Lactate and pyruvate accumulation during hypocapnia

Lactate, Base Excess, and the Pediatric Index of Mortality: Exploratory Study of an International, Multicenter Dataset

OBJECTIVES

To investigate the relationship between ICU admission blood lactate, base excess, and ICU mortality and to explore the effect of incorporating blood lactate into the Pediatric Index of Mortality.

DESIGN

Retrospective cohort study based on data prospectively collected on every PICU admission submitted to the U.K. Pediatric Intensive Care Audit Network and to the Australia and New Zealand Pediatric Intensive Care Registry.

SETTING

Thirty-three PICUs in the United Kingdom/Republic of Ireland and nine PICUs and 20 general ICUs in Australia and New Zealand.

PATIENTS

All ICU admissions between January 1, 2012, and December 31, 2015.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

One hundred twenty-three thousand two hundred fifty-two admissions were recorded in both datasets; 81,576 (66.2%) in the United Kingdom/Republic of Ireland and 41,676 (33.8%) in Australia and New Zealand. Of these 75,070 (61%) had a base excess recorded, 63,316 (51%) had a lactate recorded, and 60,876 (49%) had both base excess and lactate recorded. Median lactate value was 1.5 mmol/L (interquartile range, 1-2.4 mmol/L) (United Kingdom/Republic of Ireland: 1.5 [1-2.5]; Australia and New Zealand: 1.4 [1-2.3]). Children with a lactate recorded had a higher illness severity, were more likely to be invasively ventilated, admitted after cardiac surgery, and had a higher mortality rate, compared with admissions with no lactate recorded (p < 0.001). The relationship between lactate and mortality was stronger (odds ratio, 1.32; 95% CI, 1.31-1.34) than between absolute base excess and mortality (odds ratio, 1.13; 95% CI, 1.12-1.14). Addition of lactate to the Pediatric Index of Mortality score led to a small improvement in performance over addition of absolute base excess, whereas adding both lactate and absolute base excess achieved the best performance.

CONCLUSIONS

At PICU admission, blood lactate is more strongly associated with ICU mortality than absolute base excess. Adding lactate into the Pediatric Index of Mortality model may result in a small improvement in performance. Any improvement in Pediatric Index of Mortality performance must be balanced against the added burden of data capture when considering potential incorporation into the Pediatric Index of Mortality model.

Pulmonary lactate release in patients with acute lung injury is not attributable to lung tissue hypoxia

OBJECTIVE

To determine whether pulmonary lactate production in patients with acute lung injury is attributable to lung tissue hypoxia.

DESIGN

Prospective, controlled, clinical study.

SETTING

A multidisciplinary university intensive care unit in a general hospital.

PATIENTS

Seventy consecutive critically ill patients requiring mechanical ventilation and invasive hemodynamic monitoring. Of these patients, 18 had no acute lung injury (no ALI); 33 had acute lung injury (ALI) (Lung Injury Score [LIS] < or =2.5); and 19 had acute respiratory distress syndrome (ARDS) (LIS >2.5).

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

After hemodynamic measurements, lactate and pyruvate concentrations were assessed in simultaneously drawn arterial (a) and mixed venous (v) blood samples. Pulmonary lactate release was calculated as the product of transpulmonary a-v lactate difference (L[a-v]) times the cardiac index. Two indices of anaerobic metabolism of the lung, i.e., the transpulmonary a-v difference of lactate pyruvate ratio (L/P[a-v]) and excess lactate formation across the lungs (XL), were calculated. L(a-v) and pulmonary lactate release were higher in patients with ARDS than in the other groups (p<.001), and they were also higher in patients with ALI compared with patients with no ALI (p<.001). In patients with ALI and ARDS (n = 52), pulmonary lactate release correlated significantly with LIS (r2 = .14, p<.01) and venous admixture (r2 = .13, p<.01). When all patients were lumped together (n = 70), pulmonary lactate release directly correlated with LIS (r2 = .30, p<.001), venous admixture (r2 = .26, p<.001), and P(A-a)O2 (r2 = .14, p<.01). Neither L/P(a-v) nor XL was significantly different among the three groups.

CONCLUSION

The lungs of patients with ALI produce lactate that is proportional to the severity of lung injury. This lactate production does not seem to be attributable to lung tissue hypoxia.

Differential effects of hypothermia upon blood acid-base state and blood gases in sodium pentobarbital and urethane anaesthetised rats

1. The effects of two anaesthetics, sodium pentobarbital and urethane, and the effects of anaesthesia-associated hypothermia on acid-base status and blood gases were studied in rats without assisted ventilation. 2. Manipulation of conscious rats produces a progressive increase in arterial lactate associated with slight hyperventilation. 3. Sodium pentobarbital anaesthesia produces mild respiratory acidosis accompanied by increase in lactate arterial values. Urethane anaesthesia leads to partially compensated metabolic acidosis. 4. Hypothermia reduces metabolic acidosis and hypercapnia induced by sodium pentobarbital anaesthesia. No difference between hypothermic and normothermic values was observed in urethane anaesthesia.

Waning of panic sensations during prolonged hyperventilation

Recent theories about panic emphasize that a hyperventilatory positive feedback loop is involved in panic: catastrophic misinterpretation of bodily sensations may trigger anxiety, anxiety may stimulate hyperventilation, hyperventilation may promote the salience of feared sensations etc. Such models leave unexplained how and when panics come to an end. It was hypothesised that panic with hyperventilation may end because pronounced hyperventilation becomes, in the course of time, less powerful in generating perceivable bodily sensations. Twenty healthy subjects hyperventilated forcefully and experienced clear panic symptoms as defined by DSM IIIR. When pCO2 was kept 55% below base line for 90 min, panic symptoms waned. The mean intensity of the symptoms declined as did the number of symptoms occurring. No panic symptoms were observed in the control group (n = 20) who ventilated normally. In so far as hyperventilation is involved in the positive feedback loops that characterize panic, panic attacks may be time-limited because sensations induced by hyperventilation become less salient even if massive hyperventilation continues. As to the explanation of the reported phenomenon, it is suggested that, apart from habituation, local physiological changes due to prolonged hyperventilation may produce a decrease in interoceptive input.

The effects of hypocapnia on intracellular pH and bicarbonate

The intracellular pH (pHi) and bicarbonate concentration ([HCO3-]icw) of cardiac and skeletal muscles were monitored during respiratory alkalosis in order to further elucidate the homeostatic processes which operate in these tissues to ameliorate deviations from normal acid-base status. Rats were mechanically hyperventilated to induce hypocapnia, pHi was determined by the DMO method, and [HCO-3]icw was calculated from the Henderson-Hasselbalch equation using pHi and the partial pressure of carbon dioxide of vena caval blood. A significant intracellular alkalosis occurred in both cardiac and skeletal muscles during hypocapnia, but the changes in pHi were less in the heart than in skeletal muscle. The decreases in cardiac [HCO-3]icw were greater than those attributable to the physicochemical buffering of the heart. These data are consistent with an intramyocardial source of protons other than physicochemical buffering during respiratory alkalosis. The decreases in skeletal muscle [HCO-3]icw were less than those due to physicochemical buffering. These data are consistent with a net extrusion from skeletal muscles cells of the protons derived from physicochemical buffering during respiratory alkalosis.

Serial observations of arterial and mixed-venous blood gases after step change in ventilation

In 22 dogs, subjected to a step change in ventilation, serial changes in blood gas composition and lactate and pyruvate concentrations of arterial as well as mixed venous blood were studied. The change of PaCO2 was approximately 20 mm Hg both in hypo- and hyperventilation. During hypoventilation the difference in various forms of CO2 between arterial and mixed venous blood showed first a downward shift and then gradually increased, whereas during hyperventilation they progressively increased and reached a constant level within 10-20 min. This difference was assumed to be mainly due to more efficient CO2 elimination through lung ventilation in hyperventilation as compared with CO2 accumulation from tissue metabolism in hypoventilation. In vivo buffer slopes for CO2 during hypoventilation were about half those in vitro, whereas during hyperventilation both slopes were approximately the same. In vivo arterial buffer slope was higher during hypoventilation and lower during hyperventilation as compared to that of mixed venous blood in the respective state of ventilation.