First name
Susan
Middle name
L
Last name
Furth

Title

Using Machine Learning to Identify Metabolomic Signatures of Pediatric Chronic Kidney Disease Etiology.

Year of Publication

2022

Number of Pages

375-386

Date Published

2022 Feb

ISSN Number

1533-3450

Abstract

<p><strong>BACKGROUND: </strong>Untargeted plasma metabolomic profiling combined with machine learning (ML) may lead to discovery of metabolic profiles that inform our understanding of pediatric CKD causes. We sought to identify metabolomic signatures in pediatric CKD based on diagnosis: FSGS, obstructive uropathy (OU), aplasia/dysplasia/hypoplasia (A/D/H), and reflux nephropathy (RN).</p>

<p><strong>METHODS: </strong>Untargeted metabolomic quantification (GC-MS/LC-MS, Metabolon) was performed on plasma from 702 Chronic Kidney Disease in Children study participants (: FSGS=63, OU=122, A/D/H=109, and RN=86). Lasso regression was used for feature selection, adjusting for clinical covariates. Four methods were then applied to stratify significance: logistic regression, support vector machine, random forest, and extreme gradient boosting. ML training was performed on 80% total cohort subsets and validated on 20% holdout subsets. Important features were selected based on being significant in at least two of the four modeling approaches. We additionally performed pathway enrichment analysis to identify metabolic subpathways associated with CKD cause.</p>

<p><strong>RESULTS: </strong>ML models were evaluated on holdout subsets with receiver-operator and precision-recall area-under-the-curve, F1 score, and Matthews correlation coefficient. ML models outperformed no-skill prediction. Metabolomic profiles were identified based on cause. FSGS was associated with the sphingomyelin-ceramide axis. FSGS was also associated with individual plasmalogen metabolites and the subpathway. OU was associated with gut microbiome-derived histidine metabolites.</p>

<p><strong>CONCLUSION: </strong>ML models identified metabolomic signatures based on CKD cause. Using ML techniques in conjunction with traditional biostatistics, we demonstrated that sphingomyelin-ceramide and plasmalogen dysmetabolism are associated with FSGS and that gut microbiome-derived histidine metabolites are associated with OU.</p>

DOI

10.1681/ASN.2021040538

Alternate Title

J Am Soc Nephrol

PMID

35017168

Title

Decreased Neural Connectivity in the Default Mode Network Among Youth and Young Adults With Chronic Kidney Disease.

Year of Publication

2021

Number of Pages

455-461

Date Published

2021 Sep

ISSN Number

1558-4488

Abstract

<p>An increasing amount of literature has indicated that chronic kidney disease (CKD) is associated with cognitive deficits that increase with worsening disease severity. Although abnormalities in brain structure have been widely documented, few studies to date have examined the functioning of brain areas associated with the specific cognitive domains affected by CKD (namely, attention and executive functions). Furthermore, few studies have examined functional connectivity among CKD youth who are relatively early in the course of the disease. The present study used functional magnetic resonance imaging to examine the resting state connectivity in 67 youth with CKD (mean age, 17 y) and 58 age-matched healthy controls. Using seed-based multiple regression, decreased connectivity was observed within the anterior cingulate portion of the default mode network. In addition, decreased connectivity within the dorsolateral prefrontal cortex, paracingulate gyrus, and frontal pole were correlated significantly with disease severity. These data indicate that connectivity deficits in circuits implementing attentional processes may represent an early marker for cognitive decline in CKD.</p>

DOI

10.1016/j.semnephrol.2021.09.008

Alternate Title

Semin Nephrol

PMID

34916007

Title

Using a Multi-Institutional Pediatric Learning Health System to Identify Systemic Lupus Erythematosus and Lupus Nephritis: Development and Validation of Computable Phenotypes.

Year of Publication

2021

Date Published

2021 Nov 03

ISSN Number

1555-905X

Abstract

<p><strong>BACKGROUND AND OBJECTIVES: </strong>Performing adequately powered clinical trials in pediatric diseases, such as SLE, is challenging. Improved recruitment strategies are needed for identifying patients.</p>

<p><strong>DESIGN, SETTING, PARTICIPANTS, &amp; MEASUREMENTS: </strong>Electronic health record algorithms were developed and tested to identify children with SLE both with and without lupus nephritis. We used single-center electronic health record data to develop computable phenotypes composed of diagnosis, medication, procedure, and utilization codes. These were evaluated iteratively against a manually assembled database of patients with SLE. The highest-performing phenotypes were then evaluated across institutions in PEDSnet, a national health care systems network of &gt;6.7 million children. Reviewers blinded to case status used standardized forms to review random samples of cases (=350) and noncases (=350).</p>

<p><strong>RESULTS: </strong>Final algorithms consisted of both utilization and diagnostic criteria. For both, utilization criteria included two or more in-person visits with nephrology or rheumatology and ≥60 days follow-up. SLE diagnostic criteria included absence of neonatal lupus, one or more hydroxychloroquine exposures, and either three or more qualifying diagnosis codes separated by ≥30 days or one or more diagnosis codes and one or more kidney biopsy procedure codes. Sensitivity was 100% (95% confidence interval [95% CI], 99 to 100), specificity was 92% (95% CI, 88 to 94), positive predictive value was 91% (95% CI, 87 to 94), and negative predictive value was 100% (95% CI, 99 to 100). Lupus nephritis diagnostic criteria included either three or more qualifying lupus nephritis diagnosis codes (or SLE codes on the same day as glomerular/kidney codes) separated by ≥30 days or one or more SLE diagnosis codes and one or more kidney biopsy procedure codes. Sensitivity was 90% (95% CI, 85 to 94), specificity was 93% (95% CI, 89 to 97), positive predictive value was 94% (95% CI, 89 to 97), and negative predictive value was 90% (95% CI, 84 to 94). Algorithms identified 1508 children with SLE at PEDSnet institutions (537 with lupus nephritis), 809 of whom were seen in the past 12 months.</p>

<p><strong>CONCLUSIONS: </strong>Electronic health record-based algorithms for SLE and lupus nephritis demonstrated excellent classification accuracy across PEDSnet institutions.</p>

DOI

10.2215/CJN.07810621

Alternate Title

Clin J Am Soc Nephrol

PMID

34732529

Title

Metabolite Biomarkers of CKD Progression in Children.

Year of Publication

2021

Number of Pages

1178-1189

Date Published

2021 Aug

ISSN Number

1555-905X

Abstract

<p><strong>BACKGROUND AND OBJECTIVES: </strong>Metabolomics facilitates the discovery of biomarkers and potential therapeutic targets for CKD progression.</p>

<p><strong>DESIGN, SETTING, PARTICIPANTS, &amp; MEASUREMENTS: </strong>We evaluated an untargeted metabolomics quantification of stored plasma samples from 645 Chronic Kidney Disease in Children (CKiD) participants. Metabolites were standardized and logarithmically transformed. Cox proportional hazards regression examined the association between 825 nondrug metabolites and progression to the composite outcome of KRT or 50% reduction of eGFR, adjusting for age, sex, race, body mass index, hypertension, glomerular versus nonglomerular diagnosis, proteinuria, and baseline eGFR. Stratified analyses were performed within subgroups of glomerular/nonglomerular diagnosis and baseline eGFR.</p>

<p><strong>RESULTS: </strong>Baseline characteristics were 391 (61%) male; median age 12 years; median eGFR 54 ml/min per 1.73 m; 448 (69%) nonglomerular diagnosis. Over a median follow-up of 4.8 years, 209 (32%) participants developed the composite outcome. Unique association signals were identified in subgroups of baseline eGFR. Among participants with baseline eGFR ≥60 ml/min per 1.73 m, two-fold higher levels of seven metabolites were significantly associated with higher hazards of KRT/halving of eGFR events: three involved in purine and pyrimidine metabolism (N6-carbamoylthreonyladenosine, hazard ratio, 16; 95% confidence interval, 4 to 60; 5,6-dihydrouridine, hazard ratio, 17; 95% confidence interval, 5 to 55; pseudouridine, hazard ratio, 39; 95% confidence interval, 8 to 200); two amino acids, C-glycosyltryptophan, hazard ratio, 24; 95% confidence interval 6 to 95 and lanthionine, hazard ratio, 3; 95% confidence interval, 2 to 5; the tricarboxylic acid cycle intermediate 2-methylcitrate/homocitrate, hazard ratio, 4; 95% confidence interval, 2 to 7; and gulonate, hazard ratio, 10; 95% confidence interval, 3 to 29. Among those with baseline eGFR &lt;60 ml/min per 1.73 m, a higher level of tetrahydrocortisol sulfate was associated with lower risk of progression (hazard ratio, 0.8; 95% confidence interval, 0.7 to 0.9).</p>

<p><strong>CONCLUSIONS: </strong>Untargeted plasma metabolomic profiling facilitated discovery of novel metabolite associations with CKD progression in children that were independent of established clinical predictors and highlight the role of select biologic pathways.</p>

DOI

10.2215/CJN.00220121

Alternate Title

Clin J Am Soc Nephrol

PMID

34362785

Title

Association of Multiple Plasma Biomarker Concentrations with Progression of Prevalent Diabetic Kidney Disease: Findings from the Chronic Renal Insufficiency Cohort (CRIC) Study.

Year of Publication

2021

Number of Pages

115-126

Date Published

2021 01

ISSN Number

1533-3450

Abstract

<p><strong>BACKGROUND: </strong>Although diabetic kidney disease is the leading cause of ESKD in the United States, identifying those patients who progress to ESKD is difficult. Efforts are under way to determine if plasma biomarkers can help identify these high-risk individuals.</p>

<p><strong>METHODS: </strong>In our case-cohort study of 894 Chronic Renal Insufficiency Cohort Study participants with diabetes and an eGFR of &lt;60 ml/min per 1.73 m at baseline, participants were randomly selected for the subcohort; cases were those patients who developed progressive diabetic kidney disease (ESKD or 40% eGFR decline). Using a multiplex system, we assayed plasma biomarkers related to tubular injury, inflammation, and fibrosis (KIM-1, TNFR-1, TNFR-2, MCP-1, suPAR, and YKL-40). Weighted Cox regression models related biomarkers to progression of diabetic kidney disease, and mixed-effects models estimated biomarker relationships with rate of eGFR change.</p>

<p><strong>RESULTS: </strong>Median follow-up was 8.7 years. Higher concentrations of KIM-1, TNFR-1, TNFR-2, MCP-1, suPAR, and YKL-40 were each associated with a greater risk of progression of diabetic kidney disease, even after adjustment for established clinical risk factors. After accounting for competing biomarkers, KIM-1, TNFR-2, and YKL-40 remained associated with progression of diabetic kidney disease; TNFR-2 had the highest risk (adjusted hazard ratio, 1.61; 95% CI, 1.15 to 2.26). KIM-1, TNFR-1, TNFR-2, and YKL-40 were associated with rate of eGFR decline.</p>

<p><strong>CONCLUSIONS: </strong>Higher plasma levels of KIM-1, TNFR-1, TNFR-2, MCP-1, suPAR, and YKL-40 were associated with increased risk of progression of diabetic kidney disease; TNFR-2 had the highest risk after accounting for the other biomarkers. These findings validate previous literature on TNFR-1, TNFR-2, and KIM-1 in patients with prevalent CKD and provide new insights into the influence of suPAR and YKL-40 as plasma biomarkers that require validation.</p>

DOI

10.1681/ASN.2020040487

Alternate Title

J Am Soc Nephrol

PMID

33122288

Title

Computer-Aided Diagnosis of Congenital Abnormalities of the Kidney and Urinary Tract in Children Using a Multi-Instance Deep Learning Method Based on Ultrasound Imaging Data.

Year of Publication

2020

Number of Pages

1347-1350

Date Published

2020 Apr

ISSN Number

1945-7928

Abstract

<p>Ultrasound images are widely used for diagnosis of congenital abnormalities of the kidney and urinary tract (CAKUT). Since a typical clinical ultrasound image captures 2D information of a specific view plan of the kidney and images of the same kidney on different planes have varied appearances, it is challenging to develop a computer aided diagnosis tool robust to ultrasound images in different views. To overcome this problem, we develop a multi-instance deep learning method for distinguishing children with CAKUT from controls based on their clinical ultrasound images, aiming to automatic diagnose the CAKUT in children based on ultrasound imaging data. Particularly, a multi-instance deep learning method was developed to build a robust pattern classifier to distinguish children with CAKUT from controls based on their ultrasound images in sagittal and transverse views obtained during routine clinical care. The classifier was built on imaging features derived using transfer learning from a pre-trained deep learning model with a mean pooling operator for fusing instance-level classification results. Experimental results have demonstrated that the multi-instance deep learning classifier performed better than classifiers built on either individual sagittal slices or individual transverse slices.</p>

DOI

10.1109/isbi45749.2020.9098506

Alternate Title

Proc IEEE Int Symp Biomed Imaging

PMID

33850604

Title

Kidney Outcomes and Hypertension in Survivors of Wilms Tumor: A Prospective Cohort Study.

Year of Publication

2020

Date Published

2020 Dec 05

ISSN Number

1097-6833

Abstract

<p>Supported by a Pilot Grant from the Children's Hospital of Philadelphia Center for Pediatric Clinical Effectiveness (to D.C.). D.C. is also supported by the NIH/NIDDK (K23 DK125670). G.T. was supported by the NIH/NIDDK (K23 DK106428). Ja.G. was supported by NIH/NIDDK (K08 DK110536). M.D. was supported by the NIH/NIDDK (K23 DK093556). The NIH and NIDDK had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript. The views expressed in this article are those of the authors and do not necessarily represent the official view of the NIDDK nor NIH. G.T. serves on the scientific advisory boards for Allena Pharmaceuticals, Novome Biotechnology, and Dicerna Pharmaceuticals and serves as a consultant for Alnylam Pharmaceuticals, all of which are unrelated to this work. M.D. receives research funding from Mallinckrodt unrelated to this work. The other authors declare no conflicts of interest. Portions of this study were presented at the Pediatric Academic Society annual meeting, May 5-8, 2020, Toronto, Canada.</p>

<p><strong>OBJECTIVE: </strong>To assess the prevalence of therapy-related kidney outcomes in survivors of Wilms tumor (WT).</p>

<p><strong>STUDY DESIGN: </strong>This prospective cohort study included survivors of WT who were ≥5 years old and ≥1 year from completing therapy, excluding those with pre-existing hypertension, prior dialysis or kidney transplant. Participants completed 24-hour ambulatory blood pressure monitoring (ABPM). Abnormal blood pressure (BP) was defined as ≥90 percentile. Masked hypertension was defined as having normal office BP and abnormal ABPM findings. Urine was analyzed for KIM-1, IL-18, EGF, albumin, and creatinine. Estimated glomerular filtration rate (eGFR) was calculated using the bedside CKiD equation. Recent kidney ultrasounds and echocardiograms were reviewed for contralateral kidney size and left ventricular hypertrophy (LVH), respectively. Clinical follow-up data was collected for approximately 2 years following study enrollment.</p>

<p><strong>RESULTS: </strong>Thirty-two participants (median age 13.6 [IQR: 10.5-16.3] years; 75% ≥Stage 3 WT) were evaluated at a median of 8.7 years (IQR: 6.5-10.8) post-therapy; 29 participants underwent unilateral radical nephrectomy, two bilateral partial nephrectomy, and one radical and contralateral partial nephrectomy. 72% received kidney radiotherapy and 75% received doxorubicin. Recent median eGFR was 95.6 ml/min/1.73m (IQR: 84.6-114.0; 11 (34%) had an eGFR &lt;90). Abnormal ABPM results were found in 22/29 participants (76%), masked hypertension in 10/29 (34%), and microalbuminuria in 2/32 (6%). 22/32 (69%) participants had abnormal EGF; few had abnormal KIM-1 or IL-18. Seven participants with previous unilateral nephrectomy lacked compensatory contralateral kidney hypertrophy. None had LVH.</p>

<p><strong>CONCLUSION: </strong>In survivors of WT, adverse kidney outcomes were common and should be closely monitored.</p>

DOI

10.1016/j.jpeds.2020.12.005

Alternate Title

J Pediatr

PMID

33290810

Title

Association Between Chronic Kidney Disease-Mineral Bone Disease (CKD-MBD) and Cognition in Children: Chronic Kidney Disease in Children (CKiD) Study.

Year of Publication

2020

Number of Pages

398-406

Date Published

2020 Jul-Aug

ISSN Number

2590-0595

Abstract

<p><strong>Rationale &amp; Objective: </strong>Chronic kidney disease (CKD) in children is associated with cognitive dysfunction that affects school performance and quality of life. The relationship between CKD-mineral and bone disorder and cognitive function in children is unknown.</p>

<p><strong>Study Design: </strong>Observational study.</p>

<p><strong>Participants: </strong>702 children enrolled in the Chronic Kidney Disease in Children (CKiD) Study.</p>

<p><strong>Predictors: </strong>Plasma fibroblast growth factor 23 (FGF-23), parathyroid hormone (PTH), calcium, phosphorus, 25 hydroxyvitamin D (25[OH]D), and 1,25 dihydroxyvitamin D (1,25[OH]D).</p>

<p><strong>Outcomes: </strong>Neurocognitive tests of intelligence, academic achievement, and executive functions.</p>

<p><strong>Analytical Approach: </strong>Linear regression models to analyze the cross-sectional associations between logFGF-23, 25(OH)D, 1,25(OH)D, PTH, calcium, and phosphorus scores and the cognitive test scores of interest after adjustment for demographics, blood pressure, proteinuria, and kidney function.</p>

<p><strong>Results: </strong>At baseline, median age was 12 (95% CI, 8.3, 15.2) years and estimated glomerular filtration rate was 54 (40.5, 67.8) mL/min/1.73&nbsp;m. In fully adjusted analyses, 25(OH)D, 1,25(OH)D, PTH, calcium, and phosphorus scores did not associate with cognitive test scores. In fully adjusted analyses, logFGF-23 was associated with abnormal test scores for attention regulation (&nbsp;&lt;&nbsp;0.05); specifically, Conners' Continuous Performance Test II Errors of Omission (β&nbsp;=&nbsp;2.3&nbsp;[1.0, 3.6]), Variability (β=1.4 [0.4,&nbsp;-2.4]), and Hit Reaction Time (β&nbsp;=&nbsp;1.3&nbsp;[0.2, 2.4]). Children in the highest FGF-23 tertile group had 7% and 9% greater cognitive risk for&nbsp;Hit Reaction Time and Errors of Omission compared with those in the lowest tertile, respectively. In fully adjusted analyses, higher FGF-23 tertile was associated with increased cognitive risk (&nbsp;&lt;&nbsp;0.05) for Errors of Omission (β&nbsp;=&nbsp;0.4&nbsp;[0.1,&nbsp;0.7]) and Hit Reaction Time (β&nbsp;=&nbsp;0.4&nbsp;[0.1, 0.7]).</p>

<p><strong>Limitations: </strong>The study does not assess the cumulative&nbsp;effects of FGF-23 excess on cognitive function over time. Within-population stratified analyses were not performed due to limited sample size.</p>

<p><strong>Conclusions: </strong>In children with CKD, higher plasma FGF-23 level is associated with lower performance in targeted tests of executive function, specifically attention regulation, independent of glomerular filtration rate.</p>

DOI

10.1016/j.xkme.2020.03.005

Alternate Title

Kidney Med

PMID

32775979

Title

Magnetic resonance elastography to quantify liver disease severity in autosomal recessive polycystic kidney disease.

Year of Publication

2020

Date Published

2020 Aug 05

ISSN Number

2366-0058

Abstract

<p><strong>OBJECTIVES: </strong>To evaluate whether liver and spleen magnetic resonance elastography (MRE) can measure the severity of congenital hepatic fibrosis (CHF) and portal hypertension (pHTN) in individuals with autosomal recessive polycystic kidney disease (ARPKD), and to examine correlations between liver MRE and ultrasound (US) elastography.</p>

<p><strong>METHODS: </strong>Cross-sectional study of nine individuals with ARPKD and 14 healthy controls. MRE was performed to measure mean liver and spleen stiffness (kPa); US elastography was performed to measure point shear wave speed (SWS) in both liver lobes. We compared: (1) MRE liver and spleen stiffness between controls vs. ARPKD; and (2) MRE liver stiffness between participants with ARPKD without vs. with pHTN, and examined correlations between MRE liver stiffness, spleen length, platelet counts, and US elastography SWS. Receiver operating characteristic (ROC) analysis was performed to examine diagnostic accuracy of liver MRE.</p>

<p><strong>RESULTS: </strong>Participants with ARPKD (median age 16.8 [IQR 13.3, 18.9] years) had higher median MRE liver stiffness than controls (median age 14.7 [IQR 9.7, 16.7&nbsp;years) (2.55 vs. 1.92&nbsp;kPa, p = 0.008), but MRE spleen stiffness did not differ. ARPKD participants with pHTN had higher median MRE liver stiffness than those without (3.60&nbsp;kPa vs 2.49&nbsp;kPa, p = 0.05). Liver MRE and US elastography measurements were strongly correlated. To distinguish ARPKD vs. control groups, liver MRE had 78% sensitivity and 93% specificity at a proposed cut-off of 2.48&nbsp;kPa [ROC area 0.83 (95% CI 0.63-1.00)].</p>

<p><strong>CONCLUSION: </strong>Liver MRE may be a useful quantitative method to measure the severity of CHF and pHTN in individuals with ARPKD.</p>

DOI

10.1007/s00261-020-02694-1

Alternate Title

Abdom Radiol (NY)

PMID

32757071

WATCH THIS PAGE

Subscription is not available for this page.