Analysis of Carbon Monoxide Diffusion Capacity Threshold for Loss of Functional Pulmonary Capillary Surface Area in Patients with Scleroderma, with and Without Pulmonary Hypertension

David Langleben Center for Pulmonary Vascular Disease, Cardiology, Jewish General Hospital, Montreal, Quebec, Canada Stylianos Orfanos 2nd Dept. of Critical Care, Attikon Hospital, Athens, Greece Benjamin Fox Respirology, Assaf Harofeh Hospital, Zerifin, Israel Michele Giovinazzo Center for Pulmonary Vascular Disease, Cardiology, Jewish General Hospital, Montreal, Quebec, Canada Robert Schlesinger Center for Pulmonary Vascular Disease, Cardiology, Jewish General Hospital, Montreal, Quebec, Canada Fay Blenkhorn Center for Pulmonary Vascular Disease, Cardiology, Jewish General Hospital, Montreal, Quebec, Canada Lawrence Rudski Center for Pulmonary Vascular Disease, Cardiology, Jewish General Hospital, Montreal, Quebec, Canada John Catravas Biophysics, Old Dominion University, Norfolk, USA

Introduction: A decrease in carbon monoxide diffusing capacity (DLCO), in patients with scleroderma and minimal interstitial lung disease (ILD), may be used as a marker for the loss of perfused pulmonary functional capillary surface area (FCSA) and for the onset of clinically occult or overt pulmonary arterial hypertension (PAH). Using a unique technique that sensitively measures FCSA in humans, we have previously proved that the DLCO (%predicted) correlates linearly with FCSA in patients with PAH and scleroderma. However, we had not studied the DLCO relationship in scleroderma patients without PAH or with normal amounts of FCSA. We now describe a larger cohort, without and with PAH and determine the threshold of DLCO for abnormally low FCSA.
Methods: We assessed FCSA by measuring the first-pass transpulmonary metabolism of ³H-benzoyl-Phe-Ala-Pro, by the endothelial ectoenzyme, angiotensin converting-enzyme, in 33 patients with scleroderma, with and without PAH. DLCO was measured, and corrected for hemoglobin. 23 humans without scleroderma or pulmonary hypertension (PH) acted as controls. Correlations and receiver operating curves (ROC) were generated.
Results: 5 patients had scleroderma without PH, 2 had borderline PH and 6 had PH from left heart disease. The remainder had PAH. There was a strong linear correlation between FCSA and DLCO (r=0.74, p <<0.001). Using the range of FCSA in the 23 normal subjects, the 95% lower confidence interval for FCSA was 1581. Using that as the low value for normal, in ROC analysis the areas under the curve for DLCO to detect low FCSA were 0.86 for DLCO < 60%, 0.83 for DLCO < 70% and 0.83 for DLCO < 50%.
Conclusion: In patients with scleroderma and minimal or no ILD, DLCO is a useful parameter to predict the onset of significant loss of FCSA and early PAH. A threshold of 60% predicted is most sensitive and specific.

David Langleben

Jewish General Hospital/McGill University