Background
It is unclear which kidney function estimating equation (KFEE) is most accurate in transgender or gender diverse (TGD) adults and which sex coefficient should be applied. This study compared precision/bias/accuracy of existing KFEEs, a new sex-free equation, and measured GFR (mGFR).
Methods
TGD participants were recruited from an HIV prevention trial (ClinicalTrials.gov ID: NCT04616963). This study was approved by the institutional review board at University of California San Diego and performed in accordance with the Declaration of Helsinki.
Inclusion criteria for the parent trial were (1) age ≥18 years, (2) discrepant gender identity (GI) and sex assigned at birth (SAAB), and (3) negative HIV test and creatinine clearance ≥60 ml/min. Substudy participants were excluded if they had an iohexol allergy or were using medications that could interfere with iohexol clearance.
Iohexol was diluted 1:1 with sterile water for injection then injected (1 ml) into the subcutaneous fat overlying the triceps of fasted/semifasted participants within 1 hour of compounding. Blood/urine samples were collected 3 hours after administration.
Pre-iohexol blood samples were assayed for standardized creatinine and cystatin C, as well as estradiol, estrone, testosterone (free and total), and sex hormone binding globulin. Pre-iohexol urine samples were assayed for urinary creatinine and cystatin C. Plasma iohexol concentrations were assayed using high-performance liquid chromatography.
mGFR was calculated based on a single time point iohexol concentration using the Jacobsson equation and SAAB.1,2 All versions of CKD Epidemiology Collaboration (CKD-EPI) and European Kidney Function Consortium equations were used to calculate eGFR using serum cystatin C, serum creatinine (SCr) or both, where applicable, and SAAB or GI.3–6 All values were adjusted to body surface area.7
We employed linear regression to develop a new sex-free equation using mGFR as the dependent variable after verifying underlying model assumptions. The approach to evaluating the bias/precision/accuracy of equations was adapted from the study by Inker et al.3 Model predictive performance was iteratively assessed using a cross-validation jackknife procedure.
Results
Thirty-one individuals were enrolled. The median (interquartile range) age was 33 (26–36) years, were White race (55%), and were male SAAB (71%). 71% used gender affirming hormone therapy (GAHT) with 32% on estrogen/anti-androgen therapy, 26% on testosterone only, and 13% on estrogen only.
The relationship between mGFR and KFEEs, for the total cohort and stratified by GI, SAAB, and GAHT, are displayed in Table 1. Of the KFEEs, EKFCSCr demonstrated the lowest bias and highest accuracy. The percentage of eGFRs within 30% of mGFR (P30) was low overall for each KFEEs, and GAHT users appeared to have lower P30 compared with nonusers.
Table 1.
Performance of existing, modified, and newly developed kidney function estimating equations against measured kidney function
| Equation | GFR, Median (IQR), ml/min | Bias (95% CI) | Precision | Accuracy/P30 (95% CI) |
|---|---|---|---|---|
| Iohexol clearance mGFRa | — | — | — | |
| All participants, SAAB, (n=31) | 75 (64–96) | |||
| • AMAB, GAHT (n=14) | 75 (63–98) | |||
| • AMAB, no GAHT (n=8) | 87 (75–100) | |||
| • AFAB, GAHT (n=8) | 67 (59–76) | |||
| • AFAB, no GAHT (n=1) | 111 | |||
| All participants, GI, (n=31) | 76 (65–96) | |||
| • Female (n=15) | 78 (66–102) | |||
| • Male (n=16) | 76 (64–90) | |||
| CKD-EPI_2012CysC eGFR | ||||
| All participants, SAAB, (n=31) | 128 (114–138) | 49.7 (15.6 to 78.4) | 36.9–61.2 | 32.3 (16.7 to 51.4) |
| • AMAB, GAHT (n=14) | 133 (115–139) | 57.0 (11.3 to 75.9) | 37.2–68.9 | 28.6 (8.3 to 58.1) |
| • AMAB, no GAHT (n=8) | 136 (127–153) | 52.7 (28.7 to 70.1) | 39.4–68.1 | 62.5 (24.5 to 91.5) |
| • AFAB, GAHT (n=8) | 105 (99–125) | 47.4 (28.4 to 56.2) | 34.5–54.5 | 0 |
| • AFAB, no GAHT (n=1) | 136 | 25.5 | NC | 100 |
| All participants, GI, (n=31) | 110 (97–119) | 34.0 (−27.0 to 79.5) | 16.8–45.7 | 25.8 (11.9 to 44.6) |
| • Female (n=15) | 105 (88–117) | 34.0 (−39.9 to 60.5) | −7.1 to 45.7 | 26.7 (7.8 to 55.1) |
| • Male (n=16) | 118 (102–120) | 37.4 (2.9 to 78.8) | 26.5–46.4 | 25.0 (7.3 to 52.4) |
| CKD-EPI_2021Cr eGFR | ||||
| All participants, SAAB, (n=31) | 124 (103–138) | 43.0 (16.2 to 71.1) | 31.0–56.1 | 38.7 (21.9 to 57.8) |
| • AMAB, GAHT (n=14) | 130 (108–142) | 50.6 (9.6 to 70.4) | 30.2–63.2 | 35.7 (12.8 to 64.9) |
| • AMAB, no GAHT (n=8) | 125 (110–137) | 37.9 (20.6 to 40.8) | 28.3–40.5 | 75.0 (34.9 to 96.8) |
| • AFAB, GAHT (n=8) | 105 (102–121) | 43.7 (23.8 to 49.4) | 38.4–49.1 | 0 |
| • AFAB, no GAHT (n=1) | 175 | 64.7 | NC | 100 |
| All participants, GI, (n=31) | 100 (86–116) | 21.4 (−22.6 to 46.6) | 3.9–36.7 | 51.6 (33.1 to 70.0) |
| • Female (n=15) | 94 (80–103) | 15.7 (−33.9 to 37.8) | −3.7 to 31.9 | 60.0 (32.3 to 83.7) |
| • Male (n=16) | 108 (91–118) | 25.6 (−4.7 to 57.1) | 12.1–47.3 | 43.8 (19.8 to 70.1) |
| CKD-EPI_2021CrCysC eGFR | ||||
| All participants, SAAB, (n=31) | 121 (106–140) | 45.6 (13.3 to 74.9) | 27.4–55.6 | 35.5 (19.2 to 54.6) |
| • AMAB, GAHT (n=14) | 126 (111–142) | 52.3 (10.0 to 73.0) | 32.0–64.6 | 28.6 (8.3 to 58.1) |
| • AMAB, no GAHT (n=8) | 132 (121–146) | 51.3 (27.4 to 56.7) | 30.7–56.4 | 62.5 (24.5 to 91.5) |
| • AFAB, GAHT (n=8) | 92 (85–107) | 28.4 (19.2 to 38.6) | 22.8–36.6 | 12.5 (0.3 to 52.7) |
| • AFAB, no GAHT (n=1) | 140 | 29.3 | NC | 100 |
| All participants, GI, (n=31) | 107 (91–118) | 31.4 (−21.8 to 64.3) | 19.9–46.0 | 32.3 (16.7 to 51.4) |
| • Female (n=15) | 104 (86–108) | 26.1 (−35.9 to 47.8) | −2.7 to 38.7 | 33.3 (11.8 to 61.6) |
| • Male (n=16) | 114 (103–113) | 38.4 (4.9 to 61.4) | 21.2–47.7 | 31.3 (11.0 to 58.7) |
| EKFCCr eGFR | ||||
| All participants, SAAB, (n=31) | 100 (82–110) | 13.2 (−11.0 to 52.3) | 5.9–34.1 | 67.7 (48.6 to 83.3) |
| • AMAB, GAHT (n=14) | 109 (102–113) | 27.3 (−11.5 to 51.5) | 13.1–47.9 | 50.0 (23.0 to 77.0) |
| • AMAB, no GAHT (n=8) | 99 (85–110) | 8.8 (−7.4 to 37.6) | 6.0–17.1 | 75.0 (34.9 to 96.8) |
| • AFAB, GAHT (n=8) | 76 (68–82) | 12.2 (−10.7 to 18.7) | −1.0 to 17.6 | 87.5 (47.4 to 99.7) |
| • AFAB, no GAHT (n=1) | 110 | −0.5 | NC | 100 |
| All participants, GI, (n=31) | 94 (83–105) | 13.7 (−26.7 to 46.6) | −0.5 to 26.8 | 61.3 (42.2 to 78.2) |
| • Female (n=15) | 85 (77–94) | 6.1 (−36.6 to 27.4) | −7.9 to 21.7 | 66.7 (38.4 to 88.2) |
| • Male (n=16) | 98 (87–108) | 19.8 (−10.8 to 44.7) | 5.6–38.6 | 56.3 (29.9 to 80.3) |
| EKFCCysC eGFR | ||||
| All participants, SAAB, (n=31) | 108 (96–112) | 31.1 (8.3 to 58.9) | 8.3–44.3 | 35.5 (19.2 to 54.6) |
| • AMAB, GAHT (n=14) | 107 (93–112) | 31.7 (−33.8 to 49.0) | 3.5–44.6 | 35.7 (12.8 to 64.4) |
| • AMAB, no GAHT (n=8) | 109 (100–111) | 15.3 (3.4 to 33.3) | 6.1–31.1 | 62.5 (24.5 to 91.5) |
| • AFAB, GAHT (n=8) | 109 (97–114) | 38.7 (24.4 to 53.0) | 31.5–52.2 | 0 |
| • AFAB, no GAHT (n=1) | 96 | −15.2 | NC | 100 |
| All participants, GI, (n=31) | 108 (96–112) | 30.6 (−22.1 to 58.8) | 8.4–41.9 | 38.7 (21.9 to 57.8) |
| • Female (n=15) | 107 (93–112) | 30.4 (−32.4 to 49.9) | −7.6 to 43.9 | 40.0 (16.3 to 67.7) |
| • Male (n=16) | 109 (97–111) | 31.8 (3.8 to 56.4) | 16.2–41.6 | 37.5 (15.2 to 64.6) |
| New equationb | ||||
| All participants, SAAB, (n=31) | 81 (67–89) | −0.8 (−26.8 to 19.1) | −8.1 to 8.9 | 87.1 (70.2 to 96.4) |
| • AMAB, GAHT (n=14) | 83 (68–95) | −0.2 (−26.0 to 16.6) | −7.6 to 10.8 | 85.7 (47.4 to 99.7) |
| • AMAB, no GAHT (n=8) | 85 (74–90) | −0.5 (−28.8 to 9.8) | −10.9 to 8.6 | 87.5 (47.4 to 99.7) |
| • AFAB, GAHT (n=8) | 67 (58–76) | 1.4 (−11.7 to 10.8) | −4.7 to 9.3 | 87.5 (47.4 to 99.7) |
| • AFAB, no GAHT (n=1) | 110 | −0.9 | NC | 100 |
| All participants, GI, (n=31) | 81 (67–89) | −0.8 (−26.8 to 19.1) | −8.1 to 8.9 | 87.1 (70.2 to 96.4) |
| • Female (n=15) | 78 (64–105) | −0.8 (−26.0 to 19.4) | −7.4 to 10.4 | 84.0 (59.5 to 98.3) |
| • Male (n=16) | 75 (63–90) | 0.1 (−28.8 to 13.4) | −7.6 to 8.6 | 90.0 (68.0 to 99.8) |
AFAB, assigned female at birth; AMAB, assigned male at birth; CI, confidence interval; CKD-EPI, CKD Epidemiology Collaboration; CL, clearance; EKFC, European Kidney Function Consortium; GAHT, gender affirming hormone therapy; GI, gender identity reflected by gender affirming hormone therapy use and non-gender affirming hormone therapy users categorized as their sex assigned at birth; IQR, interquartile range; mGFR, measured GFR; NC, not computed because n=1; P30, percentage of estimated values within 30% of measured value; SAAB, sex assigned at birth.
Calculated using Jacobsson et al.1, CL = (1/[t/V+0.0016])×ln (Dose/[V×Ct]) where CL was iohexol clearance, t was minutes between iohexol administration and blood sampling for iohexol concentration and V was the apparent volume of distribution. V based on sex assigned at birth: male=166× weight (kg)+2490 and female=95×weight (kg)+6170.
New equation: eGFR=71.239+(−49.446×Serum Creatinine)+(0.661×Actual Weight)+(−0.03×Urine Creatinine)+(0.159×Age).
The new equation incorporated age, actual body weight, SCr, and urine creatinine. The median (interquartile range) clearance was 80 (68–89) ml/min per 1.73 m2. This equation had low bias (−0.8 ml/min) and high accuracy (96.8%). After cross-validation, the accuracy was 87.1%.
Discussion
This study simultaneously compared contemporary KFEEs against mGFR and illuminates discrepancies affecting TGD persons. Lower P30 values with KFEEs may be driven by GAHT, but the small sample size limited statistical confirmation. This study provides additional evidence that the sex coefficient may be removed in the transgender population without compromising precision and accuracy, but requires validation in larger cohorts. The EKFCSCys is a sex-free equation with similar accuracy to the CKD-EPI equations and may be of interest for TGD persons. In contrast to our results, Miranda et al. found high performance of KFEE using SCr and serum cystatin C.8 It is important to note that all participants were on long-term GAHT, received an intravenous infusion (versus subcutaneous injection) of iohexol, and mGFR was the mean value of the Jacobsson equation when SAAB and GI were used in the calculations.
Some limitations should be considered. Small sample sizes hinder precision. Individuals were trial participants with a creatinine clearance ≥60 ml/min because of renal dosing thresholds for tenofovir disoproxil fumarate. This study does not address kidney function estimation in TGD persons with CKD. The CKD-EPI equations were developed/validated in persons across a wider spectrum of mGFR including CKD, and this may explain, in part, the accuracy observed. We used the Jacobsson equation which includes a binary sex coefficient. Future considerations include use of multiple samples to directly calculate clearance. Finally, this equation needs to undergo more rigorous validation before widespread use, but our findings demonstrate that a sex-free and race-free equation can be developed that accurately predicts kidney function in TGD persons with a P30 that remains stable regardless of GAHT use.
In summary, existing KFEEs do not provide an accurate eGFR among TGD persons which could lead to inappropriate exclusion of medications/interventions. A simple sex-free equation demonstrated good accuracy/bias but requires additional validation in a larger cohort before widespread use.
Supplementary Material
Acknowledgments
The project described was partially supported by the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. We thank Euyhyun Lee and Lin Liu for their biostatistics support.
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/CJN/C349.
Author Contributions
Conceptualization: Linda Awdishu, Jill Blumenthal, Niamh Higgins, Sheldon Morris, Letiticia Muttera, Nimish Patel.
Data curation: Karen Chow, Nimish Patel.
Formal analysis: Nimish Patel.
Funding acquisition: Sheldon Morris, Nimish Patel.
Investigation: Leah B. Burke, Karen Chow, Letiticia Muttera, Jordan Silva.
Methodology: Linda Awdishu, Jill Blumenthal, Leah B. Burke, Karen Chow, Niamh Higgins, Sheldon Morris, DeeDee Pacheco, Nimish Patel, Sucheta Vaingankar.
Project administration: Leah B. Burke, Karen Chow, Niamh Higgins, Sheldon Morris, Letiticia Muttera, DeeDee Pacheco, Nimish Patel, Jordan Silva, Sucheta Vaingankar.
Resources: DeeDee Pacheco.
Supervision: Jill Blumenthal, Sheldon Morris.
Validation: Niamh Higgins.
Visualization: Niamh Higgins, Sucheta Vaingankar.
Writing – original draft: Linda Awdishu, Leah B. Burke, Niamh Higgins, Sheldon Morris, Nimish Patel, Sucheta Vaingankar.
Writing – review & editing: Linda Awdishu, Jill Blumenthal, Sheldon Morris, Nimish Patel.
Funding
N. Patel: California HIV/AIDS Research Program (B19-SD-002). This work was supported by National Institutes of Health (UL1TR001442).
Data Availability Statements
Partial restrictions to the data and/or materials apply. Researchers may submit reasonable data sharing request to the corresponding author after two years from date of publication. Accompanied with this request, researchers must submit a methodologically sound protocol with achievable specific aims outlined. Researchers will need to complete the requisite data use agreement paperwork from the Regents of California to gain access to deidentified data.
References
- 1.Jacobsson L. A method for the calculation of renal clearance based on a single plasma sample. Clin Physiol. 1983;3(4):297–305. doi: 10.1111/j.1475-097x.1983.tb00712.x [DOI] [PubMed] [Google Scholar]
- 2.Delanaye P Ebert N Melsom T, et al. Iohexol plasma clearance for measuring glomerular filtration rate in clinical practice and research: a review. Part 1: how to measure glomerular filtration rate with iohexol? Clin Kidney J. 2016;9(5):682–699. doi: 10.1093/ckj/sfw070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Inker LA Eneanya ND Coresh J, et al.; Chronic Kidney Disease Epidemiology Collaboration. New creatinine- and cystatin C-based equations to estimate GFR without race. N Engl J Med. 2021;385(19):1737–1749. doi: 10.1056/NEJMoa2102953 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Inker LA Schmid CH Tighiouart H, et al.; CKD-EPI Investigators. Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med. 2012;367(1):20–29. doi: 10.1056/NEJMoa1114248 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pottel H Björk J Courbebaisse M, et al. Development and validation of a modified full age spectrum creatinine-based equation to estimate glomerular filtration rate: a cross-sectional analysis of pooled data. Ann Intern Med. 2021;174(2):183–191. doi: 10.7326/M20-4366 [DOI] [PubMed] [Google Scholar]
- 6.Pottel H Björk J Rule AD, et al. Cystatin C-based equation to estimate GFR without the inclusion of race and sex. N Engl J Med. 2023;388(4):333–343. doi: 10.1056/NEJMoa2203769 [DOI] [PubMed] [Google Scholar]
- 7.Du Bois D, Du Bois EF. A formula to estimate the approximate surface area if height and weight be known. 1916. Nutrition. 1989;5(5):303–313; discussion 312-3. PMID: 2520314 [PubMed] [Google Scholar]
- 8.Turino Miranda K Dumanski SM Saad N, et al. Glomerular filtration rate estimation in transgender and gender-diverse adults using gender-affirming hormone therapy: an exploratory cross-sectional study. Kidney Int. 2024;106(4):753–756. doi: 10.1016/j.kint.2024.06.030 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Partial restrictions to the data and/or materials apply. Researchers may submit reasonable data sharing request to the corresponding author after two years from date of publication. Accompanied with this request, researchers must submit a methodologically sound protocol with achievable specific aims outlined. Researchers will need to complete the requisite data use agreement paperwork from the Regents of California to gain access to deidentified data.