Endobronchial Ultrasonography for the Quantitative Assessment of Bronchial Mural Structures in Lung Transplant Recipients*

Transcription

1 CHEST Original Research INTERVENTIONAL PULMONOLOGY Endobronchial Ultrasonography for the Quantitative Assessment of Bronchial Mural Structures in Lung Transplant Recipients* Sarosh Irani, MD; Thomas Hess, MD; Markus Hofer, MD; Ariana Gaspert, MD; Lucas M. Bachmann, MD; Erich W. Russi, MD, FCCP; and Annette Boehler, MD, FCCP Background: Endobronchial ultrasonography (EBUS) has been shown to be an adequate tool to visualize the laminar structure of the bronchial wall. The purpose of this study was to investigate the potential of EBUS to identify and quantitatively assess bronchial wall structures in lung transplant recipients. Methods: EBUS was performed with a radial 20-MHz ultrasonic miniprobe in lung transplant recipients undergoing surveillance bronchoscopies. Sequential pictures were taken of the main bronchus (proximal of anastomosis) and proximal right intermedius bronchus or proximal left lower lobe bronchus (distal of anastomosis), respectively. From every localization, five slides were chosen. The quantitative assessment of the digitized pictures was done with the aid of image analysis software. In addition to the comparison of the different layers between patients with and without infection and rejection, respectively, the intraclass correlation coefficients (ICCs) of the different measurements were calculated. Results: From 20 EBUS examinations performed in 10 lung transplant recipients, 200 slides were selected for quantitative assessment. A five-layer composition could be identified in all selected slides. The relative area of layer two (hypoechoic submucosal tissue) of the autologous part was significantly smaller in patients with graft rejection (p 0.04) compared to patients without rejection, and significantly larger in patients with graft infection (p 0.02) compared to patients without graft infection. The ICC values were calculated in 50 different slides in a subset of five consecutive patients (0.91, 0.95, 0.88, and 0.91 for layers 1, 2, 3, and 5 of the autologous and 0.70, 0.92, 0.88, and 0.84 for the allogeneic parts, respectively). Conclusions: EBUS enables to discriminate different layers of the bronchi in humans and to measure the thickness of these layers in a reproducible fashion. Therefore, EBUS may be used to investigate and quantify inflammatory alterations of bronchial wall structures in vivo. (CHEST 2006; 129: ) Key words: acute graft rejection; endobronchial ultrasonography; lung transplant recipients; quantitative assessment Abbreviations: EBUS endobronchial ultrasonography; ICC intraclass correlation coefficient; ISHLT International Society for Heart and Lung Transplantation; TBB transbronchial biopsy Endobronchial ultrasonography (EBUS) is successfully used for the detection of mediastinal lymph nodes and other mediastinal structures. 1,2 Furthermore, EBUS has been shown to be a suitable tool to visualize the laminar structure of the bronchial wall. 1,3 6 The correlation between the ultrasonographic images and the real bronchial wall structures have been clarified by means of needle puncture experiments 3,4 and by predicting the tracheobronchial wall invasion of intrathoracic malignancies with consecutive histologic confirmation. 3 5,7,8 After lung transplantation, the bronchial epithelium is considered a major target of alloimmunologic processes in both small 9 and large airways. Lymphocytic bronchitis and lymphocytic bronchiolitis represent acute airway rejection and can be CHEST / 129 / 2/ FEBRUARY,

2 graded according to the criteria of the International Society for Heart and Lung Transplantation (ISHLT). 9 In addition to histologic findings in airway biopsies of transplanted lungs, alterations of the bronchial mucosal blood flow during graft rejection have been reported in an animal model. 16 To our knowledge, it has not yet been studied whether bronchial mural structures can be quantitatively assessed by 20-MHz ultrasonic probes and whether airway alterations due to inflammatory injuries can be detected with the aid of this noninvasive technique. The rationale of the present study was to investigate the ability of EBUS to discriminate the bronchial wall structure in lung transplant recipients undergoing routine bronchoscopy and to quantitatively assess the findings in both autologous and allogeneic parts of the large airways. The comparison of these findings with the histologic and microbiological results of the surveillance bronchoscopy was a further purpose of this study. Materials and Methods Consecutive bilateral lung transplant recipients who were scheduled for routine surveillance bronchoscopy were included in the study. Before BAL and transbronchial biopsies (TBBs) were performed, ipsilaterally EBUS of the proximal main bronchus (autologous tissue, proximally of the bronchial anastomosis) and the intermediate bronchus or, on the left side, proximal lower lobe bronchus (allogeneic tissue, distally of the bronchial anastomosis) was performed. To avoid local mechanical manipulation, the anastomosis itself was not investigated. Patients were excluded if they had clinical or radiologic evidence of intercurrent lung infection or a contraindication for TBB (ie, coagulopathy); moreover, single-lung transplant recipients were excluded from the study. The study was approved by the Ethics Committee of our hospital, and written informed consent from the patients was obtained. Pulmonary Function Testing Multiple serial pulmonary function tests (spirometry and body plethysmography; Autobox; SensorMedics; Yorba Linda, CA) *From the Division of Pulmonary Medicine (Drs. Irani, Hofer, Russi, and Boehler) and Institute of Clinical Pathology (Dr. Gaspert), University Hospital, Zurich; Pneumology (Dr. Hess), Medical Clinic, Kantonsspital, Winterthur; and Horten Center (Dr. Bachmann), University Zurich, Zurich, Switzerland. This work was presented in part at the Annual Congress of the European Respiratory Society in Glasgow, Scotland, September 4 to 8, Dr. Boehler holds a professorship position from the Swiss National Science Foundation. Equipment was provided by Olympus, Switzerland. Manuscript received April 28, 2005; revision accepted June 17, Reproduction of this article is prohibited without written permission from the American College of Chest Physicians ( org/misc/reprints.shtml). Correspondence to: Annette Boehler, MD, Division of Pulmonary Medicine and Lung Transplant Program, C HOER 27, University Hospital, Ramistrasse 100, CH-8091 Zurich, Switzerland; capybara@compuserve.com were available in each patient. The latest test was performed 1 day prior to bronchoscopy. The terms baseline values and bronchiolitis obliterans syndrome were used as defined elsewhere. 17 Fiberoptic Bronchoscopy With BAL and Biopsies Bronchoscopy was performed with IV conscious sedation with midazolam (Roche; Basel, Switzerland) and hydrocodon (Knoll; Liestal, Switzerland). The upper airways were anesthetized with topical 10% lidocaine, and the lower airways were anesthetized with inhalation of 4% lidocaine and local application of 1% lidocaine. BAL of the middle lobe or lingula was performed after wedging the bronchoscope in a suitable segment or subsegment bronchus with 4 50 ml of 0.9% saline solution. BAL was followed by TBB. Seven to nine TBB specimens were taken from the upper and lower lobes of the same lung under fluoroscopic guidance (FB19C1 forceps; Olympus; Tokyo, Japan). TBB Assessment: TBB samples were stained with hematoxylineosin and Elastin van Gieson to assess acute rejection or bronchiolitis obliterans according to standard criteria. 9 Acute rejection was defined as every ISHLT grade other than A0B0. Stains for fungi and for cytomegalovirus were also performed. BAL Cell Processing Total cell counts were performed on the unfiltered BAL fluid using a hemocytometer (1: 21; Sysmex Corporation; Kobe, Japan). Cytocentrifuge preparations (1,800g for 5 min) were made using 50 L of unfiltered BAL aspirate, and cell counts were performed by counting 200 cells. One portion of BAL was sent for routine microbiologic examinations. Graft infection was defined as BAL positive for any microorganism other than oral flora. Ultrasonography and Quantitative Assessment An ultrasonography miniprobe (UM-BS20 26R; Olympus) combined with a balloon sheath was used. Additionally, our EBUS unit consists of an Olympus EU-M30 processor and an Olympus MH-240 driving unit. Under real-time conditions, the entire EBUS examination was recorded on videotape and by means of an analog/digital converter system (MovieBox DV; Pinnacle Systems; Mountain View, CA) on a laptop computer. Further analysis of the data was performed in a second phase outside the endoscopy unit. With a step-by-step analysis (Pinnacle Studio 8, Version 8.6; Pinnacle Systems), the digitalized film was screened for at least five representative slides from each autologous and allogeneic bronchial portion. From cartilagecontaining parts of the bronchial wall, those slides were selected by a blinded investigator (S.I.) that showed the most obvious and well-defined laminar structure. In order to assess as many views as possible, this procedure was applied to the complete recording available. The files were saved in tagged image file format, and the image displayed on the monitor was measured by the computer (analysis software, Version 3.1; Soft Imaging System; Munster, Germany). As mentioned, due to its well-defined configuration, we decided to take the cartilaginous portion of the bronchus into detailed consideration. Firstly, the image size was calibrated. Secondly, the largest possible sector starting at the center of the bronchus containing cartilage was defined. Thirdly, the absolute values of the thickness of each layer and the relative value of the area of each layer were measured (Fig 1). Due to the anatomic inconsistency of the diameter of the cartilage, the relative values of the other four layers were also calculated after exclusion of the area of the cartilage. For statistical analysis, the mean values of the five measurements were used. 350 Original Research

3 Figure 1. Analysis of a representative ultrasonogram of the right main bronchus. Top left, A: Calibration of the computerized image (software analysis). Top right, B: Definition of the cartilaginous portion to be measured; the sector starts in the center of the bronchus. Bottom left, C: Measurement of the absolute thickness of each of the five layers. Bottom right, D: Measurement of the relative area of each layer. Statistical Analysis Statistical software (Statistica, release 6.0; StatSoft; Tulsa, OK) was used for statistical analysis. When a range is stated, it refers to the interquartile range of the raw data. Wall thickness (absolute and relative) was compared using the nonparametric Mann-Whitney U test. Correlation coefficients (r) were obtained by Spearman rank method; p 0.05 was considered significant. To estimate the reliability of the measurements, we performed analysis of variance for each layer and localization with patient as the random factor and layer thickness percentage as the dependent variable in five consecutive patients. Based on the variance components, we calculated the intraclass correlation coefficients (ICCs) using the patient variance in the numerator and the sum of residual and patient variance in the denominator. Results Twenty EBUS examinations were performed in 10 lung transplant recipients who underwent flexible bronchoscopy for routine surveillance biopsy. A total of 200 slides were selected for further analysis. Clinical and demographic details of the lung transplant recipients are summarized in Table 1. All patients received immunosuppressive triple therapy with cyclosporine, mycophenolate mofetil, and prednisone. All procedures were performed without complications; particularly, no EBUS-associated adverse events occurred (ie, rupture or stenosis formation). The ultrasound examination prolonged the endoscopy procedure by an average of 5.4 min (range, 4.1 to 7.3 min). Fifty different slides of autologous and allogeneic parts of airways in a subset of five consecutive patients were used to determine the reliability of the quantitative results of the determination of the relative areas 1, 2, 3, and 5. The ICC values for these measurements were 0.91, 0.95, 0.88, and 0.91 for layers 1, 2, 3, and 5 of the autologous and 0.70, 0.92, 0.88, and 0.84 for the allogeneic parts, respectively. Figures 2, 3 present the quantified layer thickness (upper half) and the relative proportions after extraction of the cartilage (lower half) in both autologous and allogeneic parts of airways. The correlation of measurements between these two tissue compartments are shown in Table 2. A statistically significant correlation was found for layer three (inner hyperechoic marginal echo of the cartilage) between autologous and allogeneic airways. In Figure 2, the comparison of the assessed parameters between endoscopies of patients with acute CHEST / 129 / 2/ FEBRUARY,

4 Table 1 Clinical Details of the Lung Transplant Recipients Studied* Patient No. Sex Age, yr Time From Tx, mo Type of Tx Underlying Disease Rejection ISHLT BAL Microbiology BAL Neutrophils, % BAL Lymphocytes, % BOS Stage 1 Male 63 3 Bilateral IPF A0B A0B0 Streptococcus viridans A0B0 S viridans Ax(1)Bx(0) A0B0 S viridans Male Bilateral Emphysema A0B0 S viridans A0B0 S viridans A0B0 S viridans Male 60 5 A2B Female 61 5 Bilateral Emphysema A0B0 Cytomegalovirus Female 48 5 Bilateral LAM A1B0 Anaerobic mixed flora Ax(0)Bx(0) S viridans Male 59 2 Bilateral Emphysema A0B Male 59 7 Bilateral Emphysema Ax(1)Bx(0) A2Bx(0) A2B A2B0 S viridans Female 48 8 Bilateral Emphysema A0B1 Anaerobic mixed flora Male 43 2 Bilateral Emphysema A0B0 S viridans Female 48 2 Bilateral PH A1B0 Anaerobic mixed flora *Tx transplantation; IPF idiopathic pulmonary fibrosis; LAM lymphangioleiomyomatosis; BOS bronchiolitis obliterans syndrome; PH idiopathic pulmonary arterial hypertension. Fewer than five evaluable fragments for histology. graft rejection (defined as all TBB results other than A0B0) and examinations of patients without acute rejection is shown. The relative area of the second layer of the autologous airways was statistically significantly smaller in the rejection group (p 0.04). For the absolute values, no statistically significant difference between the two groups was found. The results of ultrasonic measurements in cases with and without graft infection (defined as BAL containing any microorganism other than oral flora) are presented in Figure 3. A statistically significant difference between these groups was found in the relative area of layer two in the autologous part (p 0.02) of airways. Discussion Several properties of EBUS make this new technique potentially useful in assessing patients after lung transplantation. EBUS is a noninvasive and safe method to assess mediastinal, peribronchial, and bronchial wall structures. Since the airways of lung transplant recipients are the key structure of alloimmunologic and nonalloimmunologic inflammatory injuries finally leading to chronic rejection, particularly the latter is a thrilling and exclusive target of EBUS. Furthermore, in contrast to invasive diagnostic tools such as lung biopsy, with EBUS any number of images from different parts of the large airways can be assessed. The dynamic property of EBUS is an additional advantage of this technique. To our knowledge, this is the first study investigating EBUS in lung transplant recipients. Using a 20-MHz ultrasonic probe, we performed EBUS in the autologous (main bronchus) and allogeneic (intermediate bronchus on the right and proximal lower lobe bronchus on the left, respectively) parts of the large airways. At both localizations, we were able to discriminate the previously described multilayer structure of the airway wall. 1,3 6 In particular, the cartilaginous parts consistently showed a five-layer laminar composition that facilitated the quantitative assessment and individual comparison. Using a computerized image analysis system (analysis), we were able to quantitatively assess these layers in a consistent manner. The ICC calculations of 50 different slides confirmed reproducible measurements in both autologous and allogeneic parts of the airways. Although due to smaller diameter and irregular cartilage formation, the latter in general showed a slightly less well-defined bronchial wall pattern in the ultrasonographic examination, we found only minor differences of the ICC values between the two localizations. Overall, only for the absolute value of layer three did we find a statistically significant correlation between the main and lower lobe/intermediate bronchi. Considering the allogeneic origin of the latter, 352 Original Research

5 Figure 2. Thickness of the five layers (1: hyperechoic marginal echo, 2: hypoechoic submucosal tissue, 3: hyperechoic inner marginal echo of the cartilage, 4: hypoechoic cartilage, 5: hyperechoic outer marginal echo of the cartilage) measured in mm and the relative area of layer 1, 2, 3 and 5 in % (see text) in patients with acute graft rejection (dots, means ISHLT rejection grades other than A0B0) vs patients without rejection (triangles). On the left half the values of the autologous airways (main bronchus) and on the right half the allogeneic (intermediate bronchus and left lower lobe bronchus, respectively) are shown, respectively. this anatomic difference of the two compartments was not unexpected. However, we found no significant difference between the proximal and distal parts either (data not separately shown). When we compared the measured and calculated (relative) values between examinations with and without acute rejection, we found a significant difference of the relative area of layer two (submucosa) of the autologous parts, whereas in the allogeneic parts no difference could be shown. Since the absolute values of the second layer did not differ significantly between the rejection and nonrejection group and we found a significant difference of the relative areas only in the autologous parts, this finding is difficult to interpret. One can speculate that the entire airway wall composition in the rejection group changed due to an alteration of the perfusion and/or inflammation characteristics in that group. Tanabe et al 16,18 showed in a dog lung transplant model that bronchial mucosal blood flow in allogeneic parts of airways was reduced in acute graft rejection. These findings may explain the difference in the sonographically assessed composition of the airway wall in patients with graft rejection. The reason for the limitation of these changes to the autologous part is speculative and may be due to altered perfusion in the allogenic parts of airways. It happened in our study population that most of the patients without rejection had infection and, hence, these findings might have been biased by other than allogeneic mechanisms as well (see below). Additionally, it should be stressed that the interrelations of the inflammatory changes of the central airways and the alterations of the lung parenchyma of lung transplant CHEST / 129 / 2/ FEBRUARY,

6 Figure 3. Thickness of the five layers and the relative area of layers 1, 2, 3, and 5 in patients with infection (dots indicate BAL positive for any microorganism other than oral flora) vs patients without infection (triangles). Left panels show the values of the autologous airways (main bronchus); right panels show the values of the allogeneic airway (intermediate bronchus and left lower lobe bronchus). recipients are a matter of ongoing debate. In particular, the alloimmune origin of inflammatory changes seen in the central airways has not been fully clarified yet. We believe that EBUS might play a role in the future research of this topic. In patients with graft infection, we found a significant increase of the relative area of layer two (submucosa) of the autologous part and a trend to an enlarged layer three (inner marginal echo of the cartilage) of the allogeneic parts. Particularly, the latter is an interesting finding, since cartilage involvement in lung allograft inflammation is well known in animal models 19 ; however, due to difficult assessment in clinical routine, data concerning changes of the cartilage in human lung allografts are scarce. However, there is evidence 20 for involvement of the cartilage in inflammatory injury in human lung transplant recipients as well. Therefore, EBUS has the potential to serve as a reliable, noninvasive tool to further investigate this mesenchymal cell compartment. The underlying reason for the increase of the relative area of the second layer of the autologous airways in patients with graft infection might be the result of a relative edema probably due to inflammatory alterations. Since no particular absolute value showed a significant difference between patients with and without graft infection, rather the entire airway composition than one single layer differs between these groups of patients. This study is limited by its cross-sectional nature and the relatively small number of patients. In particular, the uneven distribution of infections between patients with and without rejection makes a quantitative comparison difficult. Since full histologic examination of the central airway walls in vivo is not possible, comparison between endonsonographic findings of the central bronchial structures and the results of TBBs were performed. The reso- 354 Original Research

7 Table 2 Correlation Between 20 Ultrasonic Evaluations of Allogeneic (Intermediate Bronchus or Lower Lobe Bronchus) or Autologous (Main Bronchus) Bronchi in 10 Lung Transplant Recipients* Layers Allogeneic Autologous r p Value Hyperechoic marginal echo 0.33 ( ) 0.29 ( ) Relative proportion 24.8 ( ) 25.7 ( ) Hypoechoic submucosal tissue 0.28 ( ) 0.26 ( ) Relative proportion 20.2 ( ) 19.4 ( ) Hyperechoic inner marginal echo of the cartilage 0.3 ( ) 0.25 ( ) Relative proportion 17.3 ( ) 21.3 ( ) Hypoechoic cartilage 0.92 ( ) 1.05 ( ) Hyperechoic outer marginal echo of the cartilage 0.35 ( ) 0.31 ( ) Relative proportion 33.1 ( ) 33.1 ( ) *Data are presented as median (interquartile range). Relative proportion indicates proportion of the layer after extraction of the hypoechoic cartilage. lution of the 20-MHz ultrasonic probe is limited, which explains why the borders of the particular layers appear blurred. A 30-MHz probe providing a higher image resolution 6 has been developed. We plan further investigations with this new equipment. In summary, for the first time EBUS was performed in lung transplant recipients undergoing routine surveillance bronchoscopy. We showed a distinct laminar airway structure in both autologous (main bronchus) and allogeneic (intermediate bronchus and proximal lower lobe bronchus, respectively) portions. Furthermore, we were able to quantitatively and consistently assess the five different layers on both positions. Statistically significant differences in the airway composition were found between patients with graft rejection and graft infection. These preliminary results show the following: (1) EBUS is a safe procedure in lung transplant recipients; (2) both allogenous and allogeneic airways sonographically consist of a multilayer structure; (3) by computerizing these images, these distinct layers can be quantitatively assessed; and (4) there might be differences in airway composition between patients with graft rejection and graft infection, respectively. EBUS therefore potentially may play a role in future surveillance of lung transplant recipients. References 1 Herth F, Becker HD. Endobronchial ultrasound of the airways and the mediastinum. Monaldi Arch Chest Dis 2000; 55: Falcone F, Fois F, Grosso D. Endobronchial ultrasound. Respiration 2003; 70: Baba M, Sekine Y, Suzuki M, et al. Correlation between endobronchial ultrasonography (EBUS) images and histologic findings in normal and tumor-invaded bronchial wall. Lung Cancer 2002; 35: Kurimoto N, Murayama M, Yoshioka S, et al. Assessment of usefulness of endobronchial ultrasonography in determination of depth of tracheobronchial tumor invasion. Chest 1999; 115: Miyazu Y, Miyazawa T, Kurimoto N, et al. Endobronchial ultrasonography in the assessment of centrally located earlystage lung cancer before photodynamic therapy. Am J Respir Crit Care Med 2002; 165: Nakamura Y, Endo C, Sato M, et al. A new technique for endobronchial ultrasonography and comparison of two ultrasonic probes. Chest 2004; 126: Tanaka F, Muro K, Yamasaki S, et al. Evaluation of tracheobronchial wall invasion using transbronchial ultrasonography (TBUS). Eur J Cardiothorac Surg 2000; 17: Herth F, Ernst A, Schulz M, et al. Endobronchial ultrasound reliably differentiates between airway infiltration and compression by tumor. Chest 2003; 123: Yousem SA, Berry GJ, Cagle PT, et al. Revision of the 1990 working formulation for the classification of pulmonary allograft rejection: Lung Rejection Study Group. J Heart Lung Transplant 1996; 15: Boehler A, Chamberlain D, Kesten S, et al. Lymphocytic airway infiltration as a precursor to fibrous obliteration in a rat model of bronchiolitis obliterans. Transplantation 1997; 64: Fournier M, Igual J, Groussard O, et al. Mucosal T-lymphocytes in central airways of lung transplant recipients. Am J Respir Crit Care Med 1995; 151: Snell GI, Ward C, Wilson JW, et al. Immunopathological changes in the airways of stable lung transplant recipients. Thorax 1997; 52: Ward C, Snell GI, Orsida B, et al. Airway versus transbronchial biopsy and BAL in lung transplant recipients: different but complementary. Eur Respir J 1997; 10: Yousem SA. Lymphocytic bronchitis/bronchiolitis in lung allograft recipients. Am J Surg Pathol 1993; 17: Zheng L, Walters EH, Ward C, et al. Airway neutrophilia in stable and bronchiolitis obliterans syndrome patients following lung transplantation. Thorax 2000; 55: Tanabe H, Takao M, Hiraiwa T, et al. New diagnostic method for pulmonary allograft rejection by measurement of bronchial mucosal blood flow. J Heart Lung Transplant 1991; 10: Estenne M, Maurer JR, Boehler A, et al. Bronchiolitis obliterans syndrome 2001: an update of the diagnostic criteria. J Heart Lung Transplant 2002; 21: Tanabe H, Yada I, Namikawa S, et al. Early detection of lung rejection by measurement of bronchial mucosal blood flow using laser Doppler flowmeter. Transplant Proc 1989; 21: Uusitalo MH, Salminen US, Ikonen TS, et al. Alloimmune injury preceding airway obliteration in porcine heterotopic lung implants: a histologic and immunohistologic study. Transplantation 1999; 68: Yousem SA, Dauber JH, Griffith BP. Bronchial cartilage alterations in lung transplantation. Chest 1990; 98: CHEST / 129 / 2/ FEBRUARY,