An objective morphometric criterion for true versus pseudo-trifurcation of the celiac trunk, and its relationship with structures crossing the proximal trunk: a cadaveric study in a Vietnamese population

Background True versus pseudo-trifurcation of the celiac trunk (CT) is assigned by visual impression, without a numerical threshold, despite its surgical and endovascular relevance. We aimed to derive a candidate measurement-based criterion and to test whether a structure crossing the proximal trunk relates to branching pattern. Methods Thirty-six formalin-fixed cadavers were dissected. We measured the distance from the CT's aortic origin to its first branch (L1) and to complete separation of all three branches (L2), and their difference (ΔL = L2 − L1). Trifurcation was classified qualitatively before measurement and used as the reference standard for ROC analysis of ΔL. All dissections, classifications and measurements were performed once, by a single unblinded observer. The threshold, derived and evaluated in the same 27 trunks, was internally validated by bootstrap resampling (2000 replicates). Results Of 27 assessable trunks, 10 (37.0%) were true and 17 (63.0%) pseudo-trifurcations. ΔL gave an apparent AUC of 0.976 (cutoff >2.5 mm; sensitivity 100%, specificity 90%) and an optimism-corrected AUC of 0.976 (98.3%, 88.0%). Youden's J was maximal across a plateau (0.6–4.38 mm) of which 2.5 mm is only the midpoint; one specimen caused the entire loss of specificity. Variant branching (61.1%) and crossing structures (66.7%) were mutually independent (p > 0.05). Conclusions ΔL reproduced an independent qualitative classification with high derivation-sample accuracy, but remains a candidate criterion only: the threshold is poorly determined and was obtained by one unblinded observer without repeated measurements. Independent blinded confirmation and prospective validation against imaging are required before clinical use.