Discrepancies between transcutaneous and estimated glomerular filtration rates in rats with chronic kidney disease

Accurate assessment of the glomerular filtration rate (GFR) is crucial for researching kidney disease in rats. Although validation of methods that assess GFR is crucial, large-scale comparisons between different methods are lacking. Both transcutaneous GFR (tGFR) and a newly developed estimated GFR (eGFR) equation by our group provide a low-invasive approach enabling repeated measurements. The tGFR is a single bolus method using FITC-labeled sinistrin to measure GFR based on half-life of the transcutaneous signal, whilst the eGFR is based on urinary sinistrin clearance. Here, we retrospectively compared tGFR, using both 1- and 3- compartment models (tGFR_1c and tGFR_3c, respectively) to the eGFR in a historic cohort of 43 healthy male rats and 84 male rats with various models of chronic kidney disease. The eGFR was on average considerably lower than tGFR-1c and tGFR-3c (mean differences 855 and 216 μL/min, respectively) and only 20 and 47% of measurements were within 30% of each other, respectively. The relative difference between eGFR and tGFR was highest in rats with the lowest GFR. Possible explanations for the divergence are problems inherent to tGFR, such as technical issues with signal measurement, description of the signal kinetics, and translation of half-life to tGFR, which depends on distribution volume. The unknown impact of isoflurane anesthesia used in determining mGFR remains a limiting factor. Thus, our study shows that there is a severe disagreement between GFR measured by tGFR and eGFR, stressing the need for more rigorous validation of the tGFR and possible adjustments to the underlying technique.

Optimizing carboplatin dosing by an improved prediction of carboplatin clearance using a CT-enhanced estimate of renal function

AIMS

Carboplatin is generally dosed based on a modified Calvert formula, in which the Cockcroft-Gault-based creatinine clearance (CRCL) is used as proxy for the glomerular filtration rate (GFR). The Cockcroft-Gault formula (CG) overpredicts CRCL in patients with an aberrant body composition. The CT-enhanced estimate of RenAl FuncTion (CRAFT) was developed to compensate for this overprediction. We aimed to evaluate whether carboplatin clearance is better predicted by CRCL based on the CRAFT compared to the CG.

METHODS

Data of four previously conducted trials was used. The CRAFT was divided by serum creatinine to derive CRCL. The difference between CRAFT- and CG-based CRCL was assessed by population pharmacokinetic modelling. Furthermore, the difference in calculated carboplatin dose was assessed in a heterogeneous dataset.

RESULTS

In total, 108 patients were included in the analysis. Addition of the CRAFT- and CG-based CRCL as covariate on carboplatin clearance led, respectively, to an improved model fit with a 26-point drop in objective function value and a worsened model fit with an increase of 8 points. In 19 subjects with serum creatinine <50 μmol/L, the calculated carboplatin dose was 233 mg higher using the CG.

CONCLUSIONS

Carboplatin clearance is better predicted by CRAFT vs. CG-based CRCL. In subjects with low serum creatinine, the calculated carboplatin dose using CG exceeds the dose using CRAFT, which might explain the need for dose capping when using the CG. Therefore, the CRAFT might be an alternative for dose capping while still dosing accurately.

Deep learning body-composition analysis of clinically acquired CT-scans estimates creatinine excretion with high accuracy in patients and healthy individuals

Assessment of daily creatinine production and excretion plays a crucial role in the estimation of renal function. Creatinine excretion is estimated by creatinine excretion equations and implicitly in eGFR equations like MDRD and CKD-EPI. These equations are however unreliable in patients with aberrant body composition. In this study we developed and validated equations estimating creatinine production using deep learning body-composition analysis of clinically acquired CT-scans. We retrospectively included patients in our center that received any CT-scan including the abdomen and had a 24-h urine collection within 2 weeks of the scan (n = 636). To validate the equations in healthy individuals, we included a kidney donor dataset (n = 287). We used a deep learning algorithm to segment muscle and fat at the 3rd lumbar vertebra, calculate surface areas and extract radiomics parameters. Two equations for CT-based estimate of RenAl FuncTion (CRAFT 1 including CT parameters, age, weight, and stature and CRAFT 2 excluding weight and stature) were developed and compared to the Cockcroft-Gault and the Ix equations. CRAFT1 and CRAFT 2 were both unbiased (MPE = 0.18 and 0.16 mmol/day, respectively) and accurate (RMSE = 2.68 and 2.78 mmol/day, respectively) in the patient dataset and were more accurate than the Ix (RMSE = 3.46 mmol/day) and Cockcroft-Gault equation (RMSE = 3.52 mmol/day). In healthy kidney donors, CRAFT 1 and CRAFT 2 remained unbiased (MPE = − 0.71 and − 0.73 mmol/day respectively) and accurate (RMSE = 1.86 and 1.97 mmol/day, respectively). Deep learning-based extraction of body-composition parameters from abdominal CT-scans can be used to reliably estimate creatinine production in both patients as well as healthy individuals. The presented algorithm can improve the estimation of renal function in patients who have recently had a CT scan. The proposed methods provide an improved estimation of renal function that is fully automatic and can be readily implemented in routine clinical practice.

GFR estimation is complicated by a high incidence of non-steady-state serum creatinine concentrations at the emergency department

Background

Acquiring a reliable estimate of glomerular filtration rate (eGFR) at the emergency department (ED) is important for clinical management and for dosing renally excreted drugs. However, renal function formulas such as CKD-EPI can give biased results when serum creatinine (SCr) is not in steady-state because the assumption that urinary creatinine excretion is constant is then invalid. We assessed the extent of this by analysing variability in SCr in patients who visited the ED of a tertiary care centre.

Methods

Data from ED visits at the University Medical Centre Utrecht, the Netherlands between 2012 and 2019 were extracted from the Utrecht Patient Oriented Database. Three measurement time points were defined for each visit: last SCr measurement before visit as baseline (SCr-BL), first measurement during visit (SCr-ED) and a subsequent measurement between 6 and 24 hours during admission (SCr-H1). Non-steady-state SCr was defined as exceeding the Reference Change Value (RCV), with 15% decrease or 18% increase between successive SCr measurements. Exceeding the RCV was deemed as a significant change.

Results

Of visits where SCr-BL and SCr-ED were measured (N = 47,540), 28.0% showed significant change in SCr. Of 17,928 visits admitted to the hospital with a SCr-H1 after SCr-ED, 27,7% showed significant change. More than half (55%) of the patients with SCr values available at all three timepoints (11,054) showed at least one significant change in SCr over time.

Conclusion

One third of ED visits preceded and/or followed by creatinine measurement show non-stable serum creatinine concentration. At the ED automatically calculated eGFR should therefore be interpreted with great caution when assessing kidney function.

A systematic review and meta-analysis of COVID-19 in kidney transplant recipients: Lessons to be learned

Kidney transplant recipients (KTR) may be at increased risk of adverse COVID-19 outcomes, due to prevalent comorbidities and immunosuppressed status. Given the global differences in COVID-19 policies and treatments, a robust assessment of all evidence is necessary to evaluate the clinical course of COVID-19 in KTR. Studies on mortality and acute kidney injury (AKI) in KTR in the World Health Organization COVID-19 database were systematically reviewed. We selected studies published between March 2020 and January 18th 2021, including at least five KTR with COVID-19. Random-effects meta-analyses were performed to calculate overall proportions, including 95% confidence intervals (95% CI). Subgroup analyses were performed on time of submission, geographical region, sex, age, time after transplantation, comorbidities, and treatments. We included 74 studies with 5559 KTR with COVID-19 (64.0% males, mean age 58.2 years, mean 73 months after transplantation) in total. The risk of mortality, 23% (95% CI: 21%-27%), and AKI, 50% (95% CI: 44%-56%), is high among KTR with COVID-19, regardless of sex, age and comorbidities, underlining the call to accelerate vaccination programs for KTR. Given the suboptimal reporting across the identified studies, we urge researchers to consistently report anthropometrics, kidney function at baseline and discharge, (changes in) immunosuppressive therapy, AKI, and renal outcome among KTR.

Deep learning-based classification of kidney transplant pathology: a retrospective, multicentre, proof-of-concept study

BACKGROUND

Histopathological assessment of transplant biopsies is currently the standard method to diagnose allograft rejection and can help guide patient management, but it is one of the most challenging areas of pathology, requiring considerable expertise, time, and effort. We aimed to analyse the utility of deep learning to preclassify histology of kidney allograft biopsies into three main broad categories (ie, normal, rejection, and other diseases) as a potential biopsy triage system focusing on transplant rejection.

METHODS

We performed a retrospective, multicentre, proof-of-concept study using 5844 digital whole slide images of kidney allograft biopsies from 1948 patients. Kidney allograft biopsy samples were identified by a database search in the Departments of Pathology of the Amsterdam UMC, Amsterdam, Netherlands (1130 patients) and the University Medical Center Utrecht, Utrecht, Netherlands (717 patients). 101 consecutive kidney transplant biopsies were identified in the archive of the Institute of Pathology, RWTH Aachen University Hospital, Aachen, Germany. Convolutional neural networks (CNNs) were trained to classify allograft biopsies as normal, rejection, or other diseases. Three times cross-validation (1847 patients) and deployment on an external real-world cohort (101 patients) were used for validation. Area under the receiver operating characteristic curve (AUROC) was used as the main performance metric (the primary endpoint to assess CNN performance).

FINDINGS

Serial CNNs, first classifying kidney allograft biopsies as normal (AUROC 0·87 [ten times bootstrapped CI 0·85-0·88]) and disease (0·87 [0·86-0·88]), followed by a second CNN classifying biopsies classified as disease into rejection (0·75 [0·73-0·76]) and other diseases (0·75 [0·72-0·77]), showed similar AUROC in cross-validation and deployment on independent real-world data (first CNN normal AUROC 0·83 [0·80-0·85], disease 0·83 [0·73-0·91]; second CNN rejection 0·61 [0·51-0·70], other diseases 0·61 [0·50-0·74]). A single CNN classifying biopsies as normal, rejection, or other diseases showed similar performance in cross-validation (normal AUROC 0·80 [0·73-0·84], rejection 0·76 [0·66-0·80], other diseases 0·50 [0·36-0·57]) and generalised well for normal and rejection classes in the real-world data. Visualisation techniques highlighted rejection-relevant areas of biopsies in the tubulointerstitium.

INTERPRETATION

This study showed that deep learning-based classification of transplant biopsies could support pathological diagnostics of kidney allograft rejection.

FUNDING

European Research Council; German Research Foundation; German Federal Ministries of Education and Research, Health, and Economic Affairs and Energy; Dutch Kidney Foundation; Human(e) AI Research Priority Area of the University of Amsterdam; and Max-Eder Programme of German Cancer Aid.

A plasma creatinine- and urea-based equation to estimate glomerular filtration rate in rats

Monitoring renal function is a vital part of kidney research involving rats. The laborious measurement of glomerular filtration rate (GFR) with administration of exogenous filtration markers does not easily allow serial measurements. Using an in-house database of inulin clearances, we developed and validated a plasma creatinine- and plasma urea-based equation to estimate GFR in a large cohort of male rats [development cohort n = 325, R² = 0.816, percentage of predictions that fell within 30% of the true value (P30) = 76%] that had high accuracy in the validation cohort (n = 116 rats, R² = 0.935, P30 = 79%). The equation was less accurate in rats with nonsteady-state creatinine, in which the equation should therefore not be used. In conclusion, applying this equation facilitates easy and repeatable estimates of GFR in rats.NEW & NOTEWORTHY This is the first equation, that we know of, which estimates glomerular filtration rate in rats based on a single measurement of body weight, plasma creatinine, and plasma urea.

[Pharmacotherapy in patients with loss of renal function]

When dosing renally excreted drugs in patients with kidney disease, it is important to have a reliable estimate of renal function. The estimated glomerular filtration rate (eGFR) is often used in the clinic, although this measure can be inaccurate in certain situations. Choosing the appropriate drug and dosage should therefore not be solely based on the eGFR. In this review, we discuss the physiology behind renal function estimation and drug dosing and propose a step by step approach to dosing renally excreted drugs in patients with kidney disease.

Validation of multiparametric MRI by histopathology after nephrectomy: a case study

OBJECTIVES

Renal multiparametric MRI (mpMRI) is a promising tool to monitor renal allograft health to enable timely treatment of chronic allograft nephropathy. This study aims to validate mpMRI by whole-kidney histology following transplantectomy.

MATERIALS AND METHODS

A patient with kidney transplant failure underwent mpMRI prior to transplantectomy. The mpMRI included blood oxygenation level-dependent (BOLD) MRI, T₁ and T₂ mapping, diffusion-weighted imaging (DWI), 2D phase contrast (2DPC) and arterial spin labeling (ASL). Parenchymal mpMRI measures were compared to normative values obtained in 19 healthy controls. Differences were expressed in standard deviations (SD) of normative values. The mpMRI measures were compared qualitatively to histology.

RESULTS

The mpMRI showed a heterogeneous parenchyma consistent with extensive interstitial hemorrhage on histology. A global increase in T₁ (+ 3.0 SD) and restricted diffusivity (- 3.6 SD) were consistent with inflammation and fibrosis. Decreased T₂ (- 1.8 SD) indicated fibrosis or hemorrhage. ASL showed diminished cortical perfusion (- 2.9 SD) with patent proximal arteries. 2DPC revealed a 69% decrease in renal perfusion. Histological evaluation showed a dense inflammatory infiltrate and fibrotic changes, consistent with mpMRI results. Most interlobular arteries were obliterated while proximal arteries were patent, consistent with ASL findings.

DISCUSSION

mpMRI findings correlated well with histology both globally as well as locally.

Decreased native renal T1 up to one week after gadobutrol administration in healthy volunteers

Background

Gadolinium‐based contrast agents (GBCAs) are widely used in MRI, despite safety concerns regarding deposition in brain and other organs. In animal studies gadolinium was detected for weeks after administration in the kidneys, but this has not yet been demonstrated in humans.

Purpose

To find evidence for the prolonged presence of gadobutrol in the kidneys in healthy volunteers.

Study Type

Combined retrospective and prospective analysis of a repeatability study.

Population

Twenty‐three healthy volunteers with normal renal function (12 women, age range 40–76 years), of whom 21 were used for analysis.

Field Strength/Sequence

Inversion recovery‐based T₁ map at 3T.

Assessment

T₁ maps were obtained twice with a median interval of 7 (range: 4–16) days. The T₁ difference (ΔT₁) between both scans was compared between the gadolinium group (n = 16, 0.05 mmol/kg gadobutrol administered after T₁ mapping during both scan sessions) and the control group (n = 5, no gadobutrol). T₁ maps were analyzed separately for cortex and medulla.

Statistical Tests

Mann–Whitney U‐tests to detect differences in ΔT₁ between groups and linear regression to relate time between scans and estimated glomerular filtration rate (eGFR) to ΔT₁.

Results

ΔT₁ differed significantly between the gadolinium and control group: median ΔT₁ cortex –98 vs. 7 msec (P < 0.001) and medulla –68 msec vs. 19 msec (P = 0.001), respectively. The bias corresponds to renal gadobutrol concentrations of 8 nmol/g tissue (cortex) and 4 nmol/g tissue (medulla), ie, ~2.4 μmol for both kidneys (0.05% of original dose). ΔT₁ correlated in the gadolinium group with duration between acquisitions for both cortex (regression coefficient (β) 16.5 msec/day, R² 0.50, P < 0.001) and medulla (β 11.5 msec/day, R² 0.32, P < 0.001). Medullary ΔT₁ correlated with eGFR (β 1.13 msec/(ml/min) R² 0.25, P = 0.008).

Data Conclusion

We found evidence of delayed renal gadobutrol excretion after a single contrast agent administration in subjects with normal renal function. Even within this healthy population, elimination delay increased with decreasing kidney function.

Level of Evidence: 3

Technical Efficacy: Stage 3

J. Magn. Reson. Imaging 2020;52:622–631.

Histological characteristics of Acute Tubular Injury during Delayed Graft Function predict renal function after renal transplantation

Acute Tubular Injury (ATI) is the leading cause of Delayed Graft Function (DGF) after renal transplantation (RTX). Biopsies taken 1 week after RTX often show extensive tubular damage, which in most cases resolves due to the high regenerative capacity of the kidney. Not much is known about the relation between histological parameters of renal damage and regeneration immediately after RTX and renal outcome in patients with DGF. We retrospectively evaluated 94 patients with DGF due to ATI only. Biopsies were scored for morphological characteristics of renal damage (edema, casts, vacuolization, and dilatation) by three independent blinded observers. The regenerative potential was quantified by tubular cells expressing markers of proliferation (Ki67) and dedifferentiation (CD133). Parameters were related to renal function after recovery (CKD-EPI 3, 6, and 12 months posttransplantation). Quantification of morphological characteristics was reproducible among observers (Kendall's W ≥ 0.56). In a linear mixed model, edema and casts significantly associated with eGFR within the first year independently of clinical characteristics. Combined with donor age, edema and casts outperformed the Nyberg score, a well-validated clinical score to predict eGFR within the first year after transplantation (R2  = 0.29 vs. R2  = 0.14). Although the number of Ki67+ cells correlated to the extent of acute damage, neither CD133 nor Ki67 correlated with renal functional recovery. In conclusion, the morphological characteristics of ATI immediately after RTX correlate with graft function after DGF. Despite the crucial role of regeneration in recovery after ATI, we did not find a correlation between dedifferentiation marker CD133 or proliferation marker Ki67 and renal recovery after DGF.