The Role of Geriatric Nutritional Risk Index in Predicting Postoperative Pulmonary Complications in Elderly Lung Cancer Patients Undergoing Surgical Resection

The relationship between immunonutritional status (eg. Prognostic nutritional index [PNI] and Controlling Nutritional Status [COUNT] score) and risk of postoperative pulmonary complications (PPCs) after surgical resection of lung cancer had reported before. However, another immunonutritional parameter-Geriatric Nutritional Risk Index (GNRI)-had never explored. To address this issue, in this study we retrospectively analyzed patients’ characteristics and PPCs in a cohort of lung cancer patients who were treated by surgical resection at our center. The clinical utility of patients’ characteristics for predicting PPCs was evaluated by receiver operating characteristic curve analysis and the Youden index. Univariate and multivariate analysis were applied to nd the most important factors.


Introduction
Lung cancer is the leading cause of cancer-related deaths both worldwide and in China 1, 2 . Surgery is the rst-choice treatment for early-stage lung cancer 3 . The mortality and postoperative pulmonary complication (PPC) rates within 30 days after lung cancer resection are about 4.4% and 20-40%, respectively 4,5 . Many factors in uence the occurrence of PPCs, including preoperative pulmonary function and immunonutritional status 6, 7 .

Page 3/18
The immunonutritional status of lung cancer patients were an important predictor of clinical outcome.
Tumor cells alter the patient's metabolism and grow rapidly by usurping nutrients from the patient. It was previously reported that weight loss and low albumin (ALB) level were associated with the occurrence of PPCs after lung cancer resection 8 . Various instruments such as modi ed Glasgow prognostic score (mGPS), prognostic nutritional index (PNI), Controlling Nutritional Status (CONUT) score, and Geriatric Nutritional Risk Index (GNRI) have been developed to assess the preoperative immunonutritional status of cancer patients. The relationship between immunonutritional status (eg. PNI, CONUT score) and risk of PPCs after surgical resection of lung cancer had reported before 9,10 . However, another immunonutritional parameter-GNRI -had never explored.
To address the above point, in this retrospective cohort study we evaluated the clinical utility of GNRI for predicting PPCs in elderly lung cancer patients treated by surgical resection.

Patient selection
We reviewed the patient database of Shanghai Chest Hospital, Shanghai Jiao Tong University for lung cancer cases treated by resection in 2012 and 2013. The study was conducted according to the principles outlined in the Declaration of Helsinki, and the Ethics Committee of Shanghai Chest Hospital approved the protocol (no. KS1924). All patients signed the informed consent form before participating.
The inclusion criteria were as follows: (1) age between 60 and 80 years old; (2) underwent lung cancer resection at Shanghai Chest Hospital, Shanghai Jiao Tong University; (3) physical and laboratory examinations were performed within 1 week before the surgery. Exclusion criteria were as follows: (1) incomplete clinical data; (2) severe cardiac insu ciency as determined by electrocardiography and echocardiography; (3) previously underwent lung resection. Basic demographic and clinical information were collected for each patient including age, sex, smoking history, physical and laboratory examination results, major comorbidities, surgery type, pathologic diagnosis, and PPCs.

Evaluation of immunonutritional status
A standard physical examination was conducted and data including height and weight were recorded.
Laboratory examinations were also performed before surgery including a routine blood test, hepatic and renal function tests, and measurement of serum lipids. The immunonutritional parameters were determined as follows.
GNRI was calculated based on serum ALB level and body weight according to the formula GNRI=1.489×serum ALB level (g/L) + 41.7×preoperative weight/ideal weight (kg) 11 .

Surgery performance
Four surgeons and their respective teams performed all operations, using similar operative and perioperative procedures. All patients received similar medical care following the guidelines of enhanced recovery after surgery 12 . Preoperative medical care included health education, smoking cessation, preoperative pulmonary rehabilitation, comorbidities management, and prevention of deep venous thrombosis. Intraoperative management included general anesthesia combined with endotracheal intubation anesthesia and paravertebral nerve block under ultrasound guidance; precise controlled infusion of opioid frugal general anesthesia technology was used to reduce the dosage of muscle relaxants. Postoperative management included post-operative pulmonary rehabilitation and nutritional support. Each patient was given prophylactic antibiotics for 2-3 days after surgery.

Statistical analysis
We used SPSS v20.0 software (IBM, Armonk, NY, USA) for statistical analyses. Measurement data are expressed as mean ± SD. Differences between groups were evaluated by one-way analysis of variance and with the t test. Count data were analyzed with Pearson's chi-squared test and Fisher's exact probability test. Receiver operating characteristic (ROC) curve analysis and the Youden index were used to assess the predictive value of immunonutritional parameters for PPCs. Differences were considered statistically signi cant for P values <0.05.

Characteristics of the study population
A total of 128 patients met the inclusion criteria for this study (Fig. 1). Most patients were male (73.43%), underwent thoracotomy (93.75%), had no smoking history (45.31%), and were diagnosed with adenocarcinoma (54.69%) ( Table 1). There were 29 patients with 37 PPCs in the whole cohort, the incidence of PPCs was 22.48% (29/128). The most frequent PPCs were persistent air leakage or pleural effusion and pneumonia. Six patients had more than one PPC (Table 2).

Relationship between immunonutritional status and PPCs
We calculated the correlation between patient characteristics and PPCs. Signi cant differences in sex, GNRI, FEV1%, LY% were found between the PPC and non-PPC groups (all P<0.05). The difference in pathology between the 2 groups showed borderline statistical signi cance (P=0.052). (Table 3) Predictors of PPCs An ROC curve analysis was performed with GNRI, BMI, FEV1%, FVC and LY% to evaluate the predictive value of these parameters for PPCs. GNRI, FEV1%, FVC and LY% were found to be statistically signi cant based on the area under the ROC curve. We determined the best cutoff value of each parameter and calculated the corresponding sensitivity and speci city, and found that GNRI, FEV1% and LY% had similar diagnostic value. (Table 4 and Fig. 2). The we conducted the multivariate analysis. For the continuous variable, we used the cutoff value to de ne "low" or "high". In the analysis, the variable of sex, surgery method, pathology, GNRI FEV1%, FVC% and LY% were included. Finally, GNRI, sex, LY% and FEV1% were ltered to be correlated to PPCs of elderly lung cancer patients received surgery therapy.

Discussion
PPCs are a major concern in the surgical treatment of lung cancer. Pulmonary function is one of the main predictors of PPCs, and some guidelines recommend the evaluation of preoperative immune and nutritional status 15 . The PPC rate in our study is about 22%, which is similar to previous reports 4,5 . The main PPC is persistent air leakage or pleural effusion and pneumonia, which are affected by immune and nutritional status. Pneumonia is common in patients who are malnourished or immunocompromised.
The principal causes of persistent air leakage or pleural effusion are pneumothorax or hydrothorax; persistent hydrothorax may also result from pneumonia or hypoproteinemia. Thus, in theory, immune and nutritional parameters can be predictors of PPCs.
Many researches have explored the clinical meanings of single immune and nutritional in predicting PPCs. Among all, ALB and body weight related index are most studied. ALB is synthesized by hepatocytes and maintains vascular osmotic pressure; transports hormones, fatty acids, or other compounds; and regulates blood pH. Serum ALB concentration is often used as a biochemical marker of long-term nutritional status and re ects visceral protein status 16 . A correlation between ALB insu ciency and the development of early PPCs has been reported previous 17,18 . Body weight loss is also revealed correlated to high rate of PPCs. In Busch E's report, weight loss and ALB level are linked to PPCs following lung cancer resection 8, 19-21 . Globulin and LY% re ect the immune status of patients 22 . The cellular immunitymainly regulates immune function through T lymphocyte subsets 23 and in tumor surveillance, immunoglobulin contributes to target cell phagocytosis 24 .
Immunonutritional parameters are calculated based on multiple indices related to immune function (CRP, peripheral lymphocyte count, etc) and nutrition (BMI and ALB). Those parameters show to be useful for detecting sarcopenia, which has been linked to the survival of lung cancer patients 25  Among the clinical instruments included in our analysis, GNRI were found to have predictive value for PPCs, with moderate diagnostic value. GNRI are calculated based on the original index and are continuous variables. In previous study, GNRI is suggested to be not inferior to nutrition screening tool of Mini Nutritional Assessment 35 . GNRI is recognized to correlated with sarcopenia status 36 . GNRI has been widely applied in digestive tract diseases. In gastric cancer, colorectal cancer and esophageal cancer, GNRI is de ned as a prognostic factor [37][38][39] . Previous studies reveal GNRI as a prognostic factor both in early lung cancer patients treated with surgery operation 40,41 and advanced stage lung cancer [42][43][44] .
Besides, GNRI can predict the treatment response of immunotherapy 45,46 . Apart this, GNRI is suggested to be good predictor for postoperative complications after abdominal surgery or gastrointestinal malignancy 47, 48 . But, can this parameter be good predictor in elderly lung cancer surgery operation patients? There is no answer. Thus we compare the GNRI and other patients characteristic and nd GNPI has moderate sensitivity for detecting PPCs; that is, patients with low values are more likely to develop PPCs after lung cancer resection and require special care. Meanwhile, it has moderate speci city; this means that patients with a high value are not likely to experience PPCs, and routine care is therefore su cient. Thus, risk strati cation is possible based on these few parameters, which would maximize the use of limited medical resources. As mention above, ALB and body weight are the most important simple index for PPCs. When calculating GNRI, both ALB and body weight are enrolled, thus it may has excellent diagnostic value than other patient characteristics.
There were some limitations to this study that should be noted. Firstly, because of the retrospective single-center design, selection bias could not be avoided. Secondly, because the sample size is relatively small, statistical bias was inevitable.
Nonetheless, our ndings demonstrate that immunonutritional parameters can predict PPCs following lung cancer resection and can be used to identify high-risk patients who would bene t from preventive interventions. Among all, GNRI has the best performance.

Declarations
Ethics approval and consent to participate The study was conducted according to the principles outlined in the Declaration of Helsinki, and the Ethics Committee of Shanghai Chest Hospital approved the protocol (no. KS1924). All patients signed the informed consent form before participating.

Consent for publication
All author approved this publication.

Availability of data and materials
The datasets generated for this study will be made available from the corresponding author on reasonable request.

Competing interests
The authors declare that the research was conducted in the absence of any commercial or nancial relationships that could be construed as a potential con ict of interest.