Best definition of malnutrition before cancer surgery varies by tumor type

Last Updated: 2020-03-20

By Will Boggs MD

NEW YORK (Reuters Health) - Malnutrition definitions used to identify patients who need prehabilitation before oncologic surgery should be specific to each patient's cancer type, researchers suggest.

"Though malnutrition (across all definitions) is a risk factor for major morbidity and mortality after surgery, the optimal definition of malnutrition varied by cancer type," Dr. Nicholas P. McKenna of Mayo Clinic, in Rochester, Minnesota, told Reuters Health by email. "We also found it interesting that though obesity is generally considered a risk factor before surgery, it had less of an effect than the various definitions of malnutrition."

The American College of Surgeons (ACS) Strong for Surgery program encourages rehabilitation regimens before surgery in patients with multiple comorbidities. With multiple definitions of malnutrition, identifying which patients would benefit most from prehabilitation is difficult.

Particularly for patients with cancer, it remains unclear whether the best approach to identify malnutrition is one-size-fits-all or if cancer-type-specific definitions would be better.

Dr. McKenna and colleagues used ACS NSQIP data from 2005 to 2017 to determine the optimal definition of malnutrition before major oncologic resection for six cancer types (colorectal, esophageal, gastric, liver, lung, and pancreatic).

They used three classification systems (European Society for Clinical Nutrition and Metabolism, or ESPEN, the NSQIP variable of >10% weight loss during the previous six months, and the WHO BMI classification system) to divide patients into seven mutually exclusive and hierarchical nutrition categories: severe malnutrition; ESPEN 1; ESPEN 2; NSQIP; mild malnutrition; obese; and no malnutrition.

Overall, 16% of patients met one of the criteria for malnutrition; 0.6% had severe malnutrition, 1% had ESPEN 1, 2% had ESPEN 2, 6% had NSQIP and 6% had mild malnutrition. Over half (54%) had a normal preoperative nutritional status, and the remaining 31% were obese.

The overall prevalence of malnutrition ranged from a low of 14% of patients with colorectal cancer to a high of 28% of patients with pancreatic cancer.

Mortality and major morbidity varied significantly across nutrition groups, with mortality most common in the severe-malnutrition group and major morbidity most common in the severe-nutrition and ESPEN 1 groups, the researchers report in the Journal of the American College of Surgeons.

Severe malnutrition was the strongest predictor of mortality and major morbidity for colorectal cancer and of mortality for esophageal cancer. ESPEN 1 best predicted mortality and major morbidity for gastric cancer and lung cancer, whereas NSQIP was the strongest predictor of mortality in liver cancer. All six classifications of abnormal nutrition performed similarly for predicting morbidity and mortality in pancreatic cancer.

"We believe that identification of patients for prehabilitation should be cancer-specific," Dr. McKenna said. "This will enable the identification of patients who are at highest risk of postoperative morbidity and mortality and would likely benefit the most from prehabilitation. It will also avoid overtreatment of patients who may not require prehabilitation and can go to surgery sooner."

"We hope that physicians take away that multiple definitions of malnutrition exist and involve both body mass index and weight loss," he said. "We also hope they will apply cancer-specific definitions of malnutrition to their patients in the preoperative setting, and use these when deciding which patients warrant consideration of prehabilitation."

Dr. Jann Arends from the Faculty of Medicine at the University of Freiburg, in Germany, who is a member of the ESPEN expert group that developed recommendations for action against cancer-related malnutrition, told Reuters Health by email, "Weight loss is a clear predictor of worse clinical outcome and should be seen as an indication for a prehabilitation program. I am not convinced that different definitions should be selected for different cancer types. Evidence for this cannot be read from this article."

He noted that there is considerable overlap in the odds ratios for mortality associated with various malnutrition definitions. "I would like to have seen statistical comparisons proving that in any cancer entity any definition in fact yields a significantly higher odds ratio," Dr. Arends said.

SOURCE: https://bit.ly/3cKJY7G Journal of the American College of Surgeons, online February 26, 2020.