Neutrophil CD11b expression as a diagnostic marker for early-onset neonatal infection.

OBJECTIVES: To determine whether neutrophil surface expression of CD11b predicts early-onset infection or suspected infection in at-risk infants. STUDY DESIGN: CD11b expression on peripheral blood neutrophils was determined by flow cytometry of whole blood samples. Blood (0.1 ml) was obtained from a convenience sample of at-risk infants admitted to the neonatal intensive care unit, stained with antibodies detecting CD11b and CD15, chilled, and analyzed within 8 hours. Blood for culture, blood counts, and C-reactive protein (CRP) determination was obtained simultaneously. Subjects were grouped on the basis of culture results and clinical signs, and investigators were blinded to CD11b level. RESULTS: Of 106 subjects, seven had positive bacterial or viral cultures ("confirmed infection"), 17 had clinical signs of infection but negative cultures ("suspected infection"), and 82 had negative cultures and no clinical signs ("no infection"). Neutrophil CD11b was elevated in all infants with confirmed infection, 94% with suspected infection, and none with no infection. The negative and positive predictive values, sensitivity, and specificity were 100%, 99%, 96%, and 100%, respectively, for diagnosis of neonatal infection at initial evaluation. CD11b levels correlated with peak CRP (r2 = 0.76, p < 0.0001); however, CD11b was elevated at the time of admission in all five infants with proven bacterial infection, whereas CRP was normal until the second day in the neonatal intensive care unit in three of these five. Both infants with positive viral cultures had elevated CD11b, but the CRP levels remained within normal limits. The negative predictive value of neutrophil CD11b for identifying suspected or confirmed infection was 99%. CONCLUSION: This assay for neutrophil CD11b is a promising test for exclusion of early-onset neonatal infection. If validated prospectively, this assay may reduce hospital and antibiotic use in the population of neonates at risk for early-onset infection.

Relationship between breath and total body hydrogen excretion rates in neonates.

Our study examined the relationship of H2 excreted in breath to total body H2 excreted by neonates. We report simultaneously measured end-tidal H2 concentrations, plus breath H2 and total body H2 (breath H2 plus flatus H2) excretion rates in 10 neonates. End-tidal H2 concentrations varied from 2.4 to 192 ppm. Breath H2 excretion rates ranged from 0.20 to 6.5 and total body H2 excretion rates from 0.29 to 15.0 ml/h. The fractional breath H2 excretion in these infants was 48% (range 33-69%), compared with 21% reported in adults. The correlation coefficient for end-tidal derived H2 excretion and directly measured breath H2 excretion rates was 0.95 (p less than 0.001). We conclude that the proportion of total H2 excreted in the breath of neonates is increased compared with adults, suggesting that caution must be exercised when interpreting newborn breath H2 measurements and using adult norms.

Effect of ventilation on breath hydrogen measurements.

Measurement of hydrogen (H2) in expired air by interval sampling after oral administration of carbohydrate detects sugar malabsorption. Standard breath H2 tests require comparison of H2 concentrations in expired air samples obtained immediately before and after delivery of a test substrate. Comparison of interval samples assumes that minute ventilation (VE) remains constant unless H2 is independent of VE. Because healthy individuals have variable VE, we determined how H2 is influenced by changes in VE. H2 concentration was studied at different ventilatory rates in eight healthy adults. It varied inversely with VE in all subjects. We also compared the effect of changes in VE on the relationship between H2 and carbon dioxide (CO2) concentrations in expired air samples. At constant VE, the relationship between H2 and CO2 was linear (r = 0.95, P less than 0.001). As VE changed, the relationship between H2 and CO2 became nonlinear. Changes in VE altered methane concentrations in expired air samples from two subjects in a manner comparable to the effect on H2. These results demonstrate that breath H2 concentrations vary with ventilatory rate. Under conditions where frequent changes in VE are likely, independent measures for ensuring constant VE over sampling times are necessary. Use of CO2 as an internal standard to normalize H2 values to an alveolar concentration is appropriate only under conditions of constant VE.

Fasting breath hydrogen concentration: normal values and clinical application.

Excretion of hydrogen in breath commonly persists despite an overnight fast. Although elevation of hydrogen concentration above the fasting value after administration of a test sugar is evidence of malabsorption, the significance of the fasting value itself is unknown. We determined the normal limits of fasting breath hydrogen in healthy children and adults, and in patients with chronic diarrhea or recurrent abdominal pain. Fasting breath hydrogen in 221 healthy children and 9 healthy adults averaged 7.1 +/- 5.0 parts per million (mean +/- SD), exceeding 30 parts per million in less than 1%. No value exceed 42 parts per million. In 73 patients with recurrent abdominal pain and 76 patients with chronic diarrhea, fasting breath hydrogen was less than 42 parts per million in 97% and 83%, respectively. History and laboratory data were reviewed in the 15 patients where fasting breath hydrogen exceeded 42 parts per million. Seven had documented small bowel bacterial overgrowth and an additional 3 patients had radiographic evidence of intestinal stasis. Using test dinner meals, we prospectively evaluated the effect of previously ingested foods containing complex carbohydrates on fasting breath hydrogen. Dinner meals consisting of rice, wheat, or beans influenced fasting breath hydrogen values, but did not result in elevated fasting breath hydrogen in healthy individuals. Rice bread resulted in uniformly low fasting breath hydrogen values in healthy subjects (2.0 +/- 2.5 parts per million), but fasting breath hydrogen remained elevated in patients with bacterial overgrowth. Our studies indicate that conditions for measurement of the fasting breath hydrogen value may be standardized to improve discrimination between normal and abnormal values.

Glycoproteins as substrates for production of hydrogen and methane by colonic bacterial flora.

Hydrogen and methane in human breath derive entirely from bacterial fermentation in the intestinal lumen. The sources of substrates utilized for these reactions have not been completely determined. Basal excretion of both gases occurs in the fasted state, while malabsorbed carbohydrate commonly results in increased hydrogen but not methane production. Using an in vitro fecal incubation system, we have studied hydrogen and methane production from glycoproteins of endogenous as well as dietary origin. All glycoproteins tested yielded hydrogen when incubated with fecal homogenates. Mean hydrogen production from substrates containing less than 3% sugar (human serum albumin, bovine serum albumin, and alpha-casein) averaged 2.2 +/- 0.9% of hydrogen production from equivalent amounts of glucose, while carbohydrate-rich mucin yielded 46.0 +/- 6.7% of hydrogen production from glucose. Glycoproteins of intermediate carbohydrate content, including transferrin and egg white, yielded intermediate values. Methane production from glycoproteins was optimal from carbohydrate-poor protein substrates in fecal homogenates which accumulated hydrogen and became rapidly acidic when incubated with pure carbohydrate. In contrast, methane production was comparable for essentially sugar-free proteins, glycoproteins, and glucose when hydrogen did not accumulate and neutral pH was maintained. We conclude that glycoproteins are substrates for hydrogen and methane production by colonic bacteria from healthy adults. In individuals with bacterial overgrowth syndromes and in protein-losing enteropathy, bacterial catabolism of endogenous glycoproteins may cause increased basal hydrogen and methane excretion. These findings also raise the possibility that measurement of hydrogen or methane after oral administration of dietary glycoproteins may be useful in detection of protein malabsorption.

Role of pH in production of hydrogen from carbohydrates by colonic bacterial flora. Studies in vivo and in vitro.

Hydrogen produced by colonic bacteria and excreted in breath is a useful index of carbohydrate malabsorption. Since colonic contents are often acidic in individuals with carbohydrate malabsorption and in normal newborns, we determined the effect of colonic acidification on H2 production. Acidification of colonic contents by dietary means significantly reduced excess breath H2 excretion from 55.4 +/- 11.1 (SEM) to 12.2 +/- 3.1 ml/4 h (P less than 0.05) after administration of 0.3 g/kg of the nonabsorbable sugar lactulose to five normal adult subjects. Similarly, the breath H2 response to lactose was reduced or eliminated in two proven lactose malabsorbers after acidification. The correlation between pH and H2 production from carbohydrate was further investigated in adults and neonates, using an in vitro fecal incubation system. Glucose disappearance and H2 production were pH dependent and highly correlated (r = 0.94) in the pH range 5.5-7.6. Maximal production of H2 from glucose by fecal incubates occurred at pH 7.0-7.45. Inhibition of H2 production from carbohydrate occurred at acid pH. H2 per hour from glucose at pH 6.2 and 5.5 averaged 60.2% and 24.2%, respectively, of that produced at neutral pH. Rapid reversal of pH-induced inhibition by neutralization indicated a metabolic, rather than a bactericidal process. The observations indicate that the breath H2 response to malabsorbed carbohydrate is affected by colonic pH. It appears that the efficiency of bacterial carbohydrate metabolism in the colon is pH dependent.

Extreme but asymptomatic carboxyhemoglobinemia and chronic lung disease.

Carboxyhemoglobinemia is a well-known consequence of carbon monoxide exposure from smoking. However, only moderately elevated levels have been reported. We report the case of an asymptomatic man with severe chronic obstructive lung disease and carboxyhemoglobin levels repeatedly in excess of 30% (maximum, 38.0%) due to smoking. The mechanism by which such high levels were attained was primrily a combination of arterial hypoxia and a high carbon monoxide yield from tobacco. For a given level of carbon monoxide exposure, the hypoxic person will attain a higher carboxyhemoglobin level than will a person without hypoxia.

Essential fatty acid deficiency in neonates: inability to reverse deficiency by topical applications of EFA-rich oil.

Correction of essential fatty acid deficiency by transcutaneous absorption of topically applied EFA-rich oil has been reported. We measured serum EFA levels in two groups of neonates receiving fat-free total parenteral nutrition: nine control patients after 16 and 25 days of TPN, and six patients before and 12 days after beginning cutaneous application of 100 mg/kg/day of linoleic acid as sunflower seed oil. Progressive biochemical EFA deficiency occurred in all but one of the control patients. Of the six patients receiving 100 mg/kg/day of linoleic acid, one patient with mild deficiency improved, but progressive EFA deficiency occurred in the other five patients. Serum EFA levels were also measured in four patients following 76 days of TPN and daily application of high doses of topical safflower oil, all of whom had severe biochemical EFA deficiency. The topical application of EFA-rich oil cannot be assumed to be uniformly effective in reversing or preventing EFA deficiency. The transcutaneous absorption of essential fatty acids must be documented by appropriate measurements of EFA in serum lipids.

Host heme catabolism in drug-sensitive and drug-resistant malaria.

Chloroquine resistance has arisen in both human and murine forms of malaria. CR Plasmodium berghei in mice does not produce the malaria pigment which is characteristic of the CS form. Determinations of carbon monoxide production (i.e., host heme catabolism) by individual mice revealed that those infected with CS P. berghei produce only one fourth as much carbon monoxide as do CR infected mice at all levels of infection. These observations confirm the idea that malaria pigment is composed of precipitated host cell hemoglobin and suggest that drug resistance is accompanied by a basic alteration in parasite-mediated hemoglobin catabolism.