Analysis of vestibular schwannoma size in multiple dimensions: a comparative cohort study of different measurement techniques

Abstract

Objectives: In this volumetric study of the vestibular schwannoma, we evaluated the accuracy and reliability of several approximation methods that are in use, and determined the minimum volume difference that needs to be measured for it to be attributable to an actual difference rather than a retest error. We also found empirical proportionality coefficients for the different methods. Design ⁄ setting and participants: Methodological study with investigation of three different VS measurement methods compared to a reference method that was based on serial slice volume estimates. These volume estimates were based on: (i) one single diameter, (ii) three orthogonal diameters or (iii) the maximal slice area. Altogether 252 T1-weighted MRI images with gadolinium contrast, from 139 VS patients, were examined. Main outcome measures: The retest errors, in terms of relative percentages, were determined by undertaking repeated measurements on 63 scans for each method. Intraclass correlation coefficients were used to assess the agreement between each of the approximation methods and the reference method. The tendency for approximation methods to systematically overestimate ⁄ underestimate different-sized tumours was also assessed, with the help of Bland-Altman plots. Results: The most commonly used approximation method, the maximum diameter, was the least reliable measurement method and has inherent weaknesses that need to be considered. This includes greater retest errors than area-based measurements (25% and 15%, respectively), and that it was the only approximation method that could not easily be converted into volumetric units. Area-based measurements can furthermore be more reliable for smaller volume differences than diameter-based measurements. Conclusions: All our findings suggest that the maximum diameter should not be used as an approximation method. We propose the use of measurement modalities that take into account growth in multiple dimensions instead.

Tumour Resection Volumes and Facial Nerve Outcomes for Vestibular Schwannomas

“…This series of patients (Table 1) highlights the range of actual volumes produced by the same diameters, suggesting volume calculations and the use of volume software would be preferable. The diameters alone do not provide adequate volume estimates as shown by the predicted volume data and as such reinforces the need for a volumetric programme [24]. Varughese et al highlight that maximal diameters should be avoided in approximating tumour size, preferring the technique we have employed as a gold standard reference methodology, although it is more time consuming.…”

“…Volumetric measurement can lead to a better understanding of growth patterns and aid clinical management. Tumour volumes are potentially more reliable than using diameters and could be used to define total, gross total, near total and subtotal resections by volume [24]. This would then enable management algorithms to be generated to decide on long term follow up and the need for adjuvant therapy.…”

“…Varughese et al highlight that maximal diameters should be avoided in approximating tumour size, preferring the technique we have employed as a gold standard reference methodology, although it is more time consuming. Furthermore, it is well recognised that tumour diameters are measured quite differently across departments despite internationally agreed criteria, for example by inclusion of the intracanicular components of tumours [10,24].…”

Conservative Management of Vestibular Schwannoma—A Prospective Cohort Study

“…Each slice volume was estimated by multiplying the slice area by slice interval, and the tumor volume was calculated by summarizing all slices. 19 …”

MRI surveillance of vestibular schwannomas without contrast enhancement: Clinical and economic evaluation

“…Vestibular schwannoma size (and therefore growth) has been estimated using either volumetric or linear diameter measurements. Although there has been a trend toward volumetric measurements, linear diameter can serve as an adequate gauge of size and is a technique more readily available (and more likely utilized) to nonradiologists, who are ultimately responsible for determining optimal management . In this study, the intraobserver measurements of NR2 approached statistical noncorrelation in the SI plane only, although the reasons for this are unclear.…”