Original article Accuracy and precision of perfusion lung scintigraphy versus 133 Xe-radiospirometry for preoperative pulmonary functional assessment of patients with lung cancer Denis Mariano-Goulart 1, 3 , Eric Barbotte 2 , Célia Basurko 2 , F. Comte 1 , Michel Rossi 1 1 Department of Nuclear Medicine, Montpellier University Hospital, Montpellier, France 2 Department of Statistics and Epidemiology, Montpellier University Hospital, Montpellier, France 3 Service Central de Médecine Nucléaire, CHU Lapeyronie, 371 avenue du Doyen Gaston Giraud, 34295 Montpellier Cedex 5, France Received: 23 September 2005 / Accepted: 16 January 2006 / Published online: 26 April 2006 © Springer-Verlag 2006 Abstract. Purpose: This study sought to determine whether 133 Xe-radiospirometry (XRS) successfully selects patients able to undergo lung resection without postoper- ative respiratory complications and whether perfusion lung scintigraphy (PLS) is likely to provide a similar selection of patients for certain tumour stages. Methods: Two hundred and eighty-four patients with resectable lung cancer underwent preoperative assessment of postoperative forced expiratory volume in 1 s (FEV 1 ) by XRS and PLS. Correlations, Bland and Altman analysis and contingency tables were used to analyse the difference between the two predictive techniques. Results: One hundred and sixty patients underwent lung resection on the basis of XRS preoperative testing only. None of them developed respiratory insufficiency. Despite a close correlation, the limits of agreement between predicted FEV 1 by XRS and PLS exceeded ±0.3 l/s. For tumour stages T1Nx and T2N0, PLS underestimated postoperative FEV 1 whereas it overestimated this param- eter for stage III. Conclusion: XRS accurately selects patients able to undergo lung resection without postoperative pulmonary insufficiency. The agreement between XRS and PLS is unacceptable. When only PLS is available, higher thresh- olds for patients with stage III cancers and lower thresholds for those with stage I cancers should be used to decide on operability. Keywords: Radiospirometry – Xenon – Perfusion – Pulmonary resection – Lung cancer Eur J Nucl Med Mol Imaging (2006) 33:1048–1054 DOI 10.1007/s00259-006-0087-5 Introduction When indicated, surgical resection offers the best chance for cure in patients with non-small cell lung carcinoma. However, the transitory increase in the dead space to tidal volume ratio during the postoperative period may be responsible for postoperative respiratory insufficiency in patients with impaired preoperative lung function [1]. This heightens the need for both efficient postoperative care and accurate selection of patients who are likely to be able to undergo lung resection without severe postoperative res- piratory complications [2, 3]. Several studies have pointed out that global spirometric tests fail to detect those patients who are at high risk of postoperative complications, and that this is particularly true of patients with chronic obstructive pulmonary disease [1, 4]. Nevertheless, these tests are recommended by several guidelines before pulmonary resection [5, 6], and criteria for the selection of patients who should be able to tolerate lung resection have been proposed on the basis of these tests [7, 8]. When these criteria are not fulfilled, assessment of regional lung function by quantitative imaging is required. Most studies and guidelines suggest that pulmonary resection is feasible in patients with a predicted postoperative forced expiratory volume in the first second (FEV 1 ) of 30–40% or more of the normal value [1, 3, 9, 10], or 1–1.2 l/s [11, 12]. Quantitative computed tomography scanning has shown promising results, but this method is not yet widely used [2, 3, 13, 14]. More usually, radionuclide techniques, includ- ing lung ventilation or perfusion scintigraphy, are em- ployed for the assessment of postoperative pulmonary function [11, 15–17]. Since the use of functional vital capacity (FVC) does not improve the accuracy of patient selection [18], the regional measurement of FEV 1 is regarded as the most reliable spirometric index of pulmonary insufficiency and has been chosen by most investigators to assess operability. With the use of a spirometer and a large field of view gamma Denis Mariano-Goulart ()) Service Central de Médecine Nucléaire, CHU Lapeyronie, 371 avenue du Doyen Gaston Giraud, 34295 Montpellier Cedex 5, France e-mail: d-mariano_goulart@chu-montpellier.fr Tel.: +33-467-338598, Fax: +33-467-338465 European Journal of Nuclear Medicine and Molecular Imaging Vol. 33, No. 9, September 2006