Year in review 2016: Interstitial lung disease, pulmonary vascular disease, pulmonary function, paediatric lung disease, cystic fibrosis and sleep.

Abstract

Adelle S. Jee and Tamera J. Corte The year 2016 has been pivotal for researchers and clinicians with an interest in interstitial lung disease (ILD). The advent of anti-fibrotic therapy for idiopathic pulmonary fibrosis (IPF), the most common idiopathic ILD, has led to major changes to the approach towards the diagnosis and management of ILD. In appreciation of such changes, Respirology commissioned a timely series of reviews of the idiopathic interstitial pneumonias (IIPs), an important subgroup of the ILD. We review the original manuscripts from 2016, as well as highlighting some key points from this IIP review series. The increasing complexity of ILD classification and diagnostic approach underscores the importance of diagnosis by multidisciplinary discussion, including input from experienced pulmonologists, chest radiologists and lung pathologists. Jo et al. have recently demonstrated the impact of a dedicated ILD multidisciplinary team (ILD-MDT) review on clinical care. In the present study, the ILD-MDT at two specialist ILD referral centres led to a change in ILD diagnosis in 53% of patients with suspected ILD.1 The proportion of unclassifiable ILD was reduced from 42% to 12% following ILD-MDT. Importantly, 37% of patients referred with IPF had their diagnosis changed to another ILD and 8% had their diagnosis changed to IPF.1 As pharmacotherapies become increasingly targeted and the dangers of administering immunosuppression to IPF patients recognized, the ramifications of accurate diagnosis only further argue for the necessity of ILD-MDTs in the diagnostic approach to ILD.1 Accurate interpretation of HRCT findings remains pivotal in the assessment of ILD, and has major implications for diagnosis and management. In an interesting response to the review by Jacob et al. on this topic, Spagnolo et al. discuss the diagnostic challenges posed by an HRCT pattern of ‘possible’ usual interstitial pneumonia (UIP).2 The authors described the high proportion of patients with biopsy-proven UIP—thus affecting the diagnosis of IPF—despite HRCT abnormalities atypical for UIP in their study cohort.2 They suggest that in the appropriate clinical setting, an atypical or possible UIP pattern on HRCT should not exclude a diagnosis of IPF and that there is a need to continue to evolve the diagnostic criteria for IPF using real-life clinical data with the aim of safely reducing the number of patients undergoing surgical lung biopsy, without reducing diagnostic accuracy.2 The need for a gold standard to investigate for pulmonary embolism (PE) in patients with ILD is apparent. V/Q-SPECT (ventilation/perfusion single-photon emission computed tomography) and computed tomography pulmonary angiography (CTPA) are the two major modalities utilized in clinical practice to assess for PE. Leuschner et al. reported a cohort of ILD patients with suspected PE, and found a high rate of discordance between positive results as detected by V/Q-SPECT and CTPA.3 With studies now demonstrating increased mortality in IPF patients treated with warfarin without other indications for anticoagulation, and concerns that anticoagulation may provoke exacerbations of IPF, there is need for further validation of these findings.4, 5 Despite advances in medical imaging technology, up to 20% of patients still require some form of lung biopsy during the diagnostic workup of suspected ILD.6 Transbronchial cryobiopsy (TBCB) is a promising method of obtaining lung samples of a good size (average 40–50 mm2), with lower morbidity and mortality than surgical lung biopsy.7 The most common complications are pneumothorax (5–30%) and pulmonary haemorrhage (moderate in 40–56%).8-10 Echevarria-Uraga et al. described a series of 100 consecutive TBCB procedures whereby an angioplasty balloon was used for selective bronchial blockade to reduce bleeding (observed rate 13%), whilst preserving its high yield.10 Whilst TBCB continues to hold significant promise, its exact role in the diagnostic algorithm of ILD remains unclear. There is an ongoing and global search for a biomarker or panel of biomarkers that will allow early and specific diagnosis, and accurate prognostication of ILD patients. Whilst there are a wide variety of markers under development, the prognostic value of mucin 5B (MUC5B) and matrix metalloproteinase-7 (MMP-7) have more recently been validated in larger cohorts.11 van der Vis et al. investigated the prognostic implications of MUC5B allele carriage. Unlike prior studies, the authors separated truly sporadic IPF from familial interstitial pneumonia (IP), as well as idiopathic non-specific interstitial pneumonia (iNSIP) and connective tissue disease-associated ILD (CTD-ILD).12 The authors reported that MUC5B minor allele carriage conferred survival benefit in familial IP but not sporadic IPF, and was associated with worse survival in iNSIP.12 Zhang et al. analysed another biomarker, soluble L1-cell adhesion molecule (sL1-CAM), in bronchoalveolar lavage fluid (BALF) and serum in IPF patients.13 Compared with lung cancer and ‘chronic cough’ patients with no evidence of lung disease, sL1-CAM levels were higher in both BALF and serum in patients with IPF.13 It remains too early to determine the clinical utility of sL1-CAM or other candidate biomarkers, and it is essential that larger, longitudinal studies are performed to validate the most promising biomarkers from such exploratory studies. The new anti-fibrotic therapies, nintedanib and pirfenidone, have revolutionized the management of IPF and the tolerability and reduction in functional decline observed in clinical trials appear to have been borne out in real-life experience.14, 15 In the recent review by Borie et al., the authors highlight the myriad of unanswered questions with regard to anti-fibrotic therapy including when treatment should be stopped and how treatment failure is defined.14 With limited data to answer such questions, continuing efforts to formally study patients commencing anti-fibrotic therapy as well as to develop novel therapies and continue to enrol patients in clinical trials remain pertinent. ILD patients have significant symptom burden, and cough and dyspnoea have been demonstrated to negatively impact quality of life, be associated with depression and anxiety and even linked to increased mortality.16 Olukogbon et al. measured the breathing patterns in IPF patients and found that resting ventilation rate increased with worsening disease severity and declining lung function.17 These changes were also associated with increased dyspnoea and cough severity, thus providing a quantifiable link between symptoms and disease severity.17 A deeper understanding of tidal breathing patterns may inform the development of targeted approaches to breathlessness and a mechanism for monitoring disease progression in the future.17 Current approaches to breathlessness remain empirical and whilst there is growing evidence that pulmonary rehabilitation improves dyspnoea scores and quality of life, further research is still required to clarify the long-term benefits, optimal timing and protocols for exercise training programmes and the role of ambulatory oxygen in patients with exercise-induced hypoxia.18 Hanada et al. studied the effect of corticosteroids in 47 ILD patients on muscle strength and found an inverse correlation between total steroid dose and peripheral muscle strength.19 In COPD, the evidence suggests that skeletal muscle dysfunction is associated with impaired exercise capacity, quality of life and increased mortality. In ILD, there is little data exploring the role of skeletal muscle strength on longer term symptoms and outcomes. However, corticosteroids are often used in the treatment for non-IPF ILD. In an accompanying editorial, Mathur highlights the importance of further study to determine how best to balance corticosteroid dose to treat the underlying ILD pathology whilst not inducing myopathic changes.20 It appears that there is a close association between gastro-oesophageal reflux disease (GORD) and IPF, although the underlying pathogenic mechanisms of bile acid microaspiration remain unclear.21 Chen et al. investigated the farnesoid X receptor (FXR), a bile acid-activated nuclear receptor whose expression has been detected in human alveolar epithelial cells and pulmonary endothelial cells, and demonstrated activation of alveolar epithelial cells and lung fibroblasts via FXR-dependent and -independent pathways in experimental models.22 in vitro studies are required to determine the role of FXR in disease pathogenesis. Cardiovascular disease is also highly prevalent in IPF, and animal studies have suggested a possible pathogenic role of the renin–angiotensin system.23 Ekström and Bornefalk-Hermansson presented a prospective cohort of oxygen-dependent patients with pulmonary fibrosis, and found that angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs) were associated with improved survival.24 Whilst it is too early to make recommendations for clinical practice, we must continue our search for evidence-based life-prolonging treatment, inclusive of patients with severe disease. Lung transplantation remains the only definitive treatment for severe ILD refractory to medical therapy and it remains to be determined what the impact of more widespread anti-fibrotic use will have on the number of IPF patients listed and transplanted, and the timing, indications and outcomes of transplantation.25 Mesenchymal stem cells have been proposed as a novel therapeutic agent in IPF, with animal models demonstrating protective and reparative effects in experimental models of pulmonary fibrosis.26 However, with some studies showing progression of fibrosis if mesenchymal stem cells are administered in late stage disease, the safety and ideal timing of this potential treatment remains to be determined.26 Sgalla et al. reviewed IPF, the most common IIP, and highlighted the significant gains in our understanding of its epidemiology, pathophysiology, diagnosis and treatment since the original definition in 2002.27 However, there remains an unmet need to understand the mechanisms underlying accelerated disease and an accurate method of predicting future disease behaviour.27 iNSIP is now accepted as a distinct clinicopathologic entity, although its true incidence and prevalence remain to be determined.28, 29 The natural history of untreated iNSIP is unknown, and appears highly heterogeneous. Belloli et al. summarize the current management practices in their review of this topic, whereby immunosuppression is the mainstay in symptomatic or progressive disease.28 Major IIPs are now separated from rare IIPs and unclassifiable cases, the latter of whom comprise an important 10–15% of patients despite thorough investigation. Unresolved issues include lack of a uniform definition for unclassifiable ILD and paucity of evidence to guide management.30 Skolnik et al. suggest that unclassifiable ILD should remain a provisional descriptor and encourage continual re-evaluation for an underlying aetiology as new information, advancement in diagnostic techniques and the patient's clinical context evolve.30 Kokosi et al. discuss the rare IIP entities, lymphoid interstitial pneumonia (LIP) and idiopathic pleuroparenchymal fibroelastosis (PPFE), and rare histological patterns bronchiolocentric patterns of interstitial pneumonia (BPIP) and acute fibrinous and organizing pneumonia (AFOP).31 The clinical features and treatment implications of idiopathic LIP, BPIP and AFOP remain undetermined because of their low prevalence and paucity of data.31 Idiopathic PPFE, characterized by predominantly upper lobe pleural fibrosis and fibroelastosis of subjacent lung parenchyma, was first described in 1992 and acknowledged as an IIP in 2013.29 Baroke et al. challenge the notion of PPFE as an exclusively idiopathic disease in a case series of six PPFE patients, four with coexisting solid tumours.32 The present study together with previous reports of PPFE associated with a variety of other conditions support the need for larger studies to better characterize and investigate treatment options for this poorly understood patient group.32 Acute exacerbations are a significant cause of morbidity and mortality in IPF, but as discussed by Taniguchi and Kondoh in their review of acute and sub-acute IIPs, there is no established consensus definition for ‘acute exacerbations’ in IIP other than IPF, and no randomized controlled trials or treatment guidelines to direct management.33 The factors which predict exacerbations in IIP are not clear. Interestingly, Arai et al. have recently demonstrated that acute exacerbations were more frequent in patients with definite IPF on HRCT, compared with patients with inconclusive findings on HRCT.34 Cottin discusses the implications and importance of identifying underlying CTD in patients with ILD in a review of this topic.35 Patients with CTD-ILD are younger, more commonly female, non-smokers and have a better prognosis compared with IIP.36 However, there is increasing recognition that approximately 12–21% of ILD patients with CTD features but not fulfilling diagnostic criteria for a CTD may represent a distinct entity.36 In 2015, a European Respiratory Society/American Thoracic Society taskforce proposed consensus criteria for ‘interstitial pneumonia with autoimmune features’ (IPAFs) in an important first step to allow research on a uniform cohort.37 It remains to be determined what the disease behaviour and impact on prognosis and management of IPAF will be, and longitudinal, multicentre prospective studies are urgently needed. Systemic vasculitides may also be associated with ILD. Hosoda et al. investigated the clinical, radiological and pathological features of UIP associated with myeloperoxidase anti-neutrophil cytoplasmic antibody (MPO-ANCA) (ANCA/UIP). Differentiating ANCA/UIP from IPF/UIP has important implications for treatment with regards to immunosuppression, yet differentiating ANCA/UIP from IPF/UIP is traditionally difficult. In their cohort, ANCA/UIP patients had higher percentages of neutrophils detectable in BALF compared with IPF/UIP patients, and the authors conclude that further validation of features that can differentiate these two entities may help stratify which patients may benefit from immunosuppressive therapy.38 Margaritopoulos et al. review the newly classified smoking-related IIPs—respiratory bronchiolitis ILD (RB-ILD) and desquamative interstitial pneumonia (DIP)—and early changes in current smokers termed ‘interstitial lung abnormalities’ (ILAs).39 Cigarette smoking is implicated in the pathogenesis of some specific IIP, although the exact mechanisms are not well understood. Smoking cessation is crucial in the management of both ILA and smoking-related ILD, but the role of screening tests in smokers and the therapeutic and prognostic benefit of identifying a distinct group of smoking-related IIP remain to be evaluated.39 Drug-induced ILD is a complex diagnosis because of the large number of medications that can be associated with toxicity, as well as confounding factors such as infection, pulmonary oedema, radiotherapy and underlying malignancy which need to be excluded. ILD is a widely recognized complication of gemcitabine-based chemotherapy, with high associated mortality but little available data to assist risk stratification.40 In a population-based study by Hamada et al., the incidence of ILD amongst 25 924 patients undergoing gemcitabine-based chemotherapy was 1.7%.40 Identified risk factors included older age, heavy smoking, prior chemotherapy, advanced cancer stage and lung cancer.40 Further prospective studies are needed to better stratify risk prior treatment with chemotherapeutic agents associated with pulmonary toxicity. Lee et al. investigated the incidence of late-onset non-infectious ILD following autologous haematopoietic stem cell transplantation (HSCT).41 In a cohort of 340 paediatric patients, the incidence of ILD after high-dose chemotherapy (HDCT) and autologous HSCT for solid tumours was 2.5% with high mortality (six out of eight patients died during a median 4.5-year follow-up period).41 Female gender and cyclophosphamide were significant predictors of ILD.41 With such high mortality, larger studies are required to validate these findings and develop guidelines to assist the diagnosis and management of late-onset ILD after HDCT for HSCT. It has been an exciting year in ILD with major changes to our understanding of the optimal diagnostic process, and treatment options for ILD patients. We look forward to continuing efforts to accurately phenotype ILD patients, perhaps with a panel of candidate biomarkers, and wait with anticipation as the global search for definitive therapies to improve or stabilize disease and optimize symptoms continues. Stephen John Wort Idiopathic pulmonary arterial hypertension (IPAH) is an extremely rare condition associated with raised pulmonary artery pressure (≥25 mm Hg) in the context of normal left heart filling pressures (≤15 mm Hg) with no other discernible cause.42 The pathology involves remodelling of small, pre-capillary arteries (<500 µm), resulting in an increase in pulmonary vascular resistance, right heart failure and premature death. Around 1 in 10 patients will have a family history and in 80% of these patients there is a mutation in bone morphogenetic protein receptor 2 (BMPR2). Treatment of IPAH/HPAH (heritable pulmonary arterial hypertension) to date has concentrated on the associated pulmonary vasoconstriction and not on remodelling itself. It is therefore not surprising that outlook remains poor. Two articles (from the same group) and an editorial discuss an alternative approach, to target remodelling itself, by promoting wild-type BMPR2 expression.43-45 In the first, in a mouse model of pulmonary arterial hypertension (PAH) induced by doxycycline-driven expression of a dominant-negative BMPR2 mutation (R899x), the hypertensive phenotype is rescued by adenovirus targeted delivery of wild-type BMPR2.43 In the second, using the same mode of delivery, the authors demonstrate enhancement of the downstream, signalling molecules, Smads 1, 5 and 8, in normal human pulmonary microvascular endothelial cells. In fact, the same authors have now shown effective delivery of wild-type BMPR2 in three independent animal models of pulmonary hypertension. Exciting as this is, the next stage will be harder; translation into man and, in particular, when to treat and what cells to target. It should also be appreciated that only 20% of carriers of mutations in BMPR2 actually develop PAH, with a second ‘hit’ presumably needed, such as hypoxia or inflammation. There have also been related papers of note in other journals targeting this pathway; Long et al. have successfully used the natural ligand for BMPR2 and bone morphogenetic protein 9 (BMP9) to prevent and treat an animal model of PAH46 and also shown that it is possible to prevent mutation-dependent lysosomal degradation of BMPR2 with chloroquine.47 As mentioned, patients with IPAH have a terrible prognosis. Methods aimed at measuring response to treatment, as well as prognosis, are clearly important. Shi et al. examined the use of cardiopulmonary exercise testing (CPET) and, in particular, oxygen uptake efficiency (OUE) measures in patients with IPAH and chronic thromboembolic pulmonary hypertension (CTEPH).48 OUE has the advantage that it is effort independent, unlike peak oxygen uptake (VO2), and has been found to correlate with disease severity and early mortality in patients with heart failure.49 The authors found that OUE was significantly higher in patients with IPAH compared with CTEPH. It also correlated to disease severity and exercise capacity. The difference between IPAH and CTEPH probably relates to the effect of proximal vascular occlusion and consequent poorer ventilatory efficiency as described previously by Zhai et al.50 Another rare condition, and poorly understood, is isolated pulmonary capillaritis (IPC). IPC is defined as a capillaritis in the absence of an underlying systemic condition. Thompson et al. report the Mayo Clinic experience of 36 biopsy-proven cases of pulmonary capillaritis over a 17-year period.51 Most were secondary to systemic conditions such as granulomatosis with polyangiitis, only four were considered IPC. All patients presented with sub-acute haemoptysis and diffuse alveolar haemorrhage and all responded well to immunosuppression. Unfortunately, there were no obvious risk factors and diagnosis is often delayed so physicians need to remain vigilant. Finally, de Miguel-Díez et al. report on a much more common condition, PE. Using the Spanish National Hospital Database, and identifying 123 872 patient discharges with PE, they found that type 2 diabetes is associated with a high incidence of hospitalization for PE. Possible explanations are the association with other cardiovascular risk factors or direct effects more related to diabetes itself, such as chronic inflammation and decreased expression of protective endothelial factors.52 Neil D. Eves Spirometry is the cornerstone of pulmonary function testing and globally remains the gold standard for the diagnosis of respiratory disease. However, there is continued need for revision of techniques and standardization of methodology to improve the accuracy and validity of results. This year in Respirology, a position statement was published from the Australian and New Zealand Society of Respiratory Science53 regarding the utilization of reference values for spirometry. The authors rationalize why the Global Lung Function Initiative 201254 references were chosen, discuss the challenges and limitations of changing to new reference values and suggest strategies for accurately interpreting tests. While changing reference values present a number of challenges for respiratory centres, this is an important step to reduce the misinterpretation of results and standardize spirometry across Australasia.53 While pulmonary function testing has traditionally been used to monitor progressive respiratory conditions, these tests can be difficult to perform and coordinate, especially for children and patients with complex respiratory and neuromuscular conditions. As such, there is a need for novel, non-invasive tests of pulmonary function that can provide accurate clinical information. Using diaphragmatic ultrasound, Fantini et al.55 assessed diaphragmatic function in patients with amyotrophic lateral sclerosis (ALS), a progressive degenerative disease that results in loss of motor neurons and the progressive loss of muscle function including the respiratory muscles.56 The authors report that changes in a new diaphragm thickness index significantly correlated with lung impairment and demonstrated better accuracy than forced vital capacity and the sniff nasal inspiratory pressure test. As such, this index may allow early detection of diaphragm weakness and could help to guide clinicians of when to start ventilatory support in ALS patients. As highlighted in last year's review,57 the forced oscillation technique (FOT) for the assessment of pulmonary function has grown in popularity especially in paediatric populations who have difficulty performing pulmonary function manoeuvres.58 Ioan et al.59 investigated the utility of FOT for the diagnosis of vocal cord dysfunction (VCD) in children. VCD is associated with the paradoxical adduction of the vocal cords and is commonly associated with, or mistaken for, asthma because of the similarity of symptoms.60 As presented in the accompanying editorial,60 a previous study utilizing a mechanical model of VCD,61 demonstrated that the within breath respiratory resistance (Rrs) measured with FOT was significantly increased during inspiration. A similar observation was reported by Ioan et al.59 in children with VCD during exercise as Rrs was increased during inspiration when VCD was present. This phenomenon is not observed during breathing in healthy individuals or in those with asthma. These findings are promising and larger studies are now needed to confirm the clinical utility of FOT for accurately diagnosing VCD in both children and adults. In another study utilizing exercise testing,62 Johansson et al. performed strenuous exercise with dry air inhalation to investigate the time course of the decline in forced expired volume in 1 s (FEV1) in 146 adolescents with exertional dyspnoea.62 The authors reported an interesting temporal variation in the bronchoconstrictory response to exercise. More specifically, approximately half of those with exercise-induced bronchoconstriction (EIB) had the traditional reduction in FEV1 peaking 5–10 min post-exercise, while the other half demonstrated a delayed response with the greatest reduction occurring at 15–30 min after exercise. The authors conclude that when using dry air ventilation to diagnose EIB, lung function should be measured for at least 30-min post-exercise. There is growing recognition that the origins of COPD may relate back to smoking exposure in early childhood.63 Utilizing the fifth decade follow-up of the Tasmanian Longitudinal Health Study Cohort, Perret et al.64 examined the relationship between maternal smoking and airflow obstruction when subjects were 7 years and again in middle age. The authors reported a 2.7-fold higher association between heavy maternal smoking and spirometrically determined COPD compared with those not exposed. They also report that maternal smoking appears to augment the adverse effect of personal smoking on the gas transfer factor. These findings suggest that early life exposure to smoking predisposes individuals to the development of COPD, especially if they also smoke later in life.64 Claire E. Wainwright Airway microbiology has been a dominant theme this year in the cystic fibrosis (CF) world. At the beginning of the year, Shah et al.65 helped our understanding of early airway infection in CF by examining the differences between CF pigs and humans that both have early airway infection and CF mice that do not.65 CF pigs and humans both have abnormal airway acidification related to the H+ secretion by the non-gastric H+/K+ ATP12A which acidifies the airway surface liquid, and CF mice that have little ATP12A and secrete little H+ and consequently have less airway infection suggesting that ATP12A may be an important therapeutic target for CF. Salsgiver et al.66 performed a retrospective analysis of changing prevalence of selective airway pathogens in patients with CF from the CF Foundation data registry between 2006 and 2012. Pseudomonas aeruginosa and methicillin-susceptible Staphylococcus aureus (MSSA) remain the most prevalent pathogens, although there was a decreasing incidence and prevalence of P. aeruginosa and Burkholderia cepacia complex and increasing prevalence of non-tuberculous mycobacterial (NTM) species over time.66 The changing prevalence was thought to be potentially related to multifactorial factors including changes in clinical practice. Widely varying treatment practices remain between CF centres, and changes in clinical practice also seem to be related to an emerging problem of multi-antibiotic resistant P. aeruginosa (MARPA) now being seen in CF centres worldwide and complicating the management of patients. In an Australian study using data from the Australian CF data registry and the Australian Clonal P. aeruginosa in CF study, Smith et al.67 reported that in paediatric centres, MARPA prevalence was associated with the state in which treatment was delivered and use of i.v. antibiotics, whereas in adult centres, body mass index and the state were associated with MARPA.67 The implications of the present study and importance of registries combining data to better understand these issues was also discussed by Jaksic and Vyas in an editorial.68 Internationally, there has been an increasing concern around the increasing prevalence of NTM infections and especially infection with Mycobacterium abscessus species which previously were thought to be environmentally acquired. Bryant et al. recently reported in a study using whole-genome sequencing of isolates from a global collection of clinical isolates that the majority of M. abscessus isolates were dominant circulating clones acquired by transmission between individuals either by fomites or by infectious aerosols and that these clones are associated with worse clinical outcomes and have increased virulence in mouse models thus presenting a serious emerging clinical concern.69 Identification of NTM is challenging and infection may also easily be missed as organisms are easily overgrown with other bacteria and fungi, and B. cepacia agars have previously been recommended to help with identification. Preece et al. have recently reported on their development of a novel agar (RGM medium) that will greatly reduce overgrowth problems and improve the sensitivity of NTM identification.70 In addition, while clinical guidelines for the management of NTM have been published this year in Thorax, unfortunately there remain no high-quality clinical trials on which to base treatment protocols71 further highlighting the importance of collaboration between groups to tackle these challenging infections. The paucity of randomized controlled trials continues to limit the available evidence around physiotherapy techniques for airway clearance in CF because of potential ethical concerns around withholding a treatment that is part of established standards of care; however, a comprehensive clinical practice guideline of physiotherapy for CF was published by Respirology this year and 30 recommendations were highlighted based on the best evidence available.72 Of the many recommendations perhaps the greatest evi