Preoperative Controlling Nutritional Status (CONUT) Score Predicts Short-Term Surgical Prognosis in Patients with Gastric Cancer After Laparoscopy-Assisted Radical Gastrectomy

Background: The Controlling Nutritional Status (CONUT) score is an emerging nutrition assessment tool that is very useful in patients with gastric cancer who usually experience weight loss and malnutrition. The aim of our study was to assess the predictive ability of the preoperative CONUT score for short-term prognosis in patients with gastric cancer undergoing laparoscopy-assisted gastrectomy. Methods: We retrospectively reviewed medical records of 309 patients who underwent curative laparoscopy-assisted gastrectomy. The patients were divided into two groups according to the optimal cutoff value of the CONUT score. The clinical association for the CONUT score, characteristics, and postoperative complications were evaluated and analyzed. The risk factors for complications were identied by univariate and multivariate analysis. Results: The preoperative CONUT score showed a good predictive ability for postoperative complications (AUC=0.718 (cid:0) Youden index=0.343) (cid:0) with an optimal cutoff value of 2.5. The patients with high CONUT scores had a higher incidence of overall complications (P<0.001) and mild complications (P<0.001). Univariate and multivariate analysis revealed that the CONUT score was independently associated with postoperative complications (P=0.012 (cid:0) OR=2.433 (cid:0) 95%CI (cid:0) 1.218-4.862). Conclusions (cid:0) The preoperative CONUT score was identied as a reliable nutritional assessment tool for predicting short-term prognosis in patients with gastric cancer after laparoscopy-assisted gastrectomy. distal gastrectomy compared with open distal gastrectomy for clinical stage I gastric cancer relapse-free survival 37 . Our study involved clinical records of 309 patients who underwent laparoscopy-assisted radical gastrectomy with written informed consent. A certain part of patients with gastric cancer started with laparoscopic exploration (cid:0) and then transferred to open gastrectomy (cid:0) extended or palliative operation according to the surgeon’s judgement (cid:0) due to gross tumor volume (cid:0) deep tumor location or visible existence of tumor invasion or peritoneal metastasis (cid:0) which was deliberately excluded from our analysis. Based on the receiver operating characteristic curve (cid:0) we identied the optimal cutoff value as CONUT score of 2.5 to distinguish patients at disparate risk for postoperative complications. The preoperative parameters constituting CONUT score revealed an authentic and steady nutrition status of patients (cid:0) and it was reasonable for preoperative assessment. When patients underwent operation (cid:0) various factors including surgical trauma (cid:0) intraoperative hemorrhage (cid:0) venous uid utilization (cid:0) postoperative TPN support might inuence their blood components and hemodynamics (cid:0) and bring about the variations of laboratory measurements for nutritional assessment.


Introduction
Worldwide gastric cancer remains a major public health problem. Gastric cancer is fth in incidence among new cases cancer and third in mortality 1 . Adequate surgical resection is the main curative therapeutic option for gastric cancer 2 which inevitably carries some postoperative complications leading to prolonged hospitalization more expenses impaired quality of life and delayed adjuvant chemotherapy treatment.
Gastric cancer patients often suffer from malnutrition which is usually associated with humoral and cellular immune dysfunction in ammatory response alterations and poor wound healing 3 . As an independent risk factor malnutrition contributes to the occurrence of postoperative complications 4 reduces the body response to antitumor treatment and ultimately impairs long-term survival 5 . Several nutritional assessment scales or systems have been developed to detect the adverse condition of nutrition such as the modi ed Glasgow Prognostic Score (mGPS) 6 Prognostic Nutritional Index (PNI) 3 Naples Prognostic Score (NPS) 7 Platelet-Lymphocyte Ratio (PLR) 8 Skeletal Muscle Index (SMI) 9 or the Malnutrition Universal Screening Tool (MUST) 10 among others.
The Controlling Nutritional Status (CONUT) score was rst reported and validated by Ignacio de Ulíbarri J in 2005 as a tool for early detection and continuous control of malnutrition. The formula of CONUT comprises serum albumin total lymphocyte count, and cholesterol (Table.1) 11 . Recently previous studies reported that proactive assessment of nutritional status by the CONUT score could accurately predict long-term outcomes in patients with colorectal cancer 12 hepatocellular carcinoma 13 esophagus cancer 14 and gastric cancer 9,[15][16][17] . However little research work shed light on the CONUT score for shortterm complications after radical gastrectomy. Therefore the primary aim of our study was to assess the predictive ability of preoperative CONUT score for short-term prognosis in patients with gastric cancer who underwent laparoscopic radical gastrectomy.

Study patients
We retrospectively reviewed a series of consecutive clinical records of 412 patients who underwent curative laparoscopic gastrectomy at the Department of General Surgery of Sir Run Run Shaw Hospital the a liated hospital of medicine school of Zhejiang University from January 2016 to June 2019. The inclusion criteria was as followed (1)acquired pathological diagnosis con rmed with gastric carcinoma by gastroscopic biopsy (2)underwent curative laparoscopic gastrectomy (3)the age of patients>18 years old. The exclusion criteria was as followed (1)received neoadjuvant chemotherapy before gastrectomy (2)R1/2 resection (3)diagnosed as gastric stump cancer (4)combined with distant metastasis liver colon ovary and etc. (5)underwent extended or palliative operation (6)incomplete data to follow-up in 30 days. Ultimately there were 309 patients enrolled in the retrospective analysis. The detailed ow-chart was shown in Figure.1. Written informed consent for usage of clinical records was granted by each patient as required by the Institutional Review Board at hospital in accordance with ethical guidelines from the Declaration of Helsinki in 1964.

Perioperative management
Elaborate case history analysis normative physical examination and routine preoperative laboratory measurements were performed. All elder patients above 50 years old would have to take in ultrasonic cardiogram and pulmonary function test to evaluate cardiopulmonary condition before surgery. Abdominal enhanced computed tomography and endoscopy together with tissue biopsy were carried out for overall assessment of gastric tumor. The standard surgical laparoscopic gastrectomy ensured with su cient resection margin was performed followed by Japanese gastric cancer treatment guidelines 2014 ver.4 18 which was either total or distal gastrectomy coupled with systematic lymphadenectomy abiding by the D level criteria. The alimentary tract reconstruction methods were usually employed as following We performed Roux-en-Y esophagojejunostomy after total gastrectomy with regard to distal gastrectomy we selected one of three gastrointestinal methods including Billroth I gastroduodenostomy Billroth II gastrojejunostomy and Roux-en-Y gastrojejunostomy. Based on preoperative and intraoperative condition patients were transferred to intensive care unit for postoperative treatment when necessary. For all of patients reasonable perioperative management was in line with the Enhanced Recovery After Surgery (ERAS) program including preoperative disease education shrinking fasting time intraoperative utility of minimally invasive techniques uid restriction avoided from overload postoperative early drainage removing off-bed mobilization and oral feeding until discharged [19][20][21] . Thereafter patients received with pathological diagnosis of advanced gastric carcinoma were recommended to adopt subsequent adjuvant chemotherapy.

Data Collection
We retrospectively collected clinical records of baseline characteristics laboratory data imaging scanning examination and pathological diagnosis from the database. Preoperative CONUT score was calculated from precise records of serum albumin level total lymphocyte count and cholesterol level whose blood samples were obtained within 3 days before surgery. Short-term prognosis was chie y considerated as postoperative complications that occurred within 30 days after surgery or before hospital discharge.
Based on the Clavien-Dindo classi cation system 22 we ranged postoperative complications from Grade I to Grade V with Grade I and II de ned as mild complications Grade III to Grade V de ned as major complications. Among mild complications abdominal or pelvic effusion was diagnosed via ultrasonic testing or computed tomography scan excluding the case of intra-abdominal infection. When the patients developed nontransient fever over 38.5℃ after surgery we highly suspected there existed infection and adopted effective antibiotic therapy. Delayed atus and defecation denoted sluggish resuscitation of gastrointestinal function and we had to prolong the usage of TPN(total parenteral nutrition). As for major complications severe active hemorrhage after surgery called for emergency treatment. When persistent fever and purulent drainage came out we considered there was intra-abdominal abscess inside. Other intractable major complications included anastomotic leakage duodenal stump stula and etc. Mortality was regarded as any death occurring from the date of operation up to 30 days after operation. Each individual data of cancer staging involving records of the primary and regional nodal extent of tumor and the absence or presence of metastases was evaluated based on TNM Classi cation of Malignant Tumors 8 th Edition published in a liation with the Union for International Cancer Control (UICC) and the American Joint Committee on Cancer(AJCC).

Statistical Analysis
All of the data was statistically analyzed using SPSS 21.0 (IBM Corp Armonk NY). Continuous variables were presented as mean (standard deviation) or median (interquartile range) whereas categorical variables were presented as number (percentage). The Student t test or Mann-Whitney U test/ Kruskal-Wallis H test was utilized for continuous variables depending on the normality of data distribution. The Pearsonχ 2 test or Fisher exact test was applied for categorical variables as appropriate. The predictive ability of potential factors for postoperative complications was evaluated by the Receiver Operating Characteristic curve. We elaborately chose the optimal cut-off value with reference to Youden index which was set as the value maximizing the sum of sensitivity and speci city. To identify independent risk predictors for postoperative complications all signi cant associated factors (P < 0.05) on univariate analysis were assessed for multivariate analysis by logistic regression. All of P value < 0.05 was considered as statistical signi cance.

Results
Receiver operating characteristic curve of CONUT score and the optimal cutoff value identi ed The demarcated values of CONUT score correlated with prognosis differed in recent studies 9, 12-17 .
Therefore we elaborately plotted the receiver operating characteristic curve of CONUT score based on postoperative overall complications ( Figure.2). The area under the curve(AUC) was 0.718 with sensitivity of 0.549 and speci city of 0.794. The positive predictive value was 52.3% and negative predictive value was 80.8%. Youden index was 0.343 and the optimal cutoff value was identi ed as 2.5.

Study population and baseline characteristics
According to the inclusion and exclusion criteria 309 patients were enrolled in the retrospective analysis  Table.2.

Correlations of the CONUT score and clinical characteristics
On account of the optimal CONUT score cutoff value all of patients were subdivided into the low CONUT group(score<2.5 N=214) and the high CONUT group(>2.5 N=95) (Table.2 Table 4.

Discussion
Malnutrition is de ned as "a state resulting from lack of intake or uptake of nutrition that leads to altered body composition (decreased fat free mass) and body cell mass leading to diminished physical and mental function and impaired clinical outcome from disease" 25 . Among patients with cancer the condition of low BMI and weight loss causing by inadequate nutritional intake with independent prognostic signi cance 26 is quite common and may be severe. Meanwhile the depletion of skeletal muscle as the main aspect of cancer-associated malnutrition a hallmark of cancer cachexia also does damage to physical function and treatment tolerance 27,28 . The systemic in ammation syndrome is frequently activated which varies in degree but impacts all relevant metabolic pathways including protein carbohydrate and lipid metabolism 29 . The circumstance of malnutrition is particularly common among patients with gastric cancer especially advanced gastric cancer. They always have to endure unwell symptoms such as early satiety anorexia and dysphagia caused by chronic obstruction of tumor mass and hemorrhagic anemia of malignant ulcers which ultimately leads to progressive weight loss compromised immunity and perishing malnutrition 30 .
As of today several nutritional assessments have been established with aim at identifying applicable parameters of malnutrition status such as NRS(Nutritional Risk Screening) MNA(Mini Nutritional Assessment) MUST(Malnutrition Universal Screening Tool) SGA(Subjective Global Assessment) from the consensus scheme of criteria proposed by the Global Leadership Initiative on Malnutrition(GLIM) 31 . The CONUT score was originally proposed to assess nutritional status of inpatients by Ignacio de Ulíbarri J in 2005 11 precisely calculated with serum albumin total lymphocyte count and cholesterol level which subsequently was deemed to be a validated nutritional assessment approach for predicting multiple cancer outcomes in recent years 9, 12-17, 32, 33 .
We carried out this retrospective cohort study to cast attentions to the role of CONUT score played in predicting the prognosis of patients with gastric cancer to be speci c short-term prognosis. The correlation between CONUT score and postoperative complications was estimated with data from electronic medical record.
Despite prolonged operating time and fewer harvested lymph nodes laparoscopy-assisted gastrectomy is associated with minimally invasive incision less blood loss earlier healing and shorter time to oral intake and hospital stay compared with open surgery 34 . Previous studies indicated that laparoscopic gastrectomy was likely to be an alternative procedure for patients with gastric cancer with the noninferiority of short-term and long-term outcomes 35,36 . Recently an open-label multicentre non-inferiority phase III randomized controlled trial in Japan supported the non-inferiority of laparoscopy-assisted distal gastrectomy compared with open distal gastrectomy for clinical stage I gastric cancer relapse-free survival 37 . Our study involved clinical records of 309 patients who underwent laparoscopy-assisted radical gastrectomy with written informed consent. A certain part of patients with gastric cancer started with laparoscopic exploration and then transferred to open gastrectomy extended or palliative operation according to the surgeon's judgement due to gross tumor volume deep tumor location or visible existence of tumor invasion or peritoneal metastasis which was deliberately excluded from our analysis.
Based on the receiver operating characteristic curve we identi ed the optimal cutoff value as CONUT score of 2.5 to distinguish patients at disparate risk for postoperative complications. The preoperative parameters constituting CONUT score revealed an authentic and steady nutrition status of patients and it was reasonable for preoperative assessment. When patients underwent operation various factors including surgical trauma intraoperative hemorrhage venous uid utilization postoperative TPN support might in uence their blood components and hemodynamics and bring about the variations of laboratory measurements for nutritional assessment.
In order to explore potential risk factors associated with postoperative complications we conducted univariate and multivariate analysis with preoperative parameters. Consistent with what we supposed the CONUT score served as an integrated nutritional index was signi cantly associated with postoperative complications in patients with gastric cancer undergoing laparoscopic surgery(P=0.012 OR=2.433 95%CI 1.218-4.862). In addition we also identi ed age(P=0.037 OR=2.237 95%CI 1.048-4.774) preoperative RBC(P=0.003 OR=0.356 95%CI 0.180-0.707) as independent risk factors for complications. Generally speaking elderly patients anemia and malnutrition played adverse roles in short-term prognosis with patients after radical gastrectomy for cancer which was parallel to prior studies 38,39 . By the way we excluded albumin total lymphocyte count and cholesterol from multivariate analysis so as to avoid duplication.
Based on the Clavien-Dindo classi cation system 22 we ulteriorly performed strati ed statistics of postoperative complications with low and high CONUT score. As a matter of fact some patients who suffered severe complications usually underwent multiple collateral complications. One patient got a sudden bellyache and subsequent fever with abdominal tenderness and rebound tenderness for example as a result of duodenal stump rupture developed grievous intra-abdominal abscess eetly and had to suffer second laparotomy with suture irrigation and drainage. Our detailed analysis indicated that higher scale of gastric cancer patients with high CONUT score developed postoperative complications especially mild complications. The condition of hypoalbuminemia decreased lymphocytes hypocholesterolemia was prevalent among malnourished patients with complications which meant delayed wound healing increased susceptibility to infection and prolonged parenteral nutrition support. That helped to explain why the percentage of patients with mild complications was signi cantly higher in high CONUT score group. Statistic difference was not present at patients group with major complications which needed larger scale of patients to further validation. Surgical site infections(SSIs) are infections of the incision or organ or space that occur after surgery combined with complex comorbidities and antimicrobial-resistant pathogens which increase the challenge and expenses of treatment 40 . Our analysis showed that there was no signi cant difference about SSIs whether located at surface incision or deep space. The respiratory complications after surgery contained pneumonia hydrothorax which occurred more frequently in the high CONUT group as same as Song Ryo et al reported 15 . We believed that was blamed for long stay in bed and infrequent cough and sputum of malnourished patients. Therefore the CONUT score acted to furnish an evaluation strategy for precise risk strati cation oriented at short-term outcomes before surgery which allowed to implement active nutritional intervention for malnourished patients.
Despite of our ndings there were still some limitations of our present study. First the single-center study took in a homogeneous cohort of patients with a xed surgery team. Second a retrospective study could not reject selection bias. At last the follow-up assessments of CONUT score after surgery were absent which resulted in the lack of dynamic observation of nutrition status. Therefore prospective multi-center studies should be warranted to con rm predictive signi cance of the CONUT score for short-term and long-term prognosis to validate the effectivity of preoperative nutritional intervention involving with the comparison of other commonly used and well considered nutritional assessments.

Conclusion
As a simple and feasible nutrition assessment tool the CONUT score plays a reliable role in predicting postoperative complications for patients with gastric cancer after laparoscopy-assisted gastrectomy which allows precise risk strati cation and preoperative nutritional intervention before surgery.

Competing interests
The authors declare that they have no competing interests.    Figure 1 A ow chart of the inclusion process for patients with gastric cancer.

Figure 2
Receiver operating characteristic curve showing the capacity of CONUT score for predicting postoperative overall complications.