Patients undergoing surgical resection for CD are at high risk for postoperative septic complications, and early identification such patients would enable benefit from early intensive management. The results of the present study showed that levels of CRP, IL-6 and PCT were significantly different between IASC and Non-IASC group. Generally, trajectory of CRP, IL-6 and PCT had reasonably good specificity and NPV, which indicated superiority in ruling out non-IASC cases. Continued elevation of CRP, as depicted by the average of POD 3 and POD 5, was a novel parameter to predict IASCs with an AUC of 0.947.
There is currently a lack of studies comparing the trend of inflammatory biomarkers after anastomosis for CD, especially IL-6 and PCT. Most studies focused on using IL-6 and PCT on isolated time points to predict postoperative outcomes. In a meta-analysis of 1452 patients from intensive care units, PCT could not reliably differentiate sepsis from other non-infectious causes of systemic inflammatory response syndrome in critically ill adult patients.(14) However, the role of IL-6 in predicting complications has been proved in practice. IL-6 was first identified by its ability to promote the activation and reproduction of T cells, the differentiation of B cells, and regulation of the acute-phase response.(15) Another study of 50 consecutive patients who underwent elective major gastrointestinal or gynecological tumor resections8 showed the AUC of IL-6 and PCT on POD 1 in diagnosing septic complications were 82.1% and 66.4% respectively, and the optimal cutoff value of IL-6 was set at 310 pg/mL. We set the optimal cut-off value of IL-6 on POD 1 in predicting IASCs at 158.19 pg/mL, which was lower than previous literature that set optimal cut-off level for the prediction of infectious complications between 300 and 500 pg/mL.(7, 15, 16) We speculate this is due to lack of an appropriate immunological response in CD patients, as they always suffer from chronic inflammation, concurrent malnutrition and always need immunosuppressive medication, which could further compromise the patients’ immunological responses.(17–19)
Our study for the first time compared the trajectory of CRP, IL-6 and PCT in predicting postoperative IASCs for CD patients, showing CRP and IL-6 had better performance than PCT. The NPV was 0.95 and 0.91 respectively for an increase of IL-6 exceeding 30 ng/L and CRP exceeding 50 mg/L from POD 3 to 5, which means that patients who did not follow this trajectory had 95% and 91% chance of not developing IASCs. Therefore, CRP and IL-6 change from POD 3 to POD 5 might serve as an excellent biomarker to screen patients that are suitable for early discharge. Moreover, specificity of CRP increase > 50 mg/L and IL-6 increase > 30 ng/L was 0.99 and 0.97, highlighting the value of CRP and IL-6 monitoring in excluding IASCs for patients whose CRP and IL-6 level is not increasing.
Trajectory modeling seemed to provide higher specificity but lower sensitivity compared with using cut-off points in this study. The subtle difference could be explained after considering the nature of IASCs and biomarker testing. A high concentration in an isolated test should be more efficient in diagnosing IASCs, whereas lack of an increase over consecutive days are more likely to rule out a leak. Additionally, surgery might result in CRP increase even for patients without IASCs. The CRP level of these cases usually falls rapidly after recovering from surgical insult. Therefore, we innovatively calculated the average of CRP to represent two sorts of patients 1): patients with increasing CRP; 2): patients who had high CRP without sharp decrease. Combining the average calculation and trajectory testing should be informative to predict infections during hospital stay, allowing early and safe discharge from hospital, particularly in the era of fast-track or enhanced-recovery surgery.
Our study has several limitations. First, this is a single-center study based on the analysis of a series of patients who were transferred from other hospitals, and the patients generally had a more severe condition, which yielded a selection bias. Second, the serum parameters were only available on every other day, which might cause inaccuracy in reflecting the trend of inflammatory biomarkers.
In conclusion, continued elevation of CRP serves as a reliable serological indicator in predicting postoperative IASCs for CD patients undergoing intestinal anastomosis. Trajectory of CRP and IL-6 are both prominent in ruling out IASCs. Further prospective trials are needed to validate our findings.