Are volume measurements of non-functioning pituitary adenomas reliable?

Abstract

Precise radiological assessment of tumour volume is important in the follow-up of non-functioning pituitary adenomas (NFPAs). We compared the reliability of two methods for tumour volume measurements in the pre- and postoperative setting. We assessed the volume of 22 NFPAs at magnetic resonance imaging (MRI) scans before surgery and the first and third postoperative MRI obtained after submission from hospital. Volumetric assessments were performed both by summation of slices (SOS) and by diameter measures. All volumes were calculated independently by two readers. The preoperative intra- and inter-rater reliability was good for both the SOS and the diameter method (intraclass correlation coefficient (ICC) 0.996 and 0.990, and ICC: 0.982 and 0.967, respectively). The first postoperative investigation showed poorer intra- and inter-rater reliability for both methods (ICC: 0.872 and 0.791 and ICC: 0.792 and 0.810, respectively). The third postoperative MRI showed good intra-rater reliability (ICC: 0.961 and 0.962, respectively), but poorer inter-rater reliability for both methods (ICC: 0.759 and 0.703, respectively). Volume assessment by SOS presented overall slightly higher reliability than the diametric method. Overall, the reliability between the two methods was good when measured by the same reader (ICC: 0.988, 0.945 and 0.962, for the preoperative, first and third postoperative MRI, respectively). The preoperative intra- and inter-rater reliabilities were satisfactory for both the SOS and diametric method. Postoperative MRI scans showed poorer reliability, suggesting that measurements at these time points should be interpreted with care. For each MRI scan, reliability between methods was satisfactory when investigated by the same reader.