DISCUSSION
The main finding of current study is that, a significant change in renal cortical elasticity was not found between patients with DM-II versus healthy subjects or in albuminuria versus no-albuminuria subgroups.
DM-II causes microstructural changes in renal parenchyma before a significant drop in eGFR occurs (3). A moderately increased uACR indicates presence of kidney damage and is associated with future risk of progression to kidney dysfunction and cardiovascular complications independent of eGFR (16). We recognize that histological changes can be present in a number of subjects with DM-II before pathologic amount of urinary albumin can be detected (3, 7). Hence, one may argue that biopsy is required for definitive confirmation of typical diabetic changes in glomeruli (8). However, biopsy is not performed routinely in all cases with DM-II and mostly reserved for certain indications (9). In the landmark study by Fioretto P et al, microalbuminuric DM-II patients with normal eGFR underwent kidney biopsy for research purposes rather than for a clinical indication (18). They found histopathological changes (either typical glomerular or atypical tubulointerstitial and vascular changes), in 70% of DM-II subjects. That was the basis that a moderately elevated uACR in the setting of preserved eGFR was used as representative of early diabetic renal injury in the current study where biopsy was not available. We have found that, the early diabetic histologic changes in kidneys do not necessarily translate into disturbed elasticity, as we have not observed a difference in SWV values between albuminuria and no-albuminurai subjects.
Conventional kidney US findings alone are inadequate for early recognition of kidney damage in most cases. Indeed the typical US findings of CKD represent late changes (4-6), which are already irreversible. Moreover the conventional US findings of CKD are not specific for diabetic nephropathy and diverse etiologies share the common imaging changes. Elastography, on the other hand, has emerged as a tool for assessment of increased tissue stiffness caused by damages to organs including liver, breast, thyroid, prostate and kidney. The relatively novel US shear-wave elastography method was used to assess kidney stiffness in our study. 2D-SWV measurement by this technique has provided comparable results to previously described methods including transient elastography and acoustic radiation force imaging in depicting liver fibrosis (15). Although SWV values are positively correlated with amount of fibrosis in liver diseases (19), it is negatively correlated with CKD stages and eGFR (6, 12). The underlying mechanism of the inverse relationship between stiffness and eGFR in patients with CKD remains unclear. Altered renal perfusion has been proposed to affect stiffness measurements and explain this paradox (20). The ARFI derived SWV has been found to be higher in patients with DM-II versus healthy controls in one study (21). Moreover others revealed increased SWV in DM-II induced early kidney damage but no CKD (i.e. eGFR still more than 60 ml/min/1.73m²) versus diabetic CKD (i.e. eGFR already less than 60 ml/min/1.73m²) (17). Interestingly our results does not support these findings. The whole kidney average SWV’s were not found to be increased neither in DM-II vs healthy controls nor in mAlb+ vs mAlb- subjects in current study. The contradiction between both studies can be explained by the differences in study populations. Goya et al (17) did not report any exclusion criteria in terms of comorbidities that may interfere with elasticity results (hypertension, systemic diseases, other diabetic organ complications, risk factors for vascular diseases). On the other hand our population comprised a very selective group that presence of a confounder of elasticity would be very unlikely. Another factor that may influence the results is the sonographic technique used for elasticity assessment. We used 2D-SWV measurement, whereas others used the ARFI method.
The left middle and right upper portions’ elasticity seemed to be selectively disturbed in mAlb+ cases at univariate comparison. One may argue that this may be attributable to the potentially asynchronous severity of changes in different portions of kidneys (8, 10). Indeed, heterogeneous renal histopathologic changes may occur in albuminuric patients with DM-II (22, 23). However, both renal portions SWV values were not independent predictors of mAlb+ status when adjusted for age by multivariate analysis. Hence focal impaired elasticity in mAlb+ group as compared to mAlb- group can not be considered plausible.
The impairment of whole kidney elasticity probably requires more profound histological changes due to diabetic nephropathy. Previous studies have shown reduced global kidney SWV values along with decreased eGFR values with or without severe proteinuria (6, 12, 19). However these findings probably require more severe kidney damage beyond glomerulosclerosis such as tubulointerstitial fibrosis and vascular changes which are deemed irreversible. We believe early recognition of disturbed kidney elasticity as a surrogate marker of early kidney damage is not possible by 2D-SWV measurement at this stage. However due to conflicting results when compared with the previous work using ARFI technique, further controlled longitudinal studies in larger series with biopsy and histopathology being used as reference standard are required in this regard.
The reproducibility of 2D-SWV was good in the current study with an interobserver agreement ICC value of 0.66, which is close to previous reports.
The major strength of our work is that it is the first study to evaluate the ability of novel 2D-SWV US elastography technique in revealing early kidney damage in patients with DM-II. Furthermore, application of strict inclusion criteria to eliminate the potential confounders of disturbed elasticity and proteinuria is another strength of our work. The sample size was also calculated a priori, and the number of patients included are adequate to reveal a difference between the SWV values of the study groups.
The main limitations in current study is the absence of histopathological confirmation of kidney damage. We believe, a combination of moderate albuminuria with preserved eGFR is a quite pragmatic way to reveal the early diabetic kidney damage, where biopsy is not available. Although very unlikely, absence of a non-diabetic cause of moderately increased uACR cannot be guaranteed based on exclusion of diseases by history and symptoms alone.
In conclusion, the kidney elasticity does not seem to be disturbed in patients with diabetes and a preserved eGFR with or without moderate albuminuria.
CONFLICT OF INTEREST: None.