The relationship of peritubular capillary density with glomerular volume and kidney function in living kidney donors

For this retrospective cohort study, we identified 73 living kidney donors with representative kidney biopsies. Biopsies were taken right after donor nephrectomy (T1), right before implantation (T2) and/or after reperfusion (T3) and were considered representative if T1, T2 and/or T3 had a total cortical surface of minimally 0.6 mm² with at least 5 glomeruli. All donors donated between August 11, 2005 and June 17, 2008 at the University Medical Center Groningen, The Netherlands. Four donors were excluded because they were part of the Dutch “cross-over” program and only came to our center for the actual nephrectomy procedure, rendering 69 living kidney donors eligible for inclusion in this study. All donors underwent pre- and three-month post-donation clinical and laboratory measurements as part of the regular living kidney donor screening program. In 52 donors, five-year post-donation follow-up was available. In 2014, these data were added to the TransplantLines Biobank and Cohort study (ClinicalTrials.gov identifier: NCT03272841). This is an observational cohort study on short- and long-term outcomes after organ transplantation/donation, as described previously [16]. The study was approved by the institutional ethical review board (METc 2014/077). All procedures were conducted in accordance with the declaration of Helsinki and declaration of Istanbul.

All available T1, T2 and T3 biopsies were stained with periodic-acid-shiff (PAS) and, on a separate section, an immunohistochemical staining for CD34 (Monosan, Uden, the Netherlands) was performed. In brief, parafin-embedded tissue sections were incubated with primary antibody after blocking of endogeneous perioxidase and antigen retrieval by boiling in TRIS EDTA buffer. After washing, the biopsies were incubated with bright vision anti-mouse HRP (Immunologic; Duiven, The Netherlands) followed by washing and thereafter 3,3-diaminobenzidine (DAB) (DAKO cytomation, Glosturp, Denmark) was used as the chromogen. Thereafter the protocol slides were counterstained with hematoxylin (Klinipath, Duiven, The Netherlands). Periodic-acid-shiff and immunohistochemically stained slides were digitalized using a Ventana scanner (Ventana iScan HT (Roche, Basel, Switzerland), and imported in Panoramic Image Viewer (3DHistotech, Budapest, Hungary); examples are shown in Fig. 1. Microstructural parameters were measured on PAS-stained sections by one observer (ML), according to Elsherbiny et al. [14], with the exception that partial glomeruli were counted as 1 and not as 0.5. Briefly, total cortical biopsy area was annotated manually, as well as glomerular tuft surface area of all non-sclerotic glomeruli. Then the profile area of non sclerotic glomeruli was calculated by dividing the number of non sclerotic glomeruli by cortical area. The Weibel Gomez stereological model was used to calculate the non sclerotic glomeruli density. Furthermore, non sclerotic glomeruli volume (glomerular volume) was calculated as described by Elsherbiny et al. [14]. Of all CD34 stained sections a maximum of 10 pictures of 120,000 μm² were taken in a serpentine manner [17], with Panoramic Viewer 1.15.4 and exported as jpeg into Paint (Microsoft, Seattle, WA, USA). There were no glomeruli present in these pictures. In all pictures, PTCs and tubules were manually traced by one observer (ML), with the exclusion of interlobular arteries. Peritubular capillaries and Tubuli per picture were quantified by Image J. Peritubular capillary density was assessed as number of PTCs per tubule (PTC/tubule) and number of PTCs per surface area (PTC/50,000 μm²).The tubular area was determined by dividing the area of the pictures with the number of tubuli counted per biopsy.

In cases that met our inclusion criteria of at least 5 glomeruli and 600,000 μm² of cortex, the PAS stained digital section was scored histologically according to Banff by a pathologist (CPK) [18]. Grade of interstitial fibrosis and tubular atrophy (IF/TA) was determined as highest of tubular atrophy (ct) or interstitial fibrosis (ci). Also, IF/TA was assessed by the pathologist as more or less than 5% of the cortical area.

During screening, clinical parameters as weight, height, hip circumference, waist circumference and blood pressure were measured, medication use was asked as well as smoking history. Kidney function before and at three months and five years after donation was indirectly determined by measuring the clearance of the exogenous filtration marker ¹²⁵I-iothalamate (measured GFR (mGFR), described in more detail previously) [19]. In short, ¹²⁵I-Iothalamate and ¹³¹I-hippurate infusions were started and after a stabilization period, baseline measurements were performed in a steady state of plasma tracer levels. Clearances were calculated as (U*V)/P and (I*V)/P, where U*V represents the urinary excretion, I*V represents the infusion rate of the tracer and P represents the plasma tracer concentration per clearance period. We calculated mGFR from clearance levels of these tracers using (U*V)/P and corrected the renal clearance of ¹²⁵I-iothalamate for urine collection errors by multiplying the urinary ¹²⁵I-Iothalamate clearances with the ratio of plasma and urinary ¹³¹I-hippurate clearance by using the following formula:

The mGFR after stimulation with dopamine was also assessed before donation (mGFR_dopa). The mGFR_dopa was used to calculate the dopamine-induced increase in GFR (ΔmGFR_dopa, previously referred to as the renal functional reserve (RFR) [19, 20]) by subtracting the unstimulated mGFR form the mGFR_dopa. Dopamine-stimulated mGFR was missing in 4 cases. Serum creatinine was measured routinely in our central chemistry laboratory by an isotope dilution mass spectrometry (IDMS) traceable enzymatic assay on the Roche Modular (Roche Ltd., Mannheim, Germany). In addition, serum HbA1c concentration was recorded.

Data are reported as mean (standard deviation (SD)) for normally distributed variables and median [interquartile range, IQR] for skewed data. Binary variables are shown as “number (%)”. Correlations between glomerular volume, IF/TA, PTC/tubule, tubular area and PTC/50,000 µm² were assessed by scatter plots and Pearson’s correlation coefficients. In cross-sectional analyses, we investigated which pre-donation characteristics were associated with the microstructural parameters using univariable linear regression analyses. Subsequently, we used linear regression analyses to assess the association between the morphometrical parameters and pre- and post-donation kidney function outcomes. Outcomes were pre- and three months and five year post-donation mGFR. All univariable associations of the microstructural parameters with pre- and post-donation outcomes were adjusted for age using multivariable linear regression analyses, because age is a known determinant of GFR, as well as microstructural features in the kidney [2, 21]. To detect a correlation of 0.3 with an α of 0.05 and a power of 80%, 67 donors are needed. Statistical analyses were performed in SPSS version 28 for Windows (IBM, Armonk, NY), and Graphpad Prism 8 for Windows (Graphpad, San Diego, CA). p values of < 0.05 were considered statistically significant.

A total of 69 living kidney donors were included in this study. Mean age was 52 ± 11 years, 46% were female and all donors were white (Table 1). The donors had a mean BMI of 26 ± 4 kg/m² and a mean systolic blood pressure (SBP) of 130 ± 15 mmHg. Three donors had a pre-donation serum HbA1c level ≥ 6.5%, of which two donors had a BMI of 34 and 35 kg/m², respectively. Pre-donation mGFR was 119 ± 22 mL/min and decreased to 75 ± 14 at three months post-donation (Table 2). Five years after donation, mGFR was 82 ± 15 mL/min. Before donation, mean glomerular volume was 0.0024 ± 0.0007 mm³, mean number of PTC/tub was 1.97 ± 0.3, mean number of PTC/50,000 μm² was 25.9 ± 4.4, mean tubular area was 3679.2 ± 835.7 µm², and 19 donors had > 5% IF/TA (Table 3).

Scatterplots of correlations between microstructural parameters are shown in Fig. 2. The strongest correlation was observed for tubular area with PTC/50,000 μm² (R = − 0.63, p < 0.001), with fewer PTCs per 50,000 μm² in cases with larger tubular area. However, when the number of PTCs was adjusted for the number of tubules on the biopsy (PTC/tubule), we observed an increase in PTC/tubule in cases with increased tubular area (R = 0.31, p = 0.01), which is as expected because cases with larger tubules display a smaller number of tubules per surface area on the biopsy. Glomerular volume correlated positively with tubular area (R = 0.26, p = 0.03) and with PTC/tub (R = 0.24, p = 0.047), and negatively with a trend towards significance with PTC/50,000 μm² (R = − 0.21, p = 0.08). There was no correlation between PTC/tubule and PTC/50,000 μm² (R = 0.05, p = 0.70).

Univariable linear regression analyses did not reveal associations of clinical variables (e.g. age, sex, weight, blood pressure) with PTC/tubule or PTC/50,000 μm² (Table 4). Body surface area (BSA) (St.β = 0.30, p = 0.01), waist/hip-ratio (St.β = 0.25, p = 0.05), systolic blood pressure (SBP, St.β = 0.35, p = 0.004) and diastolic blood pressure (DBP, St.β = 0.30, p = 0.01) were all positively associated with glomerular volume (Table 4). A trend towards significance was shown for the association of BMI with glomerular volume (St.β = 0.23, p = 0.06). Smoking correlated negatively and significantly with tubular area (St.β = − 0.38, p = 0.004). Living kidney donors with IF/TA > 5% in their biopsy were older than donors without IF/TA (t-test p = 0.002, Table 5). Also, individuals with IF/TA > 5% had a larger tubular area (Table 5). None of the clinical parameters were associated with PTC/tubule or PTC/50,000 μm².

Peritubular capillary/tubule was significantly and independent of age associated with the ΔmGFR_dopa (= dopamine induced increase in mGFR, St.β = 0.25, p = 0.04, Table 6), but not with unstimulated mGFR (St.β = 0.17, p = 0.14). Peritubular capillary/50,000 µm² was not associated with mGFR or ΔmGFR_dopa (St.β = 0.01, p = 0.97 and St.β = 0.04, p = 0.74, respectively). Glomerular volume was significantly and positively associated with pre-donation mGFR (St.β = 0.31, p = 0.01, Table 6), but not with the ΔmGFR_dopa (St.β = − 0.13, p = 0.31). Tubular area and IF/TA were not associated with pre-donation kidney function (Table 6). In a multivariable linear regression model including glomerular volume and PTC/tubule, both were independently associated with pre-donation mGFR_dopa (R² = 0.29), with glomerular volume being associated with pre-donation mGFR, and PTC/tub with pre-donation ΔmGFR_dopa (Table 7). The association of PTC/tubule with mGFR_dopa and ΔmGFR_dopa remained significant after adjustment for tubular area (PTC/tubule with mGFR_dopa: St.β = 0.29, p = 0.01; PTC/tubule with ΔmGFR_dopa: St.β = 0.26, p = 0.045, Table 8).

There was no association of PTC/tubule with unstimulated mGFR at three months or five years post-donation. Glomerular volume was significantly and positively associated with both three-month, and five-year post-donation mGFR (St.β = 0.27, p = 0.02 and St.β = 0.30, p = 0.01, respectively, Table 9). Tubular area, PTC/50,000μm² and IF/TA were not associated with post-donation mGFR (Table 6).