Abstract: TH-OR92
Performance of Creatinine and Cystatin C-Based Equations Among Patients with Hematological or Solid Cancers: Real-World Data from a Clinical Cohort
Session Information
- Onconephrology and Precision Pharmacology
November 02, 2023 | Location: Room 107, Pennsylvania Convention Center
Abstract Time: 04:39 PM - 04:48 PM
Category: Onconephrology
- 1700 Onconephrology
Authors
- Titan, Silvia, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Lieske, John C., Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Meeusen, Jeff W., Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Rule, Andrew D., Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Herrmann, Sandra, Mayo Clinic Minnesota, Rochester, Minnesota, United States
Background
Performance of creatinine-based eGFR equations in patients with malignancy may be worse than in the general population with implications for management including drug dosing.
Methods
We retrospectively studied all cases who had mGFR (urinary iothalamate clearance) performed at Mayo Clinic 2017-2023 and ICD codes for neoplasia (excluding in situ and benign lesions) up to one year prior to the mGFR date. Serum creatinine and cystatin C within 2 days of the mGFR were used to compute CKD-EPI 2021 eGFRcr and eGFRcrcys, and 2012 eGFRCys. Bias (absolute and % mGFR - eGFR) and P30 (% with eGFR >=30% different from mGFR) and confidence intervals (bootstrap) were computed. Analyses were stratified for cancer type.
Results
Results by cancer type are in the Table. Among 3719 patients with eGFRcr, the median bias was similar to that in the literature. Among the 522 patients with available creatinine plus cystatin C, eGFRcr had less bias but eGFRcrcys (15.5% [12.6 – 18.9]) was more accurate compared to eGFRcr (26.7% [23.1 – 30.8]) or eGFRcys (25.5% [12.6 – 18.9]). Results were similar by cancer type.
Conclusion
Our study suggests that among patients with either solid or hematological malignancies, the CKDEPI eGFRcrcys better reflects mGFR than either eGFRcr or eGFRcys. When possible, mGFR may be optimal for clinical decision making in these populations.
Table 1.
| Median Bias / CI (mL/min/1.73m2 | P30 / CI (%) | |
| Creatinine and Cystatin C available | ||
| All patients (n=522) | ||
| eGFRcr | -0.81 (-1.57 , 0.44) | 26.7 (23.08 , 30.77) |
| eGFRcys | 6.29 (4.65 , 7.88) | 25.5 (22.07 , 29.07) |
| eGFRcrcys | 3.06 (2.00 , 4.27) | 15.5 (12.60 , 18.85) |
| Solid tumors (n=337) | ||
| eGFRcr | -0.64 (-2.10 , 0.60) | 24.6 (20.24 , 29.46) |
| eGFRcys | 6.29 (4.48 , 8.03) | 24.3 (20.17 , 29.06) |
| eGFRcrcys | 3.16 (1.90 - 4.83) | 15.7 (11.77 , 19.05) |
| Hematological (n=185) | ||
| eGFRcr | -1.85 (-4.24 , 1.25) | 31.0 (23.91 , 37.21) |
| eGFRcys | 6.18 (2.90 , 9.18) | 27.7 (21.2 , 33.70) |
| eGFRcrcys | 3.02 (0.95 - 5.77) | 15.2 (10.07 , 20.11) |
| Creatinine only available | ||
| All patients (n=3719) | ||
| eGFRcr | 1.42 (0.79 , 1.92) | 18.0 (16.66 - 19.08) |
| Solid tumors (n=2039) | ||
| eGFRcr | 1.51 (0.91 , 2.08) | 18.4 (16.52 , 20.16) |
| Hematological (n=1680) | ||
| eGFRcr | 1.23 (0.13 , 1.92) | 17.5 (15.6 , 19.25) |