Intraobserver and interobserver variability in computed tomography size and attenuation measurements in patients with renal cell carcinoma receiving antiangiogenic therapy: Implications for alternative response criteria

Abstract

Background Alternative response criteria have been proposed in patients with metastatic Renal Cell Carcinoma (mRCC) on Vascular Endothelial Growth Factor (VEGF)-targeted therapy, including 10% tumor shrinkage as an indicator of response/outcome. However, intraobserver and interobserver measurement variability have not been defined in this setting. We aim to determine intra- and interobserver agreement of Computed Tomography (CT) size and attenuation measurements, to establish reproducible response indicators. Methods Seventy-one mRCC patients with 179 target lesions were enrolled in Phase II and III trials of VEGF-targeted therapies and retrospectively studied with institutional review board approval. Two radiologists independently measured long axis diameter and mean attenuation of targets on baseline and follow-up CT. Concordance correlation coefficients (CCCs) and Bland-Altman plots were used to assess intra- and interobserver agreement. Results High CCCs (0.8602–0.9984) were observed in all types of measurements. The 95% limits of agreement for percent change of the sum longest diameter was (−7.30%, 7.86%) for intraobserver variability, indicating 10% tumor shrinkage represents true change in tumor size when measured by one observer. The 95% limits of interobserver variability were (−16.3%, 15.4%). In multivariate analysis, liver location significantly contributed to interobserver variability (p=0.048). The 95% limits of intraobserver agreement for percent change in CT attenuation were (−18.34%, 16.7%). Conclusion In mRCC patients treated with VEGF-inhibitors, 10% tumor shrinkage is a reproducible radiologic response indicator when baseline and follow-up studies are measured by a single radiologist. Lesion location contributes significantly to measurement variability and should be considered when selecting target lesions.

RECIST 1.1 Compared With RECIST 1.0 in Patients With Advanced Renal Cell Carcinoma Receiving Vascular Endothelial Growth Factor–Targeted Therapy

“…The range of the 95% limits of agreement, reflecting the differences in the measurements that could be attributed to measurement variation, is similar to those reported elsewhere [26]. …”

“…There were only two progressive disease or stable disease differences by RECIST 1.1 and 1.0, which would alter the length of time enrolled in the trial or receiving treatment. Although it is interesting that a few more partial responses were observed at first follow-up using RECIST 1.1, especially considering the correlation of first follow-up tumor shrinkage with survival outcomes [21, 25], the increase in mean tumor shrinkage overall (12.1% by RECIST 1.1 vs 10.8% by RECIST 1.0) is well within the range of measurement error and is unlikely to translate into a clinically meaningful difference [26]. …”

Sonographically Identified Echogenic Renal Masses Up to 1 cm in Size Are So Rarely Malignant They Can Be Safely Ignored

“…For tumors up to 1 cm in size, lesion measurement variation up to 2 mm among observers for each of the 3 orthogonal measurements was considered acceptable. 3,4 If the variation of any of the 3 diameters was greater than 2 mm, the discrepancy was resolved by consensus. The final size that was tabulated for each lesion was the single largest diameter.…”

Cancer Imaging Training in the 21st Century: An Overview of Where We Are, and Where We Need To Be

“…Response criteria have been developed that incorporate density and metabolic factors in patients who receive these drugs [5,13,20,21]. Molecularly targeted agents that are not antiangiogenic cause different appearances with response; epidermal growth factor inhibitors, such as erlotinib, tend to result in retraction of tumors, causing them to appear scarlike when responding, and to bulge at progression.…”

“…Antiangiogenic drugs and tyrosine kinase inhibitors often do not elicit size changes that qualify as response when using traditional RECIST criteria, although these drugs do have proven clinical benefits [4,19,20]. Response criteria have been developed that incorporate density and metabolic factors in patients who receive these drugs [5,13,20,21].…”