Citations

Showing 10 of 108152 citations.

Application of the CLSI EP15-A3 Guideline as an Alternative Troubleshooting Tool for Verification of Assay Precision
Aguirre, Jose Jara, Ness, Karl M. , Algeciras‐Schimnich, Alicia

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S88-S88

10.1093/ajcp/aqz117.007 Get It

Abstract Introduction The CLSI EP15-A3 guideline “User Verification of Precision and Estimation of Bias” provides a simple experimental approach to estimate a method’s imprecision and bias. The objective is to determine if the laboratory precision performance of repeatability (SR) and within-laboratory imprecision (SWL) are in accordance to the manufacturer specification claims (MSCs). Objectives Evaluate the utility of the EP15-A3 protocol to verify method precision during a troubleshooting investigation and after major instrument maintenance, using a carcinoembryonic antigen (CEA) immunoassay as an example. Methods CEA was performed on the Beckman Coulter DxI (Beckman Coulter, Brea, CA). Quality control (QC) levels (L1: 2.89; L2: 21.10; L3: 39.10 ng/mL) (Bio-Rad Laboratories, Irvine, CA) were used. Each QC level was measured before and after instrument maintenance as follows: five replicates per run, one run per day, and during 5 days. Imprecision estimates (IEs) for SR (%CVR) and SWL (%CVWL) were calculated by one-way analysis of variance using Microsoft Excel Analyse-it software. Estimated imprecision was compared to MSC and desirable imprecision specifications based on biological variation (BV). Results A change in the analytical performance of CEA was detected by a decreased sigma-metric indicator. After a bias problem was ruled out, the observed %CVR for L1, L2, and L3 were 7.2%, 3.7%, and 4.8%, respectively. The %CVWL were 8.3%, 5.0%, and 5.5%, which exceeded the MSC of %CVWL~4.0% to 4.5%. After a laboratory investigation, major instrument maintenance was performed by the manufacturer. The %CVR and %CVWL estimates for L1, L2, and L3 after maintenance were 3.2%, 3.8%, 3.5% and 3.9%, 4.2%, 4.0%, respectively. After maintenance, the CEA performance was consistent with the MSC for each of the levels analyzed and within the BV impression goal of %CV ≤6.4. Conclusion CLSI EP15-A3 guideline is an alternative troubleshooting tool that can be used to investigate and verify method precision performance before and after significant instrument maintenance.
Clinical Validation of the Lymph3Cx Assay to Distinguish Primary Mediastinal Large B-Cell Lymphoma From Diffuse Large B-Cell Lymphoma
Ramsower, Colleen , Yip, Tameson, Glinsmann‐Gibson, Betty, Robetorye, Ryan S. , Rimsza, Lisa M.

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S136-S136

10.1093/ajcp/aqz126.006 Get It

Abstract Introduction Primary mediastinal large B-cell lymphoma (PMLBCL) is recognized as a distinct lymphoid neoplasm. However, differentiation from diffuse large B-cell lymphoma (DLBCL) can be difficult due to the need for close clinicopathologic correlation and variability in morphology and immunophenotype. In addition, a recent study identified lymphomas that share molecular and morphological features with PMBCL, yet do not involve the mediastinum. Correct classification of PMBCL may impact therapeutic decision making, increasing the demand for more accurate diagnostic methods. Recent research identified a new gene expression signature that robustly differentiated PMLBCL from DLBCL in both training and independent validation cohorts using RNA isolated from routinely available formalin-fixed, paraffin-embedded (FFPE) tissues. Methods This assay, known as the “Lymph3Cx,” utilizes the nCounter platform (NanoString Technologies, Seattle, WA) and consists of probes for 58 target genes (including both discriminatory and housekeeping genes) that can distinguish PMLBCL cases from DLBCL as well as the “cell-of-origin” subtypes of DLBCL. In the current study reported here, the Molecular Diagnostics–Arizona Laboratory (MDAZL) performed a full validation of the Lymph3Cx assay for use in our clinical diagnostic laboratory. Results Assay performance included accuracy (92.9%), precision (100%), analytical sensitivity (100 ng RNA limit of detection), analytical specificity (measures are in place to prevent interfering substances), reproducibility between technologists and technical replicates (100%), specimen stability (18 months for unstained tissue sections), and RNA Isolation Kit Utility (two kits with equal performance) in order to become the first CLIA-certified, CAP-accredited clinical diagnostics laboratory to offer a molecular assay for distinction of PMLBCL from DLBCL. Conclusion The Lymph3Cx assay will be prospectively used for evaluation of Mayo Clinic patients and for use as an integrated biomarker in clinical trials.
Decreasing Troponin Redraws by Validating Tiered Hemolysis Thresholds
Harnish, Tanya L., Kaleta, Erin, Snozek, Christine, Turner, Katherine, Madani, Zahra

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S11-S12

10.1093/ajcp/aqz112.022 Get It

Abstract Cardiac troponin T (cTnT) assays are used for the diagnosis of acute myocardial infarctions and require serial measurements. Hemolysis is a common analytical interference for cTnT immunoassays, causing a false decrease in analyte concentration. Recollection of specimens that do not meet the recommended hemolysis threshold (H-index = 100) causes reporting delays and mistiming of serial measurements. This has been particularly disruptive to our emergency department and creates significant risk for patients whose diagnosis could be delayed by recollection due to hemolysis. Here we aimed to reevaluate the limits for acceptable hemolysis by determining the magnitude of cTnT concentration depression from hemolysis to evaluate whether more detailed thresholds could be established. To quantify the effects of hemolysis on cTnT, patient pools were prepared from residual serum with cTnT concentrations ranging from 10 to 100 ng/L and spiked with hemolysate prepared from lysed red blood cells to create H-indices ranging from 120 to 200. Samples were run in triplicate by the Elecsys Troponin T Gen. 5 STAT assay. Results demonstrated consistent percent decreases in cTnT across all concentrations tested for each level of hemolysis. The mean percent changes in cTnT concentrations in the presence of hemolysis for H-indices of 120, 140, 160, 180, and 200 were –3.4 ± 1.3%, –4.1 ± 1.2%, –6.3 ± 1.2%, –8.0 ± 1.7%, and –10.7 ± 1.22%, respectively. The observed decrease in cTnT was linearly related to the H index; predicted differences at higher H-indices (>200) agreed well with prior publications evaluating greater degrees of hemolysis. In our practice, a 2-hour delta ≥10 ng/L is considered significant for acute cardiac injury, <4 ng/L is a nonsignificant delta, and a delta of 4 to 9 ng/L is considered indeterminate. Baseline cTnT results of ≥100 ng/L result in immediate triage to cardiology. Approximately one-third of our patients with cTnT testing have baseline results within the reference range (≤15 ng/L males, ≤10 ng/L females). Based on the spiking data, H-index cutoffs were chosen to minimize recollections for low-risk and high-risk patients. Cutoffs for intermediate cTnT results were more restrictive to ensure delta interpretation would not change significantly. This resulted in a H-index of 300 for samples ≤8 ng/L, 200 for 9 to 40 ng/L, 160 for 41 to 70 ng/L, 140 for 71 to 99 ng/L, and 300 for ≥100 ng/L. These data quantify the percent change for cTnT in the presence of varying levels of hemolysis. At lower cTnT values, a larger degree of hemolysis can be tolerated because the percentage of depression results in a small absolute change, thus leading to less impact on the delta. The tiered H-index cutoffs allow minimal disruption to patient care for low- and high-risk patients, while maintaining the integrity of serial measurements for those with intermediate cTnT concentrations. Therefore, laboratories may consider releasing some hemolyzed cTnT specimens with a comment to decrease redraws and mistiming of serial measurements.
Tumor-Infiltrating Lymphocytes and Programmed Death-Ligand 1 Expression in Salivary Duct Carcinoma
Chatzopoulos, Kyriakos , Collins, Andrea , Garcia, Joaquin

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S40-S41

10.1093/ajcp/aqz113.009 Get It

Abstract Objectives Salivary duct carcinoma is a rare high-grade salivary gland tumor, sharing common features with invasive ductal carcinoma of the breast—particularly apocrine type. Increased density of tumor-infiltrating lymphocytes (TILs) is a favorable prognostic factor in many solid tumors and has been correlated with elevated expression of programmed death-ligand 1 (PD-L1), a targetable molecule. We investigated the prognostic role of TILs and PD-L1 expression in salivary duct carcinoma. Methods Clinical records and archived surgical pathology material were evaluated from 28 patients with a diagnosis of salivary duct carcinoma at Mayo Clinic Rochester (between 1961 and 2007). TILs were evaluated on hematoxylin and eosin–stained slides and reported as continuous variables. Immunohistochemistry for PD-L1 (clone 22C3) was performed on unstained slides and evaluated using the Combined Positive Score (CPS). Fisher’s exact test was used to assess frequency distribution differences. Log-rank test and Cox proportional hazards model were used to assess the impact of the variables on survival. Results Twenty-eight patients had salivary duct carcinoma (age range 35-87, mean 61), and follow-up averaged 60 months (range 6-252 months). At the end of the follow-up period, 22 patients (79%) had died of salivary duct carcinoma, while remaining patients were alive (4.14%) or had died of other causes (2.7%). For the entire follow-up period, levels of TILs of 40% or more were associated with favorable overall survival (log-rank test, P = .01) and were predictors of better survival in univariate analysis (hazard ratio: 0.18, P = .02). PD-L1 expression was not significantly different between the high and low TIL groups (Fisher’s exact test, P = .67). Survival analysis regarding PD-L1 expression did not yield statistically significant results (log-rank test, P = .74; hazard ratio: 0.87, P = .74). Conclusion High TIL levels, but not PD-L1 expression as determined using the CPS, predict better survival in patients with salivary duct carcinoma.
Analytical Evaluation of BioRad D-100 HPLC Analyzer and Workflow Comparison to BioRad TurboLink Variant II HPLC With Reflex to Boronate Affinity
Galior, Kornelia , Leonard, L M, Dolan, Christopher, Deiter, Ryan, Baumann, N, Block, Darci R

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S12-S13

10.1093/ajcp/aqz112.024 Get It

Abstract Introduction Measurement of hemoglobin A1c (HbA1C) is used for the diagnosis and management of patients with diabetes. Methods for measuring HbA1C are classified on the basis of charge differences (cation exchange chromatography) or structural differences (boronate affinity chromatography). Some cation exchange high-performance liquid chromatography (HPLC) analyzers may be prone to interferences from hemoglobin variants. Historically, our lab used two methods to report HbA1C results: cation exchange HPLC (VariantII) with reflex to boronate affinity HPLC (Ultra2) methods. A new analyzer (BioRad D-100) with improved interference detection and thresholds for interference was evaluated. The objectives of this study were (1) assess the comparability of HbA1c results between D-100, VariantII, and Ultra2; (2) evaluate the need for maintaining a reflex method; and (3) calculate cancellation rates before and after implementing D-100. Methods HbA1c was measured by cation exchange methods using VariantII Turbolink, D-100 (BioRad), and boronate affinity using Ultra2Affinity (Trinity Biotech) according to the manufacturer’s recommendations with the following analytical measuring ranges (AMRs), 4.0% to 17.6%, 4.0% to 18.0%, and 4.0% to 17.6%, respectively. D-100 was compared to VariantII (n = 26) and Ultra2 (n = 31) using residual patient samples from provider-ordered HbA1c having no chromatography flags and with the following chromatography flags on VariantII (n = 113): variant window (n = 41), HbA1c <4.0% (n = 24), HbA1a + HbA1b >5% (n = 19), HbF >5% (n = 9), P3 >10% (n = 6), HbA1c >17.6% (n = 4), unknown peak (n = 3), labile 5% to 15% (n = 3), labile >15% (n = 3), and HbS >60% (n = 1). The cancelation rates were calculated at baseline (January 1-31, 2018) and postimplementation of D-100 (September 20, 2018, to January 22, 2019). Results HbA1c results from D-100 were within ±0.3 or 5% with the following frequencies: Ultra2 in 96.2% (25/26) with VariantII in 96.8% (30/31) using samples with no chromatography flags. A total of 113 samples with abnormal VariantII flags were tested using D-100 with 82% (n = 92) not exceeding interference limits for result reporting. These D-100 results were within ±0.3 or 5% of Ultra2 results with the following frequencies: 38.5% (10/26) of samples below (n = 23) or above (n = 3) the AMR, 85.3% (29/34) of results with Hb variant flags, 68.0% (17/25) with unknown/minor peaks, and 71.4% (5/7) with HbF <15%. The clinical concordance was also assessed according to the following decision limits: normal <5.6%, prediabetic = 5.7% to 6.4%, and diabetic >6.5%. Results were concordant in 88% with variant peaks (12/13 w/HbA1c <5.6%, 8/11 w/HbA1c = 5.7%-6.4%, 10/10 w/HbA1c >6.5%), 84% with minor peaks (5/6 w/HbA1c <5.6%, 0/2 w/HbA1c = 5.7%-6.4%, 16/17 w/HbA1c >6.5%), and 86% with HbF (3/3 w/HbA1c <5.6%, 2/3 w/HbA1c = 5.7%-6.4%, 1/1 w/HbA1c >6.5%). Results above/below AMR had 100% concordance (>16.0% or <4.4%). The frequency of VariantII rule violation was 2.1% in January 2018, with 53 of 117 having results confirmed as <4.0% or >16.0% by Ultra2, 2 of 117 reported by Ultra2, and 62 being cancelled due to HbF >15% (n = 23) or variant Hb >60% (n = 39). The frequency of rule violation using only D-100 since implementation was 2.2%. Conclusion Implementation of D-100 has removed the need for boronate affinity reflex testing and maintained similar cancellation rates.
Assessment of Macroprolactin Persistence in a Large Clinical Population
Delaney, Sarah , Bornhorst, Joshua A. , Bornhorst, Joshua , Delaney, Sarah, Algeciras‐Schimnich, Alicia

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S12-S12

10.1093/ajcp/aqz112.023 Get It

Abstract Background Pathologic causes of hyperprolactinemia include prolactin-secreting pituitary adenoma (prolactinoma), diseases of the hypothalamus, primary hypothyroidism, and renal failure. Another cause of hyperprolactinemia is the presence of macroprolactin, a high-molecular-mass complex of prolactin and immunoglobulins, which constitutes a biologically inactive form of prolactin. Detection of macroprolactin is important to avoid misdiagnosis and inappropriate treatment; however, relatively little is known about the persistence of macroprolactin in patients over time or how macroprolactin concentrations evolve within patients. The objective of this study was to examine the changes in macroprolactin and prolactin over time in a large clinical population. Methods All physician-ordered macroprolactin tests from December 2011 to October 2018 (n = 5,318 unique patients) were examined. Data were extracted from patients with repeat collections (≥2 timepoints) during this 7-year period (n = 269). Macroprolactin testing was performed using the Roche Elecsys Prolactin II assay. Prior to analysis, an equal volume of 25% polyethylene glycol (PEG) was added to patient serum, causing macroprolactin to precipitate. Following centrifugation, unprecipitated prolactin was measured in the supernatant, as well as an aliquot of the native sample (measuring total prolactin). Macroprolactin (precipitated prolactin) was calculated by subtracting the unprecipitated prolactin from the total prolactin concentration. For this study, a specimen was considered positive for macroprolactin if the percentage of precipitated prolactin was ≥60%. Results A total of 14% (37/269) of patients tested positive for macroprolactin in at least one of the sequential determinations. For all patients with two or more timepoints, the mean interdetermination change in percent macroprolactin was +0.6% and the mean interdetermination change in total prolactin was –11.6 ng/mL. The mean time between determinations in these patients was 303 days. For the 37 patients that had a positive macroprolactin determination, 25 remained positive across all sample collections (mean interdetermination change of +0.5% macroprolactin, mean interdetermination change in total prolactin of +0.4 ng/mL), seven samples went from positive to negative (mean interdetermination change of –10% macroprolactin, mean change in total prolactin of +0.2 ng/mL), and two samples went from negative to positive (mean interdetermination change of +5.5% macroprolactin, mean change in total prolactin of +1.3 ng/mL). The remaining 3 of the 37 specimens had more than two determinations that fluctuated above and below the predefined cutoff. Conclusions The results from this study suggest that the presence of macroprolactin largely but not exclusively persists over time. Additionally, macroprolactin rarely appears after a previously negative assessment (2/235 patients). Furthermore, only small increases in sequential percent macroprolactin determinations were observed in this data set. The prevalence of the persistence of macroprolactin protein over time might have important implications for patient prolactin monitoring.
Effect of pH on the Quantification of Body Fluid Analytes for Clinical Diagnostic Testing
Nandakumar, Vijayalakshmi, Dolan, Christopher, Baumann, N, Block, Darci R

Mayo Clinic

American Journal of Clinical Pathology | September 11, 2019 | Volume: 152 | Issue: Supplement_1 | Pages: S10-S11

10.1093/ajcp/aqz112.021 Get It

Abstract Background Body fluid samples (BFs) with low pH have demonstrated underrecovery of enzyme activity. The aims of this study were (1) measure the frequency of abnormal pH in BFs received for clinical chemistry testing, (2) calculate recovery of analytes in BFs undergoing pH titration, and (3) investigate the mechanism of pH interference. Methods Residual BFs submitted for physician-ordered testing to the Central Clinical Laboratory (Mayo Clinic) were utilized for this study. BFs were centrifuged 10 minutes (3,200 rpm) prior to pH measurement (pH meter, Accumet X15; Fisher Scientific) and analysis. A Roche Cobas c501 (Roche Diagnostics) analyzer was used to measure amylase, lactate dehydrogenase (LDH), blood urea nitrogen (BUN), creatinine, glucose, albumin, total protein, bilirubin (all Roche Diagnostics reagents), and lipase (Sekisui Diagnostics). pH was measured in 122 BFs received between 10/23/2018 and 10/26/2018 within 24 hours of receipt. pH titration was performed using pooled BFs of similar source/type. Each pool was titrated over a range of pH (≤2 to ≥10) in 1-unit increments by addition of a fixed volume of acid (6M HCL) or base (6M NaOH). Analytes were measured before and after acid/base addition. Average percent recovery was calculated (measured/expected × 100%) from n = 3 to 9 pH measurements. A low pH BF pool was spiked (<5% by volume) with a high-concentration BF to investigate mechanism of enzyme underrecovery. Additionally, low pH BF pool was neutralized by addition of base and % recovery calculated. For all titration/spiking experiments, a control sample having the same volume of diluent (7% bovine serum albumin or saline) was used to account for the effects of dilution. Results BFs received (n = 122) had mean (SD) pH = 8.0 (0.6) with 6.6% (n = 8) having pH <7.0 and 2.5% (n = 3) having pH <6.0. All recovery (%) and pH data are expressed as mean (range) values. Amylase recovery (initial pH = 8.4-8.9) was 1.3 (0.9-2.2)% at pH = 1.6 (1.0-2.2) (n = 6), 55.6 (37.4-67.0)% at pH = 4.4 (3.5-5.4) (n = 6), 81.4 (74.0-85.2)% at pH = 6.2 (6.1-6.5) (n = 3), 90.1 (84.4-95.4)% at pH = 7.5 (7.3-7.7) (n = 3), 98.8 (96.2-101.7)% at pH = 10.0 (9.8-10.2) (n = 3), and 14.5 (1.4-38.9)% at pH = 11.9 (11.5-12.1) (n = 3). LDH recovery (initial pH = 8.2-8.5) was 1.4 (0.1-3.7)% at pH = 1.7 (1.0-2.4) (n = 6), 18.3 (0.5-39.7)% at pH = 4.5 (3.5-5.4) (n = 6), 63.6 (54.9-68.4)% at pH = 6.2 (6.1-6.5) (n = 3), 85.9 (80.2-90.7)% at pH = 7.2 (6.8-7.4) (n = 3), 46.1 (36.4-61.0)% at pH = 10.0 (9.8-10.2) (n = 3), and 9.8 (0.0-28.9)% at pH = 11.3 (10.4-12.1) (n = 3). Lipase recovery (initial pH = 8.2-8.9) was 10.4 (<1-18)% at pH = 1.8 (1.0-2.4) (n = 6), 73.9 (62.6-85.5)% at pH = 4.2 (3.4-5.2) (n = 3), 92.4 (90.8-93.5)% at pH = 6.2 (6.0-6.5) (n = 3), 96.3 (95.6-96.8)% at pH = 7.3 (6.8-7.7) (n = 3), and 92.9 (89.3-96.8)% at pH = 9.9 (9.8-10.1) (n = 3) and 80.9 (0.0-78.6)% at pH = 11.3 (10.4-12.1) (n = 3). Creatinine, BUN, albumin, glucose, total protein, and bilirubin recovery were 97.9 (92.2-102.9)%, 100 (96.6-103.1)%, 99.9 (97.6-102.4)%, 100.2 (97.8-103)%, 100.2 (98.7-101.9)%, and 100.1 (92.-105.9)%, respectively, between pH 1.3-11.7 (n = 9). Recovery of amylase, LDH, or lipase was <1% after spiking enzymes into BF pool with pH <3. pH adjustment to normal (pH = 8.6-9.0) also resulted in recovery of <1%. Conclusion Enzyme activity in BFs was not affected (90%-110% recovery) when pH = 7.4-8.5 for LDH, pH = 7.3-10.2 for amylase, pH = 6.0-9.9 for lipase, and pH = 1.3-11.7 for all other analytes. An irreversible loss of enzyme activity occurs at low pH. Few clinical BFs have pH <7.0, but laboratories should have awareness.
CD24hiCD38hi and CD24hiCD27+ Human Regulatory B Cells Display Common and Distinct Functional Characteristics
Hasan, Md Mahmudul , Thompson-Snipes, LuAnn , Klintmalm, Göran B. , Demetris, Anthony J. , O’Leary, Jacqueline G. , Oh, SangKon , Joo, HyeMee , Joo, HyeMee

Mayo Clinic

The Journal of Immunology | September 11, 2019 | Volume: 203 | Issue: 8 | Pages: 2110-2120

10.4049/jimmunol.1900488 | PMID 31511354 Get It

Abstract Although IL-10–producing regulatory B cells (Bregs) play important roles in immune regulation, their surface phenotypes and functional characteristics have not been fully investigated. In this study, we report that the frequency of IL-10–producing Bregs in human peripheral blood, spleens, and tonsils is similar, but they display heterogenous surface phenotypes. Nonetheless, CD24hiCD38hi transitional B cells (TBs) and CD24hiCD27+ B cells (human equivalent of murine B10 cells) are the major IL-10–producing B cells. They both suppress CD4+ T cell proliferation as well as IFN-γ/IL-17 expression. However, CD24hiCD27+ B cells were more efficient than TBs at suppressing CD4+ T cell proliferation and IFN-γ/IL-17 expression, whereas they both coexpress IL-10 and TNF-α. TGF-β1 and granzyme B expression were also enriched within CD24hiCD27+ B cells, when compared with TBs. Additionally, CD24hiCD27+ B cells expressed increased levels of surface integrins (CD11a, CD11b, α1, α4, and β1) and CD39 (an ecto-ATPase), suggesting that the in vivo mechanisms of action of the two Breg subsets are not the same. Lastly, we also report that liver allograft recipients with plasma cell hepatitis had significant decreases of both Breg subsets.
Amyloid, Vascular, and Resilience Pathways Associated with Cognitive Aging
Vemuri, Prashanthi , Lesnick, Timothy G. , Przybelski, Scott A. , Reid, Robert I. , Graff‐Radford, Jonathan , Lowe, Val J. , Petersen, Ronald C. , Jack, Clifford R. , Reid, Robert I , Jack, Clifford R , Vemuri, Prashanthi , Knopman, David S. , Mielke, Michelle M. , Machulda, Mary M.

Mayo Clinic

Annals of Neurology | September 11, 2019 | Volume: 86 | Issue: 6 | Pages: 866-877

10.1002/ana.25600 | PMID 31509621 Get It

ObjectiveTo investigate the multifactorial processes underlying cognitive aging based on the hypothesis that multiple causal pathways and mechanisms (amyloid, vascular, and resilience) influence longitudinal cognitive decline in each individual through worsening brain health.MethodsWe identified 1,230 elderly subjects (aged ≥50 years) with an average of 4.9 years of clinical follow‐up and with amyloid positron emission tomography, diffusion tensor imaging, and structural magnetic resonance imaging scans from the population‐based Mayo Clinic Study of Aging. We examined imaging markers of amyloid and brain health (white matter microstructural integrity and cortical thinning), systemic vascular health preceding the imaging markers, and early to midlife intellectual enrichment to predict longitudinal cognitive trajectories. We used latent growth curve models for modeling longitudinal cognitive decline.ResultsAll the pathways (amyloid, vascular, resilience) converged through their effects on cortical thinning and worsening cognition and together explained patterns in cognitive decline. Resilience and vascular pathways (aging process, sex differences, education/occupation, and systemic vascular health) had significant impact on white matter microstructural integrity. Education/occupation levels contributed to white matter integrity through systemic vascular health. Worsening white matter integrity contributed to significant cortical thinning and subsequently longitudinal cognitive decline. Baseline amyloidosis contributed to a significant proportion of cognitive decline that accelerated with longer follow‐up times, and its primary impact was through cortical thinning.InterpretationWe developed an integrated framework to help explain the dynamic and complex process of cognitive aging by considering key causal pathways. Such an approach is important for both better comprehension of cognitive aging processes and will aid in the development of successful intervention strategies. ANN NEUROL 2019;86:866–877
Discovery and Validation of a Biomarker Model (PRESERVE) Predictive of Renal Outcomes After Liver Transplantation
Levitsky, Josh , Asrani, Sumeet K. , Klintmalm, Göran B. , Schiano, Thomas D. , Moss, Adyr A. , Chavin, Kenneth D. , Miller, Charles C. , Guo, Kexin , Zhao, Lihui , Jennings, Linda W. , Brown, Merideth, Armstrong, Brian, Asrani, Sumeet K , Chavin, Kenneth , Abecassis, Michael M.

Mayo Clinic

Hepatology | September 11, 2019 | Volume: 71 | Issue: 5 | Pages: 1775-1786

10.1002/hep.30939 | PMID 31509263 Get It

Background and Aims A high proportion of patients develop chronic kidney disease (CKD) after liver transplantation (LT). We aimed to develop clinical/protein models to predict future glomerular filtration rate (GFR) deterioration in this population. Approach and Results In independent multicenter discovery (CTOT14) and single‐center validation (BUMC) cohorts, we analyzed kidney injury proteins in serum/plasma samples at month 3 after LT in recipients with preserved GFR who demonstrated subsequent GFR deterioration versus preservation by year 1 and year 5 in the BUMC cohort. In CTOT14, we also examined correlations between serial protein levels and GFR over the first year. A month 3 predictive model was constructed from clinical and protein level variables using the CTOT14 cohort (n = 60). Levels of β‐2 microglobulin and CD40 antigen and presence of hepatitis C virus (HCV) infection predicted early (year 1) GFR deterioration (area under the curve [AUC], 0.814). We observed excellent validation of this model (AUC, 0.801) in the BUMC cohort (n = 50) who had both early and late (year 5) GFR deterioration. At an optimal threshold, the model had the following performance characteristics in CTOT14 and BUMC, respectively: accuracy (0.75, 0.8), sensitivity (0.71, 0.67), specificity (0.78, 0.88), positive predictive value (0.74, 0.75), and negative predictive value (0.76, 0.82). In the serial CTOT14 analysis, several proteins, including β‐2 microglobulin and CD40, correlated with GFR changes over the first year. Conclusions We have validated a clinical/protein model (PRESERVE) that early after LT can predict future renal deterioration versus preservation with high accuracy. This model may help select recipients at higher risk for subsequent CKD for early, proactive renal sparing strategies.