The relationship between albumin, other plasma proteins and variables, and age in the acute phase response after liver resection in man.

A large series of plasma albumin (ALB, g/dl) and simultaneous blood and clinical measurements were prospectively performed on 92 liver resection patients, and processed to assess the correlations between ALB, other plasma proteins, additional variables and clinical events. The measurements were performed preoperatively and at postoperative day 1, 3 and 7 in all patients, and subsequently only in those who developed complications or died. In patients who recovered normally ALB was 4.3 +/- 0.4 g/dl (mean +/- SD) preoperatively, 3.7 +/- 0.7 at day 1 and 3, and 3.9 +/- 0.4 at day 7. In patients with complications its decrease was more prolonged. In non-survivors it was 3.4 +/- 0.4 preoperatively, 3.0 +/- 0.4 at day 1, and then decreased further. Regression analysis showed direct correlations between ALB and pseudo-cholinesterase (CHE, U/l, nv 5300-13000), cholesterol (CHOL, mg/dl), iron binding capacity (IBC, mg/dl), prothrombin activity (PA, % of standard reference) and fibrinogen, an inverse correlation with blood urea nitrogen (BUN, mg/dl) for any given creatinine level (CREAT, mg/dl), and weaker direct correlations with hematocrit, other variables and dose of exogenous albumin. An inverse relationship found between ALB and age (AGE, years) became postoperatively (POSTOP) also a function of outcome, showing larger age-related decreases in ALB associated with complications (COMPL: sepsis, liver insufficiency) or death (DEATH). Main overall correlations: CHE = 287.4(2.014)(ALB), r = 0.73; CHOL = 16.5(1.610)(ALB) (1.001)(ALKPH), r = 0.71; IBC = 68.6(1.391)(ALB), r = 0.64; PA = 13.8 + 16.0(ALB), r = 0.51; BUN = 21.3 + 20.2(CREAT) - 6.2(ALB), r = 0.91; ALB = 5.0-0.013(AGE) - {0.5 + 0.003(AGE)( COMPL ) + 0.012(AGE)( DEATH )}( POSTOP ), r = 0.74 [p < 0.001 for each regression and each coefficient; ALKPH = alkaline phosphatase, U/l, nv 98-279, independent determinant of CHOL; discontinuous variables in italics label the change in regression slope or intercept associated with the corresponding condition]. These results suggest that altered albumin synthesis (or altered synthesis unable to compensate for albumin loss, catabolism or redistribution) is an important determinant of hypoalbuminemia after hepatectomy. The correlations with age and postoperative outcome support the concept that hypoalbuminemia is a marker of pathophysiologic frailty associated with increasing age, and amplified by the challenges of postoperative illness.