Carbon Monoxide Predicts BPD

NeoNotes Journal Club
Andrew B. Kairalla MD, Editor

Highlights from the Southeastern Association of Neonatologists Meeting at Marco Island, May 15-18, 2003

Carbon Monoxide Predicts BPD

Measurement of End Tidal Carbon Monoxide in Preterm Infants for the Prediction of Bronchopulmonary Dysplasia. Presented by Dr. Avroy Fanaroff, Rainbow Babies & Children’s Hospital, Cleveland OH.

Background: Carbon monoxide (CO) is a by-product of heme degradation mediated by the rate limiting enzyme heme oxygenase (HO). CO binds to hemoglobin and is eliminated via the lungs. Inflammatory mediators induce HO-1 with an increase in CO production. Thus changes in CO measurements in exhaled breath are indicative of increased HO activity and cellular stress, and therefore can be correlated with the severity of some disease processes. This study was designed to determine if sequential measurements of end tidal CO corrected for background CO (ETCOc) could predict the development of BPD in preterm infants.

Methods: In 105 preterm infants (BW 1.052+/- 259 g; GA 27.9 +/- 2.4 weeks) ETCOc was measured on days of life 1,3,7,14,and 28. In ventilated patients, measurements were obtained with a catheter above the tip of the endotracheal tube. In non-ventilated patients, a nasal probe was used.

Results: ETCOc was significantly elevated on days of life 1,3,7,14 and 28 in patients who developed BPD. An ETCOc value > 2.3 ppm on day 3 was the most discriminatory measurement with a 84% sensitivity, 63% specificity, 56% positive predictive value and 87% negative predictive value.

Comments: Previous studies of the usefulness of ETCOc have focused on the ability to predict significant jaundice due to hemolytic disease (See 2-041; 2-042; 2-043; 2-044; 2-045). We now see that this measurement may also be useful in predicting which preterm infants are likely (or unlikely) to develop BPD. We learned previously that a small thymus at birth (narrow superior mediastinum on chest x-ray) was also very predictive of future development of BPD in VLBW infants (see 3-041). It is frustrating to think that these early indicators can be such good predictors for BPD. It makes me think that the predisposition for BPD is determined very early in life – possibly even in-utero! ABK.

Additional Comments:

You may add your own comments to the discussion of this topic by selecting : Submit Comments.

Return to top

Hit Counter