While IL-6 and CRP rise early after lung cancer surgery, new research shows their added value for infection prediction remains small, highlighting the need for better early-warning tools.
Does postoperative plasma IL-6 improve early prediction of infection after pulmonary cancer surgery? A two-centre prospective study. Image Credit: 3dMediSphere / Shutterstock
A recent PLOS ONE study examined the potential of IL-6 as a biomarker for predicting post-surgery infection risk.
Risks of infection after pulmonary cancer surgery
Approximately 20% of the patients who underwent pulmonary cancer surgery develop postoperative infection, which might progress into sepsis. These infections adversely affect treatment outcomes and can even increase mortality rates. Many studies have shown that patients who survive sepsis may develop persistent functional disability. To prevent adverse postoperative consequences, it is imperative to provide timely antibiotic treatment to these patients.
Early detection of patients who are more susceptible to developing postoperative infections could be beneficial, as these individuals can be closely monitored for signs of postoperative infection.
Postoperative hyperinflammation could be a valuable predictor to determine these patients. After surgery, damage-associated molecular patterns (DAMPs) are released to trigger the innate immune system, which facilitates both pro- and anti-inflammatory responses to promote tissue healing and prevent microbial infections. However, some patients develop an excessive inflammatory response to surgery that leads to hyperinflammation, which elevates infection risks.
IL-6 and C-reactive protein (CRP) are common inflammatory biomarkers associated with severe injury. Many studies have shown that compared to CRP, IL-6 reaches peak concentrations more quickly after surgery and has a shorter plasma half-life. In contrast, CRP concentrations are maximized nearly three days after surgery.
IL-6 could be exploited as a biomarker to identify patients at a higher risk of developing postoperative infections, allowing for timely antibiotic intervention to prevent severe infections. To date, few studies have investigated the potential of IL-6 as a predictor of postoperative infection in pulmonary surgery.
About the study
A prospective observational two-center cohort study was conducted at Amphia Hospital Breda and Sint Antonius Hospital Nieuwegein, the Netherlands, to assess postoperative hyperinflammation based on maximum IL-6 concentrations. Besides IL-6, the infection-predictive capacity of CRP, white blood cell count (WBC), and procalcitonin (PCT) were also evaluated.
In this study, participants were recruited between September 3rd, 2018, and April 29th, 2022. All selected participants were undergoing elective pulmonary surgery for cancer with a planned postoperative admission to the Intensive Care Unit (ICU). Blood samples were obtained at various time points, including after induction of general anesthesia (preoperative sample) and after 6, 9, 12, 24, 48, and 72 hours (postoperative samples). Plasma levels of IL-6, CRP, and PCT were measured using a standard protocol.
The development of various postoperative infections (e.g., respiratory infections, urinary tract infections, empyema, or sepsis) within 30 days of surgery was monitored.
A core model was developed based on various variables, including age, sex, surgery duration, and the Charlson Comorbidity Index (CCI). The predictive value of inflammatory biomarkers within 24 hours from the start of surgery was assessed relative to the core model.
Study findings
A total of 170 patients were included in this study, with a median age of 67 years; 45% of the study cohort consisted of females. The CCI was ≥ 3 in 30 (18%) patients. The majority of the study cohort underwent minimally invasive surgery via video-assisted thoracoscopic surgery (VATS), robot-assisted thoracoscopic surgery (RATS), lobectomy, or sleeve resection. The median surgery duration was 193 minutes, and approximately 89% of the patients received dexamethasone as PONV prophylaxis.
In this study, 22% of the cohort developed a postoperative infection. It is worth noting that some patients developed more than one infection. Most infections were related to the respiratory system (74%), followed by surgical site infections (18%) and urinary tract infections (8%). Seven patients developed postoperative sepsis. The median time to diagnosis of infection was four days after surgery.
Approximately 66% of the infections were diagnosed in the first week after surgery. Typically, patients who underwent longer surgery duration, experienced more intraoperative blood loss, and less frequently received IV dexamethasone were more prone to develop an infection.
In comparison to non-infected patients, higher preoperative concentrations of IL-6 and CRP were detected in patients with postoperative infection. In most patients, postoperative IL-6 concentrations peaked within six hours, while maximum CRP concentrations were attained three days after surgery.
It is worth noting that the median maximum concentrations of both IL-6 and CRP within 24 hours post-surgery were higher in patients with postoperative infections. However, WBC and PCT did not demonstrate any predictive capacity.
The current study observed that IL-6 was indeed associated with postoperative infection, independent of the core model predictors. Similarly, maximum CRP concentrations were also associated with postoperative infection. While both biomarkers significantly improved model fit according to statistical tests, this did not translate into a meaningful improvement in practical prediction.
Specifically, the model's ability to discriminate between patients who would contract an infection and those who would not (measured by the c-statistic) changed minimally. Both PCT and WBC failed to improve model fit.
Furthermore, when it came to reclassifying patient risk, adding IL-6 to the core model did not improve the classification of infected patients, whereas the addition of CRP actually worsened the classification for this group. While both improved the classification of non-infected patients, the overall benefit was negligible.
Conclusions
The main takeaway from the current study is that while early postoperative plasma IL-6 and CRP concentrations were independently associated with the risk of subsequent infection, the added value of these biomarkers to a simple clinical prediction model appears to be limited. The authors concluded that neither biomarker substantially improved the prognostic classification when added to a model using readily available clinical data.
The study highlighted several important limitations beyond the need for a larger cohort. The limited sample size resulted in broad confidence intervals for the performance measures, meaning a moderate incremental value may have been missed.
Additionally, the study was too small to analyze the predictive performance relative to whether patients had open versus minimally invasive surgery, a known factor affecting inflammation. Finally, the authors note that the selected times for blood sampling may not have coincided with the exact peak concentration of IL-6 for every patient.
Therefore, the study concludes that the findings must be validated in future and more extensive research before any final conclusions on the clinical utility of either biomarker can be made.