Shear wave velocities in damaged kidneys: fast and slow
- 787 Downloads
In this issue, Goya and coworkers  describe a well-performed study of sonographic elastography in relatively large groups of healthy children and children with varying degrees of renal damage as assessed by dimercaptosuccinic acid scintigraphy (DMSA). Their results are somewhat surprising: The more renal damage the slower the propagation of shear waves. This bewilders our intuitive belief that DMSA abnormality equals scarring equals fibrosis, fibrosis means stiffness, and stiffness should result in fast propagation of shear waves. It also conflicts with the report by Bruno et al.  that describes, concordant with expectations, significantly higher shear wave velocities in damaged as compared to non-damaged kidneys. The cause of the discrepancy may be multifactorial (inclusion criteria, measurement technique, etc.), and may not be easily resolved. However, several important lessons can be learned.
Unfortunately, Fig. 1 does not immediately help the interpretation of the unexpected observations made by Goya and coworkers, at least not if one expects that 1) renal damage seen on DMSA always corresponds to renal cortical thinning and 2) that a thinner cortex causes faster mechanical wave propagation. The latter assumption, based on Fig. 1, may not be possible to generalise. One could, for example, hypothesise that there are various degrees of fibrosis depending on the cause of damage and the time since damage, and that the parenchymal stiffness also depends on pressures within the renal collection system, etc. In other words, cortical thickness may not be an equally good determinant for shear wave velocity in other cohorts of patients. Extending this argument, “scarring” diagnosed by DMSA is unlikely to always mean fibrosis, so the intuitive sequence “DMSA scarring means fibrosis means stiffer parenchyma means faster wave propagation” is flawed.
The lessons learned: 1) Do not get taken by words —“DMSA scarring” may not only represent histopathological scarring, 2) Respect the data — convenient shortcuts may be invited by intuition, but at the risk of missing diagnostically important covariates, and 3) Significant relationships do not always imply added diagnostic value.
Goya et al.  offer an interesting insight into the complexity of diagnostic imaging. Their work should stimulate further investigation where the goals are improved early prediction of renal damage and less exposure of children to radionuclide modalities.
Conflicts of interest
Open AccessThis article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.