Journal of Medical Physics
: 2017  |  Volume : 42  |  Issue : 5  |  Page : 59--65

AMPI Best Paper


Correspondence Address:

How to cite this article:
. AMPI Best Paper.J Med Phys 2017;42:59-65

How to cite this URL:
. AMPI Best Paper. J Med Phys [serial online] 2017 [cited 2023 Mar 28 ];42:59-65
Available from:

Full Text

 BP-1: Applicator Commissioning for Image Based Brachytherapy

N. Siji, K. Joshi, S. Dheera1, U. Mahantshetty1, S. Chopra, A. Laishram, Swati Kumari, D. D. Deshpande1, S. V. Jamema

Advanced Centre for Treatment, Research and Education in Cancer (ACTREC), TMC, Navi Mumbai, 1Tata Memorial Hospital, TMC, Parel, Mumbai, Maharashtra, India. E-mail: [email protected]

Aim: In 3D image based brachytherapy, it is highly recommended to commission brachytherapy applicators to minimize the uncertainties during applicator reconstruction. The commercially available CT/MR compatible applicators are generally made up of two materials viz polymer and titanium. It has been reported that titanium applicators cause susceptibility artifacts especially in 3T MRI. The manufacturer specification of the magnetic field strength to be used with these applicators are 0.3-0.5T, however we have been using these applicators in 1.5T with proper commissioning. We are in the process of starting MR image based BT using 3T MRI. Hence, we carried out some QA tests using 3T MRI using both plastic and titanium applicators. The aim of this study is to report the results of the commissioning of brachytherapy applicators for MR image based Brachytherapy.

Materials and Methods: Two commercial brachytherapy applicators were evaluated viz Vienna applicator made of polymer with Ti needles (Elekta) and Fletcher- suit (Varian) applicator made of titanium. A CT/MR compatible phantom of size 30 cm x 20 cm x 12 cm was fabricated such that the applicator can be positioned in a reproducible geometry. The phantom was filled with the gel (3% Agarose + 1% Copper Sulphate) to closely resemble the tissue density in MRI. CT and MRI (1.5T and 3.0T) images of the phantom with the applicator were obtained followed by rigid registration and image analysis. CT slice thickness was 0.625 mm (GE DISCOVERY). The MR images obtained for 1.5T were as per the institutional protocol established already; however a new protocol for 3T was attempted. The MR sequences for 3T were as follows: T1 weighted FRFSE, T2 weighted FRFSE, T2 weighted SSFSE and T2 weighted FSE, all with a band width of 125 and of equal matrix of 256 x 256 pixels. The MR slice thickness was 3mm with 0.5 mm spacing. Various combination of TE (echo time), TR (repetition time), and Echo Train length were also used. The images were imported and co-registered with the CT series in Eclipse Treatment Planning System (Version 13.4) using rigid registration. Mutual Information with limited ROI method was used for the rigid registration. Manual visual correction of translation and rotation was carried out such that the registration is accurate. The MR images were analyzed for image quality and geometry.

Result: Vienna applicator did not show much difference in the image quality between 1.5 and 3.0T, however the titanium Fletcher-suit applicator resulted in artifacts which were larger in 3T as compared to 1.5T. For Titanium needles, T2 images had a better contrast between gel and the needle as compared to T1, however the needle tip could be more easily located in T1. In both the series the lumen of the needle could be identified as a void signal. When compared between T2 weighted FRFSE and FSE sequence, the former one had a grainy appearance but the needle body and tip could be well identified in both. The T2 weighted FSE sequence had a distinct artifact at the tip with a signal void region extending up to 0.8 cm but when fused with CT the needle tip was found to lie beyond the artifact at 0.5 cm in the signal void region, while in the T1 image the susceptibility artifact matched the CT tip. SSFSE series images were not of good quality. We have found that T1 weighted images with 100 kHz bandwidth, TR= 1160, TE min, Echo train length=6, slice thickness=3 mm and spacing of 0.5 mm, are best for titanium needle reconstruction with 3T.

Conclusion: The commissioning of the applicators was successfully carried out, as a part of prerequisite for the clinical implantation of MR image based Brachytherapy.

 BP-2: Radiobiological and Second Cancer Risk Estimates for Institutional Three Dimensional Conformal Planning Method: Comparison with Traditional Gap Junction and Inverse Imrt Plans of Pediatric Medulloblastoma

A. Hemalatha, M. Athiyaman, Arun Chougule1, H. S. Kumar2

Departments of Radiological Physics and 2Radiotherapy, Sardar Patel Medical College, Bikaner, 1Department of Radiological Physics, Sawai Man Singh Medical College, Jaipur, Rajasthan, India. E-mail: [email protected]

Introduction: The use of Radiation Therapy (RT) is associated with better outcome in many pediatric cancer patients but it is also associated with significant long term side effects and the risk of Second Cancers (SC). The present study is an attempt to analyze the radiobiological and SC risk associated with institutional Three Dimensional Conformal Radiotherapy Technique (3D CRT) in comparison with traditional 3DCRT gap junction and inverse Intensity Modulated RT (IMRT) for Cranio Spinal Irradiation (CSI).

Materials and Methods: The three different planning methods were created for ten pediatric medulloblastoma patients retrospectively in the eclipse treatment planning system (V13.7). The dose calculation was performed based on Anisotropic Analytical Algorithm (AAA) for 6 MV linac photon beam. The prescribed dose was 23.4Gy, 1.8Gy/fraction.

Planning Methods: The half beam blocked cranial and cervicothoracic spine field was created in both institutional and gap junction planning method. In the planning method 1 and 2 the diverging planes of the both spine field were matched using field alignment option and junction was shifted after seven fractions to reduce dose uncertainty created by error in setup. In the gap junction method the both diverging planes were separated by distance S=S1+S2.S1 = 0.5 x L1 x (d/SSD1), Where L, SSD-Length and Source to Surface Distance of the corresponding fields, d-depth of spine. For the inverse IMRT plan the symmetrical bilateral cranial fields were matched with posterior spine field and inverse optimization was performed.

The Organ Equivalent Dose (OED) was calculated by plateau dose response model using equation A:


where V is the whole organ volume, Vi & Di - volume and dose elements, which was extracted from differential dose volume histogram. And δ δorg is the organ specific model parameter taken from Schneider et al. The Lifetime Attribute Risk (LAR) of SC for age and sex dependent, site specific estimation was performed using Biological Effects of Ionizing Radiation VII parameters (R)


Results and Discussion: There is no significant difference between planning methods in case of pneumonitis. The probability of heart failure is significantly lower for IMRT compared to planning method 1 Figure 1. The LAR estimation from Figure 2 shows that SC in lung, thyroid and colon cotributed most to the overall risk in all compared modalities.{Figure 1}{Figure 2}

Conclusion: From this study it concluded that with respect to sparing organ at risk during CSI the both institutional planning method and IMRT were found to have advantages compared to gap junction method. The LAR of SC is lower for all the organs with institutional planning method compared to IMRT and gap junction method.

 BP-3: Retrospective Analysis of Geometrical Information About Parotid and Planning Target Volume to Predict the Parotid Mean Dose

Amit Nirhali, V. K. Sathiyanarayanan, M. Pooja, S. Mithun, Basu Sumit, Bhooshan Zade

Department of Radiation Oncology, Ruby Hall Clinic, Pune, Maharashtra, India. E-mail: [email protected]

Introduction: Intensity Modulated Radiotherapy (IMRT) is the widely used technique in Head and Neck cancer cases, in which PTV is surrounded by OAR structures, so IMRT planning has a challenge of reducing the OAR Doses. Mainly salvaging parotid, as many times parotid overlaps with the PTV. It's very difficult to know the level to which parotid dose can be reduced maintaining the PTV Coverage. The purpose of this study is to develop a planning tool to improve the plan quality by generating a parotid mean dose predictive model based on the retrospective analysis of geometrical information database about parotid and planning target volume (PTV), just before the planning is done.

Materials and Methods: We randomly studied 60 HN IMRT cases (115 Parotid structures) 80-training data and 35-test, data planned on Pinnacle9.8 TPS with DMPO optimization with 6MV photon beam and 7 to 9 beam directions and planned by different physicist and treated on Siemens ONCOR impression during the year of 2011 to 2016.

For each selected patient we tried further to reduce mean dose of Parotid without compromising PTV coverage and hot dose and extracted the total Parotid volume (Vp), parotid volume overlapped with PTV(Vp'), Mean dose of Parotid (Dmean) and Dose Prescribed to the PTV with which parotid is overlapping (Dpd). A retrospective analysis of data extracted from each patient has shown a correlation between the fractions of parotid overlapping the PTV and the Dmean/Dpd. Exponential regression was done on plotted graph between the fractional overlap of parotid and Dmean/Dpd. A mathematical model as shown in equation (1) was generated for predicting Dmean based upon the fractional OAR-PTV overlap (Vp'/Vp),


Where A, B, C are model parameter selected to best fit the data.

To validate this model we used the above equation to predict the mean dose of 35 parotid volumes (Dpred) compared with the planned mean dose (Dplan) of the parotid. The difference between the predicted and planned mean dose (Δ) was calculated,


Result and Discussion: The Data analysis of fraction overlap of parotid Vs Dmean/Dpd showed correlation of 94.7% (p value <0.0001).

The exponential regression is done with mathematical modeled Parameter adjusted to A=1.35, B=0.9409 and C=1.2828 to provide an best fit the mathematical model to the local plot of Dmean/DPd versus Vp'/Vp Figure 1, the R2 value for the above regression was 0.9545 (P value<0.0001). Thus equation below was derived to predict the parotid mean dose,{Figure 3}


In all the cases, data under test category showed smaller deviation (Δ) between predicted and planned parotid mean dose dose Figure 2.{Figure 4}

Conclusion: This tool helped the planner to decide the lower level of parotid dose that can be achieved and for oncologist to alter the overdrawn CTV without compromising treatment area which reduces the overlap of parotid with PTV.

 BP-4: Predicting the Impacts of Daily Translational Couch Shifts on Dose Volume Histogram and Radio Biological Parameters of Vmat Plans Using Curve Fitting Method

M. P. Noufal, K. K. Abdullah1, P. Niyas, P. Subha1

Department of Medical Physics and Radiotherapy, Baby Memorial Hospital, 1Department of Physics, Farook College, Kozhikode, Kerala, India. E-mail: [email protected]

Introduction: Dose volume histogram (DVH) plays a vital role in evaluating the treatment plans. But patient set up errors introduces variations in the daily DVH and radiobiological parameters. Our study aims to check the feasibility of predicting daily DVH by incorporating random translational couch shifts, using curve fitting method, and using this to monitor changes in the radio biological parameters on a daily basis.

Materials and Methods: Let Dj(x0, y0, z0) be the dose received by Vjth volume of a structure in the DVH of the base plan, planned with an iso-center (IC) co-ordinate (x0, y0, z0) Figure 1a. When there is a translational couch shift of 'i' on either side of the IC position (x0+i in the right or positive x direction and x0-i in the left or negative x direction), the dose received by Vjth volume due to translational couch shift now becomes Dj(x0+i, y0, z0) and Dj(x0-i , y0, z0) provided, there are no shifts in the y and z axis. Thus variations in Dj(x0, y0, z0) along the x direction can be represented by the fitted function f(x, vj) Figure 1b. Similarly variation in Dj(x0, y0, z0) along the y and z axis can be represented by functions f(y, vj) and f(z, vj) respectively. For each point on Vj, corresponding functions can be calculated and it can be used to predict effect of daily random couch shifts on DVH by the mathematical modeling of the base plan DVH without using any further dose computation on CT. To demonstrate this, we have selected 10 prostate patients treated with VMAT technology. Systematic couch translation shifts were introduced in the clinically accepted base plans with an increment of 1 mm and up to 5 mm from the IC in both positive and negative directions of each of the three axis, x (right–left), y (superior–inferior) and z (anterior–posterior). The DVHs of the base plan and the error plans were imported into the MATLAB software (2009b, The MathWorks, Natick, MA) and in-house MATLAB code was generated to find the best curve fitted functions f(x, vj), f(y, vj) and f(z, vj) for each points on the DVH and there by generating predicted DVH for PTV, CTV and OARs. It is then used to find the daily radio biological parameters such as the EUD (Equivalent uniform dose), tumor control probability (TCP) and normal tissue complication probability (NTCP). The EUD was calculated using the Niemierko model and NTCP values were calculated using the Lyman–Kutcher–Burman model . Finally, the MATLAB predicted and the T.P.S calculated DVH was compared to validate our method and percentage of variation between the two was evaluated.{Figure 5}

Results and Discussion: When T.P.S calculated and MATLAB predicted DVHs were compared, both curves almost overlapped with each other and the maximum variation between the two curves at any point was less than 0.5% in targets and OARs Figure 2a. The mean and standard deviations of the variations in the DVH and radio biological parameters due to daily shifts, calculated by T.P.S and MATLAB predicted methods also showed a good correlation Figure 2b.{Figure 6}

Conclusion: Our method of predicting the effect of couch shifts on DVH and radiobiological parameters using curve fitting method is found effective to determine the uncertainties in dose delivery on a daily basis, and to enhance the quality of treatment.

 BP-5: Study of the effects of Dwell Time Deviation Constraint on Inverse Planning Simulated Annealing optimised plans of Intracavitary Brachytherapy of cancer Cervix

Saurabh Roy1,2, V. Subramani3, Kishore Singh1, A. K. Rathi1

1Department of Radiotherapy, Lok Nayak Hospital, 3Department of Radiotherapy, Dr BRA IRCH, AIIMS, New Delhi, 2Research and Development Center, Bharathiar University, Coimbatore, Tamil Nadu, India. E-mail: [email protected]

Introduction: High Dose Rate (HDR) remote after loading brachytherapy machine and advanced treatment planning system have made it possible to make variations in individual dwell times across a catheter according to tumour density and for sparing normal structures. There is also the development of several inverse optimization methods such as Inverse Planning Simulated Annealing (IPSA) and Hybrid Inverse Planning Optimization (HIPO). But their actual application is not so frequent for treatment of cervical cancer with Intracavitary brachytherapy. One of the several reasons of it, is the limited research of inverse planning optimization for volume based intracavitary brachytherapy. Therefore very few institutions are venturing towards volume based IPSA optimised intracavitary brachytherapy. This study focuses on dwell time deviation constraint (DTDC) feature of IPSA based optimization which restricts the large variation of dwell time across the catheter. The aim to investigate the dosimetric variations in IPSA optimised intracavitary brachytherapy plans with the gradual change in DTDC values.

Materials and Methods: For this retrospective study we have generated IPSA optimised intracavitary brachytherapy plans for 20 cancer cervix patients using Oncentra Brachytherapy treatment planning system (version 4.3). The initial DTDC value of each IPSA plan was kept 0.0. Later on gradual increment was made in DTDC values in step of 0.2. The variations in dose volume histogram parameters D90%, V100%, V150%, V200%, for PTV were studied. For dose analysis of bladder and rectum, their D2cc parameters were investigated. Plan modulation index (M) defined by Ryan L. Smith et al . was used for characterising the variation of dwell time modulation with respect to gradual increase in DTDC parameter.

Results: Plan modulation index gradually decreases with increasing value of DTDC from 0.0 to 1.0. There was the 83% decrease in M value from IPSA of DTDC 0.0 to fully constrained IPSA of DTDC 1.0. Number of activated dwell positions decreases for DTDC value of 0.0 to 0.2, but then gradually increases for increasing DTDC values. Total treatment time of IPSA plan decreases with increase in DTDC value. There was 6.89 % and 18.13 % decrease in average D90% value for DTDC 0.4 and 1.0 respectively compared to average D90% for IPSA plan of DTDC 0.0. The average value of V100% decreases with increasing DTDC and rate of decrement decreases from DTDC value 0.0 to 0.6 and then increases till 1.0. It was observed that both the values decreased with increasing DTDC value for IPSA plans. There was 4.37 % and 11.23% decrease in average PTV V150% values for DTDC 0.4 and DTDC 1.0 respectively compared to DTDC 0.0. Average D2cc values for rectum and bladder decrease with increasing DTDC values. There is reduction of 8.26% and 6.95% for D2cc values of rectum and bladder respectively for DTDC 1.0 compared to DTDC 0.0.

Discussion: Increase in DTDC value restricts the deviation of dwell time from average dwell time of every catheter. A higher value of DTDC in IPSA plan removes the larger of the dwell time in isolated dwell positions and also reduces overall treatment time, but this is at the cost of decreasing D90% and V100% value which indicate lower PTV coverage.

Conclusion: The benefits of applying DTDC values in IPSA plan for intracavitary brachytherapy are that, it removes local hot spots and reduces rectum and bladder doses.

 BP-6: Volume Mis-Estimation by CT Scan: Importance of CT PERFORMANCE in Precision Lung Cancer Radiotherapy

Teerthraj Verma, Nirmal K. Painuly, Surendra P. Mishra1, M. L. B. Bhatt, Anoop K. Srivastava1, Ranjit Singh2, Dipak Shrotia3, Shraddha Srivastava, Navin Singh

King George's Medical University, 1Dr. Ram Manohar Lohia Institute of Medical Sciences, Lucknow, 3J.K. Cancer Institute, Kanpur, Uttar Pradesh, 2Postgraduate Institute of Medical Education and Research, Chandigarh, India. E-mail: [email protected]

Introduction: Accuracy of radiation delivery to a moving target found directly related with the breathing motion as well as information provided by CT acquisition. The radiation treatments are becoming increasingly conformal and the pinpoint accuracy of modern radiation treatments would be worthless if the tumor was not in the spot where the radiation beams were aimed. Any geographic miss or incorrect isodose coverage of the region of interest (tumor) may lead to significant deviation of the planned dose, which could produce catastrophic results.

Objective: The purpose of present study was to quantify the volume mis-etimation by CT of target that underwent 3D motions placed inside the CIRS 3D phantom.

Materials and Methods: First validation of wood cylinder of diameter 6.4 cm as lung equivalent inhomogeneity was performed by measuring linear attenuation coefficient and HU value of wood slabs. A cavity of dimension 2x2x2 cm3 was created in the wood cylinder. Then this cavity was filled with dental material to create the tumor of dimension 2x2x2 cm3. It was ensured that center of this tumor was lying at the line joining the two lateral external surface markers on the CIRS phantom at the depth of 7.5 cm. This position of tumor was considered rest position and various movements (AP-PA, SI & lateral) were given with respect to rest position. CT images of CIRS phantom containing tumor model were acquired after proper alignment with three lasers similar to real patient simulation. Subsequently the tumor was shifted ±5 mm, ±15 mm and ±25 mm with respect to “Rest Position” along superior and inferior directions. CT images were acquired after each movement. In total, 15 movements were given to the tumor to implement 1 D, 2D and 3D motions similar to real lung movements with the amplitude(s) as reported in various literatures.

Results and Discussion: To understand the impact of three dimensional displacement of tumor on delivered dose, a clinically relevant target having dimension 2 cm X 2 cm X 2 cm sandwiched between two wood slabs of wood cylinder (lung equivalent inhomogeneity) were put in the hollow space provided in CIRS phantom. The resulting scans showed that object shape was significantly distorted. Generally, it is the CT data that provides the tumor information (localization, volume of the target and its extension etc.) for the diagnosis as well as more importantly for radiotherapy treatment planning purposes. Therefore results of radiotherapy are affected directly by the performance of CT. In this study, the variations in tumor volume with target movements given by CT were recorded and the same was compared with its true physical volume. Trend of increase in overestimation of target volume (by CT imaging) with 3 directional movements range was found. The volume of the target was 7.8 cm3 (rest position) however variation up to 9.5 cm3 was observed in CT volume of the target with different movements. The irradiation of target with overestimated volume compared to true physical volume causes unnecessarily normal tissue irradiation.

 BP-7: Montecarlo Based Correction Factors for Small Field Detectors

P. S. Renil Mon, Sneha S. Nair, Raghavendra Holla, C. O. Clinto, M. K. Ashitha, E. Sreedevi, Bhaskaran K. Pillai

Department of Medical Physics and Radiation Safety, Amrita Institute of Medical Sciences and Research Centre, Kochi, Kerala, India. E-mail: [email protected]

Introduction: A radiation field is considered as small if the field dimension is less than the range of secondary electrons and the collimating devices are partially occluding the source. There are different types of detectors proposed for small field measurement, which includes unshielded diode, diamond detector, and small volume ion chambers. Even though the active volumes of these detectors are small, the non-water equivalent material of the detector will cause response variations. To overcome this uncertainty Alphonso et al introduce a new formalism for small field reference dosimetry. According to this formalism the relative output factor (ROF) which converts the absorbed dose to water by the standard reference field size to the given field size has to be corrected by the factor [INSIDE:1]. fmsr and fclin are the machine specific reference field and the given clinical field. Qmsr and Qclin are the corresponding beam qualities.

Objectives: Aim of our study is to calculate the [INSIDE:2] factors for our clinical detectors, Edge detector (Sun Nuclear), 60017 diode (PTW), CC01 ion chamber (IBA) for SRS cones of diameter 5 mm to 15 mm in Elekta Synergy linear accelerator using Monte Carlo simulation.

Materials and Methods: Elekta Synergy Linear accelerator treatment head was simulated using BEAMnrc Monte Carlo code as per manufactures specification. All the three detectors we used were simulated as per the manufacture's specification. Three EGSnrc user codes were used for the detector simulation based on the detector geometry. Since PTW T60017 has a cylindrical geometry DOSRZnrc was used. Because of the rectangular geometry DOSXYZnrc was used to model the Sun Nuclear Edge diode and CC01 chamber was modelled with EGSchamber user code.

The Mote Carlo model of the treatment head was validated against the measured data for standard field size of 10x10cm2. Off axis profile, PDD and were verified as validation procedure. All measurements were done in 50X50X50 cm3 water phantom (PTW RFA). PDD curves and profiles of cone fields were taken in 1mm resolution with PTW T60017 unshielded diode in integrated mode. Profiles were taken at the depth of maximum dose, at 5 cm and at 10 cm in SSD setup. ROF were also measured at the depth of maximum dose with reference field size fmsr (10X10 cm2) for all the three detectors.

Results and Discussions: Measured and Monte Carlo calculated ROF were not in agreement. Edge and PTW 60017 diodes overestimated the ROF in the smaller field sizes 5 mm and 7.5 mm. In 5 mm field size edge diode over estimated by 9.5% while 60017 diode to 6.7%. In 7.5 mm field size variation was 7.8% and 4% for edge and 60017 diode respectively. 1% to 3% variation were observed for other field size of 10 to 15 mm in both the diode detectors. This variation is mainly due to the over response of silicon and the presence of other high density material. Even though the active volume of the silicon chip of edge diode is less, its response is more than that of 60017 diode because of the presence of copper plate and brass covering. CC01 ion chamber under responded up to 10% because of its low density active volume. When applying the Monte Carlo calculated correction factor on the measured ROF it is in good agreement with the Monte Carlo calculated ROF.

 BP-8: Development and Validation of a Matlab Software Program for Real Time and Positioning Management Gating Generated Breathing Trace

Naveen Kumawat, Anil Kumar Shrotriya1, H. Malhotra Singh, Kartikeswar Patro, Anil K. Bansal, R. K. Munjal, A. K. Anand

Department of Radiation Oncology, Max Super Speciality Hospital, Saket, New Delhi, 1Department of Physics, S.P.S.B. Government P.G. College, Shahpura, Bhilwara, Rajasthan, India. E-mail: [email protected]

Introduction: In radiotherapy, the goal is to kill cancerous tumour while sparing surrounding normal tissues. This is more difficult when the target is moving with patient's breathing, so either take large PTV margin or to miss the target during treatment. To overcome this problem, AAPM TG 76 recommends, if the target displacement is more than 5 mm, motion management techniques should be considered. RPM is one of the motion management technique, provides facility to treat patients in amplitude based and phase based breathing cycles.

Objective: RPM version 1.7.5 has a limitation with its clinical application, it provides the facility to deliver radiation with either phase or amplitude based gating, but 4D-CT images are sorted only with respiratory phase. In regular breathing cycles, there is no significant difference between phase and amplitude based treatment, but in irregular breathing the performance of RPM is not robust with phase based gating and does not provide an easy way to change phase to equivalent amplitude based gating parameter. So, we developed a MATLAB Software program to calculate parameters for amplitude based treatment and analyse a RPM gating generated breathing profile.

Materials and Methods: For conversion of phase based gating window to amplitude based gating threshold, a program was designed using MATLAB (Matrix Laboratory, Math Work, R2016a) software. With this program, we analysed phase based gating profile as well as amplitude based gating profile using reference profile. Using this software program, we got predicted amplitude window for treatment, actual treated amplitude window, treatment range in terms of amplitude, duty cycle, errors in phase and amplitude. For validation of this program, we took a Varian Gating Phantom generated regular breathing trace of reference scan and found predicted amplitude parameters, and irradiated the phantom retrospectively with predicted parameters on phase and amplitude based gating. We selected 85% to 15% phases window for treatment, the phase window selection criteria were random.

Results: From reference profile, the predicted lower and upper window level was 1cm and 1.8 cm respectively from fitted trace. For phase based and amplitude treatment the actual lower amplitude was 0.99 and 1.02 cm and higher amplitude was 1.65 and 1.64 cm respectively, Treatment range was 0.65 and 0.62 cm respectively, duty cycle was 28.51% and 30.37% respectively. “Beam on” error with phase based treatment in term of phase was 1.4% and in term of amplitude was 0.98% and with amplitude based treatment the “beam on” error was 0.85%. in amplitude based treatment the treated phase window was 83.59% to 15.83%.

Discussion: The selected phase gating window in reference profile and treated phase window in amplitude based treatment was almost equal. If we set asymmetric phase window for treatment like 90% to 20% then the treated phase window may not match with set one, but it does not have significance because treatment was done with amplitude mode. The treatment range was almost equal in both mode. The “beam on” error in both modality was ~ 1%, which was due to machine lag time. The duty cycle in both modality is almost same but may vary with window section.

Conclusion: The results are encouraging and comparable for both modes of treatment under ideal breathing cycles so the developed program can be used for irregular breathing patterns to find errors and duty cycle and other parameters. This program is capable to observe inter fraction errors and variations in breathing trace.

 BP-9: Hippocampal Sparing Whole Brain Radiotherapy: A Planning Study Comparing Coplanar and Noncoplanar Volumetric Modulated Arc Therapy Plans

C. A. Muthuselvi, K. Malathi, T. K. Bijina, A. Pichandi, S. Yuvarajan1, B. Subbulakshmi

HCG Enterprises Ltd, 1HCG-MSR Centre of Oncology, Bengaluru, Karnataka, India. E-mail: [email protected]

Introduction: Whole Brain Radiotherapy (WBRT) is the most commonly used treatment option for patients with multiple brain metastases. WBRT provides rapid palliation of neurologic symptoms and improves local control as an adjuvant to resection or reduce surgery, but it also causes neurocognitive decline after therapy. Hippocampus sparing whole brain radiotherapy (HS-WBRT) can reduce neurocognitive deficits caused by radiation. Although avoiding hippocampus during whole brain irradiation is desirable, it is important to deliver a uniform dose to the rest of the brain for reducing the probability of cancer growth. The central location and unique anatomic shape of hippocampus shows great challenge in contouring and planning.

Objectives: The purpose of the study was to evaluate the feasibility of HS-WBRT using volumetric modulated arc therapy (VMAT). We performed this planning study comparing the PTV coverage, hippocampal dose as well as the unspecified tissue dose of VMAT using non- coplanar arcs and VMAT employed coplanar arcs for HS-WBRT.

Materials and Methods: Ten patients diagnosed with brain metastases underwent repeated planning of HS-WBRT. CT scan with 1.25 mm slice thickness was fused with MRI T1 & T2 weighted sequences of 1.25 mm slice thickness prior to the delineation of hippocampus, optic nerve, lens and chiasm. Hippocampal avoidance region was generated by three dimensionally expanding hippocampal contour by 5 mm. As per RTOG0933 recommendation, PTV was created by excluding the hippocampal avoidance region from CTV and prescribed to 30Gy in 10 fractions. Three VMAT plans, Coplanar VMAT (coVMAT) with three full arcs, Non-coplanar VMAT (1ncVMAT) with one coplanar full arc and two vertex arcs (150oarc, couch 90o), Non-coplanar VMAT (2ncVMAT) with one coplanar full arc and two oblique arcs (140oarc, couch 30o, 330o) were created for all ten patients using Monaco Treatment Planning System (TPS) v.5.1 for an Elekta synergy machine with agility MLC Figure 1. All three plans were normalized such that to meet RTOG acceptable compliance criteria for PTV and OAR's. Plans were evaluated using dosimetric indices such as D2%, D98%, V30Gy, HI to the target, Dmin, Dmean, Dmax to the hippocampus and V5Gy, V10Gy of unspecified tissue, excluding the target volume.{Figure 7}

Result and Discussion: All three VMAT plans achieved the RTOG0933 planning criteria Figure 2. Comparing the target doses in terms of D2%, D98%, V30Gy, HI there were no significant differences among the three plans. Comparing the hippocampus, the mean hippocampal dose was 8.3% higher in coVMAT than the non coplanar plans. Hippocampus minimum and maximum dose also lesser in non coplanar plans by 6.9% & 7.2% than the coVMAT plans. The maximum and minimum hippocampus dose+ SD were 16.64 ± 0.7Gy, 14.48 ± 0.9Gy, 14.95 ± 0.8Gy and 9.42 ± 0.3Gy, 7.34 ± 0.4Gy, 7.31 ± 0.4Gy for the coVMAT, 1ncVMAT and 2ncVMAT respectively. We thus found that the non-coplanar VMAT is better than the coplanar VMAT in hippocampal sparing. The optic nerve, chiasm and lens dose were within the tolerance limit and similar in all three plans. To quantify the normal tissue dose because of non-coplanar beams we compared V10Gy and V5Gy of unspecified tissue (body-target). V5Gy of unspecified tissue, excluding the target volume was 1174 ± 324cc, 1790 ± 274cc, 1294 ± 249cc for the coVMAT, 1ncVMAT, 2ncVMAT respectively. Unspecified tissue dose were higher in 1ncVMAT plan comparing with coVMAT and 2ncVMAT, because of the vertex field. The 2ncVMAT plan showed slightly higher unspecified tissue dose than the coVMAT plan.{Figure 8}

Conclusions: From our results, we conclude that 2ncVMAT significantly reduced the hippocampus dose without increasing the normal tissue dose. Thus, non-coplanar VMAT with oblique arcs (2ncVMAT) was better than coplanar VMAT plan (coVMAT) and non-coplanar VMAT with vertex arc (1ncVMAT) for hippocampal sparring in whole brain radiotherapy.