Year in Review 2015: Lung Cancer, Pleural Diseases, Respiratory Infections, Bronchiectasis and Tuberculosis, Bronchoscopic Intervention and Imaging
Pyng Lee,Chi Chiu Leung,Marcos I. Restrepo,Kazuhisa Takahashi,Yuanlin Song,José M. Porcel
DOI: https://doi.org/10.1111/resp.12779
IF: 6.9
2016-01-01
Respirology
Abstract:Kazuhisa Takahashi Approximately 1.5 million new lung cancer cases were diagnosed globally in 2014, and 80% were non-small cell lung cancer (NSCLC). Five-year survival for NSCLC remains dismal at 15% due to delayed diagnosis and ineffective treatment; however, if NSCLC is diagnosed early, the 5-year survival rate increases to 80%. Fong and van Zandwijk edited a series of reviews that focused on recent advances in diagnosis and treatment of lung cancer.1 Wang et al.2 reported a meta-analysis on micro-RNA (miRNA) as a non-invasive biomarker for NSCLC. Pooled sensitivity and specificity of miRNA for NSCLC were 0.75 and 0.79, respectively. Pooled positive likelihood ratio was 3.6, negative likelihood ratio was 0.32 and the diagnostic odds ratio was 12. The area under the curve (AUC) was 0.84. Interestingly, miRNA was more accurate for Caucasians than for Asians, and multiple miRNA assays were more accurate than single assay (AUC: 0.89, sensitivity: 0.83, specificity: 0.82 vs AUC: 0.81, sensitivity: 0.77 and specificity: 0.71). Serum miRNA assays were found to be more accurate than those from sputum, which led them to conclude that multiple serum miRNA assays might be a potential biomarker for NSCLC because of its relatively high diagnostic accuracy. Although fluoro-deoxy-glucose positron emission tomography/computed tomography (FDG-PET/CT) is widely used for staging of lung cancer, there are no accepted criteria for classifying thoracic nodes as malignant. Nguyen et al.3 proposed to use the maximum standardized uptake value (SUVmax) of the lymph node (SUVmaxLN), the ratio SUVmaxLN/SUVmax of the primary lung malignancy, the ratio SUVmaxLN/SUVaverage of the liver, the ratio SUVmaxLN/SUVmax of the liver and the ratio SUVmaxLN/SUVmax of the blood pool as criteria and validated this model by comparing against cytology results obtained from endobronchial ultrasound (EBUS) lymph node aspirations as well as expert visual interpretation by nuclear medicine physician. The investigators found that FDG-PET/CT-based evaluations of hilar and mediastinal lymph node metastases were useful and highly accurate but did not exceed expert visual interpretation. Implementation of evidence-based guidelines for lung cancer staging depends on expertise and resources, and FDG-PET/CT may not be available in developing countries. Some authors called for regional respiratory societies such as the Asian Pacific Society of Respirology (APSR) to take the lead in formulating minimum standards and encouraging cost-effective staging strategies such as incorporating EBUS, with national respiratory bodies.4, 5 Lau and coworkers also highlighted that not all clinically important diagnostic and staging technologies are accepted as standard of care in all Asian countries because of financial and reimbursement constraints. Several authors have stressed the importance of multidisciplinary team (MDT) approach to lung cancer management. Prabhakar et al.6 emphasized the role of the respiratory physician, while Campbell et al.7 reported an increase in lung cancer patients receiving radiotherapy when managed by MDT (66% vs 33%, P < 0.001). Kidane et al.8 debated about sublobar resection versus standard lobectomy for patients with stage I NSCLC and expounded on the role of surgery for T4 NSCLC with spinal invasion. Molecular genetics have led to the recognition of distinct subsets of NSCLC and targeted therapies. Routine tumour genotyping for epidermal growth factor receptor mutations and anaplastic lymphoma kinase rearrangement in stage 3 or 4 lung adenocarcinoma has relegated chemotherapy to second choice given only if there is no drug for driver oncogene. Obtaining tissue for epidermal growth factor receptor mutations and anaplastic lymphoma kinase rearrangement is an essential step towards personalized therapy for patients with advanced NSCLC. Comprehensive genomic study is able to detect multiple driver genes of which 75% are drug targetable within a short time, and new monoclonal antibodies that inhibit immune checkpoint pathways show promise in NSCLC.9 Nakanishi reported on the adoption of modern therapeutic approaches and research for NSCLC in Japan,10 which has witnessed more collaborative style clinical trials since the 1990s. José M Porcel Lymphangioleiomyomatosis and thoracic endometriosis are listed among the causes of secondary spontaneous pneumothorax. The former is a rare cystic lung disease affecting young and middle-aged women presenting with recurrent pneumothorax or chylothorax. In a Japanese database of 280 hospitalized patients with lymphangioleiomyomatosis, whose mean age was 41.7 years, 38% of hospitalizations were due to pneumothoraces, of which 30% were recurrent.11 Incidence of pneumothorax was 28%, chylothorax 4.6% and pleural effusion 2.5% over 33 months. Thoracic endometriosis can present as catamenial or non-catamenial pneumothorax, depending on its onset in relation to menstrual cycle. In a series of 150 women with thoracic endometriosis-related pneumothorax, 37% were catamenial, and all except one affected the right lung.12 Although all patients had endometrial deposits on the diaphragm, a significantly higher proportion in the catamenial group (62% vs 44%) also had pleural deposits, thereby suggesting that ectopic endometriosis may influence the timing of pneumothorax episodes. Lung cancer accounts for more than one-third of all malignant pleural effusions. In a comprehensive review of 556 consecutive lung cancer patients, it was found that about 40% developed pleural effusions during the course of the disease, of which half were too small to be tapped (minimal effusions) and did not progress so as to require a pleural intervention.13 The pleural effusions were malignant in 83% of 113 patients who were subjected to a diagnostic thoracentesis. Eighty-four percent of these malignant effusions needed palliative procedures, either therapeutic thoracenteses, pleurodesis or tunnelled pleural catheters. Survival was shortest among those with malignant pleural effusions (5.5 months) compared against those without pleural effusions (12.65 months) and minimal effusions (7.5 months). About 70% of cell blocks prepared from pleural fluid of 30 lung adenocarcinomas stained positive for thyroid transcription factor-1 and napsin A but negative in remaining non-lung cancer and benign effusions.14 Although there was no correlation between expression and concentration of these markers as measured by enzyme-linked immunosorbent assay, two (33%) of six double negative (thyroid transcription factor-1 and napsin A) stained lung adenocarcinomas had pleural fluid napsin A levels above 3.32 µg/mL, a 100% specific threshold for malignancy. Napsin A in the pleural fluid supernatants of patients with negative immunocytological examination warrants further evaluation. Mesothelioma is an aggressive malignancy with limited treatment options. Among targeted therapies are agents that inhibit mammalian target of rapamycin. Using a syngeneic murine mesothelioma model, temsirolimus (mammalian target of rapamycin inhibitor) curtailed mesothelioma growth by inducing tumour cell apoptosis, inhibiting tumour angiogenesis and modulating host immune response.15 Whether temsirolimus combined with standard chemotherapy (cisplatin with pemetrexed) would improve overall survival of mesothelioma patients has yet to be explored. Pleural fluid cultures identify pathogens in 60% of empyemas. Bacteriology differs between community-acquired and hospital-acquired pleural infections. Klebsiella oxytoca, which is primarily a pathogen in long-term care facilities and hospital settings, has been reported for the first time as a causative agent in 19 patients with empyema.16 The cases were nosocomial in nearly 90%, polymicrobial in 70% and two-thirds had a history of cancer. All isolates were resistant to amoxicillin/ampicillin.16 Obtaining a definitive diagnosis of tuberculous pleuritis may be challenging due to the paucibacillary nature of the disease and the limitations of conventional microbiological tests. Therefore, many pleural fluid biomarkers have been explored, and adenosine deaminase (ADA) is most robust. The diagnostic value of cytokines, chemokines and effector molecules of T-lymphocytes was assessed in 95 patients with lymphocyte-predominant exudates (35 tuberculosis (TB), 46 malignant and 14 others).17 Interferon-g ≥75 pg/mL and ADA ≥40 IU/mL were found to be highly discriminatory, with AUC of 0.905 and 0.855, respectively. Adding decoy receptor-3 and soluble tumour necrosis factor receptor-1 improved the surprisingly low sensitivities of interferon-g (66%) and ADA (40%) in isolation to 83%, at the expense of 10% reduction in specificity. Nevertheless, testing these four biomarkers simultaneously is not justified, based on the reported success with ADA. In systemic lupus erythematosus, a disease charaterized by multiorgan involvement, the prevalence of pulmonary manifestations was found to be 33% among 184 patients from Western Saudi Arabia.18 The most common was pleural effusion that affected 50% of cases and was bilateral in 60%. Although lupus pleuritis accounted for more than half of pleural effusions, TB, pneumonia and transudates should be considered, especially when the disease is not active.19 Solitary fibrous tumours of the pleura (SFTP) are rare neoplasms that arise from the mesenchymal cells beneath the mesothelium. A retrospective study of 80 patients who underwent surgical resection for SFTP showed that 59% had respiratory complaints, constitutional symptoms and/or hypertrophic pulmonary osteoarthropathy.20 About 80% arose from visceral pleura, 83% were pedunculated and 65% were classified as benign. Twenty percent of SFTP had pleural effusions, more common encountered in malignant than benign SFTP (40% vs 6%). Ten-year survival for benign and malignant SFTP was 88% and 80%, respectively. Surgical resection with clear margins was achieved in 93%, and recurrence was 7%. Marcos I Restrepo, Chi Chiu Leung, Yuanlin Song Three publications addressed the importance of identifying a pathogen in different populations of patients with pneumonia. Shi et al.21 performed a systematic review of the literature to determine the incidence of viral infection in childhood community-acquired pneumonia (CAP) using polymerase chain reaction or real-time polymerase chain reaction techniques. The authors found that respiratory viruses are widely detected in childhood CAP, but the incidence of viral infection is affected by significant heterogeneity among studies. The three most common viruses causing CAP in childhood were rhinovirus, respiratory syncytial virus and bocavirus. In addition, more than half of viral infections are probably concurrent with bacterial infections. These data suggest that there is a need to standardize the techniques to collect samples for respiratory viral pathogens in order to appropriately determine the incidence of these pathogens in children with CAP. Etherton-Beer et al.22 examined the prevalence of resistant Gram-negative bacteria in the oropharynx of 124 residents of 11 long-term care facilities. The authors found only one extended-spectrum b-lactamase producing Escherichia coli isolate representing 0.8% of multi-drug-resistant pathogens in this population. From the data, it suggested that empiric coverage against multi-drug-resistant bacteria might not be warranted in long-term care facility residents. Controversy surrounds the role of bronchoscopy in the identification of pathogens and its impact on clinical outcomes of mechanically ventilated patients. Patients who had an early bronchoscopy within 24 h after intubation had lower intensive care unit (ICU) and 90-day mortality regardless of the initial empiric antibiotic therapy. In addition, patients who survived the event and had an early bronchoscopy were more likely to have shorter duration of mechanical ventilation and length of ICU stay, respectively. The data suggest that bronchoscopic evaluation might improve clinical outcomes among patients with aspiration pneumonia that require mechanical ventilation.23 Previous studies suggested that chronic obstructive pulmonary disease (COPD) led to worse clinical outcome when they developed CAP. Jiang et al.24 performed a systematic review that included cohort and case–control studies composing of 257 958 patients and concluded that COPD was not a risk factor for increased mortality, longer length of hospital stay, admission to the ICU and need for mechanical ventilation in patients who developed CAP. The study by Sibila and colleagues 25 suggested that patients with Pseudomonas aeruginosa CAP had lower 30-day mortality when empiric antibiotics that had antipseudomonas activity were initiated within 48 h of presentation. The authors concluded that there was a need to better characterize risk factors associated with P. aeruginosa CAP in order to institute appropriate antimicrobial therapy to improve their clinical outcomes. Sanz et al.26 hypothesized that partial oxygen concentration/inspired oxygen fraction (PaO2/FiO2) from percutaneous oxygen saturation (SpO2) by the Ellis or Rice equations might correlate with PaO2/FiO2 determined by arterial blood gases. They concluded that the Ellis equation was more accurate than the Rice equation for estimating PaO2/FiO2, especially at higher levels of PaO2/FiO2 ratio, and accurate enough to assess initial oxygenation in patients with CAP. The authors recommended arterial blood gas analysis of oxygenation at the threshold for mechanical ventilation because there was misclassification using Ellis equation among those who had higher PaO2/FiO2 levels. Smith et al.27 performed a systematic review of non-pharmaceutical interventions to prevent transmission of influenza virus in adults and found seven articles with important design flaws. Diligent handwashing and dental hygiene were useful, but other interventions were not fully assessed. Marras and colleagues28 used a Delphi method to survey North American experts regarding clinical aspects related to the care of patients with non-tuberculous mycobacteria (NTM). The survey yielded extensive variability and underscored the need for more research. The editorial29 opined that the survey raised relevant questions regarding care of these patients and served as motivation for research into understanding of NTM disease, its evolution, progression, treatment responsiveness and prognosis. A retrospective cohort study showed that patients with pulmonary NTM had obstructive ventilatory defects and after 12 months of antimycobacterial therapy, there was no remarkable improvement in the pulmonary function tests. Small improvements were observed among patients with asthma, but these findings need to be evaluated further due to small numbers.30 Numata and colleagues31 evaluated 296 patients with suspected Mycobacterium avium complex pulmonary disease and found that anti-glycopeptidolipid-core IgA antibody test had diagnostic sensitivity of 59% and specificity of 98%. Malignancy was associated with false negative results, and the authors concluded that anti-glycopeptidolipid-core IgA antibody should be applied with caution in these patients. Bishwakarma et al. reviewed the epidemiological link between TB and cigarette/biomass smoke exposure.32 While a fairly large number of studies significantly associate both direct and second-hand smoke exposure with tuberculous infection, disease and/or mortality, fewer link biomass fuel smoke exposure and TB. These observational studies, as usual, might not be able to control all confounders and fell short of establishing a direct causal link. In Hong Kong, smoking was also found to adversely affect baseline disease severity, bacteriological response, treatment outcome and relapse in TB in the setting of a territory-wide TB control programme.33 Complex sociobiological interactions are likely involved at multiple stages of the pathogenesis. Further mechanistic studies are desirable to elucidate the possible links, and population-based randomized controlled trials may also be needed to provide convincing evidences for full integration of tobacco control and smoking cessation into national TB control programmes. Over 50% of pulmonary TB patients undergoing bronchoscopy were found to have endobronchial involvement in a Korean tertiary referral centre, with female gender, symptom duration >4 weeks and absence of previous TB as independent predictors. More than 20% had severe stenosis at diagnosis. Longer stricture and decreased forced expiratory volume in 1 s were risk factors for persistent bronchostenosis. Taking aside selective factors that could have contributed to the high observed prevalence of endobronchial TB, results suggest a need for early diagnostic bronchoscopy in selected patients with active pulmonary TB.34 Pulmonary TB accounted for 1.2% ICU admissions in Taiwan, and nearly two-thirds was made 3 months after discharge, indicating unrecognized nosocomial transmission risk.35 The nosocomial transmissible period was inversely related to male gender, prior airway symptoms and performing nucleic acid amplification tests and mycobacterial culture. Despite directly observed therapy short course, molecular clustering was observed among isolates of extensively drug-resistant TB in Hong Kong with most of transmission occurring apparently outside the household settings.36 Rapid diagnostic kits capable of detecting rifampin resistance would be useful to inform infection control in high-risk environments. Effective treatment of latent TB infection appears pivotal if we are to eliminate TB by 2050.37 In a prospective study among 1049 asymptomatic household contacts of smear-positive TB patients in Hong Kong, T-Spot.TB outperformed tuberculin skin test (TST) in predicting TB over the subsequent 5 years irrespective of TST cutoffs.38 The more specific TB antigens in T-Spot.TB probably minimized interference by widespread prior Bacillus Calmette-Guérin (BCG) vaccination in that locality. Low concordance between TST and QuantiFERON-TB Gold In-tube (QFT-IT) was observed.39 Smoking was an independent predictor for TST-reactive and QFT-IT-negative subjects, while age >52 years and female gender were predictors for TST non-reactive, QFT-IT-positive subjects.39 Neither TST nor interferon gamma release assays could accurately predict TB in immunocompromised patients.40 The duration of protection offered by 6 months of isoniazid in people living with HIV appears to depend on the ongoing risk of infection.41, 42 More research is needed to overcome limitations posed by available diagnostic and treatment tools for this highly prevalent but elusive infection.43 Management guidelines for bronchiectasis are lacking which can contribute to poor outcomes observed with this orphan disease although it is increasingly recognized in COPD patients.44, 45 Identifying risk factors may allow intervention and prevention. From two studies, most bronchiectasis was idiopathic, although post-infectious/post-TB accounted for about 20% and deficiency in the immune system about 9%.46, 47 Bacteria colonization plays an important role in the pathogenesis of bronchiectasis. A study investigated the role of secreted mucins (MUC2, MUC5AC and MUC5B) in sputum of patients with and without bacteria colonization. MUC2 levels were higher in those whose sputum culture yielded P. aeruginosa versus those without bacterial colonization, which might allude to potential biding effects of P. aeruginosa with MUC2.48 Yamasaki and coworkers analysed bacteria flora of bronchoalveolar lavage fluid obtained from patients with NTM-positive bronchiectasis and found anaerobes when compared against those without NTM. CT evidence of collapse/consolidation was significantly correlated with isolation of Prevotella species in the bronchoalveolar lavage fluid samples using 16sRNA detecting assay.49 NTM was also frequently isolated from the sputum of patients with diffuse panbronchiolitis, a disease associated with bronchiectasis.50 These studies suggest an etiological role of NTM in the pathogenesis of bronchiectasis, which might be exacerbated by co-infection by anaerobes that cause lung parenchyma changes on CT. Further exploration is required before therapy targeting NTM, and anaerobes can be recommended in clinical practice. Chronic obstructive pulmonary disease patients who developed bronchiectasis experienced higher mortality compared with COPD alone, which suggested that chronic airway inflammation left unabated led to airway remodelling and bronchiectasis from recurrent infection.51 A study comparing sputum matrix metalloproteinases (MMP) and tissue inhibitor metalloproteinase-1 found elevated MMP-8 and MMP-9 during exacerbation of bronchiectasis. MMP-8 and MMP-9 could be potential biomarkers, and MMP inhibitors might provide protective effects against persistent airway inflammation, remodelling and destruction.52 Hiatal hernia and gastro-oesophageal reflux were proportionally related to the severity and extent of bronchiectasis as well as forced expiratory volume in 1 s (% predicted) impairment in non-cystic fibrosis patients.53 Short-acting and long-acting bronchodilators have been prescribed for patients with spirometric evidence of ventilatory defect; however, the recommendation is not high given limited clinical trials. Although bronchodilators can relieve symptoms, use of short-acting β2 agonists (SABA) and anticholinergics (not inhaled corticosteroids (ICS)/long-acting β2 agonists (LABA) and LABA) had been shown to increase frequency of haemoptysis and must be carefully monitored.54 Pyng Lee In the final editorial for the invited review series ‘Update in Interventional Pulmonology’, Maldonado opined that as the interventional pulmonology community continues to embrace new technologies and techniques, there is an obvious need for multidisciplinary approaches for the diagnosis and treatment of pulmonary diseases that require interactions among experts of different specialties. From an educational standpoint, patients should not carry the burden of procedural training, which dispels the old adage of ‘see one, do one, teach one’ and encourages evidence-based practices such as flipped-classroom models, problem-based learning and other learner-centric interventions.55 Stather and coworkers studied the learning curves for endobronchial ultrasound-transbronchial needle aspiration (EBUS-TBNA) among fellows with >100 flexible bronchoscopy procedures but <75 clinical EBUS-TBNA encounters and <2 h with EBUS-TBNA computer simulator. All learners completed a questionnaire, underwent EBUS-TBNA knowledge review and simulator followed by evaluation of their baseline skills with two test cases. They were assessed after every 25 clinical cases with the same two test cases. There was significant variation in EBUS-TBNA learning curves of interventional pulmonology fellows, with improvement in EBUS-TBNA even after 200 clinical cases. The study can guide individualized performance assessments to ensure adequate procedures necessary for trainees' competency and proficiency.56 A systematic review and meta-analysis were undertaken to evaluate diagnostic yield of EBUS-TBNA for sarcoidosis in patients with lymphadenopathy detected on imaging. In 14 studies of 2097 patients with a median prevalence of 15%, the pooled diagnostic yield of EBUS-TBNA was 0.79, sensitivity 0.84 and specificity 1.0.57 Consecutive 135 EBUS lymph node images were evaluated for greyscale texture (pixels and entropy) using MATLAB software (MathWorks, Inc., Natick, MA, USA).58 Receiver-operating characteristic curves were generated for 94 patients and thresholds applied to subsequent 41 subjects. Entropy threshold phenotyped only 47% benign lymph nodes and 20% malignant cases leading to authors' conclusion that EBUS greyscale textural analysis was not accurate in differentiating malignant from benign lymphadenopathy in contrast to previous study.59 Another study investigated the greyscale texture of endobronchial ultrasound mini-probe images of pulmonary lesions in an attempt to find differences that discriminate benign from malignant disease. There were 85 pulmonary lesions in the prediction set and 82 in the validation set. Two physicians correctly classified 75% of pulmonary lesions concluding that greyscale texture analysis of EBUS mini-probe images may aid in aetiology with a high negative predictive value for malignancy.60 Probe-based confocal laser endomicroscopy (pCLE) is a new technology that miniaturizes the microscope to allow real-time optical imaging. Wellikoff and coworkers compared pCLE images of central and peripheral pulmonary lesions obtained during bronchoscopy against histology of biopsies of the same areas. pCLE images of NSCLC consisted of mottled elastin, septal studding and disorganization/fragmentation with increased friability, which also correlated with cancer differentiation. These features seen during pCLE could suggest malignant aetiology.61 Bronchoscopic lung volume reduction coil treatment was shown to be safe and effective for patients with severe emphysema in the short term. Hartman and colleagues reported 3-year data of 38 patients who underwent lung volume reduction coils. No coil migrations or pneumothoraces were observed on chest X-ray. At 1-year follow-up, spirometry, 6-min walk distance, modified Medical Research Council (mMRC) and the St George's Respiratory Questionnaire (SGRQ) scores improved over baseline. At 2 years, improvements in residual volume, mMRC and SGRQ persisted, and at 3 years, around 50% of the patients maintained improvements in 6-min walk distance, SGRQ and mMRC.62 Flexi-rigid pleuroscopy is a useful tool for diagnostic evaluation of pleural effusion, but biopsy with flexible forceps (FFB) can be challenging when the parietal pleura is thickened and fibrotic. There was no significant bleeding recorded with forceps (FFB) or cryo-biopsy (CB) in all 22 patients. FFB and CB successfully obtained pleural tissue suitable for histopathological analysis. CB was larger than FFB (7–15.8 mm vs 4.3–8 mm) with better preserved cellular architecture and tissue integrity and less crush artefacts (CB 2/22 vs FFB 21/22). Although there was no improvement in diagnostic yield when CB was combined with FFB, CB was shown to be safe, and its routine use would require further evaluation.63