Role of Interleukin-6 in Prediction of Early Complications After Minimally Invasive Oesophagectomy—a Pilot Study

Infectious complications following oesophagectomy are associated with significant morbidity. Early prediction of these complications may mitigate significant morbidity and mortality. Patients undergoing minimally invasive oesophagectomy for carcinoma oesophagus between January 2019 and June 2020 were included in the study. All patients underwent standard preoperative investigations and preparation. Post-operative complications including infectious complications were recorded. Association of post-operative serum interleukin-6 (IL-6) levels with post-operative complications were analysed. A total of twenty-two participants were included in the study (median age; 51 years, 13 (%) male). The tumour site was middle 1/3rd of oesophagus in 13 (59.1%), lower 1/3rd of oesophagus in 9 (40.9%). The tumour histology was squamous cell carcinoma in all patients. Eight (36.4 %) patients developed major complications and five of them developed anastomotic leak. IL-6 levels were significantly higher on POD 3 in patients who developed major complications (p = 0.009) and anastomotic leak (p = 0.031). At receiver operating characteristic curve (ROC curve) analysis, an IL-6 cut-off level of 36.4 pg/ml on POD 3 yielded a sensitivity of 87% and a specificity of 79% for the prediction of major complication and cut-off level of 44.3 pg/ml on POD 3 yielded a sensitivity of 80% and a specificity of 82% for the prediction of anastomotic leak. A high post-operative IL-6 level helps in the prediction of major complications and cervical oesophagogastric anastomotic leak.


Introduction
Oesophagectomy is associated with significant morbidity.Respiratory complications, anastomotic leak, systemic inflammatory response syndrome (SIRS), sepsis are the most serious complications following oesophagectomy.Even though minimally invasive oesophagectomy (MIE) were associated with less blood loss and decreased length of stay but pulmonary complications and anastomotic leak were comparable with open techniques [1].Early prediction of patients at high risk for developing complications helps in clinical decision making and directed therapy thereby improving their outcome.
Various markers have been investigated for their ability to predict complications after oesophagectomy.Among these different markers, serum albumin, C-reactive protein and interleukin-6 (IL-6) are regarded as the most favourable prognostic markers available.Only few studies have assessed the predictive value of IL-6 for early diagnosis of post-operative systemic inflammatory response syndrome (SIRS), pneumonia, gastroesophageal anastomotic leakage and pulmonary complications [2][3][4][5].These studies have heterogenous study population, including major abdominal surgery for both oesophageal and gastric malignancies, thoracic surgeries for lung, pleura and oesophagus.
So in this prospective study, we planned to evaluate the feasibility of IL-6 as a predictive biomarker in the early prediction of complications after minimally invasive oesophagectomy (MIE) for oesophageal cancer.

Study Design and Participants
This is a single-centre prospective study performed at a tertiary care institute in North India between January 2019 and June 2020.The study was approved by the institutional ethical committee.All participants gave written informed consent.All procedures followed were in accordance with the ethical standards and with the Helsinki Declaration of 1975, as revised in 2008.Consecutive patients of biopsy-proven carcinoma of the middle or lower third of oesophagus including gastroesophageal junction (GEJ) tumour were assessed for inclusion.Patient demographic characteristics including comorbidities were recorded all included patients.All patients underwent upper gastrointestinal endoscopy and staging was done using contrast-enhanced computed tomography (CECT) of the neck, chest and abdomen.After staging, locally advanced/ advanced cT3 N+ patients will be subjected to neoadjuvant chemoradiotherapy (NACRT) in the form of platinum/taxane doublet + 40 Gy/20# over 4 weeks.The restaging was done using CECT neck, chest and abdomen, 6-8 weeks after completion of NACRT.Patients with poor performance status, unresectable disease, previous thoracic surgery, conversion to open thoracotomy and those who refused consent were excluded.

Surgical Procedure
Oesophagectomy was done using McKeown minimally invasive oesophagectomy (MIE).First, mobilisation was done via video-assisted thoracoscopic surgery (VATS)-right thoracoscopic approach in the prone position.The abdominal dissection was done via laparoscopic or open approach as per the surgeons' discretion.Reconstruction was done using gastric conduit with a left cervical oesophago-gastric anastomosis in all patients using hand sewn or stapled technique as per surgeon's discretion.A jejunostomy tube was placed in all patients for enteral nutrition.

Post-Operative Management and Outcome Assessment
Patients were encouraged for early ambulation and chest physiotherapy.They were started on enteral nutrition via jejunostomy tube on the first post-operative day (POD 1).Daily chest drain output volume and character were monitored and recorded.Chest drain was removed when output was < 50 ml/day.
Patients are examined for infectious and anastomotic leak complications and the findings were recorded.
The complications occurring during the hospitalisation or within 30 days of the surgery were graded as per Clavien-Dindo classification into minor (grade I and II) or major (III and above) [6].

Biochemical IL-6 Analyses
Blood samples were collected from the patient in a EDTA vial of 3 ml before surgery on admission (preoperative) and then post-operatively on the morning of POD 1 and POD 3. Samples centrifuged and plasma were transferred to a polypropylene tube and stored in a freezer at -82 °C.The assays were performed in duplicate using human IL-6 ELISA kit.

Statistical Analysis
Data obtained from the study including all the demographic, operative and tumour characteristics, post-operative complications, Clavien-Dindo classification, preoperative, POD 1 and POD 3 values of IL-6 values were entered directly into the SPSS for analysis.
Statistical analysis was performed using SPSS, version 26 for Windows (SPSS Inc., Chicago, IL).Data for the clinical characteristics of patients (age, BMI, operative time, blood loss, serum albumin, CRP and IL-6) are expressed as median (IQR).Chi-square test was performed to determine the association between categorical variables.The clinical characteristics of the patients were compared using the Mann-Whitney U test according to the variable type and the data distribution.The Mann-Whitney U test was performed for two independent groups.The ROC analysis was performed to evaluate the diagnostic accuracy of IL-6.The cut-offs were determined along with sensitivity and specificity.Area under the curve (AUC) was computed and 95% confidence interval (CI) was estimated.P < 0.05 was taken as statistical significance for all the analyses.

Patient Demographic and Tumour Characteristics
A total of 29 patients were evaluated for inclusion.One patient with poor performance status, one patient with conversion to open thoracotomy, two patients with unresectable disease on thoracoscopy and three patients with technical errors on sample storage were excluded.Thus, a total of 22 patients were included in the final analysis.The demographic and tumour characteristics are summarised in Table 1.The median (IQR) age was 51 (45, 60) years.The tumour site was middle 1/3rd in 13 (59.1%),lower 1/3rd of oesophagus in 9 (40.9%).The tumour histology was squamous cell carcinoma in all patients.

Proinflammatory Markers and Complications
The relation of proinflammatory markers with post-operative complications and anastomotic leak is summarised in Table 3. Preoperative levels of IL-6, CRP and serum albumin were similar among the no major complication and major complication group.The median level of IL-6 was significantly higher in the major complication group on POD 3 (46.88pg/mL vs 15.34 pg/mL; p = 0.009) (Fig. 1A).The median IL-6 levels in patients who developed cervical anastomotic leak were significantly higher on POD3 compared to no anastomotic leak group on POD 3 (48.28pg/ mL vs 19.4 pg/mL; p = 0.031) (Fig. 1B).Using an ROC curve, an IL-6 cut-off level of 36.4 pg/ml on POD 3 yielded a sensitivity of 87% and a specificity of 79% for the prediction of major complication (AUC: 0.839; 95% CI: 0.64-1.0)(Fig. 2A).An IL-6 cut-off level of 44.3 pg/ml on POD 3 yielded a sensitivity of 80% and a specificity of 82% for the prediction of cervical anastomotic leak (AUC: 0.824; 95% CI: 0.642-1.0)(Fig. 2B).

Discussion
In our prospective observational study, we evaluated the role of IL-6 in predicting the complication after MIE for carcinoma oesophagus.In our study, major complications (grades III-V) have been reported in nine patients (40.9%) and cervical anastomotic leak in five patients (22.7%).The raised IL-6 level on POD 3 helped in predicting major complications and cervical anastomotic leak.
There is a paucity of studies using proinflammatory markers as a predictor of complications after oesophageal resection for carcinoma oesophagus and most of these are retrospective studies.Therefore, in the present study, we evaluated the role of proinflammatory markers in predicting the complications after MIE for carcinoma oesophagus prospectively.Despite improvements in patient selection, preoperative staging, surgery and perioperative care, a significant number of patients develop complication after oesophagectomy.There is an overproduction of inflammatory mediators in the perioperative period which has been shown to predict the development of post-operative complications.Various proinflammatory markers serum albumin, CRP, procalcitonin and IL-6 have been studied in predicting complications after oesophagectomy with ambiguous results   [2,[7][8][9].Only few studies were able to predict post-operative complications.The ambivalent nature of these results paved the way for further research into these proinflammatory markers.
Serum IL-6 response to surgical trauma has been characterised extensively.Surgical stress induces a rise in circulating IL-6 within hours of surgery and may remain elevated upto 72 h in uncomplicated cases [10].In one study, serum IL-6 value of 310 pg/ml on POD 1 showed a sensitivity of 90% and specificity of 58% in diagnosing post-operative sepsis [11].In another study, they did sequential measurements of serum IL-6 levels in patients who had undergone surgical resection of the oesophagus to differentiate between sepsis and post-operative SIRS.Persistently elevated serum IL-6 levels until POD 5 were used to predict and differentiate sepsis from post-operative SIRS [12].In another study, high IL-6 levels on POD 1 recognised high-risk patients for complications even before changes in CRP levels [2].
In our study, IL-6 level on POD 3 was statistically higher in the complication group.To establish a threshold for the relationship between IL-6 and the development of major complications and cervical anastomotic leak after MIE, the receiver operator characteristics (ROC) curve was plotted.A threshold value of IL-6 levels to predict major complications (36.4 pg/ml) providing a sensitivity (87%) and specificity (79%) with a diagnostic accuracy of 0.839 and cervical anastomotic leak (44.3 pg/ml) providing a sensitivity (80%) and specificity (82%) with a diagnostic accuracy of 0.824.The early detection of complications is clinically important for the initiation of early goal-directed therapy to improve the patients' outcome.Elevated IL-6 levels on POD 3 alarm further evaluation for detection and management of complications.
Though the study is limited by a small number of study population and single centre, it showed a significant association of IL-6 level in predicting early complication in minimally invasive oesophagectomy.Early initiation of directed therapy in these high-risk patients might help in improving clinical outcomes.However, a multicentre study with a large sample size is required to validate our results.

Fig. 1 Fig. 2
Fig. 1 IL-6 levels before surgery and on POD 1 and 3 dependent on the occurrence of major complications (A) and anastomotic leak (B)

Table 1
Clinical demographic, operative and tumour characteristics BMI body mass index, IQR inter quartile range, min minutes, ml millilitre

Table 2
Post-operative complications

Table 3
Proinflammatory markers compared with major complication and anastomotic leak group IQR inter quartile range, IL-6 interleukin 6, CRP C reactive protein, pg pictogram, gm gram, dl decilitre, L litre †Mann-Whitney U test