Gastric cancer is the fifth most common cancer in the world and the third most common cause of cancer-related death [1
]. Surgery is the main treatment for gastric cancer, and this treatment is one of the most common surgeries performed in South Korea [2
]. Although improvements to surgical devices and advances in surgical techniques have resulted in reduced mortality after gastrectomy, postoperative complications remain a clinically significant problem [4
Surgical site infection (SSI) is one of the most common postoperative complications of gastrectomy. SSI is associated with prolonged hospitalization and increased mortality [5
]. Therefore, it is important to investigate the risk factors for SSI.
Recently, the effects of different body compositions on SSI have been attracting attention. Especially, the effects of fat distribution and loss of skeletal muscle, called sarcopenia, have been focused on. Some studies have shown that visceral fat is a risk factor for postoperative complications in digestive surgery [6
]. Sarcopenia, is also known to be associated with complications following digestive surgery [9
]. However, each of these studies has used different body composition criteria. Because there are differences in body composition between patients from different ethnic backgrounds and regions, it is difficult to define the body composition criteria based on a specific value. To our knowledge, universally applicable criteria to define body composition have yet to be established. In addition, as almost all patients are given a computed tomographic (CT) scan before gastric cancer surgery, measuring body composition using a preoperative CT scan for predicting SSI could be worthwhile. Therefore, in this study, we aimed to investigate the most predictable body composition criteria to predict SSI after gastrectomy using a preoperative CT scan.
This study demonstrated that the VFA-to-SFA ratio had the highest predictive power for SSI after gastrectomy among the body composition criteria assessed, especially the organ/space SSI. Moreover, the VFA-to-SFA ratio could be widely applicable, because it is a value defined by a ratio, not an absolute value. The large number of patients who underwent gastrectomy in our study increased the clinical significance of this result.
A high VFA-to-SFA ratio indicates either an increased VFA or a decreased SFA. As the area of the abdominal fat and muscle is obtained through the CT scan, several studies have shown that VFA is better than BMI at predicting postoperative complications, including SSI. One reason patients with visceral obesity are more likely to develop an SSI could be due to the surgical difficulty associated with the surgeon having a deeper and poorer view of the surgical field, as well as the fragile, easily bleeding tissue in high-visceral obesity patients [7
]. In addition, visceral obesity is expected to increase post-operative morbidity, because visceral adipose tissue secretes a number of adipokines and cytokines leading to a proinflammatory, procoagulant, and an insulin-resistant state, collectively known as the metabolic syndrome [27
]. Thus, patients with visceral obesity often have higher ASA scores or more comorbidities, including diabetes and cardiovascular diseases [6
]. These are the reasons SSI, especially organ/space SSI, occurs more frequently in patients with visceral obesity. Related to these findings, Takeuchi et al. classified VFA into two groups based on a cutoff of 100 cm2
, and reported that SSI occurred more frequently in the high VFA group [8
]. However, in our study, when compared using the ROC curve and AUC value, the VFA-to-SFA ratio was more predictive for the SSI than the VFA itself.
Conversely, SFA improves insulin sensitivity. Thus, it can work as a buffer against the lipid accumulation of VFA. Consequently, SFA insufficiency causes increased lipid accumulation in visceral fat, and therefore the VFA-to-SFA ratio is effective for evaluating the risk of diseases affected by the adipose tissue [29
]. For this reason, several studies on the influence of the VFA-to-SFA ratio have been conducted in various fields in recent years. Yosuke et al. reported that the VFA-to-SFA ratio was an independent risk factor for decreased renal function in kidney transplantation recipients [30
]. Kaess et al. reported that the VFA-to-SFA ratio is correlated with cardiometabolic risk, more than BMI and VFA itself [31
]. Our study revealed that these findings can also be applied to the risk prediction of SSI after gastrectomy. Traditionally, in addition to visceral obesity, sarcopenia is known to be a major risk factor for SSI after gastrectomy. Tatsuko et al. reported that sarcopenic obesity, which was defined as a skeletal muscle mass index of ≤52.4 cm2
for men and ≤38.5 cm2
for women, and a visceral fat area of ≥100 cm2
in both sexes, is an independent risk factor for the development of SSI after laparoscopic total gastrectomy [32
]. However, the VFA-to-SFA ratio was more predictive for SSI than sarcopenia. Also, the above study defined sarcopenia and obesity based on specific values, which would be difficult to apply widely.
Recently, some studies have been published that address the risk of postoperative complications, such as SSI, by values calculated by a ratio rather than a specific value based on body composition criteria. According to Pecorelli et al., the high VFA-to-TAMA ratio was a major risk factor for postoperative complications, especially in pancreatic cancer patients [33
]. However, in our study using a ROC curve, the AUC value of the VFA-to-SFA ratio was greater than that of the VFA-to-TAMA ratio, making the VFA-to-SFA ratio a better predictor of SSI than the VFA-to-TAMA ratio after gastrectomy.
In addition to the individual fat distribution, total gastrectomy and stage III or IV cancer also influenced the occurrence of SSI. Total gastrectomy is one of the most invasive gastrointestinal surgeries, and the risk of SSI after a total gastrectomy is known to be higher than after a partial gastrectomy [35
]. Thus, this finding is consistent with the conclusions of previous studies. On the other hand, male sex, smoking, and open surgery, which were also known as risk factors of SSI, did not show statistically significant associations in multivariate analysis after adjusting the effect of the VFA-to-SFA ratio. A relatively small number of events might have limited the statistical power to detect associations. If the sample size gets larger, these variables may also be statistically significant, especially the male gender with a high OR and wide confidence interval.
This study had several limitations. First, this study was a nonrandomized retrospective study of a single country with a homogenous ethnic background. Second, we excluded 10 patients with follow-up loss or incomplete data review of the 1067 patients who underwent a gastrectomy in 2015. Although these patients did not develop SSI during the hospitalization, and the rate of SSI occurrence after discharge was low (12/1038; 1.2%), we could not completely rule out the possibility that SSI occurred after discharge in these patients. Third, active post-discharge surveillance using phone calls or family physicians has not been conducted. Thus, there was a potential for some SSI incidents to be missed. Fourth, as a result of the model comparison, the 95% confidence interval of the AUC value of each model overlapped. Fifth, in the present study, the incidence of SSI after gastrectomy was 5.6%, which is not high. Therefore, this raises the issue that the sensitivity and PPV were low for predicting the increased risk of SSI with a specific cutoff value. However, SSI is a complication that cannot be accurately predicted before surgery, so it is meaningful to find patients who are at high risk for SSI before surgery. Surgeons may consider minimal invasive surgical methods and try to reduce the operation time so as to reduce SSI in patients with a high VFA-to-SFA ratio. In this regard, even though the sensitivity and PPV were low, the specific cutoff value we determined would be beneficial. In addition, the high NPV of the VFA-to-SFA ratio cutoff value allows surgeons to identify patients at a very low risk of developing SSI before surgery. Further research is needed to determine if it will be applicable in other abdominal surgeries and other countries.