Innovations conference 2014 dr john kipritidis can we use ventilation imaging to measure predict and reduce pulmonary function loss
•Download as PPTX, PDF•
2 likes•526 views
Dr John Kipritidis - Can We Use Ventilation Imaging to Measure, Predict and Reduce Pulmonary Function Loss in Lung Cancer Radiotherapy
More Related Content
What's hot
What's hot (20)
Similar to Innovations conference 2014 dr john kipritidis can we use ventilation imaging to measure predict and reduce pulmonary function loss
Similar to Innovations conference 2014 dr john kipritidis can we use ventilation imaging to measure predict and reduce pulmonary function loss (20)
More from Cancer Institute NSW
More from Cancer Institute NSW (20)
Recently uploaded
Recently uploaded (20)
Innovations conference 2014 dr john kipritidis can we use ventilation imaging to measure predict and reduce pulmonary function loss
- 2. Can we use Ventilation Imaging to measure, predict and reduce pulmonary function loss in lung cancer radiotherapy? SYDNEY MEDICAL SCHOOL John Kipritidis, CINSW Early Career Fellow Radiation Physics Laboratory UNIVERSITY OF SYDNEY Innovations in Cancer Treatment and Care Conference October 17th 2014
- 3. Acknowledgements Nepean Cancer Care Centre Dr Fiona Hegi-Johnson Dr Roland Yeghiaian-Alvandi Jeffrey Barber Dr Chuong Bui Katrina West Kylie Unicomb Peter MacCallum Cancer Centre Prof. Michael Hofman. Dr Shankar Siva Jason Callahan Prof. Rodney Hicks Contributing authors: University of Sydney Prof Paul Keall Dr John Kipritidis Dr Enid Eslick Andy Shieh Virginia Commonwealth Univ. Jeffrey Williamson, Ph.D. Geoffrey Hugo, Ph.D. Elisabeth Weiss, Ph.D. Special thanks: University of Sydney Dr Ricky O’Brien Benjamin Cooper Royal North Shore Hospital Dr Dale Bailey Dr Jeremy Booth Disclosure Supported by a Cancer Institute NSW Early Career Fellowship, NHMRC Australia Fellowship, NHMRC project grant 1034060 and NIH/NCI P01CA116602.
- 4. (i) Why is ventilation imaging important in lung cancer radiotherapy? › 10-30% of lung cancer radiotherapy patients experience radiation-induced lung toxicity (RILT) › Functionally weighted dose-volume metrics can outperform standard dose-volume metrics as a predictor of RILT; c.f. Hoover et al. 2014. J Med Imaging Radiat Oncol 58 (2) › Functional image-guided treatment planning requires functional imaging! `Gold standard' is SPECT ventilation / perfusion. 4 Technegas SPECT Galligas PET Peter MacCallum Cancer Centre Nepean Cancer Care Centre
- 5. (ii) Innovative ventilation imaging using 4D-CT › “CT-ventilation imaging” models regional air volume change in terms of regional lung volume (or intensity) changes during the breathing cycle. c.f. Guerrero et al. (IJROBP 2005) 5 (i) Acquire 4D-CT (ii) Deformable image registration (iii) Quantify volume/intensity 4DCT data courtesy of Nepean Cancer Care Centre change high low Ventilation Main advantages: High accessibility High resolution (same as CT) No-extra cost in scan time / imaging dose (just image processing!)
- 6. (iii) How could CT-ventilation be practice-changing? › CT-ventilation will allow greater access to functional image guided radiotherapy treatment planning, with the potential to reduce functionally-weighted mean lung dose by 2-5 Gy. Yamamoto et al. 2011 IJROBP 79 (1) › In-room 4D cone beam CT could maximise sparing of functional lung by enabling adaptive functional image guidance:
- 7. (iv) How is CT-ventilation being validated? We are aiming to validate CT-ventilation imaging across multiple modalities: Modalities 4D-CT 4D-CBCT SPECT V/Q 4D-PET V/Q PFTs 4D-CT VCU VCU NCCC PMCC RNSH RNSH NCCC BH-CT RNSH RNSH RNSH 4D-CBCT VCU VCU NCCC NCCC LEGEND Under-way Happening soon Not yet • NCCC = Nepean Cancer Care Centre (Ongoing QA study) • RNSH = Royal North Shore Hospital (Ongoing prospective trial) • PMCC = Peter MacCallum Cancer Centre (Ongoing prospective trial) • VCU = Virginia Commonwealth University (Existing database from earlier study)
- 8. 8 high low Ventilation Patient 7 (Best case) 4D-CT ventilation: MC rVHU Std PET ventilation: VPET (iv) How is CT-ventilation being validated? › 12-patient comparison* using baseline Galligas 4D-PET/CT scans. › Strongest voxel-wise correlation with nuclear medicine ventilation imaging (so far!) *Kipritidis et al. Med Phys 2014 41(1) 12 patients: Data courtesy of Peter MacCallum Cancer Centre (Melbourne, VIC Australia)
- 9. (iv) How is CT-ventilation being validated? 9 › Comparing daily 4D-CBCT ventilation images to baseline Q-SPECT in lung SBRT patients: › Ongoing QA study; comparison of functional changes underway. Data courtesy of Nepean Cancer Care Centre (Penrith, NSW Australia) high low Ventilation 4D-CBCT ventilation: SPECT perfusion: 5 patients:
- 10. (iv) How is CT-ventilation being validated? 10 • Comparing CT ventilation to Technegas V-SPECT: CT ventilation Technegas V-SPECT (HU based) Data courtesy of Nepean Cancer Care Centre (Penrith, NSW Australia)
- 11. (iv) How is CT-ventilation being validated? Author Reference modality Subjects Dice similarity Voxel-wise correlation Fuld et al. (J Apply Physiol 2008) Xe-CT 4 sheep ~ 0.81 (Small ROIs) Reinhardt / Ding et al. (Med. Image Anal. 2008) Xe-CT 5 sheep ~ 0.85 (Small ROIs) Yamamoto et al. (IWPIA 2010) SPECT V/Q 1 patient ~ 0.18/0.48 (Whole lung) Castillo et al. (PMB 2010) SPECT V 7 patients 0.30-0.35 (low function) ~ Castillo et al. (PMB 2012) SPECT Q 10 patients 0.78 (low function) ~ Mathew et al. (Med. Phys. 2012) 3He-MRI 11 patients 0.86-0.88 (good function) ~ Kipritidis, Siva et al. Ga 4D PET/CT 12 patients 0.38-0.68 (low function) 0.22-0.76 (Whole lung) Hegi-Johnson et al. SPECT V/Q 30 patients (goal) TBA TBA Eslick et al. Ga PET/CT 30 patients (goal) TBA TBA
- 12. (v) New technologies, new questions 12 4D-CBCT patient study: o 19 locally advanced NSCLC patients received daily 4D-CBCTs over 4-6 weeks o We generated 56 interfraction pairs (Week 1 vs. Weeks 2, 4 and 6). o Main question: How does ventilation change during radiotherapy treatment?
- 13. 100 101 102 103 104 105 106 107 108 109 • 4D-CBCT ventilation images can exhibit a wide range of changes (both positive and negative!) during treatment. 110 111 112 114 115 116 117 118 119 First day Week 2 Week 4 Week 6 • Adaptive functional image guidance is important as poor-functioning First day Week 2 Week 4 Week 6 Scan Ventilation increase Transient change Stable Ventilation decrease Highly variable lung can re-ventilate.
- 14. (v) New technologies, new questions 14 • Patients breathe differently from day-to-day (and breath-to-breath!) • Intrafraction changes can sometimes exceed interfraction changes. • Careful normalisation of serial images is required. 4D-CBCTs courtesy Virginia Commonwealth Univ. (Richmond, VA USA)
- 15. Take home messages › CT-ventilation imaging: a potentially practice-changing technology enabling (adaptive) functional image guidance in radiotherapy treatment planning. 15 › Australian researchers are driving world-first validation studies across multiple imaging modalities. › In-room ventilation imaging: innovative technology driving new questions.
- 16. Thanks for listening! SYDNEY MEDICAL SCHOOL Radiation Physics Laboratory UNIVERSITY OF SYDNEY John Kipritidis, CINSW Early Career Fellow john.kipritidis@sydney.edu.au
Editor's Notes
- Hi everyone, Im John Kipritidis from the University of Sydney. We’re investigating an innovative new technology, CT ventilation imaging, to measure lung function, predice and potentially even reduce lung function loss in radiotherapy.
- As usual there are many people to thank..
- The main clinical driver to look at ventilation imaging in lung cancer radiotherapy is that 10-30% of patients can experience some form of radiation-induced lung toxicity (RILT), the effects of which can range from merely radiographic, to debilitating shortness of breath requiring hospitalization. There is increasing evidence that RILT correlates with the mean-lung dose, but even more strongly with functionally-weighted dose-volume metrics (e.g. fV20). This drives the study of functional image-guided treatment planning, aiming to minimize the irradiation of healthy lung. This necessitates functional lung imaging; the clinical Gold standard methods are ventilation / perfusion imaging using single-photon emission CT (SPECT), and positron emission tomography (PET).
- So then, what if we could quantify lung function using respiration-correlated, four-dimensional (or `4D’) CT? In many centres in Australia, 4D-CT is the standard of care for respiratory motion management in lung cancer radiotherapy, most patients will get a 4D-CT scan. `CT-ventilation imaging’ is an innovative technique using the following steps, (i) First a 4D-CT is acquired, (ii) Second, Using deformable image registration, we then quantify the regional lung volume change, or tissue density changes between exhale and inhale. (iii) This information is used to model regional air-volume changes in the lung, a surrogate for ventilation. What’s exciting about CT-ventilation is that it offers the potential of high accessibility, high resolution and requiring no extra-cost in scan time or imaging dose (<10 minutes of image processing!)
- Recent planning studies have shown that that CT-ventilation has the potential to reduce healthy lung dose by 2-5 Gy (modality-dependent). A challenge is that lung function can change during treatment; particularly for patients undergoing longer treatments – due to tumor growth or regression. By combining CT-ventilation with in-room 4D cone beam CT, we could achieve adaptive functional image guidance to maximise the dose-savings for healthy lung.
- So what are we doing to validate this new technology? We are correlating baseline CT-ventilation, either 4D-CT breath hold CT or 4D cone beam CT, to other modalities including Technegas ventilation SPECT, respiratory gated Galligas PET as well as pulmonary function tests. This is all patient imaging data, some resulting from QA studies collecting 4D imaging, others are prospective clinical trials. You may be wondering why, in some cases I’m comparing a modality against itself; this is referring to multiple timepoint data, and I’ll get onto that shortly.
- -At Nepean hospital in Sydney, Dr Fiona Hegi-Johnson is also looking at ventilation / perfusion V-SPECT for stereotactic lung patients. Technegas SPECT is similar to PET-Galligas except that it uses technetium instead of gallium. - The main novelty of this study is in collecting multiple 4D-cone beam CTs for every treatment fraction; we are in the progress of correlating changes in CBCT VIs during treatment, with changes in pre/post treatment SPECT imaging.
- -In this example we’re comparing CT-ventilation to Technegas V-SPECT. Technegas SPECT is similar to PET-Galligas except that it uses technetium instead of gallium.
- - We investigated 19 locally advanced NSCLC patients undergoing lung cancer radiation therapy. Each patient received multiple 4D cone beam scans over 4-6 weeks of treatment. We collated 215 4D cone beam ventilation images (or Vis) including 78 intrafraction pairs (referring to pre/post fraction 4D-CBCTs on the same day), and 56 interfraction pairs (comparing the first week of treatment to each of Weeks 2, 4 and 6). We determine if..
- Here are the interfraction ventilation images. For each patient we have four scans; Week 1, 2, 4 and 6. All the VIs were deformably registered to the corresponding planning CT, cropped to a common region of imaged lung, and normalized based on the 10th and 90th functional percentiles of contralateral lung. What we see is that patients show all sorts of different kinds of functional changes; some patients exhibit a ventilation increase, another shows a decrease, one shows what looks like a transient change. Yet others seems stable, and others quite variable.
- The same patient can also exhibit varying levels of breathing effort from scan-to-scan, leading to varying levels of deformation. To account for this, we normalize ventilation images by the relatively `healthy’ lung which is contralateral or opposite to the primary tumour..
- So what does this all mean? - The take-home messages are that: Patients can exhibit both +ve and –ve ventilation changes throughout treatment highlighting the need for adaptive functional avoidance using techniques such as 4D cone beam ventilation. At this stage, 4D-cone beam ventilation could be useful for detecting large changes in function. About ½ of interfraction ventilation variability may be due to mis-registration of changing anatomy. With better 4D-cone beam image quality, and more reliable registration of large anatomic changes, perhaps the sensitivity of this technique can be improved.