Joo-Hyun Chung; Soo Jin Lee; Ji Young Kim; Hae Jin Park

doi:10.4143/crt.2025.1248

Articles

Page Path: HOME > J Korean Cancer Assoc > Ahead-of print articles > Article

Original Article Assessing Radiation Pneumonitis through Functional Lung Imaging: A Single Photon Emission Computed Tomography–Based Approach in Lung Cancer: Joo-Hyun Chung¹, Soo Jin Lee², Ji Young Kim², Hae Jin Park³; DOI: https://doi.org/10.4143/crt.2025.1248
Published online: February 19, 2026

¹Department of Radiation Oncology, Hanyang University Medical Center, Seoul, Korea

²Department of Nuclear Medicine, Hanyang University College of Medicine, Seoul, Korea

³Department of Radiation Oncology, Hanyang University College of Medicine, Seoul, Korea

Correspondence: Hae Jin Park, Department of Radiation Oncology, Hanyang University Hospital, Hanyang University College of Medicine, 222-1 Wangsimni-ro, Seongdong-gu, Seoul 04763, Korea
Tel: 82-2-2290-8616 E-mail: haejinpark@hanyang.ac.kr

^*Joo-Hyun Chung and Soo Jin Lee contributed equally to this work.

• Received: November 13, 2025 • Accepted: February 18, 2026

This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

842 Views
25 Download

Full Article

Download PDF

Abstract

Purpose
This study aimed to develop models to assess the risk of symptomatic radiation pneumonitis (SRP) (Common Terminology Criteria for Adverse Events ver. 4.03 grade ≥ 2) in lung cancer patients by utilizing single-photon emission computed tomography (SPECT) for functional lung volume identification and dosimetric analysis.
Materials and Methods
This retrospective study included 71 lung cancer patients who underwent SPECT before radiotherapy from 2018 to 2024. Perfusion and ventilation SPECT images were co-registered with planning computed tomography to define functional and anatomical lung volumes. Functional lung was defined as voxels with ≥ 20% of the maximum intensity on SPECT. Models to assess the risk of SRP were constructed using Cox regression and evaluated using corrected Akaike information criterion (AICc) and time-dependent receiver operating characteristic analysis.
Results
At a median follow-up of 16.8 months, 19 of 71 patients (26.8%) developed SRP. Factors significantly associated with SRP risk included planning target volume ≥ 150 mL, percentage of total perfusion-defined functional lung receiving ≥ 10 Gy (pVf₁₀) exceeding that of total anatomical lung receiving ≥ 10 Gy (V₁₀), percentage of total ventilation-defined lung receiving ≥ 10 Gy (vVf₁₀) ≥ 45%, and ipsilateral vVf₁₀ ≥ 60% (p=0.004, p=0.004, p=0.024, and p=0.007, respectively). Among the three models, the model incorporating additional ventilation-based parameters demonstrated the best performance (AICc, 85.81; area under the curve, 0.819).
Conclusion
SPECT-based dosimetric parameters derived from perfusion and ventilation are significantly associated with the risk of SRP. Incorporating SPECT may improve risk stratification and enable lung-sparing strategies.
Key words: Symptomatic radiation pneumonitis, Lung SPECT, Functional lung volume, Dosimetric analysis, Lung neoplasms

Introduction

Radiation pneumonitis (RP) is a potentially severe complication of thoracic radiotherapy in lung cancer. Although severe RP is uncommon, symptomatic radiation pneumonitis (SRP) occurs in 13% to 55% of patients [1-4]. Despite adhering to clinical constraints such as mean lung dose (MLD) and the percentage of lung volume receiving ≥ 20 Gy (V₂₀), SRP still occurred in 18.4% of cases [1]. Because conventional dose-volume histograms do not fully capture the heterogeneous nature of pulmonary function, a more functionally relevant approach to dosimetry is needed.

Functional imaging, such as single-photon emission computed tomography (SPECT), enables three-dimensional assessment of pulmonary function and the generation of a dose-function histogram when combined with radiation dose distribution [5]. Preferentially sparing normally functional lung and directing dose toward less functional regions may reduce the risk of SRP. However, its clinical implementation in routine radiotherapy planning has yet to be fully established. Recent studies have suggested that functional parameters derived from SPECT-integrated radiation planning predict the development of SRP more accurately than conventional dosimetric parameters [6-11]. However, most previous studies have examined either perfusion or ventilation SPECT alone, with only a few studies evaluating both modalities simultaneously. The clinical utility of incorporating functional SPECT data into radiotherapy planning remains uncertain, as prospective studies are limited and retrospective findings have been inconsistent.

Thus, this study aimed to assess the risk of SRP defined by Common Terminology Criteria for Adverse Events (CTCAE) ver. 4.03 ≥ grade 2 in lung cancer patients by incorporating both perfusion and ventilation SPECT to identify functional lung volume and analyze dosimetric parameters.

Materials and Methods

1. Patient cohort

Patients diagnosed with lung cancer, who underwent SPECT before radiotherapy from 2018 to 2024, were enrolled in this retrospective study. Patients without lung SPECT workups were excluded from the study. The median interval between lung SPECT and the start of radiotherapy was 6 days (range, 0 to 44 days). Of 77 patients, four were lost to follow-up and three patients who received radiotherapy more than twice were excluded. Ultimately, 71 patients (58 male and 13 female) were enrolled. Perfusion SPECT was performed on 64 patients, whereas ventilation SPECT was performed on 54. The median follow-up duration was 16.8 months (range, 1.0 to 61.9 months). The diagnosis of SRP was made by a board-certified radiologist based on radiologic findings within the irradiated field and the clinical requirement for corticosteroid use according to CTCAE ver. 4.03 grade ≥ 2 criteria.

2. SPECT/computed tomography protocol

All imaging data were acquired using a hybrid SPECT/computed tomography (CT) system (NM/CT 850, GE Medical Systems). Ventilation imaging was performed following inhalation of technetium-99m (Tc-99m)–labeled carbon nanoparticles, generated using a commercially available Technegas generator. Tc-99m was first concentrated to a volume of 0.1 mL and then heated within the generator to produce nano-sized radiolabeled carbon particles with an average size of 30-60 nm. The activity administered at the time of inhalation was approximately 20 MBq.

Ventilation images were acquired during free breathing using a dual-head camera with low-energy high-resolution collimators. The acquisition parameters included a photopeak energy of 140.5 keV with a 20% energy window, 60 projections over a full 360° rotation, 12 seconds per projection (10 seconds for the perfusion scan), and a matrix size of 128×128. Images were reconstructed using an iterative ordered subset expectation maximization algorithm with attenuation correction based on the co-registered CT scan.

The CT component was performed using the following parameters: 120 kV tube voltage, 10 mA tube current, 9.37 mm/rotation table speed, pitch of 0.938:1, 10 mm collimation, and a 512×512 matrix. CT images were reconstructed with a 1.25 mm slice thickness using an adaptive statistical iterative reconstruction algorithm (ASiR, GE Healthcare).

Perfusion SPECT images were obtained following the intravenous administration of 185 MBq of Tc-99m macroaggregated albumin, with the patient in a supine position. For patients undergoing both perfusion and ventilation imaging, both scans were acquired sequentially on the same day, and only one CT acquisition was performed. Perfusion SPECT was performed immediately after the ventilation scan using the same SPECT/CT system.

3. Treatment planning and radiotherapy dosimetry

Internal target volumes were delineated on the average phase of either four-dimensional computed tomography (4DCT) (stereotactic ablative body radiotherapy [SABR]) or 3-phase CT (conventional radiotherapy) by a board-certified radiation oncologist. SPECT images were co-registered to planning CT using deformable registration (MIM Maestro ver. 7.3.7, MIM Software Inc.) (Fig. 1). Anatomical lung volumes, along with functional and nonfunctional lung volumes from perfusion and/or ventilation SPECT, were delineated. Functional lungs were defined as voxels exceeding 20% of the maximum perfusion or ventilation SPECT value. To assess the robustness of the model performance to threshold selection, a sensitivity analysis using a uniform 25% threshold applied to all patients was additionally performed (S1 Fig.). Biologically equivalent doses for various dose-fractionation schedules for SABR were converted to 2 Gy fractions (EQD2) using the linear-quadratic model. Dosimetric parameters were extracted for the anatomical and functional lungs, including MLD, and the percentage of lung volume receiving ≥ 10 Gy (V₁₀) and ≥ 20 Gy (V₂₀). For functional lungs, perfusion-based (pMLD, percentage of volume of lung in perfusion scan receiving ≥ 10 Gy [pVf₁₀], and pVf₂₀) and ventilation-based (vMLD, vVf₁₀, and vVf₂₀) parameters were similarly calculated. The dosimetric parameters of the original anatomical-based and newly created functional-based approaches were compared.

4. Statistical analysis

Clinical and dosimetric parameters between the anatomical lung (standard parameters) and the functional lung for both perfusion and ventilation (functional parameters) were compared using paired t tests. Univariate analyses were conducted using the Cox proportional hazards model and the log-rank test. A p-value less than 0.050 was considered statistically significant. When multiple dose-volume parameters reflecting similar dose-burden characteristics were significant in univariate analyses, only one representative parameter was selected for inclusion in the multivariate analysis to avoid redundancy among correlated variables. Model performance was assessed using time-dependent receiver operating characteristic (ROC) analysis. Model fit was evaluated using the corrected Akaike information criterion (AICc) to account for the small sample size, and internal validation was performed using 5-fold cross-validation with discrimination assessed by the concordance index (C-index). Each model was analyzed within the subset of patients for whom the relevant imaging data were available. All statistical analyses were performed using STATA software ver. 17.0 (Stata-Corp LLC).

Results

1. Patient characteristics

The clinical features of the 71 patients are summarized in Table 1. Nineteen patients (26.8%) developed SRP, whereas 52 (73.2%) did not. The median follow-up time was 16.8 months (range, 1.0 to 61.9 months). Forty patients (56.4%) had an underlying lung disease, whereas 31 (43.6%) did not. Forty-four patients (62.0%) received any form of chemotherapy, whereas 27 (38.0%) did not. Immunotherapy was administered to 15 patients (21.1%). Nineteen patients (26.8%) underwent surgery before the administration of radiotherapy. In terms of radiotherapy, 58 patients (81.7%) received radical treatment. Nineteen patients (26.8%) underwent SABR, whereas 52 (73.2%) underwent conventional fractionations. To account for differences in fractionation schedules among patients, the EQD2 dose was standardized using an α/β ratio of 3. The median EQD2 dose was 66 Gy.

2. Dosimetric parameters

Dosimetric parameters are listed in Table 2. The mean planning target volume (PTV) volume was 187.5 mL (range, 10.7 to 971.2 mL). The standard (anatomical lung) and functional lung parameters (perfusion and ventilation) were compared. Total and ipsilateral lung volumes were significantly lower in the functional group than in the standard group for both perfusion and ventilation scans (p < 0.001). The V₂₀ of the total lung was significantly lower in the functional parameter group on the perfusion scan (14.4% vs. 16.2%, p=0.037), whereas the other V₁₀ and V₂₀ parameters showed no significant differences. The MLD was comparable between the groups. The magnitude of dose differences between functional and anatomical lung was summarized using mismatch metrics (mean |pVf_x−V_x|), which describe the extent of dose discrepancy between functional and anatomical lung regions.

3. Univariate and multivariate analyses for assessing the risk of SRP

Table 3 summarizes the univariate and multivariate analyses for risk factors of SRP. In the univariate analysis, PTV ≥ 150 mL was significantly associated with a higher risk of SRP (hazard ratio [HR], 3.10; 95% confidence interval [CI], 1.25 to 7.71; p=0.011). Total lung V₁₀ ≥ 45% (HR, 3.01; 95% CI, 1.12 to 8.09; p=0.022), pVf₁₀ ≥ 45% (HR, 4.71; 95% CI, 1.73 to 12.9; p=0.001), vVf₁₀ ≥ 45% (HR, 3.52; 95% CI, 1.33 to 9.31; p=0.007), pVf₂₀ ≥ 20% (HR, 2.65; 95% CI, 1.00 to 6.98; p=0.041), and total lung pVf₁₀ > total lung V₁₀ (HR, 3.46; 95% CI, 1.22 to 9.82; p=0.013) were significantly associated with an increased risk of SRP. In the ipsilateral lung, V₁₀ ≥ 60% (HR, 2.57; 95% CI, 1.00 to 6.62; p=0.043) and vVf₁₀ ≥ 60% (HR, 2.88; 95% CI, 1.06 to 7.84; p=0.030) were significantly associated with an increased risk of SRP.

Total lung V₁₀ ≥ 45% and total lung pVf₂₀ ≥ 20% were excluded from the multivariate Cox regression due to overlap with related dose-volume parameters. In the multivariate analysis, PTV ≥ 150 mL (HR, 10.92; 95% CI, 2.28 to 52.26; p=0.004), total lung vVf₁₀ ≥ 45% (HR, 13.23; 95% CI, 1.40 to 125.43; p=0.024), total lung pVf₁₀ > total lung V₁₀ (HR, 8.69; 95% CI, 2.02 to 37.37; p=0.004), and ipsilateral lung vVf₁₀ ≥ 60% (HR, 23.35; 95% CI, 2.38 to 228.8; p=0.007) remained significant.

4. Choosing optimal models for identifying the risk of SRP

Statistically significant lung dose–volume parameters were grouped according to their anatomical or functional basis, resulting in anatomical (group A), perfusion (group B), and ventilation (group C) classifications (Fig. 2). PTV volume and perfusion mismatch (defined as total lung pVf₁₀ > total lung V₁₀) were included in all models as baseline covariates. Three multivariable models for identifying the risk of SRP were then constructed by combining the baseline covariates with each parameter group, as follows:

Model 1: PTV ≥ 150 mL, total lung pVf₁₀ > total lung V₁₀, group A

Model 2: PTV ≥ 150 mL, total lung pVf₁₀ > total lung V₁₀, group B

Model 3: PTV ≥ 150 mL, total lung pVf₁₀ > total lung V₁₀, group C

Models 1, 2, and 3 had AICc values of 116.38, 114.11, and 85.81, respectively, with model 3 showing the lowest AICc value. The forest plots and ROC curves of the three models are shown in Fig. 3. As shown in Fig. 3C, group C exhibited an HR of 9.80 (95% CI, 1.98 to 48.47; p=0.005) and the highest area under the curve (AUC) value of 0.819. Internal validation using 5-fold cross-validation demonstrated consistent discrimination, with C-indices of 0.78, 0.79, and 0.83 for models 1, 2, and 3, respectively. In a sensitivity analysis using a uniform 25% functional lung threshold, the relative ranking of model performance based on AICc and ROC analysis was preserved, with model 3 remaining the best-performing model (S2 Table).

Discussion

Radiation dose-volume parameters such as MLD and V₂₀ are the most commonly used dose constraints in routine clinical practice for lung protection. Generally, it is recommended to maintain MLD < 20-23 Gy and V₂₀ < 37% in conventional fractionation [12,13]. In this study, all patients met these safety criteria, with the highest MLD observed at 18.8 Gy and V₂₀ at 34.0%. Despite compliance with these guidelines, instances of SRP still occurred. This indicates that not every voxel in the anatomical lung equally represents pulmonary function. With lung SPECT, each anatomical voxel can be converted into a functional voxel, potentially offering a more accurate representation of lung functionality.

Lung SPECT enables functional assessment by visualizing both ventilation and perfusion. A perfusion scan measures the blood flow through the lung tissue, assessing the distribution and flow of blood. In contrast, a ventilation scan evaluates the amount and distribution of air entering and exiting the lungs by measuring pulmonary ventilation function [14]. Some previous studies employed 4DCT-ventilation–based functional avoidance to reduce treatment-related toxicity, often assuming that ventilation and perfusion share similar spatial distributions [15-18]. While this holds true for many cases, there are clinical scenarios, such as pulmonary edema, atelectasis, pleural effusion, and airway obstruction, where ventilation defects lead to notable spatial mismatches between ventilation and perfusion [19]. Forghani et al. [20] noted spatial mismatches in 25% of cases, and Nakajima et al. [21] reported interpatient variability of up to 6.6 Gy between ventilation and perfusion parameters. These findings emphasize the importance of incorporating both modalities for accurate functional lung evaluation.

Several studies have reported the predictive value of functional parameters derived from perfusion SPECT alone [6,8-10]. Farr et al. [6] found that functional parameters derived from SPECT were more effective in predicting the risk of RP than conventional CT-based dose-volume parameters, particularly functional MLD and functional V_5-30. Lee et al. [8] utilized both perfusion SPECT and fluorodeoxyglucose positron emission tomography/computed tomography (FDG PET/CT) to identify the risk factors associated with RP. They concluded that anatomical parameters, along with perfusion and PET metrics, were correlated. Weller et al. [10] observed that a radiation dose of 10-20 Gy to the anatomical or perfused lung reduced forced expiratory volume in 1 second.

Only a few studies have incorporated both perfusion and ventilation scans. Hoover et al. [7] reported that perfusion- and ventilation-weighted MLD, V₂₀, and V30 were significantly higher in the RP group. Owen et al. [22] found that delivering high radiation doses to low-functioning regions of the lung was significantly associated with an increased risk of radiation-induced lung toxicity. They identified combined ventilation/perfusion metrics and the average dose to the lower lung as strong independent predictors of radiation-induced lung toxicity, emphasizing the importance of minimizing radiation exposure to both well- and low-functioning lung regions to reduce the risk of toxicity. In the meta-analysis by Midroni et al. [23], functional lung avoidance planning led to statistically significant absolute reductions of 3.22% in the volume of well-ventilated lung receiving ≥ 20 Gy, 3.52% in well-perfused lung receiving ≥ 20 Gy, 1.3 Gy in the mean dose to well-ventilated lung, and 2.41 Gy in the mean dose to well-perfused lung.

To evaluate the association between functional lung parameters and SRP risk, we constructed three multivariable models. Model 1 incorporated an anatomical parameter (PTV and group A) and functional (perfusion) parameter (pVf₁₀ > V₁₀). Model 2 included an anatomical parameter (PTV) and functional (perfusion) parameter (pVf₁₀ > V₁₀ and group B). Model 3 combined an anatomical parameter (PTV) with functional parameters for both perfusion (pVf₁₀ > V₁₀) and ventilation (group C). Among the three statistically significant models, model 3, which incorporated both perfusion and ventilation parameters, demonstrated the best model performance based on the AICc and AUC values. These findings suggest that the addition of ventilation-derived parameters to anatomical and perfusion parameters provides improved risk stratification for SRP.

Functional lung regions were delineated using a uniform 20% threshold of maximum voxel intensity on SPECT images. This choice was based on previous studies, which demonstrated that thresholds in the range of 15%-25% optimize the trade-off between including adequately functional lung tissue and excluding regions with minimal function or noise [24,25]. Lower thresholds below 10% tend to overestimate functional lungs by including noise-prone regions, whereas higher thresholds above 30% to 40% risk excluding moderately functional yet clinically important areas [24,25]. Thus, the 20% cutoff balances these considerations and ensures reproducibility and comparability while minimizing misclassification.

To further evaluate the robustness of the selected threshold, we conducted an additional sensitivity analysis using a more stringent uniform 25% threshold. Although absolute model performance decreased, the relative ranking of the models was preserved, and the overall findings remained consistent despite the change in functional lung threshold. These findings suggest that modest variations within commonly used thresholds do not result in substantial changes in functional lung parameters relevant to SRP risk.

In this study, PTV was significantly associated with the development of SRP. Previous investigations have similarly demonstrated that a higher ratio of PTV to total lung volume correlates with an increased risk of radiation-induced pulmonary toxicity [26,27]. Although the specific volumetric threshold has not been clearly established, our analysis identified 150 mL as a clinically relevant cutoff, above which the incidence of SRP markedly increased. Notably, this association remained significant, independent of functional dosimetric parameters, suggesting that the absolute irradiated tumor volume is independently associated with SRP risk.

In patients with total lung pVf₁₀ > total lung V₁₀, the functional V₁₀ exceeded anatomical V₁₀ by a mean of 5.2%, which was associated with an increased risk of SRP. Because perfusion scans reflect regional lung blood flow, a higher pVf₁₀ indicates greater involvement of well-perfused, functionally active lung regions within the irradiated volume and may serve as a surrogate marker of functional lung over-irradiation. This finding supports the integration of perfusion information into anatomical planning to better identify lung regions vulnerable to radiation-induced injury. The mean discrepancy between functional and anatomical lung dose distributions shown in Table 2 reflects the degree of functional lung predominance and may provide more refined SRP risk stratification beyond a binary definition of functional lung dominance. Future studies with larger cohorts may enable analyses incorporating the magnitude of dose discrepancy to further improve risk assessment.

Similarly, the significant ventilation parameters in this study were associated with lung volumes receiving more than 10 Gy above the specific thresholds. Group C, defined as total lung vVf₁₀ ≥ 45% and ipsilateral lung vVf₁₀ ≥ 60%, demonstrated a strong association with SRP. Unlike the conventional V₂₀ threshold, a lower dose of 10 Gy was clinically relevant when applied to a ventilation-defined functional lung, likely because ventilation imaging reflects the volume of the lung that actively participates in airflow more directly. Compared with perfusion scans, ventilation imaging may delineate air-filled lung volumes more precisely, suggesting that threshold-based dose criteria can be more reliably defined using ventilation scans.

Previous studies, as summarized by Midroni et al. [23], employed heterogeneous methodologies and cutoff criteria, typically relying on either perfusion or ventilation SPECT alone, with no consensus on optimal thresholds. Our study integrated both perfusion- and ventilation-based functional dosimetric parameters into a unified multivariate model—an approach rarely explored in the existing literature.

Two prospective phase II studies further validated the clinical relevance of functional imaging in thoracic radiotherapy. An ongoing phase II trial, the ASPECT study (ClinicalTrials. gov identifier NCT04676828), is a randomized, double-blind clinical trial evaluating whether incorporating perfusion SPECT into radiation therapy planning for functional lung avoidance can reduce radiation-induced lung toxicity in lung cancer patients [28]. A total of 200 patients are being randomized in a 2:1 ratio to receive either standard or functional avoidance radiotherapy. Another prospective phase II study, the FLARE-RT trial, integrated FDG PET/CT and SPECT/CT by using SPECT to maintain the MLD below 20 Gy, while delivering dose escalation up to 90 Gy to metabolically active regions in mid-treatment non-responders identified by FDG PET/CT [29]. They reported mature outcomes at a median follow-up of 52.3 months, showing 1-year and 2-year overall survival rates of 81.6% and 54.2%, respectively.

This study had several limitations. First, the sample size was relatively small, with only 19 events, which may limit the statistical power and generalizability of the findings. Second, although internal validation using 5-fold cross-validation was performed, the models were developed and evaluated within a single retrospective cohort without external validation. Therefore, these findings should be interpreted as exploratory and require confirmation in independent datasets before clinical application. Third, although a sensitivity analysis using an alternative functional lung threshold was performed and demonstrated consistent model ranking, the lack of a universally accepted definition for functional lung segmentation remains an inherent limitation of SPECT-based approaches. Fourth, the use of MIM software for deformable registration raises concerns about the reliability of matching SPECT/CT and planning CT scans, which is crucial for accurate dose-volume assessments. Finally, the functional avoidance strategy was not implemented prospectively but was evaluated retrospectively, warranting further prospective validation.

In conclusion, functional lung dosimetric parameters derived from both perfusion and ventilation SPECT, particularly pVf₁₀ and vVf₁₀, are significantly associated with the risk of SRP in lung cancer patients receiving radiotherapy. SPECT-based planning may allow preferential dose allocation to non-functioning lung regions, reducing irradiation to normally functional lung. Accordingly, incorporating SPECT-based dosimetric analysis into routine radiotherapy planning may improve risk assessment and stratification for SRP.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).

crt-2025-1248_S1_Fig.pdf

crt-2025-1248_S2_Table.pdf

NOTES

Ethical Statement

This study was reviewed and approved by the institutional review board (IRB) of Hanyang University Hospital (IRB number: HYUH 2024-09-029). The requirement for individual patient consent was waived due to the retrospective nature of the study. The study was conducted in accordance with the Principles of the Declaration of Helsinki.

Author Contributions

Conceived and designed the analysis: Chung JH, Lee SJ, Park HJ.

Collected the data: Chung JH, Lee SJ, Kim JY.

Contributed data or analysis tools: Chung JH, Lee SJ.

Performed the analysis: Chung JH.

Wrote the paper: Chung JH, Lee SJ, Park HJ.

Conflicts of Interest

Conflict of interest relevant to this article was not reported.

Funding

This work was supported by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2021R1G1A1095275).

Declaration of Generative AI and AI-Assisted Technologies

During the preparation of this work the authors used ChatGPT (OpenAI, San Francisco, CA, USA) in order to assist with translating the manuscript from Korean and enhancing the clarity and fluency of the English writing. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Fig. 1.

Representative images of perfusion single-photon emission computed tomography (SPECT)/computed tomography (CT) fused with radiotherapy planning CT image. The perfusion SPECT image demonstrated functional lung segmentation based on a 20% threshold of maximum voxel intensity. Functional (green) and nonfunctional (cyan) lung regions were delineated and co-registered onto the planning CT image using deformable image registration. The red and yellow lines represent the 20 Gy and 10 Gy isodose levels, respectively. In this example, anatomical V₂₀ (percentage of volume of lung receiving ≥ 20 Gy)=6.68% and pVf₂₀ (percentage of volume of lung in perfusion scan receiving ≥ 20 Gy)=15.53%.

Fig. 2.

Schematic overview of group definitions and model construction. Anatomical- and perfusion-based variables (planning target volume [PTV] ≥ 150 mL and total lung pVf₁₀ > total lung V₁₀) were included as common variables. Dose-volume parameters were organized into anatomical (group A), perfusion (group B), or ventilation (group C) groups and incorporated into models 1-3, respectively. PTV, planning target volume; pVf_x, percentage of volume of lung in perfusion scan receiving ≥ x Gy; V_x, percentage of volume of lung receiving ≥ x Gy; vVf_x, percentage of volume of lung in ventilation scan receiving ≥ x Gy.

Fig. 3.

Forest plots and receiver operating characteristic curves for multivariable models assessing the risk of symptomatic radiation pneumonitis. (A) Model 1. (B) Model 2. (C) Model 3. AUC, area under the curve; CI, confidence interval; HR, hazard ratio; PTV, planning target volume.

Table 1.

Patient characteristics

	No. (%)
No. of patients	71
Sex
Male	58 (81.7)
Female	13 (18.3)
Age (yr), median (range)	73 (45-88)
ECOG PS
0	24 (33.8)
1	38 (53.5)
2	9 (12.7)
Underlying lung disease
COPD	32 (45.1)
ILD	8 (11.3)
No	31 (43.7)
Pulmonary function test, mean (range)
Baseline FEV₁ (%)	73.5 (21.6-144.9)
Baseline FVC (%)	74.9 (36.5-116.7)
FEV₁/FVC	0.69 (0.29-0.97)
Baseline DLCO (%)	54.1 (25.7-103.0)
Smoking
Yes	49 (69.0)
No	22 (31.0)
Pathology
NSCLC	50 (70.4)
SCLC	13 (18.3)
Unknown	8 (11.3)
Stage
I	20 (28.2)
II	9 (12.7)
III	40 (56.3)
IV	2 (2.8)
Tumor location
Central	44 (62.0)
Peripheral	27 (38.0)
Lung SPECT
Perfusion	64 (90.1)
Ventilation	54 (76.1)
Both	47 (66.2)
Surgery
Yes	19 (26.8)
No	52 (73.2)
Chemotherapy
Concurrent	34 (47.9)
Induction	17 (23.9)
Consolidation	19 (26.8)
No	27 (38.0)
Immunotherapy
Yes	15 (21.1)
No	56 (78.9)
Radiotherapy
Treatment aim
Radical	58 (81.7)
Postoperative	3 (4.2)
Salvage	9 (12.7)
Palliative	1 (1.4)
Treatment modality
3D-CRT	9 (12.7)
IMRT	43 (60.6)
SABR	19 (26.8)
Mediastinal irradiation
Yes	31 (43.7)
No	40 (56.3)
EQD2₃ (Gy), median (range)	66.0 (43.2-180.0)
No. of fractions	28 (4-33)
Symptomatic radiation pneumonitis
Yes	19 (26.8)
No	52 (73.2)

3D-CRT, three-dimensional conformal radiotherapy; COPD, chronic obstructive pulmonary disease; DLCO, diffusing capacity of the lung for carbon monoxide; ECOG PS, Eastern Cooperative Oncology Group performance status; EQD2₃, equivalent dose in 2 Gy fractions with an α/β ratio of 3; FEV₁, forced expiratory volume in 1 second; FVC, forced vital capacity; ILD, interstitial lung disease; IMRT, intensity-modulated radiotherapy; NSCLC, non–small cell lung cancer; SABR, stereotactic ablative body radiotherapy; SCLC, small cell lung cancer; SPECT, single-photon emission computed tomography.

Table 2.

Radiotherapy planning parameters

	Standard parameters	Functional parameters	p-value^a)
Tumor characteristics
Tumor location, n (%)
Right	44 (62.0)		-
Left	25 (35.2)
Mediastinum only	2 (2.8)
PTV volume (mL), mean (range)	187.5 (10.7-971.2)		-
Perfusion scan
Total lung, mean (range)
Volume (mL)	3,082.0 (1,374.8-6,217.1)	1,842.0 (610.5-3,469.2)	< 0.001
V₁₀ (%)	32.4 (7.3-80.1)	31.5 (6.0-84.9)	0.382
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ > V₁₀	5.2 (0.3-25.7)		-
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ ≤ V₁₀	5.2 (0.0-22.8)		-
V₂₀ (%)	16.2 (3.4-33.0)	14.4 (0.0-36.9)	0.037
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ > V₂₀	3.4 (0.2-12.6)		-
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ ≤ V₂₀	6.5 (0.3-22.6)		-
MLD (Gy)	9.5 (2.0-18.8)	9.0 (1.7-18.9)	0.150
Ipsilateral lung, mean (range)
Volume (mL)	1,588.8 (10.5-4,305.7)	841.8 (0.0-2,294.1)	< 0.001
V₁₀ (%)	47.3 (11.0-100.0)	44.3 (0.0-96.9)	0.227
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ > V₁₀	6.7 (0.3-37.3)		-
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ ≤ V₁₀	15.5 (0.0-100.0)		-
V₂₀ (%)	31.5 (0.0-100.0)	28.8 (0.0-84.1)	0.246
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ > V₂₀	6.2 (0.3-27.8)		-
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ ≤ V₂₀	15.7 (0.0-100.0)		-
MLD (Gy)	14.6 (2.8-43.9)	13.5 (0.0-51.1)	0.240
Ventilation scan
Total lung, mean (range)
Volume (mL)	2,967.4 (1,374.8-6,159.8)	1,799.6 (511.3-3,507.7)	< 0.001
V₁₀ (%)	29.4 (7.3-80.1)	27.6 (0.8-79.1)	0.218
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ > V₁₀	6.1 (0.0-22.0)		-
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ ≤ V₁₀	9.1 (0.1-32.2)		-
V₂₀ (%)	14.4 (3.4-34.0)	13.6 (0.5-36.6)	0.369
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ > V₂₀	4.3 (0.0-9.8)		-
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ ≤ V₂₀	5.9 (0.1-15.9)		-
MLD (Gy)	8.6 (2.0-18.8)	8.2 (0.4-19.8)	0.406
Ipsilateral lung, mean (range)
Volume (mL)	1,554.9 (10.5-4305.7)	924.8 (0.0-3507.7)	< 0.001
V₁₀ (%)	43.9 (11.0-100.0)	40.9 (0.0-95.6)	0.276
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ > V₁₀	9.0 (1.0-25.2)		-
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ ≤ V₁₀	14.2 (0.0-100.0)		-
V₂₀ (%)	28.4 (0.0-100.0)	26.8 (0.0-83.5)	0.516
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ > V₂₀	8.0 (0.3-23.3)		-
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ ≤ V₂₀	11.2 (0.0-100.0)		-
MLD (Gy)	13.4 (2.8-43.9)	12.6 (0.0-39.6)	0.440

MLD, mean lung dose; PTV, planning target volume; pVf_x, percentage of volume of lung in perfusion scan receiving ≥ x Gy; V_x, percentage of volume of lung receiving ≥ x Gy; vVf_x, percentage of volume of lung in ventilation scan receiving ≥ x Gy.

^a) Paired t test.

Table 3.

Univariate and multivariate analyses for assessing the risk of symptomatic radiation pneumonitis

	No. (%)	Univariate		Multivariate
	No. (%)	HR (95% CI)	p-value	HR (95% CI)	p-value
Age (yr)
≥ 70	43 (60.6)	1.26 (0.49-3.20)	0.631	-	-
< 70	28 (39.4)	1.00
Smoking
Yes	49 (69.0)	1.00
No	22 (31.0)	1.16 (0.46-2.96)	0.749	-	-
Underlying lung disease
Yes	40 (56.3)	1.00
No	31 (43.7)	1.92 (0.77-4.78)	0.152	-	-
FEV₁/FVC
≥ 0.7	31 (43.7)	1.00
< 0.7	40 (56.3)	1.47 (0.58-3.73)	0.416	-	-
FVC (%)
≥ 80	31 (43.7)	1.00
< 80	40 (56.3)	1.25 (0.52-3.02)	0.660	-	-
Surgery
Yes	19 (26.8)	1.00
No	52 (73.2)	1.23 (0.49-3.06)	0.495	-	-
Chemotherapy
Yes	44 (62.0)	1.00
No	27 (38.0)	1.01 (0.41-2.53)	0.975	-	-
Immunotherapy
Yes	15 (21.1)	1.41 (0.51-3.92)	0.508	-	-
No	56 (78.9)	1.00
Ipsilateral lung volume (L)
≥ 2	15 (21.1)	1.00
< 2	56 (78.9)	5.50 (0.73-41.2)	0.062	-	-
PTV volume (mL)
≥ 150	27 (38.0)	3.10 (1.25-7.71)	0.011	10.92 (2.28-52.26)	0.004
< 150	44 (62.0)	1.00		1.00
Tumor location
Central	44 (62.0)	1.52 (0.58-4.01)	0.386	-	-
Peripheral	27 (38.0)	1.00
Mediastinal irradiation
Yes	31 (43.7)	1.00
No	40 (56.3)	1.16 (0.46-2.97)	0.749	-	-
Radiation dose, EQD2₃ (Gy)
≥ 100	20 (28.2)	1.00
< 100	51 (71.8)	1.05 (0.40-2.76)	0.928	-	-
Total lung
V₁₀ (%)
≥ 45	16 (22.5)	3.01 (1.12-8.09)	0.022	N/A	N/A
< 45	55 (77.5)	1.00
pVf₁₀ (%)
≥ 45	15 (23.4)	4.71 (1.73-12.9)	0.001	N/A	N/A
< 45	49 (76.6)	1.00
vVf₁₀ (%)
≥ 45	12 (22.2)	3.52 (1.33-9.31)	0.007	13.23 (1.40-125.43)	0.024
< 45	42 (77.8)	1.00		1.00
V₂₀ (%)
≥ 20	27 (38.0)	1.00
< 20	44 (62.0)	1.01 (0.38-2.68)	0.978	-	-
pVf₂₀ (%)
≥ 20	16 (25.0)	2.65 (1.00-6.98)	0.041	1.00
< 20	48 (75.0)	1.00		6.06 (0.97-37.91)	0.054
vVf₂₀ (%)
≥ 20	14 (25.9)	1.23 (0.43-3.50)	0.693	-	-
< 20	40 (74.1)	1.00
pVf₁₀ > V₁₀
Yes	27 (42.2)	3.46 (1.22-9.82)	0.013	8.69 (2.02-37.37)	0.004
No	37 (57.8)	1.00		1.00
pVf₂₀ > V₂₀
Yes	30 (46.9)	1.44 (0.55-3.79)	0.455	-	-
No	34 (53.1)	1.00
vVf₁₀ > V₁₀
Yes	26 (48.2)	1.30 (0.49-3.42)	0.593	-	-
No	28 (51.8)	1.00
vVf₂₀ > V₂₀
Yes	27 (50.0)	1.65 (0.61-4.47)	0.316	-	-
No	27 (50.0)	1.00
pMLD > MLD
Yes	28 (43.7)	2.20 (0.81-5.96)	0.110	-	-
No	36 (56.3)	1.00
vMLD > MLD
Yes	27 (50.0)	1.22 (0.46-3.21)	0.685	-	-
No	27 (50.0)	1.00
Ipsilateral lung
V₁₀ (%)
≥ 60	19 (26.8)	2.57 (1.00-6.62)	0.043	1.00
< 60	52 (73.2)	1.00		10.71 (0.75-152.11)	0.080
pVf₁₀ (%)
≥ 60	19 (29.7)	2.56 (0.96-6.85)	0.052	-	-
< 60	45 (70.3)	1.00
vVf₁₀ (%)
≥ 60	12 (22.2)	2.88 (1.06-7.84)	0.030	23.35 (2.38-228.8)	0.007
< 60	42 (77.8)	1.00		1.00
pVf₁₀ > V₁₀
Yes	36 (56.3)	1.77 (0.62-5.02)	0.278	-	-
No	28 (43.7)	1.00
pVf₂₀ > V₂₀
Yes	38 (59.4)	1.41 (0.50-4.02)	0.513	-	-
No	26 (40.6)	1.00
vVf₁₀ > V₁₀
Yes	26 (48.2)	1.00
No	28 (51.8)	1.17 (0.45-3.04)	0.742	-	-
vVf₂₀ > V₂₀
Yes	27 (50.0)	1.00
No	27 (50.0)	1.77 (0.67-4.66)	0.241	-	-
pMLD > MLD
Yes	37 (57.8)	1.14 (0.42-3.08)	0.798	-	-
No	27 (42.2)	1.00
vMLD > MLD
Yes	28 (51.8)	1.00
No	26 (48.2)	1.11 (0.43-2.88)	0.832	-	-

CI, confidence interval; EQD2₃, equivalent dose in 2 Gy fractions with an α/β ratio of 3; FEV₁, forced expiratory volume in 1 second; FVC, forced vital capacity; HR, hazard ratio; MLD, mean lung dose; N/A, not available; pMLD, mean lung dose in perfusion scan; PTV, planning target volume; pVf_x, percentage of volume of lung in perfusion scan receiving ≥ x Gy; vMLD, mean lung dose in ventilation scan; V_x, percentage of volume of lung receiving ≥ x Gy; vVf_x, percentage of volume of lung in ventilation scan receiving ≥ x Gy.

REFERENCES

1. Palma DA, Senan S, Tsujino K, Barriger RB, Rengan R, Moreno M, et al. Predicting radiation pneumonitis after chemoradiation therapy for lung cancer: an international individual patient data meta-analysis. Int J Radiat Oncol Biol Phys. 2013;85:444–50. Article PubMed
2. Hanania AN, Mainwaring W, Ghebre YT, Hanania NA, Ludwig M. Radiation-induced lung injury: assessment and management. Chest. 2019;156:150–62. Article PubMed PMC
3. Rodrigues G, Lock M, D’Souza D, Yu E, Van Dyk J. Prediction of radiation pneumonitis by dose-volume histogram parameters in lung cancer: a systematic review. Radiother Oncol. 2004;71:127–38. Article PubMed
4. Senan S, Brade A, Wang LH, Vansteenkiste J, Dakhil S, Biesma B, et al. PROCLAIM: randomized phase III trial of pemetrexed-cisplatin or etoposide-cisplatin plus thoracic radiation therapy followed by consolidation chemotherapy in locally advanced nonsquamous non-small-cell lung cancer. J Clin Oncol. 2016;34:953–62. Article PubMed
5. Marks LB, Spencer DP, Sherouse GW, Bentel G, Clough R, Vann K, et al. The role of three dimensional functional lung imaging in radiation treatment planning: the functional dose-volume histogram. Int J Radiat Oncol Biol Phys. 1995;33:65–75. Article PubMed
6. Farr KP, Kallehauge JF, Moller DS, Khalil AA, Kramer S, Bluhme H, et al. Inclusion of functional information from perfusion SPECT improves predictive value of dose-volume parameters in lung toxicity outcome after radiotherapy for non-small cell lung cancer: a prospective study. Radiother Oncol. 2015;117:9–16. Article PubMed
7. Hoover DA, Reid RH, Wong E, Stitt L, Sabondjian E, Rodrigues GB, et al. SPECT-based functional lung imaging for the prediction of radiation pneumonitis: a clinical and dosimetric correlation. J Med Imaging Radiat Oncol. 2014;58:214–22. PubMed
8. Lee HJ Jr, Zeng J, Vesselle HJ, Patel SA, Rengan R, Bowen SR. Correlation of functional lung heterogeneity and dosimetry to radiation pneumonitis using perfusion SPECT/CT and FDG PET/CT imaging. Int J Radiat Oncol Biol Phys. 2018;102:1255–64. Article PubMed PMC
9. Thomas HM, Hippe DS, Forouzannezhad P, Sasidharan BK, Kinahan PE, Miyaoka RS, et al. Radiation and immune checkpoint inhibitor-mediated pneumonitis risk stratification in patients with locally advanced non-small cell lung cancer: role of functional lung radiomics? Discov Oncol. 2022;13:85.Article PubMed PMC PDF
10. Weller A, Dunlop A, Oxer A, Gunapala R, Murray I, Gray MJ, et al. Spect perfusion imaging versus CT for predicting radiation injury to normal lung in lung cancer patients. Br J Radiol. 2019;92:20190184.Article PubMed PMC PDF
11. Wang ZT, Wei LL, Ding XP, Sun MP, Sun HF, Li BS. Spect-guidance to reduce radioactive dose to functioning lung for stage III non-small cell lung cancer. Asian Pac J Cancer Prev. 2013;14:1061–5. Article PubMed
12. Timmerman R. A story of hypofractionation and the table on the wall. Int J Radiat Oncol Biol Phys. 2022;112:4–21. Article PubMed
13. Marks LB, Bentzen SM, Deasy JO, Kong FM, Bradley JD, Vogelius IS, et al. Radiation dose-volume effects in the lung. Int J Radiat Oncol Biol Phys. 2010;76(3 Suppl):S70–6. PubMed PMC
14. Zhou PX, Zhang SX. Functional lung imaging in thoracic tumor radiotherapy: application and progress. Front Oncol. 2022;12:908345.Article PubMed PMC
15. Faught AM, Miyasaka Y, Kadoya N, Castillo R, Castillo E, Vinogradskiy Y, et al. Evaluating the toxicity reduction with computed tomographic ventilation functional avoidance radiation therapy. Int J Radiat Oncol Biol Phys. 2017;99:325–33. Article PubMed PMC
16. Patton TJ, Gerard SE, Shao W, Christensen GE, Reinhardt JM, Bayouth JE. Quantifying ventilation change due to radiation therapy using 4DCT Jacobian calculations. Med Phys. 2018;45:4483–92. PubMed PMC
17. Flakus MJ, Kent SP, Wallat EM, Wuschner AE, Tennant E, Yadav P, et al. Metrics of dose to highly ventilated lung are predictive of radiation-induced pneumonitis in lung cancer patients. Radiother Oncol. 2023;182:109553.Article PubMed PMC
18. Katsuta Y, Kadoya N, Kajikawa T, Mouri S, Kimura T, Takeda K, et al. Radiation pneumonitis prediction model with integrating multiple dose-function features on 4DCT ventilation images. Phys Med. 2023;105:102505.Article PubMed
19. Carvalho P, Lavender JP. The incidence and etiology of the ventilation/perfusion reverse mismatch defect. Clin Nucl Med. 1989;14:571–6. Article PubMed
20. Forghani F, Patton T, Kwak J, Thomas D, Diot Q, Rusthoven C, et al. Characterizing spatial differences between SPECT-ventilation and SPECT-perfusion in patients with lung cancer undergoing radiotherapy. Radiother Oncol. 2021;160:120–4. Article PubMed PMC
21. Nakajima Y, Kadoya N, Kimura T, Hioki K, Jingu K, Yamamoto T. Variations between dose-ventilation and dose-perfusion metrics in radiation therapy planning for lung cancer. Adv Radiat Oncol. 2020;5:459–65. Article PubMed PMC
22. Owen DR, Sun Y, Boonstra PS, McFarlane M, Viglianti BL, Balter JM, et al. Investigating the SPECT dose-function metrics associated with radiation-induced lung toxicity risk in patients with non-small cell lung cancer undergoing radiation therapy. Adv Radiat Oncol. 2021;6:100666.Article PubMed PMC
23. Midroni J, Salunkhe R, Liu Z, Chow R, Boldt G, Palma D, et al. Incorporation of functional lung imaging into radiation therapy planning in patients with lung cancer: a systematic review and meta-analysis. Int J Radiat Oncol Biol Phys. 2024;120:370–408. Article PubMed PMC
24. Ding X, Man X, Sun M, Xing J, Fan T, Gao M, et al. Which is the optimal threshold for defining functional lung in single-photon emission computed tomography lung perfusion imaging of lung cancer patients? Nucl Med Commun. 2018;39:103–9. Article PubMed
25. Li L, Wang L, Wang M, Jin CL, Yang S, Li H, et al. Correlation of semi-quantitative analyses and visual scores in pulmonary perfusion SPECT/CT imaging with pulmonary function test parameters in patients with interstitial lung diseases. Clin Radiol. 2025;82:106795.Article PubMed
26. Ueyama T, Arimura T, Takumi K, Nakamura F, Higashi R, Ito S, et al. Risk factors for radiation pneumonitis after stereotactic radiation therapy for lung tumours: clinical usefulness of the planning target volume to total lung volume ratio. Br J Radiol. 2018;91:20170453.Article PubMed PMC PDF
27. Jiang W, Song Y, Sun Z, Qiu J, Shi L. Dosimetric factors and radiomics features within different regions of interest in planning CT images for improving the prediction of radiation pneumonitis. Int J Radiat Oncol Biol Phys. 2021;110:1161–70. Article PubMed
28. Khalil AA, Hau E, Gebski V, Grau C, Gee H, Nyeng TB, et al. Personal innovative approach in radiation therapy of lung cancer- functional lung avoidance SPECT-guided (ASPECT) radiation therapy: a study protocol for phase II randomised double-blind clinical trial. BMC Cancer. 2021;21:940.Article PubMed PMC PDF
29. Cui S, Zeng J, Hippe DS, Fu J, Yaseen F, He Y, et al. Phase II trial of risk-adaptive chemoradiation for unresectable non-small cell lung cancer: FLARE-RT mature outcomes and patterns of failure. Int J Radiat Oncol Biol Phys. 2024;120(2 Suppl):S59–60. Article

Figure & Data

REFERENCES

Citations

Citations to this article as recorded by

PubReader
ePub Link

Cite

CITE: export Copy Download Format; Close

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

RIS — For EndNote, ProCite, RefWorks, and most other reference management software
BibTeX — For JabRef, BibDesk, and other BibTeX-specific software

Include:

Citation for the content below
Citation and abstract for the content below

Assessing Radiation Pneumonitis through Functional Lung Imaging: A Single Photon Emission Computed Tomography–Based Approach in Lung Cancer

DOI: https://doi.org/10.4143/crt.2025.1248

XML Download

Abstract
Introduction
Materials and Methods
Results
Discussion
Electronic Supplementary Material
NOTES
REFERENCES

Assessing Radiation Pneumonitis through Functional Lung Imaging: A Single Photon Emission Computed Tomography–Based Approach in Lung Cancer

Fig. 1. Representative images of perfusion single-photon emission computed tomography (SPECT)/computed tomography (CT) fused with radiotherapy planning CT image. The perfusion SPECT image demonstrated functional lung segmentation based on a 20% threshold of maximum voxel intensity. Functional (green) and nonfunctional (cyan) lung regions were delineated and co-registered onto the planning CT image using deformable image registration. The red and yellow lines represent the 20 Gy and 10 Gy isodose levels, respectively. In this example, anatomical V20 (percentage of volume of lung receiving ≥ 20 Gy)=6.68% and pVf20 (percentage of volume of lung in perfusion scan receiving ≥ 20 Gy)=15.53%.

Fig. 2. Schematic overview of group definitions and model construction. Anatomical- and perfusion-based variables (planning target volume [PTV] ≥ 150 mL and total lung pVf10 > total lung V10) were included as common variables. Dose-volume parameters were organized into anatomical (group A), perfusion (group B), or ventilation (group C) groups and incorporated into models 1-3, respectively. PTV, planning target volume; pVfx, percentage of volume of lung in perfusion scan receiving ≥ x Gy; Vx, percentage of volume of lung receiving ≥ x Gy; vVfx, percentage of volume of lung in ventilation scan receiving ≥ x Gy.

Fig. 3. Forest plots and receiver operating characteristic curves for multivariable models assessing the risk of symptomatic radiation pneumonitis. (A) Model 1. (B) Model 2. (C) Model 3. AUC, area under the curve; CI, confidence interval; HR, hazard ratio; PTV, planning target volume.

Fig. 1.

Fig. 2.

Fig. 3.

Assessing Radiation Pneumonitis through Functional Lung Imaging: A Single Photon Emission Computed Tomography–Based Approach in Lung Cancer

	No. (%)
No. of patients	71
Sex
Male	58 (81.7)
Female	13 (18.3)
Age (yr), median (range)	73 (45-88)
ECOG PS
0	24 (33.8)
1	38 (53.5)
2	9 (12.7)
Underlying lung disease
COPD	32 (45.1)
ILD	8 (11.3)
No	31 (43.7)
Pulmonary function test, mean (range)
Baseline FEV₁ (%)	73.5 (21.6-144.9)
Baseline FVC (%)	74.9 (36.5-116.7)
FEV₁/FVC	0.69 (0.29-0.97)
Baseline DLCO (%)	54.1 (25.7-103.0)
Smoking
Yes	49 (69.0)
No	22 (31.0)
Pathology
NSCLC	50 (70.4)
SCLC	13 (18.3)
Unknown	8 (11.3)
Stage
I	20 (28.2)
II	9 (12.7)
III	40 (56.3)
IV	2 (2.8)
Tumor location
Central	44 (62.0)
Peripheral	27 (38.0)
Lung SPECT
Perfusion	64 (90.1)
Ventilation	54 (76.1)
Both	47 (66.2)
Surgery
Yes	19 (26.8)
No	52 (73.2)
Chemotherapy
Concurrent	34 (47.9)
Induction	17 (23.9)
Consolidation	19 (26.8)
No	27 (38.0)
Immunotherapy
Yes	15 (21.1)
No	56 (78.9)
Radiotherapy
Treatment aim
Radical	58 (81.7)
Postoperative	3 (4.2)
Salvage	9 (12.7)
Palliative	1 (1.4)
Treatment modality
3D-CRT	9 (12.7)
IMRT	43 (60.6)
SABR	19 (26.8)
Mediastinal irradiation
Yes	31 (43.7)
No	40 (56.3)
EQD2₃ (Gy), median (range)	66.0 (43.2-180.0)
No. of fractions	28 (4-33)
Symptomatic radiation pneumonitis
Yes	19 (26.8)
No	52 (73.2)

	Standard parameters	Functional parameters	p-value^a)
Tumor characteristics
Tumor location, n (%)
Right	44 (62.0)		-
Left	25 (35.2)
Mediastinum only	2 (2.8)
PTV volume (mL), mean (range)	187.5 (10.7-971.2)		-
Perfusion scan
Total lung, mean (range)
Volume (mL)	3,082.0 (1,374.8-6,217.1)	1,842.0 (610.5-3,469.2)	< 0.001
V₁₀ (%)	32.4 (7.3-80.1)	31.5 (6.0-84.9)	0.382
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ > V₁₀	5.2 (0.3-25.7)		-
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ ≤ V₁₀	5.2 (0.0-22.8)		-
V₂₀ (%)	16.2 (3.4-33.0)	14.4 (0.0-36.9)	0.037
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ > V₂₀	3.4 (0.2-12.6)		-
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ ≤ V₂₀	6.5 (0.3-22.6)		-
MLD (Gy)	9.5 (2.0-18.8)	9.0 (1.7-18.9)	0.150
Ipsilateral lung, mean (range)
Volume (mL)	1,588.8 (10.5-4,305.7)	841.8 (0.0-2,294.1)	< 0.001
V₁₀ (%)	47.3 (11.0-100.0)	44.3 (0.0-96.9)	0.227
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ > V₁₀	6.7 (0.3-37.3)		-
Mean \|pVf₁₀–V₁₀\| (%) for pVf₁₀ ≤ V₁₀	15.5 (0.0-100.0)		-
V₂₀ (%)	31.5 (0.0-100.0)	28.8 (0.0-84.1)	0.246
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ > V₂₀	6.2 (0.3-27.8)		-
Mean \|pVf₂₀–V₂₀\| (%) for pVf₂₀ ≤ V₂₀	15.7 (0.0-100.0)		-
MLD (Gy)	14.6 (2.8-43.9)	13.5 (0.0-51.1)	0.240
Ventilation scan
Total lung, mean (range)
Volume (mL)	2,967.4 (1,374.8-6,159.8)	1,799.6 (511.3-3,507.7)	< 0.001
V₁₀ (%)	29.4 (7.3-80.1)	27.6 (0.8-79.1)	0.218
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ > V₁₀	6.1 (0.0-22.0)		-
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ ≤ V₁₀	9.1 (0.1-32.2)		-
V₂₀ (%)	14.4 (3.4-34.0)	13.6 (0.5-36.6)	0.369
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ > V₂₀	4.3 (0.0-9.8)		-
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ ≤ V₂₀	5.9 (0.1-15.9)		-
MLD (Gy)	8.6 (2.0-18.8)	8.2 (0.4-19.8)	0.406
Ipsilateral lung, mean (range)
Volume (mL)	1,554.9 (10.5-4305.7)	924.8 (0.0-3507.7)	< 0.001
V₁₀ (%)	43.9 (11.0-100.0)	40.9 (0.0-95.6)	0.276
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ > V₁₀	9.0 (1.0-25.2)		-
Mean \|vVf₁₀–V₁₀\| (%) for vVf₁₀ ≤ V₁₀	14.2 (0.0-100.0)		-
V₂₀ (%)	28.4 (0.0-100.0)	26.8 (0.0-83.5)	0.516
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ > V₂₀	8.0 (0.3-23.3)		-
Mean \|vVf₂₀–V₂₀\| (%) for vVf₂₀ ≤ V₂₀	11.2 (0.0-100.0)		-
MLD (Gy)	13.4 (2.8-43.9)	12.6 (0.0-39.6)	0.440

	No. (%)	Univariate		Multivariate
	No. (%)	HR (95% CI)	p-value	HR (95% CI)	p-value
Age (yr)
≥ 70	43 (60.6)	1.26 (0.49-3.20)	0.631	-	-
< 70	28 (39.4)	1.00
Smoking
Yes	49 (69.0)	1.00
No	22 (31.0)	1.16 (0.46-2.96)	0.749	-	-
Underlying lung disease
Yes	40 (56.3)	1.00
No	31 (43.7)	1.92 (0.77-4.78)	0.152	-	-
FEV₁/FVC
≥ 0.7	31 (43.7)	1.00
< 0.7	40 (56.3)	1.47 (0.58-3.73)	0.416	-	-
FVC (%)
≥ 80	31 (43.7)	1.00
< 80	40 (56.3)	1.25 (0.52-3.02)	0.660	-	-
Surgery
Yes	19 (26.8)	1.00
No	52 (73.2)	1.23 (0.49-3.06)	0.495	-	-
Chemotherapy
Yes	44 (62.0)	1.00
No	27 (38.0)	1.01 (0.41-2.53)	0.975	-	-
Immunotherapy
Yes	15 (21.1)	1.41 (0.51-3.92)	0.508	-	-
No	56 (78.9)	1.00
Ipsilateral lung volume (L)
≥ 2	15 (21.1)	1.00
< 2	56 (78.9)	5.50 (0.73-41.2)	0.062	-	-
PTV volume (mL)
≥ 150	27 (38.0)	3.10 (1.25-7.71)	0.011	10.92 (2.28-52.26)	0.004
< 150	44 (62.0)	1.00		1.00
Tumor location
Central	44 (62.0)	1.52 (0.58-4.01)	0.386	-	-
Peripheral	27 (38.0)	1.00
Mediastinal irradiation
Yes	31 (43.7)	1.00
No	40 (56.3)	1.16 (0.46-2.97)	0.749	-	-
Radiation dose, EQD2₃ (Gy)
≥ 100	20 (28.2)	1.00
< 100	51 (71.8)	1.05 (0.40-2.76)	0.928	-	-
Total lung
V₁₀ (%)
≥ 45	16 (22.5)	3.01 (1.12-8.09)	0.022	N/A	N/A
< 45	55 (77.5)	1.00
pVf₁₀ (%)
≥ 45	15 (23.4)	4.71 (1.73-12.9)	0.001	N/A	N/A
< 45	49 (76.6)	1.00
vVf₁₀ (%)
≥ 45	12 (22.2)	3.52 (1.33-9.31)	0.007	13.23 (1.40-125.43)	0.024
< 45	42 (77.8)	1.00		1.00
V₂₀ (%)
≥ 20	27 (38.0)	1.00
< 20	44 (62.0)	1.01 (0.38-2.68)	0.978	-	-
pVf₂₀ (%)
≥ 20	16 (25.0)	2.65 (1.00-6.98)	0.041	1.00
< 20	48 (75.0)	1.00		6.06 (0.97-37.91)	0.054
vVf₂₀ (%)
≥ 20	14 (25.9)	1.23 (0.43-3.50)	0.693	-	-
< 20	40 (74.1)	1.00
pVf₁₀ > V₁₀
Yes	27 (42.2)	3.46 (1.22-9.82)	0.013	8.69 (2.02-37.37)	0.004
No	37 (57.8)	1.00		1.00
pVf₂₀ > V₂₀
Yes	30 (46.9)	1.44 (0.55-3.79)	0.455	-	-
No	34 (53.1)	1.00
vVf₁₀ > V₁₀
Yes	26 (48.2)	1.30 (0.49-3.42)	0.593	-	-
No	28 (51.8)	1.00
vVf₂₀ > V₂₀
Yes	27 (50.0)	1.65 (0.61-4.47)	0.316	-	-
No	27 (50.0)	1.00
pMLD > MLD
Yes	28 (43.7)	2.20 (0.81-5.96)	0.110	-	-
No	36 (56.3)	1.00
vMLD > MLD
Yes	27 (50.0)	1.22 (0.46-3.21)	0.685	-	-
No	27 (50.0)	1.00
Ipsilateral lung
V₁₀ (%)
≥ 60	19 (26.8)	2.57 (1.00-6.62)	0.043	1.00
< 60	52 (73.2)	1.00		10.71 (0.75-152.11)	0.080
pVf₁₀ (%)
≥ 60	19 (29.7)	2.56 (0.96-6.85)	0.052	-	-
< 60	45 (70.3)	1.00
vVf₁₀ (%)
≥ 60	12 (22.2)	2.88 (1.06-7.84)	0.030	23.35 (2.38-228.8)	0.007
< 60	42 (77.8)	1.00		1.00
pVf₁₀ > V₁₀
Yes	36 (56.3)	1.77 (0.62-5.02)	0.278	-	-
No	28 (43.7)	1.00
pVf₂₀ > V₂₀
Yes	38 (59.4)	1.41 (0.50-4.02)	0.513	-	-
No	26 (40.6)	1.00
vVf₁₀ > V₁₀
Yes	26 (48.2)	1.00
No	28 (51.8)	1.17 (0.45-3.04)	0.742	-	-
vVf₂₀ > V₂₀
Yes	27 (50.0)	1.00
No	27 (50.0)	1.77 (0.67-4.66)	0.241	-	-
pMLD > MLD
Yes	37 (57.8)	1.14 (0.42-3.08)	0.798	-	-
No	27 (42.2)	1.00
vMLD > MLD
Yes	28 (51.8)	1.00
No	26 (48.2)	1.11 (0.43-2.88)	0.832	-	-

Table 1. Patient characteristics

Table 2. Radiotherapy planning parameters

Paired t test.

Table 3. Univariate and multivariate analyses for assessing the risk of symptomatic radiation pneumonitis