Magnetic resonance elastography vs. point shear wave ultrasound elastography for the assessment of renal allograft dysfunction

Dear Editor,

In response to our recent article [1] Dr. Rueda et al wrote with some additional questions regarding the study methods [2]. In their letter they correctly point out that the small sample size is a limitation of the work which we have acknowledged. When comparing stiffness differences between functional and dysfunctional patients measured using point shear wave elastography (pSWE) we found no difference between the groups when including all ultrasound elastography measures and when considering only reliable measures (p > 0.6, manuscript section 3.2).

Reliability thresholds for renal imaging using ultrasound elastography have not yet been established therefore we applied the methods recommended for the liver i.e. the acquisition of 10 measurements per site and treating measures with interquartile range/median > 30 % to be unreliable [3]. Based on these criteria we found 13/27 patients (52 %) had reliable pSWE measurements. The low proportion of reliable measures could be due to several factors such as the influence of transducer orientation with respect to the renal pyramid axis [4,5], the effect of applied transducer pressure [6] and the use of reliability criteria established in the liver, a less anisotropic organ than the kidney. Operator training would appear to be of utmost importance when acquiring pSWE data in native or transplant kidneys, with care taken to ensure consistent transducer positioning with respect to the renal pyramid axis and applied transducer pressure.

In our study patients were considered to have functional renal allografts if estimated glomerular filtration rate (eGFR) was > 45 mL/min/1.73m² and eGFR or serum creatinine had not changed by more than 25 % over the preceding three months. Due to our institution’s policy on protocol biopsies, not all patients in the functional allograft group had pathological correlation to confirm lack of clinically significant fibrosis or inflammation which is a limitation of our study.

Skewness of corticomedullary stiffness measured with magnetic resonance elastography (MRE), not median corticomedullary stiffness, was found to be significantly lower in the dysfunctional group compared to the functional group (p = 0.041, manuscript Figure 2) in our study however there is indeed significant overlap between the groups. The limited sample size and lack of control for renal perfusion may be contributing factors to this overlap. We do not report a lower stiffness measured by pSWE or MRE in dysfunctional renal allografts based on our data. In fact, we found that patients who experienced graft loss or relist eligibility had higher cortical and corticomedullary stiffness measured with MRE compared to those who did not experience a decline in kidney function.

Logistic regression analysis was not performed in our study to determine relationships between MRE parameters and Banff pathology scores. Spearman correlation analysis in 14 subjects revealed significant correlations between the skewness of MRE corticomedullary stiffness and Banff pathology scores.

Our study is part of conflicting literature regarding the evaluation of renal stiffness using elastography. Further studies controlling for variables affecting stiffness measures, be it operator dependent (transducer position and force applied for ultrasound elastography) or physiological (renal perfusion) are needed.