Background and ObjectivesDevelopment of vascularized submental lymph node (VSLN) flap has encountered dilemmas; (a) whether to include skin paddle, (b) how to reduce the harvest area while gaining most lymph nodes. To answer, these structures were studied; submental perforator, lymph nodes in neck‐level I and anterior belly of digastric muscle (ABDM).MethodsForty VSLN flaps were harvested from 23 cadavers. The lymph nodes and arterial supply were studied macro‐ and microscopically. The nodes were classified by arterial supplies, location along the longitudinal axis and relationship with ABDM.ResultsVSLN flap had 4.4 lymph nodes by average (range 1‐8) predominantly located in the posterior three‐quarter of the flap. Half of the submental perforators were originated deep to ABDM. they circumvent the muscle, supplied much of the nodes in neck sublevel Ia before reaching the skin. While sublevel Ib located the most surgically accessible submental nodes. Most of their arterial supply was branched from submental perforator lateral to ABDM, not directly from the submental artery.ConclusionThe flap could be reduced to the posterior three‐quarter of the original area. Skin paddle should be included to serve as an indirect lymph node monitor. If Ia lymph nodes are to be included, ABDM should be sacrified.

Background: The outcome of autologous lymph node (LN) transfer has depended on the number of LNs in the donor site. Unknown accuracy of the LN counting method has thrown some doubts on the reliability of the previous statistics. This study aimed to assess the accuracy of naked eye (NK) and stereo microscopy (SM) as tools for LN count. Methods: In total, 40 vascularized submental LN flaps were harvested from 23 fresh cadavers. The colored polymer was injected into the external carotid arteries before the harvest. LNs in each flap were counted by NK, SM, and histology in sequential order. Results: An estimated 175 LNs were confirmed, 4.4 ± 1.8 per flap. NK sensitivity was 33.7% compared with that of SM at 63.5%. Both methods missed all micro-lymph nodes (micro-LNs), contributing to 5.1% (9 nodes) of all LNs. Non-LN structures (647 negative counts) were composed of fat lobules, salivary gland lobules, and muscle fibers. NK specificity was 98.0%, compared with that of SM at 96.1%. SM showed a higher false positive rate at 14.3%, compared with NK at 7.4%. False positive counts were located mostly in Ib sublevel. Conclusions: NK and SM are imperfect tools for LN count due to poor sensitivity. If the method needs to be applied, points of considerations are (1) undetectable micro-LNs, (2) interposition of LNs with the digastric muscle and submandibular salivary gland, (3) confusion of LNs with lobules of salivary gland supplied by glandular artery or fat lobules supplied by lobular artery.