THE ALBUMIN-GLOBULIN RATIO PREDICTING MORTALITY IN GASTRIC CARCINOMA

ASUDE AKSOY¹*, SERDAR DURAK² , TURKAN OZTURK³ , NILUFER AVCI⁴ , YALCIN CIRAK⁵ , AYSE NUR DEGER⁶ , YILMAZ BILGIC⁷ , SULEYMAN SERDAR KOCA² , METE OZCAN⁸ , SEVIL GOKTAS⁹ , AHMET BOZDAG¹⁰

¹ *Firat University, Medical Faculty, Medical Oncology - ² Firat University, Medical Faculty Internal Medicine - ³ Medical Oncology, Medical Faculty, Karadeniz Teknik University - ⁴ Bursa Ali Osman Sonmez Oncology Hospital, Medical Oncology - ⁵ Istanbul Bahcesehir University, Medical Faculty, Medical Oncology - ⁶ Dumlupınar University, Medical Faculty, Pathology - ⁷ Inonu University, Medical Faculty, Gastroenterology - ⁸ Firat University, Medical Faculty, Biophysic - ⁹ Denizli State Hospital, Medical Oncology - ¹⁰Elazig Education and Research Hospital, General Surgery, Turkey

Introduction: The potential effects of serum albumin and albumin-globulin rate on survival were investigated in patients with gastric carcinoma.

Materials and methods: The biochemical values of 204 patients under chemotherapy evaluation were obtained retrospectively from archive data. Albumin-globulin ratio values were calculated using albumin/total protein-albumin formula. The values of albumin-globulin ratio and albumin were categorized as 3 groups using receiver operating characteristic analysis and a descending order.

Results: The calculated cut-off value for albumin-globulin ratio was 1.14: 34.4-79.7 by receiver operating characteristic. The median survival was 20.7 months (15.9-25.5), 20.1 months (15.1-25.1), 16 months (11.8-20.1 ), respectively in 1st, 2nd, 3rt groups for albumin-globulin ratio (p=0.46). The median overall survival was 18.3 months (12.3-24.2), 21.6 months (16.8-26.3), 18.01 months (13.5-22.4) in 1st , 2nd, 3rt groups for albumin, respectively (p=0.52). There were no statistical significant differences by KaplanMeier method for albumin-globulin ratio, albumin p>0.05. The effects of clinicopathological and biochemical characteristics on mortality were assessed by Cox-proportional hazard regression analysis. The independent predictors of mortality were observed high grade, the presence of perineural invasion of the tumor, high nodal involvement (p:0.01, p:0.001, p:0.01), respectively.

Conclusion: Albumin-globulin ratio and albumin values alone may not be useful parameters to predict mortality and survival in patients with gastric carcinoma