Submental flap for vascularized lymph node transfer; a CTA‐based study on lymph node distribution

Abstract Background Amongst various options of vascularized lymph node transfers, the submental flap has the lowest risk for iatrogenic lymphedema. The aim of this study was to gain insight into distribution, number, and size of lymph nodes along the mandible using computed tomography angiography (CTA). Methods A total of 52 CTA scans of head/neck region were evaluated retrospectively. Lymph nodes in the submental and submandibular region, related to the origin of the submental artery, were recorded using a three‐dimensional coordinate system, and standardized using an iterative closest point algorithm. Results were analyzed for gender, location, size, and number. Results The mean number and size of lymph nodes were 5.30 ± 2.00 and 5.28 ± 1.29 mm, respectively. The mean distance of the lymph nodes to the origin of the submental artery was 25.53 ± 15.27 mm. There was no significant difference between both sides when comparing size (left: 5.39 ± 1.28; right: 5.17 ± 1.34; P = .19), number (left: 5.46 ± 2.10; right: 5.17 ± 1.96; P = .49), and distance (left: 24.78 ± 12.23; right: 26.32 ± 14.73; P = .19). No significance was found between males and females concerning number (P = .60), size (P = .50), and distance (P = .06). Conclusion The variance of lymph node distribution along the mandible may warrant conducting a CTA scan to maximize the number of transferred lymph nodes and aid in flap design.

can only be performed in the early stages of lymphedema, as a superficial functional lymphatic vessel is necessary for a successful procedure and outcome. 2,3 Patients who are lacking suitable functional lymphatic vessels, necessary for an LVA, can be candidates for a vascularized lymph node transfer (VLNT). In this procedure, lymph nodes are transferred from a healthy donor site to the affected limb. VLNT aims to tackle the physiologic impairment by replacing the lost lymph nodes and potentially induce the regeneration of lymphatic vessels. To date, various VLNT donor sites have been described such as the groin flap, the supraclavicular flap, the submental flap, the gastroepiploic flap, the mesenteric and the thoracic lymph node flap. 2,[5][6][7][8][9][10][11][12] With this arsenal of free lymph node flaps available, the preferred flap for treating lymphedema of the extremity is the submental flap, due to a large number of lymph nodes and low risk for iatrogenic lymphedema. 2,[13][14][15] The improvement of lymphedema after the surgery is correlated to the number of lymph nodes transferred and their vascularization, hence imparting a meticulous flap design and harvest. 15,16 The imperative of including the maximal number of the lymph nodes with perinodal tissue is traded off against the increasing size of the flap, which, in turn, results in a larger visible submental scar. The submental lymph nodes, however, may all be located proximal to the origin of the submental artery, making the distal part of the flap skin redundant, hence departing from the rationale calling for a larger flap. It would, therefore, be beneficial to gain insight into the distribution of the lymph nodes in this area to determine optimal balance between inclusion of the most lymph nodes to be transferred and size of the flap.
The aim of this study was to investigate the distribution, number, and size of lymph nodes along the mandible using computed tomography angiography (CTA) to gain insight into the variability of the lymph nodes in this area.

| MATERIALS AND METHODS
In this retrospective study, patients who had undergone a head/neck CTA for a follow-up of a head and neck malignancy, or for staging of tumors located in other regions than head and neck, were included in the study. Head/neck scans were acquired in a helical mode in craniocaudal direction at 120 kV tube voltage where Iomeron 300 mg/mL (Bracco Imaging SpA, Milan, Italy) had been used as an intravenous contrast agent. Exclusion criteria for this radiologic study comprised of lymphadenopathy or lymph node dissection in the CTA report, insufficient image quality, presence of tumor, or metastatic disease or unclear origin of the submental artery.
In total 52 CTA scans (n = 23 male, n = 29 female) of satisfactory imaging quality were included to assess 100 sides (n = 48 bilateral, n = 4 unilateral) for lymph node anatomical features. The mean age of the group was 61.77 ± 13.78 years. The mean body mass index was 26.68 ± 4.86 kg/m 2 .

| Analysis of CTA scans
The submental artery branches of the facial artery and provides the essential blood supply to the submental flap. During the procedure of harvesting the submental flap, its origin from the facial artery acts as an anatomical landmark. Therefore, lymph nodes were recorded at each side using the origin of the submental artery as a central coordinate in the CTA scans ( Figure 1). Lymph nodes in the submental and submandibular region (IA and IB) 17 were recorded between the hyoid bone, the angles of the mandible, the mental protuberance, and the bellies of the digastric muscle. In each evaluated side the coordinates of the landmarks were determined; the ipsilateral and contralateral angle of the mandible and the chin. This process was F I G U R E 1 Axial slice of a head/neck CTA scan. The origin of the right submental artery is centered in the coordinate system. Two lymph nodes encircled in red. CTA, computed tomography angiography PAULUS ET AL.
| 1227 performed on all included CTA scans using post-processing software Aquarius iNtuition Viewer 4.4.12 (TeraRecon, Foster City, CA).
The locations of lymph nodes were recorded by a first observer (VP); validation was performed by randomly selecting 10% of the recorded scans and re-evaluated by a second observer (SH). We compared the number of nodes at the levels IA and IB strictly confined within predefined anatomical landmarks (hyoid bone, angles of the mandible, mental protuberance, and belly of the anterior digastric muscle). An observed agreement of 0.90 was calculated during this validation process.

| RESULTS
A total of 530 lymph nodes were located in 100 evaluated one-sided level IA/IB cervical regions. The mean number and size of lymph nodes were 5.30 ± 2.00 and 5.28 ± 1.29 mm, respectively. The mean distance from the lymph nodes to the origin of the submental artery was 25.53 ± 15.27 mm ( showed no significant difference (P = .43 and P = .19, respectively).
When comparing the number of lymph nodes between males and females (male: 5.20 ± 2.05; female: 5.41 ± 1.96), no significant difference was found (P = .60). The size of lymph nodes did not significantly differ either (male: 5.20 ± 1.13 mm, female: 5.38 ± 1.46 mm, P = .50), nor did the distance to the origin of submental artery (male: 26.56 ± 13.32 mm, female: 24.37 ± 13.68 mm, P = .06). These results are summarized in Table 2. Figure 2 depicts the distribution of lymph nodes along the mandible. The majority of the lymph nodes were approximately found between 1 and 4 cm from the origin of the submental artery as illustrated in Figure 3.
When comparing the number of lymph nodes between left and right side for each patient, 18.8% (n = 9) no significant difference was found in our study population. Most patients (50%, n = 24), showed a difference of one lymph node between either side. Albeit much less frequent, a difference of two nodes (20.8%, n = 10), three nodes (8.3%, n = 4), and four nodes (2.1%, n = 1) between the cervical sides was also found (Table 3).

| DISCUSSION
VLNTs have gained a momentum in treatment of early and end-stage lymphedema. 18 Following the VLNT, the lymphedema accumulates in the flap and its lymph nodes, followed by redistribution of the collected fluid into the bloodstream. 19 Although the exact working mechanism remains unknown, couple of studies, both in animals and humans, showed promising results. 2   For a future study, we would like to examine the benefits of the preoperative mapping of lymphatic vessels that are entering the flap carrying vascularized lymph nodes using indocyanine green fluorescence. In theory, an LVA between lymphatic vessels of the flap and lymphatic vessels at the recipient site may further expedite and improve the outcome of VLNT.

| CONCLUSION
A retrospective analysis of the CTA scans of the head and neck region was performed to study anatomical location, number, and size of submental lymph nodes in relation to the origin of the submental artery. It was found that the lymph nodes were randomly distributed along the mandible, both in the study population as well as within a single individual. The anatomical variation found in level IA/IB lymph nodes warrants conducting a CTA scan to maximize the number of transferred lymph nodes and aid in optimal submental flap design.