The value of the pretreatment albumin/globulin ratio in predicting the long-term survival in colorectal cancer.

Publication/Presentation Date

12-1-2013

Abstract

BACKGROUND: Low serum albumin was found as a predictor of long-term mortality in colorectal cancer (CRC) patients. Our aim was to evaluate the value of the pretreatment albumin/globulin ratio (AGR) to predict the long-term mortality in CRC patients.

METHODS: Patients were included if they had comprehensive metabolic panel (CMP) before treatment (surgery or chemotherapy). The albumin/globulin ratio, routinely reported in CMP, is calculated [AGR = Albumin/(Total protein - Albumin)]. Patients were divided into three equal tertiles according to their pretreatment AGR. The primary outcome was cancer-related mortality, which was obtained from our cancer registry database.

RESULTS: A total of 534 consecutive CRC patients had pretreatment CMP. The 1st AGR tertile had a significant higher 4-year mortality compared to the second and third AGR tertiles (42 vs. 19 and 7 %, p < 0.0001 according to Fisher's exact two-tailed test). In the multivariate model, AGR remained an independent predictor of survival with 75 % decrease in mortality among the highest AGR tertile in comparison to the lowest AGR tertile, p < 0.0001. In the subset of 234 patients with normal serum albumin (albumin of >3.5 g/dl), serum AGR continues to be an independent predictor of cancer-related mortality with an adjusted hazard ratio of the third tertile compared to the first tertile equal to 0.05 (95 % confidence interval 0.01-0.33, p = 0.002).

CONCLUSION: Low AGR was a strong independent predictor of long-term cancer-specific survival among colorectal cancer patients. Additionally, among the patients with normal albumin (>3.5 g/dl), patients with lower globulins but higher albumin and AGR levels had better survival.

Volume

28

Issue

12

First Page

1629

Last Page

1636

ISSN

1432-1262

Disciplines

Medicine and Health Sciences

PubMedID

23857599

Department(s)

Fellows and Residents

Document Type

Article

Share

COinS