Abstract: SA-PO1230
Limited Accuracy of eGFR Compared with Iohexol-Measured GFR in Patients Receiving CDK4/6 Inhibitors
Session Information
- Onconephrology: Epidemiological Trends, Risk Stratification, and Clinical Outcomes
October 24, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Onconephrology
- 1600 Onconephrology
Authors
- Piko, Nejc, Univerzitetni Klinicni Center Maribor, Maribor, Administrative Unit Maribor, Slovenia
- Klavz, Jani, Univerzitetni Klinicni Center Maribor, Maribor, Administrative Unit Maribor, Slovenia
- Petreski, Tadej, Univerzitetni Klinicni Center Maribor, Maribor, Administrative Unit Maribor, Slovenia
- Stopinšek Rajšp, Mateja, Univerzitetni Klinicni Center Maribor, Maribor, Administrative Unit Maribor, Slovenia
- Bevc, Sebastjan, Univerzitetni Klinicni Center Maribor, Maribor, Administrative Unit Maribor, Slovenia
Background
Cyclin-dependent kinase 4/6 inhibitors (CDK4/6i) used in breast cancer may cause true acute kidney injury (AKI) or pseudo-AKI from impaired creatinine secretion. We prospectively assessed kidney function before and 3 months after CDK4/6i initiation, focusing on changes in kidney function and eGFR accuracy.
Methods
We established a standardized protocol (Figure 1) for assessing kidney function using iohexol-measured GFR (mGFR)and compared it with conventional assessment based on eGFR (CKD-EPI 2021 creatinine–cystatin C equation). SPSS was used to perform statistical analysis.
Results
We included 12 female patients, mean age 60.0±12.8 years (range 34–80). Mean eGFR prior to CDK4/6i initiation was 70.8±30.7 mL/min/1.73 m2 (range 25–130.2), decreasing to 59.4±29.9 mL/min/1.73 m2 at 3 months (range 16–109.5, p=0.003). In contrast, mean mGFR declined only modestly from 75.0±19.3 to 70.4±22.3 mL/min/1.73 m2 (range 43–101 and 32–101, p=0.254). Individual paired trajectories of eGFR and mGFR are presented in Figure 2. Spearman correlation demonstrated moderate association between eGFR and mGFR at baseline (rs=0.727, p=0.007) and follow-up (rs=0.605, p=0.037). Bland–Altman analysis showed minimal bias at baseline (−4.1 mL/min/1.73 m2) but wide limits of agreement (−32.1 to 23.9), indicating poor concordance. At follow-up, performance further worsened, with increased bias (−10.2 mL/min/1.73 m2) and markedly wider limits of agreement (−57.3 to 36.9). Accuracy also declined, with percentage of eGFR within ±30% of mGFR decreasing from 73.7% at baseline to 41.7% at follow-up.
Conclusion
In women receiving CDK4/6i, eGFR increasingly diverged from mGFR over time, suggesting that declines in eGFR mainly reflect pseudo-AKI rather than true kidney dysfunction. Creatinine- and cystatin C–based estimates were unreliable, supporting the use of mGFR.