Benefits of the Real-time Approach Peter Bownes HDR Prostate - PowerPoint PPT Presentation

Benefits of the Real-time Approach Peter Bownes HDR Prostate Workshop Nov 2013

Real Time US Pathway

US Based Pathway

CT Based Planning Pathway

Treatment Accuracy HDR Accuracy - time/position Dose Calculation Biological Model Target Applicator Definition Reconstruction Target Localisation Insertion 3D Imaging - Applicator “A chain is no stronger than fixation its weakest link” - Guidance

Contouring Needle Reconstruction Plan Adaption Needle Displacement

Imaging Modalities T2 Weighted MRI US - BK Flex Focus MR Image courtesy of G Lowe, Mount Vernon Hospital

Imaging Modalities CT 3mm Slice Thickness T2 Weighted MRI BK Flex Focus MR & CT images courtesy of G Lowe, Mount Vernon Hospital

Image Acquisition for Contouring Modality Slice Advantages Disadvantages Thickness MRI Typically • “Gold Standard” for prostate • Patient transfers – Legs down T2 Weighted 2 – 3mm definition – good soft tissue • Time delay defn • Resolution – SNR/Time balance • Good for OAR definition • Additional scan/fusion of needle (snap shot”) recon • Easier to fuse mpMRI • Not real-time • Non ionising US Transverse 1mm • Excellent resolution • Image quality deteriorates after • Good definition of prosatate needle insertion (artefact – needle margin/bladder/sem ves shadowing / false catheter tracks) Or • Real time feedback – aid base • Quality dependent on good and urethra definition. acoustic coupling 0.5 ° rotation US Sagittal • Information from virtual stage ECRM • Needle position as surrogate • Non-ionising CT Typically • Good for needle recon • Inferior soft tissue definition 2 – 3mm • Adequate for OAR • Patient transfers – Legs down (“snap shot”) • Time delay • Resolution – dose / noise balance • Patient dose • Not real-time

Prostate Volume Comparison • McLaughlin IJROBP 54 (3) 703-711, 2002 – 45 I-125 patients – T2 weighted MR “gold standard” – superior at: • Pelvic diaphragm • Apex v soft tissue • Base v SV • Base v Bladder CT post /MR post CT post /US pre MR post /US pre MR post /MR pre 1.34 1.40 1.07 1.10 (SD 0.35) (SD 0.35) (SD 0.26) (SD 0.2)

Aid to target definition (Virtual – Live) 3.5cm 3.5cm 4.5cm 36.5cc Needle Base Apex

OAR stability • CT/MR “snap shot” for planning • Rectal shape/size/position can change • US probe at base – Rectal position stable – Real time verification – Time between plan/treat reduced

Contouring Needle Reconstruction Plan Adaption Needle Displacement

Image Acquisition for needle reconstruction Modality Slice Advantages Disadvantages Thickness US Transverse 1mm • Fine resolution • Tip difficult to define with US alone • Real time confirmation of tip • Needle artefact (shadowing) • Free length confirms tip • False catheter tracks (blood left position when catheter retracted) • Non-ionising 0.5 ° rotation US Sagittal ECRM • Automatic CT Typically • Visualisation of catheter/tip • Patient transfers – Legs down 2 – 3mm • Time delay • Partial Volume effect – Tip location • Patient dose • Resolution – dose / noise balance • Not real-time MRI Typically • 3D sequences available • Poorer visualisation to CT 2 – 3mm • Non ionising • Patient transfers – Legs down • Time delay • Resolution – SNR/Time balance • Distortion/patient movement

US Catheter Reconstruction • Requires good probe contact with rectal wall • Main uncertainty on ultrasound is the location of tip – Needle artefacts – Blood artefacts • How is uncertainty minimised to acceptable levels – Catheter “Free-Length” – Real time guidance • Zheng et al* Average tip-detection Free Length accuracy 0.7mm [max 0.8mm] (compared to X-ray) Cradle locking device for probe Mark on Probe * Brachytherapy 10 (2011) 466-473

Inter-observer variation in applicator reconstruction on OCP MSc Project L Partridge, Uni of Leeds 2008 • 2 Cases • 6 Observers • Free Length provided • Observer 1 – reference standard

How reconstruction error effects V100 prostate 100% ± 0.7mm equates to <0.5% V100 pros 99% 98% 97% 1 96% 2 V100 prostate 3 95% 4 5 94% 6 7 93% 92% 91% 90% -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 Offset (mm) Caudal Cranial

Are there issues with CT? CT Slice • Partial volume effects • Effect Tip detection • Random errors depend on slice thickness • Slice thickness – dose/noise balance • Window settings • Effects interpretation of tip location • Hypo-intense regions

Contouring Needle Reconstruction Plan Adaption Needle Displacement

• Real Time Approach • Allows for additional needle insertion if cold spots are seen in the dosimetry

• Real Time Approach • Allows for needle position adaption • Difficulty in achieving planning aims (target coverage & OAR constraints) • Needle placement is the most critical factor in achieving good dosimetry • Knowledge of achieved dosimetry while in theatre

Contouring Needle Reconstruction Plan Adaption Needle Displacement

Needle Movement Planning to Treatment Delivery

Mean Catheter Displacement – CT Based Simnor et al Radiother. Oncol 93, 253-258 (2009) Implant • 20 Consecutive monotherapy CT plan CT QA CT QA 2 nd # 3rd # 1 st # patients • 326 catheters • 3 Fractions of 10.5Gy (30-36 hrs) 0 1-2 4-6 24 30 (Hrs) Results Mean Interfraction movement of catheters relative to the prostate (CT plan – CT 2nd ) = 7.9mm; (range 0-21mm) (CT plan – CT 3rd ) = 3.9mm; (range 0-25.5mm)

Simnor et al Radiother. Oncol 93, 253-258 (2009)

Study Mean Catheter Displacement – Real-time US Milickovic et al Med. Phys 38(9), 4982-4993 (2011) Treatment US post-irrad US pre-irrad Implant • 25 patients - monotherapy US plan • 3 Fractions of 11.5Gy • 2wks between # - separate implant 51.2 (37–65) 19.3 (16-24) 0 • Probe @base during treatment Mean Time (mins) Prostate Vol mean 35.8 cm 3 • (17.4 cm 3 to 59.6 cm 3 ) • Analysed volume changes, needle displacement and dosimetric impact Results Prostate – rigid shift mean 0.57mm (0-2.1mm) Urethra - largest shift @ base mean (plan to post) 1.1mm (0-5.1mm) Rectum mean (plan to post) 0.4mm (0-1.4mm) Needle displacement mean 1mm (<1.5mm, except for one)

Milickovic et al Med. Phys 38(9), 4982-4993 (2011)

Mean Catheter Displacement Against Time Simnor et al Radiother. Oncol 93, 253-258 (2009) Simnor (2009) 20 mono 3 Fractions 30-36h CT Scans (3mm) Tip @ 1 st # Yes Milickovic (2011) Simnor 25 mono Milickovic 3 Fractions US Tip/Free length Yes

Conclusion (1) • US Contouring • Excellent resolution of US acquisition • Real time aids definition • Virtual information can help with needle artefacts • Better soft tissue definition to CT • Needle reconstruction • Fine resolution of US acquisition, real time tracking and free-length measurements allow accurate reconstruction • Plan Adaption • Additional needles / adjustment of needle position can be made after first iteration of planning. • Dosimetry known whilst still in theatre

Conclusion (2) • Needle Displacement – Planning to Treatment • Time of procedure (implant to treat) significantly reduced (less than 2hours) • Reduces the level of oedema at time of treatment • Reduces catheter displacement (<1mm) – minimal dosimetric impact • Treatment delivery position = Planning position • Reduces movement of patient (avoids: theatre-recovery-ward-CT-ward-treatment) • OAR position stable (US probe at base) • Patient Comfort – entire treatment completed under GA; no corrective action for needle displacement • Team in same place • Uncertainties minimised if separate insertion/plan for each fraction

Thank You peter.bownes@leedsth.nhs.uk