RESUMO
BACKGROUND AND OBJECTIVES: We prospectively compared five techniques to estimate predicted postoperative function (ppo F) after lung resection: recently proposed quantitative CT scans (CT), perfusion scans (Q), and three anatomical formulae based on the number of segments (S), functional segments (FS), and subsegments (SS) to be removed. METHODS: Four parameters were assessed: FEV(1), FVC, DL(CO) and VO(2max), measured preoperatively and 6 months postoperatively in 44 patients undergoing pulmonary resection, comparing their ppo value to the postoperatively measured value. RESULTS: The correlations (r) obtained with the five methods were for CT: FEV(1) = 0.91, FVC = 0.86, DL(CO) = 0.84, VO(2max) = 0.77; for Q: 0.92, 0.90, 0.85, 0.85; for S: 0.88, 0.86, 0.84, 0.75; for FS: 0.88, 0.85, 0.85, 0.75, and for SS: 0.88, 0.86, 0.85, 0.75, respectively. The mean difference between ppo values and postoperatively measured values was smallest for Q estimates and largest for anatomical estimates using S. Stratification of the extent of resection into lobectomy (n = 30) + wedge resections (n = 4) versus pneumonectomy (n = 10) resulted in persistently high correlations for Q and CT estimates, whereas all anatomical correlations were lower after pneumonectomy. CONCLUSIONS: We conclude that both Q- and CT-based predictions of postoperative cardiopulmonary function are useful irrespective of the extent of resection, but Q-based results were the most accurate. Anatomically based calculations of ppo F using FS or SS should be reserved for resections not exceeding one lobe.