Association of perioperative cholesterol depletion with anastomotic leak after esophagectomy for cancer: a retrospective study

Introduction

Esophageal cancer ranks as the seventh most common malignancy globally, with over 470,000 new cases diagnosed each year (1). Despite advancements in the management of patients with esophageal cancer, surgical resection remains the cornerstone of treatment for esophageal cancer, particularly in cases of resectable disease (2). To reconstruct the upper digestive tract following esophagectomy, the gastric tube is widely employed as the first choice for an esophageal substitute, which is known as gastric pull-up method. However, this technique can sometimes result in postoperative anastomotic leak (AL), one of the most frequent and serious complications after esophageal cancer surgery. While AL poses a significant threat to patients’ lives, it is also associated with increased medical costs, delays in subsequent treatments, and a decline in quality of life.

Malnutrition is a well-known risk factor for adverse outcomes in patients with cancer and esophageal cancer patients frequently present with dysphagia as well as associated nutritional compromise. Shimura et al. demonstrated that indicators of malnutrition, such as hypoalbuminemia, elevated Patient-Generated Subjective Global Assessment (PG-SGA) scores, and renal insufficiency, have been shown to predict AL in patients undergoing gastrointestinal surgeries (3-5).

Beyond conventional nutritional assessment tools, quantitative nutritional indices such as the Prognostic Nutritional Index (PNI), Geriatric Nutritional Risk Index (GNRI), and Controlling Nutritional Status (CONUT) score (a higher score indicates worse nutritional status) have been developed to objectively evaluate malnutrition risk and predict adverse clinical outcomes (6). Hence, this study aimed to evaluate the relationship between these three nutritional indices and the incidence of AL, while further exploring the potential underlying mechanisms of AL development. We present this article in accordance with the STROBE reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1665/rc).

Methods

Patient criteria

This retrospective study utilized our institutional database of esophageal cancer patients treated in the Department of Thoracic Surgery at The Affiliated Cancer Hospital of Nanjing Medical University between August 2018 and January 2023. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Jiangsu Cancer Hospital (No. 2023-ke-quick103), and individual consent for this retrospective analysis was waived. Eligible participants included patients with pathologically confirmed esophageal malignancy who underwent esophagectomy during hospitalization, had complete postoperative laboratory data, and received esophagogastric anastomosis. We excluded patients with significant missing data, those with synchronous esophagogastric double primary cancers, and individuals who underwent colon-esophageal anastomosis.

Between August 2018 and January 2023, a total of 3,670 patients underwent esophagectomy in the Department of Thoracic Surgery, The Affiliated Cancer Hospital of Nanjing Medical University. After applying exclusion criteria, 1,552 (42.3%) cases were excluded due to missing critical laboratory data, 172 (4.7%) for metachronous/synchronous primary malignancies or undergoing colon-esophageal anastomosis, and 197 (5.4%) with undeterminable AL status during hospitalization. The final analytic cohort comprised 1,749 (47.7% of original population) eligible patients. For further details, see Figure 1.

Figure 1 Flowchart of patient inclusion.

Study variables

We retrospectively collected comprehensive clinical data, including baseline characteristics, diabetes mellitus history, and neoadjuvant/conversion therapy status. Detailed tumor characteristics were documented, encompassing anastomotic site, histopathological type, and TNM staging according to The TNM Classification of Malignant Tumors, 8th edition. Additionally, we obtained nutritional and metabolic parameters, including serum albumin levels (g/dL), total lymphocyte counts (/mm³), and total cholesterol concentrations (mg/dL), as well as high-density lipoprotein (HDL, mmol/L), low-density lipoprotein (LDL, mmol/L), triglycerides (TG, mmol/L), total bilirubin (TB, µmol/L), conjugated bilirubin (CB, µmol/L), and alanine aminotransferase (ALT, U/L) at both preoperative (typically within 1 week before surgery) baseline and on postoperative day 1 (POD1) at both preoperative baseline and POD1 timepoints. For each patient, we calculated three nutritional indices (PNI, GNRI, and CONUT scores) using preoperative and POD1 laboratory values, and determined their perioperative changes (Δ=postoperative−preoperative). The calculation formulas were as follows: PNI=[10×serumalbumin(g/dL)]+[0.005×totallymphocytecount(/mm3)] (7). GNRI=[1.489×albumin(g/dL)]+{41.7×[currentweight(kg)/idealweight(kg)]} (8).

Ideal weight:

Male: [height(cm)−100]−[(height(cm)−150)/4](kg)

Female: [height(cm)−100]−[(height(cm)−150)/2.5](kg)

CONUT=albuminsubscore+totallymphocytecountsubscore+totalcholesterolsubscore. Detailed scoring criteria for each component are presented in Table 1 (9).

Diagnostic criteria

AL was defined as a full-thickness gastrointestinal defect involving the esophagus, anastomotic site, suture line, or gastric conduit, regardless of clinical presentation or diagnostic method (10). Anastomotic integrity was assessed using standardized imaging protocols: either chest computed tomography (CT) or upper gastrointestinal series, with AL defined as radiographic evidence of contrast extravasation or perianastomotic air-fluid collections. Those without radiologically confirmed AL but with documented surgical evidence (e.g., wound exploration revealing a definitive anastomotic defect) or clinical findings (e.g., enteric content in drainage fluid) were also classified into the AL group. We graded the severity of AL according to the Esophagectomy Complications Consensus Group (ECCG) criteria (10).

• Type I: a localized defect requiring no change in therapeutic regimen. This includes leaks managed without pharmacologic treatment, dietary adjustments, or procedural interventions.
• Type II: a localized defect requiring interventional treatment but not surgical re-operation. Examples include percutaneous or endoscopic drainage, stent placement, or wound debridement at the bedside.
• Type III: a localized defect requiring surgical re-operation for management.

Statistical analysis

Continuous variables, including patient demographics [age, body mass index (BMI)] and nutritional assessment scores, were evaluated for normal distribution using the Kolmogorov-Smirnov test. Normally distributed parameters were analyzed with Student’s t-test. Categorical baseline characteristics were compared using Pearson’s chi-square test or Fisher’s exact test, with results presented as frequencies (n) and percentages (%).

Multivariable analysis was performed through logistic regression modeling to determine adjusted odds ratios (ORs) with 95% confidence intervals (CIs). All statistical tests were two-tailed, with a threshold of P<0.05 for statistical significance.

The data analysis and graphical representations were performed using Microsoft Excel (version 2019), SPSS Statistics (version 22.0), and GraphPad Prism (version 8.0.2).

Results

Patient characteristics

A comparative analysis of baseline characteristics was first performed between the included and excluded patients. No statistically significant differences were identified in terms of gender, age, BMI, diabetes mellitus, anastomotic site, or preoperative therapy. For more details, see Table 2. This study enrolled 1,749 patients [1,388 males (79.4%) and 361 females (20.6%)] with a mean age of 66.0±7.2 years and a mean BMI of 23.2±3.0 kg/m². The distribution of anastomotic sites was as follows: cervical region (n=504, 28.8%), supra-aortic (n=918, 52.5%), and subaortic (n=327, 18.7%). Neoadjuvant or conversion therapy had been administered to 312 patients (17.8%) and 150 (8.6%) had comorbid diabetes mellitus. Pathological staging revealed 999 cases (57.1%) in stages I–II and 750 (42.9%) in stages III–IV, with squamous cell carcinoma predominating (1,392 cases, 79.6%). Postoperative assessment via CT and gastrointestinal contrast studies identified AL in 145 patients (8.3%), with significant between-group differences observed in age (P=0.02), histopathological type (P=0.04) and anastomotic site (P<0.001) when comparing leak versus non-leak groups. For further details, see Table 3.

Cholesterol depletion can serve as an independent risk factor

We further investigated the association between perioperative nutritional status changes and AL. Preoperatively, no statistically significant differences were observed in any of the three nutritional indices between patients who subsequently developed AL and those who did not (all P>0.05), suggesting comparable baseline nutritional status between the two groups. Following esophagectomy, differential changes in nutritional parameters emerged between groups. Notably, the CONUT score demonstrated a significantly greater increase in the AL group (Δ3.9±2.2) compared to the non-AL group (Δ3.4±2.0, P=0.01). In contrast, the other two nutritional indices (PNI and GNRI) showed no significant intergroup differences in their postoperative changes (both P>0.05). A detailed analysis of the three biochemical components underlying the CONUT score revealed distinct perioperative patterns between groups. POD1 measurements demonstrated a selective metabolic divergence, while Δalbumin (−0.9±0.5 vs. −0.8±0.5 g/dL, P=0.37) and Δlymphocyte counts (−1,008.5±478.3 vs. −1,000.1±548.9 /mm³, P=0.86) were similarly in both AL and non-AL groups, the AL group showed stronger cholesterol depletion (−41.3±30.6 mg/dL) compared to non-leak patients (−33.6±26.0 mg/dL, P=0.001). More details can be found in Table 4. A multivariable-adjusted logistic model incorporating all significant univariate predictors revealed that age (per year increase: OR =1.03, 95% CI: 1.006–1.06), cholesterol depletion (per unit decrease: OR =1.01, 95% CI: 1.003–1.02), and cervical anastomosis (vs. subaortic/supra-aortic: OR =2.41, 95% CI: 1.68–3.46) remained independently associated with AL risk after mutual adjustment. For more details, see Figure 2. We further investigated the association between the severity of AL and the degree of cholesterol depletion. Among the 145 patients with AL, 25, 112, and 8 were classified as ECCG grade I, II, and III, respectively. Although no statistically significant difference in cholesterol depletion was found between grade I and the combined grades II/III, a trend was observed, with grades II/III exhibiting a greater reduction than grade I (−41.8±31.3 vs. −38.7±27.8 mg/dL, P=0.65).

Figure 2 Multivariable logistic regression analysis. CI, confidence interval; OR, odds ratio.

Acute cholesterol depletion independent of hepatic injury predicts AL

Since very-low-density lipoprotein (VLDL) cannot be measured directly but is strongly correlated with TG levels, we used TG as a surrogate for VLDL. Accordingly, we analyzed the postoperative changes in HDL, LDL, and VLDL (represented by TG). Among the components of total cholesterol, the depletion of LDL was significantly more pronounced in patients who developed AL (ΔLDL: −0.9±0.7 mmol/L) compared to those who did not (ΔLDL: −0.7±0.6 mmol/L; P<0.001). In contrast, changes in HDL and TG (representing VLDL metabolism) were comparable between the two groups (P>0.99 and P=0.89, respectively). Furthermore, key markers of liver function and injury, including TB, CB, and ALT, showed no significant intergroup differences (all P>0.05). For more details, see Table 5. This pattern of isolated cholesterol depletion, in the absence of significant hepatic injury, suggests that the observed dyslipidemia is more likely a consequence of a systemic inflammatory and metabolic stress response rather than impaired liver function. To objectively evaluate the stress-AL relationship, we employed the validated Surgical Stress Score (11), computed as:

SurgicalStressScore=−0.342+0.0139×(bloodloss/weightinmL/kg)+0.0392×(operativetimeinhours)+0.352×(incisionclass)[1]

where incision class was defined as:

• 0: minimally invasive approaches (laparoscopic/thoracoscopic, including hybrid procedures);
• 1: open laparotomy or thoracotomy;
• 2: combined thoracoabdominal access.

However, the Surgical Stress Score failed to demonstrate significant discrimination between leak and non-leak groups (P=0.16).

Discussion

Our multivariate analysis identified three independent risk factors for AL: advanced age, cervical anastomosis, and significant cholesterol depletion. These findings align with existing literature and provide new clinical insights. First, the association between advanced age and AL risk is corroborated by analysis of the Society of Thoracic Surgeons (STS) database (n=4,321 esophagectomies, 2011–2014), which identified age >65 years as a significant predictor of postoperative complications (12). Second, our demonstration of cervical anastomosis as an independent risk factor is strongly supported by evidence from van Workum et al.’s randomized controlled trial, which reported significantly higher leak rates with cervical versus intrathoracic anastomoses (34.1% vs. 12.3%, P<0.001) (13,14). This 2-fold increased risk likely reflects the anatomical challenges of cervical anastomoses, including greater anastomotic tension and compromised blood supply to the proximal esophageal stump.

Notably, neoadjuvant therapy showed no significant association with AL risk, aligning with multicenter prospective data reporting comparable leak rates between chemoimmunotherapy (8.5%) and chemoradiation (8.7%) cohorts (15).

The statistically significant intergroup difference observed specifically in cholesterol levels (P=0.001), but not in other CONUT components, suggests a potential unique role of lipid metabolism in AL development. It has been revealed that serum total cholesterol levels undergo an immediate postoperative decline, beginning within the first hour after trauma and persisting for approximately 1 week (16). The study by Asteriou et al. found that patients undergoing thoracoscopic surgery experience the peak of their stress response at 4 hours postoperatively, characterized by elevated cortisol secretion, which typically returns to baseline within 48 hours (17). This biochemical response has been clinically validated as a significant predictor of adverse outcomes, including prolonged mechanical ventilation duration, elevated infection risks, and extended hospitalization periods (18). Particularly in cardiothoracic surgical patients, sustained severe hypocholesterolemia (total cholesterol <2.5 mmol/L) has emerged as a robust marker of poor prognosis (19). The majority of clinical studies have identified prolonged corticosteroid use as an independent risk factor for AL following colorectal surgery (20,21). The underlying pathophysiology appears related to surgical stress-induced activation of the hypothalamic-pituitary-adrenal axis, leading to increased cortisol secretion and subsequent accelerated cholesterol consumption. This mechanistic understanding supports our observation of significantly greater cholesterol reduction in AL patients, suggesting an exaggerated stress response in this cohort. The study by Fan et al. identified elevated postoperative levels of total cholesterol and LDL, along with decreased HDL, as risk factors for AL (22). In contrast, our study found that a rapid decline in total cholesterol and LDL from the preoperative baseline to POD1 was associated with leak risk. This apparent discrepancy can be reconciled by the distinct focus of each study: our work captures the dynamic, acute depletion of lipids in response to surgical stress, whereas Fan et al. focused on static postoperative levels. Critically, both studies converge on the same fundamental conclusion—that dysregulated lipid metabolism is implicated in the pathogenesis of AL. While both albumin and lymphocyte count are indeed recognized markers of the stress response, Albumin’s long half-life (approximately 2–3 weeks) means its serum concentration does not decline significantly within the first postoperative day. Regarding lymphocytes, their count is exquisitely sensitive to surgical stress, plummeting rapidly in all patients undergoing major surgery like esophagectomy. While sensitive, this ubiquitous and profound drop may diminish its power to discriminate between patients with varying stress intensities.

We compared the extent of postoperative cholesterol depletion on POD1 across different ECCG grades of AL. Although the difference did not reach statistical significance, the mean cholesterol reduction was numerically greater in patients with grade II/III leaks compared to those with grade I leaks. The small cohort of grade I leaks (n=25) in our study may have underpowered this specific analysis. Future studies with larger sample sizes are warranted to confirm this observed trend. The Surgical Stress Score failed to demonstrate significant discrimination between AL and non-AL groups (P>0.05). This limited sensitivity likely stems from the fact that the Surgical Stress Score, a composite of macroscopic surgical parameters, offers an incomplete picture of the physiological stress response. It cannot fully capture the nuances of anesthesia, postoperative pain, and hemodynamic management, all of which profoundly influence surgical stress. Therefore, the acute depletion of cholesterol, as a direct biochemical readout of the body’s metabolic state, may serve as a more sensitive and integrative biomarker of the true physiological burden experienced by the patient.

This study primarily examines the association between perioperative nutritional status fluctuations and AL incidence. By analyzing nutritional indicator changes from preoperative assessment to POD1, our design specifically captures the surgical stress impact on nutritional parameters. While postoperative nutritional interventions typically require more time to demonstrate measurable effects, their exclusion in this immediate postoperative evaluation framework does not compromise the validity of our findings. This study did not incorporate surgical approach (open or minimally invasive) as a variable, existing literature consistently demonstrates no significant difference in AL rates between these techniques (23).

There are several limitations in this study. First, to ensure data accuracy for POD1 laboratory values, we included only 1,749 patients with complete data from an initial cohort of 3,670. Although a comparison of baseline characteristics between the included and excluded patients revealed no statistically significant differences, the potential for selection bias cannot be entirely ruled out. Second, due to the retrospective nature of this study, the observed association between the change in total cholesterol and AL does not establish causality. Third, a decrease in total cholesterol was identified as an independent risk factor for AL. However, the low OR (1.01 per unit decrease) indicates that, despite statistical significance, its contribution to predictive models is likely limited. Furthermore, data on preoperative lipid-lowering medication use were not collected, and the potential impact of other postoperative complications was not considered in our analysis. These omissions may have introduced confounding effects on the results. To address these limitations, we plan to conduct prospective studies that will include serial measurements of biomarkers such as C-reactive protein (CRP) and cortisol. Furthermore, animal experiments will be designed to rigorously establish the causal relationship between perioperative cholesterol metabolism and anastomotic healing. Our findings also suggest a plausible intervention: mitigating surgical stress through optimized postoperative pain and blood pressure management may consequently reduce the incidence of AL. This potential therapeutic approach warrants further investigation.

Conclusions

Our study is the first to identify that a severe cholesterol depletion on POD1 serves as an independent risk factor for AL after esophagectomy. This finding enables clinicians to early identify high-risk patients for targeted monitoring and intervention and also suggests a potential link to enhanced surgical stress response. We plan to conduct additional investigations to test this mechanistic hypothesis.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the STROBE reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1665/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1665/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1665/prf

Funding: This study was supported by Jiangsu Province Capability Improvement Project through Science, Technology and Education, Jiangsu Provincial Medical Key Laboratory (No. ZDXYS202203), Jiangsu Provincial Medical Innovation Center (No. CXZX202224), Yishan Research Project of Jiangsu Cancer Hospital (No. YSZD202403).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-1665/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of Jiangsu Cancer Hospital (No. 2023-ke-quick103) and individual consent for this retrospective analysis was waived.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Deboever N, Jones CM, Yamashita K, et al. Advances in diagnosis and management of cancer of the esophagus. BMJ 2024;385:e074962.
2. Borggreve AS, Kingma BF, Domrachev SA, et al. Surgical treatment of esophageal cancer in the era of multimodality management. Ann N Y Acad Sci 2018;1434:192-209.
3. Shimura T, Toiyama Y, Hiro J, et al. Monitoring perioperative serum albumin can identify anastomotic leakage in colorectal cancer patients with curative intent. Asian J Surg 2018;41:30-8.
4. Xu H, Kong F. Malnutrition-Related Factors Increased the Risk of Anastomotic Leak for Rectal Cancer Patients Undergoing Surgery. Biomed Res Int 2020;2020:5059670.
5. Huang J, Zhou Y, Wang C, et al. Logistic regression analysis of the risk factors of anastomotic fistula after radical resection of esophageal-cardiac cancer. Thorac Cancer 2017;8:666-71.
6. Pierce K, Gaskins J, Martin Ii RCG. The Weight of Nutrition on Post-Resection Oncologic Morbidity and Mortality: A Systematic Review and Meta-Analysis of Nutritional Indices. Nutr Rev 2025;83:988-1005.
7. Xue Y, Zhou X, Xue L, et al. The role of pretreatment prognostic nutritional index in esophageal cancer: A meta-analysis. J Cell Physiol 2019;234:19655-62.
8. Bouillanne O, Morineau G, Dupont C, et al. Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients. Am J Clin Nutr 2005;82:777-83.
9. Harimoto N, Yoshizumi T, Inokuchi S, et al. Prognostic Significance of Preoperative Controlling Nutritional Status (CONUT) Score in Patients Undergoing Hepatic Resection for Hepatocellular Carcinoma: A Multi-institutional Study. Ann Surg Oncol 2018;25:3316-23.
10. Low DE, Alderson D, Cecconello I, et al. International Consensus on Standardization of Data Collection for Complications Associated With Esophagectomy: Esophagectomy Complications Consensus Group (ECCG). Ann Surg 2015;262:286-94.
11. Haga Y, Ikei S, Ogawa M. Estimation of Physiologic Ability and Surgical Stress (E-PASS) as a new prediction scoring system for postoperative morbidity and mortality following elective gastrointestinal surgery. Surg Today 1999;29:219-25.
12. Raymond DP, Seder CW, Wright CD, et al. Predictors of Major Morbidity or Mortality After Resection for Esophageal Cancer: A Society of Thoracic Surgeons General Thoracic Surgery Database Risk Adjustment Model. Ann Thorac Surg 2016;102:207-14.
13. van Workum F, Verstegen MHP, Klarenbeek BR, et al. Intrathoracic vs Cervical Anastomosis After Totally or Hybrid Minimally Invasive Esophagectomy for Esophageal Cancer: A Randomized Clinical Trial. JAMA Surg 2021;156:601-10.
14. Biere SS, Maas KW, Cuesta MA, et al. Cervical or thoracic anastomosis after esophagectomy for cancer: a systematic review and meta-analysis. Dig Surg 2011;28:29-35.
15. Guo X, Chen C, Zhao J, et al. Neoadjuvant Chemoradiotherapy vs Chemoimmunotherapy for Esophageal Squamous Cell Carcinoma. JAMA Surg 2025;160:565-74.
16. Gui D, Spada PL, De Gaetano A, et al. Hypocholesterolemia and risk of death in the critically ill surgical patient. Intensive Care Med 1996;22:790-4.
17. Asteriou C, Lazopoulos A, Rallis T, et al. Video-assisted thoracic surgery reduces early postoperative stress. A single-institutional prospective randomized study. Ther Clin Risk Manag 2016;12:59-65.
18. Arends J, Bachmann P, Baracos V, et al. ESPEN guidelines on nutrition in cancer patients. Clin Nutr 2017;36:11-48.
19. Bozzetti F. Rationale and indications for preoperative feeding of malnourished surgical cancer patients. Nutrition 2002;18:953-9.
20. Slieker JC, Komen N, Mannaerts GH, et al. Long-term and perioperative corticosteroids in anastomotic leakage: a prospective study of 259 left-sided colorectal anastomoses. Arch Surg 2012;147:447-52.
21. Rutegård M, Westermark S, Kverneng Hultberg D, et al. Non-Steroidal Anti-Inflammatory Drug Use and Risk of Anastomotic Leakage after Anterior Resection: A Protocol-Based Study. Dig Surg 2016;33:129-35.
22. Fan Y, Bao X, Lv X, et al. Predictive value of early postoperative blood lipid metabolism for anastomotic leakage after esophageal cancer surgery. Am J Transl Res 2024;16:3794-800.
23. Patel PH, Patel NM, Doyle JP, et al. Circumferential resection margin rates in esophageal cancer resection: oncological equivalency and comparable clinical outcomes between open versus minimally invasive techniques - a retrospective cohort study. Int J Surg 2024;110:6257-67.