTV. A dramatic difference was noted among the four plans in terms of the monitor units (P = 0.001), and a significant difference was noted in post hoc comparisons of sparing VMAT and standard IMRT plans (P = 0.005). However, the conformity and homogeneity indexes were equivalent for all the VMAT and IMRT plans.
| > Discussion|| |
To investigate the feasibility of sparing CLH in radiotherapy for brain tumors, the present study compared VMAT and IMRT plans in terms of dosimetric parameters and revealed that sparing CLH is dosimetrically feasible for both the plans. Because the treatment target doses at PTV 46 were statistically higher in the standard and sparing VMAT plans than in the standard and sparing IMRT plans, with the whole target receiving at least 95% of the prescription dose but the dose to receive at least 95% of 60 Gy was similar between the two plans, it is suggested that compared with VMAT plan, IMRT plan achieved equal or better PTV coverage. However, the findings of CLH D95, CLH Dmean, and Dmax evidently proved that sparing the hippocampus by VMAT plan achieved better dose values to preserve neurocognitive functions. In terms of OAR sparing, VMAT was successful for PTV 60, especially in the left eye and left optic nerve, while higher monitoring units were found in VMAT than in sparing IMRT plan. The other investigated parameters including sparing other organs, the homogeneity index, and the conformity index could be kept stable or even improved in both the plans.
These results are consistent with those of the study by Rapole et al., who evaluated the dosimetric parameters of simultaneous integrated boost (SIB) in the treatment of malignant gliomas and compared the SIB plans of VMAT and IMRT in 28 patients. They attempted to identify influencing factors for the hippocampal dose exposure and found that GPTV (high-risk boost volume created by giving a 0.5-cm margin to the planning target volume of gross tumor volume [GTV]) coverage was statistically better in VMAT plan; however, the difference had not been implicated as clinically meaningful. They also noted that the coverage of CPTV (planning target volume of CTV given a 0.5-cm margin) was better in IMRT plan, but there was no difference in the homogeneity indexes of GPTV and CPTV annulus between the plans. Contrary to our study, they gave a higher conformity index for GPTV in the IMRT plan and found that monitor units were significantly less in the VMAT plan. They concluded that compared with the VMAT plan, the IMRT plans had better boost conformity and lower ipsilateral optic nerve and brainstem maximum doses. VMAT had better coverage, better overall PTV conformity, lower normal brain mean dose and lower monitor units. Furthermore, several studies have compared the different plans. Canyilmaz et al. compared the standard IMRT to IMRT and VMAT plans with hippocampal sparing in 20 patients. Marsh et al. compared standard IMRT to IMRT plans with sparing the contralateral neural stem cell compartment, hippocampus, and limbic circuit in five patients. The planning strategy of these studies is the same as the concept used in our study. Both studies used a dose prescription of 46 Gy in 23 fractions followed by a sequential boost of 14 Gy in seven fractions (Radiation Therapy Oncology Group consensus) and concluded that a substantial reduction of the dose to CLH is feasible with the used techniques without compromising other treatment parameters.
Other studies suggest that it is safe to reduce the applied margins in IMRT for glioblastoma. Ali et al. reported the effect of reduced PTV margins on the hippocampal dose, compared the standard margin of 2 cm (GTV to CTV) with a reduced margin of 8 mm, and found a significant reduction of the bilateral hippocampal dose when the reduced margin concept was applied. Hofmaier et al. also showed a significant correlation between the size of PTV and the dose to CLH by VMAT plan. The limitation of this study was that the extent of VMAT or IMRT better in sparing the hippocampus was not indicated compared with the standard plans from a purely technological and therapeutic perspective because the planning objectives were not the same. However, our study may provide insight into the mean, maximum, and minimum dose exposure to achieve clinical preservation of the hippocampus when the standard VMAT and IMRT planning strategies are replaced by a hippocampal sparing approach.
Although there have been few studies investigating the feasibility of hippocampal sparing in radiotherapy for brain tumors, many researches have been conducted concerning the hippocampal sparing during whole-brain radiotherapy (WBRT) for patients with brain metastases. Reported advantages of sparing both hippocampi during WBRT suggest that the approach is safe and favorable in terms of neurological outcome.,, For hippocampal sparing WBRT, the influence of patient positioning on the hippocampal dose exposure has been also investigated and an inclined head angle was suggested to be beneficial. In our study, we did not investigate whether an inclined head angle is able to improve the hippocampal dose further in radiotherapy for brain tumors because all patients were positioned in the standard way. In WBRT setting for patients with brain metastases, sparing the hippocampus is not expected to affect the therapeutic ratio because brain metastases rarely occur inside or close to the hippocampus. In contrast, sparing the hippocampus in radiotherapy is more controversial owing to the potential involvement of the subventricular zone (SVZ) in tumor genesis. Higher doses to the ipsilateral SVZ have been shown to correlate with improved progression-free survival. Thus, only the CLH was spared in different plans investigated in this study.
Our study has some limitations owing to an unexpected result of higher monitor units in the VMAT plans than in the IMRT plans. A possible explanation may be the type of planning strategy used and the small number of patients involved (n = 10), which limits the statistical power of the test and the P value for the comparison between two plans. Compared with our study, Canyilmaz et al. found higher monitor unit values of both IMRT planning and VMAT planning. They used seven fields with standard angles for PTV 46 in IMRT planning and five fields identified according to tumor localization for PTV 60, whereas we used three fields in VMAT planning and six fields in IMRT planning for PTV 46 and 60. This explains the reasonable monitor unit values in our study; hence, these contradicting results should be interpreted with caution.
| > Conclusions|| |
This study aimed to show the preservation of the hippocampus in different treatment techniques for patients who receive radiotherapy for brain tumors. In deciding a treatment plan for the patients, the dose received by CLH should be considered for the homogeneity of the therapeutic plan. Although VMAT planning is troublesome and time-consuming, the advantage of sparing the hippocampus is beneficial, preserving the hippocampus and cognitive functions during radiotherapy. Possible benefits for patients regarding the neurological toxicity of radiotherapy for brain tumors still need to be evaluated in future clinical trials.
The authors would like to thank Enago (www.enago.com) for the English language review.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| > References|| |
Diaz AZ, Choi M. Radiation-associated toxicities in the treatment of high-grade gliomas. Semin Oncol 2014;41:532-40.
Klein M, Heimans JJ, Aaronson NK, van der Ploeg HM, Grit J, Muller M, et al.
Effect of radiotherapy and other treatment-related factors on mid-term to long-term cognitive sequelae in low-grade gliomas: A comparative study. Lancet 2002;360:1361-8.
Johannesen TB, Lien HH, Hole KH, Lote K. Radiological and clinical assessment of long-term brain tumour survivors after radiotherapy. Radiother Oncol 2003;69:169-76.
Douw L, Klein M, Fagel SS, van den Heuvel J, Taphoorn MJ, Aaronson NK, et al.
Cognitive and radiological effects of radiotherapy in patients with low-grade glioma: Long-term follow-up. Lancet Neurol 2009;8:810-8.
Tsai PF, Yang CC, Chuang CC, Huang TY, Wu YM, Pai PC, et al.
Hippocampal dosimetry correlates with the change in neurocognitive function after hippocampal sparing during whole brain radiotherapy: A prospective study. Radiat Oncol 2015;10:253.
Cha J, Suh CO, Park K, Chang JH, Lee KS, Kim SH, et al.
Feasibility and outcomes of hypofractionated simultaneous integrated boost-intensity modulated radiotherapy for malignant gliomas: A preliminary report. Yonsei Med J 2014;55:70-7.
Nakamatsu K, Suzuki M, Nishimura Y, Kanamori S, Koike R, Shibata T, et al.
Treatment outcomes and dose-volume histogram analysis of simultaneous integrated boost method for malignant gliomas using intensity-modulated radiotherapy. Int J Clin Oncol 2008;13:48-53.
Kazda T, Jancalek R, Pospisil P, Sevela O, Prochazka T, Vrzal M, et al.
Why and how to spare the hippocampus during brain radiotherapy: The developing role of hippocampal avoidance in cranial radiotherapy. Radiat Oncol 2014;9:139.
Veldeman L, Madani I, Hulstaert F, De Meerleer G, Mareel M, De Neve W. Evidence behind use of intensity-modulated radiotherapy: A systematic review of comparative clinical studies. Lancet Oncol 2008;9:367-75.
Teoh M, Clark CH, Wood K, Whitaker S, Nisbet A. Volumetric modulated arc therapy: A review of current literature and clinical use in practice. Br J Radiol 2011;84:967-96.
Hofmaier J, Kantz S, Söhn M, Dohm OS, Bächle S, Alber M, et al.
Hippocampal sparing radiotherapy for glioblastoma patients: A planning study using volumetric modulated arc therapy. Radiat Oncol 2016;11:118.
Fippel M. Fast Monte Carlo dose calculation for photon beams based on the VMC electron algorithm. Med Phys 1999;26:1466-75.
Rapole PS, Karunanithi G, Kandasamy S, Prabhu S, Kumar R, Vivekanandam S. Dosimetric comparison and feasibility of simultaneous integrated boost (SIB) in treatment of malignant gliomas using intensity modulated radiotherapy (IMRT) or volumetric modulated arc therapy (VMAT) Asian Pac J Cancer Prev 2018;19:2499-506.
Canyilmaz E, Uslu GD, Colak F, Hazeral B, Haciislamoglu E, Zengin AY, et al.
Comparison of dose distributions hippocampus in high grade gliomas irradiation with linac-based imrt and volumetric arc therapy: A dosimetric study. Springerplus 2015;4:114.
Marsh JC, Godbole R, Diaz AZ, Gielda BT, Turian JV. Sparing of the hippocampus, limbic circuit and neural stem cell compartment during partial brain radiotherapy for glioma: A dosimetric feasibility study. J Med Imaging Radiat Oncol 2011;55:442-9.
Gebhardt BJ, Dobelbower MC, Ennis WH, Bag AK, Markert JM, Fiveash JB. Patterns of failure for glioblastoma multiforme following limited-margin radiation and concurrent temozolomide. Radiat Oncol 2014;9:130.
Ali AN, Ogunleye T, Hardy CW, Shu HK, Curran WJ, Crocker IR. Improved hippocampal dose with reduced margin radiotherapy for glioblastoma multiforme. Radiat Oncol 2014;9:20.
Oehlke O, Wucherpfennig D, Fels F, Frings L, Egger K, Weyerbrock A, et al.
Whole brain irradiation with hippocampal sparing and dose escalation on multiple brain metastases: Local tumour control and survival. Strahlenther Onkol 2015;191:461-9.
Awad R, Fogarty G, Hong A, Kelly P, Ng D, Santos D, et al.
Hippocampal avoidance with volumetric modulated arc therapy in melanoma brain metastases – The first Australian experience. Radiat Oncol 2013;8:62.
Mahadevan A, Sampson C, LaRosa S, Floyd SR, Wong ET, Uhlmann EJ, et al.
Dosimetric analysis of the alopecia preventing effect of hippocampus sparing whole brain radiation therapy. Radiat Oncol 2015;10:245.
Siglin J, Champ CE, Vakhnenko Y, Witek ME, Peng C, Zaorsky NG, et al.
Optimizing patient positioning for intensity modulated radiation therapy in hippocampal-sparing whole brain radiation therapy. Pract Radiat Oncol 2014;4:378-83.
Hong AM, Suo C, Valenzuela M, Haydu LE, Jacobsen KD, Reisse CH, et al.
Low incidence of melanoma brain metastasis in the hippocampus. Radiother Oncol 2014;111:59-62.
Adeberg S, Bostel T, König L, Welzel T, Debus J, Combs SE. A comparison of long-term survivors and short-term survivors with glioblastoma, subventricular zone involvement: A predictive factor for survival? Radiat Oncol 2014;9:95.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3]