Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Use of CT and MicroCT to quantify structures in human lungs McDonough, John Edward 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2012_spring_mcdonough_john.pdf [ 2.72MB ]
Metadata
JSON: 24-1.0072712.json
JSON-LD: 24-1.0072712-ld.json
RDF/XML (Pretty): 24-1.0072712-rdf.xml
RDF/JSON: 24-1.0072712-rdf.json
Turtle: 24-1.0072712-turtle.txt
N-Triples: 24-1.0072712-rdf-ntriples.txt
Original Record: 24-1.0072712-source.json
Full Text
24-1.0072712-fulltext.txt
Citation
24-1.0072712.ris

Full Text

USE OF CT AND MICROCT TO QUANTIFY STRUCTURES IN HUMAN LUNGS by John Edward McDonough B.Sc., Simon Fraser University, 2000 M.Sc., The University of British Columbia, 2006  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies (Experimental Medicine) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) April 2012  © John Edward McDonough, 2012  Abstract MicroCT is a technology that allows for obtaining images with resolution equivalent to histological sections from volumes of tissue that would be prohibitive to examine by classical sectioning techniques. In this dissertation, three studies are reported that make novel use of microCT to quantify structures within lung tissue. In the first study, microCT was used to quantify the number of alveoli within normal human lungs. The data show that alveolar number can vary throughout the lung with more alveoli per volume of lung present in the apex of the lung compared to the base. In the second study, airway dimensions from normal human lungs and lungs from patients with advanced COPD were measured on microCT images and compared to measurements made on the same airways obtained using lower resolution MDCT images. These measurements confirmed that MDCT overestimates measurements of the airway wall and underestimate measurements of the airway lumen. As well, microCT was able to show a correlation of airway wall thickness to the extent of emphysema within the lung but no correlation was found when measurements from MDCT were used. This suggests that microCT is able to detect more subtle changes in the airway wall dimensions that MDCT could not. In the third study, microCT was used to quantify the numbers and lumen area of terminal bronchioles in normal human lungs and lungs from patients with advanced COPD. We found that terminal bronchiole dimensions and number from normal lungs were similar to the classical data. As well, we found a marked reduction in the number of terminal bronchioles and a narrowing of the lumen of these airways in the lungs of patients with COPD. Remarkably, we found these changes occurred in regions of the lung with minimal emphysematous destruction, suggesting that these changes occur before parenchymal destruction takes place. These three studies show that microCT is an important and underutilized imaging technology to examine the lung and its use will likely be expanded in futures studies to examine the pathological changes that occur in the tissue structure of other lung diseases.  ii  Preface A version of chapter 4 has been published. (McDonough JE, Yuan R, Suzuki M, Seyednejad N, Elliott WM, Sanchez PG, Wright AC, Gefter WB, Litzky L, Coxson HO, Pare PD, Sin DD, Pierce RA, Woods JC, McWilliams AM, Mayo JR, Lam SC, Cooper JD, Hogg JC. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med. 2011;365(17):1567-75.) I was responsible for sampling of lung tissue, processing of lung samples, quantifying the microCT images, and wrote the first draft of the manuscript. I was also responsible for scanning lungs for MDCT and quantifying airway number per generation on these images. Ren Yuan was responsible for quantifying MDCT images for airway number. Masaru Suzuki was responsible for quantifying histological sections for airway dimensions. Nazgol Seyednejad was responsible for cutting JB4 sections. All others were involved in the collection of lungs from St. Louis, USA and Philadelphia, USA, or patients from Vancouver, Canada; or were involved with intellectual input for this study.  Ethics approval was obtained for this study under the UBC-PHC Ethics Board certificate number PHC H07-02477, UBC/PHC Research Ethics Board certificate number P00-0110, PHC REB H07-00224, and H05-50062.  iii  Table of Contents Abstract .................................................................................................................................................. ii Preface ................................................................................................................................................... iii Table of Contents .................................................................................................................................. iv List of Tables .......................................................................................................................................... vi List of Figures........................................................................................................................................ vii Acknowledgements ............................................................................................................................... ix Dedication .............................................................................................................................................. x 1  Introduction ..................................................................................................................................... 1 1.1 The anatomy and mechanics of the airways ............................................................................. 3 1.2 Principles and history of computed tomography ...................................................................... 6 1.3 CT examination of the lungs and COPD..................................................................................... 9 1.3.1 CT quantification of emphysema .................................................................................... 10 1.3.2 CT quantification of airways ............................................................................................ 12 1.4 Advances in CT and microCT ................................................................................................... 13 1.4.1 Synchrotron microCT....................................................................................................... 14 1.4.2 Small animal microCT ...................................................................................................... 15 1.4.3 Specimen microCT ........................................................................................................... 16 1.5 Principles of stereology in the lung ......................................................................................... 17  2  MicroCT Counts of Alveoli.............................................................................................................. 25 2.1 Introduction............................................................................................................................. 25 2.2 Methods .................................................................................................................................. 26 2.3 Results ..................................................................................................................................... 29 2.4 Discussion ................................................................................................................................ 30  3  CT and MicroCT Comparison of Airway Dimensions ..................................................................... 38  iv  3.1 Introduction............................................................................................................................. 38 3.2 Methods .................................................................................................................................. 39 3.3 Results ..................................................................................................................................... 42 3.4 Discussion ................................................................................................................................ 43 4  MicroCT Counts of Terminal Bronchioles in COPD ........................................................................ 53 4.1 Introduction............................................................................................................................. 53 4.2 Methods .................................................................................................................................. 54 4.3 Results ..................................................................................................................................... 60 4.4 Discussion ................................................................................................................................ 62  5  Conclusion ...................................................................................................................................... 81  References ............................................................................................................................................ 85  v  List of Tables Table 2.1  Subject demographics for alveolar number study .......................................................... 27  Table 2.2  Lung and airway measurements ..................................................................................... 30  Table 3.1  Subject demographics for airway dimension study ........................................................ 40  Table 3.2  Dimensions for microCT and MDCT airways ................................................................... 42  Table 4.1  Human subjects for airway number study ...................................................................... 55  Table 4.2  Isolated human lungs for airway number study ............................................................. 56  vi  List of Figures Figure 1.1  Electromagnetic spectrum .............................................................................................. 21  Figure 1.2  Illustration of percentile and %LAA on CT histogram ..................................................... 22  Figure 1.3  Illustration of full-width half-max method ..................................................................... 23  Figure 1.4  Illustration of the disector principle ............................................................................... 24  Figure 2.1  Alveolar openings on microCT image ............................................................................. 35  Figure 2.2  Number of alveoli per mm3 compared to lung slice ....................................................... 36  Figure 2.3  Mean linear intercept in relation to lung height ............................................................ 37  Figure 3.1  Airways located on microCT and MDCT images ............................................................. 46  Figure 3.2  Examples of paired airways on microCT and MDCT images ........................................... 47  Figure 3.3  Ray-tracing of airway wall on MDCT images and airway tracing on microCT image ...... 48  Figure 3.4  Square root wall area versus airway internal perimeter ................................................ 49  Figure 3.5  Wall area percent for airways compared to Lm measurements .................................... 50  Figure 3.6  Bland-Altman plot of MDCT-microCT wall area difference vs. microCT wall area ......... 51  Figure 3.7  Bland-Altman plot of MDCT-microCT lumen area difference vs. microCT lumen area .. 52  Figure 4.1  MDCT images analyzed by the Disector method ............................................................ 66  Figure 4.2  Lung tissue samples matched with CT images ................................................................ 67  Figure 4.3  Obtaining a representative sample of lung for microCT ................................................. 68  Figure 4.4  MicroCT Image of Terminal Bronchiole in COPD ............................................................ 70  Figure 4.5  Comparison of microCT to histological images of the same tissue ................................ 71  Figure 4.6  Counting airways by histology and comparison to microCT ........................................... 73  Figure 4.7  Numbers of small airways and airways per generation of branching, according to severity of COPD ............................................................................................................. 74  vii  Figure 4.8  Mean linear intercept and number of terminal bronchioles, according to the emphysematous phenotype of COPD ............................................................................ 76  Figure 4.9  Airway profiles and airway wall thickness, according to the severity of COPD .............. 77  Figure 4.10  Intra and interobserver comparison of airway counts ............................................... 78  Figure 4.11  Comparison of microCT terminal bronchiole number and dimensions to previous reports. ......................................................................................................................... 79  Figure 4.12  A shift to smaller diameter airways in COPD lungs ..................................................... 80  viii  Acknowledgements I offer my enduring gratitude to my supervisors, staff members at JHRC, graduate advisor, my fellow students, and those students that I have supervised, without whom I would not have been able to complete the work for this thesis. I owe particular thanks to my primary supervisor Dr. J.C. Hogg, for sparking my interest in the lung and whose breadth of knowledge continues to astound. Finally, I would like to thank my family for their support during my years of study.  ix  Dedication  Affectionately dedicated to my loving wife.  x  1  Introduction Researchers in the past centuries have worked out in detail the mechanics of breathing by  making physical measurements of the lung and developing models to explain those measurements. Fundamental to the mechanics of breathing is knowledge of the structural anatomy of the lung for which several key studies have been published. While much is known of the normal lung structure, the changes that occur in disease are less known. The history of the mechanics of breathing and of lung anatomy with specific focus on the respiratory disease, Chronic Obstructive Pulmonary Disease (COPD), will be covered in section 1.1.  Before the advent of modern imaging equipment, non-invasive examination of the human body was limited to basic measurements taken external to the body which has limited the understanding of human physiology and the pathological changes that occur with the progression of disease. The discovery of X-rays in 1896 and its application in imaging internal structures of the human body was a major advancement that has expanded the ability to diagnose disease and allowed for the development of minimally invasive treatments.(1) Advancements in the use of X-rays has led to the development of several imaging systems from fluoroscopy to modern high-resolution computed tomography, a technology that allows for the accurate three-dimensional reconstruction of internal organs. Computed tomography (CT) has been proven to be one of the greatest advances in medical imaging and its inventors, A.M. Cormack and G.N. Hounsfield, were awarded the Nobel Prize in 1979. The principles and history of computed tomography will be covered in section 1.2.  Since the first CT images were taken, the subsequent 30 years have resulted in a number of developments that have increased the resolution of these CT images and lowered the radiation dose, allowing more detailed images of anatomical structures to be captured. For researchers studying the  1  lung, modern CT imaging has changed how the progression of pulmonary diseases are assessed.(2) Chronic obstructive pulmonary disease (COPD) is a respiratory disease that is characterized by an inflammatory immune response that causes emphysematous destruction of the parenchymal tissues and increased thickness of the airway wall as the disease progresses.(3, 4) CT is currently the most common means for physicians to assess the extent of the structural damage produced by the disease in patients with COPD, specifically examining the density of the parenchymal tissues for the extent of emphysema. As well, the non-invasive nature of these scans allows for follow-up of patients so changes in the pathology of the disease over time can be assessed. More recent studies have used CT images to examine the bronchiolar tissues with measurements made on airway thickness and caliber as it relates to disease. The CT examination of the lungs and COPD are covered in section 1.3.  Advances in CT imaging have increased the speed and resolution of the images that are captured. The past several years has seen the development of extremely high resolution CT scanners that are used mainly for research purposes. This new generation of CT scanners allow for imaging the progression of disease in small animal models, such as mice and rats, as clinical CT scanners would follow the progression within a human patient. Other scanners are capable of even higher resolution on tissue specimens allowing for three-dimensional image reconstruction of lung structures that were too small to visualize previously. These scanners allow for imaging that is comparable to what is obtainable with brightfield microscopy. The ability to image structures such as the terminal bronchioles and alveolar walls allow for new questions to be asked with regard to the normal anatomical structures within the lung and how these structures change with disease. These advances in CT imaging and microCT are covered in section 1.4 and the use of microCT will be the focus of this dissertation.  Advancements in imaging technology have been matched with advancements in how structures on these images can be quantified. Identifying and counting structures is frequently used on histological  2  tissue sections as research has moved from qualitative descriptions of disease pathology to more quantitative measures. In early quantitative studies, the investigators typically counted the number of objects on a single histological section, such as the number of inflammatory leukocyte profiles, and expressed their number per area on the tissue section. Other studies have quantified tissue structures, such as counting the number of alveolar profiles on single histological sections of lung tissue. These measures make several assumptions on the structure of the object being measured, such as the size or shape of the structure that could bias the results. The development of stereological principles have allowed for unbiased measurements that make no prior assumptions and are more accurate. The use of stereology to measure structures has been extended to the analysis of CT images. The principles of stereology are covered in section 1.5.  1.1  The anatomy and mechanics of the airways  The lung is the largest organ in the body, making up several litres of which only about half a litre is tissue and the rest being air. The lung tissue forms an incredibly intricate structure with a large surface area to allow for the optimal exchange of gases with the blood. To ensure that ventilation is adequately distributed throughout this structure, the airways form a fractal branching structure termed the tracheobronchial tree due to its similarity to the branches of a tree. Weibel gave one of the first descriptions of the airway branching pattern and proposed that it be modeled as a simple symmetrical dichotomous branching structure with one trachea, leading into two bronchi, etc., for the 23 or so generations of airways found in the lung. However, an accurate and complete description of the individual branches that make up this structure has proven to be difficult with few studies having examined the lung in this regard.  To examine the airways, one commonly used method was to make a cast of the airway lumen using a liquid resin compound. Once the resin solidifies, the surrounding tissues are removed leaving behind a 3  model of the airways that can be examined. Several studies have used this technique to quantify the number of airways per generation and their dimensions.(5, 6) While Weibel’s model of a symmetrically dichotomous branching structure is true of the first several branches of the lung, this model fails to account for the non-dichotomous branching and what Horsfield et al.(6) had described as “asymmetrical dichotomous branching” that arises beyond this region. The asymmetrical dichotomous branching refers to a dichotomous branching pattern in which certain airway pathways terminate while others continue their dichotomous branching. This makes descriptions of the number of airways per generation, as is commonly used, problematic as airways can have as few as 14 branches from the trachea to the acinus or as many as 24 branches. This range of airway divisions results in an airway of a particular generation having several possible diameters depending on how close it lies to its terminal airway (see figure 4.7B for an example of the various sized airways per generation using data derived from Weibel’s monograph). Horsfield et al. proposed an alternative method of counting airways in order to correct for the asymmetrical dichotomy. This method counted airways starting at its terminus and followed the branches as it converged towards the trachea. This allowed airways of roughly similar dimensions to be grouped together as opposed to being spread across several generations as occurs when the number of airways per generation is counted starting from the trachea. However, due to the difficulty in identifying terminal airways to start the counts, most studies use the method of starting counts at the trachea despite its limitations.  When Weibel and Gomez(7) examined their lung casts, they found that the diameter of each successive generation of airways was smaller than the preceding generation by a constant factor of 0.79, this number was later revised to 0.85(8). A reduction in airway diameter by 0.79-0.85 would result in a smaller total airway resistance at each successive generation as shown by the Hess-Murray law.(9, 10) From the airway measurements made by Weibel, Malcolm Greene was able to calculate the resistance using the Hagen-Poisseuille equation of resistance in a laminar flow regime for each generation.(11) 4  While the smaller airways have a smaller diameter, their total cross-sectional area is greater than the larger airways and, based on his calculations, he showed that the small peripheral airways contribute very little to the overall airway resistance. This was later confirmed by direct measurements of airway resistance using the retrograde catheter in the lungs of dogs(12) and then humans(13).  The study using the retrograde catheter on human lungs by Hogg et al. showed that small airway resistance in normal human lungs was low and accounted for less than a quarter of the total airways resistance.(13) However, in the lungs of patients with chronic obstructive pulmonary disease (COPD),the resistance in airways smaller than 2 to 3 mm in diameter was shown to increase by 4 to 40 fold. Due to the low normal value of small airway resistance, the peripheral airway resistance can increase substantially without significantly impacting total airways resistance. In an editorial, Jere Mead proposed the concept of the lung’s “quiet zone”(14) wherein he said, “If small airways contribute so little to total airway resistance, substantial changes in their cross-sectional area, such as certainly occur in disease, can be expected to have only a small, perhaps undetectable, influence on total airways resistance, and hence on ventilator function as measured with ordinary tests”.  The ordinary tests that Jere Mead referred to are based on over a century of study spent on the mechanics of breathing and determining lung function. Maximum expiratory flow is determined by the time constant of the lung(15) which is determined by the product of the airway resistance and the elastic recoil of the lung parenchyma that drives expiratory flow. An increase in either airway resistance or lung compliance increases the time constant resulting in an increased time to empty the lungs. The measurement of how much air a person can forcibly expire from their lungs is determined by measuring the forced expiratory volume in 1 second (FEV1)which is one of the most common and important measure of lung function(16) and its ratio to the forced vital capacity (FEV1/FVC).  5  The measurement of FEV1 was used by Fletcher and Peto to conduct their classic study on the progression of COPD.(17, 18) By following several hundred working men in London over a period of six years, Fletcher and Peto were able to measure the decline in lung function in these individuals and found that a subset of 15-20% of smokers had a substantial decline in lung function leading to the diagnosis of COPD. The current criteria for diagnosis and staging the severity of COPD are based on measurements of FEV1 and FEV1/FVC as set by the Global Initiative on Obstructive Lung Disease (GOLD) committee.(4) Normal lungs are considered to have a predicted FEV1 above 80% combined with a ratio of FEV1 over forced vital capacity (FVC) above 0.7 all stages of COPD have an FEV1/FVC ratio below 0.7. The stages of COPD can be divided into mild (GOLD stage 1) COPD with FEV1 equal or above 80% predicted, moderate (GOLD stage 2) COPD with FEV1 from 50%≤FEV1<80% predicted, severe (GOLD stage 3) COPD with FEV1 from 30%≤FEV1<50% predicted, and very severe (GOLD stage 4) COPD with FEV1 below 30% predicted.  1.2  Principles and history of computed tomography Electromagnetic energy is packaged in discrete units of energy called photons created from the  movement of charged particles, primarily the negatively-charged electrons surrounding the positively charged nucleus of atoms. Photons have the characteristics of both a particle and a wave and contain more energy at higher frequencies. This spectrum of frequencies is divided such that specific frequency ranges have different properties and are named as such, e.g. visible light has a narrow range of wavelengths between 400-700 nm and their name is derived from the fact that they are visible to the naked human eye (see figure 1.1 for a more detail on the divisions of the electromagnetic spectrum). One component of this electromagnetic spectrum was only discovered in the late 19th century and was termed X-rays, named due to it being an unknown or “X” form of radiation. These X-rays comprise the  6  wavelengths between 0.01 to 10 nm which are shorter than ultraviolet (UV) waves but longer than gamma rays and possess unique properties that make it particularly useful for medical imaging.  X-ray imaging is based on the principle of transmission, absorption and scattering of photons by tissue which is dependent on the density of the tissues being imaged, in particular the elemental makeup of these materials. Images produced by X-rays have traditionally used radiographic film where the X-rays that are able to transit through the tissue, without being absorbed or scattered, react with the light sensitive chemical embedded in the film to create an image. More recently, digital sensors rather than chemical reactions are used to detect the X-rays.  When an X-ray is absorbed or scattered, several processes can occur, some of which can lead to damage of the tissue. Elastic scattering occurs when the X-ray photon encounters an atom causing the electrons within the atom to oscillate resulting in a wave of photons with the same energies being propagated in all directions. Inelastic scattering occurs when the X-ray photon with sufficient energy encounters an atom causing an electron to escape the atom resulting in a positively charged ion and a free electron. With photon energies above 100 keV, there is enough energy to completely ionize common biological elements, e.g. carbon, oxygen, nitrogen, etc. Compton scattering is another form of inelastic scattering in which an electron escapes from the atom in addition to a lower energy photon being redirected and released from the atom. Absorption and release of the free radical and ion formation are the principle causes of the damage associated with X-ray radiation. Low energy photons are more likely to be absorbed such that in diagnostic imaging systems these photons are absorbed from the source by placement of a thin aluminum sheet over the X-ray source which removes all low energy photons up to 15 keV. More energetic photons are generally equally split between elastic scattering and Compton scattering with free electron and ionized particle formed by Compton scattering being the  7  primary source of damage from X-rays. The photon energy of X-rays that are typically used for medical imaging have energies between 20-150 keV.  CT scans use X-rays to take images from multiple angles around a subject and by combining these images are able to visualize the internal structures within the subject. This technique was developed independently by Allan Cormack and Godfrey Hounsfield which eventually led to their sharing of the 1979 Nobel Prize in Physiology or Medicine.(19, 20) In principle, the idea behind CT imaging is simple; X-ray images captured from multiple angles around a subject are used to reconstruct the internal structure of the subject allowing for differentiation of areas with different densities. In practice, the mathematics for this technique are quite complex and was it not feasible until computers that allow for rapid calculations were developed.  The mathematics used to calculate the voxel densities for each slice can either be based on a simple backscatter projection calculation or more complex fourier transform techniques. Backscatter projection is based on the mathematics of linear algebra in which the scanned section makes up a matrix wherein numbers making up the matrix are unknown but the sum of each column and row are the values outputted by the sensor. Fourier transform techniques use mathematics involved in signal processing to transform the image into a waveform equation that allows for easier manipulation before transforming the processed data into a usable image.(21)  With the introduction of CT imaging, the structure of internal organs can be quantified noninvasively and in vivo in patients to determine the extent of disease. As well multiple scans can be taken of a patient to follow the progression of the disease; however, cumulative radiation dose to the patient must be taken into consideration as multiple CT scans will increase the incidence of cancer.(22) When CT was developed by Hounsfield, he had envisaged its use as a densitometer for tissue. As such, he developed a scale based on the densities of water and air to calibrate the CT scans. This scale is referred 8  to as the Hounsfield Unit (HU) and sets the density of air as -1000 HU and water as 0 HU with the simple equation to calculate density from the CT image as (Voxel HU+1000)/1000=Voxel density (g/mL).  The first CT scanners used a single beam of X-rays that was recorded using a single detector. This system took about five minutes to obtain a single image and several hours to complete a whole scan during which time the patient needs to be immobilized. Advancements made since the first scanners have involved increasing the resolution of the scans while reducing the time required obtaining the scan and the radiation dosage required for adequate imaging. Modern scanners use a fan beam with multiple detectors that rotate around the patient in a spiral pattern. This new system allows for images to be obtained in a fraction of the time that the earlier systems had required.  1.3  CT examination of the lungs and COPD Chronic obstructive pulmonary disease (COPD) is defined by the Global Obstructive Lung Disease  (GOLD) guideline as being “a preventable and treatable disease state characterized by airflow limitation that is not fully reversible.”(23) “The airflow limitation is usually progressive and is associated with an abnormal inflammatory immune response of the lungs to noxious particles or gases, primarily caused by cigarette smoking.”(24) Diagnosis for COPD is by spirometric measures of lung function, specifically airflow limitation after use of a bronchodilator. Studies on the progression of COPD are limited due to the slow decline of respiratory function in these patients and because the diagnosis typically is only made when symptoms become relatively severe. The classic study on the progression of COPD was the study by Fletcher and Peto who had followed a group of working men in London, UK for 8-years.(17, 18)  Abnormal lung function is only one manifestation of COPD and fails to accurately differentiate between the emphysematous parenchymal destruction or airway wall remodelling that is characteristic of the tissue pathology. As a measuring device to determine the density of tissues, CT imaging is ideally  9  suited to the study of emphysema of the lung as this disease is defined by its loss of tissue. Studies on quantifying the pathology of COPD have shown that CT imaging can separate emphysema-predominant and airways-predominant phenotypes of COPD.(25, 26)  1.3.1 CT quantification of emphysema Emphysema of the lung is defined by the enlargement of the airspaces distal to terminal bronchioles due to the destruction of the alveolar wall and is one of the major pathological changes that occur in patients with COPD. Two major phenotypes of emphysema exist, centrilobular emphysema and panlobular emphysema. The centrilobular emphysema phenotype is characterized by destruction of the pulmonary acinus that begins at the respiratory bronchioles near the centre of the acinus and expands from there. This is the major phenotype that is associated with the form of COPD caused by cigarette smoking. Panlobular emphysema is characterized by a more generalized and homogeneous tissue destruction that is present throughout the acinus. This phenotype is normally associated with patients who have a deficiency in the alpha1-antitrypsin enzyme due to a genetic mutation.(27)  In clinical practice, emphysema in a patient is quantified by a trained radiologist using CT images and a qualitative index of the extent of the emphysematous lesions within the lung. CT assessment of emphysema has been shown to correlate strongly with the extent of emphysema assessed on pathological specimens.(28) To determine the extent of emphysema in the lungs, the radiologist grades the emphysematous lesions on a five-point scale (0-4) with 0 being no or very little emphysema present (0-5%), a score of 1 = 5-25% emphysema , a score of 2 = 25-50%, a score of 3 = 50-75%, and a score of 4 = 75-100%. This semi-quantitative assessment of emphysema has been shown to be highly predictive of diffusing capacity but is less accurate at predicting airflow obstruction.(29-31) The issues that arise from using visual quantification are that it requires a trained radiologist and is subject to observer bias.(32, 33) 10  Use of the digital image data obtained in CT scans allows for the development of quantification methods that are independent of the observer and, as such, should be unbiased. The first study to use quantitative analysis of CT images of the lungs for emphysema relied on a percentile method.(34) The percentile method used the modal CT attenuation value and the value of the fifth percentile of the histogram and was shown to correlate with the extent of disease using histological sections. Recent studies have used the fifteenth percentile as this was shown to have the lowest variation.(35, 36)  A separate method uses the percent of lung tissue that is below a specific tissue density threshold expressed in Hounsfield Unit and is referred to as the percent low attenuation area (%LAA). The first study found that the percentage of lung with a density beyond a threshold of -910 HU on 10 mm thick sections from CT showed a good correlation with a pathological assessment of emphysema and with lung function.(32) Other threshold levels have also been studied and have shown good correlation with measurements of emphysema. Gevenois et al. used a lower threshold of -950 HU and found good correlation with parameters of emphysema on thin-slice HRCT images.(37, 38) The general consensus is to use the threshold -950 HU for severe emphysema, -910 HU for moderate emphysema, and -850 HU for mild emphysema, which corresponds to tissue densities of 20.0 mL/g tissue, 11.1 mL/g tissue, and 6.7 mL/g, respectively.(39) An illustration of both percentile and %LAA methods is shown in figure 1.2.  Both the percentile and percent area methods measure the total amount of emphysema present in the lung, however, the distribution of disease and the size of the lesions are not taken into account. Mishima et al. showed that the size and number of emphysematous lesions identified by %LAA have a distribution based on the fractal dimension (df). To measure this value, they showed that a loglog distribution of the cumulative number of clusters to the size of the clusters can be described by the power law using the equation Y=K*X-D with the exponential component “D” describing the slope of the  11  relationship. A smaller D would equate to larger emphysematous lesions present in the lungs and was shown to provide information in addition to what is provided by %LAA and allows for the separation of individuals with normal lungs and those with early emphysematous COPD(40) and has been used to detect emphysema in patients with COPD.(41) However, in another study no correlation was found between fractal dimension and macroscopic or microscopic measures of emphysema on tissue samples.(42)  1.3.2 CT quantification of airways While the extent of emphysema lends itself to quantification by CT, quantifying the extent of airways disease on CT is more difficult. One issue that limits the utility of measuring airway dimensions on CT is the resolution of the images that is possible using current CT scanners. In patients with COPD, the main site of airways obstruction is in airways less than 2 mm in diameter.(13) This places the airways to be measured below the current limit of resolution possible on these scanners. Despite being unable to precisely measure airways of this size, the past decade has produced studies using CT measurements of airways and several studies that have measured airways in patients with COPD. Nakano et al. were the first to study airway dimension in patients with COPD and showed that decreased FEV1 as measured by spirometry correlates with increased airway wall area as well as emphysema as measured by CT.(43) The association between CT measures of airway dimensions and pulmonary function tests has since been shown by other groups.(44-46) As larger airways are more accurately measured on CT than smaller airways, Nakano et al. then compared CT measurements of these larger airways to measurements made on histological sections of small airways. They found that these larger airways can be used as a surrogate measure of small airways disease as their dimensions correlate well with measurements of airways less than 2 mm in diameter.(47)  12  Several methods have been developed to measure airways on CT. Initial studies traced airways on printed images but are not an objective measure of airway dimensions.(48-50) The full-width halfmax (FWHM) method is one of the most commonly used approaches to measure airway wall thickness.(43, 51, 52) The FWHM is based on the Gaussian distribution of HU values along a line placed across airway wall. The wall thickness is the distance between the points along the line that are half of the maximum HU density in the airway wall and is illustrated in figure 1.3. While this method is useful as a measure of airway wall thickness, it has been shown to underestimate the airway lumen and overestimate the airway wall area, this effect is exaggerated in airways that are oriented obliquely to the CT scan due to volume averaging of the lumen and wall area due to the slice thickness of the CT images compared to the airway dimensions.(53) Newer techniques are available which can overcome this problem and can assess airways tangentially.(54)  Several other methods that have been tested to determine airway wall thickness. Computed Tomography Airway Morphometry (CTAM) method developed and validated by King et al.(55) uses a density threshold to determine the boundaries of the airway wall and was used to compare the airway dimensions between asthmatic and normal subjects.(56) Another method is use of the maximumlikelihood estimate to determine the boundaries of the airway wall which was found to be more accurate at estimating thin-walled airways.(57) More complicated algorithms have also been developed that include using phase congruency(58) and measuring of attenuation values within the airway wall.(59) Currently, all quantitative measurements of airway wall dimensions are used only in research; in clinical practice the radiologist gives a subjective estimate of airway wall dimensions.  1.4  Advances in CT and microCT The technology for high-resolution CT has advanced to the point where micrometer and even  nanometer-resolution images are obtainable using specialized scanners. These scanners provide higher 13  spatial resolution relative to clinical CT scanners; however, they do not provide adequate soft tissue contrast such that initial studies were primarily conducted on dense tissues such as bone and teeth.(60) These microCT images can be broadly divided into three groups based on the technologies used, each of which having different applications and costs. These three groups consist of synchrotron-based microCT, small-animal microCT, and specimen microCT. A detailed description of the design principle of the latter two systems and a comprehensive listing of manufacturers that produce these instruments has recently been published.(61)  1.4.1 Synchrotron microCT Synchrotrons are particle accelerators that speed charged particles such as electrons to near the speed of light shaping their paths with powerful magnets. Forcing the particles to change direction using these magnetic fields cause intense bursts of electromagnetic radiation to be released, termed synchrotron radiation. This radiation was initially thought an inconvenience for particle researchers but has since found an application in imaging studies. Newer generation synchrotrons have since been expressly built to harness this radiation for imaging and other purposes. An advantage of this technology is that the intensity of the photons emitted is many times greater than would be produced in a tabletop light source and it is currently the brightest source of artificial X-rays. The high photon flux is combined with a broad spectrum of emitted light covering the entire electromagnetic spectrum allowing for a variety of studies to be possible using this technology. A review of many of the different biological structures and functions imaged by synchrotron has been recently published.(62)  Studies using synchrotron imaging can theoretically have resolution of much less than 1 micron as this limit would be proportional to the wavelength of light. Combined with the high photon flux, these systems make possible rapid X-ray imaging with histological resolution. However as a particle accelerator is required for this imaging modality, few facilities exist capable of conducting studies based 14  on synchrotron microCT. To date only two studies have been published that used synchrotron based microCT to study human lung structure. One involves imaging small lung samples at 4 µm resolution to discern the structure of the human lung acinus.(63) Another study compared lungs from individuals suffering from several forms of lung pathology to normal human lungs using a lower resolution of 12 µm.(64) In both studies the lung samples were scanned ex vivo following fixation of the lung. In vivo scans would not be feasible using this method since the highly intense X-rays would cause permanent injury to the subject. One of the more interesting applications that take advantage of the highlycollimated monochromatic light generated by a synchrotron is phase contrast imaging for edge enhancement that is particularly suited for differentiating the fine lung structure between alveolar walls and airspace.(65)  1.4.2 Small animal microCT MicroCT has also been used in reference to small animal CTs which are similar in design to the clinical CTs used to examine patients but on a smaller scale and used specifically for research purposes. These scanners typically have a higher resolution relative to clinical CT scanners but a lower resolution compared to the two other microCT scanners discussed here as both the X-ray source and detector are fixed onto a ring which rotates around the subject. Based on published studies done on the lung, the resolution of these scans typically ranges from 29-150 microns, depending on the scan time and system used. The lower resolution of these scans excludes analysis of fine lung structure but still allows for general assessment of lung density and volumes to be calculated. Due to the similarity to clinical CT scanners, many of the studies using these small animal microCT scanners have a design similar to human CT studies and use the same measurements, e.g. Hounsfield density thresholds to determine percent emphysema, with the added advantage that specialized experimental models such as transgenic mice  15  can be used. In addition, using this system allows for multiple in vivo measurements to be made on animals for longitudinal data to be collected.  Published studies using this class of scanners that have focused on the lung have been based on several disease models. These include smoking models of emphysema(66-68), pulmonary cancer(69-71), lung inflammation and fibrosis(66, 72-74), and allergic lung disease(75, 76).  1.4.3 Specimen microCT The third group of technologies refers to microCT scanners in which the X-ray source and detector are not-fixed, this allows for higher-resolution images to be captured as the detector and emission source can be focused closer to the sample. In this setup, it is the sample that rotates to allow for the multiple views required to reconstruct the images. Generally this would require the specimen to be processed in a manner prohibitive of live animal imaging, though exceptions do occur(77). In addition, the higher resolution of these images allows for visualization of the fine lung structure of the alveolar wall. However, allowing for adequate contrast between tissue and air the tissue requires either extremely long scan times or a contrast agent to be used to enhance the densities of the tissue; typically a high density silver or osmium compound would be used.  In the few studies that have been published that have used this system to examine lung tissue, one group has looked at a mouse model of emphysema(78) and another has studied endotoxin-induced acute lung injury in rats.(79) A third group examined alveolar structures on formalin fixed pig lungs(80, 81) and there have been two studies of centrilobular emphysema in human lungs(82, 83). These studies have used the high resolution of these microCT scanners to provide detailed images of lung structure with images that look similar to histological sections and a voxel resolution ranging from 3.8 to 19 microns. The added advantage of this system is that multiple images can be captured per tissue sample analogous to the serial sectioning of a tissue sample by microtome. The number of images per tissue 16  sample is limited only by the time available for scanning with several hundreds of images capable of being captured per sample; a dataset which would be prohibitive to be obtained by traditional serial sectioning techniques.  1.5  Principles of stereology in the lung  Determining morphometric measurements of 3-dimensional structures using 2-dimensional sections is the basis of stereology. The relationship between geometry and statistical probability, first determined in 1777(84), allows quantification of objects distributed throughout a large structure to be determined using measurements taken from a much smaller sampling of this structure. The classical stereology studies assumed that the structure to be measured was homogeneous and based their measurements on models of this structure. In 1952, Campbell and Tomkeieff published the first study to use stereology in the lung by developing the concept of using mean linear intercept measurements to determine the surface area of the lung.(85) In the 1960’s, Weibel and colleagues published several seminal studies that advanced the use of stereology in the lung through their detailed morphometric measurements.(5, 7, 86) More recently, advances in stereology have led to the development of unbiased measurements that allow for heterogeneous structures to be measured without prior modelbased assumptions on how the structure is arranged.(87) This is particularly useful for the lung where the alveolar arrangement is complex and difficult to model and airways have an anisotropic branching structure. Due to the importance of using unbiased measurements to ensure that quantification of structures is accurate, these techniques have been adopted as standard practice by the major respiratory journals in North America and Europe.(88)  When measuring the number of objects on a thin tissue section, e.g. a histological section of a larger organ, simply counting the number of profiles per unit area is not accurate as the size and shape of the object of interest can result in biased measurements. For example, a larger object would be more likely 17  to be present on a single histological section than a smaller object. As well, if the object has a convoluted shape, such as a single vessel that snakes throughout the tissue, it may appear and be counted on a section multiple times. The principles of modern stereology were developed to provide a framework in which irregularly shaped 3-dimensional objects can be counted from 2-dimensional profiles without any inherent bias based on the objects size or shape. When examining a 2-dimensional section, we have lost a dimension. As such, 3-D objects become 2-D profiles, 2-D surface areas become 1-D lines, and 1-D lines become single points on the section. To quantify these structures, test probes must be used that compensate for this loss of a dimension such that the total equals 3. A grid of single points (1-D) to count the volume fraction of a 3-D object, a line segment (2-D) to measure the 2-D surface area, or the number of hits on a plane (3-D) to measure a 1-D line. The number of objects however is non-dimensional (0-D) and would require a 3-D probe or a volume of space for objects to be counted. A 3-D probe can be generated by using two sections a known distance apart to encompass a known volume. This 3-D probe is referred to as the disector as it uses two sections to form the volume of interest.  Quantifying the volume or number of objects within a structure requires as a first step an accurate measure of the total volume of the structure to be used as a reference volume. This can be accomplished by using Cavalieri’s principle in which the structure is cut in slices with each slice having a known thickness and the area of each slice subsequently measured. The sum of the volumes (slice area x slice thickness) of all slices equals the volume of the whole structure. For the lung, this principle is applicable to images of the lungs from computed tomography. In determining the total volume of an excised lung, any changes of the tissue from its original conformation due to processing must also be corrected for including tissue shrinkage due to fixation and embedding into paraffin. For measurements to accurately represent the whole structure, the sampling of the tissue must be random. To ensure the randomness of sampling, the tissue is sliced from a random start point and a grid is placed onto each 18  slice. Samples are systematically sampled from a random start to ensure that all structures are accounted for. Orientation of the tissue is also required to be random by using an isotropic uniform random (IUR) sampling of the tissue. An IUR sample involves embedding the tissue in a random orientation such that tissue sections for analysis do not have an inherent orientation which may bias the results.  Delesse’s principle can be applied for determining the volume of an object (or objects) within a larger structure. This principle was initially developed to determine the quantity of minerals present within rock samples but is equally applicable to determining the volume of objects within the lung. To determine the volume of objects from a histological section, a grid of points is placed on the section and the number of points that fall on the object that is to be counted is summed. The fraction of positive points over the total number of points used is the volume fraction for this object. This volume fraction is then multiplied by the reference volume to determine the total volume of the object to be counted.  The disector provides an unbiased sampling technique used to estimate the number of objects within a given volume defined by two sections separated by a known distance. These sections can either be separate histological sections known as a physical disector, or can be obtained by focusing through a tissue sections known as an optical disector. In chapter 4, a variation on the optical disector method was used for the HRCT and microCT analyses of the lung tissue which we describe as a CT disector. The disector principle involves identifying the object on one section and seeing if the object continues onto the second (i.e. look-up) section. If this object is found on both sections it is excluded from the count, otherwise it is counted (see figure 1.3 for an illustration of this principle). From these counts and the 3dimensional volume (defined as the area of the section and the distance between them), a density of objects (number of objects per unit volume) is determined. To calculate the total number of objects within a structure, the object density is multiplied by the reference volume of the whole structure being  19  measured, e.g. the whole lung. More detailed descriptions can be found in Sterio(87), Howard and Reed(89), and Hsia et al.(88).  20  Figure 1.1 Electromagnetic spectrum  Frequency  Wavelength 10 km  100 kHz Radio Waves  1 MHz 10 MHz 100 MHz  Radar Microwaves Infrared  Visible Light Ultraviolet  1 GHz  Gamma-ray  100 m 10 m 1m 100 mm  10 GHz  10 mm  100 GHz  1 mm  1 THz  100 µm  10 THz  10 µm  100 THz  1 µm  1 PHz  100 nm  10 PHz  10 nm  100 PHz  1 nm  1 EHz X-ray  1 km  10 EHz 100 EHz  100 pm 10 pm  1 pm  21  Figure 1.2 Illustration of percentile and %LAA on CT histogram  An illustration of the percentile and %LAA thresholds applied to a lung density histogram based on a CT image. The dashed line shows the threshold for the percentile method set at 15 percent. The Hounsfield Unit which encompasses 15% of pixels with the lowest attenuation values is then recorded. The solid line shows the threshold set at -910 HU where the percent of pixels that fall below this value (%LAA) are counted.  22  Figure 1.3 Illustration of full-width half-max method  The histogram shows the density of pixels across the wall of an airway. The full-width half-max (FWHM) method uses peak attenuation value (fmax) to determine the width of the airway wall. The points along the x-axis that are found at half of the fmax value (0.5*fmax) are then used to define the airway wall boundaries (x1 and x2).  23  Figure 1.4 Illustration of the disector principle  An example of 3 discrete objects (A, B, C) intersecting with a defined reference volume of space. Object A passes through both the upper and lower surface of the volume. Object B passes only through the lower surface defined as the look-down plane and object C only through the upper surface defined as the look-up plane. When comparing two planes for counting an object, object A would be excluded from the count as it is present on both planes. As object B is only on the look-down plane, if this section is the one that you were interested in counting, comparing this section to its look-up plane would show that object B is only present on the one plane and as such would be counted as a positive hit. The inverse is true of object C as this would only be counted when comparing the section where object C is located (look-up plane) with its look-down plane.  24  2  MicroCT Counts of Alveoli The number and size of alveoli and their organization within the acinus has been the focus of several  studies. While it is well known that the pleural pressure gradient within the thorax results in a regional dependency on ventilation, few studies have considered the affects this gradient would have on the development of alveolar structures. We hypothesized that the development of the upright lung would result in a gradient of alveolar density as it relates to lung height. To test this hypothesis, we measured the number of alveoli within lung samples of a known volume in 4 normal adult human lungs using microCT imaging and correlated the alveolar density to lung height. These data show that the alveolar density present in the apex of the lung at TLC is greater than the base of the lung. We hypothesize that this regional variation is a structural adaptation of the lung to optimize ventilatory flow into the lower regions of the lung during normal upright breathing.  2.1  Introduction  The role of the lung is to provide gas exchange between the environment and the blood, allowing for oxygenation of hemoglobin and release of carbon dioxide. Without this vital function, the high energy demands of a large multicellular organism would not be possible. To increase efficiency of gas exchange, a large surface area between the blood and the environment is required. As well, the anatomical structure of the lung must allow for distribution of gas throughout all regions of the lung. This is achieved by the airways of the tracheobronchial tree which conduct air from the mouth through a bifurcating tree down to the terminal bronchioles. Distal to the terminal bronchioles is where gas exchange occurs, from here the gas moves by diffusion through several generations of alveolar ducts and finally ending at the alveolar sacs. The lung acinus is the structural unit of the lung starting at the terminal bronchiole and comprised of its distal respiratory bronchioles and alveolar sacs.  25  Alveolarization of the lung begins in late gestation but the numbers of alveoli are only a fraction of the adult lung at birth as only alveolar ducts are present.(90-92) After birth, the number of alveoli increases for the first 2 years of life through septation of the alveolar ducts which is followed by a period of an increase in both alveolar size and number.(93) Investigators have used numerous techniques including lung casts(94), or histological sections(95) to count the number of alveoli within the normal human lung. The most recent study on alveolar numbers by Ochs et al.(96) used an unbiased stereological approach to arrive at an average of 480 million alveoli within a pair of adult human lungs.  While these studies provide a global assessment of the number of alveoli within a human lung, regional variation in number has not been studied. As well, these previous studies have used single histological sections or pairs of sections that form an incomplete representation of the structure of the lung. In our study, we used structural information derived from microCT images obtained from specific locations within the lung. MicroCT images provide a volumetric image of the tissue sample allowing for all individual alveoli to be counted without bias. The localization of these samples within the lung allows for regional variation of alveolar numbers to be measured.  2.2  Methods  Subject demographics: Donor lungs were collected from the Gift of Life Program. These lungs were not found to have a suitable recipient in time for transplant and were, therefore, donated for research. The donor lungs used in this study consisted of 2 lungs from non-smoking subjects and 2 from subjects with a history of smoking. Donor age, gender, height, weight, and smoking history are shown in table 2.1.  26  Table 2.1  Subject demographics for alveolar number study  Donor1  Donor2  Donor3  Donor4  Average±S.E.  M/F  M  M  M  M  4/0  Age  51  62  59  43  53.8±4.3  Height (m)  1.79  1.79  1.7  1.82  1.78±0.03  Weight (kg)  80  107  61.8  82  82.7±9.3  Pack Years  39  24  Non-smoker  Non-smoker  31.5±7.5 (n=2)  Lung Tissue Processing: Lungs were received on ice (4°C) from the Gift of Life Program in Philadelphia, USA. Lungs were immediately re-inflated using an underwater seal at 30 cmH2O pressure then maintained at 10 cmH2O on the deflation limb of the pressure-volume (PV) curve while frozen solid using liquid nitrogen vapour. This pressure will fully inflate the lung to TLC. Frozen lungs were then shipped to Vancouver, BC, Canada where they were scanned while frozen using a multidetector CT scanner (Siemens Sensation 16, Siemens Medical Solutions, Germany) using a volumetric scanning protocol at 120 kVp and 100 mA. Contiguous images were reconstructed using 1 mm slice thickness and a B60f (high spatial frequency) reconstruction kernel. Lung volume was calculated by summing the MDCT voxels that were lung. Lung mass was calculated by measuring the lung density calculated from the X-ray attenuation values of the lung in Hounsfield units (HU) and multiplying by the lung volume. Following the CT scan, the lungs were cut into 2 cm thick slices in the transverse plane. Frozen core samples were collected throughout the lung using a cork borer with 14 mm diameter. The location of samples taken was noted on photographs of slices before and after sampling with these locations marked on the MDCT scan.  27  MicroCT Sample Processing: Frozen core samples were then processed for microCT imaging by first fixing the samples in a combination of 1% glutaraldehyde in acetone at -80°C for 2 hours followed by 2 hours at -20°C then overnight at 4°C. Samples were then washed with acetone followed by contrast staining of the sample using 1% osmium tetroxide in acetone for 1 hour. This was followed by 3 washes with acetone then by several washes of anhydrous ethanol. Fixed and stained samples in anhydrous ethanol were then dried using the critical-point of liquid CO2 (Tousimis Critical-Point Dryer, Automegasamdri®-915B, Series B., MD, USA). These dried specimens (figure 1D) were scanned using a GE eXplore Locus SP microCT scanner (GE Healthcare, WI, USA) at the University of Pennsylvania. This scanning protocol provided 16.24 µm isotropic voxel resolution and 460-1000 contiguous microCT images per lung tissue sample: peak X-ray tube voltage of 80 kVp and current of 80 µA, 3 second exposure time, 500 views at 0.4° increments (shortscan), 1x1 pixel binning, and average of 4 scans.  MicroCT Image Analysis: Five lungs cores from each of the 4 donors were randomly selected from lung apex to base for analysis. A random number generator was also used to determine the starting slice number of the microCT image stack for placement of the region of interest to measure the number of alveoli. A 4x4 grid was placed on the image and a random number generator was used to select a region of parenchyma free from blood vessels or airways to be measured. Measurements were made on 10 consecutive 16.24 µm thick images using a field of view of 0.1 cm2; this equals a sampling volume of 0.001624 mL or 1.624 mm3. The alveolar openings were visually identified based on the anatomical formation of pockets within the three-dimensional volume examined (figure 2.1). These openings were counted and divided by the volume of tissue examined to determine the number of alveoli per mL lung. An average of 4 fields, evenly distributed throughout the core, was used to calculate the number of alveoli per mL in each core. For calculating total number of alveoli per lung, the volume fraction of parenchymal tissues versus blood vessels and airways was also measured. For calculating the number of alveoli per acinus, the numbers of terminal bronchioles were identified anatomically within the microCT 28  image stack with total number of terminal bronchioles per lung calculated using the total lung volume (see chapter 4 for further details on counting of terminal bronchioles). The number of alveoli per terminal bronchiole was calculated by dividing the total number of alveoli in the lung by the total number of terminal bronchioles in the lung as calculated from a previous study(97) and further described in chapter 4. Acinus volume was calculated by dividing number of alveoli per terminal bronchiole by number of alveoli per mL to equal mL per terminal bronchiole. Acinus diameter was calculated by assuming the acinus is a sphere with volume equal to the calculated acinus volume. The mean linear intercept (Lm) has a direct linear relationship with airspace size of the alveoli and alveolar ducts(98, 99) and was measured from images captured at 20 regularly spaced intervals within the microCT scans of each sample using a previously validated grid of test lines projected onto the image and a custom macro (ImagePro Plus; MediaCybernetics, Silver Spring, MD, USA).  2.3  Results  Total lung volume, lung mass, number of alveoli measured per lung, numbers of terminal bronchioles per lung, and number of alveoli per terminal bronchiole are shown in table 2.2. The percentage of parenchymal lung tissue versus blood vessels and airways for calculating total number of alveoli within the lungs was measured to be 89 ± 2%.  Ranking of lung samples from apex to base of the lung shows a significant correlation between lung height and alveolar density with increased numbers of alveoli per mL lung present in the upper lung regions compared to lower regions. In the upper lung region alveolar density is equal to 31.6 ± 6.7 alveoli per mm3, in the lower lung alveolar density is equal to 21.2 ± 3.2 alveoli per mm3 (figure 2.2). Correlation of alveolar number to lung height is equal to y = -2.5015x + 34.983; R = 0.567, P = 0.0091; where ‘y’ equals number of alveoli and ‘x’ is lung height (range: 1-5; 1 = lung apex, 5 = lung base). Despite an alveolar density gradient that occurs through the lung, there was no change in alveolar size 29  at TLC from lung apex to base when comparing the mean linear intercept values to lung height (figure 2.3).  Table 2.2  Lung and airway measurements  Donor1  Donor2  Donor3  Donor4  Average±S.E.  2,826  2,959  3,227  3,992  3,251±261  323  308  359  339  332±11  24,061  23,679  21,988  27,872  24,400±1,242  # alveoli/lung (x106)  68.0  70.1  71.0  111.3  80.1±10.4  # alveoli/lung pair (x106)  136.0  140.2  142.0  222.6  160.2±20.8  # terminal bronchioles/lung  21,306  12,472  31,343  23,898  22,255±3,893  # alveoli /terminal bronchiole  3,191  5,618  2,264  4,656  3,932±747  Acinus volume (mm3)  133  237  103  167  160±29  Acinus diameter (mm)  6.33  7.68  5.82  6.83  6.66±0.40  CT lung volume (mL) CT lung mass (g) # alveoli/mL  2.4  Discussion  The number of alveoli within the human lung has been the focus of study by multiple groups over the past several decades. Weibel and Gomez published one of the earliest reports in their classic studies on lung structure that looked at a single lung from 5 subjects, 3 males and 2 females with an age range of 8-74 years and reported a number of 296 ± 11 million alveoli within a single lung.(86) Subsequent studies have used similar techniques and found variations on the number of alveoli. Dunnill looked at a single lung of a female age 55 years and reported a number of 286 million alveoli(100); Angus and Thurlbeck looked at 42 lungs from 32 subjects, age 19-85 years, and reported a number of 375 ± 18  30  million alveoli per lung (range: 212-605).(95) These classic studies counted alveoli on two-dimensional histology sections and then used a mathematical model that made assumptions about the shape of the alveoli to determine the final number. These assumptions have since been shown to underestimate the number of ducts within the tissue sections.(101) More recently, a stereological approach was developed to allow for the counting of discrete structure within tissue samples without prior assumptions on the shape and size of these structures.(87) This approach uses two histological sections to define a volume of tissue with the number of alveolar openings within this volume counted. Using this new method on 6 adult human lungs, 2 males and 4 females, age 18-41 years, Ochs et al. reported a number of 240 ± 36 million alveoli per single lung (range: 137-395 million).(96)  The present study used microCT images to count the number of alveoli within the lungs of 4 male subjects, ages 43-62, and showed an average number of 80 ± 10 million alveoli per lung (range: 68-111 million) or 160 million per pair of lungs which is less than previous studies have reported. Several factors may explain the reduced numbers of alveoli being reported in this study. One is that the present study used microCT images with a voxel resolution of 16.24 microns while previous studies used histological sections. The higher resolution of tissue sections may allow for smaller and less obvious alveoli to be counted compared to microCT. Also, a reduction in the number of alveoli has been suggested to occur in humans as shown in a study that found increased mean linear intercept with age.(102) This may be the reason why there are fewer alveoli in the subjects from this study as these subjects were older with an average age of 53.8 compared to 28.5 years in the study by Ochs et al.  Despite the fewer numbers of alveoli measured in these lungs, acinar volumes are similar to what has been previously reported. Acinar volume was calculated by dividing the total number of alveoli per terminal bronchiole by the number of alveoli per mL to equal the number of mL per terminal bronchiole. As each acinus is associated with one terminal bronchiole, the calculated volume is equal to acinar  31  volume and was measured to be 160 ± 29 mm3. Several studies have used serial reconstruction of individual acini to measure volume, one of the earlier studies looked at a single acinus and showed an acinar volume of 182.8 mm3.(103) Another study measured 2 acini and showed acinar volume at 140 mm3 and 104 mm3.(104). A third study measured the volume of 6 acini and found an average of 185 ± 78 mm3 (range: 88.1-306.2 mm3).(105) More recently, investigators used synchrotron radiation microCT imaging to reconstruct the alveolar structure, a technique similar to the present study. Their study looked at 12 samples from a single lung and showed acinar volume of 131.3 ± 29.2 mm3 (range: 92.5171.3 mm3).(63) Finally, one other study used lung casts to measure four acini in two subjects and reported volumes of 8.7 mm3 and 1.3 mm3 for a 26-year old female and 30.9 and 14.2 mm3 for an elderly male.(106) However, these values are significantly below the values reported by all other groups and are likely the results of inadequate infiltration of the casting polymer into the alveolar regions. While the present study did not directly measure acinar volumes, our calculated acinar volume (table 2.2) was well within the consensus range of those values previously reported.  Variation in lung physiology, i.e. blood flow and ventilation, is known to correlate with lung height. This is thought to be due to the effects of gravity on the lung creating a pleural pressure gradient from the lung apex to base.(107) However, some studies have suggested that regional heterogeneity is not entirely dependent on gravity.(108) Despite the fact that the pleural pressure gradient is well described, few studies have attempted to measure regional variation in lung anatomical structures. One study had examined regional differences in the structure of dog lungs and found a small increase in the alveolar surface density in the dorsal versus ventral regions of these lungs.(109) Another study measured alveolar size in upright dog lungs and noted an increase in alveolar size in the lung apex compared to the base.(110) As dogs are naturally in the horizontal position, the normal orientation for the lung would be from dorsal to ventral positions as opposed to humans which are upright and where the lung would be oriented from apex to base. 32  Studies of frozen dogs have shown that due to the pleural pressure gradient, the lung at functional residual capacity (FRC) is equal to 80-90% TLC in the upper lung regions and only 30-40% TLC in the lower lung regions.(111) In humans, it has been shown that at FRC regional lung volume at the apex is 65% of TLC and 45% TLC at the base.(112) At TLC alveoli are a constant size throughout the lung(110) with a decrease in lung volume from TLC due to a decrease in the alveolar size and an increase in the alveolar density, i.e. the same number of alveoli within a smaller lung volume. As such, we can estimate the regional alveolar density at FRC by dividing the calculated alveolar density at TLC by the percent of FRC lung volume over TLC lung volume. This calculation provides values of 48.6 ± 10.3 alveoli per mm3 in the upper lung compared to 47.1 ± 7.1 alveoli per mm3 in the lower lung. Based on these equations, in a normally breathing upright lung at FRC, alveolar density appears to be equal between the upper and lower lung regions.  The increase in numerical density of alveoli (number of alveoli per unit volume) in the lung apex is despite no change in the average size of the alveoli as measured by the mean linear intercept. We hypothesize that this is due to the arrangement of the alveoli around the alveolar ducts such that in the lower regions of the lung, with fewer alveoli per unit volume, the remaining volume would contain an increased number of alveolar ducts. The arrangement of alveoli around the alveolar ducts can change the numerical density of alveoli without change in the tissue density of the lung. However, the number of alveolar ducts was not measured in this study and would be the subject of future research. Theoretically, the arrangement of alveoli around the alveolar ducts would affect the lung tissue mechanics. While animal studies have shown a difference in pressure-volume curves in upper versus lower lobes of monkey lungs(113), this difference was not found in humans(114).  The major limitation of this study is the small sample size of only 4 subjects. However, small sample sizes are found in most studies that have examined alveolar number or acinus size. Acinar size was only  33  estimated in this study using total number of alveoli and total number of terminal bronchioles in the lung and was not measured directly. As well, this study did not examine the branching structure of the alveolar ducts within the acinus that may explain the regional variability in alveolar numbers.  In summary, the novel finding of the present study is that regional variation of alveolar density correlated to lung height, with increased alveolar density found in the lung apex compared to the lung base in fully inflated excised human lungs.  34  Figure 2.1 Alveolar openings on microCT image  A microCT image of normal human lung parenchyma. Several alveolar openings are indicated by red arrows with the individual alveolar openings designated by the blue line. Alveoli were included in the count at the point at which they opened in the series of 10 images that were examined.  35  Figure 2.2 Number of alveoli per mm3 compared to lung slice The number of alveoli per mm3 of lung were averaged from 4 fields of view from each core with 5 cores of lung tissue per lung arranged from apex to base from 4 normal human lungs. No difference in alveolar density was found between the 2 smoker and 2 non-smoker lungs. This data shows a marked reduction in the number of alveoli per mm3 in the base of the lung compared to the apex and this trend was seen in all four lungs examined.  36  Figure 2.3 Mean linear intercept in relation to lung height  The mean linear intercept (Lm) was measured from apex to base of the four normal human lungs as a measure of average airspace size. There was little difference in Lm measurements from the apex to the base of these lungs.  37  3  CT and MicroCT Comparison of Airway Dimensions Chronic obstructive pulmonary disease is characterized by emphysematous destruction of  parenchymal tissues and thickening of the airway wall. Computed tomography is currently the most common tool to assess these changes in patients; the non-invasive nature of these scans allows for longitudinal studies on the pathological changes that occur in these patients. However, the accuracy of airway measurements is limited due to the relatively low resolution of the CT scan. The purpose of this study was to compare measurements of airway dimensions between images provided by the low resolution CT scan of the whole lung and the higher resolution microCT scans of tissue samples from these same lungs. Our findings replicate earlier studies showing an overestimation of airway wall area and an underestimation of airway lumen area on CT images compared to microCT. As well, we found a correlation between wall area % measurements of airways from microCT images and the extent of emphysema but no correlation was not found when measurements were made using multidetector computed tomography (MDCT) images. We conclude that while measurements of airway dimensions are possible using CT, the low resolution of these images limit measurements to larger airways and large changes in airway wall remodelling are required for detection due to the low sensitivity of detection from these scans.  3.1  Introduction  Chronic obstructive pulmonary disease (COPD) is characterized by an increase of resistance in airways less than 2 mm in diameter.(13) It is also known from histological studies that the airways of subjects with COPD are thicker than those that do not have airflow obstruction.(3, 115, 116) However, since histology is invasive, computed tomography (CT) has become a popular technique to measure airway dimensions. Even though CT is relatively non-invasive and allows the measurement of the lung structure in vivo it also has several limitations. CT scans can have an in-plane voxel resolution as small as 38  0.2 mm but using the standard field of view in a clinical setting the resolution is closer to 1 mm. Furthermore, CT scans also have slice thickness which is usually on the order of 1 or 1.25 mm for clinical scans. This can cause errors due to volume averaging when structures have an axial tilt that is not perpendicular to the CT scanner. Airways that have a cross-sectional profile that are not perfectly en face to the CT scan plane will produce a “thickening” of the airway wall and decrease in lumen area as the structure passes through the CT slice.(53)  Due to the difficulty of measuring small airways, studies have looked at measuring wall thickness on larger proximal airways as a surrogate to smaller distal airway measurements.(47) While airway wall thickness in larger airways may be predictive of wall thickening in smaller airways, airflow limitation remains more closely associated with small airway dimensions compared to larger airways.(117) Several studies have also made comparisons between airway measurements obtained by CT and techniques that provide higher resolution; however, few have directly compared the same airways using these two measures. Even fewer studies have looked at airway measurements in airway diseases such as COPD. In this study, we compared airways dimensions of matched airways from explanted human lungs of patients with either normal lungs or those with severe (GOLD stage 4) COPD. Dimensions of the matched airways were measured on images collected by a MDCT scanner at 1 mm voxel resolution and a higher resolution microCT scanner at 16.24 µm resolution.  3.2  Methods  Subject demographics: Airways were measured in seven lungs, two lungs were from subjects who donated their lungs for lung transplantation but no suitable recipient was found in time. Five lungs were explanted from patients who had undergone lung transplantation surgery for severe COPD. Patient demographics are shown in table 3.1.  39  Table 3.1  Subject demographics for airway dimension study  Donor1  Donor2  COPD1 COPD2 COPD3 COPD4 COPD5  M/F  M  M  F  M  F  F  M  Age  59  62  61  62  63  56  59  Height (m)  1.7  1.79  1.62  1.7  1.68  1.61  1.76  Weight (kg)  61.8  107  79  71  48  46  74  Pack Years  Non-Smoker  24  25  50  38  54  30  FEV1 (%pp)  Not Available  Not Available  15  21  12  24  15  FEV1/FVC (%)  Not Available  Not Available  28  22  26  24  35  Lung Tissue Processing: Donor lungs were received on ice (4°C) from the Gift of Life Program in Philadelphia, USA; COPD lungs were collected from surgery. All lungs were immediately re-inflated using an underwater seal at 30 cmH2O pressure then maintained at 10 cmH2O on the deflation limb of the pressure-volume (PV) curve while frozen solid using liquid nitrogen vapour. Frozen lungs were then shipped to Vancouver, BC, Canada where they were scanned while frozen using a MDCT scanner (Siemens Sensation 16, Siemens Medical Solutions, Germany) using a volumetric scanning protocol of a 1 mm slice thickness scan at 120 kVp and 100 mA and reconstructed using a high-spatial frequency kernel (B60f) and a field of view set to the edges of the whole lung. Following CT scan, lungs were cut into 2 cm thick slices in the transverse plane. Frozen core samples were collected throughout the lung using a cork borer (diameter = 14 mm). Locations of samples taken were noted on photographs of frozen lung slices before and after sampling and these locations were marked on the MDCT scan.  MicroCT Sample Processing: Frozen core samples were processed for microCT imaging by first fixing the samples in a combination of 1% glutaraldehyde in acetone at -80°C for 2 hours followed by 2 hours 40  at -20°C then overnight at 4°C. Samples were then washed with acetone followed by contrast staining of the sample with 1% osmium tetroxide in acetone for 1 hour. This is followed by 3 washes with acetone then by several washes of anhydrous ethanol. Fixed and stained samples in anhydrous ethanol are then dried using the critical-point of liquid CO2 (Tousimis Critical-Point Dryer, Automegasamdri®-915B, Series B., MD, USA). These dried specimens were scanned using a GE eXplore Locus SP microCT scanner (GE Healthcare, WI, USA) at the University of Pennsylvania. This scanning protocol provided 16.24 µm isotropic voxel resolution and 460-1000 contiguous microCT images per lung tissue sample: peak X-ray tube voltage of 80 kVp and current of 80 µA, 3 second exposure time, 500 views at 0.4° increments (shortscan), 1x1 pixel binning, and average of 4 scans.  Airway dimension measurements: A total of 199 cores were imaged by microCT, 28 airways from 7 lungs were visible and could be matched to the same airway on MDCT (figure 3.1). Matching of airways was accomplished by marking on the whole lung slices the location that each tissue core was sampled. The images of the whole lung slices were then matched to the MDCT images so that the tissue cores can be localized onto the MDCT images. Airways were selected to maintain a reasonable cross-section of the airway with a long-short diameter ratio of less than 3 with images of airways on microCT matched as closely as possible to the MDCT scan (figure 3.2). On microCT images, the airway lumen and the outer wall of the airway at the junction between the adventitial perimeter and parenchymal structures were manually traced using image analysis software (ImagePro Plus). Airway lumen area, airway wall area, and airway lumen perimeter were measured for each airway. In addition to airway measurements, mean linear intercept (Lm) measurements were made to calculate airspace size and assess the extent of emphysema in the cores. The Lm was calculated by applying a grid of line segments on 20 images evenly space throughout the tissue core. The number of intercepts between the parenchymal tissues and the line segments were counted with Lm calculated by dividing this count with the sum of the line segment lengths. 41  On MDCT images, the same airway dimensions were measured using custom software as previously described.(43, 47) The software projects 64 rays from a central seed placed on the image in the airway lumen with the X-ray attenuation values measured along each line (figure 3.3). The airway wall boundaries are defined by the full-width at half maximum principle with additional manual editing to exclude measurement artifacts (i.e. adjacent blood vessels). The wall area % (WA%) is calculated by dividing the airway wall area by the total airway area (airway wall area + airway lumen area).  3.3  Results  The number of airways and airway dimensions, including airway wall area, airway lumen area, airway lumen perimeter, wall area % and square root wall area are shown in table 3.2 (data are shown as mean ± standard deviation). Airway diameter was calculated using the airway lumen perimeter and the equation relating the diameter of a circle to its perimeter. Measurements of airways diameter obtained from microCT images ranged from 1.1-6.1 mm in diameter with 8/28 (29%) of airways <2 mm in diameter.  Table 3.2  Dimensions for microCT and MDCT airways  MicroCT Total Number of Airways Measured Number of Airways/Subject  MDCT 28  4.0±2.6  Airway Lumen Area (mm2) 4.00±2.51  1.09±0.91  Airway Wall Area (mm2) 4.76±4.17  7.05±2.51  Airway Lumen Perimeter (mm) 8.00±2.92  3.88±1.55  Wall Area % 49.6±18.5 Square Root Wall Area (mm) 2.00±0.88  88.1±5.4 2.62±0.47  42  The correlation of the wall thickness (square root wall area) to airway internal perimeter for MDCT had a steeper slope compared to microCT airways (figure 3.4). This relationship can be defined with the equations (in centimeters) for MDCT: (Square Root Wall Area) = 0.2478 × (Airway Internal Perimeter) + 0.1657 with an R2 = 0.6673 and for microCT: (Square Root Wall Area) = 0.1469 × (Airway Internal Perimeter) + 0.0848 with an R2 = 0.2785.  Measurements of WA% of airways on MDCT and microCT images is compared to the measure of airspace size (i.e. emphysema) as measured by Lm (figure 3.5). These data show that the wall area % of airways measured using MDCT have a smaller range of values (range: 77-94) compared to the same airways measured with microCT (range: 19-87). As well, regression of these measurements show a positive correlation of WA% measured using microCT images to Lm (R2 = 0.19, P = 0.02) but not using MDCT images (R2 = 0.0002, P = 0.95). Bland-Altman plots comparing the airway dimensions measured on MDCT compared to microCT images demonstrate an overestimation of wall area (figure 3.6) with greater overestimation of wall area in smaller airways compared to larger airways. Lumen area of airways was underestimated (figure 3.7) but showed a relatively consistent underestimation of -70% across lumen sizes.  3.4  Discussion  The use of CT imaging to assess airway dimensions has increased over the past decade. To ensure the accuracy of these measurements, several studies have been published to compare measurements made using CT imaging to higher resolution measurements of airway structure using either histological sections or higher-resolution imaging technologies (i.e. OCT or microCT imaging). One of the most comprehensive comparisons between histology and CT measurements measured multiple airways from 22 patients. This study compared the square-root of airway wall area to the internal perimeter measurements in individual subjects and using the slopes of the regression lines made an average value 43  from all 22 subjects.(47) For CT images, this study found an average slope of 0.23 which is similar to our slope of 0.25. Airways measured by histology had a slope of 0.28 which is much greater than our finding of 0.15, however the study by Nakano et al. was designed to measure all airways in cross-section so that not all airways were directly matched between CT and histology. This resulted in smaller airways being measured by histology compared to CT with most airways having an internal perimeter less than 0.5 cm compared to our study for which most airways had an internal perimeter greater than 0.5 cm.  Several studies have looked at airway dimensions on the same airways using CT and histology or other imaging systems. One of these studies measured airway lumen and wall area on excised and inflated sheep lungs, comparing the measurements on thin-slice CT to measurements of the same airways from the cut lung surface.(118) They found that CT imaging overestimated the wall area but had little effect on lumen area measurements. Another study used optical coherence tomography (OCT), an imaging technology that allows for visualizing the airway wall at high resolution using infrared light, to study airway dimensions in patients with varying lung functions.(119) This study found that OCT imaging was more sensitive at detecting differences in airway wall dimension compared to CT and has the same advantage as CT in that it can be used to study living subjects. Studies have also compared similar measurements made on porcine lungs(120) and found a bias in CT measurements towards a greater wall thickness compared to direct measurement made on the airways.  MicroCT is a technology that allows for much higher resolution than MDCT scans with the same advantage as MDCT in that volumetric scans can be rotated using multiplanar reconstruction to view the object from any angle. The combination of high-resolution and multiplanar reconstruction makes microCT an ideal technology to examine airway dimensions, especially in the case of airways that are not oriented parallel to the scanner. Measuring airway dimensions by microCT has been determined to be a feasible use of this technology.(121) As well, measurements of airway dimensions using microCT were  44  shown to be in good correlation with direct measurements of these same airways on tissue specimens.(80) The same study used inflation-fixed porcine lung samples to compare 59 measurements of airway lumen area and 30 measurements of airway wall area between CT and microCT. Their findings showed that CT measurements overestimated the airway wall area and underestimated airway lumen area compared to the same measurements made on microCT; the same results were found in our own data.  The major limitation of this study includes the small sample size of 7 subjects and 28 airways. MicroCT is also limited to only imaging of isolated and processed specimens so in vivo or longitudinal studies cannot be undertaken using this technology. The advent of optical coherence tomography in imaging of airways allows for in vivo and longitudinal studies of small airways but has other limitations associated with that particular technology. This study also examined small airways with an internal diameter of 1.27 mm, this is the lower limit of airways that can be resolved on MDCT and would explain the low sensitivity for MDCT measurements in this study.  To date the present study is the only one that had compared images acquired by CT and microCT on the same airways from human specimens with an airway associated disease such as COPD. The findings of our study appear to replicate the findings from earlier studies; this includes the well reported overestimation of wall area and underestimation of lumen area in airway measurements due to the low resolution of CT images. As well, the low resolution of CT images results in a reduced sensitivity for CT to changes in the wall area percent and, as such, was not able to detect the correlation to the extent of emphysema as shown in figure 3.5. Overall, CT is a useful tool to measure changes in airway dimensions due to the non-invasive nature of the scans which allow for a comparison of airway changes over time. However, due to the limitations of the airway measurements on CT, only large airways can be accurately measured and only large changes in wall area can be detected.  45  Figure 3.1 Airways located on microCT and MDCT images  Example of an airway shown on the MDCT image (left) with a resolution of 1 mm and the same airway shown on the microCT image (right) with a resolution of 16.24 µm.  46  Figure 3.2 Examples of paired airways on microCT and MDCT images  Example of images from two airways using MDCT and microCT. Note that the higher resolution of the microCT images allow for fine tissue structure to be discerned such as the alveolar tissue structure. Also note that in the bottom images, a large vessel (arrow) is present with the blood filling the lumen of the vessel clearly discernible on the microCT image (D). On the matching MDCT image (C), this vessel appears as a solid mass.  47  Figure 3.3  Ray-tracing of airway wall on MDCT images and airway tracing on microCT image  Two methods were used for defining the boundaries of the airway wall on MDCT and microCT images. On MDCT images (left), ray-tracing was used with the full-width half-max (FWHM) method to define the edges of the airway wall. On microCT images (right), manual tracing of the airway was used as the airway wall boundaries were more clearly discernible due to the higher resolution of these images.  48  Figure 3.4 Square root wall area versus airway internal perimeter  The square-root of the wall area was compared to the perimeter of the airway lumen (airway internal perimeter) on both MDCT and MicroCT images. On MDCT, measured airways showed a shift to the top and left demonstrating that these images underestimate the lumen size and overestimate the wall area compared to the microCT images.  49  Figure 3.5 Wall area percent for airways compared to Lm measurements  A positive correlation was found between microCT measurements of WA% compared to the natural log of the mean linear intercept (Lm) (R2 = 0.19 and P-value = 0.02). No correlation was found with measurements made on MDCT images (R2 = 0.0002 and P-value = 0.95).  50  Figure 3.6 Bland-Altman plot of MDCT-microCT wall area difference vs. microCT wall area  Measurements made on MDCT images overestimate the wall area of airways in comparison to measurements made on the same airways using microCT images. This overestimation in airway wall increases for smaller airways compared to larger airways when the difference in measurements are expressed as a percentage of the airway size.  51  Figure 3.7 Bland-Altman plot of MDCT-microCT lumen area difference vs. microCT lumen area  Measurements made on MDCT images underestimate the lumen area of airways in comparison to measurements made on the same airways using microCT images. This underestimation in airway lumen area is relatively consistent at approximately 70% underestimation and appears to be relatively independent of lumen size.  52  4  MicroCT Counts of Terminal Bronchioles in COPD The major site of obstruction in Chronic Obstructive Pulmonary Disease (COPD) is in airways less  than 2 mm in diameter. The purpose of this study was to determine the relationship between small airway obstruction and emphysematous destruction in COPD. For this study, multi-detector row computed tomography (MDCT) scans were used to compare the numbers of 2-2.5 mm airways per lung pair in 78 subjects at different GOLD stages of COPD, in 14 explanted lungs donated by 12 patients treated for COPD by lung transplantation and in 4 donor (control) lungs. MicroCT scans of 175 samples removed from these explanted lungs were used to compare terminal bronchiolar diameter, numbers per lung pair and total cross-sectional area to emphysematous destruction measured by mean linear intercept (Lm). MDCT scans showed that 2-2.5 mm airways per lung pair was lower in GOLD stage 1 (P=0.001) and GOLD stage 3 and 4 COPD (P<0.001) versus controls and this reduction was confirmed in MDCT scans of explanted lungs. The microCT scans of the explanted lung samples confirmed published values of terminal bronchiolar diameter, number per lung pair and total cross-sectional areas in control (donor) lungs; and showed an 81-99.7% reduction in total cross-sectional area and a 72-89% reduction in number of terminal bronchioles per lung pair in very severe (GOLD stage 4) COPD (P<0.001). Comparison of terminal bronchiolar number and dimensions at different levels of emphysematous destruction (i.e. increasing Lm) showed that terminal bronchiolar narrowing and loss may precede emphysematous destruction in COPD (P<0.001). These results show that narrowing and obliteration of small conducting airways prior to the onset of emphysematous destruction can explain the increased peripheral airways resistance reported in COPD.  4.1  Introduction  Direct measurements of the distribution of resistance in the lower respiratory tract have established that small airways less than 2 mm in diameter become the major site of obstruction in Chronic  53  Obstructive Pulmonary Disease (COPD).(13, 122, 123) Since resistance to flow through tubes is inversely related to the reduction in their radius raised to the 4th power and since loss of half of these airways will only double the total peripheral resistance due to their parallel arrangement,(124) the 4 to 40 fold increase in peripheral airway resistance originally reported in COPD(13) is more easily explained by generalized narrowing rather than loss of airways.  Diaz et al.(125) have used HRCT to demonstrate reduced airway numbers in regions of lung undergoing emphysematous destruction in persons with severe COPD. The present study was designed to extend these observations by examining the relationship between small airway dimensions and number to emphysematous destruction in COPD. These measurements were made using the 0.6-1.0 mm spatial resolution of multidetector row computed tomography (MDCT) to measure the number of 2-2.5 mm airways, the 16.24 µm spatial resolution of microCT to measure the number and cross-sectional area of the much smaller terminal bronchioles and histology to measure small airway profiles per unit area and wall thickness.  4.2  Methods  Table 4.1 shows the demographic and functional data for 78 subjects who volunteered for MDCT scans as part of a lung cancer prevention study.(126-128) Table 4.2 provides similar data on 4 persons who donated a lung for transplantation that served as a control when no suitable recipient was identified within the required time frame, plus 4 patients with centrilobular emphysema (CLE) that donated a lung and 8 patients with panlobular emphysema (PLE) that donated 10 lungs when treated by lung transplantation. Written informed consent was obtained from all participants in these studies and from the next of kin of the 4 persons whose donated lungs served as controls.  54  Table 4.1  Human subjects for airway number study  Control  GOLD1  GOLD2  GOLD3/4  20  19  19  20  10:10  11:8  9:10  7:13  Age (years)  58.7±1.1  61.4±1.8  66.0±2.5  64.6±1.4  Height (inch)  67.1±0.8  68.5±1.0  66.1±0.6  67.5±0.8  Weight (lbs)  182.3±8.4  171.7±8.3  174.5±5.4  174.5±8.1  Pack Years  43.3±2.7  45.3±2.4  49.9±5.0  54.6±3.8  FEV1 (%pred)  99.7±2.5  89.2±1.6  63.9±1.8  35.6±2.4  FEV1/FVC  78.6±1.0  65.2±0.9  62.2±1.5  46.2±2.5  4,986±313  5,884±340  5,564±309  6,747±432  846±38  832±44  803±32  788±36  4,192±284  5,099±328  4,810±288  6,008±413  177±10  129±9  136±13  54±9  Number of Patients Gender (Female:Male)  Total Lung Volume (mL) Total Lung Mass (g) Total Gas Volume (mL) Number of Airways Diameter=2-2.5mm  55  Table 4.2  Isolated human lungs for airway number study  Control  CLE  PLE  Number of Patients/Number of Lungs  4/4  4/4  8/10  Gender (Female:Male)  0:4  2:2  3:5  Age (years)  53.8±4.3  60.0±1.6  49.6±3.8  Pack Years  31.5±7.5 n=2  43.0±5.5  17.9±3.2  FEV1 (%pred)  N/A  18.0±2.7  19.0±1.6  FEV1/FVC  N/A  26.8±2.9  32.6±2.3  Total Lung Volume (%pred)  N/A  137.0±3.6  140.1±4.1  3,251±261  3,456±602  3,794±595  Lung Mass (g)  332±11  358±27  394±41  # Terminal Bronchioles/mL  6.9±1.2  0.7±0.1  1.6±0.5  22,300±3,900  2,400±600  6,200±2,100 n=7  0.1446±0.0356  0.004±0.002  0.047±0.012  424.0±48.0  51.8±30.0  210.2±48.0  3,050.3±576.6  7.7±5.1  514.1±181.9 n=7  Lung Volume (mL)  Total # Terminal Bronchioles Average Terminal Bronchioles 2  cross-sectional area (mm ) Minimum Terminal Bronchiole lumen diameter (µm) Total Terminal Bronchiole 2  cross-sectional area (mm )  Experimental design: The smaller purely conducting airways were assessed at two levels of resolution. The number of small airways (diameter: 2-2.5 mm) were measured per lung from thoracic MDCT scans performed on the 78 participants in a lung cancer prevention study selected from different 56  levels of COPD severity, on the 4 control lungs, and the 14 lungs from 12 patients with GOLD stage 4 COPD receiving lung transplants. MicroCT was used to measure Lm, terminal bronchiolar number per mL lung, their diameters and cross-sectional areas in 175 samples of lung tissue removed from these 18 explanted lungs. The total lung volumes measured by MDCT served as the reference volume to compute both total number of terminal bronchioles and total cross-sectional area of terminal bronchioles per lung and were doubled to obtain values per lung pair.  Multidetector Computed Tomography (MDCT): Volumetric MDCT scans were performed at full inspiration on each of the 78 subjects from the lung cancer prevention study. Scans were performed in the volume scan mode of a Siemens Sensation 16 scanner at 120 kVp, 125 mAs, 1.0 mm slice thickness, using B35f and B60f reconstruction filters. These scans were used to compute total lung, gas, and tissue volumes and the Disector method (see chapter 1 for further details on this method) was used to count numbers of visible small airways per mL lung (figure 4.1). Briefly, a reference volume frame was provided by 30 pairs of 1 mm thick images separated by a 2 mm distance evenly spaced between lung apex and base. The measured mean number of airways per mL lung was multiplied by the MDCT measured total lung volume to calculate the total number of 2-2.5 mm diameter airways per lung pair.  The main stem bronchus of each of the 18 isolated lungs included in this study was cannulated(129) and attached to a compressed air source with an underwater seal that allowed the lungs to be gently inflated to 30 cmH2O transpulmonary pressure (PL), deflated to 10 cmH2O on the expiratory limb of their pressure-volume (PV) curve and frozen solid in liquid nitrogen vapour (-130°C). Each specimen was kept frozen in a Styrofoam box containing dry ice while a volumetric MDCT scan was performed using the protocol described for the subjects’ thoracic MDCT scans and then stored at -80°C.  These MDCT scans were used to determine the number of small airways in each generation of branching. Software (ImageJ, NIH Bethesda, MD) was used to systematically follow each pathway from 57  the main stem bronchus to the last visible bifurcation on the MDCT scans of 4 control, 4 CLE, and 4 PLE lungs specimens. The bifurcation at the terminus of the airway defined the division from one generation to the next and the total numbers of airways counted in each generation were summed to determine the number of branches for each generation. In addition, airway location was recorded to keep track of whether the airways branched into the upper or lower lobes of the lung.  MicroCT: The frozen lung specimens were maintained on dry ice (-78.2°C) while cut into 2 cm thick slices in the same plane as the MDCT scan. Samples were removed in clusters using a sharpened 14 mm diameter cylinder to obtain the cores of lung tissue processed for microCT and a 16 mm in diameter cylinder to cut 3 companion cores of tissue adjacent to each sample removed for microCT. All of these samples were stored at -80°C and their position was recorded on the corresponding specimen’s MDCT scan by matching before and after photographs of the slices to the appropriate MDCT slice image (figure 4.2). The representative nature of these samples with respect to the entire lung was established by comparing the MDCT densities of the sampled sites to the frequency distribution of the MDCT densities in the entire lung (figure 4.3).  The cores of tissue processed for microCT (n=175) were fixed at -80°C in a 1% solution of glutaraldehyde in pure acetone (freezing point -94.7°C), warmed to room temperature overnight, washed in acetone, post-fixed in 1% osmium tetroxide in acetone, re-washed in ethanol and criticalpoint dried (Tousimis Autosamdri-815B, MD, USA). The specimens prepared for microCT (figure 4.2D) were scanned using either a GE eXplore Locus SP microCT scanner (GE Healthcare, WI, USA) at the University of Pennsylvania or a Scanco microCT 35 scanner (Scanco Medical AG, Switzerland) at the University of British Columbia. The microCT scanning protocol provided 16.24 µm isotropic voxel resolution and 460-1000 contiguous microCT images per tissue core: peak X-ray tube voltage of 80 kVp  58  and current of 80 µA, 3 second exposure time, 500 views at 0.4° increments (shortscan), 1x1 pixel binning, and average of 4 scans.  MicroCT scans were examined in contiguous sections and terminal bronchioles were identified by following small conducting airways to the point they branched into respiratory bronchioles (figure 4.2E and figure 4.4). The number terminal bronchioles per mL of lung tissue core was recorded and five randomly selected terminal bronchioles from each lung were examined using multiplanar reconstruction software (OsiriX 2.7.5; OsiriX Foundation, Switzerland) to reorient images into the plane of the airway segment and measure their diameter and lumen cross-sectional area at the narrowest point (figure 4.2F). The product of the mean number of terminal bronchioles per mL of each lung measured by microCT and total lung volume measured from the MDCT scan of the same lung provided an estimate of the total number of terminal bronchioles per lung or lung pair. The product of total number of terminal bronchioles per lung and the average cross-sectional area provided the total cross-sectional area of all terminal bronchioles in each lung. Mean linear intercept (Lm) has a direct linear relationship with airspace size(98, 99) and was measured from images captured at 20 regularly spaced intervals within the microCT scans of each sample using a previously validated grid of test lines projected onto the image and a custom macro (ImagePro Plus; MediaCybernetics, Silver Spring, MD, USA).  Histology: Portions of tissue from 74 cores examined by microCT were embedded in JB4 plastic; 4 μm thick sections were cut then stained with toluidine blue. Lm was measured on these histology sections using the same protocol as for microCT (figure 4.5). The Disector method was used to examine a subset (8/74) of the JB4 embedded blocks.(87-89) Sections 4 μm thick and 720 μm apart were used to define a volume frame of 72 μL (0.072mL). The number of bronchioles per mL was counted and compared to the number per mL determined by microCT in that same frame (figure 4.6). Portions of 16 mm diameter companion cores (n=64) cut adjacent to those examined by microCT were vacuum-  59  embedded in solution (50% vol/vol Tissue-Tek O.C.T. (Sakura Finetek USA Inc, Torrance, CA, USA) in PBS; 10% sucrose) at 1°C and immediately refrozen on dry ice. Cryosections cut from these frozen tissue blocks were used to count bronchiolar profiles per unit area and measure bronchiolar diameters and wall thickness as previously described.(3)  Statistics: The numbers of airways from the MDCT study and histological data of patients in the different GOLD categories were compared using Tukey’s method of pairwise comparison. Numbers of airways at each generation measured by CT were compared by the Mann-Whitney U-test. The numbers of terminal bronchioles were compared by Student’s t-test. Data were expressed as mean ± standard error.  4.3  Results  MDCT: Table 4.1 summarizes the demographic, lung function and MDCT scan data concerning total lung, tissue, and gas volumes for all 78 subjects that participated in this part of the study. Figure 4.7A shows that the number of 2-2.5 mm airways per lung pair was reduced in GOLD stage 1 COPD subjects compared to control subjects (P=0.001) and further reduced in the GOLD stages 3 and 4 COPD group (P<0.001). Table 4.2 summarizes the demographic data for the 16 persons that donated 18 lung specimens examined by volumetric MDCT.  Figure 4.7B compares the distribution of airways identified from MDCT reconstructions of the bronchial tree from scans of the control lungs to published data on the distribution of 4.0, 3.0, 2.5, and 2.0 mm diameter airways from airway casts.(5) These results show that this procedure identifies most 2.5 mm diameter airways but few of the 2 mm diameter airways previously reported from casts; therefore we conclude that 2.5 mm to be the minimum size of airways which can be resolved by reconstructing the airways from MDCT using this specific protocol of counting terminal bifurcations. This is with the caveat that airways smaller than 2.5 mm diameter are still visible on MDCT images but any 60  bifurcation at their terminus is below the resolution of the CT image. In contrast to the control lungs, the reconstruction of the bronchial tree of lungs from patients with the PLE phenotype of COPD showed a sharp reduction and leftward shift of the distribution of airways and those with the CLE phenotype showed an even sharper reduction in the number of visible airways. Separating the number of airways that branched into the upper (figure 4.7C) and lower (figure 4.7D) regions of the lung show that the number of airways in the CLE phenotype is reduced in both regions of the lung. As emphysema is more prevalent in the upper lung regions (figure 4.8A), the number of airways appears to be independent of the amount of emphysema in that region of lung. In contrast, the PLE phenotype shows that fewer airways are present in the lower lung regions compared to the upper lungs regions, where emphysema is more prevalent (figure 4.8B).  MicroCT: Table 4.2 shows that control lungs contain 6.9±1.2 terminal bronchioles per mL lung with an average diameter of 424 ± 76 µm and cross-sectional area of 0.1446 ± 0.0356 mm2. It also shows the total number of terminal bronchioles per lung is 22,300 ± 3,900 per lung (44,500 ± 7,800 per lung pair) and total cross-sectional area of terminal bronchioles is 3,050.3 ± 576.6 mm2 per lung (6,101 ± 1,153 mm2 per lung pair). In contrast, lungs from patients with the centrilobular emphysematous phenotype of COPD showed a 99.7% reduction in terminal bronchiolar cross-sectional area per lung pair (P<0.001) and 89% reduction in total number terminal bronchioles per lung (P<0.001). Moreover, explanted lungs from patients with the panlobular emphysema phenotype showed a 83% reduction in total crosssectional area (P< 0.001) and 72% reduction in terminal bronchiolar number (P<0.001).  Figure 4.8A shows that measurements of Lm made at regular intervals from lung apex to base varied little in control lungs and, as expected, increased in the upper regions of lungs affected by centrilobular emphysema as well as in middle and lower regions of lungs affected by panlobular emphysema (figure 4.8B). Figure 4.8C compares the frequency distribution of Lm in control lungs to those observed in both  61  the centrilobular and panlobular phenotypes of COPD. Figure 4.8D shows there was a sharp reduction in number of terminal bronchioles per mL lung in regions of diseased lungs in the centrilobular phenotypes of COPD in regions where Lm remained below the 95% confidence limits of the controls lungs (P< 0.001). Comparison of the numbers of airway profiles per cm2 in control lungs to those in lungs affected by CLE (figure 4.9A) also show a sharp reduction in profiles per cm2 in regions of the diseased lungs where Lm remained below the 95% confidence interval (≤489 µm) for Lm observed in control lungs (P=0.002). Figure 4.9B shows the remaining airways had thickened airway walls in the CLE lungs compared to controls (P<0.001).  Observer Variation: Intra and interobserver comparisons confirmed that the counting techniques used are reproducible. Figure 4.10A shows a strong correlation (R2 = 0.80) between data obtained by the same observer for measurements of airways per lung pair obtained by comparing look-up to look-down sections to measurements in the inverse direction of comparing look-down to look-up sections. Figure 4.10B shows a strong correlation (R2 = 0.76) between observations made by a second observer on 20 of the 78 cases examined by the first observer. Figure 4.10C shows that intraobserver comparison of bifurcation counts on MDCT images also has a strong correlation (R2 = 0.75). Comparison of terminal bronchiole counts on microCT images also showed strong correlation between observers (figure 4.10D) (R2 = 0.79).  4.4  Discussion  The results of the disector analysis of the subjects’ thoracic MDCT scans show that the number of 22.5 mm diameter airways per lung pair was reduced in mild (GOLD stage 1) COPD compared to controls and was further reduced in persons with severe and very severe (GOLD stage 3 and 4) COPD (figure 4.8A). For measuring airways on the MDCT images, a distance of 2 mm between images was used. This is due to stereological principles which suggest a separation between images of 30 percent of the volume  62  of the object to be counted.(89) According to measurements made on airway lengths and diameters based on casts, an airway with a diameter of 2 mm would have a length of approximately 6 mm, as such a distance of 2 mm was chosen.(6) Although comparison of the number of airways present at each generation of branching observed on the MDCT scans of the isolated control lungs (figure 4.7B) failed to identify all of the 2 mm airways reported from airways casts of normal lungs,(5) the comparison of these control results to diseased lungs showed fewer 2-2.5 mm airways in both centrilobular and panlobular emphysematous phenotypes of COPD. However, it is not possible to determine if the reduction of 2-2.5 mm airways observed by either method of MDCT analysis was a true reduction in number or simply narrowing to a point where the airways were no longer visible with the approximately 1 mm spatial resolution of the MDCT scan.  In contrast to the MDCT scans, the 16.24 μm spatial resolution of the microCT scans provided mean terminal bronchioles diameters (424 ± 76 µm) and cross-sectional areas (0.1446 ± 0.0356 mm2) in the 4 control lungs that are consistent with published normal values (figure 4.11). Furthermore, comparison of the 4 control lungs to 14 lungs donated by patients with very severe (GOLD stage 4) COPD clearly showed that the total number of terminal bronchioles and total cross-sectional area were substantially reduced in both the centrilobular and panlobular emphysematous phenotypes of COPD. Comparison of microCT measurements of the number of terminal bronchioles per mL lung to the alveolar dimensions (Lm) measured within the same lung samples (figure 4.8D) showed that narrowing and loss of terminal bronchioles occurred in regions of lung without the appearance of microscopic emphysematous destruction in the centrilobular emphysematous phenotype of COPD. Although a similar trend was present in the panlobular emphysematous phenotype of COPD, this trend did not become statistically significant until microscopic emphysema became apparent (figure 4.7C/D and 4.8D). These results extend Leopold and Gough’s(130) classical description of centrilobular emphysema that suggested these  63  lesions start in terminal bronchioles by showing that the terminal bronchioles are narrowed and destroyed prior to the onset of emphysematous destruction.  The substantial reduction in airway numbers would result in a marked increase in the airways resistance. Based on the equation of how resistance adds in parallel branches, the approximately 90% reduction in airway number would result in an increase in airways resistance by 10-fold which is well within the range of 4-40-fold increase in airways resistance as occurs in COPD. In addition, narrowing of the lumen of the remaining airways would substantially increase resistance greater than the 40-fold increase that was previously reported. It has been also been reported that collateral channel resistance is much lower in the lungs of subjects with COPD compared to normal lungs.(131) We believe that increased ventilation through collateral channels could occur in the lungs of COPD patients to compensate for the substantial loss and narrowing of the terminal bronchioles that lead to increased airways resistance.  Bignon et al.(132) were the first to measure small airway narrowing in COPD by showing that the number of bronchiolar profiles with lumen diameters <400 µm increased in post-mortem lungs from patients who died in respiratory failure. Although Matsuba and Thurlbeck(133) confirmed Bignon et al.’s finding (figure 4.12), they concluded that this shift was too small to account for the 4-40 fold increase in peripheral resistance reported in COPD. However, they added the caveat that disappearance of large numbers of the smallest bronchioles might have buffered the downward shift of mean bronchiolar diameter they observed in diseased lungs.  The data concerning profiles per unit area in this study (figure 4.9A) compare favourably to those measured by both Bignon et al. and Matsuba and Thurlbeck (figure 4.12). However, it is impossible to compute total number of terminal bronchioles per unit lung volume from profiles per unit area without following the principles of stereology and defining a volume in which the profiles are measured. In 64  contrast, microCT allows terminal bronchioles to be precisely identified (see figures 4.2E and 4.4) and counted in known volumes of lung tissue. Moreover, the multiplanar reconstruction software used to analyze these microCT images allows individual terminal bronchioles to be examined in multiple planes to accurately measure their lumen cross-sectional area at its narrowest point (figure 4.2F). The product of the mean value of terminal bronchioles per mL lung computed by microCT and total lung volume computed by MDCT provide estimates of total number of terminal bronchioles per lung.  Major limitations of microCT are that the radiation dose required can only be safely applied to isolated lung specimens, the scanner is unable to accommodate large bodies, and the costs limit its application to a relatively small number of specimens. The sampling design used in this study made it impossible to address the question as to whether airways 2-2.5 mm in diameter actually disappear or simply narrow to a point where they were no longer visible on MDCT scans. Despite these limitations, the microCT results extend earlier reports by showing that there is both widespread narrowing and loss of smaller conducting airways prior to the onset of emphysematous destruction in the centrilobular emphysema phenotypes of COPD that readily explain the 4-40 fold increase in small airway resistance that has been measured in COPD.  65  Figure 4.1 MDCT images analyzed by the Disector method  The airway of interest (arrow) show 2 airways counted in the look-up plane become one airway in the look-down plane. Therefore, counting in both the look-up and look-down directions account for both mother and 2 daughter branches of airways that divide within the reference frame, without violating the principle that only airways present in one plane and not the other are counted when calculating the number of airways within the lung.  66  Figure 4.2 Lung tissue samples matched with CT images  A) A frozen lung slice from a patient with severe centrilobular emphysema. B) The same lung slice after samples were removed. C) The matching slice from the MDCT scan of the intact lung specimen with location of samples indicated (circles). D) A single control lung sample after it had been processed for microCT. E) A 16.24 μm thick section from the microCT scan of a control lung showing a terminal bronchiole. F) The same terminal bronchiole reoriented to show the cross-section of the airway (arrow) at the plane of section indicated by the line in (E).  67  Figure 4.3 Obtaining a representative sample of lung for microCT  Comparison of the frequency distribution of voxel density in Hounsfield units for the entire lung to the frequency distribution of the Hounsfield units measured in the sampled sites of these lungs (see figure 1 of the manuscript for a description of how these sites were located). These data show that for the 4 cases of the CLE phenotype of COPD and 4 control cases, the frequency distribution of all the voxels in the lung and the frequency distribution of the voxels in the sample sites are not different from each other (P>0.05). These data provide direct evidence that the sample sites used for the microCT studies are representative of the entire lung.  68  0.14  0.12  0.12  0.1 0.08 CONTROL1 Lung 0.06  CONTROL1 Core  0.04  Frequency Distribution  Frequency Distribution  0.14  0.02  0.08 CONTROL2 Lung 0.06  CONTROL2 Core  0.04 0.02  0 -1000  0.1  0 -800  -600  -400  -1000  -800  0.14  0.14  0.12  0.12  0.1  0.08 CONTROL3 Lung 0.06  CONTROL3 Core  0.04 0.02  0.1  0.08 CONTROL4 Lung 0.06  CONTROL4 Core  0.04  0 -800  -600  -400  -1000  -800  0.14  0.14  0.12  0.12  0.1 0.08  COPD1 Lung 0.06  COPD1 Core  0.04  0.02  0.1 0.08 COPD2 Lung 0.06  COPD2 Core  0.04  0 -800  -600  -400  -1000  -800  Hounsfield Unit  0.14  0.14  0.12  0.12  0.1 0.08 COPD3 Lung  0.06  -600  -400  Hounsfield Unit  COPD3 Core  0.04 0.02  Frequency Distribution  Frequency Distribution  -400  0.02  0  0.1  0.08 COPD4 Lung  0.06  COPD4 Core 0.04 0.02  0 -1000  -600  Hounsfield Unit  Frequency Distribution  Frequency Distribution  Hounsfield Unit  -1000  -400  0.02  0 -1000  -600 Hounsfield Unit  Frequency Distribution  Frequency Distribution  Hounsfield Unit  0 -800  -600  Hounsfield Unit  -400  -1000  -800  -600  -400  Hounsfield Unit  69  Figure 4.4 MicroCT Image of Terminal Bronchiole in COPD  MicroCT images of serial cross-sectional cuts of a single terminal bronchiole from a patient with COPD. Note that the lumen progressively narrows in serial sections 1-3 and then progressively increases in size in sections 4 and 5 until it opens into the respiratory bronchiole in the section 6.  70  Figure 4.5 Comparison of microCT to histological images of the same tissue  A) (clockwise from top-left) Shows a cut surface of the tissue core that had been scanned for microCT before it was cut. A reconstruction from the microCT scan at the same level the cut was made. A single slice from the microCT image stack showing the tissue surface. A histological section cut at the same tissue level after the core was embedded in JB4 showing the tissue surface. Note the flap of tissue is seen to contain a vessel on the thick sections of tissue (arrow) and that this 3-dimensional structure becomes a 2-dimensional line on the single microCT image and histological section. B) A high degree of correlation was found for Lm measurements made on histological sections of JB4 embedded samples and microCT sections. Comparison of the Lm measurement showed R2 = 0.8323 and slope of 0.9469 when intercept was set at y = 0.  71  72  Figure 4.6 Counting airways by histology and comparison to microCT  Comparison of the number of terminal bronchioles per mL lung obtained by microCT examination of the tissue core to the number obtained using a physical disector from histological sections cut from JB4 embedded tissue from the same tissue. The physical disector method examined the volume of tissue between two sections cut 0.72 mm apart, with cross-sectional areas of 1 cm2 and a total disector volume of 0.072 mL. The average number of airways present on the look-up and look-down sections were counted and compared to matched microCT images using the same criteria for both datasets (physical disector method). In practice, this comparison was restricted to control subjects as there were too few airways in the diseased lung tissue to count using histology based methods. Although there was a trend for the histology based method to count fewer airways per mL, the difference between the two methods was not significant (P = 0.28).  8  Number of Airways/mL  7 6 5 4 3 2 1  0 Histology  MicroCT  73  Figure 4.7 Numbers of small airways and airways per generation of branching, according to severity of COPD  A) The number of airways 2-2.5 mm in diameter per lung pair obtained using a CT Disector method to analyze the MDCT scans of 78 subjects at different stages of COPD severity. Compared to the control group, GOLD stage 1 (P=0.001), GOLD stage 2 (P=0.018), and GOLD stages 3 and 4 (P<0.001) show a reduced number of small airways per lung pair. B) Data obtained by reconstructing the bronchial tree from MDCT images of explanted lung specimens. The height of the columns represents the number of airways in each generation coloured according to size as previously reported from lung casts.(5) Number of airways per generation obtained in this study shows control lungs (blue line) closely match the distribution of airways down to and including those 2.5 mm in diameter. In contrast, CLE lungs (red line) show that the number of airways at each generation is lower than predicted and airways ≤2.5 mm in diameter are largely missing or narrowed to the point they are below the resolution of the MDCT scan with PLE lungs (yellow line) falling between the CLE cases and controls. Comparison of the control lungs to lungs affected by PLE shows no difference in the (C) upper lung region and a sharp reduction in the (D) lower lung regions where PLE causes greater emphysematous destruction (see figure 4.4B). In contrast, comparison of the control lungs to the lungs affected by CLE show a sharp reduction in both (C) upper and (D) lower regions of the lung even though CLE causes greater destruction of the upper lung region (see figure 4.4A). These HRCT data are consistent with the microCT data (figure 4.4D) showing a much greater reduction in terminal bronchioles in non-emphysematous regions (i.e. ≤489 µm) in lungs from patients with CLE than lungs affected by PLE.  74  75  Figure 4.8 Mean linear intercept and number of terminal bronchioles, according to the emphysematous phenotype of COPD  MicroCT Measurements of Lm show the expected distribution of emphysema from lung apex to base in (A) centrilobular (CLE) and (B) panlobular (PLE) phenotype of COPD, with no change as a function of lung height in the 4 control lungs. C) The frequency distribution of Lm measurements in the 4 control lungs compared to the frequency distribution of Lm measured in the 4 lungs affected by centrilobular and 10 lungs affected by panlobular phenotypes of COPD. D) The regions of the diseased lungs in which Lm remained below the upper 95% confidence limit for the control lungs (489 μm) had reduced number of terminal bronchioles per mL lung tissue in the CLE group (P<0.001) and remains low in samples with Lm of 489-1000 μm and >1000 μm.  76  Figure 4.9 Airway profiles and airway wall thickness, according to the severity of COPD Compares the number of small airway profiles per cm2 (A) and thickness of their walls (B) measured from histological sections cut from samples of tissue adjacent to those examined by microCT. Note that similar to the number of terminal bronchioles per mL lung measured using microCT (figure 4.8D) the number of small airway profiles per unit area is sharply reduced in regions of diseased lungs where Lm remains below the 95% confidence limit (≤489 µm) observed in control lungs and that the surviving airways have thickened walls  77  Figure 4.10 Intra and interobserver comparison of airway counts  A) Intraobserver measurements for airway counts on the 78 subjects using MDCT. B) Interobserver measurements for airway counts on a subset (20) of the 78 subjects using MDCT. C) Shows a relatively strong correlation (R2 = 0.75) between the number of airways counted by two separate observers that followed the airways from the bronchus to their terminus and counting the number of airway bifurcations per generation from MDCT images of isolated lungs. D) Shows a similarly strong correlation (R2 = 0.79) between 2 observers that counted the number of terminal bronchioles per lung tissue core from the microCT scans.  78  Figure 4.11 Comparison of microCT terminal bronchiole number and dimensions to previous reports.  Shows that the present data on terminal bronchiole diameter (A), and cross-sectional area (B) obtained from microCT images of our control subjects closely match those previously reported by Weibel.(5, 134) Moreover, the numbers of terminal bronchioles per lung pair obtained in this study (44,510 ± 15,574; Mean±S.D.) closely matches the mean of four previous studies from casts (44,500 ± 18,574) of normal human lungs(5, 6, 135, 136) summarized in table 3 of the research paper by Horsfield and Cumming.(6)  79  Figure 4.12 A shift to smaller diameter airways in COPD lungs  Comparison of data from Matsuba and Thurlbeck(133) and Bignon et al.(132) to the present results shows that all three studies reported a left-shift in the small airway diameters in diseased lungs compared to controls. However, the results reported here (table 4.2) extend this finding by showing an 81-99.7% reduction in total lumen area and 72-89% reduction in number of terminal bronchioles that readily explain the 4-40 fold increase in small airway resistance measured in COPD.  80  5  Conclusion MicroCT imaging technology was initially developed for research on bone structure and is able to  provide visualization of lung tissue within a 3-dimensional volume with unprecedented detail allowing for novel discoveries to be made on the structure of the normal lung and the pathological changes in structure that occur in disease. In chapter 2, microCT images of normal lung tissue were used to determine the alveolar density of the lung by counting the number of alveoli per volume of lung at TLC. The novel finding of this research was that even though transpulmonary pressure was the same in all regions of the lung and there was no difference in airspace dimensions, there was a greater alveolar density (i.e. number of alveoli per unit volume) at the apex of the lung compared to the base. Based on estimations of regional differences in lung expansion, when a pleural pressure gradient is imposed at FRC, this difference in alveolar density may be abrogated by the pleural pressure gradient. The assumption is generally made that lung parenchymal anatomy is homogenous due to lung density being constant throughout the lung. Studies comparing the upper and lower lobes of both animal and human lungs have shown that the mechanical properties between these lobes differ(114), however, other studies have also shown that this may not be the case in humans.(137) In addition to these studies on lung mechanics, studies have looked at the effects of gravity on lung volume changes due to the pleural pressure gradient and have found that different orientations of the lung relative to the direction of gravity have different effects on the changes in lung volume. What these studies illustrate is that the lung is not homogenous with differences found between upper and lower regions of the lung. That we have found a gradient in alveolar density between upper and lower regions of the lung may explain these mechanical differences and why the lung does not have the same changes in lung volume when reoriented to the direction of gravity.  81  In chapter 3, microCT measurements of airway dimensions were compared to matched airway measurements made on MDCT images of the whole lung. MicroCT shares the same advantages of MDCT in that a volumetric scan of tissue can be manipulated to visualize the structure from any orientation with the added advantage of having higher resolution increasing the sensitivity to detect small changes in the airway wall compared to MDCT. This study confirmed the results that have been shown in previous studies. As well, we found a correlation between wall area % of airways measured on microCT when compared to Lm measurements with increased wall area % found in airways in more emphysematous regions of the lung. No correlation was found with measurements made using MDCT images suggesting that these images lack the resolution and sensitivity to detect changes in the airway wall.  In chapter 4, a combination of MDCT and microCT was used to count the number of airways in the lungs of normal subjects and those with COPD. Using MDCT, the numbers of larger airways measuring 22.5 mm in diameter were found to be reduced as COPD severity increases. More importantly, the microCT images showed that the terminal bronchiole in patients with advanced COPD were markedly reduced in number with severe narrowing of the lumen of those that remain. These changes were noted in relatively normal regions of the lung that were not affected by the emphysematous destruction. Based on the findings we postulate that the peripheral lung lesions in COPD begin in the terminal bronchioles and then spread into the surrounding alveolar tissue. The techniques used to quantify airway number and lumen caliber can also be used to quantify the airways in other obstructive and restrictive airflow diseases as occurs in the lungs of patients with transplanted lungs(138) or during the course of developing lung fibrosis(139).  One of the many questions that arise from this study is why terminal bronchioles appear to be especially prone to these pathological changes and what mechanisms are involved. The transition from  82  purely conducting airways to gas exchanging tissue occurs as the terminal bronchioles branch into respiratory bronchioles that lead to alveolar ducts and sacs. It is at the terminal bronchiole where the bronchopulmonary anastomoses connect the bronchiole circulation to the pulmonary circulation.(140) As well, the deep lymphatic vessels of the lung originate at the level of the terminal bronchioles and help to drain fluid from the alveolar and interstitial spaces. In addition to draining of fluid, the lymphatics are also involved in immune cell transit for presentation of antigen to lymph nodes that lead to an adaptive immune response. Early studies have also shown that tubercles or granulomas formed from infection by Mycobacterium tuberculosis are most commonly present near the terminal bronchioles.(140) These granulomas are composed of invading macrophages that are attempting to prevent the spread of the tuberculosis bacterium. In COPD, macrophages have been shown to be strongly associated with the presence of emphysema in the lungs.(141) As the terminal bronchioles form a junction for the bronchiole circulation, pulmonary circulation, and lymphatic systems, it is an ideal site for macrophages to transit from the vasculature into the airspace or from the airspace into the vasculature or lymphatics. This transit of macrophages at the terminal bronchiole may result in the narrowing and eventual destruction that was found to occur in the terminal bronchioles of patients with very severe COPD. In COPD, an inflammatory autoimmune response has also been shown to occur in several studies(142-144) that may also trigger an inflammatory response that specifically damages the terminal bronchioles. Whether the vascular and airway structures present in this region are especially vulnerable to injury from the transit of macrophages, an autoimmune response, or for other unknown reasons should be the focus of further research  In summary, microCT is a useful tool for visualizing lung tissue structure and measuring features of both bronchiolar and alveolar tissue. Unfortunately, the current methods of processing the tissue for scanning prevent further use of the tissue for other studies. Glutaraldehyde is a very strong fixative and would impede the binding of antibodies to their epitopes making techniques used for differentiating cell 83  types or proteins by immunohistochemistry difficult if not impossible. As well, the osmium used to enhance contrast would stain most tissue structures dark in colour and make histochemical stains difficult to interpret. Finally, and most damaging to modern research technologies, the treatment of the tissue by these agents would irreversibly damage the RNA and DNA within the sample preventing use of these samples in molecular studies. With advancements in the microCT instrumentation, many of these problems may be overcome. Increased sensitivity of the scanner would allow for imaging of the lung tissue without use of osmium to enhance contrast. Technology is being developed to allow for scanning of frozen tissue directly without use of fixatives or drying of the specimen. One issue that may not be overcome is that use of X-rays can directly damage RNA and DNA molecules. The extent of damage to these molecules during scanning remains unknown and as relatively short strands of these molecules remain usable for quantification, the X-rays may not damage the RNA and DNA to such an extent that would prohibit their analysis.  In conclusion, these studies show that microCT can be a useful tool for visualizing lung anatomy when tissue samples allowed the original architecture of the lung to be maintained. This dissertation presented novel use of applying established morphometric techniques on these microCT images that allowed for the quantification of the number of alveoli within the lung, to quantify the airway dimensions in comparison to CT imaging, and to quantify the number of terminal bronchioles in the lungs of patients with severe COPD. Futures studies will be to focus on the mechanisms involved in our findings that terminal bronchiole number and lumen diameters are reduced in patients with advanced COPD.  84  References 1.  Brecher R, Brecher E. The Rays: A History of Radiology in the United States and Canada. Baltimore: Williams and Wilkins; 1969.  2.  Coxson HO, Mayo J, Lam S, Santyr G, Parraga G, Sin DD. New and current clinical imaging techniques to study chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2009;180(7):588-97.  3.  Hogg JC, Chu F, Utokaparch S, Woods R, Elliott WM, Buzatu L, et al. The nature of small-airway obstruction in chronic obstructive pulmonary disease. N Engl J Med. 2004;350(26):2645-53.  4.  Pauwels RA, Buist AS, Ma P, Jenkins CR, Hurd SS. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: National Heart, Lung, and Blood Institute and World Health Organization Global Initiative for Chronic Obstructive Lung Disease (GOLD): executive summary. Respiratory care. 2001;46(8):798-825.  5.  Weibel ER. Morphometry of the Human Lung. New York: Academic Press Inc; 1963.  6.  Horsfield K, Cumming G. Morphology of the bronchial tree in man. J Appl Physiol. 1968;24(3):373-83.  7.  Weibel ER, Gomez DM. A principle for counting tissue structures on random sections. J Appl Physiol. 1962;17:343-8.  8.  Mauroy B, Filoche M, Weibel ER, Sapoval B. An optimal bronchial tree may be dangerous. Nature. 2004;427(6975):633-6.  9.  Hess WR. Das Prinzip des kleinsten Kraftverbrauchs im Dienste hamodynamischer Forschung. Achiv Anat Physiol. 1914:1-62.  10.  Murray CD. The physiological principle of minimal work. I. The vascular system and the cost of blood. Proc Natl Acad Sci U S A. 1926;12:207-14.  11.  Green M. How big are the bronchioles? St Thomas Hospital Gazette. 1965;63:136-9. 85  12.  Macklem PT, Mead J. Resistance of central and peripheral airways measured by a retrograde catheter. J Appl Physiol. 1967;22(3):395-401.  13.  Hogg JC, Macklem PT, Thurlbeck WM. Site and nature of airway obstruction in chronic obstructive lung disease. N Engl J Med. 1968;278(25):1355-60.  14.  Mead J. The lung's "quiet zone". N Engl J Med. 1970;282(23):1318-9.  15.  Rennard SI, editor. Clinical management of chronic obstructive pulmonary disease. 2 ed: CRC Press; 2008.  16.  Macklem PT. A century of the mechanics of breathing. Am J Respir Crit Care Med. 2004;170(1):10-5.  17.  Fletcher C, Peto R, Tinker C, Speizer F. The natural history of chronic bronchitis and emphysema. New York: Oxford University Press; 1976.  18.  Fletcher C, Peto R. The natural history of chronic airflow obstruction. Br Med J. 1977;1(6077):1645-8.  19.  Nobelprize.org. "Allan M. Cormack - Autobiography". 16 Jan 2012; Available from: http://www.nobelprize.org/nobel_prizes/medicine/laureates/1979/cormack-autobio.html.  20.  Nobelprize.org. "Godfrey N. Hounsfield - Autobiography". 16 Jan 2012.  21.  Guy C, Ffytche D. Introduction to the Principles of Medical Imaging: World Scientific Publishing Company; 2005. 420 p.  22.  Mayo JR. Radiation dose issues in longitudinal studies involving computed tomography. Proceedings of the American Thoracic Society. 2008;5(9):934-9.  23.  Buist AS, Anzueto A, Calverley PM, de Guia TS, Fukuchi Y, Jenkins C, et al. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease. www.goldcopd.com2007.  86  24.  Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the ATS/ERS position paper. Eur Respir J. 2004;23(6):932-46.  25.  Pistolesi M, Camiciottoli G, Paoletti M, Marmai C, Lavorini F, Meoni E, et al. Identification of a predominant COPD phenotype in clinical practice. Respir Med. 2008;102(3):367-76.  26.  Kitaguchi Y, Fujimoto K, Kubo K, Honda T. Characteristics of COPD phenotypes classified according to the findings of HRCT. Respir Med. 2006;100(10):1742-52.  27.  Eriksson S. Pulmonary Emphysema and Alpha1-Antitrypsin Deficiency. Acta medica Scandinavica. 1964;175:197-205.  28.  Bergin C, Muller N, Nichols DM, Lillington G, Hogg JC, Mullen B, et al. The diagnosis of emphysema. A computed tomographic-pathologic correlation. The American review of respiratory disease. 1986;133(4):541-6.  29.  Morrison NJ, Abboud RT, Ramadan F, Miller RR, Gibson NN, Evans KG, et al. Comparison of single breath carbon monoxide diffusing capacity and pressure-volume curves in detecting emphysema. The American review of respiratory disease. 1989;139(5):1179-87.  30.  Morrison NJ, Abboud RT, Muller NL, Miller RR, Gibson NN, Nelems B, et al. Pulmonary capillary blood volume in emphysema. The American review of respiratory disease. 1990;141(1):53-61.  31.  Biernacki W, Gould GA, Whyte KF, Flenley DC. Pulmonary hemodynamics, gas exchange, and the severity of emphysema as assessed by quantitative CT scan in chronic bronchitis and emphysema. The American review of respiratory disease. 1989;139(6):1509-15.  32.  Muller NL, Staples CA, Miller RR, Abboud RT. "Density mask". An objective method to quantitate emphysema using computed tomography. Chest. 1988;94(4):782-7.  33.  Bankier AA, De Maertelaer V, Keyzer C, Gevenois PA. Pulmonary emphysema: subjective visual grading versus objective quantification with macroscopic morphometry and thin-section CT densitometry. Radiology. 1999;211(3):851-8.  87  34.  Gould GA, MacNee W, McLean A, Warren PM, Redpath A, Best JJ, et al. CT measurements of lung density in life can quantitate distal airspace enlargement--an essential defining feature of human emphysema. The American review of respiratory disease. 1988;137(2):380-92.  35.  Dirksen A, Friis M, Olesen KP, Skovgaard LT, Sorensen K. Progress of emphysema in severe alpha 1-antitrypsin deficiency as assessed by annual CT. Acta Radiol. 1997;38(5):826-32.  36.  Shaker SB, Dirksen A, Laursen LC, Maltbaek N, Christensen L, Sander U, et al. Short-term reproducibility of computed tomography-based lung density measurements in alpha-1 antitrypsin deficiency and smokers with emphysema. Acta Radiol. 2004;45(4):424-30.  37.  Gevenois PA, de Maertelaer V, De Vuyst P, Zanen J, Yernault JC. Comparison of computed density and macroscopic morphometry in pulmonary emphysema. Am J Respir Crit Care Med. 1995;152(2):653-7.  38.  Gevenois PA, De Vuyst P, Sy M, Scillia P, Chaminade L, de Maertelaer V, et al. Pulmonary emphysema: quantitative CT during expiration. Radiology. 1996;199(3):825-9.  39.  Hoffman EA, Simon BA, McLennan G. State of the Art. A structural and functional assessment of the lung via multidetector-row computed tomography: phenotyping chronic obstructive pulmonary disease. Proceedings of the American Thoracic Society. 2006;3(6):519-32.  40.  Mishima M, Hirai T, Itoh H, Nakano Y, Sakai H, Muro S, et al. Complexity of terminal airspace geometry assessed by lung computed tomography in normal subjects and patients with chronic obstructive pulmonary disease. Proc Natl Acad Sci U S A. 1999;96(16):8829-34.  41.  Camp PG, Coxson HO, Levy RD, Pillai SG, Anderson W, Vestbo J, et al. Sex differences in emphysema and airway disease in smokers. Chest. 2009;136(6):1480-8.  42.  Madani A, Van Muylem A, de Maertelaer V, Zanen J, Gevenois PA. Pulmonary emphysema: size distribution of emphysematous spaces on multidetector CT images--comparison with macroscopic and microscopic morphometry. Radiology. 2008;248(3):1036-41.  88  43.  Nakano Y, Muro S, Sakai H, Hirai T, Chin K, Tsukino M, et al. Computed tomographic measurements of airway dimensions and emphysema in smokers. Correlation with lung function. Am J Respir Crit Care Med. 2000;162(3 Pt 1):1102-8.  44.  Orlandi I, Moroni C, Camiciottoli G, Bartolucci M, Pistolesi M, Villari N, et al. Chronic obstructive pulmonary disease: thin-section CT measurement of airway wall thickness and lung attenuation. Radiology. 2005;234(2):604-10.  45.  Grydeland TB, Dirksen A, Coxson HO, Eagan TM, Thorsen E, Pillai SG, et al. Quantitative computed tomography measures of emphysema and airway wall thickness are related to respiratory symptoms. Am J Respir Crit Care Med. 2010;181(4):353-9.  46.  Mair G, Maclay J, Miller JJ, McAllister D, Connell M, Murchison JT, et al. Airway dimensions in COPD: relationships with clinical variables. Respir Med. 2010;104(11):1683-90.  47.  Nakano Y, Wong JC, de Jong PA, Buzatu L, Nagao T, Coxson HO, et al. The prediction of small airway dimensions using computed tomography. Am J Respir Crit Care Med. 2005;171(2):142-6.  48.  McNamara AE, Muller NL, Okazawa M, Arntorp J, Wiggs BR, Pare PD. Airway narrowing in excised canine lungs measured by high-resolution computed tomography. J Appl Physiol. 1992;73(1):307-16.  49.  Seneterre E, Paganin F, Bruel JM, Michel FB, Bousquet J. Measurement of the internal size of bronchi using high resolution computed tomography (HRCT). Eur Respir J. 1994;7(3):596-600.  50.  Okazawa M, Muller N, McNamara AE, Child S, Verburgt L, Pare PD. Human airway narrowing measured using high resolution computed tomography. Am J Respir Crit Care Med. 1996;154(5):1557-62.  51.  Baxter BS, Sorenson JA. Factors affecting the measurement of size and CT number in computed tomography. Investigative radiology. 1981;16(4):337-41.  89  52.  Schwab RJ, Gefter WB, Pack AI, Hoffman EA. Dynamic imaging of the upper airway during respiration in normal subjects. J Appl Physiol. 1993;74(4):1504-14.  53.  King GG, Muller NL, Pare PD. Evaluation of airways in obstructive pulmonary disease using highresolution computed tomography. Am J Respir Crit Care Med. 1999;159(3):992-1004.  54.  Saba OI, Hoffman EA, Reinhardt JM. Maximizing quantitative accuracy of lung airway lumen and wall measures obtained from X-ray CT imaging. J Appl Physiol. 2003;95(3):1063-75.  55.  King GG, Muller NL, Whittall KP, Xiang QS, Pare PD. An analysis algorithm for measuring airway lumen and wall areas from high-resolution computed tomographic data. Am J Respir Crit Care Med. 2000;161(2 Pt 1):574-80.  56.  King GG, Carroll JD, Muller NL, Whittall KP, Gao M, Nakano Y, et al. Heterogeneity of narrowing in normal and asthmatic airways measured by HRCT. Eur Respir J. 2004;24(2):211-8.  57.  Reinhardt JM, D'Souza ND, Hoffman EA. Accurate measurement of intrathoracic airways. IEEE transactions on medical imaging. 1997;16(6):820-7.  58.  Estepar RS, Washko GG, Silverman EK, Reilly JJ, Kikinis R, Westin CF. Accurate airway wall estimation using phase congruency. Medical image computing and computer-assisted intervention : MICCAI International Conference on Medical Image Computing and ComputerAssisted Intervention. 2006;9(Pt 2):125-34.  59.  Washko GR, Dransfield MT, Estepar RS, Diaz A, Matsuoka S, Yamashiro T, et al. Airway wall attenuation: a biomarker of airway disease in subjects with COPD. J Appl Physiol. 2009;107(1):185-91.  60.  Davis GR, Wong FS. X-ray microtomography of bones and teeth. Physiol Meas. 1996;17(3):12146.  61.  Schambach SJ, Bag S, Schilling L, Groden C, Brockmann MA. Application of micro-CT in small animal imaging. Methods. 2010;50(1):2-13.  90  62.  Westneat MW, Socha JJ, Lee WK. Advances in biological structure, function, and physiology using synchrotron X-ray imaging*. Annu Rev Physiol. 2008;70:119-42.  63.  Litzlbauer HD, Korbel K, Kline TL, Jorgensen SM, Eaker DR, Bohle RM, et al. Synchrotron-based micro-CT imaging of the human lung acinus. Anat Rec (Hoboken). 2010;293(9):1607-14.  64.  Ikura H, Shimizu K, Ikezoe J, Nagareda T, Yagi N. In vitro evaluation of normal and abnormal lungs with ultra-high-resolution CT. J Thorac Imaging. 2004;19(1):8-15.  65.  Yong HS, Kang EY, Kim YK, Woo OH, Shin BK, Oh CH, et al. Phase contrast microradiography of mouse lung using synchrotron X-ray: correlation with optical microscopy. Yonsei Med J. 2009;50(3):422-6.  66.  Artaechevarria X, Blanco D, Perez-Martin D, de Biurrun G, Montuenga LM, de Torres JP, et al. Longitudinal study of a mouse model of chronic pulmonary inflammation using breath hold gated micro-CT. Eur Radiol. 2010;20(11):2600-8.  67.  Froese AR, Ask K, Labiris R, Farncombe T, Warburton D, Inman MD, et al. Three-dimensional computed tomography imaging in an animal model of emphysema. Eur Respir J. 2007;30(6):1082-9.  68.  Ford NL, Martin EL, Lewis JF, Veldhuizen RA, Holdsworth DW, Drangova M. Quantifying lung morphology with respiratory-gated micro-CT in a murine model of emphysema. Phys Med Biol. 2009;54(7):2121-30.  69.  De Clerck NM, Meurrens K, Weiler H, Van Dyck D, Van Houtte G, Terpstra P, et al. Highresolution X-ray microtomography for the detection of lung tumors in living mice. Neoplasia. 2004;6(4):374-9.  70.  Cavanaugh D, Johnson E, Price RE, Kurie J, Travis EL, Cody DD. In vivo respiratory-gated micro-CT imaging in small-animal oncology models. Mol Imaging. 2004;3(1):55-62.  91  71.  Fushiki H, Kanoh-Azuma T, Katoh M, Kawabata K, Jiang J, Tsuchiya N, et al. Quantification of mouse pulmonary cancer models by microcomputed tomography imaging. Cancer Sci. 2009;100(8):1544-9.  72.  Cavanaugh D, Travis EL, Price RE, Gladish G, White RA, Wang M, et al. Quantification of bleomycin-induced murine lung damage in vivo with micro-computed tomography. Acad Radiol. 2006;13(12):1505-12.  73.  Johnson KA. Imaging techniques for small animal imaging models of pulmonary disease: microCT. Toxicol Pathol. 2007;35(1):59-64.  74.  Shofer S, Badea C, Qi Y, Potts E, Foster WM, Johnson GA. A micro-CT analysis of murine lung recruitment in bleomycin-induced lung injury. J Appl Physiol. 2008;105(2):669-77.  75.  Lundblad LK, Thompson-Figueroa J, Allen GB, Rinaldi L, Norton RJ, Irvin CG, et al. Airway hyperresponsiveness in allergically inflamed mice: the role of airway closure. Am J Respir Crit Care Med. 2007;175(8):768-74.  76.  Lederlin M, Ozier A, Montaudon M, Begueret H, Ousova O, Marthan R, et al. Airway remodeling in a mouse asthma model assessed by in-vivo respiratory-gated micro-computed tomography. Eur Radiol. 2010;20(1):128-37.  77.  Lee HJ, Goo JM, Kim NR, Kim MA, Chung DH, Son KR, et al. Semiquantitative measurement of murine bleomycin-induced lung fibrosis in in vivo and postmortem conditions using microcomputed tomography: correlation with pathologic scores--initial results. Investigative radiology. 2008;43(6):453-60.  78.  Parameswaran H, Bartolak-Suki E, Hamakawa H, Majumdar A, Allen PG, Suki B. Threedimensional measurement of alveolar airspace volumes in normal and emphysematous lungs using micro-CT. J Appl Physiol. 2009;107(2):583-92.  92  79.  Langheinrich AC, Leithauser B, Greschus S, Von Gerlach S, Breithecker A, Matthias FR, et al. Acute rat lung injury: feasibility of assessment with micro-CT. Radiology. 2004;233(1):165-71.  80.  Dame Carroll JR, Chandra A, Jones AS, Berend N, Magnussen JS, King GG. Airway dimensions measured from micro-computed tomography and high-resolution computed tomography. Eur Respir J. 2006;28(4):712-20.  81.  Litzlbauer HD, Neuhaeuser C, Moell A, Greschus S, Breithecker A, Franke FE, et al. Threedimensional imaging and morphometric analysis of alveolar tissue from microfocal X-raycomputed tomography. Am J Physiol Lung Cell Mol Physiol. 2006;291(3):L535-45.  82.  Watz H, Breithecker A, Rau WS, Kriete A. Micro-CT of the human lung: imaging of alveoli and virtual endoscopy of an alveolar duct in a normal lung and in a lung with centrilobular emphysema--initial observations. Radiology. 2005;236(3):1053-8.  83.  Hogg JC, McDonough JE, Gosselink JV, Hayashi S. What drives the peripheral lung-remodeling process in chronic obstructive pulmonary disease? Proceedings of the American Thoracic Society. 2009;6(8):668-72.  84.  Buffon G. Essai d'arithmetique morale. Supplément à l'Histoire Naturelle. 1777;4.  85.  Campbell H, Tomkeieff SI. Calculation of the internal surface of a lung. Nature. 1952;170(4316):116-7.  86.  Weibel ER, Gomez DM. Architecture of the human lung. Use of quantitative methods establishes fundamental relations between size and number of lung structures. Science. 1962;137:577-85.  87.  Sterio DC. The unbiased estimation of number and sizes of arbitrary particles using the disector. J Microsc. 1984;134(Pt 2):127-36.  88.  Hsia CC, Hyde DM, Ochs M, Weibel ER. An official research policy statement of the American Thoracic Society/European Respiratory Society: standards for quantitative assessment of lung structure. Am J Respir Crit Care Med. 2010;181(4):394-418.  93  89.  Howard CV, Reed MG. Unbiased Stereology: Three-Dimensional Measurement in Microscopy. Oxford, UK: BIOS Scientific Publishers Ltd; 1998.  90.  Hislop AA, Wigglesworth JS, Desai R. Alveolar development in the human fetus and infant. Early Hum Dev. 1986;13(1):1-11.  91.  Thurlbeck WM. Postnatal human lung growth. Thorax. 1982;37(8):564-71.  92.  Zeltner TB, Caduff JH, Gehr P, Pfenninger J, Burri PH. The postnatal development and growth of the human lung. I. Morphometry. Respir Physiol. 1987;67(3):247-67.  93.  Balinotti JE, Tiller CJ, Llapur CJ, Jones MH, Kimmel RN, Coates CE, et al. Growth of the lung parenchyma early in life. Am J Respir Crit Care Med. 2009;179(2):134-7.  94.  Hansen JE, Ampaya EP, Bryant GH, Navin JJ. Branching pattern of airways and air spaces of a single human terminal bronchiole. J Appl Physiol. 1975;38(6):983-9.  95.  Angus GE, Thurlbeck WM. Number of alveoli in the human lung. J Appl Physiol. 1972;32(4):4835.  96.  Ochs M, Nyengaard JR, Jung A, Knudsen L, Voigt M, Wahlers T, et al. The number of alveoli in the human lung. Am J Respir Crit Care Med. 2004;169(1):120-4.  97.  McDonough JE, Yuan R, Suzuki M, Seyednejad N, Elliott WM, Sanchez PG, et al. Small-airway obstruction and emphysema in chronic obstructive pulmonary disease. N Engl J Med. 2011;365(17):1567-75.  98.  Dunnill MS. Quantitative methods in the study of pulmonary pathology. Thorax. 1962;17:320-8.  99.  Robbesom AA, Versteeg EM, Veerkamp JH, van Krieken JH, Bulten HJ, Smits HT, et al. Morphological quantification of emphysema in small human lung specimens: comparison of methods and relation with clinical data. Mod Pathol. 2003;16(1):1-7.  100.  Dunnill MS. Evaluation of a Simple Method of Sampling the Lung for Quantitative Histological Analysis. Thorax. 1964;19:443-8.  94  101.  Hansen JE, Ampaya EP. Lung morphometry: a fallacy in the use of the counting principle. J Appl Physiol. 1974;37(6):951-4.  102.  Thurlbeck WM. The internal surface area of nonemphysematous lungs. The American review of respiratory disease. 1967;95(5):765-73.  103.  Hansen JE, Ampaya EP. Human air space shapes, sizes, areas, and volumes. J Appl Physiol. 1975;38(6):990-5.  104.  Kitaoka H. Numerical density of alveoli in the human lung--measurement using serial thin slices as three-dimensional probes. Tohoku J Exp Med. 1993;170(1):45-52.  105.  Haefeli-Bleuer B, Weibel ER. Morphometry of the human pulmonary acinus. Anat Rec. 1988;220(4):401-14.  106.  Pump KK. Morphology of the acinus of the human lung. Dis Chest. 1969;56(2):126-34.  107.  Daly WJ, Bondurant S. Direct measurement of respiratory pleural pressure changes in normal man. J Appl Physiol. 1963;18(3):513-8.  108.  Glenny RW, Bernard S, Robertson HT, Hlastala MP. Gravity is an important but secondary determinant of regional pulmonary blood flow in upright primates. J Appl Physiol. 1999;86(2):623-32.  109.  Gehr P, Weibel ER. Morphometric estimation of regional differences in the dog lung. J Appl Physiol. 1974;37(5):648-53.  110.  Glazier JB, Hughes JM, Maloney JE, West JB. Vertical gradient of alveolar size in lungs of dogs frozen intact. J Appl Physiol. 1967;23(5):694-705.  111.  Hogg JC, Nepszy S. Regional lung volume and pleural pressure gradient estimated from lung density in dogs. J Appl Physiol. 1969;27(2):198-203.  112.  Sutherland PW, Katsura T, Milic-Emili J. Previous volume history of the lung and regional distribution of gas. J Appl Physiol. 1968;25(5):566-74.  95  113.  Pare PD, Boucher R, Michoud MC, Hogg JC. Static lung mechanics of intact and excised rhesus monkey lungs and lobes. J Appl Physiol. 1978;44(4):547-52.  114.  Silvers GW, Petty TL, Stanford RE, Filley GF. The elastic properties of lobes of excised human lungs. The American review of respiratory disease. 1979;120(1):207-9.  115.  Matsuba K, Wright JL, Wiggs BR, Pare PD, Hogg JC. The changes in airways structure associated with reduced forced expiratory volume in one second. Eur Respir J. 1989;2(9):834-9.  116.  Bosken CH, Wiggs BR, Pare PD, Hogg JC. Small airway dimensions in smokers with obstruction to airflow. The American review of respiratory disease. 1990;142(3):563-70.  117.  Hasegawa M, Nasuhara Y, Onodera Y, Makita H, Nagai K, Fuke S, et al. Airflow limitation and airway dimensions in chronic obstructive pulmonary disease. Am J Respir Crit Care Med. 2006;173(12):1309-15.  118.  Berger P, Perot V, Desbarats P, Tunon-de-Lara JM, Marthan R, Laurent F. Airway wall thickness in cigarette smokers: quantitative thin-section CT assessment. Radiology. 2005;235(3):1055-64.  119.  Coxson HO, Quiney B, Sin DD, Xing L, McWilliams AM, Mayo JR, et al. Airway wall thickness assessed using computed tomography and optical coherence tomography. Am J Respir Crit Care Med. 2008;177(11):1201-6.  120.  Achenbach T, Weinheimer O, Brochhausen C, Hollemann D, Baumbach B, Scholz A, et al. Accuracy of automatic airway morphometry in computed tomography-Correlation of radiological-pathological findings. Eur J Radiol. 2010.  121.  Sera T, Fujioka H, Yokota H, Makinouchi A, Himeno R, Schroter RC, et al. Three-dimensional visualization and morphometry of small airways from microfocal X-ray computed tomography. J Biomech. 2003;36(11):1587-94.  96  122.  Van Brabandt H, Cauberghs M, Verbeken E, Moerman P, Lauweryns JM, Van de Woestijne KP. Partitioning of pulmonary impedance in excised human and canine lungs. J Appl Physiol. 1983;55(6):1733-42.  123.  Yanai M, Sekizawa K, Ohrui T, Sasaki H, Takishima T. Site of airway obstruction in pulmonary disease: direct measurement of intrabronchial pressure. J Appl Physiol. 1992;72(3):1016-23.  124.  Pedley TJ, Schroter RC, Sudlow MF. The prediction of pressure drop and variation of resistance within the human bronchial airways. Respir Physiol. 1970;9(3):387-405.  125.  Diaz AA, Valim C, Yamashiro T, Estepar RS, Ross JC, Matsuoka S, et al. Airway count and emphysema assessed by chest CT imaging predicts clinical outcome in smokers. Chest. 2010;138(4):880-7.  126.  Lam S, leRiche JC, McWilliams A, Macaulay C, Dyachkova Y, Szabo E, et al. A randomized phase IIb trial of pulmicort turbuhaler (budesonide) in people with dysplasia of the bronchial epithelium. Clin Cancer Res. 2004;10(19):6502-11.  127.  Lam S, MacAulay C, Le Riche JC, Dyachkova Y, Coldman A, Guillaud M, et al. A randomized phase IIb trial of anethole dithiolethione in smokers with bronchial dysplasia. J Natl Cancer Inst. 2002;94(13):1001-9.  128.  McWilliams AM, Mayo JR, Ahn MI, MacDonald SL, Lam SC. Lung cancer screening using multislice thin-section computed tomography and autofluorescence bronchoscopy. J Thorac Oncol. 2006;1(1):61-8.  129.  Choong CK, Haddad FJ, Martinez C, Hu DZ, Pierce JA, Meyers BF, et al. A simple, reproducible, and inexpensive technique in the preparation of explanted emphysematous lungs for ex vivo studies. J Thorac Cardiovasc Surg. 2005;130(3):922-3.  130.  Leopold JG, Gough J. The centrilobular form of hypertrophic emphysema and its relation to chronic bronchitis. Thorax. 1957;12(3):219-35.  97  131.  Hogg JC, Macklem PT, Thurlbeck WM. The resistance of collateral channels in excised human lungs. The Journal of clinical investigation. 1969;48(3):421-31.  132.  Bignon J, Khoury F, Even P, Andre J, Brouet G. Morphometric study in chronic obstructive bronchopulmonary disease. Pathologic, clinical, and physiologic correlations. The American review of respiratory disease. 1969;99(5):669-95.  133.  Matsuba K, Thurlbeck WM. The number and dimensions of small airways in emphysematous lungs. Am J Pathol. 1972;67(2):265-75.  134.  Weibel ER. What makes a good lung? Swiss Med Wkly. 2009;139(27-28):375-86.  135.  Rohrer F. Der Stromungswiderstand in der menschlichen Atemwegen und der Einfluss der unregelmässigen Verzweigung es Bronchial-systems auf der Atmungsverlauf in vershiedenen Lungenbezinken. Arch Ges Physiol. 1915;162:225-9.  136.  Findeisen W. Uber das Absetzen kleiner, in dur Luft suspendierter Teilchen in der menschlichen Lunge bei der Atmung. Arch Ges Physiol. 1935;236:367-79.  137.  Berend N, Skoog C, Thurlbeck WM. Lobar pressure-volume characteristics of excised human lungs. Thorax. 1981;36(4):290-5.  138.  Verleden GM, Vos R, De Vleeschauwer SI, Willems-Widyastuti A, Verleden SE, Dupont LJ, et al. Obliterative bronchiolitis following lung transplantation: from old to new concepts? Transplant international : official journal of the European Society for Organ Transplantation. 2009;22(8):771-9.  139.  Raghu G, Collard HR, Egan JJ, Martinez FJ, Behr J, Brown KK, et al. An official ATS/ERS/JRS/ALAT statement: idiopathic pulmonary fibrosis: evidence-based guidelines for diagnosis and management. Am J Respir Crit Care Med. 2011;183(6):788-824.  140.  Miller WS. Key points in lung structure. Radiology. 1925;4(3):173-7.  98  141.  Niewoehner DE, Kleinerman J, Rice DB. Pathologic changes in the peripheral airways of young cigarette smokers. N Engl J Med. 1974;291(15):755-8.  142.  Lee SH, Goswami S, Grudo A, Song LZ, Bandi V, Goodnight-White S, et al. Antielastin autoimmunity in tobacco smoking-induced emphysema. Nature medicine. 2007;13(5):567-9.  143.  Hemminki K, Liu X, Ji J, Sundquist K, Sundquist J. Subsequent COPD and lung cancer in patients with autoimmune disease. Eur Respir J. 2011;37(2):463-5.  144.  Stefanska AM, Walsh PT. Chronic obstructive pulmonary disease: evidence for an autoimmune component. Cellular & molecular immunology. 2009;6(2):81-6.  99  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0072712/manifest

Comment

Related Items