Lung cancer is the leading cause of cancer mortality worldwide with non-small cell lung
cancers (NSCLC) accounting for approximately 85% of all lung cancers diagnosed. Surgery,
immunotherapy, chemotherapy, targeted therapy, and radiation therapy may be used alone or in
combination for lung cancer treatment depending on the cancer staging. Radiation therapy is a
primary component of lung cancer treatment with 77% of lung cancer patients having
evidence-based indication for external beam radiotherapy at some point through their
treatment journey.
Radiotherapy is associated with radiation-induced toxicities such as radiation pneumonitis
and fibrosis that adversely impacts a patient's quality of life and limits the dose that can
be safely delivered. Radiation pneumonitis is the inflammation of lung tissue with symptoms
such shortness of breath, cough, and fever, manifesting 4 to 12 weeks following the
completion of a radiotherapy course. Symptomatic pneumonitis, defined as grade 2 or higher,
has an overall incidence of 29.8% with conventional radiotherapy.
Current radiotherapy treatment planning assumes that lung function is homogeneous throughout
the organ and does not account for regional differences in lung function. These regional
differences may arise from the cancer growth itself via tumour compression of lung
structures, or from pre-existing, often smoking-induced factors which can also impact the
lung health of cancer patients. Pre-existing conditions that can impact lung health include
chronic obstructive pulmonary disease (COPD), fibrosis and thickening of the lymphatics in
the lung, emphysema, asbestosis and partial lung collapse or side effects from previous
cancer treatment such as fluid in the lungs, inflammation and pneumonitis, or even partial
lung removal. In order to give patients the best quality of life after radiotherapy, it is
crucial to preserve as much healthy lung as possible during the course of treatment.
Functional lung avoidance treatment planning for radiotherapy has the potential to reduce
pulmonary toxicity by minimising irradiation of healthy lung tissue based on functional maps
of the lung. Nuclear medicine functional lung images have been previously incorporated into
radiotherapy treatment planning for the purposes of reducing mean doses to functional lung
and hence radiation-induced toxicity. Normal tissue complication probability (NTCP) models
have also predicted an overall reduction of 6% and 3% for grade 2+ and 3+ radiation
pneumonitis, respectively, with functional planning while honouring dose constraints to the
target and organs at risk.
Nuclear medicine scans are expensive, time consuming and not available in all institutions.
CT ventilation imaging has been developed as a cheaper and more accessible alternative to
nuclear medicine for mapping the healthy areas of the lung as part of routine pre-treatment
CT acquisition.
The key steps in CT ventilation imaging are:
1. Acquire CT images of the lung at exhale and inhale states, using 4DCT or BHCT. 4DCT is a
technology that forms the standard of care for lung cancer radiation therapy throughout
the world. 4DCT acquires a set of 8 to 10 CT images from peak inhale to peak exhale
showing the motion of the lungs as a patient breathes while BHCT acquires static
end-inspiration and end-expiration images of the lung as the patient holds their breath
for around 10 seconds.
2. Deformable image registration is used to determine a spatial mapping (deformation map)
between the different CT images (from peak inhale to peak exhale).
3. Application of a ventilation metric to quantify high and low functioning lung which
involves quantitative analysis based on the information from the deformable image
registration.
The resulting ventilation image is superimposed directly onto the anatomic image, providing
an added dimension of functional information which is easy to understand and can be of direct
benefit to radiotherapy treatment planning.
There are existing commercial products, CT Lung Ventilation Analysis Software and XV
Technology Lung Ventilation Analysis Software (4DMedical Limited, Melbourne), for producing
ventilation maps from CT scans and fluoroscopy imaging sequences. Unlike CT which produces
one or more 3D images, fluoroscopy produces a fast sequence of 2D X-ray images. These images
are usually acquired by dedicated C-arm scanners in radiology departments or surgical
theatres and has not previously been assessed with relation to radiotherapy. Cine-fluoroscope
sequences capturing at least one complete, continuous breath are acquired at five distinct
angles across the chest during spontaneous breathing and a lung motion field is reconstructed
to produce a ventilation map. As Therapeutic Goods Administration (TGA)-approved modalities
of lung ventilation assessment, comparison of these techniques offers valuable insights into
their robustness.
Functional lung avoidance treatment planning based on pre-treatment ventilation images can be
compromised by interfraction (week-to-week) ventilation changes during treatment, such as
those resulting from tumour regression. 4DCBCT allows convenient acquisition of anatomical
images on the radiotherapy system to ensure accurate patient positioning prior to treatment
delivery. Existing techniques developed for computed tomography ventilation imaging (CTVI)
can be applied to 4DCBCT images to produce ventilation images that can be used to adapt
treatment planning and minimise irradiation of healthy lung. However, 4DCBCT suffers from
poor image quality and the physiological accuracy of 4DCBCT-based ventilation images remains
to be assessed.
Previous studies have validated the physiological accuracy of CT ventilation imaging against
the existing clinical gold standard Galligas PET ventilation imaging. These studies found
strong correlation at the lobar level (several cm) and moderate positive correlations at the
regional level (2 to 5 mm).
Clinical evidence for efficacy of the functional lung avoidance technique using CT
ventilation images is beginning to be gathered. Vinogradskiy et al. performed a Phase II
clinical trial, comparing radiation pneumonitis rates in a cohort treated with functional
lung avoidance against historic controls. The authors found the rate of grade ≥2 radiation
pneumonitis to be 14.9% of patients in the functional lung avoidance cohort, compared to an
historical rate of 25%, reporting a positive trial outcome with a power of 80% and
significance of 0.05. Using Xenon-enhanced CT ventilation, Huang et al performed a Phase II
study comparing functional lung avoidance to historic controls. With a 17% rate of grade ≥2
radiation pneumonitis compared to the historical control of 30%, the authors also concluded a
positive trial outcome with a power of 80% and significance of 0.05.
Despite this encouraging clinical evidence, there are still gaps in knowledge on assessing
the best technical implementation of CT ventilation imaging, quality assurance of the process
and the clinically acceptable threshold of accuracy required. To this end, the focus of our
proposed clinical trial is to determine the best implementation of fluoroscopic and CT
ventilation imaging. To achieve this, we will assess the physiological accuracy of different
CT and fluoroscopy-based ventilation imaging techniques, using Galligas as a ground truth for
true ventilation. Secondary aims of our study are to assess the dosimetric variation in
functional avoidance radiotherapy plans produced using these ventilation imaging techniques,
establish a quality assurance procedure for functional lung avoidance RT and to evaluate the
clinical acceptable thresholds for accuracy of the method.
To compare the impact of functional lung avoidance planning on the radiation therapy
treatment workflow, functional avoidance radiation therapy treatment plans will be created in
addition to standard of care treatment plans and the difference in planning time quantified.
The predicted reduction in radiation pneumonitis rates will be quantified using the method of
Faught et al. and the predicted reduction in number of hospitalisations and medical costs
will be calculated.
CT perfusion imaging methods have been recently developed to allow the computation of
perfusion information from non-contrast inhale/exhale CT image pairs. CTPI is based on the
principle that there is a mass change within the lung during the respiratory cycle due to
changes in blood volume. Perfusion is derived from this mass change which is calculated at
the voxel resolution using intensity-based material density estimates and spatial mapping
between different CT images. Computed tomography perfusion imaging (CTPI) is a nascent
technology that requires further validation of its accuracy. However, there is currently a
lack of paired BHCT-perfusion datasets for validation. To date, the only validation study
conducted by Castillo et al. reported a median correlation of 0.57 between CTPI and SPECT
perfusion imaging.
The University of Sydney and Northern Sydney Local Health District have recently installed a
new Total Body PET scanner at the Royal North Shore Hospital which can acquire scans in a
single bed position with higher sensitivity, faster acquisition time, and longer scan length.
The higher sensitivity of this new device allows faster imaging, including 4D (time series)
images, giving more detailed ventilation information. To support this, an additional low-dose
4DCT scan, across the lungs only rather than the full scan length, will be acquired in
pre-treatment imaging for the purpose of investigating 4D attenuation correction.
Choice of comparators Control.
- - Galligas PET ventilation images will be acquired for each participant and used as the
ground truth.
Galligas PET is an established modality for acquiring ventilation images
with higher spatial resolution than SPECT imaging.
- - Tc-99m MAA SPECT images will be acquired for each participant and used as the ground
truth for perfusion.
99mTc-MAA SPECT is an established modality for acquiring perfusion
images.
- - 4D PET images reconstructed with 3D attenuation CT.
- - Standard of care anatomical based treatment plans will be generated for each
participant.
Comparators.
- - Ventilation images derived from BHCT, fluoroscopy, 4DCBCT, and 4DCT scans acquired from
each participant.
X-ray-based ventilation imaging has been developed as a cost-effective
and accessible alternative for visualising lung ventilation.
- - Perfusion images derived from BHCT scans acquired from each patient.
CT perfusion
imaging is under development as a cost-effective and accessible alternative for
visualising lung perfusion.
- - 4D PET images reconstructed with 4D attenuation CT.
4D attenuation is proposed to give
more accurate PET imaging in regions where breathing motion is expected.
- - Functional based radiotherapy plans created using the ventilation images, in terms of
predicted radiation pneumonitis rates.