Interfractional reproducibility of lung tumor location using various methods of respiratory motion mitigation

Machine-learning-based prediction of the effectiveness of the delivered dose by exhale-gated radiotherapy for locally advanced lung cancer: The additional value of geometric over dosimetric parameters alone

Purpose

This study aimed to assess interfraction stability of the delivered dose distribution by exhale-gated volumetric modulated arc therapy (VMAT) or intensity-modulated arc therapy (IMAT) for lung cancer and to determine dominant prognostic dosimetric and geometric factors.

Methods

Clinical target volume (CTV_Plan) from the planning CT was deformed to the exhale-gated daily CBCT scans to determine CTV_i, treated by the respective dose fraction. The equivalent uniform dose of the CTV_i was determined by the power law (gEUD_i) and cell survival model (EUD_iSF) as effectiveness measure for the delivered dose distribution. The following prognostic factors were analyzed: (I) minimum dose within the CTV_i (D_{min_i}), (II) Hausdorff distance (HDD_i) between CTV_i and CTV_Plan, (III) doses and deformations at the point in CTV_Plan at which the global minimum dose over all fractions per patient occurs (PD_{min_global_i}), and (IV) deformations at the point over all CTV_i margins per patient with the largest Hausdorff distance (HDPw_orst). Prognostic value and generalizability of the prognostic factors were examined using cross-validated random forest or multilayer perceptron neural network (MLP) classifiers. Dose accumulation was performed using back deformation of the dose distribution from CTV_i to CTV_Plan.

Results

Altogether, 218 dose fractions (10 patients) were evaluated. There was a significant interpatient heterogeneity between the distributions of the normalized gEUD_i values (p<0.0001, Kruskal-Wallis tests). Accumulated gEUD over all fractions per patient was 1.004-1.023 times of the prescribed dose. Accumulation led to tolerance of ~20% of fractions with gEUD_i <93% of the prescribed dose. Normalized D_min >60% was associated with predicted gEUD values above 95%. D_min had the highest importance for predicting the gEUD over all analyzed prognostic parameters by out-of-bag loss reduction using the random forest procedure. Cross-validated random forest classifier based on D_min as the sole input had the largest Pearson correlation coefficient (R=0.897) in comparison to classifiers using additional input variables. The neural network performed better than the random forest classifier, and the gEUD values predicted by the MLP classifier with D_min as the sole input were correlated with the gEUD values characterized by R=0.933 (95% CI, 0.913-0.948). The performance of the full MLP model with all geometric input parameters was slightly better (R=0.952) than that based on D_min (p=0.0034, Z-test).

Conclusion

Accumulated dose distributions over the treatment series were robust against interfraction CTV deformations using exhale gating and online image guidance. D_min was the most important parameter for gEUD prediction for a single fraction. All other parameters did not lead to a markedly improved generalizable prediction. Dosimetric information, especially location and value of D_min within the CTV _i , are vital information for image-guided radiation treatment.

Dose assessment for patients with stage I non-small cell lung cancer receiving passive scattering carbon-ion radiotherapy using daily computed tomographic images: A prospective study

BACKGROUND AND PURPOSE

This study aimed to assess dose distributions for stage I non-small cell lung cancer (NSCLC) with passive scattering carbon-ion radiotherapy (C-ion RT) using daily computed tomography (CT) images.

MATERIALS AND METHODS

We enrolled 10 patients with stage I NSCLC and acquired a total of 40 pre-fractional CT image series under the same settings as the planning CT images. These CT images were registered with planning CT images for dose evaluation using both bone matching (BM) and tumor matching (TM). Using deformable image registration, we generated accumulated doses. Moreover, the volumetric dose parameters were compared in terms of tumor coverage and lung exposure and statistical analyses were performed.

RESULTS

Overall, 25% of 40 fractional dose distributions were unacceptable with BM, compared with 2.5% with TM (P < 0.001). Using BM, three patients' accumulated dose distributions were unacceptable; however, all were satisfactory with TM (P < 0.001). No differences were observed in water-equivalent path length (WEL). The required margins in patients with poor dose distribution were 5.9 and 4.4 mm for BM and TM, respectively.

CONCLUSIONS

This study establishes that CT image-based TM is robust compared with conventional BM for both daily and accumulated dose distributions. The effects of changes in WEL seem to be limited. Hence, daily CT alignment is recommended for patients with stage I NSCLC receiving C-ion RT.

The need for, and implementation of, image guidance in radiation therapy

The introduction of image guidance in radiation therapy and its subsequent innovations have revolutionised the delivery of cancer treatment. Modern imaging systems can supplement and often replace the historical practice of relying on external landmarks and laser alignment systems. Rather than depending on markings on the patient's skin, image-guided radiation therapy (IGRT), using techniques such as computed tomography (CT), cone beam CT, MV on-board imaging (OBI), and kV OBI, allows the patient to be positioned based on the internal anatomy. These advances in technology have enabled more accurate delivery of radiation doses to anatomically complex and temporally changing tumour volumes, while simultaneously sparing surrounding healthy tissues. While these imaging modalities provide excellent bony anatomy image quality, magnetic resonance imaging (MRI) surpasses them in soft tissue image contrast for better visualisation and tracking of soft tissue tumours with no additional radiation dose to the patient. However, the introduction of MRI into a radiotherapy facility has a number of complications, including the influence of the magnetic field on the dose deposition, as well as the effects it can have on dosimetry systems. The development and introduction of these new IGRT techniques will be reviewed, and the benefits and disadvantages of each will be described.

Lung volume reproducibility under ABC control and self-sustained breath-holding

An Active Breathing Coordinator (ABC) can be employed to induce breath-holds during CT imaging and radiotherapy of lung, breast and liver cancer, and recently during lung cancer MRI. The apparatus measures and controls respiratory volume, hence subject lung volume reproducibility is its principal measure of effectiveness. To assess ABC control quality, the intra-session reproducibility of ABC-induced lung volumes was evaluated and compared with that reached by applying the clinical standard of operator-guided self-sustained breath-holds on healthy volunteers during MRI. Inter-session reproducibility was investigated by repeating ABC-controlled breath-holds on a second visit. Additionally, lung volume agreement with ABC devices used with different imaging modalities in the same institution (MR, CT), or for a breast trial treatment, was assessed. Lung volumes were derived from three-dimensional (3D) T1-weighted MRI datasets by three observers employing semiautomatic lung delineation on a radiotherapy treatment planning system. Inter-observer variability was less than 6% of the delineated lung volumes. Lung volume agreement between the different conditions over all subjects was investigated using descriptive statistics. The ABC equipment dedicated for MR application exhibited good intra-session and inter-session lung volume reproducibility (1.8% and 3% lung volume variability on average, respectively). MR-assessed lung volumes were similar using different ABC equipment dedicated to MR, CT, or breast radiotherapy. Overall, lung volumes controlled by the same or different ABC devices agreed better than with self-controlled breath-holds, as suggested by the average ABC variation of 1.8% of the measured lung volumes (99 mL), compared to the 4.1% (226 mL) variability observed on average with self-sustained breath-holding.

High quality machine-robust image features: identification in nonsmall cell lung cancer computed tomography images

PURPOSE

For nonsmall cell lung cancer (NSCLC) patients, quantitative image features extracted from computed tomography (CT) images can be used to improve tumor diagnosis, staging, and response assessment. For these findings to be clinically applied, image features need to have high intra and intermachine reproducibility. The objective of this study is to identify CT image features that are reproducible, nonredundant, and informative across multiple machines.

METHODS

Noncontrast-enhanced, test-retest CT image pairs were obtained from 56 NSCLC patients imaged on three CT machines from two institutions. Two machines ("M1" and "M2") used cine 4D-CT and one machine ("M3") used breath-hold helical 3D-CT. Gross tumor volumes (GTVs) were semiautonomously segmented then pruned by removing voxels with CT numbers less than a prescribed Hounsfield unit (HU) cutoff. Three hundred and twenty eight quantitative image features were extracted from each pruned GTV based on its geometry, intensity histogram, absolute gradient image, co-occurrence matrix, and run-length matrix. For each machine, features with concordance correlation coefficient values greater than 0.90 were considered reproducible. The Dice similarity coefficient (DSC) and the Jaccard index (JI) were used to quantify reproducible feature set agreement between machines. Multimachine reproducible feature sets were created by taking the intersection of individual machine reproducible feature sets. Redundant features were removed through hierarchical clustering based on the average correlation between features across multiple machines.

RESULTS

For all image types, GTV pruning was found to negatively affect reproducibility (reported results use no HU cutoff). The reproducible feature percentage was highest for average images (M1 = 90.5%, M2 = 94.5%, M1∩M2 = 86.3%), intermediate for end-exhale images (M1 = 75.0%, M2 = 71.0%, M1∩M2 = 52.1%), and lowest for breath-hold images (M3 = 61.0%). Between M1 and M2, the reproducible feature sets generated from end-exhale images were relatively machine-sensitive (DSC = 0.71, JI = 0.55), and the reproducible feature sets generated from average images were relatively machine-insensitive (DSC = 0.90, JI = 0.87). Histograms of feature pair correlation distances indicated that feature redundancy was machine-sensitive and image type sensitive. After hierarchical clustering, 38 features, 28 features, and 33 features were found to be reproducible and nonredundant for M1∩M2 (average images), M1∩M2 (end-exhale images), and M3, respectively. When blinded to the presence of test-retest images, hierarchical clustering showed that the selected features were informative by correctly pairing 55 out of 56 test-retest images using only their reproducible, nonredundant feature set values.

CONCLUSIONS

Image feature reproducibility and redundancy depended on both the CT machine and the CT image type. For each image type, the authors found a set of cross-machine reproducible, nonredundant, and informative image features that would be useful for future image-based models. Compared to end-exhale 4D-CT and breath-hold 3D-CT, average 4D-CT derived image features showed superior multimachine reproducibility and are the best candidates for clinical correlation.

A serial 4DCT study to quantify range variations in charged particle radiotherapy of thoracic cancers

Weekly serial 4DCT scans were acquired under free breathing conditions to assess water-equivalent path length (WEL) variations due to both intrafractional and interfractional changes in tissue thickness and density and to calculate proton dose distributions resulting from anatomical variations observed in serial 4DCT. A template of region of interests (ROIs) was defined on the anterior-posterior (AP) beam's eye view, and WEL measurements were made over these ROIs to quantify chest wall thickness variations. Interfractional proton dose distributions were calculated to assess changes in the expected dose distributions caused by range variations. Mean intrafractional chest wall WEL changes during respiration varied by: -4.1 mm (<-10.2 mm), -3.6 mm (<-7.1 mm), -3.2 mm (<-5.6 mm) and -2.5 mm (<-5.1 mm) during respiration in the ITV, upper, middle and lower lung regions, respectively. The mean interfractional chest wall WEL variation at Week 6 decreased by -4.0 mm (<-8.6 mm), -9.1 mm (<-17.9 mm), -9.4 mm (<-25.3 mm) and -4.5 mm (<-15.6 mm) in the ITV, upper, middle and lower lung regions, respectively. The variations were decomposed into anterior and posterior chest wall thickness changes. Dose overshoot beyond the target was observed when the initial boli was applied throughout the treatment course. This overshoot is due to chest wall thickness variations and target positional variations. The radiological path length can vary significantly during respiration as well as over the course of several weeks of charged particle therapy. Intrafractional/interfractional chest wall thickness changes can be a significant source of range variation in treatment of lung tumors with charged particle beams, resulting in dose distribution perturbations from the initial plan. Consideration of these range variations should be made in choosing the therapeutic charged particle beam range.

Potential benefits and pitfalls of respiratory-gated radiotherapy in the treatment of thoracic malignancy

AIM

Despite advances in radiotherapy delivery, the prognosis of lung cancer remains poor. Higher doses of radiation have been associated with improved outcomes but may result in higher toxicities. Respiratory gated radiotherapy (RGRT) has the potential to reduce pulmonary toxicity but there are significant limitations and pitfalls to its use. The aim of this article is to (i) describe the RGRT technique currently employed at Nepean and Westmead Hospitals; (ii) discuss the practical issues of implementing such a program; (iii) present the results of our RGRT program and (iv) review the potential uncertainties in using this technique and the methods we have used to overcome these.

METHODS

A retrospective review of all patients who had a 4D-computed tomography (4D-CT) scan was undertaken. Records from treatment planning systems were used to assess the prospective gating program.

RESULTS

Between September 2007 and June 2011, 53 patients at Nepean and 26 patients at Westmead Hospital underwent a 4D-CT. Between April and August 2011, 26 patients at Westmead Hospital underwent a prospective 4D-CT scan as treatment verification. Two of the 26 patients (7.7%) were found to have incomplete coverage of the planning target volume. Both patients underwent respiratory re-coaching, alleviating the need for replanning.

CONCLUSION

RGRT may reduce doses to organs at risk with the potential for dose escalation. However its implementation requires significant staff training, treatment time and resources. Treatment verification with image guided radiation therapy are essential for safe delivery.

Reproducibility of image quality for moving objects using respiratory-gated computed tomography: a study using a phantom model

To investigate the reproducibility of computed tomography (CT) imaging quality in respiratory-gated radiation treatment planning is essential in radiotherapy of movable tumors. Seven series of regular and six series of irregular respiratory motions were performed using a thorax dynamic phantom. For the regular respiratory motions, the respiratory cycle was changed from 2.5 to 4 s and the amplitude was changed from 4 to 10 mm. For the irregular respiratory motions, a cycle of 2.5 to 4 or an amplitude of 4 to 10 mm was added to the base data (i.e. 3.5-s cycle, 6-mm amplitude) every three cycles. Images of the object were acquired six times using respiratory-gated data acquisition. The volume of the object was calculated and the reproducibility of the volume was decided based on the variety. The registered image of the object was added and the reproducibility of the shape was decided based on the degree of overlap of objects. The variety in the volumes and shapes differed significantly as the respiratory cycle changed according to regular respiratory motions. In irregular respiratory motion, shape reproducibility was further inferior, and the percentage of overlap among the six images was 35.26% in the 2.5- and 3.5-s cycle mixed group. Amplitude changes did not produce significant differences in the variety of the volumes and shapes. Respiratory cycle changes reduced the reproducibility of the image quality in respiratory-gated CT.

Real-time 4-D radiotherapy for lung cancer

Respiratory motion considerably influences dose distribution, and thus clinical outcomes in radiotherapy for lung cancer. Breath holding, breath coaching, respiratory gating with external surrogates, and mathematical predicting models all have inevitable uncertainty due to the unpredictable variations of internal tumor motion. The amplitude of the same tumor can vary with standard deviations > 5 mm occurring in 23% of T1-2N0M0 non-small cell lung cancers. Residual motion varied 1-6 mm (95th percentile) for the 40% duty cycle of respiratory gating with external surrogates. The 4-D computed tomography is vulnerable to problems relating to the external surrogates. Real-time 4-D radiotherapy (4DRT), where the temporal changes in anatomy during the delivery of radiotherapy are explicitly considered in real time, is emerging as a new method to reduce these known sources of uncertainty. Fluoroscopic, real-time tumor-tracking technology using internal fiducial markers near the tumor has ± 2 mm accuracy, and has achieved promising clinical results when used with X-ray therapy. Instantaneous irradiation based on real-time verification of internal fiducial markers is considered the minimal requisite for real-time 4DRT of lung cancers at present. Real-time tracking radiotherapy using gamma rays from positron emitters in tumors is in the preclinical research stage, but has been successful in experiments in small animals. Real-time tumor tracking via spot-scanning proton beam therapy has the capability to cure large lung cancers in motion, and is expected to be the next-generation real-time 4DRT.

Multiple breath-hold CBCT for online image guided radiotherapy of lung tumors: simulation with a dynamic phantom and first patient data

BACKGROUND AND PURPOSE

Computer controlled breath-hold effectively reduces organ motion for image-guided precision radiotherapy of lung tumors. However, the acquisition time of 3D cone-beam-CT (CBCT) exceeds maximum breath-hold times. We have developed an approach enabling online verification using CBCT image acquisition with ABC®-based breath-hold.

METHODS

Patient CBCT images were acquired with ABC®-based repeat breath-hold. The clinical situation was also simulated with a Motion Phantom. Reconstruction of patient and phantom images with selection of free-breathing and breath-hold projections only was performed.

RESULTS

CBCT-imaging in repeat breath-hold resulted in a precisely spherical appearance of a tumor-mimicking structure in the phantom. A faint "ghost" structure (free-breathing phases) can be clearly discriminated. Mean percentage of patient breath-hold time was 66%. Reconstruction based on free-breathing-only shows blurring of both tumor and diaphragm, reconstruction based on breath-hold projections only resulted in sharp contours of the same structures. From the phantom experiments, a maximal repositioning error of 1mm in each direction can be estimated.

DISCUSSION AND CONCLUSION

CBCT during repetitive breath hold provides reliable soft-tissue-based positioning. Fast 3D-imaging during one breath-hold is currently under development and has the potential to accelerate clinical linac-based volume imaging.