BACKGROUND

No study has clarified the clinical significance of albumin-bilirubin (ALBI) grade in a large cohort of pancreatic cancer patients.

METHODS

A total of 1006 consecutive patients diagnosed with pancreatic cancer and deemed eligible for surgical resection were analyzed. The ALBI score was calculated as: ALBI score = (log bilirubin [µmol/L] × 0.66) + (albumin [g/L]  × - 0.0852). ALBI grade was assigned as grade 1, 2a, 2b, and 3. ALBI grade 1 was assigned to the ALBI low group (N = 566), and grades 2a, 2b, and 3 to the ALBI high group (N = 440).

RESULTS

The primary lesion could not be resected in 129 patients. Among all patients, overall survival (OS) was significantly worse in the ALBI high group than in the ALBI low group (P = 0.024). Overall, 877 patients underwent pancreatectomy. In these patients, the ALBI high group was associated with high CA19-9 level (P < 0.001), lower morbidity rate (P < 0.001), and pancreatic head tumor (P = 0.001). Patients' OS after resection was significantly worse in the ALBI high group than in the ALBI low group (P < 0.001). Cox proportional hazard analysis revealed ALBI grade as an independent predictor for prognosis (hazard ratio, 1.33; P = 0.015). Even in the CA19-9 negative patients, OS was significantly worse in the ALBI high group than in the ALBI low group (P = 0.046).

CONCLUSIONS

The ALBI grade is a clinically useful predictor for prognosis in pancreatic cancer patients.