The Prognosis And Influence Factors Of Stress Hyperglycemia Within Gastric Cancer Patients Receiving Post-surgical Nutritional Support

BackgroundGastric cancer has a high morbidity and mortality in China. Most of the time, surgery is preferred for the treatment of gastric cancer, and therefore improving upon post-surgical outcomes is a key focus. The non-diabetic patients with gastric cancer will develop stress hyperglycemia （SHG） in response to the surgical trauma, compromising their recovery from surgery. Because the glucose metabolism is disrupted in patients with SHG, infection becomes more prevalent and wound healing is compromised, thereby worsening surgical outcomes. SHG induces chronically-elevated blood glucose and increases glycemic variability （GV） arising from an unstable fluctuation of blood glucose. Previous studies have shown that GV poses greater harm to patients than to constantly elevated blood glucose. Enteral nutrition （EN） or parenteral nutrition （PN） is provided to non-diabetic gastric cancer patients during the post-surgery fasting period, which may increases GV. While the significance of GV within diabetic patients has been highlighted, the GV within non-diabetic patients is largely overlooked. The gold standard for evaluation GV has not been unified. The glycemic liability index （GLI）, which is an index of GV, measures both the amplitude and rate of blood glucose changes, enabling the temporal assessment of GV, and hence promises to be a valuable tool. Therefore, in order to promote non-diabetic gastric cancer patients’ recovery, it is necessary to investigate the influence factors contributing to SHG post-surgery within non-diabetic gastric cancer patients receiving nutritional support, choose a suitable nutrition support, examine blood glucose index as a prognostic marker and determine a suitable index for monitoring GV within non-diabetic gastric cancer patients receiving post-surgical nutritional support. All those will do great good for patients to stable their blood glucose level and can promote their prognosis.Objectives1. To investigate the incidence of, and influencing factors for, SHG post-surgery within non-diabetic gastric cancer patients receiving nutritional support as a reference for developing appropriate management strategies.2. To quantify GV using GLI and determine its post-surgical predictive value for SHG in non-diabetic gastric cancer patients.3. To identify an typical blood glucose index for monitoring and informing clinical intervention for non-diabetic gastric cancer patients suffering SHG during post-surgery nutritional support.MethodsTwo-hundred and twenty patients with no previous diagnosis of diabetes and treated for gastric cancer between November 2014 and July 2015 at the general surgery unit of a tertiary hospital in Jiangsu, China, were enrolled. Patient characteristics （age, gender, educational background, past history, and smoking and alcohol consumption habits）, their treatment regimens （surgery type, method and duration, and mode of nutritional support） and their prognosis（had complication or not） were obtained. Patients’blood glucose levels were monitored at admission （BGAdm）, on the morning of surgery （06:00）, immediately after returning to ward post-surgery, and 4 times daily during the first 3 days post-surgery （00:00,06:00, 12:00,18:00）. Patients’nutritional status was assessed according to the nutrition risk screening 2002 （NRS 2002）. Patient anxiety and depression levels were scored according to the self-rating anxiety scale （SAS） and self-rating depressive scale （SDS）, respectively. _Average fasting blood glucose (FBG_Ave), average of blood glucose (BG_Ave), and GLI were calculated. Factors impacting upon non-diabetic patients during the post-surgery nutritional period were examined using univariate and binary logistic regression analyses. The prognostic value of FBG_Ave, BG_Ave and GLI were modelled with binary logistic regression and assessed according to the area under the receiver operating characteristic curve （AUC）. Finally, the threshold of GLI at which complications occurred was further determined by the median and interquartile range of GLI. Patients were assigned to PN-only or combined EN-PN nutritional support groups, the prognostic values of FBG_Ave, BG_Ave, and GLI within these two groups and the better nutritional support which could maintain GV were determined by t test or chi-square testResults1. The incidence of SHG in patients with gastric cancer during the post-surgical nutritional support period was 32.87%. Random glucose testing revealed that patients who experienced SHG had the highest blood glucose levels when returned to the ward after surgery; while fasting blood glucose was the highest on the first day post-surgery, and decreased gradually with time.It showed that duration of surgery （P= 0.044）, nutritional risk （P< 0.001）, level of depression （P= 0.002）, anxiety status （P= 0.003）, age （P= 0.008）, mode of nutritional support （P= 0.009）, and method of surgery （P= 0.038） were influence factors for SHG in gastric cancer patients during the post-surgical nutritional support period.2. Patients were divided into two groups according to their complications. There were no significant differences in the age, gender, method of surgery or BGAdm between the 2 subsets of patients with or without surgical complications （P> 0.05）, whilst the FBG_Ave BG_Ave and GLI were higher in patients with complications （P< 0.05）. Modelling the individual associations of FBG_Ave BG_Ave and GLI on the prognosis of non-diabetic gastric cancer patients receiving surgery using binary logistic regression reveals that the FBG_Ave, BG_Ave, and GLI were all independent predictors of patients’prognosis （P< 0.05）. The AUC for FBG_Ave, BG_Ave, and GLI in non-diabetic gastric cancer patients receiving surgery were 0.657,0.711, and 0916, respectively. There were significant differences in AUC between GLI vs FBG_Ave and GLI vs BG_Ave but there was no difference in AUC between FBG_Ave vs BG_Ave (GLI vs. FBG_Ave, Z= 4.35, P= 0.000; GLI vs. BG_Ave, Z= 3.50, P= 0.000; FBG_Ave vs. BG_Ave, Z= 0.71, P= 0.475).The incidence of complications, FBG_Ave BG_Ave, and GLI were higher within patients experiencing SHG compared to those without, with the differences in incidence of complications and blood glucose index being statistically significant （P < 0.01）. Patients with SHG were divided into 4 subgroups according to the interquartile range (2.38（mmol/L）2·h^-1·d^-1), 6.69（mmol/L）2·-h·d^-1) and the median (3.19 （mmol/L） 2·h^-1·d^-1) of GLI. There were 1,2,9, and 15 cases, and an incidence rate of 5.88%,11.11%,50.00%and 83.33%within subgroups 1-4, respectively. The incidence rate of complications increased with GLI. Only subgroups 2 and 3 showed a significant difference in the incidence of complications when consecutive subgroups were compared in a pairwise manner （χ2= 6.415, P= 0.011）.Patients who did not develop SHG were divided in two subgroups about the median GLI (1.14 （mmol/L） 2·h^-1·d^-1). There were 1 and 11 cases within the below and above-median subgroups, respectively, for a complication rate of 1.37%and 15.28%. The incidence of complications increased as GLI increased. The difference in incidence rates between the below and above-median subgroups was statistically significant （χ2= 9.237, P= 0.002）.3. Patients were divided into two subgroups based on their mode of nutritional support:PN-only and EN-PN combined. There was no statistical difference in the baseline characteristics （age, gender, and past history） between the PN and EN-PN subgroups （P> 0.05）. The incidence of SHG and GLI were higher in patients within the EN-only than those in the EN-PN subgroups （P< 0.05）, but there were no significant differences for FBG_Ave and BG_Ave（P> 0.05）.Conclusions1. The incidence of SHG in patients with gastric cancer during the post-surgical nutritional support period was high. The duration of surgery, nutritional risk, levels of depression and anxiety, age, mode of nutritional support, and method of surgery were significant predictors for the presence of SHG within non-diabetic gastric cancer patients receiving post-surgical nutritional support. Clinical staff should therefore treat with an appropriate surgical method, minimize surgical time, be attentive to the patients’ nutritional status. More frequent observations should be also provided for the senior or weak patients. Furthermore, nutritional support should be provided as early as possible after surgery and due care should be given to a patients’ depression or anxiety scores, so that appropriate psychological support can be given to help alleviate their stress.2. GLI appears to be a superior post-surgery prognostic marker for non-diabetic gastric cancer patients. Clinical staff should therefore monitor a patients’ blood glucose for at least 3 days, increase the monitoring of blood glucose levels and focus on GLI fluctuations. GLI of SHG gastric cancer patients should be maintained at below 3.19 （mmol/L） 2·h^-1·d^-1, while that of patients experiencing non-SHG should be maintained at below 1.14 （mmol/L） 2·h^-1·d^-1 in order to stabilize the overall blood glucose levels.3. Nutritional support with EN-PN was superior to PN-only in managing blood glucose levels and reducing the incidence of GV. GLI was a superior indicator of GV in non-diabetic gastric cancer patients during the post-surgical nutritional support period compared to FBG_Ave and BG_Ave-Medical staff should provide EN support as early as possible to patients who meet the criteria as to minimize the risk of developing SHG. GLI should be used for GV monitoring in patients during post-surgical nutritional support.