Treatment versus observation for borderline T-Cell–mediated rejection in kidney transplantation: a systematic review and meta-analysis with trial sequential analysis

Abstract Background Kidney transplantation markedly improves survival and quality of life in end-stage kidney disease, and advances in immunosuppression have reduced acute rejection. Nonetheless, long-term graft survival has not improved proportionally, partly attributed to subclinical inflammation - “borderline” T-cell–mediated rejection (TCMR)- characterized by modest interstitial inflammation with tubulitis. Evidence on the prognostic impact and optimal management of borderline TCMR is conflicting. To address this uncertainty, we performed a systematic review and meta-analysis comparing renal function and survival outcomes between treated and untreated patients with borderline rejection after kidney transplantation. Methods A comprehensive literature search was conducted across various databases until September 2025 to identify relevant studies. The quality was assessed using the Newcastle–Ottawa Scale (NOS) tool and the analysis was performed using RevMan 5 software. Results The pooled results showed higher 1-month serum creatinine with treatment versus observation at 1 month post-biopsy (SMD = 0.40, 95% CI 0.15–0.65; favoring no treatment), while no differences were seen at 6 months (SMD = 0.10, 95% CI − 0.13 to 0.33) or 12 months (SMD = 0.07, 95% CI − 0.20 to 0.34) after diagnosis. Estimated glomerular filtration rate showed no difference at 6 months (SMD = − 0.16, − 0.45 to 0.12); at 12 months, eGFR favored the treated group (MD = 3.44 mL/min/1.73 m², 0.86–6.02). Death-censored graft survival did not differ between the two groups, whereas acute rejection was more frequent in treated patients. Heterogeneity was low to negligible across analyses. Leave-one-out tests indicated that the eGFR findings (6 and 12 months) and the acute-rejection signal were sensitive to single-study omission, while serum creatinine (all time points) and graft survival were robust. Trial-sequential analysis for the primary endpoint did not reach the required information size or monitoring boundaries, indicating current evidence trends toward no-treatment benefit but remains insufficient for definitive conclusions. Conclusion Treatment of borderline rejection was associated with a transient increase in serum creatinine at 1 month compared to non-treated patients, with no improvement in renal function (eGFR) at 6 months. Although the pooled 12-month eGFR estimate favored treatment, this finding was driven predominantly by Min 2012 (89.8% weight) and lost statistical significance when that study was omitted, indicating that the result was not robust. While treatment of borderline rejection identified on protocol biopsies showed no functional benefit and potential for increased fibrosis, its impact on indication-based biopsies remains inconclusive. The pooled data suggest that clinical context is a primary determinant of treatment response, and a universal treatment strategy for all borderline changes is not supported by current evidence.