Dataset Information

Radiomics features of the primary tumor fail to improve prediction of overall survival in large cohorts of CT- and PET-imaged head and neck cancer patients.

ABSTRACT: Radiomics studies require many patients in order to power them, thus patients are often combined from different institutions and using different imaging protocols. Various studies have shown that imaging protocols affect radiomics feature values. We examined whether using data from cohorts with controlled imaging protocols improved patient outcome models. We retrospectively reviewed 726 CT and 686 PET images from head and neck cancer patients, who were divided into training or independent testing cohorts. For each patient, radiomics features with different preprocessing were calculated and two clinical variables-HPV status and tumor volume-were also included. A Cox proportional hazards model was built on the training data by using bootstrapped Lasso regression to predict overall survival. The effect of controlled imaging protocols on model performance was evaluated by subsetting the original training and independent testing cohorts to include only patients whose images were obtained using the same imaging protocol and vendor. Tumor volume, HPV status, and two radiomics covariates were selected for the CT model, resulting in an AUC of 0.72. However, volume alone produced a higher AUC, whereas adding radiomics features reduced the AUC. HPV status and one radiomics feature were selected as covariates for the PET model, resulting in an AUC of 0.59, but neither covariate was significantly associated with survival. Limiting the training and independent testing to patients with the same imaging protocol reduced the AUC for CT patients to 0.55, and no covariates were selected for PET patients. Radiomics features were not consistently associated with survival in CT or PET images of head and neck patients, even within patients with the same imaging protocol.

SUBMITTER: Ger RB

PROVIDER: S-EPMC6752873 | biostudies-literature | 2019

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Radiomics features of the primary tumor fail to improve prediction of overall survival in large cohorts of CT- and PET-imaged head and neck cancer patients.

Ger Rachel B RB Zhou Shouhao S Elgohari Baher B Elhalawani Hesham H Mackin Dennis M DM Meier Joseph G JG Nguyen Callistus M CM Anderson Brian M BM Gay Casey C Ning Jing J Fuller Clifton D CD Li Heng H Howell Rebecca M RM Layman Rick R RR Mawlawi Osama O Stafford R Jason RJ Aerts Hugo H Court Laurence E LE

PloS one 20190919 9

Radiomics studies require many patients in order to power them, thus patients are often combined from different institutions and using different imaging protocols. Various studies have shown that imaging protocols affect radiomics feature values. We examined whether using data from cohorts with controlled imaging protocols improved patient outcome models. We retrospectively reviewed 726 CT and 686 PET images from head and neck cancer patients, who were divided into training or independent testin ...[more]

PMID: 31536526

Similar Datasets

Project description:This multi-center study aims to investigate the prognostic value of context-aware saliency-guided radiomics in 18F-FDG PET/CT images of head and neck cancer (HNC). 806 HNC patients (training vs. validation vs. external testing: 500 vs. 97 vs. 209) from 9 centers were collected from The Cancer Imaging Archive (TCIA). There were 100/384 and 60/123 oropharyngeal carcinoma (OPC) patients with human papillomavirus (HPV) status in training and testing cohorts, respectively. Six types of images were used for radiomics feature extraction and further model construction, namely (i) the original image (Origin), (ii) a context-aware saliency map (SalMap), (iii, iv) high- or low-saliency regions in the original image (highSal or lowSal), (v) a saliency-weighted image (SalxImg), and finally, (vi) a fused PET-CT image (FusedImg). Four outcomes were evaluated, i.e., recurrence-free survival (RFS), metastasis-free survival (MFS), overall survival (OS), and disease-free survival (DFS), respectively. Multivariate Cox analysis and logistic regression were adopted to construct radiomics scores for the prediction of outcome (Rad_Ocm) and HPV-status (Rad_HPV), respectively. Besides, the prognostic value of their integration (Rad_Ocm_HPV) was also investigated. In the external testing cohort, compared with the Origin model, SalMap and SalxImg achieved the highest C-indices for RFS (0.621 vs. 0.559) and MFS (0.785 vs. 0.739) predictions, respectively, while FusedImg performed the best for both OS (0.685 vs. 0.659) and DFS (0.641 vs. 0.582) predictions. In the OPC HPV testing cohort, FusedImg showed higher AUC for HPV-status prediction compared with the Origin model (0.653 vs. 0.484). In the OPC testing cohort, compared with Rad_Ocm or Rad_HPV alone, Rad_Ocm_HPV performed the best for OS and DFS predictions with C-indices of 0.702 (p = 0.002) and 0.684 (p = 0.006), respectively. Saliency-guided radiomics showed enhanced performance for both outcome and HPV-status predictions relative to conventional radiomics. The radiomics-predicted HPV status also showed complementary prognostic value.

Project description:To compare the interobserver agreement and degree of confidence in anatomical localisation of lesions using 2-[fluorine-18]fluoro-2-deoxy-D-glucose ((18)F-FDG) positron emission tomography (PET)/computed tomography (CT) and (18)F-FDG PET alone in patients with head and neck tumours. A prospective study of 24 patients (16 male, eight female, median age 59 years) with head and neck tumours was undertaken. (18)F-FDG PET/CT was performed for staging purposes. 2D images were acquired over the head and neck area using a GE Discovery LS PET/CT scanner. (18)F-FDG PET images were interpreted by three independent observers. The observers were asked to localise abnormal (18)F-FDG activity to an anatomical territory and score the degree of confidence in localisation on a scale from 1 to 3 (1=exact region unknown; 2=probable; 3=definite). For all (18)F-FDG-avid lesions, standardised uptake values (SUVs) were also calculated. After 3 weeks, the same exercise was carried out using (18)F-FDG PET/CT images, where CT and fused volume data were made available to observers. The degree of interobserver agreement was measured in both instances. A total of six primary lesions with abnormal (18)F-FDG uptake (SUV range 7.2-22) were identified on (18)F-FDG PET alone and on (18)F-FDG PET/CT. In all, 15 nonprimary tumour sites were identified with (18)F-FDG PET only (SUV range 4.5-11.7), while 17 were identified on (18)F-FDG PET/CT. Using (18)F-FDG PET only, correct localisation was documented in three of six primary lesions, while (18)F-FDG PET/CT correctly identified all primary sites. In nonprimary tumour sites, (18)F-FDG PET/CT improved the degree of confidence in anatomical localisation by 51%. Interobserver agreement in assigning primary and nonprimary lesions to anatomical territories was moderate using (18)F-FDG PET alone (kappa coefficients of 0.45 and 0.54, respectively), but almost perfect with (18)F-FDG PET/CT (kappa coefficients of 0.90 and 0.93, respectively). We conclude that (18)F-FDG PET/CT significantly increases interobserver agreement and confidence in disease localisation of (18)F-FDG-avid lesions in patients with head and neck cancers.

Project description:Background: Radiomics has been widely used to non-invasively mine quantitative information from medical images and could potentially predict tumor phenotypes. Pathologic grade is considered a predictive prognostic factor for head and neck squamous cell carcinoma (HNSCC) patients. A preoperative histological assessment can be important in the clinical management of patients. We applied radiomics analysis to devise non-invasive biomarkers and accurately differentiate between well-differentiated (WD) and moderately differentiated (MD) and poorly differentiated (PD) HNSCC. Methods: This study involved 206 consecutive HNSCC patients (training cohort: n = 137; testing cohort: n = 69). In total, we extracted 670 radiomics features from contrast-enhanced computed tomography (CT) images. Radiomics signatures were constructed with a kernel principal component analysis (KPCA), random forest classifier and a variance-threshold (VT) selection. The associations between the radiomics signatures and HNSCC histological grades were investigated. A clinical model and combined model were also constructed. Areas under the receiver operating characteristic curves (AUCs) were applied to evaluate the performances of the three models. Results: In total, 670 features were selected by the KPCA and random forest methods from the CT images. The radiomics signatures had a good performance in discriminating between the two cohorts of HNSCC grades, with an AUC of 0.96 and an accuracy of 0.92. The specificity, accuracy, sensitivity, positive predictive value (PPV), and negative predictive value (NPV) of the abovementioned method with a VT selection for determining HNSCC grades were 0.83, 0.92, 0.96, 0.94, and 0.91, respectively; without VT, the corresponding results were 0.70, 0.83, 0.88, 0.80, and 0.84. The differences in accuracy, sensitivity and NPV were significant between these approaches (p < 0.05). The AUCs with VT and without VT were 0.96 and 0.89, respectively (p < 0.05). Compared to the combined model and the radiomics signatures, The clinical model had a worse performance, and the differences were significant (p < 0.05). The combined model had the best performance, but the difference between the combined model and the radiomics signature weren't significant (p > 0.05). Conclusions: The CT-based radiomics signature could discriminate between WD and MD and PD HNSCC and might serve as a biomarker for preoperative grading.

Project description:ImportanceThe mechanisms underlying the association between chronic stress and higher mortality among individuals with cancer remain incompletely understood.ObjectiveTo test the hypotheses that among individuals with active head and neck cancer, that higher stress-associated neural activity (ie. metabolic amygdalar activity [AmygA]) at cancer staging associates with survival.DesignRetrospective cohort study.SettingAcademic Medical Center (Massachusetts General Hospital, Boston).Participants240 patients with head and neck cancer (HNCA) who underwent 18F-FDG-PET/CT imaging as part of initial cancer staging.Measurements18F-FDG uptake in the amygdala was determined by placing circular regions of interest in the right and left amygdalae and measuring the mean tracer accumulation (i.e., standardized uptake value [SUV]) in each region of interest. Amygdalar uptake was corrected for background cerebral activity (mean temporal lobe SUV).ResultsAmong individuals with HNCA (age 59±13 years; 30% female), 67 died over a median follow-up period of 3 years (IQR: 1.7-5.1). AmygA associated with heightened bone marrow activity, leukocytosis, and C-reactive protein (P<0.05 each). In adjusted and unadjusted analyses, AmygA associated with subsequent mortality (HR [95% CI]: 1.35, [1.07-1.70], P = 0.009); the association persisted in stratified subset analyses restricted to patients with advanced cancer stage (P<0.001). Individuals within the highest tertile of AmygA experienced a 2-fold higher mortality rate compared to others (P = 0.01). The median progression-free survival was 25 months in patients with higher AmygA (upper tertile) as compared with 36.5 months in other individuals (HR for progression or death [95%CI], 1.83 [1.24-2.68], P = 0.001).Conclusions and relevanceAmygA, quantified on routine 18F-FDG-PET/CT images obtained at cancer staging, independently and robustly predicts mortality and cancer progression among patients with HNCA. Future studies should test whether strategies that attenuate AmygA (or its downstream biological consequences) may improve cancer survival.

Project description:Background Systemic inflammation is predictive of the overall survival in cancer patients and is related to the density of immune cells in the tumor microenvironment of cancer, which in turn correlates with 18F -fluorodeoxyglucose (FDG)-positron emission tomography/computed tomography (PET/CT) metabolic parameters (MPs). The density of tumor-infiltrating lymphocytes (TILs) in the microenvironment has the potential to be a biomarker that can be used clinically to optimize patient selection in oropharyngeal head and neck squamous cell carcinoma (HNSCC). There is little to no data regarding the association of systemic inflammation with PET/CT-MPs, especially in HNSCC. This study aimed to evaluate the correlation between markers of host inflammation, namely blood neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR), with the PET/CT-MPs standardized uptake value (SUV), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) of the primary tumor, derived from FDG-PET/CT in patients with nonmetastatic (cM0) HNSCC before treatment. We hypothesized that NLR and PLR at baseline are positively correlated with PET/CT-MPs. Methods A retrospective review of consecutive patients with HNSCC with a pretreatment PET/CT was performed. NLR and PLR were computed using complete blood counts measured within 10 days before the start of any treatment. The correlation between NLR and PLR with PET/CT-MPs was evaluated with Spearman's rho test. Results Seventy-one patients were analyzed. Overall survival (OS) at 1, 2, and 3 years was 86%, 76%, and 68%. PLR was found to be correlated with MTV (rho = 0.26, P = .03) and TLG (rho = 0.28, P = .02) but not with maximum SUV or mean SUV. There was no correlation between NLR and the analyzed PET/CT-MPs. TLG was associated with worse survival in uni- and multivariable analysis, but no other PET/CT-MPs were associated with either OS or disease-specific survival (DSS). NLR and PLR were associated with OS and DSS on uni- and multivariable analysis. Conclusions In patients with HNSCC before any treatment such as definitive radio (chemo)therapy or oncologic surgery followed by adjuvant RT, baseline PLR correlated with MTV and TLG but not with SUV. NLR was not correlated with any PET/CT-MPs analyzed in our study. Confirmatory studies are needed, and a potential interaction between tumor microenvironment, host inflammation, and FDG-PET/CT measures warrants further investigation. Supplementary Information The online version contains supplementary material available at 10.1186/s13014-022-02112-4.

Dataset Information

Radiomics features of the primary tumor fail to improve prediction of overall survival in large cohorts of CT- and PET-imaged head and neck cancer patients.

Publications

Radiomics features of the primary tumor fail to improve prediction of overall survival in large cohorts of CT- and PET-imaged head and neck cancer patients.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets