Changes in serum prealbumin as a marker for nitrogen balance in surgical oncology patients

Background: Serum prealbumin has long been used as a marker of nutritional status. However, prealbumin is a negative acute phase reactant influenced by several non-nutritional-related factors including surgery, infection, and cancer. An increasing prealbumin has been correlated with a positive nitrogen balance in general surgery patients receiving parenteral nutrition (PN) with 88% specificity and 70% sensitivity. To date, no trial has evaluated the effect of concurrent cancer and surgery on the value of prealbumin in predicting nitrogen balance. Methods: This study is a concurrent retrospective design of post-operative patients (≥ 19 years of age) identified by the nutrition support service who received PN for ≥ 5 days, had a baseline and follow-up serum prealbumin and C-reactive Protein (CRP) measured, as well as a 24-h urinary urea nitrogen (UUN) performed between days 5–10 of PN. Exclusion criteria include anuric renal failure, Child–Pugh Class C liver failure, pregnancy, and corticosteroid use. Prealbumin was correlated to nitrogen balance, measuring sensitivity, specificity, and negative and positive predictive values. Information was collected regarding patient demographics and presence or absence of metastatic cancer. Results: Thirty patients were identified and evaluated for this study from December 1st, 2010 to July 15th, 2011. Patients included in the study had a mean age of 57 years old (range 20–82), 53% male, with a mean weight of 84 kg (range 42–140) and body mass index (BMI) of 29 kg/m² (range 14.9–56.8). The mean daily caloric dose of PN per actual body weight was 21 kcal/kg (range 10–34) and the mean daily protein dose was 1.4 g/kg (range 1–2). Forty seven percent of patients were obese (BMI > 30 kg/m²) and were prescribed high-protein hypocaloric PN. The most common indication for PN was post-operative ileus (23/30 patients). 24-h urine collection for UUN was performed on average of day 8 after PN initiation (range 5–10 days). Nitrogen balance as calculated from 24-h UUN was positive in 17/30 patients. A positive prealbumin change of greater than 2.8 mg/dL was found to have a statistically significant association with positive nitrogen balance (p = 0.02). At the cut off level of positive 2.8 mg/dL, the likelihood of a positive nitrogen balance had a sensitivity of 82% (95% confidence interval (CI) 64–100%); specificity of 62% (95% CI 35–88%); positive predictive value of 74% (95% CI 54–93%); negative predictive value of 73% (95% CI 46–99%). No absolute value for prealbumin level (e.g., > 20 mg/dL) was found to be a significant predictor of positive nitrogen balance. CRP levels at initiation of PN were significantly elevated with a mean level of 147 mg/dL. Conclusion: These results indicate a positive change in serum prealbumin (> 2.8 mg/dL) has sufficient sensitivity (82%) to predict positive changes in nitrogen balance in the surgical oncology population. However, the low specificity (62%) makes it less useful in predicting a negative nitrogen balance. Absolute prealbumin levels were greatly affected by inflammation, as evidenced by CRP levels, and single values were not useful in predicting positive nitrogen balance. Clinical relevancy: Positive changes in serum prealbumin levels have previously been associated with a positive nitrogen balance (NB) in surgical patients receiving parenteral nutrition (PN); however, it is unclear if this is true in oncologic surgery patients. This study highlights how changing levels of serum prealbumin and C-reactive protein correlates to NB for cancer patients in the post-operative period requiring PN. Changes in prealbumin levels from baseline showed sufficient sensitivity, but not specificity to utilize routinely for predicting NB in this population.