Contemporary Concise Review 2018: Lung cancer and pleural disease

Role of EBUS in peripheral tumours

Standard initial staging of lung cancer is made on the basis of radiological findings, frequently following positron emission tomography/computed tomography imaging (PET/CT). However, there is increasing evidence that PET-CT alone is insufficient for appropriate nodal staging of lung cancer.1 With advances in therapy for lung cancer and in particular novel ablative techniques, accurate staging of nodal involvement is more crucial than ever. In the absence of evidence of metastatic disease, it is generally accepted that patients with PET-avid hilar or mediastinal nodes should proceed to minimally invasive sampling of mediastinal and/or hilar lymph nodes, commonly performed through EBUS-transbronchial needle aspiration (EBUS-TBNA).

In patients with PET negative nodes, the role of EBUS is controversial, and previous work has demonstrated a discrepancy between results from PET-CT and EBUS.2 The most recent American College of Chest Physicians (ACCP) guidelines suggest, that if the tumour is central or hilar N1 nodes are present, to proceed to minimally invasive sampling of the nodes.3 If the tumour is peripheral, and there is no radiological suspicion of mediastinal or hilar nodes, these guidelines advise that there is no role for EBUS.

Vial et al. bring into question the appropriateness of this recommendation in their prospective cohort study published in Respirology.4 They consecutively recruited patients with non-small cell lung cancer (NSCLC) between April 2009 and February 2014. A total of 120 participants were enrolled, of whom 75 patients' samples were included. EBUS-TBNA was performed under general anaesthetic in a specialist centre with the availability of rapid on-site evaluation (ROSE). The work demonstrated that the sensitivity of EBUS-TBNA in this cohort was 40% in the identification of N2 disease and 20% of patients with N0 or N1 disease by radiographic criteria were found to have N2 disease. In addition, N2 disease was also present in 5 of 61 (8%) patients with radiologically staged N0 disease and peripheral tumours, which is a group of patients which the ACCP guidelines do not recommend sampling. False negative results arose from lymph nodes which did not meet sampling criteria, for example, due to their small size. The study demonstrates the difficulty in proving N1 disease, which in some cases was due to the inability to sample intraparenchymal lymph nodes inaccessible to EBUS. This highlights potential limitations of current patient and node sampling selection, and whether we should be systematically sampling more nodes from each station, or routinely sample stations 4 and 7 regardless of size.

The safety of EBUS has been reviewed in a review published in Respirology in 2017 by Vaidya et al., confirming very low rates of significant complications, with only six pneumothoraces reported out of 10 000 patients undergoing EBUS-TBNA only procedures.5 The review also states that there were five deaths attributed to EBUS to date at the time of the review. This suggests that where there is diagnostic doubt regarding nodal involvement, the risk–benefit will often fall in favour of performing minimally invasive sampling of the nodes to achieve accurate staging, as this alters prognosis and treatment options for the patient. Of course, this will be on an individual basis considering other patient-specific factors.

Diagnosing peripheral pulmonary nodules

Diagnosis of small peripheral pulmonary lesions can be challenging. Techniques such as radial EBUS, virtual navigational bronchoscopy (VNB) and electromagnetic navigation bronchoscopy (ENB) have been developed to try to improve the diagnostic yield of sampling such lesions. However, results presented in 2016 from the AQuIRE (American College of Chest Physicians Quality Improvement Registry, Evaluation, and Education) Registry suggested that conventional bronchoscopy performed better than guided bronchoscopy, even after adjustment for baseline factors.6 Diagnostic yield was 63.7% when no radial EBUS and no ENB were used, 57.0% with radial EBUS alone, 38.5% with ENB alone and 47.1% when the techniques were combined. This involved longitudinal data collection from 15 centres into a database; however, as the authors state, this was not a randomized control trial (RCT) directly comparing the techniques. These figures are in contrast to the results of the systematic review and meta-analysis published here in 2017 by Ali et al., showing a 70.6% diagnostic rate of radial EBUS at diagnosing peripheral lung lesions.7

The role of navigation techniques in the diagnosis of peripheral lesions remains uncertain, with several techniques and systems available to assist the bronchoscopist in obtaining diagnostic samples. Studies comparing different approaches are therefore very important in defining the role of these methods in the diagnostic work-up of patients. In 2018, Respirology published the first prospective RCT evaluating the use of virtual bronchoscopic navigation to assist CT-guided transbronchial biopsy in the diagnosis of peripheral pulmonary lesions.8 Kato et al. showed that virtual bronchoscopic navigation improves diagnostic yield of CT-guided transbronchial biopsy.8 Patients were recruited with small peripheral pulmonary nodules <20 mm in size and were randomized to CT-guided transbronchial biopsy (n = 50) or virtual bronchoscopic navigation transbronchial biopsy (n = 50). They demonstrated that the diagnostic yield was significantly higher in the VBN group versus the standard CT-guided biopsy group (84% vs 58%, respectively (P = 0.013)). The accompanying editorial by Asano reviews this along with other studies and discusses pros and cons of different imaging modalities in virtual bronchoscopic navigation.9

Malignant airway obstruction

Malignant airway obstruction may arise from a primary bronchogenic carcinoma, secondary metastatic deposits or from extrathoracic compression or local invasion from nearby tumours such as oesophageal or thyroid tumours. Symptoms can be distressing and debilitating and it can represent a poor prognosis in terms of quality and quantity of life. Presentation may be rapid, requiring immediate treatment, or slower, allowing for a considered approach thus giving rise to a variety of treatment modalities.

Management is usually performed with palliative intent, and intervention performed in the absence of significant symptom burden is often ill-advised. It is however necessary to offer active intervention to patients with significant distress to relieve their symptoms, even (and often) in the presence of advanced disease. With this in mind, it is unsurprising that therapeutic bronchoscopy is not without complication. A large prospective review of complications following therapeutic bronchoscopy for malignant central airways obstruction as part of the AQuIRE registry found that complications occurred in 3.9% of patients, with death occurring in 0.5%.10

Phillips’ editorial11 recognizes that the type (extraluminal or intraluminal), site and extent of obstruction are crucial in determining the appropriate treatment, and discusses potential therapeutic options. Options may include bronchoscopic debulking of the tumour (by laser therapy, cryotherapy, electrocautery or mechanical debridement), endobronchial stent placement and external beam radiotherapy (EBRT). These treatment modalities may be used alone or in combination. Bronchoscopic intervention is the mainstay of treatment for acute central airway obstruction, as benefit can be immediate, with successful recanalization being achieved in around 90% when performed by experienced clinicians.10 Complete or severe obstruction is likely to require immediate intervention, and hence is unlikely to be suited to EBRT alone, and conversely small cell tumours which may be highly radio- and chemo-sensitive, may be more suitable for sole treatment with EBRT.11

Respirology has published work by Mallow et al. comparing treatment outcomes in malignant airway obstruction.12 They performed a retrospective single-centre analysis of 237 patients with malignant airway obstruction, and compared EBRT, endobronchial stents and a combination of the two. It showed better survival in the combination group than in either single treatment group; however, the retrospective nature of the study introduces several biases. Also, as acknowledged by the authors, the retrospective design of the study meant it was not possible to assess the impact of treatment on quality of life or symptoms. Further prospective studies are needed to do so, as quality of life is crucial for such patients in their terminal phase of life, and prolonging life at the expense of quality of life is unlikely to be in patients’ best interests. However, previous trials set up to address these questions prospectively have had difficulty recruiting.

Nonetheless, there is previous work showing that intervention improves symptoms,13 and it is possible that the same may be found in the combined treatment approach mentioned above. A previous prospective observational study of consecutive patients undergoing therapeutic bronchoscopy for central airway obstruction analysed health-related quality of life (HRQOL) and quality-adjusted survival in 102 patients.13 Improved HRQOL after therapeutic bronchoscopy was demonstrated as compared with baseline, resulting in approximately a 5.8% improvement in HRQOL per day of life.13

Endobronchial drug delivery

With the development of targeted therapies against specific genetic mutations in lung cancer, treatment outcomes have improved. Nonetheless, conventional chemotherapy remains at the core of many treatment regimens. The non-specific nature of these cytotoxic drugs leads to significant adverse effects limiting their utility. Direct administration of chemotherapy to the tumour in theory should allow targeted treatment of that tumour, with reduced systemic side effects. Tsukada et al. demonstrate safety and feasibility of paclitaxel delivery into porcine bronchial tissue.14 Paclitaxel is administered into bronchial tissue via a bronchoscopic catheter involving a balloon and a needle. Following instillation of the drug, sustained tissue concentrations were subsequently recorded up to 28 days following injection, and importantly the drug remained localized to the lung tissue. As no animal models of tumours exist, they were administered into healthy bronchial tissue rather than into a tumour. The group will continue their work looking at paclitaxel in human patients with malignant airway obstruction in an upcoming trial (NCT02066103). Steinfort's accompanying editorial discusses the potential uses of endobronchial intratumoral chemotherapy (EITC) with the possibility of a role in neoadjuvant therapy in NSCLC as well as in malignant airway obstruction.15 The use of intratumoral chemotherapy in severe malignant airway obstruction has been described previously, with work dating back more than 20 years, where a combination of chemotherapeutic agents were administered via flexible bronchoscopy, and relieved the obstruction in 81 of 93 patients.16

Surgery and post-operative follow-up

Lobectomy and lymph node dissection is the established surgical management of early stage thoracic cancer, whilst the suitability of limited resection remains controversial.17 At present, the only available RCT data comparing the two approaches was published by the Lung Cancer Study Group (LCSG) in 1995 and demonstrated that patients undergoing limited resection suffered increased recurrence rates versus lobectomy.18 The evaluation of segmentectomy in this study was complicated by the inclusion of patients undergoing wedge resection in the same ‘limited resection’ arm, and the 23-year interval has seen the advent of widespread CT and PET/CT availability. With this in mind, Nguyen's editorial in our July edition discusses the importance of these findings in the context of modern practice, and summarizes conflicting results from recent studies.19

One such study published in this journal, a retrospective observational review of the Surveillance, Epidemiology, and End Results (SEER) cancer registry, found no significant difference in the 5-year overall survival rates of 15 358 patients with primary NSCLC ≤ 20 mm, N0, treated with either lobectomy or anatomical segmentectomy (76% vs 74.4%, P = NS).20 Elderly patients were more likely to have undergone segmentectomy, with a subgroup analysis of those >75 years old, demonstrating comparable overall and lung cancer specific survival rates compared with lobectomy. These important findings add considerable weight to the need to reconsider the role of anatomical segmentectomy, which preserves pulmonary function and allows for mediastinal lymph node dissection.20 Several large RCTs (Japanese Clinical Oncology Group trial in Japan (JCOG0802); Alliance trial (CALGB140503) in North America) investigating the role of anatomical segmentectomy are ongoing, which will help establish whether segmentectomy will become the new surgical standard for early stage lung cancer.19, 21

Post-operative pneumonia is a significant cause of morbidity and mortality following lung cancer resection surgery, with laryngeal pathology through injury to the left recurrent laryngeal nerve or double-lumen tube-related vocal cord trauma often implicated.22-24 A prospective study published in Respirology by Fourdrain et al.,22 and considered in an accompanying editorial,23 demonstrated that endoscopic laryngeal assessment performed post-operatively is effective in diagnosing laryngeal pathology, with vocal cord paralysis found in 13 of 250 patients (5.2%). Furthermore, such findings were associated with a higher rate of post-operative pneumonia (P = 0.03), reintubation (P = 0.007) and a trend towards an increased 90-day mortality rate (P = 0.09). Dysphonia was found to be associated in all patients with vocal cord paralysis.

The authors propose that post-operative laryngeal endoscopic assessment should be routinely performed in all patients undergoing surgical resection for lung cancer, as the early identification of vocal cord injury can lead to the instigation of measures to reduce the risk of aspiration such as vocal cord medialization, careful swallowing management and speech therapy.

Lung cancer MDT meeting

MDT meetings play a crucial role in determining the most appropriate, individualized management for patients with suspected and confirmed lung cancer. Stone et al. recently published an article in Respirology describing how they sought expert consensus opinion to formulate a minimum required data set for the lung cancer MDT.25 Their aim was to improve the management decision-making of the MDT, but also to have a systematic way of being able to evaluate the performance of the MDT and to allow benchmarking between different centres. They included lung MDT healthcare professionals in different states of Australia, and their work involved two survey rounds followed by a case conference to create a data set. Criteria suggested inclusion of patient details, risk factors including smoking and asbestos exposure, details on biopsy data and staging from various modalities (including CT, PET, EBUS and surgical pathology). They also include what the authors refer to as ‘timeliness data’, covering date of confirmed diagnosis and date of first treatments, and treatment details. The authors recognize that experts in certain fields were under-represented, including radiologists, pathologists and palliative care clinicians, and they postulate that future work could also involve patient participants. Nonetheless, the suggested data set gives a frame work through which the MDT can operate, and will be implemented in St Vincent's Hospital (Sydney) Lung Cancer MDT. Work such as this is important in constructing national and international guidelines around minimum standards for MDT discussion, and others are encouraged to share their experiences.

Pneumothorax

Pleural effusions: Pleurodesis and IPC

Non-specific pleuritis

In patients with active malignancy who develop a pleural effusion with negative fluid cytology, the histological finding of non-specific pleuritis at pleuroscopy presents a diagnostic dilemma. Vakil et al. retrospectively report on long-term outcomes of 199 patients with suspected malignancy and an undiagnosed exudative pleural effusion of 1–1.5 L who underwent pleuroscopy between 2005 and 2015 at a single institution.47 A median of 10 biopsies were performed per patient and mean follow-up was 24.2 ± 10.0 months. Ninety patients (52%) were diagnosed with non-specific pleuritis. The negative predictive value for malignancy at 2-year follow-up was 97%: only three patients subsequently developed malignancy at follow-up. Over half of the cases of non-specific pleuritis were attributable to the effects of chemotherapy and radiotherapy. This pivotal study not only serves to corroborate the early teachings of Boutin et al.48 emphasizing the importance of histopathology, but also the finding of non-specific pleuritis can be relied on as robust evidence to commence aggressive therapy. As highlighted in Fielding's editorial, the use of pleural ultrasound to target smaller effusions, surface imaging such as tissue autofluorescence to better interrogate the pleural surface for abnormalities, and improved biopsy techniques to optimize diagnostic yield, will all serve towards establishing a confident diagnosis of benign disease at pleuroscopy.49

Pleural infection

Bacterial PPE is a condition with high morbidity and mortality.50 The pathogenesis reflects a complex interplay of host, bacterial and cellular factors that trigger a cascade of inflammatory pathways that may culminate in empyema and pleural loculation, a therapeutic challenge, but increasingly regarded as a protective mechanism to wall off infection. Popowicz et al. provide an excellent overview of the intricate processes involved in their January 2018 editorial in Respirology.51 The clinical phenotype of affected patients is heterogeneous in presentation and in disease severity, and host susceptibility may be influenced by underlying co-morbidity and timing of therapy. Bacterial virulence is in part determined by a variety of products expressed by the causative organism. Lipoteichoic acid (LTA) is such an example of a major soluble cell wall component of Gram-positive bacteria (GPB) that binds to target cells inducing inflammation and coagulation.52, 53 Pleural mesothelial cells (PMC) are the first line of defence against invading pathogens and their armamentarium.54 In response to infection, they produce plasminogen activator inhibitor-1 (PAI-1) which impedes fibrinolysis and results in loculation.55, 56 PAI-1 is significantly elevated in complicated PPE (i.e. PPE which requires intervention) compared to uncomplicated PPE and is associated with residual pleural thickening.57, 58 However, no studies have implicated LTA in PAI-1 production. Lee et al. studied thirty consecutive patients with PPE of whom 12 had GPB (Staphylococcus aureus in 9 and Streptococcus pneumoniae in 3), 7 had Gram-negative bacteria (Klebsiella pneumoniae in 3, Pseudomonas aeruginosa in 3 and Escherichia coli in 1) and 11 had no growth in their pleural fluid culture.59 Their study showed that culture-positive PPE, especially that caused by GPB, was associated with significantly higher levels of PAI-1 compared to uncomplicated culture-negative PPE. Furthermore, in vitro, LTA upregulates PAI-1 expression through activation of Toll-like receptor 2 (TLR2)/c-Jun N-terminal kinase (JNK)/activator protein 1 (AP-1) signalling in human PMC. It is tempting to speculate that modulation of this pathway might prevent pleural organization. Conversely, advances in this field may help to develop novel pleurodesing agents for those patients with recurrent effusion or pneumothorax utilizing similar principles (e.g. LTA and staphylococcal superantigen).