Prediction of postoperative lung function in patients with lung cancer: comparison of quantitative CT with perfusion scintigraphy.

Published

Journal Article

OBJECTIVE: Prediction of postoperative lung function is important in preoperative evaluation of patients with lung cancer. Perfusion scintigraphy is the current method to assess the fractional contribution of lung function of the remaining lung. We developed a quantitative CT method and compared it with perfusion scintigraphy for predictions of postoperative forced expiratory volume in 1 sec (FEV1) in patients with lung cancer. SUBJECTS AND METHODS: Forty-four patients with lung cancer undergoing lung resection with preoperative CT and perfusion scintigraphy were enrolled. Quantitative CT used a dual threshold (-500 and -910 H) on standard preoperative CT to semiautomatically extract lung volume without emphysema or tumor and atelectasis, which we defined as "functional lung volume." Prediction was calculated from preoperative FEV1 multiplied by the fractional contribution of functional lung volume of the remaining lung by quantitative CT. Perfusion scintigraphy was the standard method. Predictions were correlated with postoperatively measured FEV1. RESULTS: Both quantitative CT and perfusion scintigraphy predicted postoperative FEV1 well in patients who underwent pneumonectomy (n = 28, r = 0.88 vs r = 0.86) and in lobectomy (n = 16, r = 0.90 vs r = 0.80) (both, p < 0.001). There was good agreement between the two methods by the Bland-Altman method. In the four patients with low measured postoperative FEV1 (<40% predicted normal), quantitative CT had true-positive prediction in four and perfusion scintigraphy, in only two. CONCLUSION: Given its simplicity, we proposed that quantitative CT be widely used in predicting postoperative FEV1.

Full Text

Cited Authors

  • Wu, M-T; Pan, H-B; Chiang, AA; Hsu, H-K; Chang, H-C; Peng, N-J; Lai, P-H; Liang, H-L; Yang, C-F

Published Date

  • March 2002

Published In

Volume / Issue

  • 178 / 3

Start / End Page

  • 667 - 672

PubMed ID

  • 11856695

Pubmed Central ID

  • 11856695

International Standard Serial Number (ISSN)

  • 0361-803X

Digital Object Identifier (DOI)

  • 10.2214/ajr.178.3.1780667

Language

  • eng

Conference Location

  • United States