Assessment of the Prognostic Value of Radiomic Features in 18F-FMISO PET Imaging of Hypoxia in Postsurgery Brain Cancer Patients: Secondary Analysis of Imaging Data from a Single-Center Study and the Multicenter ACRIN 6684 Trial.

Hypoxia is associated with resistance to radiotherapy and chemotherapy in malignant gliomas, and it can be imaged by positron emission tomography with 18F-fluoromisonidazole (18F-FMISO). Previous results for patients with brain cancer imaged with 18F-FMISO at a single center before conventional chemoradiotherapy showed that tumor uptake via T/Bmax (tissue SUVmax/blood SUV) and hypoxic volume (HV) was associated with poor survival. However, in a multicenter clinical trial (ACRIN 6684), traditional uptake parameters were not found to be prognostically significant, but tumor SUVpeak did predict survival at 1 year. The present analysis considered both study cohorts to reconcile key differences and examine the potential utility of adding radiomic features as prognostic variables for outcome prediction on the combined cohort of 72 patients with brain cancer (30 University of Washington and 42 ACRIN 6684). We used both 18F-FMISO intensity metrics (T/Bmax, HV, SUV, SUVmax, SUVpeak) and assessed radiomic measures that determined first-order (histogram), second-order, and higher-order radiomic features of 18F-FMISO uptake distributions. A multivariate model was developed that included age, HV, and the intensity of 18F-FMISO uptake. HV and SUVpeak were both independent predictors of outcome for the combined data set (P < .001) and were also found significant in multivariate prognostic models (P < .002 and P < .001, respectively). Further model selection that included radiomic features showed the additional prognostic value for overall survival of specific higher order texture features, leading to an increase in relative risk prediction performance by a further 5%, when added to the multivariate clinical model..

Combining structural and textural assessments of volumetric FDG-PET uptake in NSCLC.

Numerous studies have reported the prognostic utility of texture analyses and the effectiveness of radiomics in PET and PET/CT assessment of non-small cell lung cancer (NSCLC). Here we explore the potential, relative to this methodology, of an alternative model-based approach to tumour characterization, which was successfully applied to sarcoma in previous works. The spatial distribution of 3D FDG-PET uptake is evaluated in the spatial referential determined by the best-fitting ellipsoidal pattern, which provides a univariate uptake profile function of the radial position of intratumoral voxels. A group of structural features is extracted from this fit that include two heterogeneity variables and statistical summaries of local metabolic gradients. We demonstrate that these variables capture aspects of tumour metabolism that are separate to those described by conventional texture features. Prognostic model selection is performed in terms of a number of classifiers, including stepwise selection of logistic models, LASSO, random forests and neural networks with respect to two-year survival status. Our results for a cohort of 93 NSCLC patients show that structural variables have significant prognostic potential, and that they may be used in conjunction with texture features in a traditional radiomics sense, towards improved baseline multivariate models of patient overall survival. The statistical significance of these models also demonstrates the relevance of these machine learning classifiers for prognostic variable selection.

Positron emission tomography-based assessment of metabolic gradient and other prognostic features in sarcoma.

Intratumoral heterogeneity biomarkers derived from positron emission tomography (PET) imaging with fluorodeoxyglucose (FDG) are of interest for a number of cancers, including sarcoma. A range of radiomic texture variables, adapted from general methodologies for image analysis, has shown promise in the setting. In the context of sarcoma, our group introduced an alternative model-based approach to the measurement of heterogeneity. In this approach, the heterogeneity of a tumor is characterized by the extent to which the 3-D FDG uptake pattern deviates from a simple elliptically contoured structure. By using a nonparametric analysis of the uptake profile obtained from this spatial model, a variable assessing the metabolic gradient of the tumor is developed. The work explores the prognostic potential of this new variable in the context of FDG-PET imaging of sarcoma. A mature clinical series involving 197 patients, 88 of whom have complete time-to-death information, is used. Texture variables based on the imaging data are also evaluated in this series and a range of appropriate machine learning methodologies are then used to explore the complementary prognostic roles for structure and texture variables. We conclude that both texture-based and model-based variables can be combined to achieve enhanced prognostic assessments of outcome for patients with sarcoma based on FDG-PET imaging information.

Prospective Study of Serial 18F-FDG PET and 18F-Fluoride PET to Predict Time to Skeletal-Related Events, Time to Progression, and Survival in Patients with Bone-Dominant Metastatic Breast Cancer.

Assessing therapy response of breast cancer bone metastases is challenging. In retrospective studies, serial 18F-FDG PET was predictive of time to skeletal-related events (tSRE) and time to progression (TTP). 18F-NaF PET improves bone metastasis detection compared with bone scanning. We prospectively tested 18F-FDG PET and 18F-NaF PET to predict tSRE, TTP, and overall survival (OS) in patients with bone-dominant metastatic breast cancer (MBC). Methods: Patients with bone-dominant MBC were imaged with 18F-FDG PET and 18F-NaF PET before starting new therapy (scan1) and again at a range of times centered around approximately 4 mo later (scan2). Maximum standardized uptake value (SUVmax) and lean body mass adjusted standardized uptake (SULpeak) were recorded for a single index lesion and up to 5 most dominant lesions for each scan. tSRE, TTP, and OS were assessed exclusive of the PET images. Univariate Cox regression was performed to test the association between clinical endpoints and 18F-FDG PET and 18F-NaF PET measures. mPERCIST (Modified PET Response Criteria in Solid Tumors) were also applied. Survival curves for mPERCIST compared response categories of complete response+partial response+stable disease versus progressive disease for tSRE, TTP, and OS. Results: Twenty-eight patients were evaluated. Higher 18F-FDG SULpeak at scan2 predicted shorter time to tSRE (P = <0.001) and TTP (P = 0.044). Higher 18F-FDG SUVmax at scan2 predicted a shorter time to tSRE (P = <0.001). A multivariable model using 18F-FDG SUVmax of the index lesion at scan1 plus the difference in SUVmax of up to 5 lesions between scans was predictive for tSRE and TTP. Among 24 patients evaluable by 18F-FDG PET mPERCIST, tSRE and TTP were longer in responders (complete response, partial response, or stable disease) than in nonresponders (progressive disease) (P = 0.007, 0.028, respectively), with a trend toward improved survival (P = 0.1). An increase in the uptake between scans of up to 5 lesions by 18F-NaF PET was associated with longer OS (P = 0.027). Conclusion: Changes in 18F-FDG PET parameters during therapy are predictive of tSRE and TTP, but not OS. mPERCIST evaluation in bone lesions may be useful in assessing response to therapy and is worthy of evaluation in multicenter, prospective trials. Serial 18F-NaF PET was associated with OS but was not useful for predicting TTP or tSRE in bone-dominant MBC.

Assessment of a statistical AIF extraction method for dynamic PET studies with 15O water and 18F fluorodeoxyglucose in locally advanced breast cancer patients.

Blood flow-metabolism mismatch from dynamic positron emission tomography (PET) studies with [Formula: see text]-labeled water ([Formula: see text]) and [Formula: see text]-labeled fluorodeoxyglucose (FDG) has been shown to be a promising diagnostic for locally advanced breast cancer (LABCa) patients. The mismatch measurement involves kinetic analysis with the arterial blood time course (AIF) as an input function. We evaluate the use of a statistical method for AIF extraction (SAIF) in these studies. Fifty three LABCa patients had dynamic PET studies with [Formula: see text] and FDG. For each PET study, two AIFs were recovered, an SAIF extraction and also a manual extraction based on a region of interest placed over the left ventricle (LV-ROI). Blood flow-metabolism mismatch was obtained with each AIF, and kinetic and prognostic reliability comparisons were made. Strong correlations were found between kinetic assessments produced by both AIFs. SAIF AIFs retained the full prognostic value, for pathologic response and overall survival, of LV-ROI AIFs.

18F-Fluoromisonidazole Quantification of Hypoxia in Human Cancer Patients Using Image-Derived Blood Surrogate Tissue Reference Regions.

UNLABELLED: (18)F-fluoromisonidazole ((18)F-FMISO) is the most widely used PET agent for imaging hypoxia, a condition associated with resistance to tumor therapy. (18)F-FMISO equilibrates in normoxic tissues but is retained under hypoxic conditions because of reduction and binding to macromolecules. A simple tissue-to-blood (TB) ratio is suitable for quantifying hypoxia. A TB ratio threshold of 1.2 or greater is useful in discriminating the hypoxic volume (HV) of tissue; TBmax is the maximum intensity of the hypoxic region and does not invoke a threshold. Because elimination of blood sampling would simplify clinical use, we tested the validity of using imaging regions as a surrogate for blood sampling. METHODS: Patients underwent 20-min (18)F-FMISO scanning during the 90- to 140-min interval after injection with venous blood sampling. Two hundred twenty-three (18)F-FMISO patient studies had detectable surrogate blood regions in the field of view. Quantitative parameters of hypoxia (TBmax, HV) derived from blood samples were compared with values using surrogate blood regions derived from the heart, aorta, or cerebellum. In a subset of brain cancer patients, parameters from blood samples and from the cerebellum were compared for their ability to independently predict outcome. RESULTS: Vascular regions of heart showed the highest correlation to measured blood activity (R(2) = 0.84). For brain studies, cerebellar activity was similarly correlated to blood samples. In brain cancer patients, Kaplan-Meier analysis showed that image-derived reference regions had predictive power nearly identical to parameters derived from blood, thus obviating the need for venous sampling in these patients. CONCLUSION: Simple static analysis of (18)F-FMISO PET captures both the intensity (TBmax) and the spatial extent (HV) of tumor hypoxia. An image-derived region to assess blood activity can be used as a surrogate for blood sampling in quantification of hypoxia.

Sarcoma mid-therapy [F-18]fluorodeoxyglucose positron emission tomography (FDG PET) and patient outcome.

BACKGROUND: Our previous research investigated the ability of [F-18]fluorodeoxyglucose (FDG) positron emission tomography (PET) imaging results to predict outcome in patients with sarcoma. Tumor uptake of FDG before and after neoadjuvant chemotherapy was predictive of patient outcome. With this background, a prospective clinical study was designed to assess whether tumor FDG uptake levels in the middle of neoadjuvant chemotherapy added additional prognostic information to pre-therapy imaging data. METHODS: Sixty-five patients with either bone or soft-tissue sarcoma were treated with neoadjuvant-based chemotherapy according to the standard clinical practice for each tumor group. All patients had FDG PET studies before therapy, mid-therapy (after two cycles of chemotherapy), and before resection. Tumor FDG uptake (SUVmax, the maximum standardized uptake value) at each imaging time point, tumor type (bone or soft-tissue sarcoma), tumor size, and histopathologic grade were recorded for each patient. The time from the pre-therapy FDG PET study to events of local tumor recurrence, metastasis, or death were extracted from the clinical records for comparison with the imaging data. Univariate and multivariate analyses of the imaging and clinical data were performed. RESULTS: Univariate and multivariate data analyses showed that the difference (measured as the percentage reduction) between the pre-therapy and mid-therapy maximum tumor uptake values added prognostic value to patient outcome predictions independently of other patient variables. CONCLUSIONS: The utility of a tumor pre-therapy FDG PET scan as a biomarker for the outcome of patients with sarcoma was strengthened by a mid-therapy scan to evaluate the interim treatment response.

Nonparametric Residue Analysis of Dynamic PET Data With Application to Cerebral FDG Studies in Normals.

Kinetic analysis is used to extract metabolic information from dynamic positron emission tomography (PET) uptake data. The theory of indicator dilutions, developed in the seminal work of Meier and Zierler (1954), provides a probabilistic framework for representation of PET tracer uptake data in terms of a convolution between an arterial input function and a tissue residue. The residue is a scaled survival function associated with tracer residence in the tissue. Nonparametric inference for the residue, a deconvolution problem, provides a novel approach to kinetic analysis-critically one that is not reliant on specific compartmental modeling assumptions. A practical computational technique based on regularized cubic B-spline approximation of the residence time distribution is proposed. Nonparametric residue analysis allows formal statistical evaluation of specific parametric models to be considered. This analysis needs to properly account for the increased flexibility of the nonparametric estimator. The methodology is illustrated using data from a series of cerebral studies with PET and fluorodeoxyglucose (FDG) in normal subjects. Comparisons are made between key functionals of the residue, tracer flux, flow, etc., resulting from a parametric (the standard two-compartment of Phelps et al. 1979) and a nonparametric analysis. Strong statistical evidence against the compartment model is found. Primarily these differences relate to the representation of the early temporal structure of the tracer residence-largely a function of the vascular supply network. There are convincing physiological arguments against the representations implied by the compartmental approach but this is the first time that a rigorous statistical confirmation using PET data has been reported. The compartmental analysis produces suspect values for flow but, notably, the impact on the metabolic flux, though statistically significant, is limited to deviations on the order of 3%-4%. The general advantage of the nonparametric residue analysis is the ability to provide a valid kinetic quantitation in the context of studies where there may be heterogeneity or other uncertainty about the accuracy of a compartmental model approximation of the tissue residue.

Fluorodeoxyglucose positron emission tomography in leiomyosarcoma: imaging characteristics.

OBJECTIVE: Leiomyosarcoma, a malignant neoplasm of smooth muscle, accounts for 7% of the sarcomas. Patients with leiomyosarcoma tumors have an average survival of 5 years. These tumors, which are derived from mesenchymal tissues, are difficult to diagnose, and treatment options remain controversial. The relatively rare incidence of this soft tissue sarcoma subtype has limited the number of patients available for studies and research. This study examines whether the imaging characteristics of positron emission tomography (PET) with radiolabeled fluorodeoxyglucose (FDG) provide a reliable, noninvasive means to predict tumor behavior in patients with leiomyosarcomas. METHODS: [18F]-FDG-PET was performed on the tumors of participating patients before the neoadjuvant chemotherapy or resection, and a maximum tumor standard uptake value (SUVmax) was calculated. RESULTS: The SUVmax was correlated with tumor grade (P=0.001) and tumor size as greatest dimension (P=0.004). Analysis of these data indicated the potential effectiveness of FDG-PET imaging in predicting tumor grade. CONCLUSION: In leiomyosarcoma, the SUVmax from FDG-PET is a likely predictor of tumor behavior. The results of this study suggest that a large (by greatest dimension) intermediate grade tumor is expected to have the same predicted outcome as a high-grade tumor and should be treated in the same manner, as they share the same prognosis by definition of tumor grade. Improvements made in the clinical treatment of leiomyosarcomas by use of FDG-PET imaging data may lead to an increase in patient survival.

Risk assessment based on FDG-PET imaging in patients with synovial sarcoma.

UNLABELLED: Synovial sarcoma generally is associated with poor prognosis. With recent advances in molecular biology, it has become apparent not all synovial sarcomas share the same tumor biology. (18)F-fluorodeoxyglucose positron emission tomography (FDG-PET) is useful for risk assessment in several types of sarcomas. We therefore assessed the clinical value of (18)F-FDG-PET-derived maximum standard uptake value (SUV(max)) for predicting survival in patients with synovial sarcoma. (18)F-FDG-PET was performed in 44 patients with synovial sarcoma before therapy and resection. SUV(max) was calculated for each tumor and then evaluated for prognostic usefulness along with metastasis at presentation, tumor grade, histopathologic subtype, age, gender, postsurgical margins, anatomic location, and tumor size for overall survival and progression-free survival. SUV(max) ranged from 1.2 to 13.0 (median, 4.35). Pretherapy tumor SUV(max) predicted overall survival and progression-free survival. Patients presenting with a SUV(max) greater than 4.35 had a decreased disease-free survival and were therefore at high risk for having local recurrences and metastatic disease. LEVEL OF EVIDENCE: Level I, diagnostic study. See the Guidelines for Authors for a complete description of levels of evidence.

Spatial heterogeneity in sarcoma 18F-FDG uptake as a predictor of patient outcome.

UNLABELLED: (18)F-FDG PET images of tumors often display highly heterogeneous spatial distribution of (18)F-FDG-positive pixels. We proposed that this heterogeneity in (18)F-FDG spatial distribution can be used to predict tumor biologic aggressiveness. This study presents data to support the hypothesis that a new heterogeneity-analysis algorithm applied to (18)F-FDG PET images of tumors in patients is predictive of patient outcome. METHODS: (18)F-FDG PET images from 238 patients with sarcoma were analyzed using a new algorithm for heterogeneity analysis in tumor (18)F-FDG spatial distribution. Patient characteristics, tumor histology, and patient outcome were compared with image analysis results using univariate and multivariate analysis. Cox proportional hazards models were used to further analyze the significance of the data associations. RESULTS: Statistical analyses show that heterogeneity analysis is a strong independent predictor of patient outcome. CONCLUSION: The new (18)F-FDG PET tumor image heterogeneity analysis method is validated for the ability to predict patient outcome in a clinical population of patients with sarcoma. This method can be extended to other PET image datasets in which heterogeneity in tissue uptake of a radiotracer may predict patient outcome.

Tumor hypoxia imaging with [F-18] fluoromisonidazole positron emission tomography in head and neck cancer.

PURPOSE: Advanced head and neck cancer shows hypoxia that results in biological changes to make the tumor cells more aggressive and less responsive to treatment resulting in poor survival. [F-18] fluoromisonidazole (FMISO) positron emission tomography (PET) has the ability to noninvasively quantify regional hypoxia. We investigated the prognostic effect of pretherapy FMISO-PET on survival in head and neck cancer. EXPERIMENTAL DESIGN: Seventy-three patients with head and neck cancer had pretherapy FMISO-PET and 53 also had fluorodeoxyglucose (FDG) PET under a research protocol from April 1994 to April 2004. RESULTS: Significant hypoxia was identified in 58 patients (79%). The mean FMISO tumor/bloodmax (T/Bmax) was 1.6 and the mean hypoxic volume (HV) was 40.2 mL. There were 28 deaths in the follow-up period. Mean FDG standard uptake value (SUV)max was 10.8. The median time for follow-up was 72 weeks. In a univariate analysis, T/Bmax (P=0.002), HV (P=0.04), and the presence of nodes (P=0.01) were strong independent predictors. In a multivariate analysis, including FDG SUVmax, no variable was predictive at P<0.05. When FDG SUVmax was removed from the model (resulting in n=73 with 28 events), nodal status and T/Bmax (or HV) were both highly predictive (P=0.02, 0.006 for node and T/Bmax, respectively; P=0.02 and 0.001 for node and HV, respectively). CONCLUSIONS: Pretherapy FMISO uptake shows a strong trend to be an independent prognostic measure in head and neck cancer.

Modelling multivariate biomechanical measurements of the spine during a rowing exercise.

OBJECTIVE: To investigate the ability of statistical techniques to detect systematic changes in rowing technique during a rowing session and to discriminate between rowers of different abilities with and without back pain. DESIGN: Statistical techniques were applied to kinematic datasets of elite level rowers, in order to construct an empirical model of the rowing stroke. BACKGROUND: The size and complexity of datasets generated by biomechanical kinematics evaluations has led to opportunities for analysing pathology whilst introducing substantial challenges for statistical analysis. METHODS: Spinal motion and load output of 18 International and National standard competitive rowers were monitored during ergometer rowing sessions. International rower data were used to construct an empirical model of this activity. Linear stroke models were derived using principal components and a generalized cross-validation procedure. Performance characteristics of the identified models were calculated for all rowing groups. The stroke model was applied to distinguishing pattern variations within and between rowers. A multivariate logistic regression analysis was carried out to examine the relationship between stroke model parameters on the incidence of low back pain. RESULTS: 90% of the variability in the data was explained by the first three principal component variables. Stroke models with three basis functions were selected for each variable. The models performed well on the National rowers, providing validation of the models. A 2-variable model showed a significant difference between the rowing stroke characteristics of rowers with and without low back pain (P<0.01). CONCLUSIONS: A parsimonious collection of empirical models effectively describes motion and load characteristics of ergometer rowing. Patterns in rowing technique are found to be strongly associated with the incidence lower back pain. RELEVANCE: Empirical statistical models can be used to track changes in rowing technique, and discriminate between different rowing groups. This may impact rowing training, and rehabilitation.

Factorial design considerations.

PURPOSE: Factorial designs may be proposed to test extra questions within a clinical trial. A common approach to sample size and analysis for factorial trials assumes no statistical interactions and does not adjust for multiple testing. This investigation considered the trade-off between potential gains from testing more questions with fewer patients versus how often a factorial trial might arrive at an incorrect conclusion. METHODS: A simulation study of a 2 x 2 design (observation v chemotherapy v radiation therapy v the combination) was performed under various conditions, including effect of one, both, or neither treatment and absence or presence of statistical interaction (effect of one treatment differed according to the presence of the other). Three analysis approaches were investigated, one assuming no interaction, a second testing first for interaction, and the third testing for interaction as well as adjusting for multiple testing. The approaches were compared with respect to the probability of selecting the correct treatment arm. RESULTS: No one approach was superior. Testing for interaction was beneficial in some settings but detrimental in others. Under some scenarios, the factorial design improved efficiency, but under others, all three approaches resulted in poor probability of selecting the correct treatment arm at the end of the trial. CONCLUSION: Extra efficiency is possible, but it is difficult to predict when favorable conditions exist. If a factorial design is used, potential efficiency gains should be weighed against potential loss of power to arrive at the correct conclusion under possible scenarios of interest.