8/12/2011 Disclosure Patient Positioning in Head & Neck Treatment Lei Dong, Ph.D. Dept. of Radiation Physics University of Texas MD Anderson Cancer Center, Houston, Texas Sponsored Research: Varian Sponsored Research: Philips NIH Grant: Optimization of Proton Therapy CPRIT Grant: Advanced Volumetric Imaging and Adaptive RT for Lung Cancer AAPM Therapy Educational Symposium Aug. 2nd, 2011 Outline Learning Objectives • • • Learn about essential components of Head & Neck Immobilization Learn about IGRT applications and limitations Suggest IGRT/ART strategies Immobilization IGRT Adaptive RT 1 8/12/2011 Immobilization Principles for H&N Head Aquaplastic mask Bite block Mouth-piece Neck Clavicles Thermal Mask Immobilization Cutout over eyes and mouth, bolus when needed Chin Custom head mold Aquaplastic mask Shoulder restraining methods Mouth-piece/Bite-block for Tongue/Chin Immobilization 2 8/12/2011 Civico How effective is an immobilization device? How much margin do we need if IGRT is not used Orfit Daily Setup Shifts Evaluated by IGRT CIVICO ORFIT 10/12 patients and 350 measurements 3 8/12/2011 Mask doesn’t properly fit Margin Considerations Direction Systematic Error (cm) Random Error (cm) Van Herk’s Margins (cm) Civco Orfit Civco Orfit Civco Orfit AP 0.163 0.202 0.124 0.125 0.494 0.593 SI 0.156 0.122 0.156 0.117 0.499 0.387 ML 0.127 0.161 0.107 0.109 0.392 0.479 Planning CT Daily CT09 Measuring set up using MatriXX system Mask doesn’t properly fit • Base Frame • 9o Wedge • Foam Head Rest (Medium Density): 1 to 6 B C Energy: 6MV F Daily CT09 Planning CT Author: Luis Fong, Ph.D. Mayo Clinic 4 8/12/2011 No. 1 MatriXX Measurement Rt From IMRT to IGRT % Transmission 98.0-100.0 96.0-98.0 94.0-96.0 92.0-94.0 90.0-92.0 Inf Transmission through Central Area Average 95.8 Max 93.7 Min 97.0 Author: Luis Fong, Ph.D. Mayo Clinic Image Guidance Technologies for H&N IMRT Direct Portal Verification Daily orthogonal KV x-rays Implanted fiducials and kV x-rays Optic-based surface alignment In-room volumetric imaging Tomotherapy MV CT Cone-beam CT (KV and MV) kV CT-on-rails MR (in development) To Isocenter Verification 5 8/12/2011 Isocenter Verification KV DRR MV Isocenter Verification KV DRR MV AHE enhancement KV-KV What is the planning CT scan slice spacing in your practice? 1. 2. 3. 4. 5. kV DRR 5 mm 3 mm 2.5 mm 2 mm < 1.5 mm On-board kV Imaging 6 8/12/2011 2D planar x-rays = volume alignment? • Traditional alignment assumes rigid-body for the entire H&N region • Can you combine multiple 2D ROI for a 3D ROI? • Setup Uncertainties In Head & Neck Treatment 2 x 2D = 3D? Elapsed Days Patient with Tongue Base Carcinoma 19 CT Scans over 47 Days Lei Dong et al. (MDACC) DailyVoluntary Setup Uncertainties JunctionCaptured Shifts by Orthogonal kV Images 2D vs. 3D Alignment 7 8/12/2011 Weekly film discovered: IGRT is off? 3D vs. 2D In-room CT Plan Daily IGRT Reduces Margins by Half Imaging Dose Planning CT 28 consecutive patients 1013 CBCT scans Pre-alignment margins: ~ 4-5 mm Post-alignment margins: ~ 1.5 – 2.5 mm More margins may be needed larynx cancers or patients with significant weight loss Den et al., IJROBP v76 No. 5, 2010 CBCT “Standard Head” 0.4 cGy Varian OBI ver 1.4 8 8/12/2011 Planning Planning Daily Daily 3D IGRT Challenge Which structure to align? How to handle non-rigid setup error? Examples of anatomy rotation between the planning CT (first row) and the daily CT (second row). The axial CT images shown on the left indicate a roll in the patient’s head; the coronal CT images on the right show yaw (rotation of the spinal column). Change in the Neck Curvature Planning CT PPM Daily Cone-beam CT with planning contour overlay C2 C6 L. Zhang et al. "Multiple regions-of-interest analysis of setup uncertainties for head and neck cancer radiotherapy," Int. J. Radiat. Oncol. Biol. Phys. 64 (5), 1159-69 (2006). 9 8/12/2011 40 Histogram of PPM Measured Random Shifts in AP Direction 1 .5 0 .5 0 .0 -0 .5 C6 PPM -1 .0 -1 .5 SI 1 .0 0 .5 0 .0 -0 .5 C6 PPM -1 .0 -1 .5 -1 .5 -1 .0 -0 .5 0 .0 0 .5 1 .0 1 .5 C2 Shifts (cm) -1 .5 -1 .0 -0 .5 0 .0 0 .5 1 .0 1 .5 C2 Shifts (cm) RL 1 .0 Zhang LF et al. Multiple regionsof-interest analysis of setup uncertainties for head-and-neck cancer radiotherapy. Int J Radiat Oncol Biol Phys 2006;64:15591569. 0 .5 0 .0 -0 .5 C6 PPM -1 .0 -1 .5 -1 .5 -1 .0 -0 .5 0 .0 0 .5 1 .0 30 20 10 Stent (AP) Without Stent (AP) 0 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 40 Histogram of PPM Measured Random Shifts in SI Direction 30 20 10 Stent (SI) Without Stent (SI) 0 Red: no mouth-piece -1.0 Blue: with mouth-piece 1 .5 C2 Shifts (cm) -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 40 Percent C6/PPM Shifts (cm) 1 .5 Histogram comparison of PPM measured random shift between stent group and nonstent group Percent 1 .0 Percent AP C6/PPM Shifts (cm) C6/PPM Shifts (cm) 1 .5 Histogram of PPM Measured Random Shifts in RL Direction 30 20 10 Stent (RL) Without Stent (RL) 0 -1.0 -0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 PPM Measured Random Shifts (cm) Setup uncertainties in subregions of H&N anatomy Increased deformation with longer distances from the reference (C1-C3) Kranen et al., IJROBP, V73, pp1566 (2009) Kranen et al., IJROBP, V73, pp1566 (2009) 10 8/12/2011 Impact of ROI Selection for Patient Setup A Region-of-Interest based image guidance strategy is preferred The selection of ROI depends on the clinical case Kranen et al., IJROBP, V73, pp1566 (2009) Primary Tumor Response To RT 140 HN01 120 HN03 100 HN04 HN05 Volume (cc) Anatomy Changes During Radiotherapy HN02 HN06 80 HN07 HN08 60 HN09 HN10 HN13 40 HN14 HN15 20 0 0 7 14 21 28 35 42 49 Elapsed Treatment Days Barker et al. Int J Radiat Oncol Biol Phys 2004; 59 (4):960-970. 11 8/12/2011 Significant Anatomic Variations Planning CT Lee, C., K. M. Langen, et al. (2008). Int J Radiat Oncol Biol Phys 71(5): 1563-71. A RTOG H0022 case During Treatment CTV is in the air! C2-bone Setup and Anatomy Change Image-guided Adaptive Radiotherapy for H&N Dong/MDACC 12 8/12/2011 Benefit of IGRT Alone Parotid gland mean dose above the planned dose by 5 to 7 Gy in 45% of the patients Bone (C2) alignment led to reductions relative to BB alignment in 91% of patients (median, 2 Gy; range, 0.3– 8.3 Gy). The parotid dose from bone alignment was still greater than planned (median, 1.0 Gy, p = 0.007). Neither approach affected tumor dose coverage. Who is going to draw these contours in adaptive radiotherapy? O’Daniel et al. IJROBP vol.69 No.4 (2007) What is Deformable Image Registration? Auto-Segmentation of H&N Anatomy Planning CT Image Mid-Course CT Image Geometric transformation that brings one image in precise spatial correspondence with another image. Registration Planning Contours Deformed Contours 13 8/12/2011 Head and Neck Case Replanning Challenges upper Planning contours were automatically transformed to subsequent CT images upper lower lower Volume DVH Dong/MDACC Re-Plan Activities 20% What are the benefits of adaptive replanning? Re-Plan1 Re-Plan2 10% 0% -10% -20% CTV70 % Change CTV63 % Change CTV57 % Change -30% Left Parotid % Change Right Parotid % Change -40% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 14 8/12/2011 Re-Plan for Fraction #11 Re-Plan for Fraction #11 Fraction #11: 1st replan Fraction #11: 1st replan Solid lines: 1st replan Dotted Lines: original plan Target Coverage is not a problem Original Plan Re-calculated on Fraction #11 Replan on Fraction #11 CT 15 8/12/2011 Fraction #11: 1st replan Solid lines: 1st replan Dotted Lines: original plan Fraction #11: 1st replan Contra-lateral parotid can be reduced by replanning Fraction #11: 1st replan Solid lines: 1st replan Dotted Lines: original plan Solid lines: 1st replan Dotted Lines: original plan Replanning benefits both parotids Fraction #23: 2nd replan Solid lines: 2nd replan Dotted Lines: 1st replan Oral cavity 16 8/12/2011 Fraction #23: 2nd replan Tumor Progression During Treatment Planning CT 1st replan re-calculated on Fraction #23 2 weeks into treatment Replan on Fraction #23 CT Practical Issues in Adaptive RT Mask is not fitting How do you improve immobilization? Planning First Fraction Last Fraction 17 8/12/2011 H&N Setup Uncertainties Adaptive Mask? 1.0 pat1 pat2 pat3 pat4 pat5 pat6 Linear (pat1) Linear (pat2) Linear (pat3) Linear (pat4) Linear (pat5) Linear (pat6) 0.9 0.8 3D Shifts (cm) 0.7 Cranial Stopper 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0 5 10 15 20 25 30 35 Fraction # Summary for H&N patients • Setup uncertainties – Need to worry about residual setup error at different parts of the H&N anatomy Acknowledgement • MD Anderson H&N Team • ROI-based alignment is strongly recommended – Immobilization technique is important to minimize non-rigid changes. – Systematic uncertainties appear to be larger than random setup uncertainties. • Offline correction strategy is beneficial. • Image Guidance – 3D alignment can resolve out-of-plane rotations and 3D shifts of ROI better than 2D projection method • Trends in volumetric and positional variations – Adaptive radiotherapy strategy 18