Reliability of prenatal sonographic lung biometry in the diagnosis of pulmonary hypoplasia.

OBJECTIVE

The purpose of the present study was to assess the value of biometric lung measurements for the diagnosis of severe fetal pulmonary hypoplasia by investigating whether a significant correlation between two-dimensional lung biometry measurements and autopsy findings could be established.

METHODS

This was a prospective study carried out between 1995 and 1997. Nomograms for normal fetuses of the anterior-posterior and transverse inner thoracic diameters, which describe the growth and shape of the lung, were used as a basis for diagnosis of pulmonary hypoplasia in fetuses at high risk of developing the condition (the fetuses had bilateral renal agenesis or multicystic kidneys; chronic PROM <25 gestational weeks or hydrothorax). Pregnancy was terminated by abortion or intrauterine death in 29/43 high-risk fetuses and autopsies were performed. Only the 29 fetuses for which there were autopsy findings were included in the study.

RESULTS

The best plane for diagnosing pulmonary hypoplasia was the four-chamber view. The diagnostic accuracy for this view as expressed by the sensitivity was 57% for the anterior-posterior diameter and 44% for the transverse diameter; as expressed by the specificity it was 42% for the anterior-posterior diameter and 50% for the transverse diameter. The results for the four-chamber view for the various high-risk conditions were as follows: for fetuses with chronic PROM we obtained sensitivities of 75% and 50% (anterior-posterior and transverse dimensions, respectively) and specificities of 80% and 60% (anterior-posterior and transverse dimensions, respectively). The sensitivities of lung biometry in fetuses with hydrothorax were 1% and 80% for the two diameters, but there was a low specificity. In fetuses with bilateral renal agenesis or multicystic kidneys we obtained sensitivities of 36% and 30% (anterior-posterior and transverse dimensions, respectively) and a specificity of 50% (anterior-posterior dimension).

CONCLUSIONS

The present results show that two-dimensional lung biometry is not a suitable method for antenatal detection of pulmonary hypoplasia. However, in individual cases with high risk for pulmonary hypoplasia, lung biometry might prove to be an additional diagnostic parameter.