Comparison of Aptamer-Based and Antibody-Based Assays for Protein Quantification in Chronic Kidney Disease : Clinical Journal of the American Society of Nephrology

Journal Logo

Original Article: Chronic Kidney Disease

Comparison of Aptamer-Based and Antibody-Based Assays for Protein Quantification in Chronic Kidney Disease

Lopez-Silva, Carolina1; Surapaneni, Aditya2,3; Coresh, Josef1,2,3; Reiser, Jochen4; Parikh, Chirag R.1,2,3; Obeid, Wassim1; Grams, Morgan E.1,2,3; Chen, Teresa K.1,2

Author Information
CJASN 17(3):p 350-360, March 2022. | DOI: 10.2215/CJN.11700921
  • Free
  • Infographic
  • SDC



Proteomics, the systematic study of proteins in a biological system, has emerged as a promising tool to improve the diagnosis and management of CKD (1). Antibody-based technologies, such as ELISAs or multiplexed immunoassays, have long been used in clinical research to quantify biomarkers that may play a role in human pathophysiology (2). However, these traditional immunoassays can only analyze a small fraction of the human proteome, thereby limiting researchers’ abilities to discover new biomarkers of potential clinical relevance (3,4).

Novel aptamer-based technologies, such as SOMAscan (SomaLogic), provide greater coverage of the proteome. SOMAscan uses slow off-rate modified aptamers (SOMAmers), which are small, single-stranded deoxyoligonucleotides that bind with high affinity to target proteins in their native conformation (5). One major advantage of SOMAscan over immunoassays is its capability to analyze approximately 7000 analytes in a small-volume sample. Numerous studies have taken advantage of the increased throughput that SOMAscan offers to identify biomarkers in specific diseases (6–9). Nonetheless, the use of SOMAmer-based assays for the identification of proteins relevant in kidney disease remains in question, as the specificity of aptamers for distinct proteins has not been thoroughly studied in the context of CKD.

In this study, we sought to compare the performance of the SOMAscan platform with immunoassays in quantifying eight immune activation biomarkers (IFN-γ, IL-6, IL-8, IL-10, TNF-α, soluble TNF receptor superfamily member 1A [TNFRSF1A], soluble TNF receptor superfamily member 1B [TNFRSF1B], and soluble urokinase-type plasminogen activator receptor [suPAR]) and one biomarker of kidney function (cystatin C) in the African American Study of Kidney Disease and Hypertension (AASK). Increasing evidence suggests that many of these biomarkers are associated with CKD progression, kidney failure, and mortality and could serve as prognostic markers in the management of CKD (10–13). Method comparison was conducted via three approaches: (1) correlating serum concentrations of the proteins measured by SOMAscan versus immunoassays, (2) comparing correlations of SOMAscan and immunoassay biomarker measurements with measured GFR, and (3) comparing associations between proteins measured via SOMAscan and immunoassays with risks of incident kidney failure and all-cause mortality.

Materials and Methods

Study Population

AASK was a 3 × 2 factorial, double-blinded, randomized controlled trial of self-reported Black participants with CKD attributed to hypertension. Participants were randomized to one of three BP medications (metoprolol, ramipril, or amlodipine) and one of two BP goals (mean arterial pressure of 102–107 or <92 mm Hg). For inclusion in the trial, participants had to be 18–70 years old and have a diastolic BP >95 mm Hg and measured GFR of 20–65 ml/min per 1.73 m2. Individuals with diabetes, a urine protein-creatinine ratio >2.5 g/g, or CKD attributable to a cause other than hypertension were excluded (14–16). Participants were followed from trial enrollment (February 1995 to September 1998) until September 2001 (15,16), at which point those who had not developed kidney failure were invited to continue on with the cohort phase (April 2002 to June 2007). During this second phase, all participants received ramipril and had a BP goal <140/90 mm Hg (after 2004, <130/80 mm Hg) (15). All protocols were approved by institutional review boards at each participating site, and informed consent was obtain from participants (10,15,16). Among 1094 AASK trial participants, 705 had SOMAscan data, of whom 498 had available immunoassay data (Figure 1).

Figure 1.:
Flow chart of study population. Final study population included 498 African American Study of Kidney Disease and Hypertension (AASK) participants with available SOMAscan and immunoassay data for soluble TNF receptor superfamily member 1A (TNFRSF1A), soluble TNF receptor superfamily member 1B (TNFRSF1B), TNF-α, IFN-γ, IL-6, IL-8, and cystatin C; 495 AASK participants with available SOMAscan and immunoassay data for IL-10; and 483 AASK participants with available SOMAscan and immunoassay data for soluble urokinase-type plasminogen activator receptor (suPAR).

Biomarker Measurement

Proteins of interest were measured from stored samples collected at the AASK trial baseline visit. Cystatin C was measured from 2005 to 2006 using particle-enhanced nephelometric immunoassay (Dade/Siemens) (17). suPAR was measured in 2017 using Virogates’ suPARnostic enzyme-linked immunoassay absorbent assay kit (12). The remaining biomarkers were measured from December 2019 to January 2020 using the Mesoscale Discovery Platform (Meso Scale Diagnostics), which utilizes electrochemiluminescence and multiarray technology (10). Aptamer-based measurements were performed in January 2021 using the SOMAscan v.4.1 platform (SomaLogic). Immunoassay cystatin C was reported as milligrams per liter, and immunoassay biomarker data were reported as picograms per milliliter; SOMAscan data were reported as relative fluorescence units. For this study, we chose available biomarkers with interassay coefficients of variation on the immunoassay platform that were <10%. Interassay coefficients of variation were calculated as the SD over the mean of a population of blind duplicate samples (10,12,13,18).


The main outcomes of interest were measured GFR, incident kidney failure, and all-cause mortality. Measured GFR was assessed at baseline via125I-iothalamate clearance (15). Diagnosis of kidney failure was defined by initiation of dialysis or receipt of a kidney transplant, and both kidney failure and mortality were determined through active follow-up during the trial and cohort phases (15).

Statistical Analyses

Proteins were log base 2 transformed to achieve a more normal distribution. Correlations between immunoassay and SOMAscan measurements were evaluated using both Pearson (r) and Spearman (rs) correlation coefficients. Agreement between immunoassay and SOMAscan was assessed by Bland–Altman plots. We correlated protein levels measured by each method with measured GFR assessed at the same visit. Cox proportional hazards models were constructed to study the association of proteins, as measured by immunoassay versus SOMAscan, with incident kidney failure and all-cause mortality. Linearity was evaluated via cubic spline plots, with associations being fairly linear for all significant biomarkers. Covariates included age, sex, and randomized treatment groups. Seemingly unrelated regressions were used to compare the strength of the association between biomarkers measured via immunoassay versus SOMAscan with kidney failure and all-cause mortality (19–21). In order to determine whether one measure added information over and above the other measure, likelihood ratio tests were performed to compare models that included biomarkers measured via both immunoassay and SOMAscan versus each individual method. For example, a likelihood ratio test P value for “immunoassay” would indicate whether the additional use of immunoassay provides additional prognostic information after SOMAscan analysis and vice versa. Finally, to compare risk discrimination between protein measurements from each platform with kidney failure and mortality, the Harrell's C statistic was calculated for a clinical model that adjusted for age, sex, and randomized treatment groups; clinical model plus immunoassay-based biomarker measurement; and clinical model plus SOMAscan-based biomarker measurement. Differences in C statistic between clinical models that included immunoassay versus SOMAscan biomarker measurements were then determined. Analyses were conducted using Stata 15 (StataCorp, College Station, TX), and a P value of <0.05 was considered statistically significant.


Patient Characteristics

The mean age for participants was 54 years, 37% were women, and 50% had a history of heart disease. Mean systolic BP was 151 mm Hg, mean measured GFR was 45 ml/min per 1.73 m2, and median urine protein-creatinine ratio was 92 mg/g (interquartile range, 30–384) (Table 1).

Table 1. - Patient characteristics
Characteristic Overall
N 498
Age, yr 54 (11)
Women 186 (37%)
History of heart disease 250 (50%)
Current smoker 131 (26%)
Former smoker 150 (30%)
Measured GFR, ml/min per 1.73 m2 45 (13)
UPCR, mg/g 92 [30–384]
Systolic BP, mm Hg 151 (25)
Total cholesterol, mg/dl 213 (45)
HDL cholesterol, mg/dl 48 (16)
BMI, kg/m2 30.9 (6.5)
Summary of patient characteristics at the baseline visit of the African American Study of Kidney Disease and Hypertension. Data are presented as mean (SD), number (percentage), or median [interquartile range]. UPCR, urine protein-creatinine ratio; BMI, body mass index.

Immunoassay versus SOMAscan Biomarker Measurements

All immunoassay and SOMAscan measurements exhibited interassay coefficients of variation below 10% (Table 2). Four proteins (IL-8, TNFRSF1B, cystatin C, and TNFRSF1A) had strong correlations between immunoassay and SOMAscan measurements, with Pearson and Spearman correlation coefficients above 0.80. One IL-6 aptamer was moderately correlated (r=0.46, rs=0.50), whereas two proteins (another IL-6 aptamer and suPAR) had low correlations, with Pearson and Spearman correlation coefficients ranging from 0.22 to 0.33. All other proteins were poorly correlated across the two measurement methods (Figure 2).

Table 2. - Interassay coefficients of variation
Protein Immunoassay Interassay Coefficient of Variation, % SOMAscan Interassay Coefficient of Variation, %
Cystatin C 5.05 and 4.87 4.89
IFN-γ (#1) 6.17 2.15
IFN-γ (#2) 6.17 3.81
IL-6 (#1) 5.05 2.82
IL-6 (#2) 5.05 3.01
IL-8 5.09 2.52
IL-10 (#1) 6.72 2.06
IL-10 (#2) 6.72 3.34
TNF-α (#1) 7.52 2.41
TNF-α (#2) 7.52 1.92
TNFRSF1A 3.33 3.25
TNFRSF1B (#1) 2.96 2.07
TNFRSF1B (#2) 2.96 1.88
suPAR 3.90 2.24
Interassay coefficients of variation for immunoassays and SOMAscan. Coefficients of variation for cystatin C immunoassay were 5.05% and 4.87% at mean concentrations of 0.97 and 1.90 mg/L, respectively, from a pooled study of the African American Study of Kidney Disease and Hypertension and three other cohorts (18). Herein, we report two SOMAscan measurements for IFN-γ, IL-6, IL-10, TNF-α, and TNFRSF1B, which result from different aptamers binding to separate IFN-γ, IL-6, IL-10, TNF-α, and TNFRSF1B epitopes. TNFRSF1A, soluble TNF receptor superfamily member 1A; TNFRSF1B, soluble TNF receptor superfamily member 1B; suPAR, soluble urokinase-type plasminogen activator receptor.

Figure 2.:
Correlation of immunoassay and SOMAscan biomarker measurements. Scatterplots of immunoassay measurements (milligrams per liter for cystatin C and picograms per milliliter for other biomarkers) versus SOMAscan measurements (relative fluorescence units) for (D) cystatin C and (A–C and E–N) eight immune activation biomarkers. Two SOMAscan measurements are reported for (B and C) TNFRSF1B, (F and G) IL-6, (I and J) IL-10, (K and L) IFN-γ, and (M and N) TNF-α due to different aptamers binding to separate protein epitopes.

Measurements of IL-8, TNFRSF1B, cystatin C, and TNFRSF1A exhibited high agreement between methods across biomarker level. There was poorer agreement between both assays for IL-6, suPAR, IFN-γ, IL-10, and TNF-α (Figure 3).

Figure 3.:
Bland–Altman plots. Difference between normalized immunoassay measurements and SOMAscan measurements versus average of normalized immunoassay and SOMAscan measurements. (A–E) Agreement between both methods for IL-8, TNFRSF1B, cystatin C, and TNFRSF1A quantification remains consistent throughout different protein concentration levels. (F–N) There was poorer agreement between methods for IL-6, suPAR, IFN-γ, IL-10, and TNF-α. Two SOMAscan measurements are reported for TNFRSF1B, IL-6, IL-10, IFN-γ, and TNF-α due to aptamer binding of separate protein epitopes.

Correlation of Immunoassay versus SOMAscan Measurements with Measured Glomerular Filtration Rate

Correlations between protein and measured GFR are shown in Table 3 and Supplemental Table 1. Immunoassay measurement of cystatin C, a commonly used diagnostic marker in kidney disease (17), had a strongly negative correlation with measured GFR (r=−0.79). The correlation between SOMAmer-based measurement of cystatin C and measured GFR was also strongly negative (r=−0.76).

Table 3. - Analysis results for proteins with non-negligible correlations
Biomarker Correlations between Assays Correlations with Measured GFR, r Associations with Kidney Failure Associations with Mortality
r rs Hazard Ratio (95% Confidence Interval) P Value Likelihood Ratio Test P Value Hazard Ratio (95% Confidence Interval) P Value Likelihood Ratio Test P Value
 Immunoassay 0.94 0.95 0.06 0.89 (0.75 to 1.06) 0.18 0.83 1.41 (1.20 to 1.65) <0.001 0.005
 SOMAscan 0.09 0.87 (0.72 to 1.05) 0.15 0.51 1.31 (1.13 to 1.53) 0.001 0.08
SUR 0.50 0.004
 Immunoassay 0.93 0.93 −0.66 2.77 (2.34 to 3.27) <0.001 0.003 1.58 (1.31 to 1.90) <0.001 0.26
 SOMAscan −0.71 2.59 (2.21 to 3.04) <0.001 0.12 1.54 (1.28 to 1.84) <0.001 0.52
SUR 0.09 0.56
 Immunoassay 0.92 0.92 −0.66 2.77 (2.34 to 3.27) <0.001 <0.001 1.58 (1.31 to 1.90) <0.001 0.11
 SOMAscan −0.70 2.37 (2.05 to 2.73) <0.001 0.76 1.50 (1.26 to 1.79) <0.001 0.74
SUR <0.001 0.28
Cystatin C
 Immunoassay 0.89 0.89 −0.79 3.08 (2.57 to 3.69) <0.001 <0.001 1.52 (1.26 to 1.84) <0.001 0.94
 SOMAscan −0.76 2.64 (2.22 to 3.14) <0.001 0.84 1.63 (1.35 to 1.98) <0.001 0.01
SUR 0.01 0.19
 Immunoassay 0.85 0.85 −0.76 3.43 (2.87 to 4.10) <0.001 <0.001 1.51 (1.25 to 1.82) <0.001 0.002
 SOMAscan −0.72 2.81 (2.38 to 3.32) <0.001 0.08 1.34 (1.12 to 1.60) 0.001 0.32
SUR 0.03 0.02
IL-6 (#1)
 Immunoassay 0.46 0.50 0.03 0.97 (0.83 to 1.13) 0.69 0.54 1.23 (1.05 to 1.46) 0.01 0.03
 SOMAscan −0.03 1.03 (0.88 to 1.20) 0.71 0.56 1.09 (0.94 to 1.27) 0.25 0.91
SUR 0.46 0.14
IL-6 (#2)
 Immunoassay 0.22 0.33 0.03 0.97 (0.83 to 1.13) 0.69 0.70 1.23 (1.05 to 1.46) 0.01 0.02
 SOMAscan −0.08 0.99 (0.87 to 1.13) 0.91 0.98 1.02 (0.85 to 1.22) 0.85 0.69
SUR 0.79 0.08
 Immunoassay 0.23 0.29 −0.34 1.80 (1.52 to 2.13) <0.001 <0.001 1.55 (1.27 to 1.89) <0.001 <0.001
 SOMAscan −0.17 1.19 (1.00 to 1.42) 0.05 0.57 1.39 (1.09 to 1.77) 0.007 0.04
SUR 0.001 0.43
Pearson (r) and Spearman (rs) correlation coefficients of protein levels were measured via immunoassays versus SOMAscan. Pearson correlation coefficients of protein measurements are with measured GFR. Hazard ratios for protein measurements (per two-fold higher baseline level) with incident kidney failure/mortality were adjusted for age, sex, and randomized treatment groups. SUR compares strength of association between protein measurements and kidney failure/mortality. The likelihood ratio test compares models using both assays versus a single assay. SUR, seemingly unrelated regression; TNFRSF1B, soluble TNF receptor superfamily member 1B; TNFRSF1A, soluble TNF receptor superfamily member 1A; suPAR, soluble urokinase-type plasminogen activator receptor.

Similar trends were noted for TNFRSF1A, with both assays showing strong negative correlations with measured GFR (r=−0.76 for immunoassay versus −0.72 for SOMAscan). Both aptamer-based measures of TNFRSF1B were more negatively correlated with measured GFR compared with immunoassay (r=−0.70 and −0.71 versus −0.66, respectively). Correlations of suPAR with measured GFR were slightly weaker, more so when measured by SOMAscan (r=−0.17) than by immunoassay (r=−0.34). TNF-α was negatively correlated with measured GFR when measured via immunoassay (r=−0.30), and one aptamer of IFN-γ was modestly positively correlated with measured GFR when measured via SOMAscan (r=0.20). Otherwise, there was minimal correlation between IL-8, IL-6, IFN-γ, or IL-10 levels and measured GFR with either assay.

Associations of Immunoassay versus SOMAscan Biomarker Measurements with Kidney Failure and All-Cause Mortality

Over a median follow-up of 8.5 years (interquartile range, 4.5–10.3), there were 160 (32%) kidney failure events. In demographic- and randomized treatment group–adjusted analyses, higher levels of cystatin C were significantly associated with a higher risk of kidney failure, with a stronger association when using the immunoassay (hazard ratio [HR], 3.08; 95% confidence interval [95% CI], 2.57 to 3.69 for immunoassay and HR, 2.64; 95% CI, 2.22 to 3.14 for SOMAscan; P value for comparison =0.01). Both TNFRSF1A and TNFRSF1B were associated with a higher risk of kidney failure, with immunoassay values exhibiting stronger associations than SOMAscan for TNFRSF1A and one measure of TNFRSF1B. suPAR, IL-10, and TNF-α were only associated with a higher risk of kidney failure when measured via immunoassay but not SOMAscan. One aptamer of IFN-γ was associated with a lower risk of kidney failure when measured via SOMAscan (HR, 0.77; 95% CI, 0.65 to 0.91). IL-8 and IL-6 were not associated with a higher risk of developing kidney failure via either assay (Table 3, Supplemental Figure 1, Supplemental Table 1).

Over a median follow-up of 9.6 years (interquartile range, 6.8–11.0), there were 112 (22%) deaths. Both immunoassay and SOMAmer-based measurements of IL-8, TNFRSF1B, cystatin C, TNFRSF1A, and suPAR were associated with a higher risk of all-cause mortality. For IL-8 and TNFRSF1A, associations with all-cause mortality were stronger when measured by immunoassay compared with SOMAscan. Lastly, IL-6 and TNF-α were only associated with a higher risk of mortality when measured via immunoassay but not SOMAscan.

Likelihood ratio tests were used to compare models that included both immunoassay and SOMAscan-measured biomarkers versus each individual method. On average, SOMAscan-measured biomarkers provided no additional prognostic information for kidney failure or mortality over immunoassays. Exceptions were cystatin C and suPAR, whose SOMAscan-measured values improved upon the models that only used the proteins’ immunoassay-measured values to estimate mortality risk, and one aptamer of IFN-γ, whose SOMAscan-measured values improved models to estimate kidney failure risk. Conversely, immunoassay-measured TNFRSF1B, cystatin C, TNFRSF1A, suPAR, IL-10, and TNF-α provided additional prognostic information for kidney failure above their SOMAscan-measured counterparts, as did IL-8, TNFRSF1A, IL-6, suPAR, and TNF-α for mortality.

Evaluation of Prediction Model Discrimination by Immunoassay versus SOMAscan Biomarker Measurements

The C statistic for a clinical model containing age, sex, and randomized treatment groups was 0.65 (95% CI, 0.61 to 0.70) for kidney failure and 0.62 (95% CI, 0.57 to 0.67) for mortality (Supplemental Table 2). The addition of TNFRSF1B, cystatin C, and TNFRSF1A, measured by both immunoassay and SOMAscan, to the clinical model improved risk prediction for kidney failure, with immunoassay measurements of cystatin C and TNFRSF1A having greater improvements in risk prediction than SOMAscan measurements. The addition of immunoassay- but not SOMAscan-based measurements of TNF-α and suPAR to the clinical model also improved risk prediction for kidney failure. The addition of IL-8, TNFRSF1B, cystatin C, and TNFRSF1A by both measurement methods improved model discrimination for mortality; clinical models that included measurements of IL-8 and TNF-α by immunoassay performed better than SOMAscan.


In this study of Black adults with CKD attributed to hypertension, we compared the quantification of proteins as measured by SOMAscan, an aptamer-based assay, versus traditional immunoassays in the context of CKD. We found strong correlations between aptamer-based and immunoassay-based measurements of IL-8, cystatin C, TNFRSF1B, and TNFRSF1A but less so for other proteins. More specifically, we report inconsistencies between the two methods when quantifying IL-6, suPAR, IL-10, IFN-γ, and TNF-α, thus raising caution regarding the use of SOMAscan for these particular proteins. In general, immunoassay-based biomarker measurements had stronger associations with adverse outcomes than SOMAscan-based measurements, suggesting that after initial screening with SOMAscan, there may be benefit in remeasuring candidate proteins with immunoassays to confirm associations.

Overall, our findings concur with previous studies comparing the performance of SOMAmer-based assays with traditional antibody-based techniques. SOMAscan quantification of serum cystatin C, IL-8, TNFRSF1A, and TNFRSF1B has been shown to have excellent to moderate correlations with immunoassays, whereas IFN-γ has exhibited poor correlations, and IL-6 and TNF-α have had mixed findings (22–25). A number of other proteins not included in this study (e.g., C-reactive protein, E-selectin, neutrophil gelatinase–associated lipocalin, TNF-β) have also been shown to have high and moderate correlations when quantified via SOMAscan versus traditional clinical assays (22–25). Other studies have demonstrated significant discrepancies between aptamer-based and antibody-based assays in measuring various clinically relevant biomarkers (24). For example, the quantification and predictive power of suPAR for kidney outcomes have been shown to vary significantly among aptamer-based assays, multiplexed antibody assays, and traditional ELISAs (26,27) (S.S. Hayek, et al., unpublished article). We similarly showed that five of nine biomarkers examined (IL-6, suPAR, IL-10, IFN-γ, and TNF-α) had low to negligible correlations between methods. Our study adds to the current literature by validating aptamer-based quantification of specific biomarkers in the setting of CKD, a condition known to greatly affect the proteome, while also showing that not all proteins were accurately measured with SOMAscan (28).

Differential performance of SOMAscan versus multiplex assays across proteins may be explained by a combination of technical and biologic factors. Protein values below the assays’ detection limits or differences in synthetic peptides used for aptamer and antibody reagent selection may contribute to observed differences between assays (M. Pietzner et al., unpublished article). In addition, part of the SOMAscan normalization pipeline includes scaling protein dilution bins to match their medians to those measured in an external healthy reference population (29). In individuals and populations with large deviations from a healthy plasma proteome, such as patients with advanced CKD or kidney failure, this normalization could potentially attenuate the strength of associations by reducing more extreme values if they are associated with a clinical outcome. Perhaps this could be contributing to the attenuated prognostic associations observed with SOMAscan. On the other hand, genetic polymorphisms between individuals as well as the presence of post-translational modifications and transmembrane domains could all lead to structural differences that further affect binding affinities between aptamers and target protein epitopes (M. Pietzner et al., unpublished article).

Our data support prior research reporting associations of cystatin C and immune activation biomarkers with clinical outcomes in CKD. Cystatin C is commonly used to estimate GFR, with higher serum concentrations associated with a higher risk of kidney failure and mortality among patients with CKD (17,30). Similarly, TNFRSF1A, TNFRSF1B, IL-6, IL-8, and suPAR have been independently associated with a higher risk of kidney failure and/or mortality in CKD (10,12,13,31). Our findings support these known associations while providing novel information on the utility and limitations of using SOMAscan to quantify these prognostic biomarkers. In comparing models that included both methods of measurement versus either method alone, immunoassay-based measurements generally added additional information to SOMAmer-based measurements but not vice versa, particularly for the outcome of kidney failure. Compared with SOMAscan, the addition of immunoassay-based biomarker measurements to a clinical model also had greater improvements in risk prediction for select biomarkers. These findings suggest that targeted immunoassays play an important role in the evaluation of candidate proteins after screening by SOMAscan.

Our study has some notable strengths. First, the utilization of both the trial and cohort phases of AASK allowed for a long follow-up period with data that were prospectively collected. Second, we used direct measurements of GFR via125I-iothalamate clearance. Third, we validated SOMAscan using multiple approaches, including direct comparisons with immunoassay measurements and assessment of associations with measured GFR, kidney failure, and all-cause mortality.

We also acknowledge some limitations. The AASK trial enrolled exclusively self-identified Black adults with CKD attributable to hypertension. This may limit the generalizability of our findings to other ethnic groups or to patients with CKD attributable to other causes. The lack of a second replication cohort did not allow us to verify our findings or analyze how the performance of aptamer-based assays may vary across different patient populations. Moreover, the samples used in our analysis had been in storage since 1995 and 1998, with biomarker measurements taking place 19–26 years later. Differences in the long-term storage conditions and the number of freeze-thaw cycles could have contributed to varying extents of biomarker degradation, perhaps accounting for observed differences in assay performance. However, previous studies have demonstrated that although long-term storage and repeated freeze-thaw cycles can affect the stability of IL-6, IL-8, IL-10, IFN-γ, and TNF-α, this effect can be attenuated by storing samples at −80°C (32–34). Serum samples of TNFRSF1A, TNFRSF1B, and suPAR have been shown to be resistant to significant degradation after long-term storage and undergoing at least one freeze-thaw cycle (32,35,36). Similarly, levels of serum cystatin C stored for multiple years at −70°C have been shown to remain stable after one freeze-thaw cycle (37). Finally, the majority of our immunoassay measurements were performed using a multiplex protein assay and not ELISA, which is considered the gold standard of antibody-based techniques (38). However, both particle-enhanced nephelometric immunoassays and multiplex protein assays have been thoroughly validated against ELISA, and the Mesoscale Discovery Platform used in our study has been shown to have comparable or better performance than other commercially available multiplex assays (38–40).

High-throughput SOMAmer-based assays have gained significant attention in recent years as they have the potential to greatly increase our understanding of the human proteome. In just the past decade, SOMAscan has been used to identify hundreds of biomarkers in a wide range of conditions (6–9). Moreover, the rapid pace at which SOMAmer-based assays are being incorporated into clinical research calls for more studies to evaluate their performance as a biochemical assay and clinical instrument. Our study is one of the first to examine the use of novel aptamer-based technologies in quantifying immune activation biomarkers and evaluate their performance as prognostic markers in CKD. We showed that although SOMAscan is a reliable technique for the measurement of specific biomarkers in the setting of kidney disease, not all proteins were accurately quantified with aptamer-based assays. Additional research is needed to further validate SOMAmer-based technology as a research and clinical tool in a variety of disease contexts.


T.K. Chen reports research funding from the National Institutes of Health (NIH)/the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) and Yale University. J. Coresh reports consultancy agreements with; ownership interest in; research funding from the National Kidney Foundation (NKF; which receives industry support) and NIH; an approximately $3000 honorarium from Abbott on May 23, 2012 for a presentation on the topic of the interaction of eGFR and renal cardiovascular disease; and serving as a scientific advisor or member of and NKF. M.E. Grams reports honoraria from academic institutions for giving grand rounds and American Diabetes Association for reviewing abstracts; serving in an advisory or leadership role for American Journal of Kidney Diseases, CJASN, the JASN Editorial Fellowship Committee, the Kidney Disease Improving Global Outcomes Executive Committee, the NKF Scientific Advisory Board, and the United States Renal Data System Scientific Advisory Board; and other interests or relationships with NKF, which in turn receives funding from Abbvie, Relypsa, and Thrasos, among others. C. Lopez-Silva received Dean’s Summer Research Funding from Johns Hopkins University School of Medicine while completing this manuscript. C.R. Parikh reports consultancy agreements with Genfit Biopharmaceutical Company and Novartis; is a member of the advisory board of and owns equity in RenalytixAI; reports research funding from the National Heart, Lung and Blood Institute and NIDDK; and reports serving in an advisory or leadership role for Genfit Biopharmaceutical Company. J. Reiser reports consultancy agreements with Mantra Bio, Visterra, and Walden Biosciences; ownership interest in Walden Biosciences; research funding from Walden Biosciences; honoraria from Visterra; serving in an advisory or leadership role for Walden Biosciences (cochair of the scientific advisory board); and other interests or relationships with Nephcure Kidney International. J. Reiser is an inventor on issued and pending patents pertinent to novel methods and treatments for proteinuric kidney diseases and stands to gain royalties from future commercialization. J. Reiser is also a scientific cofounder and shareholder of Walden Biosciences (formerly TRISAQ), a biotechnology company. Parts of J. Reiser's intellectual property have been outlicensed to Miltenyi Biotech. All remaining authors have nothing to disclose.


T.K. Chen is supported by a George M. O’Brien Center for Kidney Research Pilot and Feasibility Grant from Yale University (National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases grant P30DK079310) and National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases grant K08DK117068. M.E. Grams is supported by National Institutes of Health, National Heart, Lung and Blood Institute grant K24HL155861 and National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases grant R01DK108803. C.R. Parikh is supported by National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases grant U01DK106962. Data and samples for TNFRSF1A, TNFRSF1B, TNF-α, IFN-γ, IL-6, IL-8, and IL-10 immunoassay measurements were supplied by the National Institute of Diabetes and Digestive and Kidney Diseases Central Repositories viaNational Institute of Diabetes and Digestive and Kidney Diseases grant X01DK118497 and measured by the Translational Research Core of the George M. O’Brien Kidney Center.

Published online ahead of print. Publication date available at


The authors thank the staff and participants of AASK.

Portions of this work were presented at the 2021 American Society of Nephrology Kidney Week (November 2021).

AASK was conducted by the AASK investigators and supported by the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK). The data and samples from the AASK trial reported here were supplied by the NIDDK Central Repositories. The AASK trial and cohort were supported by National Institutes of Health (NIH), NIDDK institutional grants M01 RR-00080, M01 RR-00071, M0100032, P20-RR11145, M01 RR00827, M01 RR00052, 2P20 RR11104, RR029887, DK 2818-02, DK057867, and DK048689 and the following pharmaceutical companies: AstraZeneca, Forest Laboratories, GlaxoSmithKline, King Pharmaceuticals, Pfizer, Pharmacia, and Upjohn.

This manuscript was not prepared in collaboration with investigators of AASK and does not necessarily reflect the opinions or views of AASK, the NIDDK Central Repositories, NIH, or NIDDK.

Author Contributions

T.K. Chen, M.E. Grams, C. Lopez-Silva, and A. Surapaneni conceptualized the study; W. Obeid and C.R. Parikh were responsible for investigation; A. Surapaneni was responsible for formal analysis; T.K. Chen, J. Coresh, M.E. Grams, C. Lopez-Silva, and A. Surapaneni were responsible for methodology; T.K. Chen and C. Lopez-Silva were responsible for project administration; T.K. Chen, J. Coresh, M.E. Grams, and J. Reiser were responsible for resources; C. Lopez-Silva and A. Surapaneni were responsible for visualization; T.K. Chen and M.E. Grams were responsible for funding acquisition; T.K. Chen and M.E. Grams provided supervision; C. Lopez-Silva wrote the original draft; and T.K. Chen, J. Coresh, M.E. Grams, C. Lopez-Silva, W. Obeid, C.R. Parikh, J. Reiser, and A. Surapaneni reviewed and edited the manuscript.

Data Sharing Statement

Preexisting data access policies for the parent cohort study specify that research data requests can be submitted to the steering committee; these will be promptly reviewed for confidentiality or intellectual property restrictions and will not unreasonably be refused. Please refer to the data sharing policies of this study. Individual-level patient or protein data may further be restricted by consent, confidentiality, or privacy. These policies apply to both clinical and proteomics data.

Supplemental Material

This article contains the following supplemental material online at

Supplemental Figure 1. Forest plots of biomarkers and risk of kidney failure and mortality.

Supplemental Table 1. Analysis results for uncorrelated proteins.

Supplemental Table 2. Harrell's C statistics for the clinical model with or without biomarkers for prediction of kidney failure and mortality.


1. Dubin RF, Rhee EP: Proteomics and metabolomics in kidney disease, including insights into etiology, treatment, and prevention. Clin J Am Soc Nephrol 15: 404–411, 2020
2. Ellington AA, Kullo IJ, Bailey KR, Klee GG: Antibody-based protein multiplex platforms: Technical and operational challenges. Clin Chem 56: 186–193, 2010
3. Juncker D, Bergeron S, Laforte V, Li H: Cross-reactivity in antibody microarrays and multiplexed sandwich assays: Shedding light on the dark side of multiplexing. Curr Opin Chem Biol 18: 29–37, 2014
4. Pertea M, Salzberg SL: Between a chicken and a grape: Estimating the number of human genes. Genome Biol 11: 206, 2010
5. Kraemer S, Vaught JD, Bock C, Gold L, Katilius E, Keeney TR, Kim N, Saccomano NA, Wilcox SK, Zichi D, Sanders GM: From SOMAmer-based biomarker discovery to diagnostic and clinical applications: A SOMAmer-based, streamlined multiplex proteomic assay. PLoS One 6: e26332, 2011
6. Ostroff RM, Mehan MR, Stewart A, Ayers D, Brody EN, Williams SA, Levin S, Black B, Harbut M, Carbone M, Goparaju C, Pass HI: Early detection of malignant pleural mesothelioma in asbestos-exposed individuals with a noninvasive proteomics-based surveillance tool. PLoS One 7: e46091, 2012
7. Petek LM, Rickard AM, Budech C, Poliachik SL, Shaw D, Ferguson MR, Tawil R, Friedman SD, Miller DG: A cross sectional study of two independent cohorts identifies serum biomarkers for facioscapulohumeral muscular dystrophy (FSHD). Neuromuscul Disord 26: 405–413, 2016
8. Sattlecker M, Kiddle SJ, Newhouse S, Proitsi P, Nelson S, Williams S, Johnston C, Killick R, Simmons A, Westman E, Hodges A, Soininen H, Kłoszewska I, Mecocci P, Tsolaki M, Vellas B, Lovestone S, Dobson RJ; AddNeuroMed Consortium: Alzheimer’s disease biomarker discovery using SOMAscan multiplexed protein technology. Alzheimers Dement 10: 724–734, 2014
9. Webber J, Stone TC, Katilius E, Smith BC, Gordon B, Mason MD, Tabi Z, Brewis IA, Clayton A: Proteomics analysis of cancer exosomes using a novel modified aptamer-based array (SOMAscan™) platform. Mol Cell Proteomics 13: 1050–1064, 2014
10. Chen TK, Estrella MM, Appel LJ, Coresh J, Luo S, Reiser J, Obeid W, Parikh CR, Grams ME: Biomarkers of immune activation and incident kidney failure with replacement therapy: Findings from the African American Study of Kidney Disease and Hypertension. Am J Kidney Dis 78: 75–84.e1, 2021
11. Ishigami J, Taliercio J, I Feldman H, Srivastava A, Townsend R, L Cohen D, Horwitz E, Rao P, Charleston J, Fink JC, Ricardo AC, Sondheimer J, Chen TK, Wolf M, Isakova T, Appel LJ, Matsushita K; CRIC Study Investigators: Inflammatory markers and incidence of hospitalization with infection in chronic kidney disease. Am J Epidemiol 189: 433–444, 2020
12. Luo S, Coresh J, Tin A, Rebholz CM, Chen TK, Hayek SS, Tracy M, Lipkowitz MS, Appel LJ, Levey AS, Inker LA, Reiser J, Grams ME: Soluble urokinase-type plasminogen activator receptor in Black Americans with CKD. Clin J Am Soc Nephrol 13: 1013–1021, 2018
13. Chen TK, Estrella MM, Appel LJ, Coresh J, Luo S, Obeid W, Parikh CR, Grams ME: Serum levels of IL-6, IL-8 and IL-10 and risks of end-stage kidney disease and mortality. Nephrol Dial Transplant 36: 561–563, 2021
14. Wright Jr. JT, Kusek JW, Toto RD, Lee JY, Agodoa LY, Kirk KA, Randall OS, Glassock R: Design and baseline characteristics of participants in the African American Study of Kidney Disease and Hypertension (AASK) pilot study. Control Clin Trials 17 [Suppl]: 3S–16S, 1996
15. Appel LJ, Wright Jr. JT, Greene T, Agodoa LY, Astor BC, Bakris GL, Cleveland WH, Charleston J, Contreras G, Faulkner ML, Gabbai FB, Gassman JJ, Hebert LA, Jamerson KA, Kopple JD, Kusek JW, Lash JP, Lea JP, Lewis JB, Lipkowitz MS, Massry SG, Miller ER, Norris K, Phillips RA, Pogue VA, Randall OS, Rostand SG, Smogorzewski MJ, Toto RD, Wang X; AASK Collaborative Research Group: Intensive blood-pressure control in hypertensive chronic kidney disease. N Engl J Med 363: 918–929, 2010
16. Wright Jr. JT, Bakris G, Greene T, Agodoa LY, Appel LJ, Charleston J, Cheek D, Douglas-Baltimore JG, Gassman J, Glassock R, Hebert L, Jamerson K, Lewis J, Phillips RA, Toto RD, Middleton JP, Rostand SG; African American Study of Kidney Disease and Hypertension Study Group: Effect of blood pressure lowering and antihypertensive drug class on progression of hypertensive kidney disease: Results from the AASK trial. JAMA 288: 2421–2431, 2002
17. Bhavsar NA, Appel LJ, Kusek JW, Contreras G, Bakris G, Coresh J, Astor BC; AASK Study Group: Comparison of measured GFR, serum creatinine, cystatin C, and beta-trace protein to predict ESRD in African Americans with hypertensive CKD. Am J Kidney Dis 58: 886–893, 2011
18. Stevens LA, Coresh J, Schmid CH, Feldman HI, Froissart M, Kusek J, Rossert J, Van Lente F, Bruce 3rd RD, Zhang YL, Greene T, Levey AS: Estimating GFR using serum cystatin C alone and in combination with serum creatinine: A pooled analysis of 3,418 individuals with CKD. Am J Kidney Dis 51: 395–406, 2008
19. Zellner A: An efficient method of estimating seemingly unrelated regressions and tests for aggregation bias. J Am Stat Assoc 57: 348–368, 1962
20. Zellner A, Huang DS: Further properties of efficient estimators for seemingly unrelated regression equations. Int Econ Rev 3: 300–313, 1962
21. Zellner A: Estimators for seemingly unrelated regression equations: Some exact finite sample results. J Am Stat Assoc 58: 977–992, 1963
22. Kukova LZ, Mansour SG, Coca SG, de Fontnouvelle CA, Thiessen-Philbrook HR, Shlipak MG, El-Khoury JM, Parikh CR: Comparison of urine and plasma biomarker concentrations measured by aptamer-based versus immunoassay methods in cardiac surgery patients. J Appl Lab Med 4: 331–342, 2019
23. Lim SY, Lee JH, Welsh SJ, Ahn SB, Breen E, Khan A, Carlino MS, Menzies AM, Kefford RF, Scolyer RA, Long GV, Rizos H: Evaluation of two high-throughput proteomic technologies for plasma biomarker discovery in immunotherapy-treated melanoma patients. Biomark Res 5: 32, 2017
24. Raffield LM, Dang H, Pratte KA, Jacobson S, Gillenwater LA, Ampleford E, Barjaktarevic I, Basta P, Clish CB, Comellas AP, Cornell E, Curtis JL, Doerschuk C, Durda P, Emson C, Freeman CM, Guo X, Hastie AT, Hawkins GA, Herrera J, Johnson WC, Labaki WW, Liu Y, Masters B, Miller M, Ortega VE, Papanicolaou G, Peters S, Taylor KD, Rich SS, Rotter JI, Auer P, Reiner AP, Tracy RP, Ngo D, Gerszten RE, O’Neal WK, Bowler RP; NHLBI Trans-Omics for Precision Medicine (TOPMed) Consortium: Comparison of proteomic assessment methods in multiple cohort studies. Proteomics 20: e1900278, 2020
25. Tin A, Yu B, Ma J, Masushita K, Daya N, Hoogeveen RC, Ballantyne CM, Couper D, Rebholz CM, Grams ME, Alonso A, Mosley T, Heiss G, Ganz P, Selvin E, Boerwinkle E, Coresh J: Reproducibility and variability of protein analytes measured using a multiplexed modified aptamer assay. J Appl Lab Med 4: 30–39, 2019
26. Abraham AG, Xu Y, Roem JL, Greenberg JH, Weidemann DK, Sabbisetti VS, Bonventre JV, Denburg M, Warady BA, Furth SL: Variability in CKD biomarker studies: Soluble urokinase plasminogen activator receptor (suPAR) and kidney disease progression in the Chronic Kidney Disease in Children (CKiD) study. Kidney Med 3: 712–721.e1, 2021
27. Winnicki W, Sunder-Plassmann G, Sengölge G, Handisurya A, Herkner H, Kornauth C, Bielesz B, Wagner L, Kikić Ž, Pajenda S, Reiter T, Schairer B, Schmidt A: Diagnostic and prognostic value of soluble urokinase-type plasminogen activator receptor (suPAR) in focal segmental glomerulosclerosis and impact of detection method. Sci Rep 9: 13783, 2019
28. Christensson A, Ash JA, DeLisle RK, Gaspar FW, Ostroff R, Grubb A, Lindström V, Bruun L, Williams SA: The impact of the glomerular filtration rate on the human plasma proteome. Proteomics Clin Appl 12: e1700067, 2018
29. Williams SA, Kivimaki M, Langenberg C, Hingorani AD, Casas JP, Bouchard C, Jonasson C, Sarzynski MA, Shipley MJ, Alexander L, Ash J, Bauer T, Chadwick J, Datta G, DeLisle RK, Hagar Y, Hinterberg M, Ostroff R, Weiss S, Ganz P, Wareham NJ: Plasma protein patterns as comprehensive indicators of health. Nat Med 25: 1851–1857, 2019
30. Shlipak MG, Matsushita K, Ärnlöv J, Inker LA, Katz R, Polkinghorne KR, Rothenbacher D, Sarnak MJ, Astor BC, Coresh J, Levey AS, Gansevoort RT; CKD Prognosis Consortium: Cystatin C versus creatinine in determining risk based on kidney function. N Engl J Med 369: 932–943, 2013
31. Barreto DV, Barreto FC, Liabeuf S, Temmar M, Lemke H-D, Tribouilloy C, Choukroun G, Vanholder R, Massy ZA; European Uremic Toxin Work Group (EUTox): Plasma interleukin-6 is independently associated with mortality in both hemodialysis and pre-dialysis patients with chronic kidney disease. Kidney Int 77: 550–556, 2010
32. Hardikar S, Song X, Kratz M, Anderson GL, Blount PL, Reid BJ, Vaughan TL, White E: Intraindividual variability over time in plasma biomarkers of inflammation and effects of long-term storage. Cancer Causes Control 25: 969–976, 2014
33. Gong Y, Liang S, Zeng L, Ni Y, Zhou S, Yuan X: Effects of blood sample handling procedures on measurable interleukin 6 in plasma and serum. J Clin Lab Anal 33: e22924, 2019
34. Kordulewska NK, Topa J, Tańska M, Cieślińska A, Fiedorowicz E, Jarmołowska B: Stability of interleukin-1β, -4, -6, -8, -10, -13, interferon-γ and tumor necrosis factor-α in human sera after repetitive freeze-thaw cycles and long storage. J Pharm Biomed Anal 196: 113900, 2021
35. Kofoed K, Schneider UV, Scheel T, Andersen O, Eugen-Olsen J: Development and validation of a multiplex add-on assay for sepsis biomarkers using xMAP technology. Clin Chem 52: 1284–1293, 2006
36. Odden N, Henriksen T, Mørkrid L: Serum soluble urokinase plasminogen activator receptor (suPAR) in early pregnancy prior to clinical onset of preeclampsia. Acta Obstet Gynecol Scand 91: 1226–1232, 2012
37. Ishigami J, Karger AB, Lutsey PL, Coresh J, Matsushita K: Stability of serum bone-mineral, kidney, and cardiac biomarkers after a freeze-thaw cycle: The ARIC study [published online ahead of print October 13, 2021]. Am J Epidemiol 10.1093/aje/kwab251
38. Elshal MF, McCoy JP: Multiplex bead array assays: Performance evaluation and comparison of sensitivity to ELISA. Methods 38: 317–323, 2006
39. Hossain MA, Emara M, El Moselhi H, Shoker A: Comparing measures of cystatin C in human sera by three methods. Am J Nephrol 29: 381–391, 2009
40. Dabitao D, Margolick JB, Lopez J, Bream JH: Multiplex measurement of proinflammatory cytokines in human serum: Comparison of the Meso Scale Discovery electrochemiluminescence assay and the Cytometric Bead Array. J Immunol Methods 372: 71–77, 2011

AASK (African American Study of Kidney Disease and Hypertension); chronic kidney disease; end-stage renal disease; mortality; chronic inflammation; antibodies; biological assay

Copyright © 2022 by the American Society of Nephrology