Impact of enhanced chest CT with three-dimensional reconstruction combined with preoperative olive oil administration on the incidence of postoperative chylothorax: a retrospective study

Abstract

Background

Chylothorax is a rare but potentially life-threatening complication after esophageal cancer resection. This study aims to investigate the impact of a strategy combining preoperative three-dimensional (3D) reconstruction of contrast-enhanced chest CT with olive oil ingestion on the incidence of chylothorax following esophageal cancer resection.

Methods

We retrospectively analyzed the clinical data of 357 patients who underwent esophageal cancer resection in the Department of Thoracic Surgery at Gansu Provincial People’s Hospital from January 2019 to June 2024. Among them, 179 patients who underwent preoperative 3D reconstruction of contrast-enhanced chest CT combined with oral olive oil administration were assigned to the 3D reconstruction group, while 158 patients who did not receive these interventions served as the control group. The primary outcome measures included the incidence of chylous leakage and related perioperative outcomes.

Results

All patients in both groups underwent surgeries successfully without any perioperative deaths. Compared to the control group, the 3D reconstruction group exhibited a significant decrease in the incidence of postoperative chylothorax (0% versus 8.2%, P < 0.001). The visualization rate of the thoracic duct during surgery was higher in the 3D reconstruction group compared to the control group (93.85% vs. 48.73%). Furthermore, patients in the 3D reconstruction group had less intraoperative bleeding (P = 0.003), shorter postoperative hospital stays (P = 0.003), and reduced chest drainage in the first three postoperative days (P < 0.001).

Conclusion

This study reveals that preoperative three-dimensional reconstruction of chest using enhanced CT allows for the anticipation of thoracic duct anatomy. Additionally, the combination of preoperative oral olive oil administration significantly enhances the intraoperative visualization of the thoracic duct. This integrated strategy effectively reduces the incidence of postoperative chylothorax, demonstrating promising clinical application prospects.

Introduction

Esophageal cancer is one of the most aggressive gastrointestinal malignancies, with a global 5-year overall survival rate ranging from 15 to 25%, making it the sixth leading cause of cancer-related deaths worldwide [1]. Esophagectomy, the primary surgical approach for esophageal cancer patients, is associated with major postoperative complications including pneumonia, recurrent laryngeal nerve paralysis, anastomotic leakage, and chylothorax, with the latter occurring in approximately 2.2% of cases [2]. The thoracic duct (TD) is involved in the drainage of about 75% of the body’s lymph, and chylothorax resulting from TD injury can lead to significant dehydration or respiratory failure if severe. Furthermore, persistent chylothorax may result in depletion of immunoglobulins or T-cells and is associated with a high risk of sepsis, posing a challenging clinical management scenario [3, 4].

To prevent chylothorax, some studies advocate for routine ligation of the supra-diaphragmatic TD during surgery. However, the efficacy of prophylactic TD ligation in preventing postoperative chylothorax remains controversial, as this invasive procedure may increase pressure and potentially cause TD rupture. Additionally, the long-term outcomes of prophylactic TD ligation in patients undergoing esophagectomy are unclear [5,6,7,8,9]. Studies have reported that preoperative oral administration of olive oil or dairy products can enhance intraoperative TD visualization and avoid TD injury. However, in our clinical practice, we have observed that it is still difficult to identify the TD intraoperatively when it is surrounded by thick fat or pleura [10, 11]. Oguma et al. proposed the use of preoperative MRI for TD imaging, which can effectively identify variant TDs and allow for preoperative planning of a reasonable surgical approach to avoid TD injury and reduce the risk of chylous leakage. Nevertheless, the routine application of preoperative MRI for TD imaging faces challenges due to cost-effectiveness considerations [12].

Personalized three-dimensional (3D) models are often used in clinical practice to assist with preoperative planning and intraoperative navigation. Interactive 3D models have been shown to significantly enhance surgeons’ spatial perception and reduce surgical complications by predicting anatomical structures preoperatively [13, 14]. Furthermore, 3D reconstructed images offer advantages such as low cost and ease of acquisition, making them more acceptable to patients. In our study, we investigated the impact of a strategy combining preoperative three-dimensional reconstruction of contrast-enhanced chest CT with oral olive oil administration on clinical outcomes such as the incidence of chylothorax after esophageal cancer resection, exploring the safety and clinical application value of this approach.

Patients and methods

Clinical information

This study retrospectively analyzed 357 patients who underwent esophageal cancer resection at Gansu Provincial People’s Hospital from January 2019 to June 2024.

Inclusion criteria were as follows: (1) Pathological diagnosis of esophageal squamous cell carcinoma; (2) Pathological staging of T1-T3 according to the 8th edition of the American Joint Committee on Cancer (AJCC), with no distant metastasis (M0) detected in preoperative examinations or intraoperative explorations, and clinical staging of I-III; (3) No history of esophageal or gastric surgery; (4) Complete clinical data (including comprehensive surgical records, pathological data, and postoperative follow-up data). Exclusion criteria were: (1) Patients unable to tolerate surgery due to advanced age, cardiopulmonary or hepatorenal insufficiency; (2) Patients who received preoperative radiotherapy or chemotherapy; (3) Patients with concurrent other malignancies; (4) Patients with a history of previous thoracic surgery.

Patients were divided into two groups based on whether they received preoperative three-dimensional (3D) reconstruction of contrast-enhanced chest CT combined with olive oil administration: the 3D reconstruction group (received 3D reconstruction of contrast-enhanced chest CT combined with olive oil administration) and the control group (did not receive 3D reconstruction of contrast-enhanced chest CT combined with olive oil administration). In the 3D reconstruction group, patients were instructed to orally consume 100 mL of olive oil before the contrast-enhanced chest CT examination and 8 h before surgery. Apart from this, other preoperative preparations were the same for both groups. This study was reviewed and approved by the Ethics Committee of Gansu Provincial People’s Hospital (Approval No. 2024 − 556). All patients provided written informed consent before surgery.

Chest enhanced CT scanning and three-dimensional reconstruction method

All patients underwent dynamic enhanced CT scanning using a GE LightSpeed VCT scanner and a dual-syringe high-pressure injector from Ulrich, Germany. The scanning range covered the chest and abdomen, and the scanning was typically completed 3–7 days before surgery. Patients orally consumed 100 mL of olive oil 8 h before the enhanced CT scan. Scanning parameters were as follows: tube voltage of 120 kVp, tube current automatically adjusted using automatic control technology, collimator width of 64 × 0.625 mm, X-ray tube rotation speed of 0.5 s/rotation, and pitch of 0.929. The contrast agent used was Iopamidol containing iodine (320 mg/mL, GE), with the dosage calculated based on the patient’s weight at 0.9 mL/kg. The contrast agent was injected at a rate of 3 mL/s, followed immediately by an injection of 20 mL of saline at the same rate. The three-dimensional reconstruction software, uAI Surgical Planning Portal-Thoracic, was used to store the reconstructed data and generate three-dimensional reconstructed images. Preoperatively, a thoracic surgeon and two radiologists analyzed and created the three-dimensional reconstructed images (as shown in Fig. 1) to clarify the anatomical relationships between the thoracic duct and surrounding tissues. We performed three-dimensional reconstruction of the esophagus, esophageal tumor, and thoracic duct before surgery, and labeled them with different colors. Based on the lesion location and tumor staging, the resection plane and area were manually designed. The operator simulated the surgery and developed the optimal plan based on the three-dimensional reconstructed images.

Fig. 1

Schematic diagram of chest enhanced CT three-dimensional reconstruction

Perioperative management

All surgical patients underwent operations under general anesthesia with endotracheal intubation. The surgeries were performed using three common approaches: left thoracotomy, Ivor-Lewis, and McKeown procedures. During the operations, thoracoscopic computer-assisted three-dimensional (3D) image reconstruction technology was utilized in the operating room. An assistant off the operating table, based on the surgeon’s requirements, rotated, displayed, or hid the 3D reconstructed images in real-time, providing the surgeon with detailed anatomical relationships of important structures such as the esophagus, esophageal tumor, and thoracic duct. Patients were anesthetized with a combination of intravenous and inhalation general anesthesia using a single-lumen endotracheal tube and positioned in the left lateral decubitus position. The thoracoscopic port (observation port) was established at the 7th intercostal space along the anterior axillary line of the right chest. Carbon dioxide was used to create an artificial pneumothorax with a pressure of 6–8 mmHg for thoracic manipulation. One operating port was placed 2 cm anterior to the inferior angle of the scapula at the 5th intercostal space, another at the 8th intercostal space, and a third 2 cm anterior to the anterior axillary line at the 3rd intercostal space. The tumor was explored to confirm its resectability, and then the umbilical vein arch was dissected and divided. During the operation, the thoracic duct, marked with olive oil, appeared crystal clear and full due to the presence of chyle (as shown in Fig. 2). The esophagus was dissected along its posterior wall towards the upper thoracic segment, with careful protection of the thoracic duct. Subsequently, the esophagus was dissected along the anterior wall space towards the right subclavian artery, and lymph nodes adjacent to the right recurrent laryngeal nerve were cleared. When separating the posterior wall of the middle and lower thoracic esophagus below the umbilical vein arch, the course of the thoracic duct between the umbilical vein and the aorta was first observed. The esophagus was then separated along the posterior wall of the thoracic aorta and dissected along its anterior aspect, with clearance of enlarged lymph nodes adjacent to the left recurrent laryngeal nerve and below the carina. Before completing the surgery, the thoracic duct was inspected for any milky chyle extravasation, and if present, the thoracic duct was ligated. After closing the chest, the patient was placed in the supine position, and the stomach was dissected either through an open or laparoscopic approach, preserving the right gastroepiploic vascular arch. A tubular stomach was created, and the left cervical esophagus and stomach were anastomosed using surgical instruments. A mediastinal drainage tube was routinely placed.

Fig. 2

Identification of the thoracic duct (TD)

Definition and postoperative management of chylothorax

The presence of ≥ 400 mL of milky, non-coagulating, and turbid fluid in chest drainage should raise suspicion for chylothorax. If the fluid becomes clear after adding ether and a subsequent chyle test is positive, the diagnosis can be confirmed. Patients initially receive conservative treatment, including a diet rich in high-quality protein, low in fat, and containing medium-chain triglycerides, or total parenteral nutrition if fasting is required. Surgical intervention is considered if conservative treatment fails after one week.

Statistical analysis

Statistical analysis was performed using SPSS 26.0 software. Measurement data are expressed as mean ± standard deviation (X̅±s), while count data are expressed as number of cases and percentages. Independent sample t-tests were used for comparisons between groups for measurement data, and chi-square tests or Fisher’s exact test were used for count data.

Results

Baseline pathological characteristics of the 3D reconstruction group and control group

The clinical and pathological background characteristics of the 3D reconstruction group and the control group are presented in Table 1. A total of 357 patients with esophageal cancer met the criteria and were included in the study, with 179 patients in the 3D reconstruction group and 158 patients in the control group. There were no significant differences between the two groups in terms of age (P = 0.683), gender (P = 0.42), smoking history (P = 0.079), BMI (P = 0.302), tumor location (P = 0.21), tumor size (P = 0.197), or pathological stage (P = 0.97). Patients in the 3D reconstruction group tolerated oral olive oil well, and no adverse gastrointestinal symptoms were reported.

Table 1 Patient characteristics

Surgical outcomes

The surgical data for both patient groups are presented in Table 2. No cases of chylothorax occurred in the 3D reconstruction group after surgery, while there were 13 cases of chylothorax in the control group, resulting in a significantly lower incidence of postoperative chylothorax in the 3D reconstruction group compared to the control group (0% vs. 8.2%, χ²=15.32, P < 0.001). The visualization rate of the thoracic duct during surgery was higher in the 3D reconstruction group compared to the control group (93.85% vs. 48.73%). The 3D reconstruction group also had shorter operative times (P = 0.008), less intraoperative bleeding (P = 0.003), shorter postoperative hospital stays (P = 0.003), and less chest drainage in the first three days after surgery (P < 0.001). There were no significant differences between the two groups in terms of anastomotic leakage (P = 0.88), anastomotic stricture (P = 0.33), dysphagia (P = 0.34), or avoidance of recurrent laryngeal nerve injury (P = 0.623). In the control group, chylothorax occurred due to thoracic duct injury during surgery. Six patients recovered after conservative treatment, including fasting, nutritional support, closed chest drainage, and intravenous administration of octreotide. Seven patients did not respond to conservative treatment and underwent reoperation through the original surgical incision for thoracic duct ligation.

Table 2 Comparison of clinical outcomes related to chylothorax between the 3D reconstruction group and the control group

Discussion

In our study, individualized surgical plans were formulated based on three-dimensional reconstruction of the thoracic duct using preoperative contrast-enhanced chest CT to understand its anatomical structure. Additionally, patients were instructed to orally consume olive oil 8 h before surgery to fill and enhance the visibility of the thoracic duct during the procedure. The results demonstrated that this strategy effectively avoided thoracic duct injury and reduced the incidence of chylothorax.

In the control group, 13 patients (8.2%) had chest drainage with a milky, non-coagulating, turbid fluid of ≥ 400 mL, and further chylous tests were positive, leading to a diagnosis of chylothorax. Studies have shown that contrast-enhanced CT can identify the location of chylous leakage and determine whether the chylous fluid in non-traumatic chylothorax originates from a tumor, providing valuable information for subsequent diagnostic and therapeutic strategies [15]. In our study, chylous leakage was controlled in 6 of the 13 patients with chylothorax through conservative treatment, while the remaining 7 patients underwent unplanned second surgeries. Conservative treatment principles include reducing chyle formation and promoting natural healing of the thoracic duct, with dietary recommendations of high-quality protein, low-fat, and medium-chain triglycerides, or total parenteral nutrition with fasting [16]. Intravenous octreotide was included in our conservative treatment regimen, which promotes spontaneous healing of the fistula by reducing triglyceride content in the chyle. Studies have reported that although octreotide may not immediately seal the fistula, its effects of reducing visceral blood flow and inhibiting intestinal triglyceride absorption are beneficial for thoracic duct wound healing, with a cure rate for traumatic chylous leakage of up to 38% [17]. Additionally, some studies have treated chylous leakage by intrathoracic injection of IL-2, which recruits inflammatory cells to form severe adhesions in the thorax, thereby sealing the damaged thoracic duct [18, 19].

In our study, for patients in the control group with chylothorax that was unresponsive to conservative treatment or had a daily chylous pleural effusion drainage volume > 1000 mL, the thoracic cavity was entered through the original surgical incision to thoroughly clean the chyle and locate and ligate the damaged thoracic duct. In 1998, Gregory et al. first proposed antegrade catheterization through the abdomen under lymphangiography, using a 15–20 cm Chiba needle to pierce the cisterna chyli and enter the thoracic duct, and then filling the duct with embolic coils or liquid embolics to treat chylous leakage. For some patients with larger body sizes or inconspicuous cisterna chyli, retrograde catheterization through the jugular vein can also be attempted. However, due to the technical difficulty and the lack of high-quality, multicenter clinical studies to verify its safety and clinical value, this method has not been widely adopted [20].

Since Lampson et al. [21] first reported the successful treatment of chylothorax with thoracic duct ligation, routine thoracic duct ligation has become an option for preventing postoperative chylothorax. However, the discussion around thoracic duct ligation has never ceased. Some studies suggest that not performing prophylactic thoracic duct ligation is significantly associated with the occurrence of postoperative chylothorax, and the incidence of chylothorax and secondary surgery rates are lower in the ligation group than in the non-ligation group [6]. A meta-analysis including 11 clinical studies found that the incidence of chylous leakage was similar between patients with routine thoracic duct ligation and those without, and patients without ligation had a higher 5-year survival rate [22]. As an invasive procedure, thoracic duct ligation can lead to increased free fatty acids in the liver tissue interstitium, resulting in worsened liver function [23]. Furthermore, a retrospective study of 230 esophageal cancer patients with type 2 diabetes found that thoracic duct ligation may exacerbate nutritional metabolic disorders in these patients, increase the incidence of pulmonary complications and anastomotic leakage, and suggested that thoracic duct ligation should be performed cautiously in such patients, and routine prophylactic ligation is not recommended [24].

Studies have reported reducing the incidence of chylothorax by orally consuming olive oil or dairy products before surgery [10, 11]. Their methods all emphasized intraoperative visualization of the thoracic duct. However, in our clinical practice, we found that when the thoracic duct is surrounded by thick adipose tissue or pleura, it remains difficult to identify during surgery. Similarly, to enhance thoracic duct visualization, Yang et al. administered indocyanine green intraperitoneally before surgery and observed the green fluorescent thoracic duct using near-infrared fluorescence imaging during surgery. However, indocyanine green is rapidly metabolized in the body, and the thoracic duct no longer fluoresces in some patients after about an hour [25, 26]. Tadashi et al. [27] identified variant thoracic ducts preoperatively using MRI thoracic duct imaging. Among the 228 patients enrolled, 7 (3%) had thoracic ducts on both sides of the thorax, posing a high risk of chylothorax. Preoperative MRI thoracic duct imaging allowed surgeons to comprehensively understand the anatomy and course of the thoracic duct, reducing the incidence of postoperative chylothorax. Considering cost-effectiveness, standardizing preoperative MRI thoracic duct imaging is challenging. CT imaging has potential advantages over MRI lymphangiography, including faster imaging speed, higher density resolution, broader applicability, and higher cost-effectiveness. Importantly, in resource-limited areas or medical institutions, CT scans are relatively low-cost and easy to maintain, effectively reducing medical costs and improving the accessibility and efficiency of medical services.

Previously, experienced surgeons mentally reconstructed the 3D structures from what they observed in 2D images, so that all details could be presented as they would appear during surgery. With the development of artificial intelligence technology, 3D models can simplify this reconstruction process and assist surgeons in a less technically demanding manner. Clinical studies have shown that preoperative 3D anatomical reconstruction models can provide comprehensive anatomical structures, facilitating the safe completion of Endoscopic Surgery [28]. Past research has found that 3D models can accurately predict the resection volume and margin in liver surgeries [29] and significantly reduce the abnormality rate of femoral anteversion restoration in hip replacement surgeries [30]. Hong et al. applied 3D reconstruction technology in lung segmentectomy and found that it significantly reduced the risk of accidental injury to intrapulmonary vessels and trachea, minimized unnecessary tissue dissection, and alleviated surgical trauma. Additionally, it markedly shortened the surgical duration [31].

In this study, the 3D reconstruction group not only significantly reduced the incidence of postoperative chylothorax but also demonstrated advantages in surgical duration, intraoperative bleeding, postoperative hospital stay, and chest drainage within three days after surgery. We utilized preoperative enhanced chest CT combined with olive oil consumption for 3D reconstruction. After olive oil was fully absorbed by the body, it enhanced the visualization of the thoracic duct under CT. Based on the 3D reconstruction model, the optimal surgical plan was devised for each patient preoperatively. During surgery, the assistant adjusted the rotation, display, or hiding status of the 3D reconstruction images in real-time according to the surgeon’s needs. Furthermore, the thoracic duct appeared distended due to the preoperative olive oil consumption, facilitating its identification. Surgical duration, one of our key indicators, was effectively reduced due to the rapid and accurate localization of lesions and vital organs, thereby avoiding prolonged surgery. Several published studies have observed results similar to ours [31, 32]. The reduction in surgical duration is closely related to decreased intraoperative bleeding. A shorter surgical duration reduces tissue exposure time, mitigating tissue damage and inflammatory responses caused by the surgical procedure. For patients, this means less trauma and faster postoperative recovery. These factors significantly contributed to the shorter postoperative hospital stay and reduced chest drainage in the 3D reconstruction group compared to the control group.

This study has some limitations. First, due to individual differences, olive oil absorption, and limitations in CT scanning technology, the delineation of the thoracic duct in 3D reconstruction was not always completely successful or encountered difficulties, limiting the number of participants. We optimized image quality by adjusting scanning parameters, changing patient positions, or increasing scan slice thickness to improve the accuracy of thoracic duct delineation as much as possible. Second, this study was a single-center study with a relatively small sample size (n = 357) and was retrospective, subject to potential distortions in medical record documentation and recall bias. Prospective, large-sample, multi-center studies are needed to validate our conclusions. Third, this study focused on short-term perioperative clinical outcomes. Therefore, further long-term follow-up evaluations are necessary to compare and analyze differences in long-term clinical outcomes between the two groups. Fourth, this study was conducted in China, and all participants were of Asian ethnicity, making it difficult to generalize the results to other populations.

In conclusion, the strategy of combining preoperative enhanced chest CT 3D reconstruction with preoperative olive oil consumption can effectively reduce the incidence of chylothorax after esophageal cancer resection, shorten surgical duration, decrease intraoperative bleeding and postoperative chest drainage, and shorten hospital stay, facilitating rapid postoperative recovery for patients. This method is also cost-effective and convenient, making it worthy of further promotion and application in clinical practice.