Introduction
Esophageal squamous cell carcinoma (ESCC) is a highly malignant disease characterized by aggressive tumor growth. Because the esophagus is centrally located in the mediastinum and anatomically adjacent to vital organs, tumor progression frequently results in invasion of surrounding structures, including the trachea, main bronchus, and aorta. According to the Japanese Esophageal Cancer Registry in 2016, 9.2% of patients were initially diagnosed with cT4b disease [
1]. In most such cases, intensive induction therapy—such as chemoradiotherapy and/or triplet chemotherapy—is indicated with curative intent. When sufficient downstaging is achieved and the tumor is considered potentially resectable, conversion surgery may be performed, provided the patient’s general condition allows. However, accurate intraoperative assessment of tumor resectability remains challenging, particularly at an early stage of the operation. As a result, combined resection of infiltrated adjacent organs may ultimately be required in some cases, and it has been reported that 2.8% of esophagectomies result in non-curative (R2) resection [
1].
When tumor infiltration does not extend to the tracheal or bronchial smooth muscle, the connective tissue surrounding the airway membranous portion may be resected en bloc with the tumor to secure an adequate surgical margin. In such cases, the resected membranous portion becomes thin and fragile due to loss of supporting structures, resulting in direct exposure to the post-esophagectomy cavity. This loss of structural support predisposes the airway to excessive respiration-dependent deformation and potential collapse. In contrast, in cases resulting in R2 resection, the ulcer bed of residual tumor—contaminated with bacterial flora from the oral cavity and pharynx—remains within the mediastinum and is exposed to postoperative exudate. This environment promotes bacterial proliferation and can lead to severe deep mediastinitis. Once any of these complications develops, they may be fatal and therefore must be avoided whenever possible. Accordingly, it is critically important to fill the posterior mediastinal dead space created after esophagectomy with viable tissue. This is necessary both to reinforce the fragile airway membranous portion after combined resection of peritracheal connective tissue and to prevent the spread of infection by covering residual tumor tissue in cases of R2 resection.
In esophageal reconstruction, the stomach is typically mobilized to the neck to restore alimentary continuity, together with the greater omentum. When posterior mediastinal dead space requires filling, the gastric conduit is usually elevated via the posterior mediastinal route. However, because priority is given to packing the central posterior mediastinal cavity with the greater omentum, the gastric conduit is often displaced dorsomedially into the right thoracic cavity and remains subject to intrathoracic negative pressure [
2]. As a result, dilation and/or displacement of the gastric conduit may occur, leading to impaired transit and delayed gastric emptying [
3,
4].
To overcome these issues, we developed a novel procedure in which the greater omentum is elevated together with the gastric conduit via the retrosternal route and subsequently pulled down from the anterior mediastinum to fill the posterior mediastinal dead space. In this study, we report the feasibility, safety, and clinical usefulness of this technique, which we have termed the “waterfall” method.
Discussion
We developed a novel surgical technique, termed the “waterfall method”, to fill the posterior mediastinal dead space created after esophagectomy by transposing the greater omentum via the retrosternal route rather than the conventional posterior mediastinal route. In this procedure, the omental flap is elevated to the neck together with the gastric conduit and subsequently pulled down from the anterior mediastinum to fill the posterior mediastinal cavity. We refer to this technique as the “waterfall method” because the greater omentum naturally drapes downward along the gastric conduit, covering the posterior mediastinal space in a manner resembling a waterfall. This technique was applied in 20 selected patients who required combined resection of infiltrated tissues or who underwent R2 resection. No serious postoperative complications—such as tracheobronchial necrosis, airway collapse, or exacerbation of mediastinal infection—were observed, suggesting that the greater omentum may function effectively as a reinforcing and space-filling tissue.
After esophagectomy, resection of the surrounding peritracheobronchial tissues may result in loss of structural support for the membranous portion of the airway, leaving it exposed over the post-esophagectomy cavity. Because this cavity can change in volume with respiration, the unsupported membranous portion may be subjected to dynamic deformation driven by airway pressure fluctuations. In the absence of normal connective tissue restraint, the thin residual layer composed of mucosa and smooth muscle may bulge inward during inspiration and distend outward during expiration, particularly during forceful respiration or coughing. Such repetitive deformation could potentially compromise microvascular perfusion of the membranous portion, thereby increasing the risk of ischemia and subsequent airway–mediastinal fistula formation. The tracheal mucosa is known to be vulnerable to ischemic injury caused by mechanical compression, as demonstrated in studies of cuff-related tracheal injury during prolonged intubation. In such settings, mucosal ischemia may progress to tracheoesophageal fistula formation, as previously reported [
14]. In cases of R2 resection, if a contaminated ulcer bed of residual tumor remains within the mediastinum, accumulation of postoperative exudate promotes bacterial proliferation, placing the patient at high risk for severe mediastinitis. In both scenarios, it is essential to obliterate the dead space after esophagectomy with well-vascularized tissue to reinforce the airway and eliminate a potential nidus for infection.
An ideal tissue for filling the dead space should have abundant blood supply, resistance to infection, and sufficient volume to pack the cavity. Although muscles such as the latissimus dorsi or pectoralis major have been used for this purpose, the greater omentum offers several advantages, including rich vascularity, immunologic activity due to abundant milky spots [
15], angiogenic potential [
16], and absorptive capacity for excess fluid [
17]. These properties have led to its use in the treatment of empyema and mediastinitis, reinforcement of bronchial stumps after pulmonary resection [
18,
19], and prevention of anastomotic leakage after esophagectomy [
20,
21]. In addition, the greater omentum produces vascular endothelial growth factor [
22], facilitating revascularization and tissue healing [
23,
24]. Given these characteristics, the greater omentum is particularly well suited for reinforcement and infection control in the posterior mediastinum.
Although the dead space after esophagectomy is located in the posterior mediastinum, the high mobility of the omental flap allows transposition from the anterior mediastinum even when the gastric conduit is reconstructed via the retrosternal route. Our waterfall method successfully achieved this in all patients without serious complications such as airway fistula, tracheobronchial stenosis due to compromised blood flow, or deep mediastinitis. A major advantage of this technique over the conventional posterior mediastinal approach is that the gastric conduit is positioned centrally within the anterior mediastinum and outside the thoracic cavity. This configuration protects the conduit from intrathoracic negative pressure, preventing dilation, displacement, and regurgitation, thereby ensuring favorable postoperative transit. In addition, the conduit remains uninvolved in the event of local recurrence or regrowth of residual tumor, and postoperative radiotherapy can be administered without concern for conduit positioning. Because of its technical simplicity, this procedure is well suited for minimally invasive esophagectomy, including thoracoscopic or robotic approaches.
Good candidates for this technique include patients who are at high risk of developing gastro-airway fistula, such as those who have undergone resection of the connective tissue surrounding the membranous portion of the airway, those with a bronchial stump after simultaneous lung lobectomy, or those with contaminated residual tumor ulcer beds that may promote mediastinal infection and subsequent fistula formation. In the present series, no cases of airway–mediastinal fistula were observed, although the relatively small number of patients precludes definitive conclusions.
Several technical points are essential for successful application of this method. First, a notch should be created on the right side of the greater omentum by dividing the rightmost and second epiploic vessels to form a sickle-shaped flap (Fig.
1a), thereby maximizing reach into the thoracic cavity (Figs.
2d and
3). Second, the tip of the omental flap should be fixed with a thread at least 15 cm distal to the tip of the gastric conduit to ensure correct positioning in the anterior mediastinum during elevation of the gastric conduit (Fig.
1b). If the fixation point is too proximal, the omental flap may become impacted in the narrow thoracic inlet together with the gastric conduit, resulting in considerable difficulty in pulling the omental flap downward. This represents an important technical point to avoid during the procedure. Third, the omental flap should be placed to the right, passing in front of the gastric conduit, to facilitate smooth transposition. However, because the greater omentum is originally attached to the greater curvature of the stomach on the left side, a theoretical concern exists regarding the potential risk of gastric conduit torsion. In practice, the gastric conduit is stabilized within the retrosternal space, and only the omental flap located on the right side of the conduit is gently drawn down into the right thoracic cavity. Because excessive traction is not required, this procedure is considered unlikely to cause gastric conduit torsion, and no such complication was observed in the present series. Fourth, creation of the retrosternal tunnel and the pleural incision in the anterior mediastinum should be performed carefully and kept as small as possible to minimize the influence of intrathoracic negative pressure (Fig.
2a). Finally, even in lean patients, the volume of the greater omentum is usually sufficient to fill the posterior mediastinal dead space after esophagectomy. However, if the omental volume is insufficient, alternative approaches such as the use of a pedicled muscle flap may be necessary.
This study has several limitations, including its retrospective, single-center design, small sample size, and lack of objective quantitative assessment of delayed gastric emptying. In particular, objective functional or morphological assessment of the gastric conduit would be desirable in future studies. Moreover, it is ethically difficult to determine whether this technique definitively prevents serious postoperative complications compared with no dead space filling or the conventional method using posterior mediastinal reconstruction. This study is a retrospective single-arm experience, and the absence of a control group may introduce potential bias in evaluating the true efficacy of the technique. Furthermore, the target population is relatively rare, and severe complications such as tracheal membranous rupture are often life-threatening, making controlled comparative studies challenging in clinical practice. In recent years, retrosternal reconstruction has been increasingly adopted because of its potential advantages, including a lower risk of gastro-tracheal fistula, reduced reflux, and easier creation of a gastrostomy. Therefore, rather than switching to posterior mediastinal reconstruction only when omental filling is required, it may be advantageous to have a technique that allows omental filling even during standard retrosternal reconstruction.
In conclusion, the waterfall method is a simple, feasible, and potentially safe technique that allows effective filling of the posterior mediastinal dead space while preserving optimal positioning and function of the gastric conduit. This approach may represent a valuable option for surgeons managing complex cases of esophageal cancer requiring combined resection or non-curative surgery. Further clinical experience will help to better define the indications and clinical utility of this technique.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.