Outcomes of Free Flap Surgery in Head and Neck Cancer Patients With Chronic Kidney Disease: A Single-Institution Retrospective Study of 61 Cases.

Introduction The number of head and neck cancer patients with chronic kidney disease (CKD) is increasing. Although free flap reconstruction has become more feasible with advances in renal replacement therapy, perioperative risks in CKD remain uncertain. Prior reports are limited and inconclusive. We evaluated free flap outcomes in CKD patients undergoing head and neck reconstruction. Materials and methods We performed a single-institution retrospective analysis of 61 patients with CKD, operationalized by preoperative estimated glomerular filtration rate (eGFR) within the presurgical window (<60 mL/min/1.73 m²), who underwent free flap reconstruction (2011-2022). The primary outcome was flap complications (postoperative thrombosis requiring thrombectomy or flap necrosis). Failure cases underwent histopathologic review of anastomosed vessels. A prespecified descriptive comparison contrasted flap events by diabetes mellitus (DM) status. Statistical analysis used Fisher's exact test; owing to only five events, no multivariable models were fitted. Postoperative continuous hemodiafiltration (CHDF) use was recorded. Results The most common eGFR category was G3a (65.1%), followed by G3b, G5, and G4. Flap complications occurred in 5/61 (8.2%) patients: necrosis (3/61; 4.9%) and thrombosis salvaged by thrombectomy (2/61; 3.3%). All events arose in G3a; no events were seen in G3b-G5 or among CHDF cases. Histopathology showed severe occlusive change in one necrosis case (~80% luminal stenosis), while other failures had at most mild intimal thickening; two failures occurred without diabetes mellitus (DM)/cardiovascular disease or evident vascular pathology. DM was present in 14/61 (23%) patients; event proportion was 14.3% with DM vs. 6.4% without DM (Fisher p = 0.32), a non-significant difference consistent in direction with prior reports. Discussion Events were infrequent and clustered in G3a; the absence of events in higher categories is not interpretable, given very small subgroup sizes and limited power. Vascular pathology among failures was heterogeneous, indicating that overt vasculopathy or glycemic status alone does not fully explain failure. Careful perioperative management may support flap viability, but causal inference is unwarranted. Conclusion Free flap reconstruction appears feasible in selected CKD patients with optimized perioperative care; however, findings are hypothesis-generating and require confirmation in larger controlled cohorts with standardized operative reporting and adjustment for confounding.