Correlation of histopathologic and dynamic tissue perfusion measurement findings in transplanted kidneys

Acute and chronic damage to renal transplants requires an early and accurate diagnosis to start the appropriate treatment. Renal transplant biopsies are an established method to evaluate such changes. These biopsies are performed as routine protocol biopsies in regular intervals in asymptomatic recipients or are performed when a clinical demand emerges. Biopsies are invasive and therefore a need for less potentially harmful procedures exists. Sonography is widely used to describe morphological changes of the transplant. Doppler sonography is used to describe changes of the blood flow to transplants. Today, only limited information can be drawn from Doppler sonographic measurements. To refine the impression of a transplant’s perfusion contrast enhanced sonography was introduced. Nevertheless, a need for a more differentiated evaluation remains. Dynamic tissue perfusion measurement (DTPM) has demonstrated the potential to describe perfusion in renal cortex in thin cortical slices with more than 50 parameters [1‐5] as well as in other organs [6‐10].

This study was approved by Institutional Review Board. Totally, 96 DTPM studies and renal transplant biopsies were performed in 78 Taiwanese patients (single study for 63 patients; two studies for 12 patients; three studies for 3 patients) between December 2005 and September 2008. The mean duration between the patient receiving renal transplantation and DTPM study was 44.8 months. Among 78 patients, there were 41 male and 37 female; the mean age when receiving living or deceased donors’ renal transplantation was 41.0 ± 12.6 years (range 6–72 years); 68 received renal transplantation from deceased donor, 6 received simultaneous pancreas and kidney transplantation, 4 received living-related renal transplantation. The indication for grafting was end stage renal disease; the transplant surgery was performed in Taiwan or People’s Republic of China. Details of the transplant recipients are outlined below:

Among 96 DTPM studies, 36 belonged to the PTC 0-group, 46 to the PTC 1-group, 11 to the PTC 2-group and 3 to the PTC 3-group; 47 belonged to the IF/TA 0-group, 32 to the IF/TA 1-group, 17 to the IF/TA 2-group and 0 to the IF/TA 3-group.

The mean perfusion intensities were 0.65 (cm/s) in PTC 0, 0.48 in PTC 1 (p < 0.001 vs. PTC 0), 0.39 in PTC 2 (p < 0.001 vs. PTC 0, p = 0.046 vs. PTC1), 0.22 in PTC 3 (p < 0.001 vs. PTC 0, p = 0.002 vs. PTC 1) (Figure 3, Tables 3 and 4). The conventional RI (resistance index) were 0.73 in PTC 0, 0.78 in PTC 1 (p = 0.018 vs. PTC 0), 0.77 in PTC 2, 0.87 in PTC 3 (Figure 3, Table 3). RI showed a weaker connection to peritubular inflammation than did cortical tissue perfusion with only one significant difference between the PTC-classes compared to five significant differences among the DTPM measurements (Figure 3).

Increasing peritubulitis caused a perfusion loss from central (p20) to distal (d20) layers from 0.96 to 0.20 (79% reduction) in PTC 0 transplants, from 0.75 to 0.11 (85% reduction) in PTC 1 transplants, from 0.67 to 0.04 (94% reduction) in PTC 2 transplants, from 0.25 to 0.02 (94% reduction) in PTC 3 transplants (Figure 4). As peritubulitis increased from PTC 0 to PTC 3, the perfusion intensity decreased from 0.96 to 0.25 (64% reduction) in p20, 0.86 to 0.32 (63% reduction) in p50, 0.42 to 0.10 (76% reduction) in d50, 0.20 to 0.02 (90% reduction) in d20 layer (Figure 4).

In those without peritubulitis (PTC 0), perfusion intensity was 0.272 in IF/TA 0, 0.535 in IF/TA 1 (p < 0.001 vs. IF/TA 0), 0.726 in IF/TA 2 (p < 0.001 vs. IF/TA 0) (Figure 5). IF/TA dependent changes were significantly higher than PTCs influences if PTCs reached values of 1 or 2 and IF/TA remained at zero-level compared to the mirror situation with IF/TA reaching levels of 1 and 2 and no PTCs (PTCs 0). Polyomavirus infection was accompanied by a significant increase in cortical perfusion in the outer cortical layer (distal 20%), albeit less pronounced in stage B than stage A (Figure 6).

Renal transplant biopsies are evaluated according to the Banff scoring system since it was shown that a high Banff score is predictive for the graft survival [13].

Color Doppler ultrasound of the outer renal cortex provides the basis for a quantitative measurement of the perfusion of renal transplants [1, 2, 14]. Vascular changes, which are mirrored by perfusion changes, may precede manifest function loss of the kidney and can be described histologically.

Peritubular inflammation is a criterion of humoral rejection in renal transplants. According to the Banff classification, this feature is scored as peritubular capillaritis (PTCs) [15]. It might be important to detect peritubular capillaritis in renal transplants early since it can precede the development of a full blown chronic antibody-mediated rejection (CAMR). Others propose to assume a chronic antibody-mediated rejection (CAMR) even in the absence of C4d deposition in cases with peritubular capillaritis, transplant glomerulitis, thickening of the peritubular capillaries basement membrane, and circulating anti-HLA antibodies [16, 17]. Detection of peritubular inflammation might be significant for indicating the subsequent poor prognosis [18, 19] and thus could be regarded an early sign of graft deterioration. Our study showed a significant decrease of cortical perfusion as the inflammatory cells accumulation in peritubular capillaries increased. The reduction of tissue perfusion due to peritubular inflammation depended on the position of the cortical layer too. The more distal the layer the stronger was the effect of peritubular inflammation. Cortical perfusion was also influenced by interstitial fibrosis and tubular atrophy (IF/TA). Unexpectedly, increasing IF/TA scores were connected to an increased perfusion. This might be explained by the short-circuiting of blood via shunts connecting the afferent and efferent vascular pools in juxtamedullary glomeruli [20, 21]. Recently, in a mouse model the overexpression of EphB4 receptor tyrosine kinase in the kidney was shown to promote the development of such shunts [22].

Conventional color Doppler sonography with calculation of Resistance Index (RI) and Pulsatility Index (PI) were used for many years to evaluate the perfusion of renal transplants. RI measurements reflect a complex interaction of renal as well as systemic vascular influences and are therefore disregarded to be suitable for evaluation of the renal vascular state by some investigators [23, 24]. An elevated RI alone cannot reliably differentiate the causes of transplanted kidney dysfunction and it is considered a nonspecific marker [25].

RI does not inform about the real blood flow velocities, since a ratio of two velocities (systolic and diastolic) is calculated. It represents spatially restricted information from a single vessel and gives no data on the width of the vessel, let alone the perfused area in a certain tissue. Thus RI is basically less suitable for describing the perfusion status of the renal parenchyma. Furthermore, the pathological changes inside the kidney can be distributed inhomogeneously [26], and RI at a single point of an artery may not represent the status of perfusion in the region of interest. In contrast, Dynamic Tissue Perfusion Measurement (DTPM) is based upon a heart cycle triggered measurement of a tissue’s mean flow velocity of all vessels inside the region of interest multiplied by the mean perfused area (respecting changing width of all vessels during the heart cycle) [3, 4]. Thus, DTPM offers a more sophisticated investigation of renal perfusion by using velocity as well as perfused area information, referring all measurements to all vessels inside the tissue and is capable in differentiating all perfusion data between millimeter-thin slices. It reflects histological changes as PTCs better than conventional RI (Figure 3: left diagram compared to the right one). DTPM thus has the potential to go beyond the limitations of usual vascular resistance calculations by RI measurements. The detection of subtle perfusion changes simultaneously to a polyomavirus infection (Figure 6) clearly demonstrates that the slice wise perfusion measurement in consecutive layers of renal microvasculature offers a sophisticated access to relevant histological changes even beyond the Banff-classification. It was demonstrated in earlier investigations that with time elapsed after transplantation the cortical perfusion in renal transplants of children decreased significantly [2] and that simultaneously the pulsatility of renal transplants’ microvascular perfusion increased significantly [14]. Reduction of smaller subcapsular transplant vessels occurs in acute as well as chronic rejected organs [27]. Cortical vessel loss, as determined visually in power Doppler images, was predictive in contrary to RI as for the transplant function 1 year after transplantation [28]. Recent investigations also showed that microvascular damage is the key process linked to the loss of transplants’ function [29, 30]. Our findings might be especially noteworthy since we demonstrated a significant and gradual decline of cortical perfusion along with increasing peritubular capillaritis using DTPM. DTPM clearly showed the potential to measure the progressive obliteration of the transplant’s microvessels. This process starts in the periphery of renal cortex where the vessel diameters are naturally small. Our study did not investigate the relationship of DTPM with the clinical outcome of the graft. Further studies are needed to elucidate if specific histological findings or the destruction of the microvasculature are more predictive for the transplant’s function.

In the long run many renal disorders lead to a gradual decline of cortical perfusion [31‐33]. DTPM might thus be useful in other chronic diseases of the kidney too.

One limitation of our study is that we only assess the relationship between cortical perfusion and the histopathological changes of transplant kidney. Future studies are needed to confirm the role of DTPM as a surrogate of outcome of renal transplant. Then DTPM could become a non-invasive tool to reduce the need for invasive procedures such as biopsies.

Microvascular perfusion changes in renal transplants can be demonstrated by DTPM. A significant reduction of cortical perfusion in the transplant kidney, especially in its peripheral layers, is related to peritubular capillary inflammation.