After publication of this article [1], it came to light that there were errors in the reported glomerular filtration rate (GFR) estimates.
The two-fold purpose of this paper [1] was to 1) compare accuracy and bias of widely used glomerular filtration rate (GFR) estimating equations to a gold-standard GFR measure (iohexol disappearance from plasma) in HIV-positive and HIV-negative volunteers, and 2) to assess factors associated with bias and accuracy of the creatinine-based and cystatin C-based equations. Recently, our co-investigators, who performed the laboratory analyses and calculations for the iohexol GFR, identified a drift that occurred in their measurement of iohexol (prior to this study) that led to an across-the-board underestimation of iohexol concentrations from blood samples, which produced a systematic overestimation of GFR by approximately 10%. This measurement error in this laboratory was described in a publication in 2017[2]. We subsequently repeated the analyses in the PLOS ONE paper using recalibrated (corrected) iohexol GFR values provided here in an updated version of Table 1.
Table 1. Clinical characteristics of HIV-positive and HIV-negative participants.
| Clinical characteristics | HIV-positive (n = 187) | HIV-negative (n = 98) | P value | |
|---|---|---|---|---|
| Age, years, median (P25, P75) | 49 (45, 53) | 49 (45, 54) | 0.58 | |
| Body mass index, kg/m2, median (P25, P75) | 26 (23, 31) | 27 (23, 33) | 0.21 | |
| Sex | Female, n (%) | 66 (35) | 18 (18) | 0.0027 |
| Male, n (%) | 121 (65) | 80 (82) | ||
| Race | White, n (%) | 11 (6) | 8 (8) | 0.46 |
| Black, n (%) | 176 (94) | 90 (92) | ||
| Current smoker, n (%) | 124 (66) | 60 (61) | 0.44 | |
| History of hypertension, n (%) | 65 (35) | 21 (21) | 0.021 | |
| History of cardiovascular disease, n (%) | 21 (11) | 4 (4) | 0.048 | |
| Hepatitis C seropositive, n (%) | 100 (54) | 28 (29) | 0.0001 | |
| Systolic blood pressure, mm Hg, median (P25, P75) | 120 (108, 131) | 126 (113, 135) | 0.0074 | |
| Diastolic blood pressure, mm Hg, median (P25, P75) | 71 (65, 77) | 73 (66, 82) | 0.058 | |
| Glycosylated hemoglobin, %, median (P25, P75) | 5.4 (5.1, 5.7) | 5.5 (5.3, 5.8) | 0.038 | |
| High-sensitivity C-reactive protein, mg/dL, median (P25, P75) | 1.7 (0.6, 4.2) | 1.9 (0.7, 5.5) | 0.43 | |
| Percentage activateda CD4 cells, median (P25, P75) | 8.3 (5.4, 14.1) | 3.8 (3.1–5.9) | <0.0001 | |
| Percentage activateda CD8 cells, median (P25, P75) | 30.7 (19.2, 46.9) | 10.8 (7.7, 20.5) | <0.0001 | |
| Urine albumin-creatinine ratio, mg/g, median (P25, P75) | 7 (3, 19) | 5 (3,11) | 0.18 | |
| Urine albumin-creatinine ratio > 30 mg/g, n (%) | 36 (19) | 9 (9) | 0.027 | |
| Serum creatinine, mg/dL, median (P25, P75) | 0.9 (0.8, 1.1) | 1.0 (0.8, 1.1) | 0.19 | |
| Serum cystatin C, mg/L, median (P25, P75) | 0.93 (0.82,1.10) | 0.84 (0.76, 1.10) | 0.0002 | |
| Measured glomerular filtration rate, ml/min/1.73m2, median (P25, P75) |
90 (76, 103) |
97 (84, 111) |
0.0044 |
|
| eGFRcr, ml/min/1.73m2, median (P25, P75) | 103 (85, 118) | 103 (92, 114) | 0.84 | |
| eGFRcys, ml/min/1.73m2, median (P25, P75) | 87 (70,103) | 101 (81, 112) | 0.0001 | |
| eGFRcr-cys, ml/min/1.73m2, median (P25, P75) | 95 (81, 109) | 100 (89, 114) | 0.012 | |
| Taking antiretroviral therapy, n (%) | 171 (91) | - | - | |
| Taking tenofovir, n (%) | 127 (68) | - | - | |
| Nadir CD4 count, cells/mm3, median (P25, P75) | 145 (42, 301) | - | - | |
| Current CD4 count, cells/mm3, median (P25, P75) | 464 (248, 627) | - | - | |
| HIV RNA > 400 copies/mL, n (%) | 38 (20) | - | - | |
| HIV RNA in subjects with values > 400 copies/mL, median (P25, P75) | 11,680 (4,562, 62,084) | - | - | |
P25 and P75, 25th and 75th percentiles, respectively; eGFRcr, eGFRcys, and eGFRcr-cys are glomerular filtration rates estimated by CKD-EPI equations using plasma creatinine, cystatin C, and both biomarkers, respectively.
a Activated CD4 or CD8 T-cells defined as expressing both CD38 and HLA-DR surface markers
Because mGFR was recalibrated approximately 10% lower and mGFR was central to analyses, almost all estimates in Table 2, Table 3, and Table 4 have been revised, with substantive changes described below. We also revised all Figs 1–3, although the clinical inferences from the figures are unchanged.
Table 2. Performance of glomerular filtration rate estimating equations in HIV-positive and HIV-negative participants.
| Performance measure | HIV-positive | HIV-negative | P valuea | |
|---|---|---|---|---|
| Accuracyb (95% CI) | eGFRcr |
79 (72, 85) |
88 (80, 94) |
0.075 |
| eGFRcys |
86 (81, 91) |
88 (80, 94) |
0.85 |
|
| eGFRcr-cys |
91 (86, 95) |
93 (86, 97) |
0.82 |
|
| P valuec |
eGFRcr vs. eGFRcys |
0.06329 |
1.00 |
- |
| eGFRcr vs. eGFRcr-cys |
0.000032 |
0.05878 |
- | |
| eGFRcys vs eGFRcr-cys |
0.08326 |
0.0587 |
- | |
| Biasd (P25, P75) | eGFRcr |
9.1 (-0.8, 21.0) |
3.5 (-6.1, 14.7) |
0.00496 |
| eGFRcys |
-4.6 (-17.1, 8.3) |
0.5 (-11.2, 13.3) |
0.0404 |
|
| eGFRcr-cys |
3.6 (-8.2, 12.2) |
2.5 (-6.0, 13.3) |
0.821 |
|
| P valuec | eGFRcr vs. eGFRcys |
<0.0001 |
0.01557 |
- |
| eGFRcr vs. eGFRcr-cys |
<0.0001 |
0.242 |
- | |
| eGFRcys vs eGFRcr-cys |
<0.0001 |
0.0002 |
- | |
| Precisione (95% CI) | eGFRcr | 22.8 (18.4, 27.3) | 20.9 (15.1, 26.7) | 0.50 |
| eGFRcys | 25.9 (22.1, 29.7) | 24.5 (18.9, 30.2) | 0.61 | |
| eGFRcr-cys | 22.0 (18.1, 25.9) | 19.8 (13.2, 26.4) | 0.49 | |
| P valuec | eGFRcr vs. eGFRcys | 0.43 | 0.65 | - |
| eGFRcr vs. eGFRcr-cys | 0.43 | 0.27 | - | |
| eGFRcys vs eGFRcr-cys | 0.12 | 0.10 | - | |
CI, confidence interval; eGFRcr, eGFRcys, and eGFRcr-cys are glomerular filtration rates estimated by CKD-EPI equations using plasma creatinine, cystatin C, and both biomarkers, respectively; P25 and P75, 25th and 75th percentiles, respectively
a Comparisons of a single equation between the HIV-positive and HIV-negative groups. P values in bold font indicate difference is statistically significant accounting for multiple comparisons (see text).
b Accuracy defined as percentage of estimated GFR values within 30% of measured GFR.
c Comparisons of a different equations within the HIV-positive or HIV-negative group. P values in bold font indicate difference is statistically significant accounting for multiple comparisons (see text).
d Bias defined as difference between estimated GFR and measured GFR (mL/min/1.73m2).
e Precision defined as interquartile range of bias.
Table 3. Factors associated with glomerular filtration rate estimating equation accuracya in HIV-positive and HIV-negative participants.
| Factor | HIV-positive | HIV-negative | |||
|---|---|---|---|---|---|
| eGFRcr | eGFRcys | eGFRcr | eGFRcys | ||
| Age, years | ≤ 49 |
78 (69, 86) |
85 (76, 91) |
83 (70, 92) |
87 (74, 94) |
| > 49 |
80 (70, 87) |
88 (80, 94) |
94 (82, 99) |
89 (77, 96) |
|
| P valueb |
1.00 |
0.67 |
0.13 |
0.76 |
|
| Body mass index, kg/m2 | ≤ 26 |
76 (67, 84) |
82 (73, 89) |
93 (82, 98) |
87 (73, 95) |
| > 26 |
82 (72, 89) |
91, 83, 96) |
83 (71, 92) |
89 (77, 96) |
|
| P valueb |
0.37 |
0.13 |
0.22 |
0.77 |
|
| Sex | Female |
72 (60, 83) |
86 (75, 93) |
74 (49, 91) |
89 (67, 99) |
| Male |
83 (75, 89) |
87 (79, 82) |
91 (83, 96) |
88 (78, 94) |
|
| P valueb |
0.13 |
0.82 |
0.050 |
1.00 |
|
| mGFR, mL/min/1.73m2 | < 90 |
65 (55, 75) |
81 (71, 88) |
68 (51, 82) |
79 (63, 90) |
| ≥ 90 |
93 (86, 98) |
92 (85, 97) |
100 (94, 100) |
93 (84, 98) |
|
| P valueb |
<0.0001 |
0.030 |
<0.0001 |
0.054 |
|
| Hepatitis C serostatus | Negative |
76 (66, 85) |
88 (79, 94) |
87 (77, 94) |
87 (77, 94) |
| Positive |
81 (72, 88) |
85 (76, 91) |
93 (76, 99) |
93 (76, 99) |
|
| P valueb |
0.47 |
0.67 |
0.50 |
0.50 |
|
| High-sensitivity C-reactive protein, mg/dL | ≤ 1.8 |
79 (70, 87) |
88 (80, 94) |
92 (80, 98) |
90 (78, 97) |
| > 1.8 |
79 (69, 87) |
84 (75, 91) |
84 (71, 93) |
86 (73, 94) |
|
| P valueb |
1.00 |
0.52 |
0.36 |
0.76 |
|
| Percentage activated CD4 cells | ≤ Medianc |
82 (72, 89) |
92 (84, 97) |
82 (69, 91) |
82 (69, 91) |
| > Medianc |
77 (70, 85) |
81 (71, 89) |
94 (83, 99) |
94 (83, 99) |
|
| P valueb |
0.46 |
0.047 |
0.12 |
0.12 |
|
| Percentage activated CD8 cells | ≤ Mediand |
82 (72, 89) |
93 (86, 97) |
82 (68, 91) |
84 (70, 93) |
| > Mediand |
77 (67, 85) |
80 (70, 88) |
94 (83, 99) |
92 (81, 98) |
|
| P valueb |
0.46 |
0.015 |
0.071 |
0.23 |
|
| Taking antiretroviral therapy | Yes |
78 (71, 84) |
88 (82, 93) |
||
| No |
94 (70, 100) |
69 (41, 89) |
|||
| P valueb |
0.20 |
0.016 |
|||
| Nadir CD4, cells/ mm3 | > 150 |
81(71, 89) |
88 (79, 94) |
||
| ≤ 150 |
77 (67, 85) |
85 (77, 92) |
|||
| P valueb |
0.59 |
0.67 |
|||
| Current CD4, cells/mm3 | > 450 |
84 (75, 90) |
91 (83, 96) |
||
| ≤ 450 |
74 (64, 83) |
82 (72, 89) |
|||
| P valueb |
0.15 |
0.13 |
|||
| HIV RNA, copies/ml | ≤ 400 |
80 (72, 86) |
90 (84, 99) |
||
| > 400 |
76 (60, 89) |
74 (57, 87) |
|||
| P valueb |
0.66 |
0.047 |
|||
eGFRcr and eGFRcys are glomerular filtration rates estimated by CKD-EPI equations using plasma creatinine and cystatin C, respectively; mGFR, measured glomerular filtration rate by iohexol clearance.
a Accuracy shown as percent of estimated GFR values within 30% of measured GFR values (95% confidence interval).
b P values in bold font indicate difference is statistically significant accounting for multiple comparisons (see text).
c Medians 8.3% and 3.8% in HIV-positive and HIV-negative groups, respectively.
d Medians 30.7% and 10.7% in HIV-positive and HIV-negative groups, respectively.
Table 4. Factors associated with glomerular filtration rate equation biasa in HIV-positive and HIV-negative participants.
| Factor | HIV-positive |
HIV-negative |
|||
|---|---|---|---|---|---|
| eGFRcr | eGFRcys | eGFRcr | eGFRcys | ||
| Age, years | ≤ 49 |
8.3 (-2.6, 22.5) |
-4.4 (-16.2, 9.9) |
3.3 (-6.5, 15.0) |
0.7 (-7.4, 15.0) |
| > 49 |
10.2 (2.2, 19.6) |
-6.2 (-17.9, 7.0) |
6.3 (-5.1, 13.9) |
-4.4 (-12.9, 10.9) |
|
| P valueb |
0.53 |
0.46 |
0.81 |
0.29 |
|
| Body mass index, kg/ m2 | ≤ 26 |
8.0 (-1.1, 22.2) |
-6.5 (-18.4, 8.2) |
2.1 (-7.8, 13.7) |
-0.5 (-9.5, 13.3) |
| > 26 |
10.2 (0.2, 20.8) |
-3.0 (-14.3, 9.7) |
4.2 (-3.2, 15.3) |
0.6 (-11.7, 12.9) |
|
| P valueb |
0.93 |
0.36 |
0.15 |
0.89 |
|
| Sex | Female |
15.5 (4.3, 25.1) |
-3.0 (-16.8, 8.0) |
12.1 (2.4, 23.4) |
0.5 (-9.3, 8.8) |
| Male |
6.9 (-1.9, 18.7) |
-5.9 (-17.1, 9.8) |
1.7 (-7.8, 13.6_ |
0.8 (-11.6, 15.1) |
|
| P valueb |
0.0042 |
0.96 |
0.0019 |
0.94 |
|
| mGFR, mL/min/1.73 m2 | < 90 |
15.1 (2.7, 29.2) |
0 (-13.4, 9.9) |
13.4 (-1.9, 24.1) |
8.1 (-3.1, 21.7) |
| ≥ 90 |
5.7 (-4.5, 17.1) |
-9.5 (-20.9, 5.7) |
1.0 (-7.8, 9.3) |
-5.5 (-13.7, 4.7) |
|
| P valueb |
0.0001 |
0.0029 |
0.0005 |
0.0001 |
|
| Hepatitis C serostatus | Negative |
9.1 (-1.1, 22.6) |
2.9 (-11.6, 12.7) |
3.4 (-6.3, 15.3) |
3.5 (-7.4, 16.1) |
| Positive |
8.6 (0.9, 20.3) |
-9.8 (-18.4, 4.4) |
3.3 (-5.6, 12.7) |
-9.5 (-21.4, -0.4) |
|
| P valueb |
0.85 |
0.0008 |
0.91 |
0.0003 |
|
| High-sensitivity C-reactive protein, mg/dl | ≤ 1.8 |
9.1 (1.7, 21.0) |
-3.4 (18.9, 8.1) |
3.4 (-8.5, 14.4) |
-0.5 (-11.5, 14.9) |
| >1.8 |
8.6 (-1.1, 22.2) |
-6.2 (-15.4, 9.2) |
4.2 (-4.7, 15.3) |
0.6 (-8.7, 9.1) |
|
| P valueb |
0.90 |
0.80 |
0.43 |
0.76 |
|
| Percentage activated CD4 cells | ≤ Medianc |
8.7 (-1.9, 18.6) |
2.9 (-9.9, 10.9) |
3.3 (-6.6, 19.8) |
1.3 (-10.1, 15.8) |
| > Medianc |
10.2 (1.2, 22,8) |
-12.2 (-22.7, 1.3) |
3.8 (-3.6, 10.1) |
-5.5 (-11.6, 7.6) |
|
| P valueb |
0.25 |
<0.0001 |
0.86 |
0.10 |
|
| Percentage activated CD8 cells | ≤ Medianc |
7.0 (-2.1, 19.4) |
0.1 (-10.6, 12.4) |
6.3 (-6.3, 17.7) |
0.7 (-10.1, 15.8) |
| > Medianc |
13.0 (1.3, 22.6) |
-9.9 (-21.2, 4.5) |
2.1 (-5.0 10.1) |
-0.6 (-11.6, 7.6) |
|
| P valueb |
0.11 |
0.0002 |
0.35 |
0.23 |
|
| Taking antiretroviral therapy | Yes |
9.1 (-0.8, 21.4) |
-3.3 (-15.6, 9.4) |
||
| No |
9.9 (0.8, 21.3) |
-16.2 (-28.9, -9.7) |
|||
| P valueb |
0.89 |
0.0022 |
|||
| Nadir CD4 count, cells/mm3 | >150 |
5.9 (-5.0, 19.6) |
-8.3 (-17.8, 8.0) |
||
| ≤150 |
13.2 (3.7, 22.8) |
-2.8 (-15.5, 8.4) |
|||
| P valueb |
0.0035 |
0.15 |
|||
| CD4 count, cells/mm3 | > 450 |
8.6 (-1.3, 19.4) |
-3.0 (-13.3, 9.8) |
||
| ≤ 450 |
10.2 (1.9, 23.4) |
-7.9 (-21.7, 7.6) |
|||
| P valueb |
0.13 |
0.027 |
|||
| HIV RNA, copies/ml | ≤ 400 |
9.5 (-0.9, 20.4) |
-0.8 (-12.7, 10.2) |
||
| > 400 |
7.9 (1.3, 25.4) |
-16.8 (-31.3, -7.1) |
|||
| P valueb |
0.89 |
<0.0001 | |||
eGFRcr and eGFRcys are glomerular filtration rates estimated by CKD-EPI equations using plasma creatinine and cystatin C, respectively; mGFR, measured glomerular filtration rate by iohexol clearance.
a Bias defined as median difference between estimated glomerular filtration rate (GFR) and measured GFR (25th percentile, 75th percentile)
b P values in bold font indicate difference is statistically significant accounting for multiple comparisons (see text).
c Medians 8.3% and 3.8% in HIV-positive and HIV-negative groups, respectively.
d Medians 30.7% and 10.7% in HIV-positive and HIV-negative groups, respectively
Fig 1.
Bland-Altman plots for estimated and measured glomerular filtration rate (GFR) in HIV-positive participants using the CKD-EPI equations for serum creatinine (A), cystatin C (B), or both biomarkers (C). The average GFR (measured and estimated) is shown on the X axes. Bias, defined as the difference between estimated and measured GFR, is displayed on the Y axes. The average biases are represented by the horizontal solid lines and the horizontal dashed lines represent 2 standard deviations above and below the averages.
Fig 3.
Correlation of estimated glomerular filtration rate (eGFR) bias, defined as the difference between eGFR and measured GFR, with percentage of activated CD8 T cells (CD38+ and HLA-DR+) using the creatine-based CKD-EPI equation in HIV-negative (A) and HIV-positive (B) subjects, and the cystatin C-based CKD-EPI equation in HIV negative (C) and HIV-positive (D) subjects. The percentage of CD8+ T cells with an activated phenotype is shown on the X axes (note, different scales for HIV-positive and HIV-negative groups). Rho is the spearman rank correlation coefficient, which may vary between -1 and 1. The dashed lines represent least-squares regression lines.
Fig 2.
Bland-Altman plots for estimated and measured glomerular filtration rate (GFR) in HIV-negative participants using the CKD-EPI equations for serum creatinine (A), cystatin C (B), or both biomarkers (C). The average GFR (measured and estimated) is shown on the X axes. Bias, defined as the difference between estimated and measured GFR, is displayed on the Y axes. The average biases are represented by the horizontal solid lines and the horizontal dashed lines represent 2 standard deviations above and below the averages.
In the original paper, we reported that the cystatin C-based equation (eGFRcys) was the least accurate and most biased of the three CKD-EPI equations in HIV-positive participants. In the revised analysis, we found that the creatinine-based equation (eGFRcr) was the least accurate and most biased of the three equations. This is relevant because eGFRcr is the most commonly used equation in clinical practice. Consistent with the original analysis, the combined biomarker equation (eGFRcr-cys) remained the most accurate and least biased equation.
In contrast to the original analysis, we found that the accuracy and bias of eGFRcr varied significantly by stratum of mGFR (<90 vs. ≥90 mL/min/1.73m2) in both the HIV-positive and HIV-negative groups, such that this equation was more biased and less accurate at lower levels of kidney function than at higher kidney function. This is important, because accurate GFR estimation may be more important at lower compared with higher levels of kidney function.
Consistent with the original analysis, we found that the bias of eGFRcys was influenced by immune activation and HIV viremia, whereas eGFRcr performance was not affected by these factors (Fig 3). However, in contrast to the original analysis, these factors were no longer statistically significantly associated with the accuracy of eGFRcys.
Please see the revised Figs 1–3 and revised Tables 2–4 here.
A member of PLOS ONE's Editorial Board reviewed the new results and underlying data and confirmed that they support the overall conclusions reported in the article.
Supporting information
Clinical Variables eGFR and mGRF.
(DTA)
References
- 1.Bhasin B, Lau B, Atta MG, Fine DM, Estrella MM, Schwartz GJ, et al. (2013) HIV Viremia and T-Cell Activation Differentially Affect the Performance of Glomerular Filtration Rate Equations Based on Creatinine and Cystatin C. PLoS ONE 8(12): e82028 10.1371/journal.pone.0082028 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Schwartz GJ, Wang H, Erway B, et al. Multicenter Laboratory Comparison of Iohexol Measurement. The Journal of Applied Laboratory Medicine: An AACC Publication 2017: jalm. 2017.024240. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Clinical Variables eGFR and mGRF.
(DTA)