Determination of 'unmeasured' anions in acidotic ICU patients

Metabolic acidosis is one of the most frequent acid-base disorders occurring in the ICU. Major causes of metabolic acidosis in critically ill patients are hyperchloremia, hyperlactatemia and the presence of anions of unknown identity, the so-called 'unmeasured' anions. The latter is associated with increased mortality and several diseases: sepsis, shock, liver dysfunction and renal failure. The physicochemical approach described by Stewart can be applied to quantify metabolic acidosis. Accordingly, the strong ion gap (SIG) is a quantitative measure of 'unmeasured' anions. We hypothesised that derangements in amino acid and organic acid metabolism and abnormal uric acid concentrations could be an explanation for the SIG.

From 32 adult ICU patients with metabolic acidosis, defined as a pH less than 7.35 and a base excess less than -5 mmol/l, the SIG was calculated in a single arterial blood sample. Two groups were compared: patients with SIG <2 mEq/l and patients with SIG >5 mEq/l. 'Unmeasured' anions were examined quantitatively by ion-exchange column chromatography, reverse-phase HPLC and gas chromatography/mass spectrometry measuring, respectively, 25 amino acids, uric acid and organic acids. Some organic acids were determined semi-quantitatively. The Mann–Whitney U test was applied for significance (considered P < 0.05) in all cases. For nominal data, the chi-square test was used.

Aspartic acid, isoleucine, ornithine, uric acid, succinic acid, fumaric acid, p-OH-phenyllactic acid and the semi-quantified organic acids 3-OH-isobutyric acid, pyroglutamic acid and homovanillic acid were all significantly elevated in the SIG >5 group (n = 12, mean = 8.3 mEq/l) compared with the SIG <2 group (n = 8, mean = 0.6 mEq/l). Generally, no major differences in organic acid spectra between both groups were observed. However, in one patient in the SIG >5 group who was in a prolonged fasted state at ICU admission, 3-OH-butyric acid was extremely high: 4.0 mEq/l, corresponding to 25% SIG. Overall, the averaged difference between both groups in total amino acid, uric acid and organic acid concentration contributed to the SIG for, respectively, 3.5% (268 μEq/l, not significant), 2.2% (169 μEq/l, P = 0.021) and 1.0% (79 μEq/l, P = 0.025). The total organic acid concentration consisted of glycolic acid, oxalic acid, methylmalonic acid, succinic acid, fumaric acid, malic acid, adipic acid and p-OH-phenyllactic acid. Comparison of patient characteristics of both groups showed that age, sex, APACHE II score, pH, base excess and lactate were not significant. However, renal insufficiency, sepsis and mortality were more prominent in the SIG >5 group. Also, the apparent strong ion difference (due to a significantly lower plasma chloride), phosphate and urea were significantly elevated in the SIG >5 group.

This study demonstrates that total amino acids, uric acid and organic acids form a minor contribution (6.8%, corresponding to 517 μEq/l) to the SIG in acidotic ICU patients.