Abstract: FR-PO1182
Beyond Diagnosis: Post-Treatment Histological Phenotypes Improve Risk Stratification in Antibody-Mediated Rejection
Session Information
- Transplantation: Clinical - Rejection, Biomarkers, and Pharmacology
October 23, 2026 | Location: Exhibit Hall A, Convention Center
Abstract Time: 10:00 AM - 12:00 PM
Category: Transplantation
- 2002 Transplantation: Clinical
Authors
- Camacho Murillo, Luis Agustín, Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, CDMX, Mexico
- Zuñiga Gonzalez, Erick Yasar, Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, CDMX, Mexico
- Uribe-Uribe, Norma O., Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, CDMX, Mexico
- Morales-Buenrostro, Luis E., Instituto Nacional de Ciencias Medicas y Nutricion Salvador Zubiran, Mexico City, CDMX, Mexico
Background
Antibody-mediated rejection (ABMR) remains a leading cause of kidney allograft loss. While the Banff classification at diagnosis (T0) provides a baseline risk, it fails to capture the dynamic response to treatment. We aimed to evaluate the incremental prognostic value of follow-up biopsies (T1) by identifying histological phenotypes through unsupervised learning and assessing their impact on graft survival.
Methods
We analyzed 138 kidney transplant recipients with biopsy-proven ABMR. Unsupervised PCA-based clustering was used to integrate Banff Activity (IA) and Chronicity (IC) indices into risk phenotypes (Low, Intermediate, High) at T0 and T1. Multivariable Cox proportional hazards models, adjusted for age at rejection and eGFR at diagnosis, were used to assess graft loss. Incremental value was determined via the Likelihood Ratio Test (LRT) and changes in the C-statistic.
Results
The median follow-up was 53 (IQR 31-84) months with 24 graft losses (17.4%). At T0, the high-risk phenotype was a strong predictor of failure (HR: 7.37, p<0.001; C-index: 0.821) Figure 1A-D. However, the addition of T1 data significantly improved the model's fit (LRT: p=0.015) and discrimination (C-index: 0.846). In the combined multivariable model, a persistent high-risk phenotype at T1 remained a potent independent predictor of graft loss (HR: 4.08, 95%CI: 1.32-12.53, p=0.014), while eGFR at rejection showed marginal significance (p=0.049). Notably, patients who transitioned from High to Low/Intermediate risk ("Responders") achieved survival rates comparable to stable low-risk cases, whereas "Persistent High Risk" patients had a 56% failure rate.
Conclusion
Follow-up biopsy is essential for dynamic risk re-stratification in ABMR. Identifying persistent high-risk histological phenotypes enables the detection of non-responders at imminent risk of graft loss, guiding personalized therapeutic interventions.
Figure 1.