Predicting lung radiotherapy-induced pneumonitis using a model combining parametric Lyman probit with nonparametric decision trees

Utilization of machine learning methods for prediction of acute and late rectal toxicity due to curative prostate radiotherapy

OBJECTIVE

Rectal toxicity is one of the primary dose-limiting side effects of prostate cancer radiotherapy, and consequential impairment on quality of life in these patients with long survival is an important problem. In this study, we aimed to evaluate the possibility of predicting rectal toxicity with artificial intelligence model which was including certain dosimetric parameters.

MATERIALS AND METHODS

One hundred and thirty-seven patients with a diagnosis of prostate cancer who received curative radiotherapy for prostate +/- pelvic lymphatics were included in the study. The association of the clinical data and dosimetric data between early and late rectal toxicity reported during follow-up was evaluated. The sample size was increased to 274 patients by synthetic data generation method. To determine suitable models, 15 models were studied with machine learning algorithms using Python 2.3, Pycaret library. Random forest classifier was used with to detect active variables.

RESULTS

The area under the curve and accuracy were found to be 0.89-0.97 and 95%-99%, respectively, with machine learning algorithms. The sensitivity values for acute and toxicity were found to be 0.95 and 0.99, respectively.

CONCLUSION

Early or late rectal toxicity can be predicted with a high probability via dosimetric and physical data and machine learning algorithms of patients who underwent prostate +/- pelvic radiotherapy. The fact that rectal toxicity can be predicted before treatment, which may result in limiting the dose and duration of treatment, makes us think that artificial intelligence can enter our daily practice in a short time in this sense.

Embedding machine learning based toxicity models within radiotherapy treatment plan optimization

Objective.This study addresses radiation-induced toxicity (RIT) challenges in radiotherapy (RT) by developing a personalized treatment planning framework. It leverages patient-specific data and dosimetric information to create an optimization model that limits adverse side effects using constraints learned from historical data.Approach.The study uses the optimization with constraint learning (OCL) framework, incorporating patient-specific factors into the optimization process. It consists of three steps: optimizing the baseline treatment plan using population-wide dosimetric constraints; training a machine learning (ML) model to estimate the patient's RIT for the baseline plan; and adapting the treatment plan to minimize RIT using ML-learned patient-specific constraints. Various predictive models, including classification trees, ensembles of trees, and neural networks, are applied to predict the probability of grade 2+ radiation pneumonitis (RP2+) for non-small cell lung (NSCLC) cancer patients three months post-RT. The methodology is assessed with four high RP2+ risk NSCLC patients, with the goal of optimizing the dose distribution to constrain the RP2+ outcome below a pre-specified threshold. Conventional and OCL-enhanced plans are compared based on dosimetric parameters and predicted RP2+ risk. Sensitivity analysis on risk thresholds and data uncertainty is performed using a toy NSCLC case.Main results.Experiments show the methodology's capacity to directly incorporate all predictive models into RT treatment planning. In the four patients studied, mean lung dose and V20 were reduced by an average of 1.78 Gy and 3.66%, resulting in an average RP2+ risk reduction from 95% to 42%. Notably, this reduction maintains tumor coverage, although in two cases, sparing the lung slightly increased spinal cord max-dose (0.23 and 0.79 Gy).Significance.By integrating patient-specific information into learned constraints, the study significantly reduces adverse side effects like RP2+ without compromising target coverage. This unified framework bridges the gap between predicting toxicities and optimizing treatment plans in personalized RT decision-making.

Role of the cGAS-STING pathway in radiotherapy for non-small cell lung cancer

One of the most important therapeutic interventions for non-small cell lung cancer is radiotherapy. Ionizing radiation (IR) is classified by traditional radiobiology principles as a direct cytocidal therapeutic agent against cancer, although there is growing recognition of other antitumor immunological responses induced by this modality. The most effective therapeutic combinations to harness radiation-generated antitumor immunity and enhance treatment results for malignancies resistant to existing radiotherapy regimens could be determined by a more sophisticated understanding of the immunological pathways created by radiation. Innate immune signaling is triggered by the activation of cGAS-STING, and this promotes adaptive immune responses to help fight cancer. This identifies a molecular mechanism radiation can use to trigger antitumor immune responses by bridging the DNA-damaging ability of IR with the activation of CD8 + cytotoxic T cell-mediated killing of tumors. We also discuss radiotherapy-related parameters that affect cGAS-STING signaling, negative consequences of cGAS-STING activation, and intriguing treatment options being tested in conjunction with IR to support immune activation by activating STING-signaling. Improved therapeutic outcomes will result from a better understanding of how IR promotes cGAS-STING signaling in immune-based treatment regimens that maximize radiotherapy's anticancer effectiveness.

Hyperpolarized 129Xe Magnetic Resonance Imaging for Functional Avoidance Treatment Planning in Thoracic Radiation Therapy: A Comparison of Ventilation- and Gas Exchange-Guided Treatment Plans

PURPOSE

To present a methodology to use pulmonary gas exchange maps to guide functional avoidance treatment planning in radiation therapy (RT) and evaluate its efficacy compared with ventilation-guided treatment planning.

METHODS AND MATERIALS

Before receiving conventional RT for non-small cell lung cancer, 11 patients underwent hyperpolarized ¹²⁹Xe gas exchange magnetic resonance imaging to map the distribution of xenon in its gas phase (ventilation) and transiently bound to red blood cells in the alveolar capillaries (gas exchange). Both ventilation and gas exchange maps were independently used to guide development of new functional avoidance treatment plans for every patient, while adhering to institutional dose-volume constraints for normal tissues and target coverage. Furthermore, dose-volume histogram (DVH)-based reoptimizations of the clinical plan, with reductions in mean lung dose (MLD) equal to the functional avoidance plans, were created to serve as the control group. To evaluate each plan (regardless of type), gas exchange maps, representing end-to-end lung function, were used to calculate gas exchange-weighted MLD (fMLD), gas exchange-weighted volume receiving ≥20 Gy (fV20), and mean dose in the highest gas exchanging 33% and 50% volumes of lung (MLD-f33% and MLD-f50%). Using each clinically approved plan as a baseline, the reductions in functional metrics were compared for ventilation-optimization, gas exchange optimization, and DVH-based reoptimization. Statistical significance was determined using the Freidman test, with subsequent subdivision when indicated by P values less than .10 and post hoc testing with Wilcoxon signed rank tests to determine significant differences (P < .05). Toxicity modeling was performed using an established function-based model to estimate clinical significance of the results.

RESULTS

Compared with DVH-based reoptimization of the clinically approved plans, gas exchange-guided functional avoidance planning more effectively reduced the gas exchange-weighted metrics fMLD (average ± SD, -78 ± 79 cGy for gas exchange, compared with -45 ± 34 cGy for DVH-based; P = .03), MLD-f33% (-135 ± 136 cGy, compared with -52 ± 47 cGy; P = .004), and MLD-f50% (-96 ± 95 cGy, compared with -47 ± 40 cGy; P = .01). Comparing the 2 functional planning types, gas exchange-guided planning more effectively reduced MLD-f33% compared with ventilation-guided planning (-64 ± 95; P = .009). For some patients, gas exchange-guided functional avoidance plans demonstrated clinically significant reductions in model-predicted toxicity, more so than the accompanying ventilation-guided plans and DVH-based reoptimizations.

CONCLUSION

Gas exchange-guided planning effectively reduced dose to high gas exchanging regions of lung while maintaining clinically acceptable plan quality. In many patients, ventilation-guided planning incidentally reduced dose to higher gas exchange regions, to a lesser extent. This methodology enables future prospective trials to examine patient outcomes.

Predictive Modeling of Thoracic Radiotherapy Toxicity and the Potential Role of Serum Alpha-2-Macroglobulin

Background: To investigate the impact of alpha-2-macroglobulin (A2M), a suspected intrinsic radioprotectant, on radiation pneumonitis and esophagitis using multifactorial predictive models. Materials and Methods: Baseline A2M levels were obtained for 258 patients prior to thoracic radiotherapy (RT). Dose-volume characteristics were extracted from treatment plans. Spearman's correlation (Rs) test was used to correlate clinical and dosimetric variables with toxicities. Toxicity prediction models were built using least absolute shrinkage and selection operator (LASSO) logistic regression on 1,000 bootstrapped datasets. Results: Grade ≥2 esophagitis and pneumonitis developed in 61 (23.6%) and 36 (14.0%) patients, respectively. The median A2M level was 191 mg/dL (range: 94-511). Never/former/current smoker status was 47 (18.2%)/179 (69.4%)/32 (12.4%). We found a significant negative univariate correlation between baseline A2M levels and esophagitis (Rs = -0.18/p = 0.003) and between A2M and smoking status (Rs = 0.13/p = 0.04). Further significant parameters for grade ≥2 esophagitis included age (Rs = -0.32/p < 0.0001), chemotherapy use (Rs = 0.56/p < 0.0001), dose per fraction (Rs = -0.57/p < 0.0001), total dose (Rs = 0.35/p < 0.0001), and several other dosimetric variables with Rs > 0.5 (p < 0.0001). The only significant non-dosimetric parameter for grade ≥2 pneumonitis was sex (Rs = -0.32/p = 0.037) with higher risk for women. For pneumonitis D15 (lung) (Rs = 0.19/p = 0.006) and D45 (heart) (Rs = 0.16/p = 0.016) had the highest correlation. LASSO models applied on the validation data were statistically significant and resulted in areas under the receiver operating characteristic curve of 0.84 (esophagitis) and 0.78 (pneumonitis). Multivariate predictive models did not require A2M to reach maximum predictive power. Conclusion: This is the first study showing a likely association of higher baseline A2M values with lower risk of radiation esophagitis and with smoking status. However, the baseline A2M level was not a significant risk factor for radiation pneumonitis.

Machine Learning-Based Models for Prediction of Toxicity Outcomes in Radiotherapy

In order to limit radiotherapy (RT)-related side effects, effective toxicity prediction and assessment schemes are essential. In recent years, the growing interest toward artificial intelligence and machine learning (ML) within the science community has led to the implementation of innovative tools in RT. Several researchers have demonstrated the high performance of ML-based models in predicting toxicity, but the application of these approaches in clinics is still lagging, partly due to their low interpretability. Therefore, an overview of contemporary research is needed in order to familiarize practitioners with common methods and strategies. Here, we present a review of ML-based models for predicting and classifying RT-induced complications from both a methodological and a clinical standpoint, focusing on the type of features considered, the ML methods used, and the main results achieved. Our work overviews published research in multiple cancer sites, including brain, breast, esophagus, gynecological, head and neck, liver, lung, and prostate cancers. The aim is to define the current state of the art and main achievements within the field for both researchers and clinicians.

Survey on deep learning for radiotherapy

More than 50% of cancer patients are treated with radiotherapy, either exclusively or in combination with other methods. The planning and delivery of radiotherapy treatment is a complex process, but can now be greatly facilitated by artificial intelligence technology. Deep learning is the fastest-growing field in artificial intelligence and has been successfully used in recent years in many domains, including medicine. In this article, we first explain the concept of deep learning, addressing it in the broader context of machine learning. The most common network architectures are presented, with a more specific focus on convolutional neural networks. We then present a review of the published works on deep learning methods that can be applied to radiotherapy, which are classified into seven categories related to the patient workflow, and can provide some insights of potential future applications. We have attempted to make this paper accessible to both radiotherapy and deep learning communities, and hope that it will inspire new collaborations between these two communities to develop dedicated radiotherapy applications.

GSTP1 Ile105Val polymorphism might be associated with the risk of radiation pneumonitis among lung cancer patients in Chinese population: A prospective study

Background: Growing data suggest that DNA damage repair and detoxification pathways play crucial roles in radiation-induced toxicities. To determine whether common functional single-nucleotide polymorphisms (SNPs) in candidate genes from these pathways can be used as predictors of radiation pneumonitis (RP), we conducted a prospective study to evaluate the associations between functional SNPs and risk of RP.

Methods: We recruited a total of 149 lung cancer patients who had received intensity modulated radiation therapy (IMRT). GSTP1 and XRCC1 were genotyped using the SurPlexTM-xTAG method in all patients. RP events were prospectively scored using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE), version 4.0. Kaplan-Meier analysis was used to determine the cumulative probability of RP of grade ≥ 2. Cox proportional hazard regression was performed to identify clinical variables and SNPs associated with risk of RP grade ≥ 2, using univariate and multivariate analysis, respectively.

Results: With a median follow-up of 9 months, the incidence of RP of grade ≥ 2 was 38.3%. A predicting role in RP was observed for the GSTP1 SNP (adjusted hazard ratio 3.543; 95% CI 1.770-7.092; adjusted P< 0.001 for the Ile/Val and Val/Val genotypes versus Ile/Ile genotype). Whereas, we found that patients with XRCC1 399Arg/Gln and Gln/Gln genotypes had a lower risk of RP compares with those carrying Arg/Arg genotype (adjusted HR 0.653; 95% CI 0.342-1.245), but with no statistical significance observed (adjusted P = 0.195).

Conclusions: Our results suggested a novel association between GSTP1 SNP 105Ile/Val and risk of RP development, which suggests the potential use of this genetic polymorphism as a predictor of RP. In addition, genetic polymorphisms of XRCC1 399Arg/Gln may also be associated with RP.

Association of DNA repair gene polymorphisms with the risk of radiation pneumonitis in lung cancer patients

A total of 149 lung cancer patients were recruited to receive intensity modulated radiation therapy (IMRT). The association of developing radiation pneumonitis (RP) with genetic polymorphism was evaluated. The risks of four polymorphic sites in three DNA repair related genes (ERCC1, rs116615:T354C and rs3212986:C1516A; ERCC2, rs13181:A2251C; XRCC1, rs25487:A1196G) for developing grade ≥ 2 RP were assessed respectively. It was observed that ERCC1 T354C SNP had a significant effect on the development of grade ≥ 2 RP (CT/TT vs. CC, adjusted HR = 0.517, 95% CI, 0.285-0.939; adjusted P = 0.030). It is the first time demonstrating that CT/TT genotype of ERCC1 354 was significantly associated with lower RP risk after radio therapy.

Statistical-learning strategies generate only modestly performing predictive models for urinary symptoms following external beam radiotherapy of the prostate: A comparison of conventional and machine-learning methods

PURPOSE

Given the paucity of available data concerning radiotherapy-induced urinary toxicity, it is important to ensure derivation of the most robust models with superior predictive performance. This work explores multiple statistical-learning strategies for prediction of urinary symptoms following external beam radiotherapy of the prostate.

METHODS

The performance of logistic regression, elastic-net, support-vector machine, random forest, neural network, and multivariate adaptive regression splines (MARS) to predict urinary symptoms was analyzed using data from 754 participants accrued by TROG03.04-RADAR. Predictive features included dose-surface data, comorbidities, and medication-intake. Four symptoms were analyzed: dysuria, haematuria, incontinence, and frequency, each with three definitions (grade ≥ 1, grade ≥ 2 and longitudinal) with event rate between 2.3% and 76.1%. Repeated cross-validations producing matched models were implemented. A synthetic minority oversampling technique was utilized in endpoints with rare events. Parameter optimization was performed on the training data. Area under the receiver operating characteristic curve (AUROC) was used to compare performance using sample size to detect differences of ≥0.05 at the 95% confidence level.

RESULTS

Logistic regression, elastic-net, random forest, MARS, and support-vector machine were the highest-performing statistical-learning strategies in 3, 3, 3, 2, and 1 endpoints, respectively. Logistic regression, MARS, elastic-net, random forest, neural network, and support-vector machine were the best, or were not significantly worse than the best, in 7, 7, 5, 5, 3, and 1 endpoints. The best-performing statistical model was for dysuria grade ≥ 1 with AUROC ± standard deviation of 0.649 ± 0.074 using MARS. For longitudinal frequency and dysuria grade ≥ 1, all strategies produced AUROC>0.6 while all haematuria endpoints and longitudinal incontinence models produced AUROC<0.6.

CONCLUSIONS

Logistic regression and MARS were most likely to be the best-performing strategy for the prediction of urinary symptoms with elastic-net and random forest producing competitive results. The predictive power of the models was modest and endpoint-dependent. New features, including spatial dose maps, may be necessary to achieve better models.

Genetic Variants in MTHFR Gene Predict ≥ 2 Radiation Pneumonitis in Esophageal Squamous Cell Carcinoma Patients Treated with Thoracic Radiotherapy

Reactive oxygen species (ROS), formed as an indirect production of radiotherapy (RT), could cause DNA damage of normal tissues. Meanwhile, our body possesses the ability to restore the damage by DNA repair pathways. The imbalance between the two systems could finally result in radiation injury. Therefore, in this prospective cohort study, we explored the association of genetic variants in ROS metabolism and DNA repair pathway-related genes with radiation pneumonitis (RP). A total of 265 locally advanced esophageal squamous cell carcinoma (ESCC) patients receiving RT in Chinese Han population were enrolled. Five functional single nucleotide polymorphisms (SNPs) (rs1695 in GSTP1; rs4880 in SOD2; rs3957356 in GSTA1; and rs1801131, rs1801133 in MTHFR) were genotyped using the MassArray system, and rs1801131 was found to be a predictor of ≥ 2 RP. Our results showed that, compared with TT genotype, patients with GG/GT genotypes of rs1801131 had a notably lower risk of developing ≥ 2 RP (HR = 0.339, 95% CI = 0.137–0.839, P = 0.019). Further independent studies are required to confirm this findings.

Pitfalls in Prediction Modeling for Normal Tissue Toxicity in Radiation Therapy: An Illustration With the Individual Radiation Sensitivity and Mammary Carcinoma Risk Factor Investigation Cohorts

PURPOSE

To identify the main causes underlying the failure of prediction models for radiation therapy toxicity to replicate.

METHODS AND MATERIALS

Data were used from two German cohorts, Individual Radiation Sensitivity (ISE) (n=418) and Mammary Carcinoma Risk Factor Investigation (MARIE) (n=409), of breast cancer patients with similar characteristics and radiation therapy treatments. The toxicity endpoint chosen was telangiectasia. The LASSO (least absolute shrinkage and selection operator) logistic regression method was used to build a predictive model for a dichotomized endpoint (Radiation Therapy Oncology Group/European Organization for the Research and Treatment of Cancer score 0, 1, or ≥2). Internal areas under the receiver operating characteristic curve (inAUCs) were calculated by a naïve approach whereby the training data (ISE) were also used for calculating the AUC. Cross-validation was also applied to calculate the AUC within the same cohort, a second type of inAUC. Internal AUCs from cross-validation were calculated within ISE and MARIE separately. Models trained on one dataset (ISE) were applied to a test dataset (MARIE) and AUCs calculated (exAUCs).

RESULTS

Internal AUCs from the naïve approach were generally larger than inAUCs from cross-validation owing to overfitting the training data. Internal AUCs from cross-validation were also generally larger than the exAUCs, reflecting heterogeneity in the predictors between cohorts. The best models with largest inAUCs from cross-validation within both cohorts had a number of common predictors: hypertension, normalized total boost, and presence of estrogen receptors. Surprisingly, the effect (coefficient in the prediction model) of hypertension on telangiectasia incidence was positive in ISE and negative in MARIE. Other predictors were also not common between the 2 cohorts, illustrating that overcoming overfitting does not solve the problem of replication failure of prediction models completely.

CONCLUSIONS

Overfitting and cohort heterogeneity are the 2 main causes of replication failure of prediction models across cohorts. Cross-validation and similar techniques (eg, bootstrapping) cope with overfitting, but the development of validated predictive models for radiation therapy toxicity requires strategies that deal with cohort heterogeneity.

HSPB1 polymorphisms might be associated with radiation-induced damage risk in lung cancer patients treated with radiotherapy

Several studies investigating the association between heat shock protein beta-1 (HSPB1) polymorphisms and radiation-induced damage in lung cancer patients administrated with radiotherapy have derived conflicting results. This meta-analysis aimed to assess the association between the HSPB1 genes' (rs2868370 and rs2868371) polymorphisms and the risk of radiation-induced damage in lung cancer patients. After an electronic literature search, four articles including six studies were found to be eligible for this meta-analysis. No association was observed between rs2868370 genotypes and radiation-induced damage risk. However, rs2868371 showed a statistically increased risk of radiation-induced damage under CC vs. CG/GG model (OR = 1.59, 95 % CI = 1.10-2.29). Subgroup analysis by ethnicity showed that the genotypes of rs2868371 were also associated with a significantly increased risk of radiation-induced damage in CC vs. CG/GG model (OR = 1.86, 95 % CI = 1.21-2.83) among mixed ethnicities which are mainly comprised of white people. When the data was stratified by organ-damaged, a significant association was only observed in the esophagus group (OR = 2.94, 95 % CI = 1.35-6.37, for CC vs. CG/GG model). In conclusion, the present study demonstrated that the rs2868371 genotypes of HSPB1 might be associated with radiation-induced esophagus damage risk, especially in Caucasians but not in the Asian population.

Machine Learning Approaches for Predicting Radiation Therapy Outcomes: A Clinician's Perspective

Radiation oncology has always been deeply rooted in modeling, from the early days of isoeffect curves to the contemporary Quantitative Analysis of Normal Tissue Effects in the Clinic (QUANTEC) initiative. In recent years, medical modeling for both prognostic and therapeutic purposes has exploded thanks to increasing availability of electronic data and genomics. One promising direction that medical modeling is moving toward is adopting the same machine learning methods used by companies such as Google and Facebook to combat disease. Broadly defined, machine learning is a branch of computer science that deals with making predictions from complex data through statistical models. These methods serve to uncover patterns in data and are actively used in areas such as speech recognition, handwriting recognition, face recognition, "spam" filtering (junk email), and targeted advertising. Although multiple radiation oncology research groups have shown the value of applied machine learning (ML), clinical adoption has been slow due to the high barrier to understanding these complex models by clinicians. Here, we present a review of the use of ML to predict radiation therapy outcomes from the clinician's point of view with the hope that it lowers the "barrier to entry" for those without formal training in ML. We begin by describing 7 principles that one should consider when evaluating (or creating) an ML model in radiation oncology. We next introduce 3 popular ML methods--logistic regression (LR), support vector machine (SVM), and artificial neural network (ANN)--and critique 3 seminal papers in the context of these principles. Although current studies are in exploratory stages, the overall methodology has progressively matured, and the field is ready for larger-scale further investigation.

Decision Trees Predicting Tumor Shrinkage for Head and Neck Cancer: Implications for Adaptive Radiotherapy

OBJECTIVE

To develop decision trees predicting for tumor volume reduction in patients with head and neck (H&N) cancer using pretreatment clinical and pathological parameters.

METHODS

Forty-eight patients treated with definitive concurrent chemoradiotherapy for squamous cell carcinoma of the nasopharynx, oropharynx, oral cavity, or hypopharynx were retrospectively analyzed. These patients were rescanned at a median dose of 37.8 Gy and replanned to account for anatomical changes. The percentages of gross tumor volume (GTV) change from initial to rescan computed tomography (CT; %GTVΔ) were calculated. Two decision trees were generated to correlate %GTVΔ in primary and nodal volumes with 14 characteristics including age, gender, Karnofsky performance status (KPS), site, human papilloma virus (HPV) status, tumor grade, primary tumor growth pattern (endophytic/exophytic), tumor/nodal/group stages, chemotherapy regimen, and primary, nodal, and total GTV volumes in the initial CT scan. The C4.5 Decision Tree induction algorithm was implemented.

RESULTS

The median %GTVΔ for primary, nodal, and total GTVs was 26.8%, 43.0%, and 31.2%, respectively. Type of chemotherapy, age, primary tumor growth pattern, site, KPS, and HPV status were the most predictive parameters for primary %GTVΔ decision tree, whereas for nodal %GTVΔ, KPS, site, age, primary tumor growth pattern, initial primary GTV, and total GTV volumes were predictive. Both decision trees had an accuracy of 88%.

CONCLUSIONS

There can be significant changes in primary and nodal tumor volumes during the course of H&N chemoradiotherapy. Considering the proposed decision trees, radiation oncologists can select patients predicted to have high %GTVΔ, who would theoretically gain the most benefit from adaptive radiotherapy, in order to better use limited clinical resources.

Interstitial Lung Change in Pre-radiation Therapy Computed Tomography Is a Risk Factor for Severe Radiation Pneumonitis

PURPOSE

We examined clinical and dosimetric factors as predictors of symptomatic radiation pneumonitis (RP) in lung cancer patients and evaluated the relationship between interstitial lung changes in the pre-radiotherapy (RT) computed tomography (CT) and symptomatic RP.

MATERIALS AND METHODS

Medical records and dose volume histogram data of 60 lung cancer patients from August 2005 to July 2006 were analyzed. All patients were treated with three dimensional (3D) conformal RT of median 56.9 Gy. We assessed the association of symptomatic RP with clinical and dosimetric factors.

RESULTS

With a median follow-up of 15.5 months (range, 6.1 to 40.9 months), Radiation Therapy Oncology Group grade ≥ 2 RP was observed in 14 patients (23.3%). Five patients (8.3%) died from RP. The interstitial changes in the pre-RT chest CT, mean lung dose (MLD), and V30 significantly predicted RP in multivariable analysis (p=0.009, p < 0.001, and p < 0.001, respectively). MLD, V20, V30, and normal tissue complication probability normal tissue complication probability (NTCP) were associated with the RP grade but less so for grade 5 RP. The risk of RP grade ≥ 2, ≥ 3, or ≥ 4 was higher in the patients with interstitial lung change (grade 2, 15.6% to 46.7%, p=0.03; grade 3, 4.4% to 40%, p=0.002; grade 4, 4.4% to 33.3%, p=0.008). Four of the grade 5 RP patients had diffuse interstitial change in pre-RT CT and received chemoradiotherapy.

CONCLUSION

Our study identified diffuse interstitial disease as a significant clinical risk for RP, particularly fatal RP. We showed the usefulness of MLD, V20, V30, and NTCP in predicting the incidence and severity of RP.

Nondosimetric risk factors for radiation-induced lung toxicity

The decision to administer a radical course of radiotherapy (RT) is largely influenced by the dose-volume metrics of the treatment plan, but what are the patient-related and other factors that may independently increase the risk of radiation lung toxicity? Poor pulmonary function has been regarded as a risk factor and a relative contraindication for patients undergoing radical RT, but recent evidence suggests that patients with poor spirometry results may tolerate conventional or high-dose RT as well as, if not better than, patients with normal function. However, caution may need to be exercised in patients with underlying interstitial pulmonary fibrosis. Furthermore, there is emerging evidence of molecular markers of increased risk of toxicity. This review discusses patient-related risk factors other than dosimetry for radiation lung toxicity.

Predicting 2-year survival for radiation regimens in advanced non-small cell lung cancer

AIMS

Total dose, dose per fraction, number of fractions and treatment time are important determinants of the biological effect of a radiation regimen. Several randomised clinical trials (RCTs) have tested a variety of dosing regimens in advanced unresected non-small cell lung cancer, but survival remains poor. This work used past RCT data to develop and validate a predictive model that could help in designing new radiation regimens for successful testing in RCTs.

MATERIALS AND METHODS

Eleven RCTs that compared radiation regimens alone were used to define the relationship between radiation regimens and 2-year survival. On the basis of this relationship, predictive models were developed. Predicted values were internally and externally validated against observed values from the same 11 RCTs and 21 other RCTs. Scatter plots and Pearson's correlation coefficient (r) were used for validation. Finally, regimens were explored that could improve survival.

RESULTS

Increments in the total dose, dose per day and the number of treatment days were associated with improved survival; increments in dose-squared and treatment weeks were associated with reduced survival. The observed and predicted values were similar on internal (r = 0.96) and external validation (r = 0.76). Regimens that delivered a higher total dose over a shorter time had higher survival rates compared with the standard (60 Gy, 30 fractions, 6 weeks); survival may be improved by delivering the standard treatment in 5 weeks rather than 6 weeks.

CONCLUSION

The developed model can predict the effect of thoracic radiation on survival in advanced non-small cell lung cancer patients. It is a useful tool for designing new radiation regimens for clinical trials.

Functional promoter variant rs2868371 of HSPB1 is associated with risk of radiation pneumonitis after chemoradiation for non-small cell lung cancer

PURPOSE

To date, no biomarkers have been found to predict, before treatment, which patients will develop radiation pneumonitis (RP), a potentially fatal toxicity, after chemoradiation for lung cancer. We investigated potential associations between single nucleotide polymorphisms (SNPs) in HSPB1 and risk of RP after chemoradiation for non-small cell lung cancer (NSCLC).

METHODS AND MATERIALS

Subjects were patients with NSCLC treated with chemoradiation at 1 institution. The training data set comprised 146 patients treated from 1999 to July 2004; the validation data set was 125 patients treated from August 2004 to March 2010. We genotyped 2 functional SNPs of HSPB1 (rs2868370 and rs2868371) from all patients. We used Kaplan-Meier analysis to assess the risk of grade ≥2 or ≥3 RP in both data sets and a parametric log-logistic survival model to evaluate the association of HSPB1 genotypes with that risk.

RESULTS

Grade ≥3 RP was experienced by 13% of those with CG/GG and 29% of those with CC genotype of HSPB1 rs2868371 in the training data set (P=.028); corresponding rates in the validation data set were 2% CG/GG and 14% CC (P=.02). Univariate and multivariate analysis confirmed the association of CC of HSPB1 rs2868371 with higher risk of grade ≥3 RP than CG/GG after adjustment for sex, age, performance status, and lung mean dose. This association was validated both in the validation data set and with Harrell's C statistic.

CONCLUSIONS

The CC genotype of HSPB1 rs2868371 was associated with severe RP after chemoradiation for NSCLC.

A literature-based meta-analysis of clinical risk factors for development of radiation induced pneumonitis

INTRODUCTION

The risk of developing side effects after radiotherapy is not only dependent on radiation dose, but may also be affected by patient-related risk factors. Here we perform a literature-based meta-analysis to estimate the effect of various clinical risk factors on the incidence of symptomatic radiation pneumonitis (RP).

MATERIAL AND METHODS

A systematic review of English language articles in the Pubmed, Embase and Cochrane controlled trials registers. Studies with the mesh term "radiation pneumonitis" or the search term "radiation pneumonitis" were included. Additional studies were identified by manual searching of the references. Studies reporting crude incidence or odds ratios (OR) for radiation pneumonitis vs. age, disease location, smoking status, chemotherapy schedule or comorbidity were included. A systematic overview (meta-analysis) was conducted to synthesize data across multiple studies.

RESULTS

Significant risk factors for RP were: older age (OR = 1.7, p < 0.0001); disease located in mid-lower lung (OR = 1.9, p = 0.002); presence of comorbidity (OR = 2.3, p = 0.007). Ongoing smoking was found to protect against RP (OR = 0.6, p = 0.008). History of smoking tended to protect against RP (OR = 0.7, p = 0.06). Sequential (rather than concomitant) chemotherapy scheduling (OR = 1.6, p = 0.01) increased RP risk, but treatment intensity and patients selection are likely confounders.

CONCLUSION

This systematic overview revealed several clinical risk factors for RP that have not been unambiguously identified in the literature. These risk factors should be considered when defining dose-volume constraints for radiation treatment plan optimization.

An improved model for predicting radiation pneumonitis incorporating clinical and dosimetric variables

PURPOSE

Single dose-volume metrics are of limited value for the prediction of radiation pneumonitis (RP) in day-to-day clinical practice. We investigated whether multiparametric models that incorporate clinical and physiologic factors might have improved accuracy.

METHODS AND MATERIALS

The records of 160 patients who received radiation therapy for non-small-cell lung cancer were reviewed. All patients were treated to the same dose and with an identical technique. Dosimetric, pulmonary function, and clinical parameters were analyzed to determine their ability to predict for the subsequent development of RP.

RESULTS

Twenty-seven patients (17%) developed RP. On univariate analysis, the following factors were significantly correlated with the risk of pneumonitis: fractional volume of lung receiving >5-20 Gy, absolute volume of lung spared from receiving >5-15 Gy, mean lung dose, craniocaudal position of the isocenter, transfer coefficient for carbon monoxide (KCOc), total lung capacity, coadministration of angiotensin converting enzyme inhibitors, and coadministration of angiotensin receptor antagonists. By combining the absolute volume of lung spared from receiving >5 Gy with the KCOc, we defined a new parameter termed Transfer Factor Spared from receiving >5 Gy (TFS(5)). The area under the receiver operator characteristic curve for TFS(5) was 0.778, increasing to 0.846 if patients receiving modulators of the renin-angiotensin system were excluded from the analysis. Patients with a TFS(5) <2.17 mmol/min/kPa had a risk of RP of 30% compared with 5% for the group with a TFS(5) ≥ 2.17.

CONCLUSIONS

TFS(5) represents a simple parameter that can be used in routine clinical practice to more accurately segregate patients into high- and low-risk groups for developing RP.

ATM polymorphisms are associated with risk of radiation-induced pneumonitis

PURPOSE

Since the ataxia telangiectasia mutated (ATM) protein plays crucial roles in repair of double-stranded DNA breaks, control of cell cycle checkpoints, and radiosensitivity, we hypothesized that variations in this gene might be associated with radiation-induced pneumonitis (RP).

METHODS AND MATERIALS

A total of 253 lung cancer patients receiving thoracic irradiation between 2004 and 2006 were included in this study. Common Terminology Criteria for Adverse Events version 3.0 was used to grade RP. Five haplotype-tagging single nucleotide polymorphisms (SNPs) in the ATM gene were genotyped using DNA from blood lymphocytes. Hazard ratios (HRs) and 95% confidence intervals (CIs) of RP for genotypes were computed by the Cox model, adjusted for clinical factors. The function of the ATM SNP associated with RP was examined by biochemical assays.

RESULTS

During the median 22-month follow-up, 44 (17.4%) patients developed grade > or = 2 RP. In multivariate Cox regression models adjusted for other clinical predictors, we found two ATM variants were independently associated with increased RP risk. They were an 111G > A) polymorphism (HR, 2.49; 95% CI, 1.07-5.80) and an ATM 126713G > A polymorphism (HR, 2.47; 95% CI, 1.16-5.28). Furthermore, genotype-dependent differences in ATM expression were demonstrated both in cell lines (p < 0.001) and in individual lung tissue samples (p = 0.003), which supported the results of the association study.

CONCLUSIONS

Genetic polymorphisms of ATM are significantly associated with RP risk. These variants might exert their effect through regulation of ATM expression and serve as independent biomarkers for prediction of RP in patients treated with thoracic radiotherapy.

A genetic algorithm for variable selection in logistic regression analysis of radiotherapy treatment outcomes

A given outcome of radiotherapy treatment can be modeled by analyzing its correlation with a combination of dosimetric, physiological, biological, and clinical factors, through a logistic regression fit of a large patient population. The quality of the fit is measured by the combination of the predictive power of this particular set of factors and the statistical significance of the individual factors in the model. We developed a genetic algorithm (GA), in which a small sample of all the possible combinations of variables are fitted to the patient data. New models are derived from the best models, through crossover and mutation operations, and are in turn fitted. The process is repeated until the sample converges to the combination of factors that best predicts the outcome. The GA was tested on a data set that investigated the incidence of lung injury in NSCLC patients treated with 3DCRT. The GA identified a model with two variables as the best predictor of radiation pneumonitis: the V30 (p=0.048) and the ongoing use of tobacco at the time of referral (p=0.074). This two-variable model was confirmed as the best model by analyzing all possible combinations of factors. In conclusion, genetic algorithms provide a reliable and fast way to select significant factors in logistic regression analysis of large clinical studies.

The importance of patient characteristics for the prediction of radiation-induced lung toxicity

PURPOSE

Extensive research has led to the identification of numerous dosimetric parameters as well as patient characteristics, associated with lung toxicity, but their clinical usefulness remains largely unknown. We investigated the predictive value of patient characteristics in combination with established dosimetric parameters.

PATIENTS AND METHODS

Data from 438 lung cancer patients treated with (chemo)radiation were used. Lung toxicity was scored using the Common Toxicity Criteria version 3.0. A multivariate model as well as two single parameter models, including either V(20) or MLD, was built. Performance of the models was expressed as the AUC (Area Under the Curve).

RESULTS

The mean MLD was 13.5 Gy (SD 4.5 Gy), while the mean V(20) was 21.0% (SD 7.3%). Univariate models with V(20) or MLD both yielded an AUC of 0.47. The final multivariate model, which included WHO-performance status, smoking status, forced expiratory volume (FEV(1)), age and MLD, yielded an AUC of 0.62 (95% CI: 0.55-0.69).

CONCLUSIONS

Within the range of radiation doses used in our clinic, dosimetric parameters play a less important role than patient characteristics for the prediction of lung toxicity. Future research should focus more on patient-related factors, as opposed to dosimetric parameters, in order to identify patients at high risk for developing radiation-induced lung toxicity more accurately.

Combining multiple models to generate consensus: application to radiation-induced pneumonitis prediction

The fusion of predictions from disparate models has been used in several fields to obtain a more realistic and robust estimate of the "ground truth" by allowing the models to reinforce each other when consensus exists, or, conversely, negate each other when there is no consensus. Fusion has been shown to be most effective when the models have some complementary strengths arising from different approaches. In this work, we fuse the results from four common but methodologically different nonlinear multivariate models (Decision Trees, Neural Networks, Support Vector Machines, Self-Organizing Maps) that were trained to predict radiation-induced pneumonitis risk on a database of 219 lung cancer patients treated with radiotherapy (34 with Grade 2+ postradiotherapy pneumonitis). Each model independently incorporated a small number of features from the available set of dose and nondose patient variables to predict pneumonitis; no two models had all features in common. Fusion was achieved by simple averaging of the predictions for each patient from all four models. Since a model's prediction for a patient can be dependent on the patient training set used to build the model, the average of several different predictions from each model was used in the fusion (predictions were made by repeatedly testing each patient with a model built from different cross-validation training sets that excluded the patient being tested). The area under the receiver operating characteristics curve for the fused cross-validated results was 0.79, with lower variance than the individual component models. From the fusion, five features were extracted as the consensus among all four models in predicting radiation pneumonitis. Arranged in order of importance, the features are (1) chemotherapy; (2) equivalent uniform dose (EUD) for exponent a=1.2 to 3; (3) EUD for a=0.5 to 1.2, lung volume receiving >20-30 Gy; (4) female sex; and (5) squamous cell histology. To facilitate ease of interpretation and prospective use, the fused outcome results for the patients were fitted to a logistic probability function.

Using patient data similarities to predict radiation pneumonitis via a self-organizing map

This work investigates the use of the self-organizing map (SOM) technique for predicting lung radiation pneumonitis (RP) risk. SOM is an effective method for projecting and visualizing high-dimensional data in a low-dimensional space (map). By projecting patients with similar data (dose and non-dose factors) onto the same region of the map, commonalities in their outcomes can be visualized and categorized. Once built, the SOM may be used to predict pneumonitis risk by identifying the region of the map that is most similar to a patient's characteristics. Two SOM models were developed from a database of 219 lung cancer patients treated with radiation therapy (34 clinically diagnosed with Grade 2+ pneumonitis). The models were: SOM(all) built from all dose and non-dose factors and, for comparison, SOM(dose) built from dose factors alone. Both models were tested using ten-fold cross validation and Receiver Operating Characteristics (ROC) analysis. Models SOM(all) and SOM(dose) yielded ten-fold cross-validated ROC areas of 0.73 (sensitivity/specificity = 71%/68%) and 0.67 (sensitivity/specificity = 63%/66%), respectively. The significant difference between the cross-validated ROC areas of these two models (p < 0.05) implies that non-dose features add important information toward predicting RP risk. Among the input features selected by model SOM(all), the two with highest impact for increasing RP risk were: (a) higher mean lung dose and (b) chemotherapy prior to radiation therapy. The SOM model developed here may not be extrapolated to treatment techniques outside that used in our database, such as several-field lung intensity modulated radiation therapy or gated radiation therapy.

Investigation of the support vector machine algorithm to predict lung radiation-induced pneumonitis

The purpose of this study is to build and test a support vector machine (SVM) model to predict for the occurrence of lung radiation-induced Grade 2+ pneumonitis. SVM is a sophisticated statistical technique capable of separating the two categories of patients (with/without pneumonitis) using a boundary defined by a complex hypersurface. Despite the complexity, the SVM boundary is only minimally influenced by outliers that are difficult to separate. By contrast, the simple hyperplane boundary computed by the more commonly used and related linear discriminant analysis method is heavily influenced by outliers. Two SVM models were built using data from 219 patients with lung cancer treated using radiotherapy (34 diagnosed with pneumonitis). One model (SVM(all)) selected input features from all dose and non-dose factors. For comparison, the other model (SVM(dose)) selected input features only from lung dose-volume factors. Model predictive ability was evaluated using ten-fold cross-validation and receiver operating characteristics (ROC) analysis. For the model SVM(all), the area under the cross-validated ROC curve was 0.76 (sensitivity/specificity = 74%/75%). Compared to the corresponding SVM(dose) area of 0.71 (sensitivity/specificity = 68%/68%), the predictive ability of SVM(all) was improved, indicating that non-dose features are important contributors to separating patients with and without pneumonitis. Among the input features selected by model SVM(all), the two with highest importance for predicting lung pneumonitis were: (a) generalized equivalent uniform doses close to the mean lung dose, and (b) chemotherapy prior to radiotherapy. The model SVM(all) is publicly available via internet access.