Radiologists and Clinical Trials: Part 2: Practical Statistical Methods for Understanding and Monitoring Independent Reader Performance.

Though many clinical trials rely on medical image evaluations for primary or key secondary endpoints, the methods to monitor reader performance are all too often mired in the legacy use of adjudication rates. If misused, this simple metric can be misleading and sometimes entirely contradictory. Furthermore, attempts to overcome the limitations of adjudication rates using de novo or ad hoc methods often ignore well-established research conducted over the last half-century and can lead to inaccurate conclusions or variable interpretations. Underperforming readers can be missed, expert readers retrained, or worse, replaced. This paper aims to standardize reader performance evaluations using proven statistical methods. Additionally, these methods will describe how to discriminate between scenarios of concern and normal medical interpretation variability. Statistical methods are provided for inter-reader and intra-reader variability and bias, including the adjudicator's bias. Finally, we have compiled guidelines for calculating correct sample sizes, considerations for intra-reader memory recall, and applying alternative designs for independent readers.

Guidelines for quality control of PET/CT scans in a multicenter clinical study.

To date, there is no published detailed checklist with parameters referencing the DICOM tag information with respect to the quality control (QC) of PET/CT scans. The aims of these guidelines are to provide the know-how for effectively controlling the quality of PET/CT scans in multicenter studies, to standardize the QC, to give sponsors and regulatory agencies a basis for justification of the data quality when using standardized uptake values as an imaging biomarker, to document the compliance with the imaging guidelines, to verify the per protocol population versus intent to treat population, and to safeguard the validity of multicenter study conclusions employing standardized uptake value (SUV) as an imaging biomarker which is paramount to the scientific community. Following the proposed guidelines will ensure standardized prospective imaging QC of scans applicable to most studies where SUVs are used as an imaging biomarker. The multitude of factors affecting SUV measurements when not controlled inflicts noise on the data. Decisions on patient management with substantial noise would be devastating to patients, ultimately undermine treatment outcome, and invalidate the utility of SUV as an imaging biomarker usefulness. Strict control of the data quality used for the validation of SUV as an imaging biomarker would ensure trust and reliability of the data.

Retrospective quality control review of FDG scans in the imaging sub-study of PALETTE EORTC 62072/VEG110727: a randomized, double-blind, placebo-controlled phase III trial.

PURPOSE: (18)F-Labelled fluorodeoxyglucose (FDG) can detect early changes in tumour metabolism and may be a useful quantitative imaging biomarker (QIB) for prediction of disease stabilization, response and duration of progression-free survival (PFS). Standardization of imaging procedures is a prerequisite, especially in multicentre clinical trials. In this study we reviewed the quality of FDG scans and compliance with the imaging guideline (IG) in a phase III clinical trial. METHODS: Forty-four cancer patients were enroled in an imaging sub-study of a randomized international multicentre trial. FDG scan had to be performed at baseline and 10-14 days after treatment start. The image transmittal forms (ITFs) and Digital Imaging and Communications in Medicine (DICOM) [1] standard headers were analysed for compliance with the IG. Mean liver standardized uptake values (LSUVmean) were measured as recommended by positron emission tomography (PET) Response Criteria in Solid Tumors 1.0 (PERCIST) [2]. RESULTS: Of 88 scans, 81 were received (44 patients); 36 were properly anonymized; 77/81 serum glucose values submitted, all but one within the IG. In 35/44 patients both scans were of sufficient visual quality. In 22/70 ITFs the reported UT differed by >1 min from the DICOM headers (max. difference 1 h 4 min). Based on the DICOM, UT compliance for both scans was 31.4%. LSUVmean was fairly constant for the 11 patients with UT compliance: 2.30 ± 0.33 at baseline and 2.27 ± 0.48 at follow-up (FU). Variability substantially increased for the subjects with unacceptable UT (11 patients): 2.27 ± 1.04 at baseline and 2.18 ± 0.83 at FU. CONCLUSION: The high attrition number of patients due to low compliance with the IG compromised the quantitative assessment of the predictive value for early response monitoring. This emphasizes the need for better regulated procedures in imaging departments, which may be achieved by education of involved personnel or efforts towards regulations. LSUVmean could be monitored to assess quality and compliance in an FDG PET/CT study.

Potent long-term cardioprotective effects of single low-dose insulin-like growth factor-1 treatment postmyocardial infarction.

BACKGROUND: Insulin-like growth factor-1 (IGF-1) is recognized as an important regulator of cardiac structure and cardiomyocyte homeostasis. The prosurvival and antiapoptotic effects of IGF-1 have been investigated in vitro and in rodent models of myocardial infarction (MI). However, the clinical application of IGF-1 has been hampered by dose-dependent side effects both acutely and during chronic administration. We hypothesized that single, low-dose IGF-1 (LD-IGF-1) administered locally and early in the reperfusion phase after acute MI in a large animal model would avoid significant side effects but would have prosurvival effects that would manifest in long-term structural and functional improvement after MI treatment. METHODS AND RESULTS: Forty-four female Landrace pigs underwent intracoronary administration of LD-IGF-1 or saline 2 hours into the reperfusion phase of acute left anterior descending artery occlusion MI. In the area of infarction, IGF-1 receptor and signaling responses were activated at 30 minutes and cardiomyocyte cell death attenuated at 24 hours after LD-IGF-1 but not saline treatment. Hemodynamic and structural studies using pressure-volume loop, CT, and triphenyltetrazolium chloride analysis 2 months post-MI confirmed a marked reduction in infarct size, attenuation of wall thinning, and augmentation of wall motion in the LD-IGF-1-treated but not in the saline-treated animals. These regional structural benefits were associated with global reductions in left ventricular volumes and significant improvement in left ventricular systolic and diastolic function. CONCLUSIONS: One-time LD-IGF-1 effects potent acute myocardial salvage in a preclinical model of left anterior descending artery occlusive MI, extending to long-term benefits in MI size, wall structure, and function and underscoring its potential as an adjunctive therapeutic agent.