skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Interrater Reliability of the Categorization of Late Radiographic Changes After Lung Stereotactic Body Radiation Therapy

Abstract

Purpose: Radiographic changes after lung stereotactic body radiation therapy (SBRT) have been categorized into 4 groups: modified conventional pattern (A), mass-like fibrosis; (B), scar-like fibrosis (C), and no evidence of increased density (D). The purpose of this study was to assess the interrater reliability of this categorization system in patients with early-stage non-small cell lung cancer (NSCLC). Methods and Materials: Seventy-seven patients were included in this study, all treated with SBRT for early-stage (T1/2) NSCLC at a single institution, with a minimum follow-up of 6 months. Six experienced clinicians familiar with post-SBRT radiographic changes scored the serial posttreatment CT images independently in a blinded fashion. The proportion of patients categorized as A, B, C, or D at each interval was determined. Krippendorff's alpha (KA), Multirater kappa (M-kappa), and Gwet's AC1 (AC1) scores were used to establish interrater reliability. A leave-one-out analysis was performed to demonstrate the variability among raters. Interrater agreement of the first and last 20 patients scored was calculated to explore whether a training effect existed. Results: The number of ratings ranged from 450 at 6 months to 84 at 48 months of follow-up. The proportion of patients in each category was as follows: A, 45%; B, 16%; C, 13%; andmore » D, 26%. KA and M-kappa ranged from 0.17 to 0.34. AC1 measure range was 0.22 to 0.48. KA increased from 0.24 to 0.36 at 12 months with training. The percent agreement for pattern A peaked at 12 month with a 54% chance of having >50% raters in agreement and decreased over time, whereas that for patterns B and C increased over time to a maximum of 20% and 22%, respectively. Conclusion: This post-SBRT radiographic change categorization system has modest interrater agreement, and there is a suggestion of a training effect. Patterns of fibrosis evolve after SBRT and alternative categorization systems should be evaluated.« less

Authors:
 [1];  [1]; ;  [1];  [2];  [3]; ; ; ; ;  [1]
  1. Department of Radiation Oncology, Princess Margaret Cancer Centre, Toronto, ON (Canada)
  2. Department of Radiology, University Health Network, Toronto, Ontario (Canada)
  3. Department of Biostatistics, Princess Margaret Cancer Centre, Toronto, Ontario (Canada)
Publication Date:
OSTI Identifier:
22420396
Resource Type:
Journal Article
Resource Relation:
Journal Name: International Journal of Radiation Oncology, Biology and Physics; Journal Volume: 89; Journal Issue: 5; Other Information: Copyright (c) 2014 Elsevier Science B.V., Amsterdam, The Netherlands, All rights reserved.; Country of input: International Atomic Energy Agency (IAEA)
Country of Publication:
United States
Language:
English
Subject:
62 RADIOLOGY AND NUCLEAR MEDICINE; CAT SCANNING; FIBROSIS; LUNGS; NEOPLASMS; PATIENTS; RADIOTHERAPY

Citation Formats

Faruqi, Salman, Giuliani, Meredith E., E-mail: meredith.giuliani@rmp.uhn.on.ca, Raziee, Hamid, Yap, Mei Ling, Roberts, Heidi, Le, Lisa W., Brade, Anthony, Cho, John, Sun, Alexander, Bezjak, Andrea, and Hope, Andrew J. Interrater Reliability of the Categorization of Late Radiographic Changes After Lung Stereotactic Body Radiation Therapy. United States: N. p., 2014. Web. doi:10.1016/J.IJROBP.2014.04.042.
Faruqi, Salman, Giuliani, Meredith E., E-mail: meredith.giuliani@rmp.uhn.on.ca, Raziee, Hamid, Yap, Mei Ling, Roberts, Heidi, Le, Lisa W., Brade, Anthony, Cho, John, Sun, Alexander, Bezjak, Andrea, & Hope, Andrew J. Interrater Reliability of the Categorization of Late Radiographic Changes After Lung Stereotactic Body Radiation Therapy. United States. doi:10.1016/J.IJROBP.2014.04.042.
Faruqi, Salman, Giuliani, Meredith E., E-mail: meredith.giuliani@rmp.uhn.on.ca, Raziee, Hamid, Yap, Mei Ling, Roberts, Heidi, Le, Lisa W., Brade, Anthony, Cho, John, Sun, Alexander, Bezjak, Andrea, and Hope, Andrew J. Fri . "Interrater Reliability of the Categorization of Late Radiographic Changes After Lung Stereotactic Body Radiation Therapy". United States. doi:10.1016/J.IJROBP.2014.04.042.
@article{osti_22420396,
title = {Interrater Reliability of the Categorization of Late Radiographic Changes After Lung Stereotactic Body Radiation Therapy},
author = {Faruqi, Salman and Giuliani, Meredith E., E-mail: meredith.giuliani@rmp.uhn.on.ca and Raziee, Hamid and Yap, Mei Ling and Roberts, Heidi and Le, Lisa W. and Brade, Anthony and Cho, John and Sun, Alexander and Bezjak, Andrea and Hope, Andrew J.},
abstractNote = {Purpose: Radiographic changes after lung stereotactic body radiation therapy (SBRT) have been categorized into 4 groups: modified conventional pattern (A), mass-like fibrosis; (B), scar-like fibrosis (C), and no evidence of increased density (D). The purpose of this study was to assess the interrater reliability of this categorization system in patients with early-stage non-small cell lung cancer (NSCLC). Methods and Materials: Seventy-seven patients were included in this study, all treated with SBRT for early-stage (T1/2) NSCLC at a single institution, with a minimum follow-up of 6 months. Six experienced clinicians familiar with post-SBRT radiographic changes scored the serial posttreatment CT images independently in a blinded fashion. The proportion of patients categorized as A, B, C, or D at each interval was determined. Krippendorff's alpha (KA), Multirater kappa (M-kappa), and Gwet's AC1 (AC1) scores were used to establish interrater reliability. A leave-one-out analysis was performed to demonstrate the variability among raters. Interrater agreement of the first and last 20 patients scored was calculated to explore whether a training effect existed. Results: The number of ratings ranged from 450 at 6 months to 84 at 48 months of follow-up. The proportion of patients in each category was as follows: A, 45%; B, 16%; C, 13%; and D, 26%. KA and M-kappa ranged from 0.17 to 0.34. AC1 measure range was 0.22 to 0.48. KA increased from 0.24 to 0.36 at 12 months with training. The percent agreement for pattern A peaked at 12 month with a 54% chance of having >50% raters in agreement and decreased over time, whereas that for patterns B and C increased over time to a maximum of 20% and 22%, respectively. Conclusion: This post-SBRT radiographic change categorization system has modest interrater agreement, and there is a suggestion of a training effect. Patterns of fibrosis evolve after SBRT and alternative categorization systems should be evaluated.},
doi = {10.1016/J.IJROBP.2014.04.042},
journal = {International Journal of Radiation Oncology, Biology and Physics},
number = 5,
volume = 89,
place = {United States},
year = {Fri Aug 01 00:00:00 EDT 2014},
month = {Fri Aug 01 00:00:00 EDT 2014}
}
  • Purpose: To estimate the risk of radiation-induced changes in the lung before single-dose treatment (stereotactic body radiation therapy [SBRT]) of lung cancer, the quantitative dose-response and volume-response relations must be known. Methods and Materials: A total of 64 patients treated for non-small-cell lung cancer with single doses of 20-30 Gy were classified according to the occurrence or nonoccurrence of perifocal changes in the lung detected by CT. Patients without toxic events in the lung were required to have {>=}6 months of follow-up. The mean dose (D{sub mean}) in the ipsilateral lung and the volume receiving >7 or 10 Gy (V{submore » 7} and V{sub 10}, respectively) were used to calculate the dose-response and volume-response curves. The predictive value of additional variables was also investigated. Results: Of the 64 patients, 83% exhibited the selected endpoint. The tolerance values at a 50% probability of toxic events were 1.2 {+-} 0.7 Gy for the D{sub mean} and 5.8 {+-} 3.0% and 3.1 {+-} 2.0% for V{sub 7} and V{sub 10}, respectively. A nonsignificant shift to higher doses was seen for the dose-response curve for the upper compared with the lower part of the lung. Conclusion: The D{sub mean}, V{sub 7}, and V{sub 10} can be used to predict the risk of lung toxicity after SBRT treatment of non-small-cell lung cancer. Because of the lack of patients with low prescribed doses, however, the related uncertainty of this prediction is still relatively large. The D{sub mean}, V{sub 7}, and V{sub 10} are equally well suited. The additional investigated variables did not provide significant advantages. The lower part of the lung appears to be more radiosensitive than the upper.« less
  • Purpose: Stereotactic body radiation therapy (SBRT) is becoming the standard of care for early stage nonoperable lung cancers. Accurate dose–response modeling is challenging for SBRT because of the decreased number of clinical toxicity events. As a surrogate for a clinical toxicity endpoint, studies have proposed to use radiographic changes in follow up computed tomography (CT) scans to evaluate lung SBRT normal tissue effects. The purpose of the current study was to use local fibrotic lung regions to spatially and dosimetrically evaluate lung changes in patients that underwent SBRT.Methods: Forty seven SBRT patients treated at our institution from 2003 to 2009more » were used for the current study. Our patient cohort had a total of 148 follow up CT scans ranging from 3 to 48 months post-therapy. Post-treatment scans were binned into intervals of 3, 6, 12, 18, 24, 30, and 36 months after the completion of treatment. Deformable image registration was used to align the follow up CT scans with the pretreatment CT and dose distribution. Areas of visible fibrotic changes were contoured. The centroid of each gross tumor volume (GTV) and contoured fibrosis volume was calculated and the fibrosis volume location and movement (magnitude and direction) relative to the GTV and 30 Gy isodose centroid were analyzed. To perform a dose–response analysis, each voxel in the fibrosis volume was sorted into 10 Gy dose bins and the average CT number value for each dose bin was calculated. Dose–response curves were generated by plotting the CT number as a function of dose bin and time posttherapy.Results: Both fibrosis and GTV centroids were concentrated in the upper third of the lung. The average radial movement of fibrosis centroids relative to the GTV centroids was 2.6 cm with movement greater than 5 cm occurring in 11% of patients. Evaluating dose–response curves revealed an overall trend of increasing CT number as a function of dose. The authors observed a CT number plateau at doses ranging from 30 to 50 Gy for the 3, 6, and 12 months posttherapy time points. There was no evident plateau for the dose–response curves generated using data from the 18, 24, 30, and 36 months posttherapy time points.Conclusions: Regions of local fibrotic lung changes in patients that underwent SBRT were evaluated spatially and dosimetrically. The authors found that the average fibrosis movement was 2.6 cm with movement greater than 5 cm possible. Evaluating dose–response curves revealed an overall trend of increasing CT number as a function of dose. Furthermore, our dose–response data also suggest that one of the possible explanations of the CT number plateau effect may be the time posttherapy of the acquired data. Understanding normal tissue dose–response is important for reducing toxicity after SBRT, especially in cases where larger tumors are treated. The methods presented in the current work build on prior quantitative studies and further enhance the understanding of normal lung dose–response after SBRT.« less
  • Purpose: To evaluate the influence of tumor size, prescription dose, and dose to the lungs on posttreatment pulmonary function test (PFT) changes after stereotactic body radiation therapy (SBRT) for early-stage non-small cell lung cancer (NSCLC). Methods and Materials: The analysis is based on 191 patients treated at 5 international institutions: inclusion criteria were availability of pre- and post-SBRT PFTs and dose-volume histograms of the lung and planning target volume (PTV); patients treated with more than 1 SBRT course were excluded. Correlation between early (1-6 months, median 3 months) and late (7-24 months, median 12 months) PFT changes and tumor size, planning targetmore » volume (PTV) dose, and lung doses was assessed using linear regression analysis, receiver operating characteristics analysis, and Lyman's normal tissue complication probability model. The PTV doses were converted to biologically effective doses and lung doses to 2 Gy equivalent doses before correlation analyses. Results: Up to 6 months after SBRT, forced expiratory volume in 1 second and carbon monoxide diffusion capacity changed by −1.4% (95% confidence interval [CI], −3.4% to 0) and −7.6% (95% CI, −10.2% to −3.4%) compared with pretreatment values, respectively. A modest decrease in PFTs was observed 7-24 months after SBRT, with changes of −8.1% (95% CI, −13.3% to −5.3%) and −12.4% (95% CI, −15.5% to −6.9%), respectively. Using linear regression analysis, receiver operating characteristic analysis, and normal tissue complication probability modeling, all evaluated parameters of tumor size, PTV dose, mean lung dose, and absolute and relative volumes of the lung exposed to minimum doses of 5-70 Gy were not correlated with early and late PFT changes. Subgroup analysis based on pre-SBRT PFTs (greater or equal and less than median) did not identify any dose-effect relationship. Conclusions: This study failed to demonstrate a significant dose-effect relationship for changes of pulmonary function after SBRT for early-stage non-small cell lung cancer.« less
  • Purpose: The purpose of this study was to analyze the computed tomographic (CT) appearance of radiation injury to the lung and clinical symptoms after stereotactic body radiation therapy (SBRT) and evaluate the difference by the presence of pulmonary emphysema (PE) for small lung cancers. Methods and Materials: In this analysis, 45 patients with 52 primary or metastatic lung cancers were enrolled. We evaluated the CT appearance of acute radiation pneumonitis (within 6 months) and radiation fibrosis (after 6 months) after SBRT. Clinical symptoms were evaluated by Common Terminology Criteria for Adverse Events, version 3.0. We also evaluated the relationship betweenmore » CT appearance, clinical symptoms, and PE. Results: CT appearance of acute radiation pneumonitis was classified as follows: (1) diffuse consolidation, 38.5%; (2) patchy consolidation and ground-glass opacities (GGO), 15.4%; (3) diffuse GGO, 11.5%; (4) patchy GGO, 2.0%; (5) no evidence of increasing density, 32.6%. CT appearance of radiation fibrosis was classified as follows: (1) modified conventional pattern, 61.5%; (2) mass-like pattern, 17.3%; (3) scar-like pattern, 21.2%. Patients who were diagnosed with more than Grade 2 pneumonitis showed significantly less no evidence of increased density pattern and scar-like pattern than any other pattern (p = 0.0314, 0.0297, respectively). Significantly, most of these patients with no evidence of increased density pattern and scar-like pattern had PE (p = 0.00038, 0.00044, respectively). Conclusion: Computed tomographic appearance after SBRT was classified into five patterns of acute radiation pneumonitis and three patterns of radiation fibrosis. Our results suggest that SBRT can be also safely performed even in patients with PE.« less
  • Purpose: To investigate the serial changes of tumor hypoxia in response to single high-dose irradiation by various clinical and preclinical methods to propose an optimal fractionation schedule for stereotactic ablative radiation therapy. Methods and Materials: Syngeneic Lewis lung carcinomas were grown either orthotopically or subcutaneously in C57BL/6 mice and irradiated with a single dose of 15 Gy to mimic stereotactic ablative radiation therapy used in the clinic. Serial [{sup 18}F]-misonidazole (F-MISO) positron emission tomography (PET) imaging, pimonidazole fluorescence-activated cell sorting analyses, hypoxia-responsive element-driven bioluminescence, and Hoechst 33342 perfusion were performed before irradiation (day −1), at 6 hours (day 0), and 2 (daymore » 2) and 6 (day 6) days after irradiation for both subcutaneous and orthotopic lung tumors. For F-MISO, the tumor/brain ratio was analyzed. Results: Hypoxic signals were too low to quantitate for orthotopic tumors using F-MISO PET or hypoxia-responsive element-driven bioluminescence imaging. In subcutaneous tumors, the maximum tumor/brain ratio was 2.87 ± 0.483 at day −1, 1.67 ± 0.116 at day 0, 2.92 ± 0.334 at day 2, and 2.13 ± 0.385 at day 6, indicating that tumor hypoxia was decreased immediately after irradiation and had returned to the pretreatment levels at day 2, followed by a slight decrease by day 6 after radiation. Pimonidazole analysis also revealed similar patterns. Using Hoechst 33342 vascular perfusion dye, CD31, and cleaved caspase 3 co-immunostaining, we found a rapid and transient vascular collapse, which might have resulted in poor intratumor perfusion of F-MISO PET tracer or pimonidazole delivered at day 0, leading to decreased hypoxic signals at day 0 by PET or pimonidazole analyses. Conclusions: We found tumor hypoxia levels decreased immediately after delivery of a single dose of 15 Gy and had returned to the pretreatment levels 2 days after irradiation and had decreased slightly by day 6. Our results indicate that single high-dose irradiation can produce a rapid, but reversible, vascular collapse in tumors.« less