ORIGINAL ARTICLE
Krypton for Computed Tomography Lung Ventilation Imaging Preliminary Animal Data Andreas H. Mahnken, MD, MBA, MME,* Gregor Jost, MSc,† and Hubertus Pietsch, PhD† Objectives: The objective of this study was to assess the feasibility and safety of krypton ventilation imaging with intraindividual comparison to xenon ventilation computed tomography (CT). Materials and Methods: In a first step, attenuation of different concentrations of xenon and krypton was analyzed in a phantom setting. Thereafter, 7 male New Zealand white rabbits (4.4–6.0 kg) were included in an animal study. After orotracheal intubation, an unenhanced CT scan was obtained in end-inspiratory breath-hold. Thereafter, xenon- (30%) and krypton-enhanced (70%) ventilation CTwas performed in random order. After a 2-minute wash-in of gas A, CT imaging was performed. After a 45-minute wash-out period and another 2-minute wash-in of gas B, another CT scan was performed using the same scan protocol. Heart rate and oxygen saturation were measured. Unenhanced and krypton or xenon data were registered and subtracted using a nonrigid image registration tool. Enhancement was quantified and statistically analyzed. Results: One animal had to be excluded from data analysis owing to problems during intubation. The CT scans in the remaining 6 animals were completed without complications. There were no relevant differences in oxygen saturation or heart rate between the scans. Xenon resulted in a mean increase of enhancement of 35.3 ± 5.5 HU, whereas krypton achieved a mean increase of 21.9 ± 1.8 HU in enhancement (P = 0.0055). Conclusions: The use of krypton for lung ventilation imaging appears to be feasible and safe. Despite the use of a markedly higher concentration of krypton, enhancement is significantly worse when compared with xenon CT ventilation imaging, but sufficiently high for CT ventilation imaging studies. Key Words: computed tomography, lung, ventilation, xenon, krypton (Invest Radiol 2015;50: 305–308)
I
maging of airways and lung tissue is a common indication for computed tomography (CT) of the chest. Values of CT as measured in Hounsfield units (HUs) provide a quantitative measure of lung ventilation. Assessment of regional ventilation by means of CT, however, requires inhalation of radiodense xenon. Xenon ventilation CT has shown potential in the assessment of regional ventilation.1,2 It has proven potentially useful in various pathologic conditions such as asthma, chronic obstructive pulmonary disease, bronchiolitis obliterans, and pulmonary embolism.3–6 Limitations regarding the quantification of xenon enhancement caused by different lung volumes between repeated scans have mostly been overcome by the introduction of dual-energy CT imaging because material decomposition allows for the direct assessment of xenon distribution within the lungs.7–9 Because dual-energy CT is not ubiquitously available, alternative means such as advanced image subtraction techniques are needed to assess the relatively small attenuation changes. Received for publication September 21, 2014; and accepted for publication, after revision, November 5, 2014. From the *Department of Diagnostic and Interventional Radiology, Philipps-University, Marburg, and †MR & CT Contrast Media Research, Bayer Healthcare, Berlin, Germany. Conflicts of interest and sources of funding: The project was in parts funded by Bayer Healthcare, Berlin, Germany. The authors report no other conflicts of interest. Reprints: Andreas H. Mahnken, MD, MBA, MME, Department of Diagnostic and Interventional Radiology, University Hospital, Philipps-University Marburg Baldingerstrasse, D-35043 Marburg, Germany. E-mail: [email protected]. Copyright © 2014 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0020-9996/15/5005–0305
Investigative Radiology • Volume 50, Number 5, May 2015
Nevertheless, xenon ventilation CT has mostly been limited to clinical research because it has several relevant problems. Most importantly, xenon has relevant adverse effects including somnolence, headaches, and nausea. To keep these adverse effects under control, its concentration is limited to 30% of a xenon-oxygen mixture, but even then up to 70% of patients describe transient adverse effects.4 With krypton, a 70% krypton concentration was shown to provide an acceptable enhancement in dual-energy CT.10 In addition, precision of quantitative CT ventilation measurements depends on the separation of lung and soft tissue compartments. The moderate solubility of xenon in blood will limit the precision of regional ventilation measurements because the uptake of xenon into blood alters the CT values of the portion of a region of interest that does not contain air. If images are obtained in the supine position, gravity-dependent gradients in lung ventilation may appear on xenon maps with relatively high xenon content in the gravity-dependent lung areas.8,11 This problem seems to be less pronounced in the prone position.11 Finally, xenon is expensive. Krypton has been recognized as an alternative means to assess regional ventilation using CT imaging.12 It is a noble gas with a sufficiently high atomic number (Z = 36) to visibly alter CT values when inhaled. In fact, this property has been described as early as 1977,13 but it has rarely been used, despite some obvious advantages.12 It has no narcotic potential, and high concentrations can be inhaled without known adverse effects.14 In fact, its safety in low concentrations is well known from nuclear medicine, where 81mKr has been used for decades.15 Its solubility in blood is much lower when compared with xenon, and it only costs a tenth of xenon.13 However, there is a single relevant disadvantage. Its atomic number is much lower than the atomic number of xenon (Z = 54). Therefore, it will provide a markedly lower change in attenuation if administered at identical concentrations. The goal of this experimental study was to assess the feasibility and safety of krypton ventilation CT imaging with digital subtraction for postprocessing.
MATERIALS AND METHODS Phantom Study In a first step, a calibration curve for different concentrations of krypton and xenon mixed with oxygen was generated. For this purpose, pure xenon or krypton was obtained from a pressurized metal storage container (Air Liquide, Düsseldorf, Germany) via a pressure valve (FMD650-03; DruVa, Eppelheim, Germany) connected to a calibrated mass flow meter (Voegtlin, Aesch, Switzerland). The latter allows for a precise measurement and dosing of gas used for the phantom studies. With the pressure valve of the metal storage container set to 1.0 bar, xenon or krypton was continuously insufflated into a 100-mL plastic container until a target concentration balanced with air was achieved. The container was placed inside the gantry of a multislice computed tomography scanner (SOMATOM Definition Flash; Siemens, Forchheim, Germany) and scanned at different tube voltages with the effective tube current-time product normalized to a computed tomography dose indexVol of 4.65 mGy. A standardized scan protocol with 2 128 0.6-mm collimation, a table feed of 0 mm per rotation (normalized pitch = 0), and a gantry rotation time of 270 milliseconds was used. Tube voltages were 80 kV(p), 100 kV(p), 120 kV(p), and 140 kV (p) with corresponding effective tube current-time products of 245 www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
305
Investigative Radiology • Volume 50, Number 5, May 2015
Mahnken et al
Animal Preparation
FIGURE 1. Graphic display of the study design: After an unenhanced CT scan, either krypton- or xenon-enhanced CT ventilation imaging was performed in random order after a 2-minute wash-in phase. Figure 1 can be viewed online in color at www.investigativeradiology.com.
mAseff (at 80 kV[p]), 110 mAseff (at 100 kV[p]), 65 mAseff (at 120 kV [p]), and 42 mAseff (at 140 kV[p]). For image reconstruction, a slice thickness of 3 mm with an increment of 2 mm and a medium smooth convolution kernel (B31f) were chosen with a field of view of 150 mm 150 mm. Computed tomographic values inside the container were measured on 4 consecutive slices. Mean attenuation values as measured in HU were computed and expressed as mean ± standard deviation. Results were displayed graphically, and linear equations for each kilovolt (peak) setting and its coefficient of determination were computed.
Study Design For animal studies, a crossover design was used to allow for intraindividual comparison. First, an unenhanced baseline scan with normal air ventilation was obtained. Thereafter, the animals were randomly assigned to start with either krypton or xenon ventilation CT. After a 2-minute wash-in phase, either krypton- or xenon-enhanced CT scans were obtained. Thereafter, the animals underwent a 45-minute washout phase at room air to clear the airways from residual gas. Finally, a second CT scan was performed with either krypton or xenon ventilation (depending on the gas used for the first scan) after another 2-minute washin phase. After the experiment, each animal had undergone unenhanced (air), krypton-, and xenon-enhanced lung CT (Fig. 1).
A total of 7 New Zealand white rabbits (4.4–6.0 kg) were included in this study after approval from the local animal care committee. For imaging procedures, all rabbits were anesthetized with intramuscular administration of 35 mg/kg of ketamine hydrochloride (Ketavet; Pfizer, Karlsruhe, Germany) and 5 mg/kg of xylazine (Rompun; Bayer Vital, Leverkusen, Germany). The animals were orotracheally intubated and mechanically ventilated at a respiratory rate of 25/min with room air. Anesthesia was continued with 0.9 mg/kg per hour of propofol 1% (Disoprivan 1%; AstraZeneca, London, United Kingdom) throughout the procedure. For contrast-enhanced lung imaging, either 70 vol% of krypton or 30 vol% of xenon was added to the ventilated air via the calibrated mass flow meter. Pulse and oxygen saturation were continuously monitored and recorded during the experiment.
Scan Protocol During the entire experiment, the animals remained on the CT table. All CT (SOMATOM Definition Flash; Siemens, Forchheim, Germany) scans were performed in the prone position during endinspiratory breath-hold. A standardized scan protocol with 2 128 0.6-mm collimation, pitch of 0.9, and a gantry rotation time of 270 milliseconds was used. Tube voltage was 100 kV(p) with an effective tube current-time product of 110 mAseff. For image reconstruction, a slice thickness of 0.6 mm with an increment of 0.4 mm was used. A field of view of 250 mm 250 mm and a medium smooth convolution kernel (B31f) were applied.
Data Processing and Analysis For postprocessing, data were transferred to an external workstation equipped with a commercially not available software prototype developed for computing the arterial enhancement fraction of the liver (Hepacare; Siemens Healthcare). This tool is equipped with a nonrigid registration algorithm for registering different contrast phases from a multiphasic contrast-enhanced CT scan.16 With the unenhanced CT volume serving as reference volume, either krypton or xenon images
FIGURE 2. Corresponding sections showing xenon (left) and krypton (middle) enhancement after postprocessing with registration and digital subtraction. There is an excellent correspondence between enhancement patterns between both image sets. The black areas surrounded by lung tissue on postprocessed images correspond to blood vessels as seen on routine CT images (right).
306
www.investigativeradiology.com
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 5, May 2015
Krypton CT Ventilation Imaging
TABLE 2. Mean CT Difference in CT Values Between Unenhanced and Krypton and Xenon-Enhanced CT
Total Lung Left Lung Right Lung
Xenon, HU
Krypton, HU
35.3 ± 5.5 35.3 ± 5.3 35.5 ± 5.7
21.9 ± 1.8 22.1 ± 3.1 21.5 ± 2.1
There were significant differences between xenon and krypton enhancement (P = 0.0055), whereas there were almost identical results for the left and right lungs. CT indicates computed tomography.
FIGURE 3. Krypton (dashed lines) and xenon (black lines) enhancement as determined from phantom experiments. There is markedly higher enhancement of xenon when compared with krypton.
were registered on the unenhanced CT images. After the registration, the matching of the CT data sets was visually assessed. Thereafter, the unenhanced volume was subtracted from either the krypton- or the xenon-enhanced data sets, providing a subtraction data set that contains only either krypton or xenon enhancement (Fig. 2). These image data were used for further analysis. For quantifying contrast enhancement, a dedicated software tool for semiautomated measurement of pulmonary attenuation was used (CT Pulmo; Siemens Healthcare). First, the lung was semiautomatically detected by the software. Automated lung detection was then corrected manually by one of the investigators. Average CT values were computed for the entire lung as well as for the left and right lungs separately. For assessing homogeneity of gas distribution, attenuations of the left and right lungs were compared. The CT values were compared between krypton- and xenon-enhanced CT scans. Results were compared with the data from the phantom experiments. In addition, all postprocessed images were visually inspected for inhomogeneity of lung enhancement. Results are given as mean ± standard deviation. After testing for normal distribution, paired Student t test was used for assessing differences in CT values. P values of less than 0.05 were considered statistically significant. All statistics were computed with MedCalc 12.5 (MedCalc Software, Mariakerke, Belgium).
RESULTS The phantom studies showed a roughly 4 times higher, linear enhancement for xenon when compared with krypton (Fig. 3). At 100 kV, the linear equation was y = 1.98x − 1000.3 for xenon and y = 0.45x − 999.4 for krypton (Table 1). One animal had to be excluded from the study owing to problems during intubation. All other animal procedures were completed without complications. There were no relevant changes in heart rate or TABLE 1. Linear Equations for Krypton and Xenon as Measured From Phantom Data in Free Air Xenon 80 kV(p) 100 kV(p) 120 kV(p) 140 kV(p)
Krypton 2
y = 2.50–999.5 (R = 1) y = 1.98–1000.3 (R2 = 1) y = 1.43–1000.9 (R2 = 1) y = 1.21–999.3 (R2 = 1)
y = 0.57–1000.4 (R2 = 1) y = 0.45–999.4 (R2 = 1) y = 0.365–1000.2 (R2 = 1) y = 0.318–998.5 (R2 = 1)
The coefficient of determination is given in brackets.
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
oxygen saturation between unenhanced, krypton-, or xenon-enhanced CT scans. Corresponding heart rates were 128.7 ± 32.3 beats per minute, 128.4 ± 35.5 beats per minute, and 131.6 ± 37.7 beats per minute (P > 0.05), whereas oxygen saturation was 99.1% ± 0.9%, 99.3% ± 1.1%, and 99.1% ± 1.6% (P > 0.05). Visual assessment showed no registration errors using the nonrigid registration algorithm. On visual assessment, there was no inhomogeneity in lung enhancement and neither in craniocaudal nor in ventrodorsal direction. Mean attenuation difference between unenhanced and xenon-enhanced CT was 35.3 ± 5.5 HU for xenon and 21.9 ± 1.8 HU for krypton (Table 2). There were no differences if xenon or krypton was used first. There was no significant difference in attenuation between the left and right lungs with a mean difference of 2.2 ± 1.0 HU in xenon and 2.1 ± 1.1 HU in krypton ventilation (P > 0.05). The enhancement difference between krypton and xenon ventilation was statistically significant (P = 0.0055).
DISCUSSION Computed tomographic ventilation imaging using xenon as contrast agent has first been described in the late 1970s.1 Despite promising results, it did not become a clinical routine technique, probably because of limitations in scanning speed, lack of easy-to-use postprocessing software, and high costs of xenon. With the introduction of dualenergy CT, xenon CT ventilation imaging was rediscovered. Several groups proved its use feasible.3,5,6 Its clinical application, however, is limited by the high price of xenon and its anesthetic potential. Only recently, krypton has been rediscovered as an alternative without the potential deleterious adverse effects of xenon such as respiratory depression. So far, there is only a single patient study proving krypton ventilation imaging with dual-energy CT feasible.12 In this study, we proved the feasibility of krypton ventilation CT imaging with additional digital subtraction for visualization. The technique proved feasible and robust in a healthy animal model. Although the tool was initially designed for liver imaging, it performed well for lung imaging without obvious misregistration. A nonrigid registration was used to overcome potential differences in breath-hold position and subsequent differences in lung volumes on repeated scans. The CT values as determined from postprocessed images were higher when compared with the initial patient data on krypton ventilation dualenergy CT,12 which were almost identical to recently reported animal data10 but lower as expected from phantom data. The difference in comparison to the phantom data is likely caused by the mixing of gas with air from the animal's lungs and airways. The differences with the patient study are most likely caused by differences in krypton application; whereas the animals in this study and those in a previous study were intubated10 and had a prolonged washin phase, all patients in the study by Hachulla et al12 received krypton via a mask after only a short washin phase. Despite efforts to seal the mask, it potentially allows for some leakage of gas and consequently lowers enhancement. In addition, variations in regional ventilation are known to contribute to this effect. www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
307
Investigative Radiology • Volume 50, Number 5, May 2015
Mahnken et al
To minimize this effect, a 2-minute wash-in phase was used, but it is well known that CT values will fall short of the theoretically feasible attenuation values and that prolonged wash-in phases are used to overcome this problem.7 The length of the wash-in phase will contribute to the height of intrapulmonary enhancement. Accordingly, the enhancement in this animal study with a long wash-in period was approximately 10 HU higher when compared with a short 5 respiratory maneuvers of wash-in phase as it was used in a patient study. In this patient study, the mean enhancement on dual-energy krypton maps after 80% krypton inhalation was approximately 12.4 HU.12 In addition to the length of the wash-in period, other factors such as inaccuracies of dual-energy CT are likely to contribute to these differences. In fact, dual-energy CT loses some information to technical effects such as cross-scatter or partial volume effects. Moreover, quantitative results of dual-energy CT strongly depend on the setting of the material decomposition. From theory, a perfect subtraction imaging as it was used in this study should overcome these problems. Moreover, it could be more widely used because it does not require a dedicated hardware such as a dual-source CT scanner. This study has some limitations. First, the rabbit model may limit transfer to humans because the lung volume of rabbits is comparably small. Nevertheless, the use of rabbits is an established model for lung ventilation10 and permitted sufficient testing of the registration software. Second, all scans were obtained only in the prone position. However, from previous animal studies, it is known that there are no significant differences between both positions,17 with the prone position being somewhat superior to the supine position.11 Moreover, in clinical routine, most patient studies only scan at a single patient position to minimize radiation exposure. Third, the subtraction technique requires 2 scans, whereas a dual-energy CT scan can be performed at a lower radiation exposure, when compared with 2 single-energy scans. The described subtraction technique, however, seems to work with a single baseline scan for multiple krypton-/xenon-enhanced scans as an approach toward dose reduction In contrast to this feasibility study, there is much more potential for dose reduction. For further studies, lower tube voltages such as 80 kV(p) or even 70 kV(p) need to be applied to bring down radiation exposure. This would also result in a relevant increase of CT attenuation values and therefore improve image quality. Nevertheless, in clinical routine practice, only a limited number of patients are eligible for 80-kV scanning; therefore, the animal studies were limited to 100-kV imaging. Finally, only a small number of animals and only healthy animals were included in this study without lung pathology such as air trapping. This issue needs to be addressed by further studies. In conclusion, the use of krypton for contrast enhancement in subtraction CT is feasible and seems to be safe. Although the changes in attenuation are less pronounced than those with xenon, krypton
308
www.investigativeradiology.com
provides sufficient attenuation to permit contrast-enhanced airway imaging. Further studies are needed to investigate the use of kryptonenhanced subtraction CT in a clinical setting. REFERENCES 1. Gur D, Shabason L, Borovetz HS, et al. Regional pulmonary ventilation measurements by xenon enhanced dynamic computed tomography: an update. J Comput Assist Tomogr. 1981;5:678–683. 2. Murphy DM, Nicewicz JT, Zabbatino SM, et al. Local pulmonary ventilation using nonradioactive xenon-enhanced ultrafast computed tomography. Chest. 1989;96:799–804. 3. Chae EJ, Seo JB, Lee J, et al. Xenon ventilation imaging using dual-energy computed tomography in asthmatics: initial experience. Invest Radiol. 2010;45: 354–361. 4. Goo HW, Yang DH, Hong SJ, et al. Xenon ventilation CT using dual-source and dual-energy technique in children with bronchiolitis obliterans: correlation of xenon and CT density values with pulmonary function test results. Pediatr Radiol. 2010;40:1490–1497. 5. Park EA, Goo JM, Park SJ, et al. Chronic obstructive pulmonary disease: quantitative and visual ventilation pattern analysis at xenon ventilation CT performed by using a dual-energy technique. Radiology. 2010;256:985–997. 6. Zhang LJ, Zhou CS, Schoepf UJ, et al. Dual-energy CT lung ventilation/perfusion imaging for diagnosing pulmonary embolism. Eur Radiol. 2013;23:2666–2675. 7. Chae EJ, Seo JB, Goo HW, et al. Xenon ventilation CT with a dual-energy technique of dual-source CT: initial experience. Radiology. 2008;248:615–624. 8. Kong X, Sheng HX, Lu GM, et al. Xenon-enhanced dual-energy CT lung ventilation imaging: techniques and clinical applications. AJR Am J Roentgenol. 2014;202:309–317. 9. Fuld MK, Halaweish AF, Newell JD Jr, et al. Optimization of dual-energy xenoncomputed tomography for quantitative assessment of regional pulmonary ventilation. Invest Radiol. 2013;48:629–637. 10. Chung YE, Hong SR, Lee MJ, et al. Krypton-enhanced ventilation CT with dual energy technique: experimental study for optimal krypton concentration. Exp Lung Res. 2014;40:439–446. 11. Marcucci C, Nyhan D, Simon BA. Distribution of pulmonary ventilation using Xe-enhanced computed tomography in prone and supine dogs. J Appl Physiol (1985). 2001;90:421–430. 12. Hachulla AL, Pontana F, Wemeau-Stervinou L, et al. Krypton ventilation imaging using dual-energy CT in chronic obstructive pulmonary disease patients: initial experience. Radiology. 2012;263:253–259. 13. Winkler SS, Holden JE, Sackett JF, et al. Xenon and krypton as radiographic inhalation contrast media with computerized tomography: preliminary note. Invest Radiol. 1977;12:19–20. 14. Cullen SC, Gross EG. The anesthetic properties of xenon in animals and human beings, with additional observations on krypton. Science. 1951;113:580–582. 15. Goris ML, Daspit SG, Walter JP, et al. Applications of ventilation lung imaging with 81mKrypton. Radiology. 1977;122:399–403. 16. Mahnken AH, Klotz E, Schreiber S, et al. Volumetric arterial enhancement fraction predicts tumor recurrence after hepatic radiofrequency ablation of liver metastases: initial results. AJR Am J Roentgenol. 2011;196:W573–W579. 17. Chon D, Beck KC, Simon BA, et al. Effect of low-xenon and krypton supplementation on signal/noise of regional CT-based ventilation measurements. J Appl Physiol (1985). 2007;102:1535–1544.
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Krypton for Computed Tomography Lung Ventilation Imaging Preliminary Animal Data Andreas H. Mahnken, MD, MBA, MME,* Gregor Jost, MSc,† and Hubertus Pietsch, PhD† Objectives: The objective of this study was to assess the feasibility and safety of krypton ventilation imaging with intraindividual comparison to xenon ventilation computed tomography (CT). Materials and Methods: In a first step, attenuation of different concentrations of xenon and krypton was analyzed in a phantom setting. Thereafter, 7 male New Zealand white rabbits (4.4–6.0 kg) were included in an animal study. After orotracheal intubation, an unenhanced CT scan was obtained in end-inspiratory breath-hold. Thereafter, xenon- (30%) and krypton-enhanced (70%) ventilation CTwas performed in random order. After a 2-minute wash-in of gas A, CT imaging was performed. After a 45-minute wash-out period and another 2-minute wash-in of gas B, another CT scan was performed using the same scan protocol. Heart rate and oxygen saturation were measured. Unenhanced and krypton or xenon data were registered and subtracted using a nonrigid image registration tool. Enhancement was quantified and statistically analyzed. Results: One animal had to be excluded from data analysis owing to problems during intubation. The CT scans in the remaining 6 animals were completed without complications. There were no relevant differences in oxygen saturation or heart rate between the scans. Xenon resulted in a mean increase of enhancement of 35.3 ± 5.5 HU, whereas krypton achieved a mean increase of 21.9 ± 1.8 HU in enhancement (P = 0.0055). Conclusions: The use of krypton for lung ventilation imaging appears to be feasible and safe. Despite the use of a markedly higher concentration of krypton, enhancement is significantly worse when compared with xenon CT ventilation imaging, but sufficiently high for CT ventilation imaging studies. Key Words: computed tomography, lung, ventilation, xenon, krypton (Invest Radiol 2015;50: 305–308)
I
maging of airways and lung tissue is a common indication for computed tomography (CT) of the chest. Values of CT as measured in Hounsfield units (HUs) provide a quantitative measure of lung ventilation. Assessment of regional ventilation by means of CT, however, requires inhalation of radiodense xenon. Xenon ventilation CT has shown potential in the assessment of regional ventilation.1,2 It has proven potentially useful in various pathologic conditions such as asthma, chronic obstructive pulmonary disease, bronchiolitis obliterans, and pulmonary embolism.3–6 Limitations regarding the quantification of xenon enhancement caused by different lung volumes between repeated scans have mostly been overcome by the introduction of dual-energy CT imaging because material decomposition allows for the direct assessment of xenon distribution within the lungs.7–9 Because dual-energy CT is not ubiquitously available, alternative means such as advanced image subtraction techniques are needed to assess the relatively small attenuation changes. Received for publication September 21, 2014; and accepted for publication, after revision, November 5, 2014. From the *Department of Diagnostic and Interventional Radiology, Philipps-University, Marburg, and †MR & CT Contrast Media Research, Bayer Healthcare, Berlin, Germany. Conflicts of interest and sources of funding: The project was in parts funded by Bayer Healthcare, Berlin, Germany. The authors report no other conflicts of interest. Reprints: Andreas H. Mahnken, MD, MBA, MME, Department of Diagnostic and Interventional Radiology, University Hospital, Philipps-University Marburg Baldingerstrasse, D-35043 Marburg, Germany. E-mail: [email protected]. Copyright © 2014 Wolters Kluwer Health, Inc. All rights reserved. ISSN: 0020-9996/15/5005–0305
Investigative Radiology • Volume 50, Number 5, May 2015
Nevertheless, xenon ventilation CT has mostly been limited to clinical research because it has several relevant problems. Most importantly, xenon has relevant adverse effects including somnolence, headaches, and nausea. To keep these adverse effects under control, its concentration is limited to 30% of a xenon-oxygen mixture, but even then up to 70% of patients describe transient adverse effects.4 With krypton, a 70% krypton concentration was shown to provide an acceptable enhancement in dual-energy CT.10 In addition, precision of quantitative CT ventilation measurements depends on the separation of lung and soft tissue compartments. The moderate solubility of xenon in blood will limit the precision of regional ventilation measurements because the uptake of xenon into blood alters the CT values of the portion of a region of interest that does not contain air. If images are obtained in the supine position, gravity-dependent gradients in lung ventilation may appear on xenon maps with relatively high xenon content in the gravity-dependent lung areas.8,11 This problem seems to be less pronounced in the prone position.11 Finally, xenon is expensive. Krypton has been recognized as an alternative means to assess regional ventilation using CT imaging.12 It is a noble gas with a sufficiently high atomic number (Z = 36) to visibly alter CT values when inhaled. In fact, this property has been described as early as 1977,13 but it has rarely been used, despite some obvious advantages.12 It has no narcotic potential, and high concentrations can be inhaled without known adverse effects.14 In fact, its safety in low concentrations is well known from nuclear medicine, where 81mKr has been used for decades.15 Its solubility in blood is much lower when compared with xenon, and it only costs a tenth of xenon.13 However, there is a single relevant disadvantage. Its atomic number is much lower than the atomic number of xenon (Z = 54). Therefore, it will provide a markedly lower change in attenuation if administered at identical concentrations. The goal of this experimental study was to assess the feasibility and safety of krypton ventilation CT imaging with digital subtraction for postprocessing.
MATERIALS AND METHODS Phantom Study In a first step, a calibration curve for different concentrations of krypton and xenon mixed with oxygen was generated. For this purpose, pure xenon or krypton was obtained from a pressurized metal storage container (Air Liquide, Düsseldorf, Germany) via a pressure valve (FMD650-03; DruVa, Eppelheim, Germany) connected to a calibrated mass flow meter (Voegtlin, Aesch, Switzerland). The latter allows for a precise measurement and dosing of gas used for the phantom studies. With the pressure valve of the metal storage container set to 1.0 bar, xenon or krypton was continuously insufflated into a 100-mL plastic container until a target concentration balanced with air was achieved. The container was placed inside the gantry of a multislice computed tomography scanner (SOMATOM Definition Flash; Siemens, Forchheim, Germany) and scanned at different tube voltages with the effective tube current-time product normalized to a computed tomography dose indexVol of 4.65 mGy. A standardized scan protocol with 2 128 0.6-mm collimation, a table feed of 0 mm per rotation (normalized pitch = 0), and a gantry rotation time of 270 milliseconds was used. Tube voltages were 80 kV(p), 100 kV(p), 120 kV(p), and 140 kV (p) with corresponding effective tube current-time products of 245 www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
305
Investigative Radiology • Volume 50, Number 5, May 2015
Mahnken et al
Animal Preparation
FIGURE 1. Graphic display of the study design: After an unenhanced CT scan, either krypton- or xenon-enhanced CT ventilation imaging was performed in random order after a 2-minute wash-in phase. Figure 1 can be viewed online in color at www.investigativeradiology.com.
mAseff (at 80 kV[p]), 110 mAseff (at 100 kV[p]), 65 mAseff (at 120 kV [p]), and 42 mAseff (at 140 kV[p]). For image reconstruction, a slice thickness of 3 mm with an increment of 2 mm and a medium smooth convolution kernel (B31f) were chosen with a field of view of 150 mm 150 mm. Computed tomographic values inside the container were measured on 4 consecutive slices. Mean attenuation values as measured in HU were computed and expressed as mean ± standard deviation. Results were displayed graphically, and linear equations for each kilovolt (peak) setting and its coefficient of determination were computed.
Study Design For animal studies, a crossover design was used to allow for intraindividual comparison. First, an unenhanced baseline scan with normal air ventilation was obtained. Thereafter, the animals were randomly assigned to start with either krypton or xenon ventilation CT. After a 2-minute wash-in phase, either krypton- or xenon-enhanced CT scans were obtained. Thereafter, the animals underwent a 45-minute washout phase at room air to clear the airways from residual gas. Finally, a second CT scan was performed with either krypton or xenon ventilation (depending on the gas used for the first scan) after another 2-minute washin phase. After the experiment, each animal had undergone unenhanced (air), krypton-, and xenon-enhanced lung CT (Fig. 1).
A total of 7 New Zealand white rabbits (4.4–6.0 kg) were included in this study after approval from the local animal care committee. For imaging procedures, all rabbits were anesthetized with intramuscular administration of 35 mg/kg of ketamine hydrochloride (Ketavet; Pfizer, Karlsruhe, Germany) and 5 mg/kg of xylazine (Rompun; Bayer Vital, Leverkusen, Germany). The animals were orotracheally intubated and mechanically ventilated at a respiratory rate of 25/min with room air. Anesthesia was continued with 0.9 mg/kg per hour of propofol 1% (Disoprivan 1%; AstraZeneca, London, United Kingdom) throughout the procedure. For contrast-enhanced lung imaging, either 70 vol% of krypton or 30 vol% of xenon was added to the ventilated air via the calibrated mass flow meter. Pulse and oxygen saturation were continuously monitored and recorded during the experiment.
Scan Protocol During the entire experiment, the animals remained on the CT table. All CT (SOMATOM Definition Flash; Siemens, Forchheim, Germany) scans were performed in the prone position during endinspiratory breath-hold. A standardized scan protocol with 2 128 0.6-mm collimation, pitch of 0.9, and a gantry rotation time of 270 milliseconds was used. Tube voltage was 100 kV(p) with an effective tube current-time product of 110 mAseff. For image reconstruction, a slice thickness of 0.6 mm with an increment of 0.4 mm was used. A field of view of 250 mm 250 mm and a medium smooth convolution kernel (B31f) were applied.
Data Processing and Analysis For postprocessing, data were transferred to an external workstation equipped with a commercially not available software prototype developed for computing the arterial enhancement fraction of the liver (Hepacare; Siemens Healthcare). This tool is equipped with a nonrigid registration algorithm for registering different contrast phases from a multiphasic contrast-enhanced CT scan.16 With the unenhanced CT volume serving as reference volume, either krypton or xenon images
FIGURE 2. Corresponding sections showing xenon (left) and krypton (middle) enhancement after postprocessing with registration and digital subtraction. There is an excellent correspondence between enhancement patterns between both image sets. The black areas surrounded by lung tissue on postprocessed images correspond to blood vessels as seen on routine CT images (right).
306
www.investigativeradiology.com
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Investigative Radiology • Volume 50, Number 5, May 2015
Krypton CT Ventilation Imaging
TABLE 2. Mean CT Difference in CT Values Between Unenhanced and Krypton and Xenon-Enhanced CT
Total Lung Left Lung Right Lung
Xenon, HU
Krypton, HU
35.3 ± 5.5 35.3 ± 5.3 35.5 ± 5.7
21.9 ± 1.8 22.1 ± 3.1 21.5 ± 2.1
There were significant differences between xenon and krypton enhancement (P = 0.0055), whereas there were almost identical results for the left and right lungs. CT indicates computed tomography.
FIGURE 3. Krypton (dashed lines) and xenon (black lines) enhancement as determined from phantom experiments. There is markedly higher enhancement of xenon when compared with krypton.
were registered on the unenhanced CT images. After the registration, the matching of the CT data sets was visually assessed. Thereafter, the unenhanced volume was subtracted from either the krypton- or the xenon-enhanced data sets, providing a subtraction data set that contains only either krypton or xenon enhancement (Fig. 2). These image data were used for further analysis. For quantifying contrast enhancement, a dedicated software tool for semiautomated measurement of pulmonary attenuation was used (CT Pulmo; Siemens Healthcare). First, the lung was semiautomatically detected by the software. Automated lung detection was then corrected manually by one of the investigators. Average CT values were computed for the entire lung as well as for the left and right lungs separately. For assessing homogeneity of gas distribution, attenuations of the left and right lungs were compared. The CT values were compared between krypton- and xenon-enhanced CT scans. Results were compared with the data from the phantom experiments. In addition, all postprocessed images were visually inspected for inhomogeneity of lung enhancement. Results are given as mean ± standard deviation. After testing for normal distribution, paired Student t test was used for assessing differences in CT values. P values of less than 0.05 were considered statistically significant. All statistics were computed with MedCalc 12.5 (MedCalc Software, Mariakerke, Belgium).
RESULTS The phantom studies showed a roughly 4 times higher, linear enhancement for xenon when compared with krypton (Fig. 3). At 100 kV, the linear equation was y = 1.98x − 1000.3 for xenon and y = 0.45x − 999.4 for krypton (Table 1). One animal had to be excluded from the study owing to problems during intubation. All other animal procedures were completed without complications. There were no relevant changes in heart rate or TABLE 1. Linear Equations for Krypton and Xenon as Measured From Phantom Data in Free Air Xenon 80 kV(p) 100 kV(p) 120 kV(p) 140 kV(p)
Krypton 2
y = 2.50–999.5 (R = 1) y = 1.98–1000.3 (R2 = 1) y = 1.43–1000.9 (R2 = 1) y = 1.21–999.3 (R2 = 1)
y = 0.57–1000.4 (R2 = 1) y = 0.45–999.4 (R2 = 1) y = 0.365–1000.2 (R2 = 1) y = 0.318–998.5 (R2 = 1)
The coefficient of determination is given in brackets.
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
oxygen saturation between unenhanced, krypton-, or xenon-enhanced CT scans. Corresponding heart rates were 128.7 ± 32.3 beats per minute, 128.4 ± 35.5 beats per minute, and 131.6 ± 37.7 beats per minute (P > 0.05), whereas oxygen saturation was 99.1% ± 0.9%, 99.3% ± 1.1%, and 99.1% ± 1.6% (P > 0.05). Visual assessment showed no registration errors using the nonrigid registration algorithm. On visual assessment, there was no inhomogeneity in lung enhancement and neither in craniocaudal nor in ventrodorsal direction. Mean attenuation difference between unenhanced and xenon-enhanced CT was 35.3 ± 5.5 HU for xenon and 21.9 ± 1.8 HU for krypton (Table 2). There were no differences if xenon or krypton was used first. There was no significant difference in attenuation between the left and right lungs with a mean difference of 2.2 ± 1.0 HU in xenon and 2.1 ± 1.1 HU in krypton ventilation (P > 0.05). The enhancement difference between krypton and xenon ventilation was statistically significant (P = 0.0055).
DISCUSSION Computed tomographic ventilation imaging using xenon as contrast agent has first been described in the late 1970s.1 Despite promising results, it did not become a clinical routine technique, probably because of limitations in scanning speed, lack of easy-to-use postprocessing software, and high costs of xenon. With the introduction of dualenergy CT, xenon CT ventilation imaging was rediscovered. Several groups proved its use feasible.3,5,6 Its clinical application, however, is limited by the high price of xenon and its anesthetic potential. Only recently, krypton has been rediscovered as an alternative without the potential deleterious adverse effects of xenon such as respiratory depression. So far, there is only a single patient study proving krypton ventilation imaging with dual-energy CT feasible.12 In this study, we proved the feasibility of krypton ventilation CT imaging with additional digital subtraction for visualization. The technique proved feasible and robust in a healthy animal model. Although the tool was initially designed for liver imaging, it performed well for lung imaging without obvious misregistration. A nonrigid registration was used to overcome potential differences in breath-hold position and subsequent differences in lung volumes on repeated scans. The CT values as determined from postprocessed images were higher when compared with the initial patient data on krypton ventilation dualenergy CT,12 which were almost identical to recently reported animal data10 but lower as expected from phantom data. The difference in comparison to the phantom data is likely caused by the mixing of gas with air from the animal's lungs and airways. The differences with the patient study are most likely caused by differences in krypton application; whereas the animals in this study and those in a previous study were intubated10 and had a prolonged washin phase, all patients in the study by Hachulla et al12 received krypton via a mask after only a short washin phase. Despite efforts to seal the mask, it potentially allows for some leakage of gas and consequently lowers enhancement. In addition, variations in regional ventilation are known to contribute to this effect. www.investigativeradiology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
307
Investigative Radiology • Volume 50, Number 5, May 2015
Mahnken et al
To minimize this effect, a 2-minute wash-in phase was used, but it is well known that CT values will fall short of the theoretically feasible attenuation values and that prolonged wash-in phases are used to overcome this problem.7 The length of the wash-in phase will contribute to the height of intrapulmonary enhancement. Accordingly, the enhancement in this animal study with a long wash-in period was approximately 10 HU higher when compared with a short 5 respiratory maneuvers of wash-in phase as it was used in a patient study. In this patient study, the mean enhancement on dual-energy krypton maps after 80% krypton inhalation was approximately 12.4 HU.12 In addition to the length of the wash-in period, other factors such as inaccuracies of dual-energy CT are likely to contribute to these differences. In fact, dual-energy CT loses some information to technical effects such as cross-scatter or partial volume effects. Moreover, quantitative results of dual-energy CT strongly depend on the setting of the material decomposition. From theory, a perfect subtraction imaging as it was used in this study should overcome these problems. Moreover, it could be more widely used because it does not require a dedicated hardware such as a dual-source CT scanner. This study has some limitations. First, the rabbit model may limit transfer to humans because the lung volume of rabbits is comparably small. Nevertheless, the use of rabbits is an established model for lung ventilation10 and permitted sufficient testing of the registration software. Second, all scans were obtained only in the prone position. However, from previous animal studies, it is known that there are no significant differences between both positions,17 with the prone position being somewhat superior to the supine position.11 Moreover, in clinical routine, most patient studies only scan at a single patient position to minimize radiation exposure. Third, the subtraction technique requires 2 scans, whereas a dual-energy CT scan can be performed at a lower radiation exposure, when compared with 2 single-energy scans. The described subtraction technique, however, seems to work with a single baseline scan for multiple krypton-/xenon-enhanced scans as an approach toward dose reduction In contrast to this feasibility study, there is much more potential for dose reduction. For further studies, lower tube voltages such as 80 kV(p) or even 70 kV(p) need to be applied to bring down radiation exposure. This would also result in a relevant increase of CT attenuation values and therefore improve image quality. Nevertheless, in clinical routine practice, only a limited number of patients are eligible for 80-kV scanning; therefore, the animal studies were limited to 100-kV imaging. Finally, only a small number of animals and only healthy animals were included in this study without lung pathology such as air trapping. This issue needs to be addressed by further studies. In conclusion, the use of krypton for contrast enhancement in subtraction CT is feasible and seems to be safe. Although the changes in attenuation are less pronounced than those with xenon, krypton
308
www.investigativeradiology.com
provides sufficient attenuation to permit contrast-enhanced airway imaging. Further studies are needed to investigate the use of kryptonenhanced subtraction CT in a clinical setting. REFERENCES 1. Gur D, Shabason L, Borovetz HS, et al. Regional pulmonary ventilation measurements by xenon enhanced dynamic computed tomography: an update. J Comput Assist Tomogr. 1981;5:678–683. 2. Murphy DM, Nicewicz JT, Zabbatino SM, et al. Local pulmonary ventilation using nonradioactive xenon-enhanced ultrafast computed tomography. Chest. 1989;96:799–804. 3. Chae EJ, Seo JB, Lee J, et al. Xenon ventilation imaging using dual-energy computed tomography in asthmatics: initial experience. Invest Radiol. 2010;45: 354–361. 4. Goo HW, Yang DH, Hong SJ, et al. Xenon ventilation CT using dual-source and dual-energy technique in children with bronchiolitis obliterans: correlation of xenon and CT density values with pulmonary function test results. Pediatr Radiol. 2010;40:1490–1497. 5. Park EA, Goo JM, Park SJ, et al. Chronic obstructive pulmonary disease: quantitative and visual ventilation pattern analysis at xenon ventilation CT performed by using a dual-energy technique. Radiology. 2010;256:985–997. 6. Zhang LJ, Zhou CS, Schoepf UJ, et al. Dual-energy CT lung ventilation/perfusion imaging for diagnosing pulmonary embolism. Eur Radiol. 2013;23:2666–2675. 7. Chae EJ, Seo JB, Goo HW, et al. Xenon ventilation CT with a dual-energy technique of dual-source CT: initial experience. Radiology. 2008;248:615–624. 8. Kong X, Sheng HX, Lu GM, et al. Xenon-enhanced dual-energy CT lung ventilation imaging: techniques and clinical applications. AJR Am J Roentgenol. 2014;202:309–317. 9. Fuld MK, Halaweish AF, Newell JD Jr, et al. Optimization of dual-energy xenoncomputed tomography for quantitative assessment of regional pulmonary ventilation. Invest Radiol. 2013;48:629–637. 10. Chung YE, Hong SR, Lee MJ, et al. Krypton-enhanced ventilation CT with dual energy technique: experimental study for optimal krypton concentration. Exp Lung Res. 2014;40:439–446. 11. Marcucci C, Nyhan D, Simon BA. Distribution of pulmonary ventilation using Xe-enhanced computed tomography in prone and supine dogs. J Appl Physiol (1985). 2001;90:421–430. 12. Hachulla AL, Pontana F, Wemeau-Stervinou L, et al. Krypton ventilation imaging using dual-energy CT in chronic obstructive pulmonary disease patients: initial experience. Radiology. 2012;263:253–259. 13. Winkler SS, Holden JE, Sackett JF, et al. Xenon and krypton as radiographic inhalation contrast media with computerized tomography: preliminary note. Invest Radiol. 1977;12:19–20. 14. Cullen SC, Gross EG. The anesthetic properties of xenon in animals and human beings, with additional observations on krypton. Science. 1951;113:580–582. 15. Goris ML, Daspit SG, Walter JP, et al. Applications of ventilation lung imaging with 81mKrypton. Radiology. 1977;122:399–403. 16. Mahnken AH, Klotz E, Schreiber S, et al. Volumetric arterial enhancement fraction predicts tumor recurrence after hepatic radiofrequency ablation of liver metastases: initial results. AJR Am J Roentgenol. 2011;196:W573–W579. 17. Chon D, Beck KC, Simon BA, et al. Effect of low-xenon and krypton supplementation on signal/noise of regional CT-based ventilation measurements. J Appl Physiol (1985). 2007;102:1535–1544.
© 2014 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
