Positron Emission Tomography (PET) for Assessing Aerosol Deposition of Orally Inhaled Drug Products

Abstract

The topical distribution of inhaled therapies in the lung can be viewed using radionuclides and imaging. Positron emission tomography (PET) is a three-dimensional functional imaging technique providing quantitatively accurate localization of the quantity and distribution of an inhaled or injected PET radiotracer in the lung. A series of transaxial slices through the lungs are obtained, comparable to an X-ray computed tomography (CT) scan. Subsequent reformatting allows coronal and sagittal images of the distribution of radioactivity to be viewed. This article describes procedures for administering [¹⁸F]-fluorodeoxyglucose aerosol to human subjects for the purpose of determining dose and distribution following inhalation from an aerosol drug delivery device (ADDD). The advantages of using direct-labeled PET drugs in the ADDD are discussed with reference to the literature. The methods for designing the inhalation system, determining proper radiation shielding, calibration, and validation of administered radioactivity, scanner setup, and data handling procedures are described. Obtaining an X-ray CT or radionuclide transmission scan to provide accurate geometry of the lung and also correct for tissue attenuation of the PET radiotracer is discussed. Protocols for producing accurate images, including factors that need to be incorporated into the data calibration, are described, as well as a proposed standard method for partitioning the lung into regions of interest. Alternate methods are described for more detailed assessments. Radiation dosimetry/risk calculations for the procedures are appended, as well as a sample data collection form and spreadsheet for calculations. This article should provide guidance for those interested in using PET to determine quantity and distribution of inhaled therapeutics.

Introduction

Positron emission tomography (PET) is one of the major current diagnostic imaging tools, utilizing molecular markers labeled with positron-emitting radionuclides and synthesized with a variety of pharmaceutical agents. These unique products along with specialized equipment and specific methodologies allow investigations of biologic processes.⁽¹⁾ As the radionuclide is incorporated into the drug molecule, the measurement of drug deposited in and absorbed from the lung will reflect the deposition of the pharmaceutical itself and not just the free radionuclide. Although clinical oncology is the area where PET is used routinely to diagnose malignancies, monitor disease progression, and stage treatment,^(2,3) the development of PET-labeled drugs also allows the biodistribution and pharmacokinetics (PK) of these labeled drugs to be determined in vivo.^(4–6) PET has the advantages over single photon-emission computed tomography (SPECT) of being easily quantitative and allowing for tomographic dynamic acquisition (very limited for SPECT), as well as offering the possibility of performing acquisitions physiologically gated to the respiratory cycle, thereby reducing the blurring due to the expanding chest and lungs during breathing. PET investigations with tracer microdoses are also increasingly being used to establish “proof of concept” in drug development.^(7,8)

The PET imaging technique relies on the coincidence detection of two 511 keV photons, the by-products of the annihilation of a positron with its antiparticle, the ordinary, negatively charged electron, found in tissue. The two photons that are emitted travel in opposite directions, and are detected by the fixed radiation detectors mounted in a ring or partial ring structure within the scanner head that surround the source of radioactivity (lungs) when the subject is maneuvered into position in the scanner. When the two photons are registered by the detectors, the line along which the annihilation has taken place is defined, thus giving the location of the labeled PET molecule within the lung. With the accumulation of a number of these events, the distribution of radioactivity in the lung is defined.

A series of transaxial slices through the lungs are subsequently obtained, comparable to a computed tomography (CT) scan. This transaxial information is used to reconstruct the lung activity for the other two orthogonal planes, enabling coronal and sagittal images of the lungs to be viewed. PET spatial resolution is approximately 4–6 mm in plane, with each plane of the order of 2–5 mm thick in the z or axial direction. Correction for photon attenuation and scattering is made directly by the scanner after obtaining a transmission scan using an external radionuclide source (such as ⁶⁸Ge or ¹³⁷Cs) or using a CT scan. These corrections are applied to each voxel in each slice of each plane, allowing absolute amounts of radioactivity to be measured per milliliter of tissue, and resulting in the actual topographic distribution of drug throughout the lung. The transmission image or CT scan of the thorax is also used to define the lung borders, providing landmarks from which to define regions of interest (ROIs).

Combination PET/CT scanners are now widely used in nuclear medicine.⁽⁹⁾ The high-resolution CT (HRCT) component of these hybrid systems is used to provide low-resolution electron density maps. These are used to obtain tissue attenuation factors that are applied to the PET image during data reconstruction.⁽¹⁰⁾ The HRCT is also used to delineate lung and lobar boundaries, and thus provide precise anatomical location of tumors and help define individual airways when investigating changes in airway geometry in disease or with treatment.^(11,12)

The protocols below describe the process of acquiring three-dimensional (3D) images of an inhaled radioaerosol in the lungs by utilizing molecular markers and functional imaging with PET. The radioaerosol administered will be a product containing a positron-emitting radionuclide (511 keV). There are several possible PET radionuclides for use with inhaled medications, with either direct-labeled active pharmaceutical ingredients (APIs)^(13,14) or APIs prepared in saline.⁽⁶⁾ The choice for the former approach is dependent on whether a molecular marker can be inserted into a drug molecule, the production facilities available, and the research question. In the latter, where the objective is to evaluate lung deposition from a nebulizer, the most common product to use is [¹⁸F]-fluorodeoxyglucose ([¹⁸F]-FDG), prepared in saline. [¹⁸F]-FDG can also be spray-dried to produce a powder for inhalation,⁽¹⁵⁾ thus allowing deposition from dry powder inhalers (DPIs) to be assessed using PET technology. Table 1 lists the possible radionuclides/compounds for various PET investigations, some of which are applicable to measurements of lung deposition and clearance from inhalers ([¹¹C]-triamcinolone acetonide,⁽¹⁶⁾ [¹¹C]-zanamavir,⁽¹⁷⁾ [¹⁸F]-fluticasone propionate^(17,18)). Studies of lung function using PET-labeled drugs require radiochemistry and extensive validation/calibration before use in animal models and in human subjects. The following text will address the use of nebulized [¹⁸F]-FDG for lung deposition measurements, but can easily be modified to evaluate drug delivery from other portable inhalers.

Table 1.

Characteristics of Radiolabels for Use with Pet to Evaluate Lung Deposition, PK, and Absorption ⁽⁵⁴⁾

Radionuclide	Half-Life	Radiopharmaceutical	Use
¹⁵O	2 min	[¹⁵O]-H₂O	Blood flow
		[¹⁵O]-CO	Blood volume
¹³N	10 min	[¹³N]-N₂	Ventilation
		[¹³N]-NH₃	Blood flow
¹¹C	20 min	[¹¹C]-Acetate	Metabolism
		[¹¹C]-CGP12177	β-Receptors
		[¹¹C]-MQNB	Muscarinic receptors
		[¹¹C]-PK11195	Microglial label
		[¹¹C]-5HT	Amine extraction
		[¹¹C]-Zanamivir	Aerosol deposition, PK
		[¹¹C]-Triamcinolone acetonide	Aerosol deposition, PK
		[¹¹C]-Albuterol	Aerosol deposition, PK
		[¹¹C]-Salmeterol	Aerosol deposition, PK
¹⁸F	110 min	[¹⁸F]-FDG	Metabolism
			Inflammation
			Aerosol deposition
		[¹⁸F]-Dopamine	Sympathetic innervation
		[¹⁸F]-Fluticasone dipropionate	Aerosol deposition, PK
		[¹⁸F]-Fluorocaptopril	ACE^a inhibition
		[¹⁸F]-Insulin	Aerosol deposition, PK
⁶⁸Ga	68 min	[⁶⁸Ga]-Galligas	Ventilation

ACE, angiotensin-converting enzyme.

A general schema for planning PET deposition studies is shown in Figure 1.⁽¹⁹⁾ The production of PET radionuclides usually requires the use of a cyclotron facility and specialized chemistry.

FIG. 1.

Schematic representation detailing major processes for producing and using a PET-labeled aerosol for deposition imaging. PET studies are logistically complex, requiring a number of methods to be in place prior to exposure of human subjects to the PET tracers.⁽²⁰⁾

As the PET label is directly attached to the drug molecule, radiolabeled PET drugs must be manufactured to Good Manufacturing Practice (GMP) standards and with regulatory approvals in place prior to administering to human subjects.⁽²⁰⁾ As with two-dimensional (2D) planar and 3D SPECT imaging, systems must be calibrated and validated prior to initiation of the study protocol. The aerosol delivery system may be the same ones as used in gamma camera studies; however, the major difference is in the shielding around the source and in the handling of the PET radiopharmaceuticals. Table 2 lists the various processes and steps that an investigator should be aware of and that need to be taken prior to preparing for and initiating a PET radioaerosol study.

Table 2.

Description of Processes Undertaken Prior to Initiating a PET Clinical Aerosol Deposition Study

• Licenses and permits for handling, storage, and use of radioactive substances

• PET radionuclide selection and radiolabeling/production of drug aerosol product

• Equipment calibration/SOPs defined for preparation of drug delivery system

• Validation of radiolabeled aerosol product

– Performance and characterization using selected aerosol drug delivery system (emitted dose of tracer±drug, aerosol particle size distribution of tracer±drug)

• Ethics, regulatory, radiation safety approvals for clinical trial

– Study protocol approvals (REB, RPC, Regulatory Agency for NDA or ANDA)

– Sample size/statistical analysis protocol

– Dosimetry estimation

• Acquisition protocols—imaging protocols

– Calibration, validation of scanners

– Transmission scan or CT

– Emission scan

• Blood sampling (if PK part of study outcomes)

– Protocol development

– Validation of measurements: tracer±drug

• Clinical trial

– Clinical coordinator

• Clinical SOPs for study

• CRF development/approval

– Recruitment (healthy volunteers or subjects with respiratory disease)

– Randomization code/enrollment of subjects/consent form

– Screening (demographics, history, physical exam, blood work, PFTs, sputum, current medications, etc.)

– PD/PK measurements pre- and post-inhalation maneuvers

– Preparation/inhalation of radiolabeled drug dose

• Data processing and analysis protocols

– Image reconstruction

• Decay correction

• Attenuation correction

• Partial volume correction

• Image fusion: transmission or CT with emission

– Definition of lung edge and ROIs/shells

– Calculation of deposition results, time-activity curves (if a study outcome), etc.

1. Prestudy preparation

1.1. Validation of radiolabeling of drug^a

The in vitro validation protocol for a positron-emitting radioaerosol is similar to that for single photon-emitting radionuclides, with the exception of shielding requirements, appropriate counting equipment for the PET radionuclide, and standards preparation. The in vitro emitted dose of radiolabeled test drug from the inhaler (ED_I), as well as the particle size distributions (PSD_I) of test drug and radiolabeled aerosol emitted from the inhaler, needs to be determined prior to these same measurements on the day of the in vivo imaging measurements. A shielded dose collection unit (DCU) and cascade impactor are used for emitted dose and particle size distribution measurements from the nebulizers, pressurized metered dose inhalers (pMDIs), and DPIs used to deliver the PET-labeled aerosolized drug. When using [¹⁸F]-FDG, a range of ¹⁸F standards in saline should be prepared, with known quantities of ¹⁸F spanning the expected values on the inlet, impactor plates, and filter. These standards are used to determine the linearity, reproducibility, and efficiency of the system used to measure the radioactivity in the DCU and that for counting the impactor plates. When using a ¹⁸F-labeled formulation, for example, fluticasone dipropionate, assays of radioactivity are performed first, and when the radionuclide has decayed sufficiently (usually after 6 half-lives; the half-life of ¹⁸F is 109.5 min), the drug can be assayed from the plates and DCUs. Full recovery from the impactor plates, filter, and inlet should be demonstrated prior to each actual sizing measurement. If a PET-labeled drug is being tested, both the drug and radioactivity of each sample need to be measured when determining particle size metrics and emitted dose, as with a conventional single photon-emitting radionuclide product.⁽²¹⁾ As the production of a radiolabeled inhaler is never exactly reproducible from day to day, it is critical to the accuracy of the analysis of lung deposition data to measure both ED_I and PSD_I for each subject's inhaler on each study day.

When a direct-labeled PET drug is used in a deposition study, PK measurements on the individual study subjects are likely to be undertaken to determine the kinetics of the labeled drug. Standards preparation and calibration of the various systems used with a PET-labeled drug product then become more complex compared with a study using [¹⁸F]-FDG. PK studies involve whole blood sampling, and thus calibration/validation methods for detecting both radioactivity and drug in blood would need to be developed prior to study initiation. ¹⁸F levels in blood can be measured with a conventional gamma counter and drug levels with HPLC. Procedures are the same for all PET radionuclides used in these types of investigations.

1.2. Documentation

An example of a study worksheet is attached (Appendix 1).

1.3. Shielding

Extensive lead shielding is required for the inhalation system, whether positioned in a separate dosing room or beside the scanner, and serves to protect and reduce exposure to staff and subjects from the radioactive material used in the study.⁽²²⁾ The thickness of the lead around the inhalation system will be determined by the amount of [¹⁸F]-FDG placed in the nebulizer reservoir, the nebulizer output, and the particle size. However, approximately 4 mm of lead or equivalent is required to reduce the photon transmission rate by 50%. The equivalent figure for ^99mTc used in SPECT studies is 0.3 mm, and 4 mm of lead reduces the Tc^99m exposure rate by a factor of 10,000.

The objective is to inhale the PET aerosol for as short a time as possible to load the lung with approximately 10–50 MBq. Inhalation time should take no longer than 1–1.5 min. Subjects can inhale in either the upright or supine position. The latter requires that the system be adjacent to the scanner head and able to be removed quickly before scanning commences. The lead shielding may render this not possible due to the weight of the shield, and so inhalation would be done in the department's routine dosing area. Figure 2 shows a shield of 5-cm-thick lead bricks built to accommodate two nebulizers. One nebulizer is shown inside the shield on the left. On the right is a volunteer subject inhaling from this nebulizer. In our (M.B.D.) setup, access to the nebulizer reservoirs for loading with radioactivity was accomplished using a butterfly needle and attached catheter threaded between the lead bricks into the nebulizer bowl. Using a shielded syringe, the radioactive liquid is injected into the bowl via the catheter. The amount of radioactivity required is based on the efficiency of nebulization at the particular nebulizer operating flow rate for a specific fill volume, simulated breathing pattern, and targeted inhalation time (Appendix 1). These data, which characterize the delivery system, need to be determined in advance of the actual study day. This can be done using a low-energy radionuclide or drug as tracer. Note that inhalation times may have to be modified if (1) the subject's inhaled volume is very different from the simulated volume, and (2) there is a delay in the start time for the experiment, as the physical half-life of ¹⁸F is 110 min.

FIG. 2.

(A) Lead shield constructed of 2-in lead bricks to contain nebulizers for generating PET-labeled aerosols. Depending on the nebulizer used, the reservoir dose of radioactivity can be 800–2,400 MBq in a 2-mL fill volume. Subjects inhale with tidal breathing for 60 sec to deposit approximately 40 MBq in the lung, receiving a fraction of the starting nebulizer dose. The remainder is left to decay over time. Radioactive exhalate is captured on the absolute filter. (B) Subject is shown inhaling from the nebulizer placed inside the shield. (Images obtained from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON, Canada)

1.4. Calibration of camera

Calibration factor to convert scanner counts to kBq/mL

This calibration is done routinely for quality control of the PET scanner. After the above validation steps are completed, a quality assurance image should be acquired prior to any study. Using a standard lung phantom, a transmission image or HRCT can be acquired to define the phantom specific volume. The phantom is then filled with a reference activity concentration and imaged. Both structural and emission images are processed with existing scanner software and should constitute the basis of the quality assurance check.

1.5. Dosimetry and risk calculation

See Appendix 2 for calculation of PET radiation exposure. Dosimetry for procedures using CT requires a separate calculation. The total radiation exposure to the subject is the sum of the doses from both scans.⁽²³⁾

2. Preparations on study day

2.1. Documentation

Prepare the study worksheet (Appendix 1).

2.2. Subject preparation

Subject preparation is similar to that which is described under this topic in the SPECT imaging protocol (see Fleming et al., this issue). After any questions and concerns that the subject may have regarding the study protocol have been addressed, the consent form is then signed by the subject on the screening day when demographics and baseline (clinical) measurements are obtained. The subject will also have the imaging procedure explained, and will be shown the scanner and given the opportunity to ask questions. The subject is assigned a Study ID on the screening day. Training in the use of the inhaler occurs on the screening day, with these instructions reinforced on each subsequent scanning day. If medications are to be withheld for the study, the conditions are reviewed by the clinical coordinator with the subject on the screening day. These conditions are reviewed again on each actual study day(s).

Note that conventional clinical measurements, such as spirometry [forced expiratory volume in 1 sec (FEV₁)], should be made prior to initiating the PET scan. Expiratory volumes post scan could be reduced by up to 20% as a result of lying in the supine position on the scanner bed for the duration of the acquisition protocol.

2.3. Dose preparation and measurement of device activity

See section on dose preparation and measurement of device activity in the planar imaging protocol (see Newman et al., section 2.4, this issue). Although the PET procedures are similar, as noted, shielding requirements will be different. In addition, in our (M.B.D.) facility because of government licensing requirements, the preparation of the PET dose will likely be handled by nuclear medicine technologists, rather than by research personnel. The amount of radioactivity withdrawn from source should be measured just prior to inserting into the nebulizer or test inhaler. Note that while the 2-mL nebulizer loading dose can contain 400–1,500 MBq, the subject will inhale only a small fraction (∼10–50 MBq) of this dose. It is critical that the times at which all measurements are made be recorded on the study data sheet.

3. Image acquisition

3.1. Aerosol inhalation prior to imaging

The subject breathes the radioaerosol from the device containing an amount of activity that will deposit approximately 10–50 MBq (∼0.25–1 mCi) in the lungs, in the shortest time possible. The subject typically inhales in the upright, seated position in a separate dosing room. However, the subject could be positioned in the scanner and inhale from the source of radioactivity in the supine position. This means that all tubing carrying aerosol including the expiratory filter will need to be shielded during the inhalation maneuver and removed immediately after the inhalation, as any remaining activity will interfere with the acquired scans. The acquired data are corrected for radionuclide decay during the time of imaging. A further correction will be needed to take into account the time between the start of inhalation of the labeled aerosol and initiation of the PET scan.

3.2. Image acquisition

• Scanner is programmed to acquire axial images from 2–4×15–22 cm bed positions containing sections of lung, corresponding to the axial extent of the field of view of the PET system, beginning above the mouth or at the level of the ears to ensure the oropharynx is included. Older PET scanners are likely to require up to 10 min per bed position for a total of 40 min to complete a whole chest acquisition. Newer machines may require only 10 min (or less) for the complete emission scan. The end-position of the last section should be below the diaphragm. The PET emission acquisition protocol will be programmed from the scanner's computer using the following parameters: scan duration; scan direction; output image type; reconstruction type; scatter correction; image size/zoom; filter full-width half-maximum; iteration/subsets. Newer systems may use time-of-flight (ToF) technology to further improve image quality.

Imaging times will vary for other PET radionuclides. For ¹³N and ¹¹C PET products, with a half-life of approximately 10 and 20 min, respectively, imaging time must be as short as possible and may require higher doses to be administered. This, however, is dependent on the resource and efficiency of production of ¹¹C. The sequence of events for acquisition is:

• Subject is positioned supine on scanner bed and made comfortable;

• Scanning bed is moved into scanner head and subject is positioned within the gantry, usually using lasers built into the scanner;

• Horizontal and vertical positions of the scanner bed are noted for a repeat scan if required;

• Acquisition is initiated. Subject breathes normally throughout the scan.

Figures 3 and 4 show panels of transmission (lung structure) and emission (deposited radioactivity) coronal slices obtained from a healthy volunteer. Similar images in a patient with airways obstruction can be found in an earlier report.⁽²⁴⁾

FIG. 3.

Visual presentation of an emission scan for a healthy volunteer showing the distribution of radioactivity in 48/87 slices. Coronal slices are matched with the corresponding transmission coronal slices (Fig. 4) and similarly for the slices in the sagittal and transaxial planes. (Images obtained from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

FIG. 4.

PET transmission scan in a healthy volunteer showing a section of the available coronal slices from the scanner. The geometry of each slice varies as the scanner moves from the anterior to the posterior lung. There are approximately 128 slices in the axial plane, 87 in the coronal plane, and 47 in the sagittal plane. Data files are acquired in the axial plane with coronal and sagittal slice images available from postreconstruction data. (Images obtained from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

3.2.1. Acquisition of volume image and definition of lung boundaries with CT scans

Current PET scanners are mainly hybrid scanners, with CT capabilities integral with PET hardware/software in the same scanner framework. Since their introduction in 2001, 80% or more of new installations are PET/CT systems, and since 2005 it has become virtually 100%. These dual-purpose machines allow CT and PET scans to be performed sequentially with the patient positioned only once on the same scanner table. Co-registration of data from the two sets of images is an integral part of the hybrid scanner commercial software providing an accurate aligning of the transmission (density map, structure) and emission (radioactivity) data maps. In addition, corrections for tissue attenuation within the imaging field are applied to the emission data on a voxel-by-voxel basis, as well as internal corrections for radionuclide decay occurring during imaging. The lung boundary is defined from the CT scan for the whole lung and specific ROIs within the lungs. As the spatial resolution of current CT scanners is 1 mm or lower, and the PET resolution is 4–6 mm, errors in edge definition may result. Dosimetry calculations for the CT scan (e.g., CT dose index) need to be made or obtained directly from CT scanner software.⁽²⁵⁾

3.2.2. Positioning

A scout image, or CT topogram, is usually acquired for positioning the subject in the scanner and acquired after the emission scan. CT protocols are usually set up by the radiology staff or nuclear medicine technologists. These settings are scanner-dependent, taking into account scanner sensitivity and resolution. In our department (M.B.D.) for the Siemens Biograph16, the CT settings for the scout (low dose) image typically are: tube voltage, 120 kVp; effective beam current, 100 mAs; 0.75 sec gantry rotation; 0.75 mm detector collimation; slice thickness, 1 mm; 15 mm feed/rotation; 1.0 mm reconstruction increment; “B80f” filter kernel. The settings are based on the need for sufficient resolution to delineate the lung and accurately determine volumes and density.

3.2.3. Image acquisition protocol for low-dose CT scan

• The low-dose CT protocol is programmed by a nuclear medicine or radiology technologist. CT parameters are defined by the scanner technologists; the parameters for reconstruction of the images are usually defined by the nuclear medicine physicist. For a PET/CT scan, the sequence of events for acquisition is:

• Subject is in supine position on scanner bed and made comfortable;

• Arms are placed above head or positioned by side, separated from the body with foam bolsters;

• The bed is moved into the CT scanner;

• The subject is asked (trained) to breath-hold at functional residual capacity (FRC) for 10 sec while a 10-sec CT image is acquired. This attempts to match the average breathing pattern during PET acquisition;

• Subject then relaxes and breathes normally during the PET acquisition.

3.2.4. Image acquisition protocol for transmission scan

Older PET cameras were built with an external transmission source, usually ⁶⁸Ge, which decayed in situ to ⁶⁸Ga, a positron-emitting radionuclide. The source is rotated around the patient, providing a lower-resolution density or volume scan but with the required tissue attenuation factors. The resolution of this type of transmission scan is the same as that of the PET emission scan, namely, 4–6 mm. The need to image deposition and uptake as soon as possible after inhalation of the radiolabeled aerosol means that this PET transmission scan is acquired after the emission scan.⁽²⁶⁾ While acquisition times are approximately half for this type of transmission scan compared with the emission scan, maintaining the same position on the scanner bed as for the emission scan is critical to avoid misalignment errors in the data reconstruction.

The following protocol is for those PET scanners without an integral CT, but containing a transmission source. As mentioned above, the transmission source is an external ¹³⁷Cs or ⁶⁸Ge source mounted within the scanner that rotates around the subject. The transmission scan is typically acquired after the emission scan with the subject remaining on the scanner bed in the same position as for the emission scan.

• Scanner is programmed to acquire axial images from 2–4×15–22 cm sections of lung (1–4 min per section) beginning above the mouth, or at the level of the ears, to ensure the oropharynx is included. The last section should be below the diaphragm. The PET transmission setup protocol is usually programmed to be sequential with and after the emission scan protocol so that there is no delay between the two scans;

• Subject remains in the supine position on scanner bed;

• Bed is moved into scanner and subject is positioned within the scanner head using the lasers;

• Horizontal and vertical bed positions are noted;

• Acquisition is initiated. Subject breathes normally throughout the 12–16-min scan;

• Emission data are reconstructed with the transmission data, and both files are available for importing into a display program for image viewing.

3.2.5. Image acquisition protocol for HRCT scan

An HRCT scan provides the anatomical location of areas of increased uptake of the PET marker in the lung, or other organ being scanned. Use of a high-resolution scan may provide details of individual airways of varying caliber and position within the lung, perhaps up to the seventh or ninth generation of airway, depending on the scanner resolution.

The particular settings are somewhat dependent on the properties of the CT scanner. Typical settings for HRCT are: tube voltage, 120 kVp; effective beam current, 70–130 mAs; 0.5 sec gantry rotation; 0.75 mm detector collimation; 5 mm slice thickness; 15 mm feed/rotation; 5.0 mm reconstruction increment; “B40” filter kernel.

3.2.6. Aerosol administration and acquisition protocol for a dynamic scan following inhalation

An advantage of PET investigations with direct-labeled pharmaceutical products is the possibility of measuring uptake of the drug from the lung following inhalation. Imaging, along with blood sampling, can yield information on the regional kinetics of the drug. The following brief protocol describes a method used to measure the absorption kinetics of the inhaled drug from the lung.

• Select the section of lung to be imaged (usually the mid-section of lung);

• Subject is positioned in the scanner head so that the selected section only is imaged. Horizontal and vertical bed coordinates are noted;

• Scanner is programmed to acquire emission counts only from the selected section. Imaging time should bracket the time just prior to the dose of radioactivity being given to the subject (if inhalation is performed while the subject is in position in the scanner) and stop at an appropriate time when sufficient uptake information for the tracer has been acquired. This can be followed by static whole-chest imaging to extend the acquisition time, particularly when the radioactivity count rate is low;

• Frame rates should reflect the expected change in tracer distribution over time—short framing rates initially and longer rates after peak lung activity;

• Amount of radioactivity given needs to be high enough to acquire sufficient detected events at the most rapid framing rate;

• The emission scan is either preceded by a CT scan or followed by a CT or PET transmission scan of the area imaged during the emission scan. A further whole-body CT or PET transmission scan is also acquired to use when reconstructing the whole-chest static PET scans.

4. Image Analysis

4.1. Regional analyses for whole lung and subdivisions

Both SPECT and PET studies of aerosol deposition provide a 3D spatial description of the aerosol deposition pattern. As a result, there are similarities in the approaches to data analysis between the two imaging techniques.

4.1.1. Segmentation of the lung

The first step in the data-handling program is to segment the lungs, and this requires a method to define the lung border. In x-ray imaging, the lung itself is generally referred to as a high-contrast organ, because the density differences are high compared with those of the external soft tissue. This is relatively straightforward when working with a CT scan where the resolution is approximately 1 mm, but not so for a radionuclide-based PET transmission scan, with a resolution closer to 4 mm. Section 4.1 in the SPECT chapter (see Fleming et al., this issue) provides details for several algorithms that can be used to define the 3D border from either a transmission scan or CT scan.

Another approach involves an adaptive thresholding technique. This can also be used to segment (extract) the lung from its surrounding high-density soft tissue. However, an accurate segmentation of lung from soft tissue on the basis of transmission PET data sets is, in some cases, challenging due to the limited spatial resolution of the PET transmission scan, and to the overlapping of densities at the edges of the lung. With a thresholding method based on a single level, it can be difficult to determine with good accuracy the exact edge of the lung. The outcome of the segmentation therefore depends strongly on the chosen density threshold level. A low threshold level will underestimate the lung volume, and a higher level will overestimate the lung size and may lead to “leaking out” of the segmented volume, particularly in areas with a thin chest wall.

Using MATLab 7.0, we (M.B.D.) developed an algorithm based on Otsu's method⁽²⁷⁾ to threshold the thoracic PET transmission images. The original images are converted to a set of two main groups of pixels: high-intensity pixels located in the body, and low-intensity pixels that are in the lung and the surrounding air. Due to the large difference in intensity between these two groups, this adaptive thresholding has led to a better separation. An example of a 2D slice using this method is shown in Figure 5. The accuracy of the software algorithm was tested using different lung phantoms with a known volume, and the accuracy was within 2%.

FIG. 5.

Original transverse image (left) and the corresponding thresholded result. (Images obtained from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

4.1.2. Defining ROIs within the lung: concentric shells

The volume image of the total lung imaged by PET can be further segmented into 3D concentric shells, centered about the carina, analogous to what is described in the SPECT section (see Fleming et al., section 4.5.2, this issue), but where the hilum is used as the starting point. The volume of each shell in both lungs, as well as the total volume for each lung, is calculated by computer programs. In our model, the 3D shell generation program begins by drawing a number of vectors on the mid-transaxial slice of the transmission map and projecting each vector to the outer edge of each lung (Fig. 6). The entry, or focus point, is positioned by the user to be 1 cm above the carina. Each vector is divided into 10 equal sections and contours generated to form 10 shells (right lung) and 10 shells (left lung), concentric about the carina. The volume per shell is calculated as the product of voxel volume and the number of voxels per shell.

FIG. 6.

Shell generation. (A) Cartoon showing sample vectors drawn from the focus point within the trachea, 1 cm above the carina to the lung edge. Generated shells are shown in grayscale. The focus point can also be set to a point within the main stem bronchus. (B) Examples of vectors superimposed on an emission scan of the right lung. The 3D shells are automatically generated by the software program. (Images obtained from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

The number of shells per lung can vary. This choice is made by selecting the number of shells in the software program when processing the data.^(28,29) Similarly, the position of the focus point can be relocated to the hilum of the first bifurcation of the right main stem bronchus, generating 10 shells for the right lung only. This latter approach is similar to that described in the SPECT section for shell generation (see Fleming et al., this issue).

4.2. Quantifying radioactivity in the defined lung ROIs

The third step in the program is to apply the segmented lung edges to the emission images to extract and quantify aerosol deposition within the areas of interest. In PET, image degradation effects such as photon attenuation, scatter, and dead-time need to be accurately accounted for, as well as the correction for the partial volume effect.^(30,31) Although the scanner applies corrections for such effects, the static and dynamic imaging protocols used (framing duration and rate) and the amount of radioactivity may be different (lower) from the usual clinical situation, and these differences can lead to imaging artifacts. There is also a possibility for a malfunctioning of the scanner at the time of conducting a study. Therefore, for accuracy in determining deposited dose, it is a good practice to perform dose verification and validate the reconstructed images on the basis of the amount of activity present within a certain test volume (i.e., activity concentration). Figure 7A shows an example of a segmented lung, and Figure 7B the 3D shells (10 per lung) generated by the Analyze software analysis program.

FIG. 7.

(Top) Illustration of segmented lungs and generated 3D shells within each lung. (Bottom) Volumes of each lung and the individual shells can be obtained from our MATLab program (The MathWorks, Inc., Natick, MA), allowing a normalization of the emission data from each lung and shell. (Images were obtained from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON, Canada.)⁽³⁶⁾ The 3D shells, concentric about the carina, have been generated using our in-house (Analyze) program written in MATLab. The transmission scan contour of the lung volume is used to define the lung edge and the concentric shells are then generated, each following this outline. Each panel represents one lung slice with the generated shells in the specific designated plane. It is evident from the slices shown in the example in Figure 4 that the geometry changes from the anterior to posterior sections of the lung. As a result, the number of shells that are accommodated by each slice will vary. The shell structure is applied to the matching emission slices (Figure 3), and the regional distribution of the inhaled dose is calculated.

4.3. Calculations

Once the reconstructed data are obtained from the scanner, the following calculations can be made and the data entered into or sent directly from within the software program to a spreadsheet for further manipulation.⁽³²⁾

a. Volumes (V, mL) are calculated from the number of voxels, obtained from the PET transmission or CT scan and the voxel size: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*} {\rm V_{Lung}} = \mathop{\Sigma\rm V}_{i = 1}^{i= n} {\rm Shells} \ or \ \mathop{\Sigma\rm V}_{i = 1}^{i= s} {\hbox{transaxial slices}}\end{align*} \end{document}

where n=number of shells; s=number of slices.

The voxel volume for the ECAT/ART scanner is 0.324 mm³, but these dimensions will vary depending on the resolution of the scanner used.

b. The radioactivity per shell or per lung slice (A, in kBq) is calculated from the counts per second×scale factor×calibration factor for the PET scanner used: \documentclass{aastex}\usepackage{amsbsy}\usepackage{amsfonts}\usepackage{amssymb}\usepackage{bm}\usepackage{mathrsfs}\usepackage{pifont}\usepackage{stmaryrd}\usepackage{textcomp}\usepackage{portland, xspace}\usepackage{amsmath, amsxtra}\pagestyle{empty}\DeclareMathSizes{10}{9}{7}{6}\begin{document} \begin{align*}{\rm A_{Lung}} = \mathop{\Sigma\rm A}_{i = 1}^{i= n} {\rm Shells} \ or \ \mathop{\Sigma\rm A}_{i = 1}^{i= s} {\hbox{transaxial slices}}\end{align*} \end{document} \end{document}

where n=number of shells; s=number of slices.

c. Shell volume (V_s) and activity (A_s) data should be normalized to the total lung volume (%V_L) or total lung activity (%A_L) as the deposited radioactivity is likely to vary between subjects and also within the same subject imaged multiple times.

Figure 8 shows an example of the plot of cumulative lung volume calculated from summing the individual shell volumes, and Figure 9 illustrates the relative radioactivity of each slice in each plane for one subject who underwent two PET scans to map the deposition distribution from two different nebulizers. The total activity is the sum of the activity per slice that would be determined from the scanner data imported into a spreadsheet. Both lungs are evident in the sagittal plane plot. These profile plots can be generated when the distribution of the inhaled tracer throughout the lung is of interest and need not necessarily be part of a standardized analysis protocol.

FIG. 8.

A sample plot of cumulative lung volumes calculated from volumes of shells 1–10 for two different PET scans. The dashed line, indicating 50% cumulative lung volume, shows that this occurs for shells 1–8 for these subjects. Shell 10, representing peripheral lung, comprises approximately 30% of the total lung volume. (Data from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

FIG. 9.

Relative dose of [¹⁸F]-FDG per lung slice (normalized to maximum slice dose for each test nebulizer) shown for each plane. These profile scans are for interest in visualizing the distribution of radioactivity throughout the lung and are not necessarily part of a standard imaging analysis protocol.⁽⁶⁾ (Data from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

d. The activity per unit volume (%A/%V), or activity density, can be calculated using normalized activities and volumes.

A sample spreadsheet showing volume and activity for shells and total lung and generated from our MATLab program is given in Appendix 3. Investigators should design a spreadsheet based on their own needs. Program imaging files can also be imported into a viewing program (e.g., Amide, http://amide.sourceforge.net) to view images in each plane and all slices for a study subject. A panel of coronal slices from an Amide file is shown in Figure 10.

FIG. 10.

Examples of coronal slices from one subject taken from an Amide image file. (Data from study files, Dolovich Aerosol Research Laboratory, McMaster University, St. Joseph's Healthcare, Hamilton, ON.)

e. Outer and inner lung regions can be identified using a two-shell method as described in the SPECT chapter (see Fleming et al., sections 4.5.2 and 4.5.3, this issue). However, if desired, more than two shells can be defined, but comparison with the two-shell model should be done. In the example shown in Figure 8, which plots cumulative volume data for a 10-shell model, shell 10, the most peripheral shell, represents 30% of the total lung volume. The volume of the adjacent shell, shell 9, could be added to shell 10 to increase this volume to what might represent an outer lung region of a two-shell model. Similarly, data from shells positioned closer to the carina could similarly be summed until the value targeted for the inner lung region is obtained.⁽³³⁾ The corresponding radioactivity calculated for the combined shells, expressed as a percentage of the total radioactivity in the lung, would also be summed to provide the outer and inner regional radioactivity for a two-shell model. Ratios of outer/inner regional activity, normalized to total volume and to total activity, can then be calculated.⁽³⁴⁾ However, it should be noted that information about the topical distribution of the inhaled aerosol across the lung is likely to be lost by combining shell volumes and shell radioactivity to achieve a two-shell model of the lung.

4.4. Dynamic imaging analysis

The distribution of an inhaled radiolabeled drug in the lung is obtained by imaging the radioactivity over time in a specified area of the lung field.^(35–37) Data acquired for accumulated activity over time from the selected lung section can be analyzed and plotted for the total lung, lung slices within the defined section, and lung shells. Note that both the total volume and shell volumes calculated from the dynamic scan will be less than for the total lung obtained with static (whole-chest) imaging.

4.5. Measuring radioactivity in the inhaler post inhalation

The time the radioactivity (A₀) was loaded into the inhaler system (t₀), the time the inhalation was taken (t_i), and the time at the end of inhalation (t_f) should be noted. Handling the inhaler after inhalation should occur only after allowing for sufficient decay of the radionuclide. An estimate of the amount of activity released from the inhaler during the inhalation maneuver can be made from the above times and A₀. The amount of radioactivity in the inhaler can be measured using a dose calibrator (Capintec Radioradionuclide Calibrator; Capintec, Inc., Ramsey, NJ), a standard piece of equipment in nuclear medicine. If the inhaler does not fit into the calibrator, it can be imaged on the PET scanner. A rectangular ROI would be drawn around the inhaler and the counts of radioactivity obtained through one of the resident protocols. A CT should be obtained to document the volume. A calibration would need to be done prior to these procedures using a known quantity of [¹⁸F]-FDG or whatever radionuclide is being used as well as a known (container) volume.

5. Summary of Key Recommendations

5.1. Study preparation

• Prior to the study, acquire a PET image of a 3D phantom containing a known amount of the radionuclide being used, to check for PET uniformity and provide quantitative calibration of the PET image.

• A decision should be made prior to the study on the acquisition time to be used for the PET imaging procedure, based on the estimated inhaled dose of radioactivity, the rate of clearance, and absorption of the tracer.

• The inhaler device should be shielded during inhalation.

• On each study day, the following quality controls should be carried out: (i) PET scanner sensitivity, and (ii) PET scanner uniformity. Standard quality control checks should also be applied to any imaging device used to provide supplementary images, e.g., CT.

5.2. Radioaerosol administration

• Inhalation training is usually required prior to aerosol administration.

• Subject should normally be sitting upright during aerosol administration.

5.3. Image acquisition and reconstruction

• The subject should be moved to the scanner and positioned on the scanner bed as soon as possible after radioaerosol administration.

• Each transaxial reconstruction matrix should consist of a minimum of 512×512 pixels.

• Images should be obtained of (a) the chest including all the lung fields and the stomach area, and (b) the head to include all the extrathoracic airways. Scanner programmed to acquire images from 4–5 sections.

• The need for subjects to remain in the same position on the scanner bed for the whole procedure in order to obtain good images should be clearly communicated to them. To optimize comfort, it is recommended that subjects are imaged with their arms by their sides supported in a sling or what means are used for the particular scanner bed.

• PET data reconstruction is performed using either the Ordered Subsets Expectation Maximisation algorithm (OSEM) or filtered back projection (FBP)⁽³⁸⁾. The former method may be preferred because of the highly nonuniform deposition patterns that sometimes occur, which may lead to artifacts with FBP. If OSEM is the technique used, two iterations with eight subsets should be viewed as the default setting, which can be modified for local needs. Both attenuation and scatter correction should be incorporated into the reconstruction protocol for each study. This should be done for each paired PET and CT study; CT data are not transferable to a repeat study on the same subject. As corrections are done on a voxel-by-voxel basis, any change in subject position between studies, however slight, means that the alignment of the PET image and an earlier acquired CT image will not be the same.

• Combined PET/CT is considered to be the optimal technique for PET acquisition, as both imaging data sets are co-registered in the machine software, allowing accurate attenuation correction, definition of the volume image, and the best anatomical detail for use in interpreting the deposition distribution.

5.4. Analysis of whole-lung deposition and extrathoracic deposition

• Corrections to administered dose of radioactivity should be made to allow for radioactive decay between the time the dose is drawn to the time it is inhaled.

• Images should be corrected for attenuation and scatter. These corrections are usually integrated into the PET acquisition software protocol.

• Definition of volumes of interest should ideally be derived from anatomical images.

• Data should be corrected for the partial volume effect either by dilating the volume of interest or by using the geometric transfer matrix technique.

▪ When data are presented as percentages of the dose administered, it is vital to state clearly whether these are % ex-valve dose, % delivered dose, or % of some other quantity. It is recommended that a statement of mass deposition is also made whenever possible, for instance: “the mean lung deposition was X% of the dose of drug placed in the nebulizer reservoir, equivalent to Y μg of drug available at the mouth.”

• Mass balance should be estimated, by acquiring counts from all sites at which radiolabel can be located [e.g., the lungs, the extrathoracic airways, esophagus/trachea, and stomach (to account for activity deposited in the extrathoracic airways and then swallowed), retention on the device, the exhaled air filter, and any other items such as syringes, used to load the delivery system]. The mass balance or dose recovery is a measure of the total radioactivity measured versus the expected total radioactivity, expressed as a percentage. The mean value across the study population should lie within 90% and 110% of the expected value.

5.5. Analysis of regional lung deposition

• For both right and left lungs, the position of the carina, or the hilum, and a volume of interest representing the lung envelope should be defined (segmented) using the anatomical image.

• Each segmented lung volume of interest should then be divided into 3D shell-shaped subvolumes using a radial transform, with the minimum number of shells used being two.

• Outer (O) and inner (I) volumes should be defined from the outer and inner half of the shells, respectively.

• The 3D O/I ratios are calculated from the counts in the outer and inner volumes, respectively, preferably after correction for the partial volume effect. If no correction is used, then the full-width half-maximum resolution of the PET images should be recorded with the results.

• The 3D O/I ratio should be normalized for regional volume by dividing it by the corresponding ratio for an air volume scan or functional ventilation [⁶⁸Ga] Galligas or perfusion image [⁶⁸Ga] MAA to give 3D penetration indices PI_3D. Functional images only approximate to a distribution of air volume and should therefore be considered alternatives only to be used if a true volume image is not available.

• Where possible, curves of total deposition and concentration per shell should be calculated. Curves of concentration per milliliter should be normalized for the voxel volume of the shell.

• Further analysis of deposition pattern per lobe or segment, by generation or using one-dimensional profiles, may also be performed.

5.6. Study reporting

It is recommended that authors should report in their publications the extent to which they have adhered to the above key recommendations when conducting their studies. A checklist of suggested minimum data and information to be presented in study reports or manuscripts is given in Appendix 4.

6. Limitations

Although the advantages of PET for investigations of lung deposition and also functional measurements are many, there are nonetheless a number of limitations to this imaging modality.

PET is a major clinical functional imaging tool, used mainly to investigate oncology patients, and the software protocols residing on the scanner control unit are specific to the needs of these investigations. Some of this software can be applied to measuring lung deposition, but tends to be limited. Hence, a medical physicist with an interest in 3D PET imaging and the lung, plus the ability to write software to manipulate the acquired data, would be a valuable addition to the research team.

Development of specific molecular tracers of pharmaceutical products for inhalation requires specialized radiochemistry, proximity to a cyclotron to produce the short-lived radionuclides, lengthy validation and toxicology studies for the product once developed, sufficient production capacity for the radionuclide to be used, and much paperwork for approvals of the tracer for human use.⁽³⁶⁾

The costs of development are high, but once a product has been approved, the research questions are many and varied. Other issues are the need for implementing PET and CT software for co-registration of data if two separate scanners are used,⁽³⁹⁾ the development of segmentation tools and nonstandard software for handling the data, PET scanner resolution and slice thickness, exponentially greater data outcomes to analyze, and high operating costs. Similar issues need to be considered for SPECT and CT scanners used separately or as a unit.

The standard PET protocol for acquisition of [¹⁸F]-FDG uptake data usually requires anywhere from 10 to 15 min for new models of PET scanners to up to 45–60 min of scanning time for older equipment. In all cases, the patient is supine and motionless on the narrow scan table. Inevitably, there will be patient movement, movement of internal organs, and respiratory motion during scanning, all resulting in alignment errors and giving rise to fusion errors between CT and PET. As both cardiac and respiratory gating are options available with PET, but not yet with all CT scanners, artifacts due to motion can be minimized by asking patients to breath-hold at FRC during the 10–15 sec required to obtain the CT scan of the lung. However, maintaining a breath-hold for even 10 sec could be difficult for those patients with severe lung disease, even with coaching from technical personnel. Misalignment between PET and CT may be adjusted by comparing activity in other organs, such as the kidneys and liver.

Further harmonization of analytical approaches between SPECT and PET is needed so that there is consistency between investigative centers. One area that needs consensus is how to define the shells and how many shells there should be.

7. Future Directions

7.1. Areas for investigation of orally inhaled products using PET imaging

The following remain open areas for investigation using PET in studies of aerosol deposition:

• In vivo assessment of dose of drug per airway generation;

• Measurement of differential absorption of drug from the airways, parenchyma:

– rate of absorption versus airway surface area;

– DPIs where drug is not released from carrier (in radiolabeling the drug, the carrier is not labeled but mixed into the formulation prior to loading the inhaler. Thus, if a portion of the drug is not released from the carrier, oropharyngeal dose would be high and possibly greater than lung dose);

• Monitoring the development of lung inflammation using [¹⁸F]-FDG uptake and clinical measurements;

• Displacement studies—drug trafficking⁽⁴⁰⁾:

– studies to determine if certain drugs are sequestered in the lung;

• Use or develop specific PET radiopharmaceuticals (current respiratory therapies):

– to enable receptor mapping;

– to define outcomes for the pharmaceutical;

– drug absorption from the lung and mucociliary clearance;

– deposition kinetics of an inhaled drug using dynamic imaging;

– effect of particle size on deposition distribution (e.g., males versus females, underlying disease, disease resolution);

– tracers for cell trafficking;

– ⁶⁸Ga-labeled carbon nanoparticles (“Galligas”) for ventilation scanning.

Recently, a number of groups have investigated the potential for ventilation and perfusion lung imaging using PET to replace SPECT. PET, having superior spatial resolution and sensitivity, offers many potentially attractive features for ventilation studies using inhaled agents such as radioaerosols. When used with ⁶⁸Ga (t_½=68 min),^(41,42) which can be produced on-site using a radionuclide generator (⁶⁸Ge-⁶⁸Ga) with a useful life of approximately 12 months, PET/CT can produce respiratory gated (both CT and PET) images that are quantitatively accurate and of high quality. The short half-life permits multiple ventilation studies to be conducted on the same day a number of hours apart. Figure 11 shows inhaled ⁶⁸Ga-labeled carbon nanoparticles (“Galligas”) that has been produced using a commercially available conventional “Technegas” generator (Cyclomedica Pty. Ltd., Sydney, Australia). ⁶⁸Ga-labeled lung perfusion agents (macroaggregated albumin) are also being developed.⁽⁴³⁾

FIG. 11.

PET/CT ventilation scan on a healthy volunteer using [⁶⁸Ga]-labeled “Galligas,” substituting ⁶⁸Ga instead of ^99mTc in a Technegas generator. The total acquisition time for the PET was 10 min. The PET ventilation scan offers some advantages over a SPECT Technegas scan, such as higher spatial resolution and sensitivity. In addition, PET systems are often equipped with the ability to perform respiratory-gated acquisitions, which are less common on SPECT systems, and which further improves spatial resolution by removing blurring due to motion. In the images shown, there is an apparent increase in radioactivity seen on the posterior lung from base to apex, which is attributed to gravity-dependent compression (micro-atelectasis) of the lung when imaged in the supine position. (Image courtesy of D. Bailey, University of Sydney, Sydney, Australia)

• Combined PET with magnetic resonance imaging (MRI) scanners. PET/MRI scanners are beginning to be installed in clinics and research institutes, and these may offer exciting new opportunities in the field of aerosol deposition. As MRI is able to measure dynamic ventilation with hyperpolarized gases such as ³He and ¹²⁹Xe, it may be possible to image changes in ventilation resulting from an inhaled pharmaceutical.

Footnotes

Acknowledgments

Professor Dolovich wishes to thank Dr. Ahmed Benelfassi for his efforts in writing the MATLab code for the PET Analyze program and for summarizing the required steps for its validation and use.

Author Disclosure Statement

The authors declare that no financial conflicts of interest exist.

a

Please also refer to validation of radiolabeling of drug in the radiology section (see Devadason et al., this issue).

b

The Medical Internal Radiation Dose (MIRD) Committee of the Society of Nuclear Medicine⁽⁵¹⁾ produces tables that list the S factor dose constants for trachea, liver, kidneys, bladder wall, etc. These doses need to be calculated using the same formulae as above for the different organs and listed when applying for the various approvals/licences.

References

Jones

. The enabling technologies needed for PET-based molecular imaging to support drug development. Drug Discov Today Technol, 2005; 2:305–309.

Basu

, Alavi

. Unparalleled contribution of ¹⁸F-FDG PET to medicine over 3 decades. J Nucl Med, 2008; 49:17N–21N37N.

Rhodes

, Hughes

. Pulmonary studies using positron emission tomography. Eur Respir J, 1995; 8:1001–1017.

Elsinga

. Radiopharmaceutical chemistry for positron emission tomography. Methods, 2002; 27:208–217.

Elsinga

, Doze

, van Waarde

, Pieterman

, Blanksma

, Willemsen

, Vaalburg

. Imaging of β-adrenoceptors in the human thorax using (S)-[¹¹C]CGP12388 and positron emission tomography. Eur J Pharmacol, 2001; 433:173–176.

Dolovich

, Nahmias

, Coates

. Unleashing the PET: 3D imaging of the lung. Byron

, Dalby

, Farr

. Respiratory Drug Delivery VII, 7th. Serentec Press: Raleigh, NC, 215–230. 2000.

Paans

. The use of radiotracers in drug discovery and development. Drug Discov Today, 2003; 8:734.

Eckelman

, Rohatagi

, Krohn

, Vera

. Are there lessons to be learned from drug development that will accelerate the use of molecular imaging probes in the clinic? Nucl Med Biol, 2005; 32:657–662.

Coleman

, Delbeke

, Guiberteau

, Conti

, Royal

, Weinreb

, Siegel

, Federle

, Townsend

, Berland

. Concurrent PET/CT with an integrated imaging system: intersociety dialogue from the joint working group of the American College of Radiology, the Society of Nuclear Medicine, and the Society of Computed Body Tomography and Magnetic Resonance. J Nucl Med, 2005; 46:1225–1239.

10.

Lambrou

, Groves

, Erlandsson

, Screaton

, Endozo

, Win

, Porter

, Hutton

. The importance of correction for tissue fraction effects in lung PET: preliminary findings. Eur J Nucl Med Mol Imaging, 2011; 38:2238–2246.

11.

Nishimura

. Application of three-dimensional airway algorithms in a clinical study. Proc Am Thorac Soc, 2008; 5:910–914.

12.

Nishino

, Washko

, Hatabu

. Volumetric expiratory HRCT of the lung: clinical applications. Thorac Surg Clin, 2010; 20:121–127viii–ix.

13.

Guenther

, Yoganathan

, Garofalo

, Kawabata

, Strack

, Labiris

, Dolovich

, Chirakal

, Valliant

. Synthesis and in vitro evaluation of ¹⁸F- and ¹⁹F-labeled insulin: a new radiotracer for PET-based molecular imaging studies. J Med Chem, 2006; 49:1466–1474.

14.

Yanai

, Hatazawa

, Ojima

, Sasaki

, Itoh

, Ido

. Deposition and clearance of inhaled 18FDG powder in patients with chronic obstructive pulmonary disease. Eur Respir J, 1998; 11:1342–1348.

15.

Ojima

, Ido

, Takahashi

, Hatazawa

, Ito

, Sasaki

, Yanai

, Aikawa

, Takashima

, Abiko

et al. Preparation of a fine powder of 2-deoxy-2-[¹⁸F]fluoro-D-glucose suitable for inhalation to diagnose lung diseases by means of PET. Ann Nucl Med, 1989; 3:143–147.

16.

Berridge

, Lee

, Heald

. Pulmonary distribution and kinetics of inhaled [¹¹C]triamcinolone acetonide. J Nucl Med, 2000; 41:1603–1611.

17.

Bergstrom

, Cass

, Valind

, Westerberg

, Lundberg

, Gray

, Bye

, Langstrom

. Deposition and disposition of [¹¹C]zanamivir following administration as an intranasal spray. Evaluation with positron emission tomography. Clin Pharmacokinet, 1999; 36,Suppl 1:33–39.

18.

Rahman

, Rhodes

, Constantinou

, Waters

, Aigbirho

, Osman

. Lung deposition of ¹⁸F-fluticasone propionate in normal subjects using positron emission tomography. Am J Respir Crit Care Med, 2000; 161,Part 2:A177.

19.

Scheuch

, Bennett

, Borgström

, Clark

, Dalby

, Dolovich

, Fleming

, Gehr

, Gonda

, O'Callaghan

, Taylor

, Newman

. Deposition, imaging, and clearance: what remains to be done? J Aerosol Med Pulm Drug Deliv, 2010; 23:S39–S57.

20.

Henderson

, Alexander

, Smith

. The Food and Drug Administration and molecular imaging agents: potential challenges and opportunities. J Am Coll Radiol, 2005; 2:833–840.

21.

Dolovich

. In vitro measurements of delivery of medications from MDIs and spacer devices. J Aerosol Med, 1996; 9,Suppl 1:S49–S58.

22.

Vallabhajosula

. Radiation dosimetry, protection. Molecular Imaging. Springer: Berlin, HeidelbergChapter 19339–349. 2009.

23.

Brix

, Lechel

, Glatting

, Ziegler

, Münzing

, Müller

, Beyer

. Radiation exposure of patients undergoing whole-body dual-modality ¹⁸F-FDG PET/CT Examinations. J Nucl Med, 2005; 46:608–613.

24.

Dolovich

. Measuring total and regional lung deposition using inhaled radiotracers. J Aerosol Med, 2001; 14,Suppl 1:S35–S44.

25.

Huda

. Radiation dosimetry in CT: the role of the manufacturer. Imaging Med, 2011; 3:247–259.

26.

, Nahmias

. Single-photon transmission measurements in positron tomography using ¹³⁷Cs. Phys Med Biol, 1995; 40:1255–1266.

27.

Otsu

. A threshold selection method from gray-level histograms. IEEE Trans Syst Man Cybern, 1979; 9:62–66.

28.

Dolovich

. Practical aspects of imaging techniques employed to study aerosol deposition, clearance. Hickey

. Pharmaceutical Inhalation and Aerosol Technology, 2nd. Marcel Dekker: New York, 171–213. 2004.

29.

Dolovich

, Yuki

, Rhem

. Defining 3D regions of interest in PET lung images using a shell-generating segmentation process. Proc Am Thorac Soc, 2006; 3:A705.

30.

Rousset

, Ma

, Evans

. Correction for partial volume effects in PET: principle and validation. J Nucl Med, 1998; 39:904–911.

31.

Wahl

, Asselin

, Nahmias

. Regions of interest in the venous sinuses as input functions for quantitative PET. J Nucl Med, 1999; 40:1666–1675.

32.

Dolovich

, Rhem

, Guo

, Labiris

, Thabane

. Does particle size influence distribution of inhaled therapy in the presence of airways obstruction? A 3D PET study of the lung in cystic fibrosis. Proc Am Thorac Soc, 2005; 2:A377.

33.

Sanchis

, Dolovich

, Chalmers

, Newhouse

. Quantitation of regional aerosol clearance in the normal human lung. J Appl Physiol, 1972; 33:757–762.

34.

Dolovich

, Sanchis

, Rossman

, Newhouse

. Aerosol penetrance: a sensitive index of peripheral airways obstruction. J Appl Physiol, 1976; 40:468–471.

35.

Jones

, Donovan

, Goddard

, McNeil

, Atkinson

, Clark

, White

, Chivers

. Use of ¹⁸FDG-PET to discriminate between infection and rejection in lung transplant recipients. Transplantation, 2004; 77:1462–1464.

36.

Dolovich

. ¹⁸F-Fluorodeoxyglucose positron emission tomographic imaging of pulmonary functions, pathology, and drug delivery. Proc Am Thorac Soc, 2009; 6:477–485.

37.

Chen

, Bedient

, Kozlowski

, Rosenbluth

, Isakow

, Ferkol

, Thomas

, Mintun

, Schuster

, Walter

. [¹⁸F]Fluorodeoxyglucose positron emission tomography for lung antiinflammatory response evaluation. Am J Respir Crit Care Med, 2009; 180:533–539.

38.

Bailey

. 3D acquisition and reconstruction in positron emission tomography. Ann Nucl Med, 1992; 6:123–130.

39.

Cohade

, Wahl

. Applications of positron emission tomography/computed tomography image fusion in clinical positron emission tomography—clinical use, interpretation methods, diagnostic improvements. Semin Nucl Med, 2003; 33:228–237.

40.

Cunningham

, Parker

, Rabiner

, Gee

, Gunn

. PET studies in drug development: methodological considerations. Drug Discov Today Technol, 2005; 2:311–315.

41.

Hofman

, Beauregard

, Barber

, Neels

, Eu

, Hicks

. ⁶⁸Ga PET/CT ventilation–perfusion imaging for pulmonary embolism: a pilot study with comparison to conventional scintigraphy. J Nucl Med, 2011; 52:1513–1519.

42.

Kotzerke

, Andreeff

, Wunderlich

. PET aerosol lung scintigraphy using Galligas. Eur J Nucl Med Mol Imaging, 2010; 37:175–177.

43.

Ament

, Maus

, Reber

, Buchholz

, Bausbacher

, Brochhausen

, Graf

, Miederer

, Schreckenberger

. PET lung ventilation/perfusion imaging using ⁶⁸Ga aerosol (Galligas) and ⁶⁸Ga-labeled macroaggregated albumin. Recent Results Cancer Res, 2012; 194:395–423.

44.

Almeida

, Bendriem

, de Dreuille

, Peltier

, Perrot

, Brulon

. Dosimetry of transmission measurements in nuclear medicine: a study using anthropomorphic phantoms and thermoluminescent dosimeters. Eur J Nucl Med, 1998; 25:1435–1441.

45.

Towson

JEC

. Radiation dosimetry and protection in PET. Valk

, Bailey

, Townsend

, Maisey

. Positron Emission Tomography: Basic Science & Clinical Practice. Springer-Verlag: London, 265–279. 2003.

46.

Bailey

, Roach

, Bailey

, Hewlett

, Keijzers

. Development of a cost-effective modular SPECT/CT scanner. Eur J Nucl Med Mol Imaging, 2007; 34:1415–1426.

47.

Radiation dose to patients from radiopharmaceuticals. A report of Committee 3 adopted by the International Commission on Radiological Protection. Ann ICRP, 1991; 22,3:i–18.

48.

Radiological protection in biomedical research. A report of Committee 3 adopted by the International Commission on Radiological Protection. Ann ICRP, 1991; 22,3:A1–B28.

49.

Eberlein

, Bröer

, Vandevoorde

, Santos

, Bardiès

, Bacher

, Nosske

, Lassmann

. Biokinetics and dosimetry of commonly used radiopharmaceuticals in diagnostic nuclear medicine—a review. Eur J Nucl Med Mol Imaging, 2011; 38:2269–2281.

50.

Bacher

, Thierens

. Accurate dosimetry: an essential step towards good clinical practice in nuclear medicine. [Report] Nucl Med Commun, 2005; 26:581–586.

51.

Loevinger

, Budinger

, Watson

. MIRD primer for absorbed dose calculations. Society of Nuclear Medicine: New York, 1988.

52.

Cohen

, Lee

. A catalog of risks. Health Phys, 1979; 36:707–722.

53.

Hendrick

, Dodd

3rd , Fullerton

, Hendee

, Borgstede

, Larke

. The University of Colorado Radiology Adult Dose-Risk Smartcard. J Am Coll Radiol, 2012; 9:290–292.

54.

Dolovich

, Labiris

. Imaging drug delivery and drug responses in the lung. Proc Am Thorac Soc, 2004; 1:329–337.

Positron Emission Tomography (PET) for Assessing Aerosol Deposition of Orally Inhaled Drug Products

Abstract

Abstract

Introduction

1. Prestudy preparation

1.1. Validation of radiolabeling of drug a

1.2. Documentation

1.3. Shielding

1.4. Calibration of camera

Calibration factor to convert scanner counts to kBq/mL

1.5. Dosimetry and risk calculation

2. Preparations on study day

2.1. Documentation

2.2. Subject preparation

2.3. Dose preparation and measurement of device activity

3. Image acquisition

3.1. Aerosol inhalation prior to imaging

3.2. Image acquisition

3.2.1. Acquisition of volume image and definition of lung boundaries with CT scans

3.2.2. Positioning

3.2.3. Image acquisition protocol for low-dose CT scan

3.2.4. Image acquisition protocol for transmission scan

3.2.5. Image acquisition protocol for HRCT scan

3.2.6. Aerosol administration and acquisition protocol for a dynamic scan following inhalation

4. Image Analysis

4.1. Regional analyses for whole lung and subdivisions

4.1.1. Segmentation of the lung

4.1.2. Defining ROIs within the lung: concentric shells

4.2. Quantifying radioactivity in the defined lung ROIs

4.3. Calculations

4.4. Dynamic imaging analysis

4.5. Measuring radioactivity in the inhaler post inhalation

5. Summary of Key Recommendations

5.1. Study preparation

5.2. Radioaerosol administration

5.3. Image acquisition and reconstruction

5.4. Analysis of whole-lung deposition and extrathoracic deposition

5.5. Analysis of regional lung deposition

5.6. Study reporting

6. Limitations

7. Future Directions

7.1. Areas for investigation of orally inhaled products using PET imaging

Footnotes

Acknowledgments

Author Disclosure Statement

a

b

References

1.1. Validation of radiolabeling of drug^a