PET image reconstruction methods include modeling of resolution degrading phenomena, often referred to as point-spread function (PSF) reconstruction. The aim of this study was to develop a clinically relevant phantom and characterize the reproducibility and accuracy of high-resolution PSF reconstructed images of small lesions, which is a prerequisite for using PET in the prediction and evaluation of responses to treatment. Sets of small homogeneous 18F-spheres (range 3–12 mm diameter, relevant for small lesions and lymph nodes) were suspended and covered by a 11C-silicone, which provided a scattering medium and a varying sphere-to-background ratio. Repeated measurements were made on PET/CT scanners from two vendors using a wide range of reconstruction parameters. Recovery coefficients (RCs) were measured for clinically used volume-of-interest definitions. For non-PSF images, RCs were reproducible and fell monotonically as the sphere diameter decreased, which is the expected behavior. PSF images converged slower and had artifacts: RCs did not fall monotonically as sphere diameters decreased but had a maximum RC for sphere sizes around 8 mm, RCs could be greater than 1, and RCs were less reproducible. To some degree, post-reconstruction filters could suppress PSF artifacts. High-resolution PSF images of small lesions showed artifacts that could lead to serious misinterpretations when used for monitoring treatment response. Thus, it could be safer to use non-PSF reconstruction for quantitative purposes unless PSF reconstruction parameters are optimized for the specific task.
Semi-automated quantitative measurement of metabolic tumor volume (MTV) for prognosis in diffuse large B-Cell lymphoma (DLBCL) has gained considerable interest lately. However, simple tumor volume measures may be inadequate for assessment of prognosis in DLBCL as other characteristics such as growth pattern and metabolic heterogeneity may be just as important. In addition, MTV measurements require delineation of tumor lesions by semi-automated software, which can be time-consuming. We hypothesized that a simple visual assessment of tumor volume performs as well as standardized MTV measurements in DLBCL prognostication. Quantitative and visual analyses of pre-therapy 18F-FDG PET/CT scans in 118 patients with newly diagnosed DLBCL were conducted. Quantitative analyses were performed using Hermes TumourFinder® to obtain MTV2.5 (SUV 2.5 cut-off) and MTV41 (41% SUVmax isocontour cut-off). Visual assessments included a binary prediction (good/poor prognosis) as well as tumor burden based on a visual analog scale (MTVVAS) and an estimated volume (eMTV). Three experienced nuclear medicine physicians who were blinded to clinical outcome performed visual evaluations. Progression-free survival was evaluated by Kaplan-Meier curves and log-rank test. Inter-observer variability was evaluated by Fleiss’ kappa for multiple observers. In the quantitative analysis, a ROC-determined MTV2.5 cut-off (log-rank p = 0.11) seemed to outperform MTV41 (log-rank p = 0.76) for PFS prediction. TLG2.5 (log-rank p = 0.14) and TLG41 (log-rank p = 0.34) were not associated with outcomes. By visual analysis, all three reviewers were able to stratify patients into good/poor prognosis (reviewer A log-rank p = 0.002, reviewer B log-rank p = 0.016, and reviewer C log-rank p = 0.012) with fair inter-observer agreement (Fleiss’ kappa 0.47). MTVVAS and eMTV were not consistently correlated with the outcome. Predictions of outcome after first-line treatment for DLBCL were surprisingly good when left to the unsupervised, subjective judgment of experienced readers of lymphoma 18F-FDG-PET/CT. The study highlights the importance of non-standardized clinical judgments and shows potential loss of valuable prognostic information when relying solely on semi-automated MTV measurements.