Higher BMI Indicated Better Overall Survival Prognosis of Pancreatic Ductal Adenocarcinoma: A Real-World Study Based on 2,010 Cases


 Background: Whether body mass index (BMI) was associated with the overall survival (OS) of pancreatic ductal adenocarcinoma (PDAC) remained controversial and uncertain.Method: A total of 2,010 patients from single high-volume center were enrolled in the study. OS of PDAC patients was evaluated based on restricted cubic spline (RCS), propensity score (PS) and multivariable risk adjustment analyses. Result: BMI was discovered linear related with OS (total P=0.004, non-linear P=0.124). BMI was analyzed as categorical data based on X-tile software defined cutoffs and World Health Organization (WHO) recommended cutoffs, respectively. Adjusted with confounding covariates, higher BMI manifested as a positive prognostic predictor. (PXtile=0.003, PWHO=0.002) Furtherly, BMI was proven associated with OS in PS analysis. (UnderweightXtile vs. NormalXtile P=0.003, OverweightXtile vs. NormalXtile P=0.019; UnderweightWHO vs. NormalWHO P<0.001, OverweightWHO vs. NormalWHO P=0.024). It was also revealed that patients with higher BMI benefitted more from chemotherapy.(adjusted hazard ratio (aHR): UnderweightXtile: 0.565 (0.389-0.819), NormalXtile: 0.474 (0.395-0.567), OverweightXtile: 0.409 (0.337-0.496); UnderweightWHO: 0.613 (0.401-0.940), NormalWHO: 0.464 (0.387-0.557), OverweightWHO: 0.425 (0.353-0.512)). Conclusion: Among PDAC patients, higher BMI manifested as a favorable OS indicator, and the protective impact was probably based on chemotherapy administration. Patients with higher BMI were also observed with more chemotherapy administration and more OS benefits from chemotherapy.


Background
Pancreatic cancer, one of the most lethal malignancies, was estimated to cause 47,050 deaths in 2020. [1] The 5year survival rate is merely 9%. [1] Pancreatic ductal adenocarcinoma (PDAC) is the main pattern of pancreatic cancer, taking approximately 85%. [2] To improve and predict the overall survival (OS) of those patients was an everlasting hotspot for clinicians and researchers.
Body mass index (BMI), an easily accessible and inexpensive parameter, was ever reported associated with the incidence of pancreatic cancer. [3] However, in our previous study, it was incidentally discovered that there existed a relevance between higher BMI and better OS among PDAC patients. [4] And so far, the role of BMI in the longterm prognoses of PDAC patients still remained controversial. [5][6][7][8][9][10] Limited by sample size and retrospective research nature, those studies could not convincingly elaborate the role of BMI. In this study, relying on propensity score (PS) analysis and multivariable risk adjustment analysis, we discussed the association between BMI and OS of among 2,010 PDAC patients. Based on our results, further researches could be designed and excavated aiming at the mechanism beneath and the relationship between metabolic status and the prognosis of PDAC.

Data Collection
All eligible patients pathologically diagnosed as PDAC treated according to National Comprehensive Cancer Network (NCCN) guideline were consecutively enrolled from Pancreatic Disease Center, Ruijin Hospital A liated to Shanghai Jiao Tong University School of Medicine during 2013.1 to 2019.12. The exclusive criteria included: 1) Without BMI data at diagnosis. 2) Without complete oncological data. 3) Without regular follow-up. 4) heterogenous carcinoma. Finally a total of 2,010 patients were admitted in our research. (Supp. Figure 1) The study protocol was approved by the institutional review board at the authors' a liated hospital. The local ethics committee waived the need for informed consent because the study was observational and retrospective.
The study was undertaken according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [11] and in accordance with the latest version of the Declaration of Helsinki.

De nition
Height and weight information were recorded when diagnosed. Body mass index was calculated as weight in kilograms divided by the square of height in meters. The cutoffs de ned by X-tile were 18.9 kg/m 2 & 23.3 kg/m 2 for trinomial categorizing. Also BMI was categorized in three groups based on the cutoffs (18.5 kg/m 2 and 23 kg/m 2 ) de ned by World Health Organization (WHO) for Asians. [12] Patients below 18.9 kg/m 2 , between 18. coherence of surgical intervention and TNM stages, surgical intervention was hided in multivariable Cox models and PS analyses. The knot number for RCS curves was set as 5 because of the sample size. X-tile was utilized for determining the cut-off value. Association between categorical data was assessed using c2 test, while association between continuous data using Spearman rank test, and association between categorical data and continuous data using Kruskal-Wallis test. For interaction evaluation, product term in Cox model was utilized to assess the multiplicative interaction, while the relative excess risk due to interaction (RERI), the attributable proportion due to interaction (AP) and the synergy index (S) were used for the assessment of additive interaction.
To eliminate the imbalance of chemotherapy administration among BMI categories, inverse probability of treatment weighting (IPTW) and standardized mortality ratio weighting (SMRW) were also introduced in for further analysis. The matching parameters included age, gender, ASA physical status, CA199, total bilirubin (TB), fasting blood glucose (FBG), albumin (ALB), biliary drainage, stage, differentiation and chemotherapy administration.
Statistical analysis of this study was performed by SPSS (IBM SPSS Statistics 26.0) , R Studio and X-Tile software. P< 0.050 was regarded statistically signi cant.

Patients and Characteristics
A total of 2,010 patients were enrolled in this study. (

Relationship of BMI with OS
The association of BMI at diagnosis and OS is depicted in RCS curves adjusted by Cox model. (Figure 1). The P value of total correlation and non-linear correlation were 0.004 and 0.124 respectively, suggesting that there existed a negatively correlated linear relationship between BMI and OS. The sharp slope was observed at the beginning of the curve, and the curve tends to be smooth when BMI exceeded around 23 kg/m 2 .
BMI was a Positive Prognostic Predictor in Multivariable Risk Adjustment Analysis.   Propensity Score Analysis Revealed the Role of BMI Additionally Not only between Underweight category and Normal category, IPTW and SMRW analyses were performed between Overweight category and Normal category as well. The characteristics of weighted cohorts were shown.
( Table 4, Supp. Table 4, Supp. Table 5, Supp. Table 6) It was discovered that BMI was still an independent prognostic predicter with chemotherapy factor balanced. (Figure 3, Figure 4 Discussion BMI, a simple, inexpensive and readily available parameter, was broadly applied in the evaluation of metabolic diseases, endocrinal diseases, oncological diseases and so on. [13][14][15][16][17] The role of BMI in pancreatic cancer oncogenesis was studied and discussed in previous studies. Higher BMI was ever discovered related with increased incidence of pancreatic cancer. [3] However, for patients diagnosed as pancreatic cancer already, how BMI in uenced their prognoses was still unclear and controversial.
In RCS analysis, a negative linear relationship between BMI and OS was observed. There was a strong association with higher risk of death in underweight patients when compared to Normal group individuals. The curve tends to be smooth when BMI above around 23 kg/m 2 .
Because the included patients were treated according to NCCN guideline, the surgical intervention was closely associated with TNM stages. Therefore we hided the surgical intervention in multivariable Cox models and PS analyses. With univariate analysis, the tumor location and the diagnostic year were ltered out for multivariable analysis. In multivariable risk adjustment analysis, TNM stages and differentiation were signi cantly related with OS as expected. Both categorical BMI were also observed signi cantly related with OS. FBG and ALB were additional 2 parameters to evaluate the nutrition condition as the supplementation of BMI, whereas neither signi cant impact on the OS (Table 3) nor correlation with BMI (FBG: P=0.087, ALB: P=0.060) was observed in this study.
Chemotherapy administration was discovered related with both continuous BMI and categorical BMI. No signi cant association or interaction was observed between TNM stages and BMI. More patients were treated with chemotherapy in higher BMI groups. Strati ed with BMI level, we compared the impact of chemotherapy in different categories. And chemotherapy was discovered with more positive impact on OS in higher BMI categories. Strati ed with chemotherapy administration, the protective impact of BMI was only observed in the chemotherapy subcategory, indicating that BMI might prolong the OS via the administration of chemotherapy.
Given that there existed interaction and association between BMI and chemotherapy, IPTW analysis was introduced for further certi cation of BMI's role. With pairwise comparisons, compared with normal weight patients, underweight was discovered a risk factor while overweight a protective factor on OS. The consistence of IPTW and SMRW analysis results indicated that no signi cant confounding/interacting factor was excluded.
No signi cant differences of chemotherapy administration were observed anymore in different pairs.
Some limitations in this study should be stressed out. First, though with PS analysis and multivariable risk adjustment, the nature of retrospective single center data still limited the reliability and extrapolation of this study. A multi-center prospective study containing BMI analysis could help well facilitate the results. Secondly, given that all enrolled subjects were Chinese Han population, more researches aiming at patients from different districts and with different races should be planned.
In this study, we studied and elaborated the relationship between BMI and OS of PDAC patients based on IPTW, RCS and multivariable risk adjustment analyses. Higher BMI generally indicated better OS among PDAC patients.
More patients were treated with chemotherapy in higher BMI groups. And for different BMI categories, higher BMI indicated better OS bene ts from chemotherapy, which could be the possible mechanism of higher BMI indicating better OS. The relevance between BMI and chemotherapy demanded further studies. We are also curious of the association among BMI changes, chemotherapy and OS. And further researches targeting the mechanism beneath and the relations between metabolic status and malignant diseases could be organized and performed.

Declarations
Ethics approval and consent to participate: All clinical decisions or interventions were in accordance with National Comprehensive Cancer Network (NCCN) guideline for pancreatic cancer. The ethics committee of Shanghai Jiao Tong University School of Medicine a liated Ruijin Hospital waived the need for informed consent because the study was observational and retrospective. The ethics committee of Shanghai Jiao Tong University School of Medicine a liated Ruijin Hospital approved this retrospective study including patients from Pancreatic Disease Center .

Consent for publication:
All co-authors declared no con icts on publication.
Availability of data and materials: The raw data should be inaccessible to public. The statistic data and results could be available. Anyone requesting data could contact fnz01b74@rjh.com.cn.
Competing interests: All co-authors declared no con icts of interests. Funding: This study was supported in part by the National Natural Science Foundation of China (Grant number: 81871906