The present study showed an increased risk of AL in patients with a preoperative high inflammatory state and visceral fat index after CRC surgery and evaluated the possible protective effect of PPN against the occurrence of AL.
The presence of increased activation of the systemic inflammatory response has been linked to visceral fat18. Our analysis showed no differences in mean CRP values prior to surgery based on VFI, but patients with high VFI were more frequent in the CRP groups ≥ 5 mg/dL and ≥ 10.2 mg/dL. Himber C et al.8 showed in their analysis a correlation between CRP values and VFA measured by CT, although this correlation was weak.
In the postoperative period, we observed a difference in CRP values as a function of VFI, reaching statistical significance on the second day. Okamura A. et al.19 reported a similar relationship between CRP values and VFA in postoperative patients undergoing esophagectomy.
Visceral fat has also an important influence on postoperative outcomes20. One meta-analysis21 found a relationship between anastomotic leakage and visceral fat area.
Similarly, Kuritzkes et al.22 performed an ROC curve model that linked the VFA of patients with the ability to predict major complications. Its area under the curve was higher than ours (0.66 vs. 0.56). The cutoff point with the best sensitivity/1-specificity ratio was 191 cm2, which presented an OR of 2 to predict major complications. However, they did not differentiate the analysis according to the patient’s sex. Females are not expected to have a large amount of visceral fat, although having a greater amount of total fat is located preferably in the subcutaneous cell tissue. In males, however, there is a higher ratio of visceral fat/subcutaneous fat10. This feature can justify why our ROC curve model has a greater area under the curve in females (0.683 vs. 0.438).
CRP measurement allowed for early detection of anastomotic leak in colorectal cancer surgery, us to anticipate their treatment by reducing the impact and morbidity caused by them23. CRP is the most commonly used biomarker for its ability to anticipate AL after colorectal surgery and is considered the gold standard in postoperative follow-up24.
The use of biomarkers of the inflammatory response is also beginning to be used in the preoperative period. AL could be related to an altered inflammatory response, expressed by elevated serological levels of inflammatory markers prior to surgery, such as CRP25.
The identification of models predicting postoperative complications or specifically anastomotic leakage is one of the current focuses26,27. However, most of them use a large number of variables that make it difficult to implement them in normal clinical practice. The determination of factors related to body composition or to patients' presurgery inflammatory state is beginning to be included in these prediction models28,29.
The PPN complementary to the oral route has demonstrated its benefit in reducing mortality and complications in situations of high energy demand12,13, enhancing its protective effect in fragile patients or with low muscle mass30,31. PeriOlimel N4-E is a peripheral parenteral nutrition (PPN) with a high content of oleic acid, which in addition to allowing supplementation via the oral route, could help reduce oxidative stress and the inflammatory response after surgery32.
The current guidelines for clinical practice in surgery advise starting an NP in situations of high risk of malnutrition or after 5 days without achieving a satisfactory oral tolerance and do not individualize the treatment according to muscle reserves or the degree of catabolic response presented by the patient15,33.
Our analysis shows that the reduction in the rate of anastomosis leakage in patients who have received a complementary PPN is greater in patients with a higher degree of inflammatory activation prior to surgery.
To the best of our knowledge, this is the first study to analyze the risk of AL in CRC surgery within an ERAS program based on the patient’s inflammatory state and visceral fat index and to evaluate the possible protective effect of PPN. As limitations of our study, we mainly highlight the low sample size that prevented reaching significant differences in risk group divisions and being a unicentric study. In addition, our risk model for AL is easy to perform, allowing the detection of at-risk patients and the individualization of treatments.