Assessment of ventilation inhomogenity with Krypton SPECT and planar imaging Trine Stavngaard and Jann Mortensen Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark Correspondence Trine Stavngaard, Department of Clinical Physiology and Nuclear Medicine 4011, Blegdamsvej 9, DK-2100 Copenhagen, Denmark E-mail: stavngaard@dadlnet.dk Accepted for publication Received 18 August 2004; accepted 29 November 2004 Key words chronic obstructive pulmonary disease; lung healthy; pulmonary function test; pulmonary scintigraphy; ventilation scintigraphy Summary In 29 chronic obstructive pulmonary disease (COPD) patients and nine lung healthy volunteers, above the age of 50 years, ventilation defects were examined by 81m Kr planar scintigraphy and 81m Kr single photon emission computed tomography (SPECT) to investigate if SPECT adds information regarding size and extent of visually scored ventilation defects, and to correlate the extent of defects obtained from the two imaging settings with standard pulmonary function tests performed in these patients/volunteers. For testing the reproducibility of the visual defect score of 81m Kr scintigraphy additionally 13 patients suspected for pulmonary embolism or lung cancer were included. Each series of planar or SPECT studies were read for the extent (% abnormal lung) and severity (0–3) of ventilation abnormalities. Seventeen scans were read twice for reproducibility studies. The extent of ventilation defect assessed by 81m Kr SPECT was higher than by 81m Kr planar (slope of regression line 0Æ60, P<0Æ0001), likewise severity score (rank signed test: P<0Æ0001). Correlation between ventilation inhomogeneity and pulmonary function test (residual volume and T L,CO ) in the COPD group revealed only significance for the SPECT acquisition. We found good reproducibility of visual assessment of ventilation defect extent (correlation: 0Æ97, P<0Æ0001) and severity (Kappa 0Æ62). In conclusion, visual scoring of extent and severity of ventilation defects was reproducible. Ventilation defects were better demonstrated with SPECT than planar imaging. The correlation to pulmonary function was better with SPECT than planar imaging. Introduction Radionuclide imaging of the lung is central in the visualization of regional lung function. However, with the development of other modalities of functional lung imaging e.g. magnetic resonance imaging with hyperpolarized 3 helium, nuclear medicine techniques are often described inferior because of limited spatial and temporal resolution (de Lange et al., 1999; Eberle et al., 2001; Kauczor et al., 2001). Planar ventilation scanning remains the standard, first line method of assessing patients with ventilation abnormalities. However, single photon emission computed tomography (SPECT) provides improved contrast resolution and more complete three-dimensional spatial information of the ventilation and perfusion distribution than can be achieved with planar imaging. SPECT became available for routine clinical use approximately 20 years ago, and although Osborn et al. already in 1983 in animal studies of pulmonary embolism showed higher sensitivity of SPECT compared with planar lung perfusion scintigraphy (Osborne et al., 1983), it is only in the last few years that the technique has been more widely applied in lung analysis, especially for lung perfusion in the diagnosis of pulmonary embolism where SPECT is superior compared with planar acquisition (Sando et al., 1997; Magnussen et al., 1999; Palmer et al., 2001; Bajc et al., 2002; Meignan, 2002). Ventilation SPECT has also been investigated. Whereas the majority of lung-perfusion scintigraphy is per- formed with 99m Technetium (Tc) Macro Aggregated Albumin the radiotracers used in ventilation scintigraphy varies. The most widely used are: radioactive gases ( 127 Xenon, 133 Xenon, 81m Krypton), ultra fine radioaerosols [ 99m Tc Technegas and 99m Tc Pertechnegas (particle size 0Æ02 to 0Æ2 lm)] and radioaerosols [diethylene triamine penta-acetic acid ( 99m Tc- DTPA), 99m Tc sulphur colloid, 99m Tc pyrophosphate] (Alderson & Line, 1980; Worsley & Gottschalk, 2000). The different radiotracers have dissimilar physiological distribution patterns and T 1/2 , and the conclusions regarding gaining additional information from SPECT compared with planar imaging diverge between the radiotracers (Jamadar et al., 1999; Palmer et al., 2001; Bajc et al., 2002). The relation between aerosol deposition and ventilation is weak, especially in patients with chronic obstructive pulmonary diseases (COPD), who have centrally deposited tracer (Jamadar et al., 1999). The ultra fine radio- Clin Physiol Funct Imaging (2005) 25, pp106–112 Ó 2005 Blackwell Publishing Ltd • Clinical Physiology and Functional Imaging 25, 2, 106–112 106