Year in review 2014: Lung cancer, pleural diseases, respiratory infections and tuberculosis, bronchoscopic intervention and imaging.

Abstract

This section reviews papers published in Respirology in 2014 regarding prevention, novel diagnostic techniques, translational research and treatments for lung cancer. Other important developments in this field are also addressed. Cannabis is the most commonly used illicit substance worldwide.1 Although cannabis is known to contain harmful and carcinogenic substances, little is understood with regard to its effects on respiratory health, such as the potential for chronic obstructive pulmonary disease (COPD) and lung cancer. Gates et al.2 reported the results of an excellent review regarding cannabis smoking and respiratory health. Based on the previous literature, chronic cannabis use is associated with an increased prevalence of symptoms of chronic bronchitis.3 In contrast, the relationship between cannabis and lung cancer is contentious, although some papers have suggested a positive association. What about electronic cigarettes (e-cigarettes)? Are they safe or potentially harmful? E-cigarettes or electronic nicotine delivery system (ENDS) devices deliver a vaporized liquid to the lungs that usually contains nicotine and other chemicals. While some studies have shown that individuals who do not intend to quit smoking may reduce their intake of combustible cigarettes by smoking e-cigarettes, other studies have failed to show the superiority of this approach over the use of nicotine replacement medicine in individuals trying to quit smoking.4 Therefore, the potential benefits of ENDS are not well proven. In contrast, the inhalation of vapor from e-cigarette cartridges has been shown to enhance inflammatory changes in mouse airways, and the short-term use of e-cigarettes has been shown to adversely affect the lung function. Therefore, e-cigarettes are considered to be potentially harmful. Lam D et al.4 mentioned, in an editorial published in Respirology, that the Asian Pacific Society of Respirology (APSR) fully supports the joint statement by the Forum of International Respiratory Societies (FIRS) on the use of e-cigarettes, that electronic nicotine delivery devices should be restricted or banned until more information about their safety is available. The incidence of adenocarcinoma (Ad) has increased, while that of squamous cell carcinoma (Sq) and small cell lung carcinoma (SCLC) has decreased in Europe, North America and Japan.5 However, few studies have assessed the incidence and histological type of lung cancer in China. Kong et al.6 reported the findings of a population-based study of the incidence of lung cancer in Tianjin, the third largest city in China, during the period of 1981–2005. According to their report, the most common histological type is Sq in males and Ad in females. However, during the 25-year study period, the age-adjusted incidence of Sq declined sharply, while that of Ad continued to increase among younger groups. The authors concluded that tailored strategies for prevention and control should be developed to meet the needs of various populations. In Respirology, Alcada et al.7 evaluated the features of mediastinal lymphadenopathy on computed tomography (CT) and the clinical data of 217 human immunodeficiency virus (HIV) patients in the era of combination antiretroviral therapy. Among these patients, 52 were identified to have mediastinal lymphadenopathy, and 17 (33%) were diagnosed with pulmonary malignancy, including lung cancer. Larger lymph nodes were associated with increased odds of malignancy (OR 2.89; 95% confidence interval 1.24–6.1) according to a multivariate analysis. The reason for the high association between lung malignancy and HIV was discussed; however, the mechanisms remain unknown. The authors emphasized the need for a histological diagnosis in patients exhibiting swollen mediastinal lymph nodes. Although the National Lung Cancer Screening Trial (NLST) demonstrated that the application of low-dose computed tomography (LDCT) is useful for reducing lung cancer-related mortality,8 whether the detection of subcentimeter non-calcified nodules without changes in size over a 2-year follow-up period are indicative of malignancy or should be further followed up remains unknown. Shin et al.9 performed a retrospective study to answer this question. The authors investigated 635 subjects who received follow-up for LDCT for an initial 2-year screening period with three additional years thereafter in whom non-calcified subcentimeter nodules were evaluated. A total of 1107 subcentimeter nodules (1037 solid, 70 ground-glass opacity nodules) were detected. All solid subcentimeter nodules exhibiting stability for the initial 2-year period on screening LDCT were considered benign, as none of these lesions showed growth in the subsequent period. In contrast, subcentimeter GGNs have a greater potential for growth than solid nodules and require further follow-up with CT for more than 2 years, as some of them could be malignant. As a result of the increasing rate of detection of lung cancer nodules during lung cancer screening using CT, the development of a non-invasive diagnostic method for reducing the risks associated with biopsies is warranted. Confocal laser endoscopy (CLE) allows for real-time non-invasive histological imaging, and a probe-based CLE (pCLE) method has been developed in which the probe can be passed into the distal airway via the working channel of the bronchoscope, thus obtaining an optical biopsy sample. Meanwhile, Sorokina et al.10 performed an ex vivo study to evaluate the correlations between light microscopic findings and pCLE imaging findings of primary lung cancer. The authors examined 18 lobectomy samples from 18 different patients and found that pCLE can be used to identify lung carcinoma in ex vivo samples. Certain light microscopic features of lung carcinoma may be visualized on pCLE. These results suggest that pCLE is a non-invasive diagnostic method for making the histological diagnosis of nodules detected on CT scans. The volume doubling time (VDT) can be calculated using serial CT scans and may be applied to evaluate indeterminate pulmonary nodules detected at lung cancer screening. A retrospective study of VDT in surgically resected non-small cell lung cancer patients was recently reported by Mackintosh et al.11 The authors investigated 109 eligible scans in 46 patients with lung cancer (36 Ad, six Sq, two large cell and two carcinoids) and demonstrated that the non-small cell lung carcinoma (NSCLC) growth rate appears to be highly variable and related to both the histological subtype and smoking history, but not the presence of symptoms at diagnosis. Relatively slow-growing Ad lesions frequently metastasize. Araz et al.12 examined the correlations between the Ki-67, p53, transforming growth factor (TGF)-b and lysyl oxidase (LOX) values and the metastatic stage in various types of lung cancers. Consequently, high levels of cellular LOX and TGF-b were found to be related to an increased incidence of distant metastasis in Ad patients. LOX and TGF-b may therefore be useful markers of metastatic disease in patients with Ad. Furthermore, Ming et al.13 evaluated the diagnostic utility of the vascular endothelial growth factor (VEGF) messenger ribonucleic acid (mRNA) and specificity protein-1 (SP-1) mRNA expression levels in cells obtained via bronchial brushing in 93 patients with lung cancer and 51 benign pulmonary lesions as a control. In the cancer group, the VEGF mRNA levels were significantly correlated with the SP-1 mRNA levels (P < 0.01) and showed the highest diagnostic rate, with a sensitivity of 89.2% and accuracy of 90.3%. These values were significantly better than those for cytology (P < 0.01). The authors concluded that the detection of VEGF mRNA and SP-1 mRNA in bronchial brushing cells may be useful for identifying early-stage lung cancer. Ost et al. and Yasufuku et al. both published excellent reviews with regard to multi-modality systematic approaches to mediastinal lymph node staging and the role of the bronchoscopist in identifying molecular profiles in cases of non-small cell lung cancer, respectively14, 15 Accurate and efficient lymph node staging is essential for selecting the treatment modality, and determining the molecular profile of the tumour is necessary for choosing appropriate molecular target drugs and prolonging the patient's survival. Qvale et al.16 showed that paraneoplastic Hu and collapsing response mediator protein 5 antibodies, which were originally thought to be specific markers for paraneoplastic syndrome, are found only in smokers without cancer or neurological disease using sera obtained from 552 smokers (379 smokers with and 173 without COPD) and 300 healthy controls. Preoperative physiological assessments are important for accurate patient selection and to provide appropriate treatment. Although the Eastern Cooperative Oncology Group (ECOG) performance status (PS) is usually used in clinical practice, its accuracy is not satisfactory. Therefore, Roman et al.17 compared the peak oxygen consumption (VO2 peak) with the ECOG PS in order to evaluate the clinical utility of the VO2 peak in operable patients with non-small cell lung cancer. The authors employed 392 NSCLC patients. PS scoring systems do not provide sensitive measurements of the functional status. Therefore, the authors concluded that the VO2 peak may be useful in the clinical management of oncology patients. Patients with advanced adenoid cystic carcinoma (ACC) of the central airway often develop fibrotic airway stenosis following radiotherapy. Eom et al.18 revealed the clinical utility of bronchoscopic intervention, including silicone airway stenting, in such patients. Forty-seven patients with ACC who received radiotherapy were analysed. Twenty-three per cent of the patients suffered from fibrotic airway stenosis after radiotherapy and underwent bronchoscopic intervention. The authors demonstrated this intervention to be safe and useful for treating airway stenosis after radiotherapy in patients with ACC. The intrapleural delivery of a commercially available compound made up of proteins produced by Staphylococcus aureus is used clinically to induce pleurodesis. Lansley et al.19 reported the findings of a proof-of-principle study showing the efficacy of inhibiting mesothelioma growth with a compound consisting of Staphylococcus aureus bioproducts. The authors demonstrated that in addition to its pleurodesing effect, this bacterial compound has anti-tumoural activities against pleural malignancies and concluded that its anti-tumoural activity against mesothelioma warrants clinical validation.20 Finally, a very attractive treatment strategy, stem cell therapy, was recently introduced in an editorial published in Respirology.21 It has been reported that cellular therapy using stem cells may constitute a major component of future therapies. Pleural procedures such as thoracentesis and chest drain insertion are commonly performed worldwide. Complications, including iatrogenic pneumothorax, bleeding and re-expansion pulmonary oedema, may occur although probably at a lower rate than previously thought. In a 3.5-year audit of 529 bedside procedures using safety checklists, and ultrasound-guidance in 86% of the cases, the complication rate was only 3% and was similar whether being performed by either pulmonologist or non-pulmonologist operators.22 The presence of COPD independently increased the risk of complications by nearly sevenfold. In another large series of 9320 ultrasound-guided thoracenteses, the overall complication rate (0.98%) and pneumothorax rate (0.61%) was also lower than many published studies.23 Specifically, the incidence of re-expansion pulmonary oedema was extremely low (0.01%). Recently, an intriguing preliminary observation of cough-related changes in pleural pressure during therapeutic thoracentesis in three patients showed that elevation of pleural pressure not only coincided with coughing episodes but persisted shortly thereafter.24 Thus, the authors hypothesized that cough might have a beneficial effect preventing the excessive drop in pleural pressure. Determination of epidermal growth factor receptor (EGFR) and Kirsten rat sarcoma viral oncogen homologue (KRAS) mutational status, as well as anaplastic lymphoma kinase rearrangement, has become an essential part of the evaluation of lung cancer patients because of its prognostic and therapeutic implications. Obtaining tissue samples for this purpose is often difficult, whereas analysing pleural fluid is easy and less invasive. In a study of 57 patients with malignant pleural effusions (84% with lung cancer), detection of KRAS mutations by peptide nucleic acid clamping was higher in pleural fluid samples (14%) and their cellular blocks (16.7%) than in matched tumour tissues (5%) and serum (3.6%) specimens, thus reinforcing the suitability of the former for mutation analyses.25 An excellent review addressed the current controversies and trends in the management of malignant pleural effusions.26 The main goal when dealing with this condition should be to relieve patient's symptoms with the least invasive and cost-effective strategies, while reducing hospitalization time. Therapeutic thoracentesis, due to its temporary benefits, is indicated in patients with: (i) very short expected survival (e.g. ≤4 weeks) or poor performance status; (ii) potential multifactorial causes of breathlessness in order to establish the relative contribution of the effusion to this symptom; and (iii) tumours in which a favourable short-term response to targeted therapies or chemotherapy is expected (eg. small cell lung cancer, lung cancer with EGFR mutations, lymphoma). Even though the risk of re-expansion pulmonary oedema is extremely low, recommended fluid removal is generally limited to 1.5 L unless pleural pressure is monitored during the procedure. Pleurodesis and placement of an indwelling pleural catheter (IPC) are the common definitive procedures to control symptomatic malignant effusions, but the preferred choice is controversial.26 The ideal timing for pleurodesis, that is, whether it should be performed routinely at the diagnosis of malignant pleurisy or deferred until symptomatic recurrence occurs, has not been established. Talc poudrage pleurodesis through thoracoscopy has not demonstrated superiority over chest tube talc slurry.26 If talc is selected as the sclerosant agent, large particle size preparations are recommended to avoid the development of an acute respiratory distress syndrome. Finally, IPC allows the ambulatory drainage of malignant effusions with similar symptomatic benefits as pleurodesis. It is indicated either as a primary approach or when pleurodesis fails (up to 30%) or is unsuitable (trapped lung). The procedure has few minor complications (eg. catheter blockage, cellulitis, symptomatic loculation) and spontaneous pleurodesis ensues in almost half the patients at an average time of 2 months, a situation in which IPC is definitely removed.26 Future treatment strategies for malignant pleural effusions should encompass advances in translational and experimental medicine. For example, one study evaluated the anti-tumoural effects of a S. aureus bio-product which has been used as a pleurodesing agent.19 This commercially available compound killed mesothelial cells in vitro and retarded tumour growth in a murine model of mesothelioma. A review described the role of interventional pulmonology in the management of bacterial infections of the pleural space.27 All patients with community-acquired parapneumonic effusions or empyema should receive empirical antibiotics covering Gram-positive cocci and anaerobes, but usually not atypical bacterial pathogens such as Legionella or Mycoplasma. The optimal duration of antibiotic therapy is unknown, although a period of 4 to 6 weeks is not uncommon.27 Nutritional supplementation is advised if poor nutrition is a concern. A decision on whether to drain non-purulent effusions is challenging and generally based on radiological (large effusions) and pleural fluid biochemical (low pH or glucose) or microbiological (positive Gram stains or cultures) data. As there are no randomized controlled studies offering guidance, it may be wise to follow the recommendation of Dr Light in initially performing a therapeutic rather than a diagnostic thoracentesis.28 The rationale is that if no fluid re-accumulates, no additional therapy will be necessary. Alternatively, when indicators for pleural drainage exist, small-bore chest drains placed under ultrasound guidance are the procedure of choice.27 The rightful place for intrapleural fibrinolytics remains a matter of intense debate. Meta-analyses of randomized controlled trials that have included the negative First Multicenter Intrapleural Sepsis Trial 1 (MIST1) and second MIST2 studies still conclude that urokinase or alteplase might be potentially effective for reducing the need for surgery.29 According to MIST2 study, the administration of alteplase and DNase intrapleurally should be considered whenever patients, particularly those who are not good surgical candidates, fail to respond to thoracostomy drainage. Surgical (i.e. video-assisted thoracoscopic surgery or open thoracotomy) rather than medical thoracoscopy is favoured when sepsis is uncontrolled despite the previously instituted therapies or when lung entrapment develops as a complication of the pleural infection. A systematic review reported the clinical characteristics and treatment of patients with yellow-nail syndrome (YNS).30 It is probably an acquired disease affecting lymphatic drainage, which is diagnosed when at least two of the following three characteristics are met: yellow nails, lymphoedema and chronic respiratory symptoms including pleural effusions in 40–50% of the cases. The authors compiled 150 YNS patients with pleural effusions.30 The median age was 60 years, without gender predominance. All patients had lymphoedema, yet 14% did not exhibit yellow nails. Pleural effusions were bilateral in two thirds, had a serous appearance in 75% of the cases and milky in 20%, and met exudative criteria with lymphocytic predominance in 95%. On pleural biopsy, findings were either unspecific or consistent with chronic pleuritis. For symptomatic persistent effusions, pleurodesis or decortication/pleurectomy is effective in most cases. Acute respiratory infection is a major cause of morbidity and mortality among children, especially in developing countries. In a study by Wu et al.31 in Hubei, China, indirect immunofluorescence assays for immunoglobulin M antibodies were positive in 7046 (67.5%) of 10 435 serum specimens collected from hospitalized children presenting with acute respiratory infection symptoms against at least one of the following nine pathogenic viruses and atypical bacteria: Mycoplasma pneumoniae, influenza virus B, respiratory syncytial virus, parainfluenza virus, adenovirus, influenza virus A, Legionella pneumophila, Coxiella burnetii and Chlamydophila pneumoniae. M. pneumoniae had the highest detection rate (56.9%), followed by influenza virus B (35.4%) and respiratory syncytial virus (18.9%). Selective factors for hospitalization and exact diagnostic criteria for study inclusion might have partly contributed to the relatively high detection rates. As many as 4136 (39.6%) specimens were positive for at least two pathogens, possibly reflecting an intrinsic limitation of the serological test in differentiating current and recent infections. Notwithstanding these, characterization of the pathogen profile with their specific seasonal and age patterns may still help to inform the clinical management of children hospitalized for acute respiratory infection. Gidaris et al. reviewed the potential role of acute bronchiolitis, a common paediatric disease of infancy, in the inception of asthma.32 Acute bronchiolitis in infancy has been associated with an increased risk of recurrent wheezing throughout the primary school years. However, it remains uncertain whether acute bronchiolitis leads to the subsequent development of asthma, or it just represents the first clinical presentation in persons otherwise predisposed to asthma. Complex interplay between various host factors and viral factors is likely to be involved, perhaps by varying extent in different individuals. Further elucidation of these issues may carry important prognostic implications and help to inform prevention and/or treatment. In a meta-analysis by Ye et al. involving 13 studies with a total of 1138 patients, the pooled sensitivity, specificity, positive likelihood ratio, negative likelihood ratio and diagnostic odds ratio of the soluble triggering receptor expressed on myeloid cells-1 (sTREM-1) for diagnosis of lower respiratory tract infection (LRTI) were 0.84, 0.77, 3.6, 0.21 and 17, respectively.33 The overall diagnostic value was similar for community-acquired LRTI and hospital-acquired LRTI in subgroup analyses. Sensitivity, but not specificity, was significantly affected by the assay method in univariate meta-regression analysis. Different cut-offs for sTREM-1 were adopted in individual studies using different assay methods on either serum or bronchoalveolar lavage fluid as the sample source. LRTI is a heterogeneous clinical entity. Further studies are required to realize the diagnostic utility of sTREM-1 in different clinical settings. Mycoplasma pneumoniae cannot be reliably diagnosed through clinical symptoms and signs alone in children and adolescents with community-acquired pneumonia (CAP).34 Miyasita et al. reported abnormal findings on high-resolution CT (HRCT) of the chest in 24 (41%) out of 53 cases of laboratory-confirmed M. pneumoniae infection with cough lasting 7 or more days and a normal chest radiograph. Such observation highlights the suboptimal sensitivity of chest radiograph in detecting bronchiolitis or bronchopneumonia due to M. pneumoniae.35 An apparently normal chest radiograph alone should not be relied upon to exclude M. pneumoniae in children or adolescents presenting with relevant acute respiratory symptoms. A simple clinical risk score using weighing factors of 4 for recent hospitalization, 3 for nursing home, 2 for chronic hemodialysis and 1 for critically ill has previously been found to perform moderately well at classifying patients regarding their risk for methicillin-resistant S. aureus (MRSA), Pseudomonas aeruginosa, or extended-spectrum β-lactamase-producing organisms in adult patients admitted via the emergency department with bacterial pneumonia.36 Ma et al. derived another risk score for predicting pneumonia caused by potentially drug-resistant (PDR) pathogens using weighing factors of 14 for bronchiectasis, 5 for recent hospitalization, 2 for severe pneumonia and 0.5 for the presence of other risk factors for health care-associated pneumonia (HCAP).37 The risk score outperformed the HCAP definition in predicting PDR pathogens in the derivation cohort with 48 PDR isolates among 354 patients hospitalized with non-nosocomial bacterial pneumonia and in a subsequent validation cohort with 21 PDR isolates among 96 similar patients. An optimal balance between sensitivity and specificity was observed at a relatively low cut-off score of ≥2.5. A history of bronchiectasis or recent hospitalization alone may thus prompt the consideration for appropriate broad-spectrum antibiotic coverage against PDR in the empirical treatment of non-nosocomial bacterial pneumonia. The isolation of MRSA from bacterial culture does not always point to a clinically significant infection. In the study by Nagoka et al., among 337 consecutive patients with MRSA isolation from respiratory specimens, ‘true’ MRSA nosocomial pneumonia was diagnosed by anti-MRSA treatment effects, Gram-staining or bronchoalveolar lavage fluid culture in 36 patients (11%), whereas nosocomial pneumonia with MRSA colonization was diagnosed in another 34 (10%).38 Patients with MRSA nosocomial pneumonia were independently associated with a Pneumonia Patient Outcomes Research Team score of 5, single cultivation of MRSA, MRSA quantitative cultivation yielding more than 106 CFU/mL, radiological findings other than lobar pneumonia and a history of head, neck, oesophageal or stomach surgery. In a prospective, multicentre study by Sanz et al., 152 (14.6%) of 1038 patients diagnosed of CAP had prolonged onset of symptoms as defined by a time interval of over 7 days from symptom onset to pneumonia diagnosis.39 The longer time taken from symptom onset to diagnosis was independently associated with previous corticosteroids and antibiotic therapy, alcoholism and less severe pneumonia. While it was associated with a higher rate of suppurative complications, there was less need for intensive care unit admission and no significant effect on 30-day mortality. Urinary pneumococcal antigen test has been shown to be a useful addition to the conventional diagnostic workup for community-acquired pneumonia.40 A higher 30-day mortality was observed in bacteraemic pneumococcal CAP as compared to non-bacteraemic but urinary antigen test-positive pneumococcal CAP in a previous study.41 In a study by Zalacain et al. involving 350 consecutive adult immunocompetent patients admitted for bacteraemic pneumococcal pneumonia, the urinary pneumococcal antigen test performed within the first 24 h of admission showed a sensitivity of 74.6%.42 Positive results were independently associated with a higher risk of intensive care unit admission, treatment failure and adverse outcome, thus implying prognostic implication of a positive urinary antigen test in pneumococcal CAP even in presence of bacteraemia. Use of procalcitonin to guide initiation and duration of antibiotic treatment for acute respiratory infections has been shown to reduce antibiotic consumption without adverse impact on treatment failure or mortality across different clinical settings.43 In a prospective, nested case-control study conducted by Tokman et al., elevated serum procalcitonin >0.5 ng/mL was found to be an independent predictor of in-hospital mortality among 241 HIV-infected adults admitted to a hospital with cough for 2 or more weeks in Kampala, Uganda.44 Elevated serum procalcitonin, tachypnea (rate ≥30 breaths/minute) and hypoxemia (oxygen saturation <90%) could be combined into a prognostic model to stratify patients with different mortality risks from 1% (no variable present) to 42% (all variables present). Such a model may be useful in identifying patients at high risk of dying from lower respiratory tract infections in low and middle-income countries where HIV infection is prevalent. In a study by Ferrer et al., previous use of inhalational steroid (ICS) in patients hospitalized for CAP was independently associated with a reduced systemic inflammatory response as reflected by lower tumour necrosis factor (TNF)-α level, but there was no significant impact on long-term mortality. Such association persisted in a subgroup of patients with COPD.45 A previous prospective study demonstrated a different early inflammatory pattern with a lower TNF-α level in CAP among patients with COPD as compared to those without, but such difference was not completely steroid mediated.46 Both ICS and COPD therefore appear to modify host inflammatory response in CAP, with potential impact on disease manifestations and/or clinical course. Prior ICS treatment has also been associated with a lower incidence of parapneumonic effusion in pneumonic patients with different chronic respiratory disorders.47 The clinical course of pulmonary aspergilloma is often difficult to predict. In an observational study by Lee et al. involving a total of 143 adult pulmonary aspergilloma patients in multiple centres, the size of aspergilloma decreased in 13.3% and increased in another 25.9% on serial chest CT during a median follow-up duration of 1 year.48 Size reduction was associated with higher C-reactive protein, more severe bronchiectasis and tuberculosis-destroyed lung. Clinically significant haemoptysis was associated with absolute size of cavity and mass of aspergilloma. Reliable prognostic markers are lacking in patients with non-cystic fibrosis bronchiectasis and Mycobacterium avium complex (MAC) infection. Zoumot et al. conducted a retrospective analysis of 52 adults with non-cystic fibrosis bronchiectasis and coexisting MAC infection.49 Chronic pulmonary aspergillosis, cavitation within nodules and emphysema on HRCT at presentation were found to be independent predictors of mortality. Anti-MAC chemotherapy was more likely to lead to MAC culture conversion as compared with patients managed with observation, but there was no significant improvement in survival. The latter finding has to be interpreted with caution as it is often difficult to exclude selection bias, especially in relation to who to treat and when to treat, in an observational study. Ho et al. conducted an experiment using magnetic nanoparticles and confocal microscopy to investigate the role of Aminopeptidase N (CD13), an ectoenzyme located in the outer membrane of a variety of cells, on the internalization and intracellular survival of Mycobacterium tuberculosis in monocytes.50 M. tuberculosis was found to be capable of binding to either soluble or membranous CD13. Pretreatment of monocytes with anti-CD13 antibody (WM15 and WM47) reduced the number of intracellular M. tuberculosis, thus lending support to a possible role of the enzyme in the pathogenesis. Huang et al. evaluated the performance of microscopic observation drug susceptibility (MODS) assay in 173 samples (pleural fluid, 112; cerebrospinal fluid, 61) collected from patients suspected to have extrapulmonary tuberculosis.51 Notwithstanding its shorter turnover time as compared to the Lowenstein–Jensen methods, the clinical utility of the MODS assay was still suboptimal