Prognostic potential of nutritional risk screening and assessment tools in predicting survival of patients with pancreatic neoplasms: a systematic review

Mengxia Yu,Xiaoxuan Li,Mingxia Chen,Linglong Liu,Tianying Yao,Jiarong Li,Wang Su

DOI: https://doi.org/10.1186/s12937-024-00920-w

2024-02-05

Nutrition Journal

Abstract:The nutritional evaluation of pancreatic cancer (PC) patients lacks a gold standard or scientific consensus, we aimed to summarize and systematically evaluate the prognostic value of nutritional screening and assessment tools used for PC patients.

nutrition & dietetics

What problem does this paper attempt to address?

### What problem does this paper attempt to solve?

This paper aims to solve the problem of the lack of a gold standard or scientific consensus in nutritional assessment for patients with pancreatic cancer (PC). Specifically, the author hopes to summarize and evaluate the prognostic value of nutritional screening and assessment tools for pancreatic cancer patients through systematic review and meta - analysis. The main objectives of the study include:

1. **Evaluate the predictive ability of different nutritional screening and assessment tools for all - cause mortality in pancreatic cancer patients**: By comparing multiple nutritional assessment tools (such as Prognostic Nutritional Index (PNI), Geriatric Nutritional Risk Index (GNRI), Controlling Nutritional Status Score (CONUT), Nutrition Risk Screening (NRS2002) and Glasgow Prognostic Score (GPS), etc.), determine which tools are most effective in predicting the survival rate of pancreatic cancer patients.

2. **Determine the best nutritional assessment tool**: By comparing these tools, find the most suitable nutritional assessment tool for pancreatic cancer patients, so as to provide a reference for clinical application.

3. **Explore the relationship between nutritional status and the prognosis of pancreatic cancer patients**: Verify whether poor nutritional status is an independent predictor of all - cause mortality in pancreatic cancer patients and explore the underlying mechanisms.

### Main conclusions

Studies have shown that poor nutritional status is significantly associated with all - cause mortality in pancreatic cancer patients. In particular, CONUT (HR = 1.978, 95% CI 1.345 - 2.907, P = 0.001) shows the highest risk ratio in predicting the mortality of pancreatic cancer patients and may become the best nutritional assessment tool. In addition, the study also found that the Modified Glasgow Prognostic Score (mGPS) cannot significantly predict the all - cause mortality of pancreatic cancer patients.

### Brief description of methods

- **Literature search**: Relevant studies from January 2010 to December 2023 were retrieved from major databases such as PubMed, Embase, Web of Science and Cochrane Library.
- **Inclusion criteria**: Included pancreatic cancer patients aged ≥ 18 years, studies that used the nutritional screening tools recommended by ESPEN 2017 to assess nutritional risk, and studies that reported all - cause mortality.
- **Data analysis**: Meta - analysis was performed using STATA 14.0, and statistical analysis was performed when three or more studies used the same tool.

### Significance

This study provides an important reference for the nutritional screening and assessment of pancreatic cancer patients, which is helpful for early identification and intervention of malnourished patients, thereby improving clinical outcomes, shortening hospital stay and reducing medical costs.