| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Yale University School of Medicine, New Haven, Connecticut
DIAGNOSTIC IMAGING
Theresa McLoud
Professor of Medicine
Harvard University School of Medicine
Department of Radiology
Boston Massachusetts
The relative benefits and cost effectiveness of computerized tomography (CT) screening for lung cancer remain controversial (1). Swensen and coworkers (2) assessed their own five-year prospective experience with CT lung cancer screening at the Mayo Clinic in this study of 1,520 subjects. Subjects who were screened had a minimum of 20 pack-years of tobacco exposure and were 50 years of age or older. After a five-year period, 3,356 noncalcified lung nodules were identified, with 74% of subjects displaying one or more nodules (1,118 participants). Sixty-eight lung cancers were diagnosed in 66 of the participants with 31 discovered within the initial year of scanning (prevalence cases), 34 on subsequent years of scanning (incidence cases), and 3 discovered between annual screening CTs (interval cases). Twenty-eight of the 34 incidence cases were non–small cell in histology. Of these, 17 were stage I tumors. A lung cancer mortality rate of 1.6 per 1,000 person-years was noted in the incidence portion of the trial (between Years 2 and 5). Overall there was no statistically significant difference in the lung cancer mortality rates after subsequent examinations compared to the earlier Mayo Lung Project (2.5 vs. 2.0 per 1,000 person-years; p = 0.43). The authors concluded that CT screening does allow detection of earlier state lung cancers but that the rate of lung nodule detection was extremely high.
The determination of nodules on CT scan by radiologists carries with it some subjective inter-reader variability. Rubin and his colleagues (3) sought to explore the use of computer-aided detection (CAD) in assisting the radiologist in more accurately "calling" nodules. In this study of 20 multi-detector lung CT scans, three radiologists recorded the locus of each nodule candidate they saw independently. The also assigned a confidence score for each such nodule recorded. Next, a CAD algorithm was applied to the same 20 CT scans with predetermined parameters chosen by leave-one-out cross validation. A reference standard was established by two experienced thoracic radiologists in consensus, and double reading performance defined a "true positive" by either reader. Receiver operating characteristic (ROC) plots were generated from true and false positives and confidence levels accordingly. One hundred ninety-five noncalcified nodules 3 mm or greater in size were determined. Mean sensitivity for individual readers was 50%, and when double reading was applied this increased to 63%. CAD algorithm allowed a mean sensitivity increase to 76%. The authors conclude that CAD algorithms compliment individual readers by detecting individual nodules more effectively than did a second reader.
The use of thin section CT to narrow clinician's differential diagnosis in cases of interstitial lung disease (ILD) was examined in a 2006 study by Aziz and colleagues (4). One hundred sixty-eight patients suspected of having ILD were separately assessed by six pulmonologists. They used patient data sheets containing symptom and exam findings along with pulmonary function test (PFT) data and plain chest radiographs to assimilate differential diagnoses and offer first choice diagnoses. Thin section CTs were then offered and the clinicians were again asked to offer first choice diagnoses and differential diagnoses. The first choice diagnosis changed in 51% of 1,008 cases, and agreement on first choice diagnosis increased from 0.42 to 0.72 after thin section CT data was introduced. In addition, confidence in the first choice diagnosis increased especially with usual interstitial pneumonitis (UIP) or idiopathic pulmonary fibrosis (IPF) determinations. Thoracoscopic lung biopsy requests decreased from 26.8% to 11.2% with the addition of thin section CT.
The prospective use of multi-detector CT to diagnose pulmonary embolism (PE) has achieved widened acceptance in recent years. However, incidental PE detection occurs in patients undergoing CT for other indications with unclear frequency. Storto and coworkers (5) assessed their own experience with incidental PE detection in 589 patients retrospectively. None of these studies were done with suspicion of PE. Overall 3.4% (20) of the study population had CT evidence of PE (4% inpatients and 0.9% of the outpatients). Seventy percent of these patients had malignancy as an underlying diagnosis. The authors conclude that routine multi-detector CT scanning uncovered PE unexpectedly in a significant number of patients, especially in those with underlying cancers.
ASTHMA
Sally Wenzel
Professor of Medicine
National Jewish Medical and Research Center
Division of Pulmonary and Critical Care
Denver Colorado
While imbalances of Th2 lymphocyte populations clearly play a role in asthma development, increasing recognition that innate immune response to viral infection may also contribute motivated an interesting study by Wark and colleagues (6). These investigators obtained primary bronchial epithelial cells from individuals with asthma (with and without inhaled corticosteroid [ICS] use) as well as normal subjects and grew them (x3 passages) in vitro. After the third passage, these cells were exposed to rhinovirus and then monitored for virus persistence as well as immune response to the viral infection. Viral messenger ribonucleic acid (mRNA) was not cleared as efficiently in asthmatic cells as it was by "normal" subject cells. Greater epithelial cell killing by virus along with decreased epithelial cell apoptosis was also noted in asthmatic cells compared with normal cells. Deficiency in interferon beta (IFN-ß) was clearly demonstrated in the asthmatic cell populations, and replacement of IFN-ß reversed the aforementioned defects. Use of ICS had no effect on these measures. The authors conclude that a defect in innate immune response exists in subjects with asthma, limiting ability to clear viral infections.
Another area of emerging interest in the field of asthma is the use of biomarkers to define better phenotyping of the various syndromes that collectively constitute asthma. Smith and coworkers (7) examined the role of exhaled nitric oxide (eNO) measurement in the management of asthma. Ninety-seven subjects had their ICS modulated in a placebo-controlled single-blind fashion based on an algorithm developed by current guidelines versus by serial measurement of eNO. Inhaled steroid dose was adjusted over 1 yr (phase I) and then followed every 2 mo for an additional year (phase II). A threshold of 15 parts per billion (ppb) eNO was used as the threshold to increase ICS (at an exhaled flow rate of 250 ml/s). No statistically or clinically significant difference in total exacerbation rates were noted between the two groups. However, the eNO group achieved this rate of control at a lower mean ICS dose (50% lower). No difference in sputum eosinophilia was seen in either phase I or phase II.
In a related study, Smith and colleagues (8) examined 60 patients with respiratory symptoms not previously diagnosed as having asthma to see if eNO would predict steroid-responsive phenotypes. Diaries, peak flows, eNO, and pulmonary function tests (PFTs) were followed for 4 wk. Patients were then given ICS (fluticasone). Those with subsequent improvements in forced expiratory volume in the first second (FEV1) of 12% or more were considered a responder phenotype. Asthma was diagnosed by relevant history, > 12% response to ICS trial, or bronchial hyperresposiveness of 
20% (BHR). Overall, asthma by this definition was diagnosed in 27 of 52 subjects who completed the study. Fifteen of these had eNO levels in the highest tertile. These same subjects had the highest response to ICS. Exhaled NO was generally more sensitive and specific for identifying ICS responders than was FEV1% predicted, peak flow variability, or BHR. The authors conclude that eNO may have utility in defining phenoytypes of asthma that are likely to be ICS-responsive.
Another potential way to better understand asthma phenotypes is to assess the inflammatory cells in the sputum of individuals with asthma. Simpson and coworkers (9) studied 93 stable, nonsmoking, ICS-treated subjects with asthma. These patients were characterized as eosinophilic (> 1%), neutrophilic (> 61%), or paucigranulocytic (all others) on the basis of sputum cell counts. The authors found that total and active matrix metalloprotein 9 (MMP-9) were greatest in subjects with eosinophilic asthma, although subjects with neutrophilic asthma had high levels of pro (but not active) MMP-9. In contrast to this, active neutrophil elastase was highest in the subjects with neutrophilic asthma, along with interleukin-8 (IL-8). As such, sputum cellular profiling may help identify phenotypes of asthma with different underlying pathobiologies. The authors caution that further studies are needed to determine whether these cell-defined populations can predict differing long-term clinical outcomes or responses to therapy.
While long-acting ß-agonists (LABA) have been shown to have a multitude of positive effects, concern has been raised that they may also contribute to negative outcomes (including death) in subpopulations. In the salmeterol multicenter asthma research trial (SMART), Nelson and collegues (10) examined data from 26,355 subjects enrolled and randomized to salmeterol or placebo and followed for 28 wk for asthma-related deaths or life-threatening exacerbations. The study was terminated prematurely, as mid-term analysis suggested that the use of salmeterol was associated with higher numbers of asthma-related deaths or life-threatening exacerbations. Further analysis revealed that this was predominantly seen in African Americans. There was a suggestion of ICS being protective in whites, but conclusions were difficult to draw regarding the role of ICS in African Americans. In general, African Americans in the study had a higher rate of urgent health care use and a lower rate of ICS use reported. The study concluded that while LABA in combination with ICS continue to be a useful strategy in the majority of subjects, a small percentage of patients may be posed with the threat of worsening and even life-threatening asthma. Whether such populational differences in response are due to differences in genetic polymorphisms could not be addressed by this study.
Targeted anti-inflammatory therapy for asthma may allow for more specific disease-modifying agents than the traditional approach with ICS. Earlier observations that tumor necrosis factor alpha (TNF-
) is increased in individuals with severe asthma (11) prompted Berry and colleagues (12) to study the effect of blocking TNF- with entanercept in a double-blind, placebo-controlled, crossover study of 10 subjects. Disease-related outcomes examined included FEV1 and BHR. Exhaled NO, quality-of-life, and sputum inflammatory markers were also studied. Entanercept was added to "usual care" in patients and compared to "usual care" with patients serving as their own controls in the crossover phase of the study. The investigators found significant improvements in FEV1 and BHR (2.5-fold) as well as asthma quality-of-life measures. Sputum inflammatory cells and eNO were unchanged, but sputum histamine levels showed significant reductions in the entanercept-treated group. The authors conclude that TNF-–blocking strategies may offer a new therapeutic approach in asthma but caution that larger trials are warranted to determine if "disease modifying" efforts long-term can truly be achieved.
LUNG TRANSPLANTATION
Robert M. Kotloff
Professor of Medicine
University of Pennsylvania Medical Center
Division of Pulmonary Allergy and Critical Care
Philadelphia Pennsylvania
The field of lung transplantation has been hampered by insufficient organs to transplant those in need. Gamez and colleagues (13) reported on their initial experience with non–heart-beating donors (NHBD) with ischemic times ranging from 9 to 11 h. Five procedures were performed. All recipients survived to discharge and were alive from 2 to 13 mo after the procedure. This study is a proof-of-concept demonstration of the feasibility of using NHBD as a potential new source of lung allografts.
Others have suggested that lobar transplantation rather than conventional bilateral cadeveric grafts may expand the number of patients potentially benefiting from transplantation. Bowdish and colleagues (14) compared outcomes of 59 living-donor bilateral lobar transplant recipients to 43 recipients of standard bilateral cadaveric grafts. At 3 mo mortality was significantly higher in the lobar recipients (20% vs. 6.5%; p = 0.009), but scrutiny suggests that this was likely due to increased acuity of illness in the lobar recipients, with 73% being hospitalized at the time of transplantation versus 2% of the conventional group. After the 3-mo mark, survivals were similar between the two groups (5-yr survival: 75% in cadeveric group vs. 62% in lobar group; p = 0.32), as were pulmonary functions and maximum exercise capacities.
Since organ shortage will like remain an issue clarification of double- versus single-lung transplantation strategies is of great importance. Meyer and coworkers (15) examined the UNOS national database, comparing survival of patients undergoing either single-lung (n = 636) or double-lung (n = 185) transplantations for pulmonary fibrosis from the period 1994 to 2000. Single-lung transplantation afforded superior survival in patients younger that 60 yr of age, mostly due to increased early mortality in the bilateral group. When survival statistics for patients surviving to 3 mo or longer were examined, no survival difference was seen between single- and double-lung transplantation. No difference was seen in patients aged greater than 60 yr, but the small number of bilateral transplantations in this age group limit meaningful conclusions in this subset. Pulmonary artery diastolic pressure was also an independent predictor of survival in this study.
Examination of the UNOS/ISHLT registry also allowed Christie and colleagues (16) to assess the incidence of primary graft dysfunction (PGD), a dreaded early complication of lung transplantation. An overall incidence of 10.2% was noted in 5,262 patients. A 30-d mortality rate of 42% with PGD versus only 6% of those without PGD was observed. Patients with PGD had significantly worse survival beyond the first year as well (hazards ratio,1.35; confidence interval, 1.07–1.70). The risk of subsequent death even beyond the first year in those experiencing PGD suggests lingering consequences even beyond the acute event.
Avoiding rejection remains the number one challenge in lung transplantation. Two contributions to the literature were offered this year in this arena. Iacono and coworkers (17) studied 58 patients in a single-center randomized, double-blind placebo-controlled trial of inhaled cyclosporine versus saline over a 2-yr period. No difference in acute rejection (the primary endpoint) was observed; however, use of inhaled cyclosporine was associated with improved overall survival and with improved chronic rejection-free survival. Brochiolitis obliterans (BO) rates were reduced by nearly 50% at 3 yr after transplant in cyclosporine compared with controls. In another study examining the development of BO after transplantation, Yates and colleagues (18) tested the hypothesis that low-dose azithromycin could impact on the course of BO in 20 transplant recipients. At the time of azithromycin administration (250 mg 3 times weekly) the breakdown of BO stages (BOS) were 2 with BOS0p, 6 with BOS 1, 2 with BOS 2, and 18 in BOS 3. At the 3-mo time point, mean FEV1 increased by 110 ml (14%; –70 to 730 ml range). Twelve of 17 patients who initially responded had sustained benefits beyond 3 mo. Although limited in size, this study raises the possibility that macrolide antibiotics may exert benefits beyond their antimicrobial effects that may mitigate BO development. The authors note that larger studies are warranted.
INTERVENTIONAL PULMONOLOGY
Udaya B. S. Prakash
Scripps Professor of Medicine
Mayo Clinic College of Medicine
Pulmonary and Critical Care
Mayo Medical Center
Rochester Minnesota
The use of bronchoscopic mediastinal lymph node needle aspiration (BNA) to stage non–small cell lung cancer has become an accepted technique in centers worldwide. Overall positive rates of diagnostic accuracy vary widely in numerous studies, but are about 75% in most. Holty and coworkers (19) performed a meta-analysis of previously presented studies to clarify our understanding of this technique. A literature search identified 525 potential studies; 398 were deemed not relevant after careful review by the authors. Of the 127 remaining articles, a review of their bibliographies identified an additional 203 studies of potential relevance. A preliminary review of these 330 articles eliminated 268, leaving 62 for detailed analysis. In the end, 13 studies enrolling at least 10 patients with and/or without mediastinal metastasis (range, 10–183) were included. Six of them provided surgical confirmation of the BNA result. Studies were graded as tier 1 if they enrolled at least 10 subjects with and 10 subjects without mediastinal lymph node involvement, surgically confirmed all BNA results, and used the patient as a unit of analysis. All others were considered non–tier 1. In tier 1 studies the median prevalence of mediastinal metastasis was 34%. Pooled sensitivity and specificity were 39% and 99%, respectively. Compared with tier 1 studies, the median prevalence of mediastinal metastases were higher (81%; p = 0.002) and pooled sensitivity higher (78%; p = 0.0009) in non–tier 1 studies. Rates of major complications overall were low (0.26%) but included two major bleeds, one pneumothorax requiring a chest tube, and two pneumothoraces that spontaneously resolved. The authors conclude that BNA is highly specific but that sensitivity depends greatly on study methods and the population being studied. Lower prevalence of mediastinal metastasis is associated with markedly lower sensitivity.
The application of thermal energy to treat asthma is a novel application explored by Miller and coworkers (20). Nine patients scheduled to undergo lung resection for suspected or proven cancer underwent bronchothermoplasty (BT) during routine preoperative bronchoscopy up to 3 wk before resection. BT was applied to areas of segmental bronchi within the lobe to be resected. Twelve minutes or less were required to perform the BT, with a total of 12 activations at 55°C performed in two patients and a total of 41 activations at 65°C performed in six. The number of activations ranged from three to nine per subject. Treated airways were inspected bronchoscopically at the time of resection and histologically after resection in all patients. No new symptoms or unscheduled physician visits were observed in the interval from BT to resection. Treated sites exhibited slight erythema and edema for up to 2 wk but appeared normal at later time points. Narrowing of 25 to 50% (visually estimated) was seen in four airways of two subjects. No bronchosopic evidence of scarring was seen in any airways. Airway smooth muscle reduction was seen on histology limited to the treated airways. The authors conclude that this proof-of-concept study suggests BT was well tolerated and resulted in significant reduction of smooth muscle mass in the treated airways. More extensive experience was offered by Cox and colleagues (21). In this study out of the same center, 16 patients with mild to moderate asthma underwent BT. Baseline and 12-wk post-treatment measures of spirometry, methacholine challenge, and daily diary recordings of peak flow, symptoms, and medication use were recorded. The authors found that all patients demonstrated improvement in airway responsiveness after BT. The mean PC20 increased from 2.37 ± 1.72 (p < 0.001), to 2.77 ± 1.53 (p < 0.007) and 2.64 ± 1.52 (p < 0.001) at 12 wk, 1 yr, and 2 yr after procedure, respectively. Symptom-free days, morning peak flow, and evening peak flow also significantly improved (p = 0.015, 0.01, and 0.007, respectively). The procedure was well tolerated in all subjects.
The use of interventional bronchoscopy to address bronchopleural fistulas (BPF) was the topic of work by Lois and coworkers (22) this past year. BPF incidence after lung resection varies from 1.5% to 28% and is more likely after lung surgery for malignancy. Pneumonectomy results in BPF in anywhere from 4.5 to 20%, while after lobectomy it is distinctly unusual. Right-sided resections carry the highest risk of this complication. Preoperative risk factors include fever, steroid use, Hemophilus influenza colonization in the sputum, elevated sedimentation rates, and anemia. Flexible bronchoscopy is the best technique to visualize the BPF track after pneumonectomy or lobectomy. Smaller BPFs may require thin guidewire probing or methylene blue instillation. Fiberoptic techniques used to repair BPFs include tissue glue instillation, ethanol, silver nitrate, cryanoacrylate compounds, coils, lead plugs, gel foam, spigots, and autologous blood patching, which often requires multiple bronchoscopic sessions. Size of the BPF is an important factor in predicting the success of bronchoscopic interventions, with those more than 8 mm not suitable to be managed this way, while those less than 1 mm in size have the greatest chance of successful closure.
Haussinger and colleagues (23) performed a prospective randomized multicenter trial in Europe from 1999 to 2003 designed to evaluate the prevalence of moderate to severe dysplasia in smokers using flexible bronchoscopy with either traditional white light (WB) or autofluorescence bronchoscopy (AB). Smokers more than 40 yr old and with over 20 pack-years of tobacco abuse were stratified into one of four risk groups and investigated with either WB + AB (arm A) or WB alone (arm B). Five hundred eighty-nine patients were recruited to the WB + AB group, while 584 patients were recruited to the WB alone arm. Mean age of the 1,173 patients was 58; 916 of the patients were men. Overall (both arms A and B) preinvasive lesions (dysplasia grades II–III and carcinoma in situ [CIS]) were noted in 3.9%. Preinvasive lesions were noted in 2.7% of the WB arm and 5.1% of the WB+AB arm (p = 0.037). When dissected out further, for dysplasia grades II–III WB+AB increased detection by a factor of 2.1 (p = 0.03), while only a factor of 1.24 (p = 0.70) was seen for CIS lesions. Biopsy-based sensitivity for detection of dysplasia grades II and III and CIS was 57.9% with WB alone and 82.3% with WB +AB, with corresponding specificities of 62.1% and 58.4%, respectively.
Finally, Mehta and coworkers (24) and colleagues underscored the concerning variability seen in compliance with flexible bronchoscope reprocessing recommendations between uses to minimize bronchoscopic-associated infections (BAI). Although the exact incidence of BAI is unknown, in recent years it is estimated that more than 800 patients may have been potentially affected. A survey of bronchoscopists in the United Kingdom revealed that national guidelines were not followed consistently, that minimum disinfection times were achieved only 35% of the time, and that in 34% of "emergent" bronchoscopies no disinfection was performed at all. A survey of bronchoscopists in the United States revealed that two-thirds of the respondents acknowledged that they were unfamiliar with the national reprocessing guidelines. As such, this consensus statement underscores the extent of the problem and identifies specific risk factors inherent in the problem.
FOOTNOTES
Conflict of Interest Statement: T.K.T. does not have a financial relationship with a commercial entity that has an interest in the subject of this manuscript.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
