Congenital defects of large airways

• Congenital defects should be considered in all patients with bronchiectasis. [D]

Foreign bodies and aspiration

• Gastric aspiration should be considered as a cause in all patients. [D]

What is the current relevance of previous severe lower respiratory tract infections to patients with bronchiectasis?

• A history of previous severe lower respiratory tract infections due to bacterial and viral pneumonia, pertussis or tuberculosis should be sought in all patients with bronchiectasis. [C]

• Where possible, the temporal relationship of identified infections to the onset of chronic respiratory symptoms should be determined. [D]

Mycobacterium tuberculosis and opportunist mycobacteria

• All patients with repeated isolates of opportunist mycobacteria should have regular follow-up in secondary care. [D]

Immune deficiency and bronchiectasis

• The possibility of underlying immune deficiency, particularly antibody deficiency, should be considered in all children and adults with bronchiectasis. [A]

• Serious, persistent or recurrent infections, particularly involving multiple sites, or infections with opportunist organisms should raise the suspicion of immune deficiency. [D]

• The possibility of symptomatic or clinically silent bronchiectasis should be considered as a potential complication in all patients with immune deficiency, particularly primary antibody deficiency. [D]

• In patients with immune deficiency and patients with bronchiectasis, features in the history or clinical examination which may support the coexistence of both conditions should be considered and adequately assessed. [D]

• In patients with suspected or proven immune deficiency and bronchiectasis in combination, specialist aspects of diagnosis, monitoring and management should optimally be provided within a shared specialist care arrangement (joint working between chest physician and immunologist). [D]

What is the relationship of other airway diseases to bronchiectasis?

What is the relationship of bronchiectasis to cystic fibrosis?

• For all patients with bronchiectasis, the possibility of underlying cystic fibrosis should be considered (see section 3). [D]

Which connective tissue disorders are associated with bronchiectasis?

• A history of rheumatoid arthritis should be sought in all patients with bronchiectasis. [D]

• Closer follow-up of patients with rheumatoid arthritis-related bronchiectasis is warranted in view of a poorer prognosis. [C]

Inflammatory bowel diseases

• Bronchiectasis should be considered in patients with inflammatory bowel disease who develop a chronic productive cough. [D]

Disorders of ciliary function

• In all children with bronchiectasis, a detailed history of the neonatal period should be taken. [D]

• In children and adults with bronchiectasis, a history of chronic upper respiratory tract problems, particularly otitis media, should be sought. [D]

• Adults should be questioned about any history of infertility. [D]

Is α₁-antitrypsin deficiency a cause of bronchiectasis?

• Routine screening for α₁-antitrypsin deficiency is not required unless the radiological investigations suggest basal emphysema. [D]

Yellow nail syndrome

• The assessment of patients with bronchiectasis should include a search for features of yellow nail syndrome. [D]

The upper respiratory tract in patients with bronchiectasis

• Every patient with bronchiectasis should have an assessment of upper respiratory tract symptoms. [D]

Who to investigate for bronchiectasis

Clinical presentation of bronchiectasis

Investigations directed at underlying cause

Radiological investigations

Sputum microbiology

Lung function tests

General approach and treatment of the specific underlying cause

• Identify and treat underlying cause to prevent disease progression. [D]

• Maintain or improve pulmonary function. [D]

• Reduce exacerbations. [D]

• Improve quality of life by reducing daily symptoms and exacerbations. [D]

• In children, achieve normal growth and development. [D]

• Patients with primary or secondary immune deficiency should be under joint care with a clinical immunologist. [D]

• Patients with cystic fibrosis should be referred to a cystic fibrosis specialist centre. [D]

Role of primary care

Role of nurses

Physiotherapy: airway clearance techniques and exercise

Airway pharmacotherapy

Defining and managing exacerbations

Antibiotic resistance

Surgery for bronchiectasis

Reason for BTS bronchiectasis guideline

The diagnosis, investigation and particularly management of bronchiectasis has been largely empirical and the subject of relatively few controlled clinical trials. There are no clear guidelines, although an Australian position statement has been published concerning bronchiectasis in children.212

The purposes of these guidelines were therefore threefold: (1) to identify relevant studies in non-cystic fibrosis (CF) bronchiectasis; (2) to provide guidelines on management based on published studies where possible or a consensus view; and (3) to identify gaps in our knowledge and identify areas for future study.

Guideline group members

The membership of the BTS Bronchiectasis (non-CF) Guideline Group is as follows: Dr Mark C Pasteur, Dr Diana Bilton, Dr Adam T Hill, Professor Andrew Bush, Dr Charles Cornford, Dr Steven Cunningham, Dr Xavier Emmanuel, Jane French, Dr Mike Greenstone, Professor David M Hansell, Alex Harvey, Dr Richard Herriot, Karen Heslop, Dr Pota Kalima, Frances Sinfield, Dr Samantha Sonnappa, Dr David A Spencer, Professor Robert A Stockley, Lorna Willcox, Dr Robert Wilson, Mr G Wyn Parry. The assistance of Julia Bott, Jennifer Pryor and Dr Colin Wallis is gratefully acknowledged. The full list of contributors for each section of the guideline is given in Appendix 4.

How has the guideline been designed?

The guideline is divided into sections covering different aspects of the management of the condition. Guidance for children and adults is dovetailed together throughout to avoid repetition while acknowledging differences between these groups. Areas of particular or sole relevance to one or other of these groups are indicated. Sections 2 and 3 cover the background, clinical assessment and investigation of patients (including appropriate radiological and laboratory investigations). The principles and broad approach to management are discussed in Section 4 including recommendations for physiotherapy and non-antibiotic drug treatment. The use of antibiotics is covered in Section 5 and surgery and the management of advanced disease is covered in Section 6.

Definition

This guideline refers to the investigation and management of patients with symptoms of persistent or recurrent bronchial sepsis related to irreversibly damaged and dilated bronchi—namely, clinical bronchiectasis. It does not cover the management of cystic fibrosis (CF) and, for the purposes of the guideline, ‘bronchiectasis’ is synonymous with the term ‘non-CF bronchiectasis’. Likewise, it does not focus on traction bronchiectasis secondary to other lung pathologies, particularly the interstitial lung diseases, which is commonly asymptomatic.

Methods

A literature search was performed using the following databases: Pubmed, Cinahl, Embase and AMED using the combined search terms ‘bronchiectasis’ and ‘not CF’ which resulted in 1803 references published in English. The abstract for each reference was retrieved and reviewed by each of the three members of the steering committee. The steering committee then decided, on the basis of the abstract, whether the full paper should be reviewed. Papers were the assigned to one or more of the three working groups based on their relevance to each section and read by each member of the group. Those papers with information relevant to the guideline were included in the final document. The methods for assessing the level of evidence for each paper and the grading of recommendations were those developed by the Scottish Intercollegiate Guidelines Network (SIGN) and as used in the British Thoracic Society (BTS)/SIGN British guideline on the management of asthma (see table 1). The evidence level for each paper is given at the end of each text section and the grade of recommendation follows each recommendation statement. This guideline is due for revision in 2011.

How common is bronchiectasis in adults and children in the 21st century?

The incidence of bronchiectasis in a given community is largely unknown. There is a general belief that the incidence is falling, although the evidence for such a belief is unclear. Several studies of hospital admission with bronchiectasis have shown a reduction since the 1950s.20 478 479 Most of this change has been attributed to the introduction of antibiotics and, for this reason, bronchiectasis is no longer considered a major healthcare problem.

However, the incidence, when assessed, does vary widely between populations from 3.7/100 000 children in New Zealand480 to 52/100 000 adults in the USA.481 In the UK there are no recent studies, although mass chest x-ray features of bronchiectasis in the 1950s suggested a prevalence of 100/100 000.482 The prevalence increases with age.481 Some of these population studies, many of which were conducted years ago, have not included modern diagnostic techniques and specifically high-resolution CT scanning (HRCT). The importance of this issue is that, whereas the disease severity may have shown a gradual shift, the true incidence remains unknown in most populations. For instance, the pathological changes of bronchiectasis have been identified using HRCT in up to 15–30% of patients diagnosed in primary care with chronic bronchitis and chronic obstructive pulmonary disease (COPD).219 225 This may represent a new case load of patients with clinically significant bronchiectasis or a mild pathological change in patients whose primary condition requires standard therapy alone. In children, non-CF-related bronchiectasis was identified in 1% of all secondary care referrals in a UK population.8

What are the pathology and underlying causes?

Bronchiectasis is a persistent or progressive condition characterised by dilated thick-walled bronchi. The symptoms vary from intermittent episodes of expectoration and infection localised to the region of the lung that is affected to persistent daily expectoration often of large volumes of purulent sputum. Bronchiectasis may be associated with other non-specific respiratory symptoms including dyspnoea, chest pain and haemoptysis, and may progress to respiratory failure and cor pulmonale.

The underlying pathological process is damage to the airways which results from an event or series of events where inflammation is central to the process. This is easy to understand as part of the ‘vicious circle’ hypothesis which has been applied to bronchiectasis and has been the major factor influencing current disease management.

The lung is continuously exposed to inhaled pathogens and (in many countries) environmental pollutants. The lung has a sophisticated primary and secondary defence system that maintains sterility of the normal lung. If this defence system is breached as in disorders of mucociliary clearance or specific antibody deficiencies, the lung becomes susceptible to infection, colonisation (the persistence of bacteria in the lower respiratory tract) occurs and the subsequent inflammation that causes airway damage further impairing host defences. Thus, once established, the defective defences can lead to a self-perpetuating cycle of events that facilitate bacterial colonisation and airway sterility becomes unlikely.

Primary defects in the lung defences are uncommon in patients investigated as adults, suggesting they are either subtle or do not influence the primary event. However, immunodeficiency may be more common when bronchiectasis presents in childhood.232 Episodes causing clear lung damage such as previous pneumonias, gastric aspiration or viral illnesses in childhood would represent such initiating events, although recent evidence suggests these may be less common. The damage to the airway by such episodes would particularly impair the normal mucociliary function and hence clearance of inhaled pathogens initiating the inflammatory cycle.

However, despite many studies over the years using modern immunological techniques, not only have few primary deficiencies of host defences been found but up to 40% of patients do not even have a clear defining event that appears to initiate the process.

What is the outlook for these patients?

The long-term outcome is also generally regarded as good, although this tends to be taken as little progressive loss of lung function with no effect on mortality. Indeed, studies that have been conducted generally support this view. Again, most of the information is from historical data and suggests that antibiotic therapy has had an effect. For instance, in the 1940s most patients diagnosed with bronchiectasis died before the age of 40 years but, by the 1960s, the average age of death had risen to 55 years. Nevertheless, this still indicates a significant reduction in life expectancy in patients with bronchiectasis. More recent data suggest a better prognosis,373 although it is recognised that the general health of patients with bronchiectasis can be poor251 and certain subsets (particularly those colonised with Pseudomonas) are particularly affected, with continued ill health and progressive deterioration.

Causes and associations

Good practice point

• Underlying causes of bronchiectasis should be assessed in all patients.

Congenital defects of large airways

Congenital bronchiectasis is much rarer than previously considered. Specific causes include Williams–Campbell syndrome (bronchial cartilage deficiency),7 11 12 tracheobronchomegaly (Mounier–Kuhn syndrome),1 6 Marfan’s syndrome,10 late presenting H-type tracheobronchial and oesophagobronchial fistula.3 Bronchiectasis has been reported in congenital lung malformations such as sequestration2 8 or rarely with a rib malformation.9 Familial causes of undefined aetiology have been described.4 5 It should be remembered that congenital defects, particularly tracheobronchomegaly, may first present in adulthood.1 6

Foreign bodies and aspiration

What is the current relevance of previous severe lower respiratory tract infections to patients with bronchiectasis?

Although never subjected to a formal case–control study until recently,39 numerous case series from the last century point to the importance of lower respiratory tract infection in the aetiology of bronchiectasis, the most frequent association being with bacterial pneumonia.20 52 53 Pertussis,20 51 53 pulmonary tuberculosis,20 47 48 53 mycoplasma40 51 and viral pneumonia (particularly adenoviruses41 50 217 and measles20 43 52 but also influenza44 and respiratory syncytial45 viruses) have, in addition, all been linked directly to permanent lung damage and bronchiectasis. Post-tuberculous bronchiectasis is often segmental or lobar in the area primarily affected.47 48 Studies of both children and adults in the mid part of the 20th century found that severe lower respiratory tract infection was the most frequently identified cause, ranging from 41% to 69% of cases.20 52 53

The importance of asking about a history of such infections with particular note taken of early childhood illnesses is therefore clearly relevant in assessing patients with bronchiectasis. If the infectious insult can be directly linked to the onset of chronic respiratory symptoms, then the link between the two can be taken with more certainty. Past and recent studies indicate that many patients are able to recollect such a direct link.46 53 These patients may require a less intensive search for other causes. A difficulty arises if there is a significant gap between an identified infection and the onset of chronic respiratory symptoms years later with a period of normal health in between. Can symptomatic presentation of bronchiectasis be delayed years after the original insult? One study which looked at the sequelae of adenovirus pneumonia suggests that it can.217

With the decline in the incidence of pneumonia and other infections, particularly in children, these aetiologies might be expected to be less relevant in current patients. While this is indeed the case, three recent studies looking at populations with mean ages >50 years have shown that it is still an important cause with 28%,54 29%22 and 42%46 of patients with this aetiology. One study which investigated predominantly young adults with bronchiectasis found only 6% had an infectious cause,49 suggesting less relevance in this age group. However, in children, a post-infectious aetiology was suspected in 47% of one UK series23 and 25–34% in two non-UK series.19 42 A single case–control study of Australian children has confirmed that pneumonia requiring hospitalisation is linked to the onset of bronchiectasis, with a particular association with severe or recurrent pneumonia.39 In children from deprived communities there is a high incidence of post-pneumonic bronchiectasis.19 39 42

Good practice point

• Identifying a post-infectious cause may limit the need for further investigations, particularly in elderly subjects.

Mycobacterium tuberculosis and opportunist mycobacteria

Bronchiectasis may result from pulmonary Mycobacterium tuberculosis infection, with the incidence reflecting the prevalence of tuberculosis in the population.37 42 48 54 56 It is also increasingly recognised that opportunist mycobacteria are associated with localised or widespread bronchiectasis.57 60 Bronchiectasis, like other forms of lung damage, makes patients prone to picking up environmental mycobacterial species and bronchial damage may occur as a result of opportunist mycobacterial infection. Opportunist mycobacteria have been isolated in 2%59 and 10%58 of random sputum specimens from patients with bronchiectasis, but the clinical significance is unclear. Patients with Mycobacterium avium complex infection may develop bronchiectasis over years. Middle-aged or elderly women seem particularly prone to this disease.57 However, isolation of an opportunist mycobacterial species should not necessarily be interpreted as pathogenic. A ‘one-off’ isolate may have been inhaled shortly before the sample was provided. Persistent isolation (colonisation) may occur without any change in clinical status. HRCT scan features can be helpful in confirming infection. One series of adults with primary ciliary dyskinesia (PCD) found repeated isolation of opportunist mycobacteria in 5% of cases.55

Once an opportunist organism has been isolated, prolonged follow-up may be required to decide whether this represents colonisation or infection. Careful follow-up is mandatory because colonisation can change to infection. This will include clinical features (deterioration favouring infection), sputum examination (repeated culture, smear positive, heavy growth), lung function (rapid decline), HRCT (exudative ‘tree-in-bud’ bronchiolitis, mucus plugging, cavitating nodules, rapid progression of disease), as well as failure to respond to usual treatment. The species isolated will also influence the likelihood of infection (M avium complex,60 61 M kansasii,60 M malmoense62).

Immune deficiency and bronchiectasis

Bronchiectasis can complicate most defined primary8 22 68–73 77 90–98 484 and secondary66 67 74–76 79 81 82 84 86 87 immune deficiency disorders. The mechanism is presumed to involve defective immune clearance with repeated, persistent or severe infection leading to recurrent episodes of airways inflammation, regeneration, repair and ultimately structural damage.99 The most frequent—and clinically most important—association between bronchiectasis and underlying immune deficiency occurs with primary antibody deficiency syndromes, a link that has been recognised for almost 50 years.100 With other primary and secondary immune defects, development of bronchiectasis—although important on an individual disease and case basis—is less significant in absolute numbers and, in many published series and cases, associated only with poorly characterised defects of immunity.

Chronic suppurative lung disease and bronchiectasis constitute major causes of morbidity and mortality in patients with primary antibody deficiency disorders.91–93 101 Such disorders encompass a heterogeneous group of conditions characterised by defective production or function of all immunoglobulin classes, individual classes or subclasses, defects in production of antibodies against specific antigens such as bacterial capsular polysaccharides102 or combinations of these patterns. The presentation of immune deficiency may occur initially in adult life and is not confined to infancy or childhood.92 Bronchiectasis occurs, with varying degrees of frequency, as a complication in all forms of the condition from severe panhypogammaglobulinaemia to subtle defects of specific antibody production. The three most commonly encountered disease variants are common variable immune deficiency (CVID), X-linked agammaglobulinaemia (XLA) and IgA deficiency. Bronchiectasis is reported as an established disease complication in 18–68% of patients with CVID85 93–96 103 and 7–20% of patient cohorts with XLA,93 95 104 and is therefore a significant characteristic of these patient groups. Bronchiectasis appears to complicate isolated selective IgA deficiency relatively rarely,105–107 but may occur at greater frequency when IgA deficiency is part of, or evolves into, a more clinically significant and complex antibody deficiency disorder (specific antibody/IgG subclass deficiency or CVID).108 109 In addition to these more common antibody deficiency disorders, there is an established and growing recognition of an important association between bronchiectasis and defects of specific antibody production.22 71 83 88–90 Additional disease co-factors, such as α₁-antitrypsin deficiency, may play a role which is contributory to, and cumulative with, infection-associated airway damage in some immunocompromised patients.78 80 The frequency and importance of such interactions remain to be fully elucidated, particularly in progressive disease which responds poorly to treatment. It has been suggested that selective antibody deficiency with recurrent respiratory infections may account for the recognised occurrence of bronchiectasis in yellow nail syndrome.118 Significant antibody deficiency confers a particular susceptibility to mucosal infection with encapsulated organisms. Recurrent upper and lower respiratory tract infections with Streptococcus pneumoniae, Haemophilus influenzae (frequently untypable) and Moraxella catarrhalis are characteristic of antibody deficiency in both children and adults.63 94 96 114 117 119–121 This is seen before diagnosis/institution of therapy but also as breakthrough episodes in patients receiving immunoglobulin replacement therapy.485

An additional facet underlining the relationship between antibody deficiency and bronchiectasis is the consistent identification of the former as a significant aetiological factor in large-scale cohort studies of patients who have bronchiectasis of undefined aetiology (BUA) where other causative factors have been excluded.8 22 68 71 73 83 88 89 252 486 The more recent of these studies, using modern immunological diagnostic techniques, have shown rates of definable antibody deficiency in BUA ranging between 6% and 48%,8 22 71 73 83 89 although the precise clinical significance of many of the disorders described is uncertain and probably limited. Specific defects identified have encompassed the full range of clinical antibody disorders with IgG subclass deficiency being the most commonly described finding. If such deficiencies of IgG subclasses are excluded in light of their limited direct clinical significance,90 111 112 the minimum frequency with which significant underlying antibody deficiency contributes to bronchiectasis as a whole, and to BUA, is about 5% and 10% respectively.

The importance of detecting bronchiectasis in patients with clinically significant antibody deficiency and of actively considering the possibility of antibody deficiency as an underlying factor in patients with bronchiectasis is emphasised by the high incidence of established bronchiectasis (silent or clinically apparent) by the time that compromised immunity is recognised. A significant delay in the diagnosis and treatment of antibody deficiency and underdiagnosis within the population is common in the UK.95 101 113 A diagnostic delay of >2 years is associated with an increased risk of developing bronchiectasis in antibody-deficient patients,95 and the strongest predictor of chronic progressive pulmonary disease in antibody-deficient patients, even after starting treatment, is established lung disease at the time of initial presentation.114

Early identification of antibody deficiency and effective therapy are essential factors in preventing the development of bronchiectasis or in retarding progression of established disease. Effective immunoglobulin replacement, especially at higher doses, can substantially improve pulmonary function in hypogammaglobulinaemia.63–65 104 115 In some patients, however, even adequate immunoglobulin replacement does not prevent the silent insidious progression of bronchiectasis.116 Such patients may require higher than standard doses of immunoglobulin, although other aspects of management may also have central importance in retarding the advance of lung damage. The optimal dose of immunoglobulin for replacement therapy in patients with hypogammaglobulinaemia, whether or not complicated by bronchiectasis, is not defined and is best determined on the basis of individual clinical response to treatment and development of disease complications rather than on the basis of an arbitrary target serum level of IgG. Central to effective care of patients who have bronchiectasis associated with immune deficiency is increased awareness of the close link between these conditions. This should be accompanied by development of robust pathways which facilitate access to diagnostic and specialist expertise and which encourage integrated assessment, monitoring and treatment of this complex patient group by both respiratory and immunology teams.

What is the relationship of other airway diseases to bronchiectasis?

What is the relationship of bronchiectasis to cystic fibrosis (CF)?

As recurrent lower respiratory tract infection is a feature common to both CF and non-CF-related bronchiectasis, consideration should be given to the possibility of CF being the cause in all patients found to have bronchiectasis. The importance of this has been emphasised with recent findings that atypical CF may present with pulmonary problems in the absence of other manifestations such as pancreatic failure, gastrointestinal disease or raised sweat chloride levels.156 In children presenting with bronchiectasis, CF will be high on the list of differential diagnoses42 and late first presentation of CF in adults as bronchiectasis has been described repeatedly.22 151 152 Most will be pancreatic-sufficient, and even those who are not will rarely have gastrointestinal symptoms.152 157

The prevalence of CF in patients presenting with bronchiectasis has been examined in a number of studies, most of which used a combination of sweat chloride measurement and genetic screening for CF transmembrane regulator (CFTR) mutations. In three unselected adult cohorts CF, diagnosed on the basis of homozygosity for a CFTR mutation or heterozygosity for a mutation and a raised sweat chloride, was found in 0 of 100 patients (0%),153 46 of 601 (7.6%)152 and 4 of 150 (2.6%). Some studies have focused on patients with bronchiectasis of unknown cause, including a study of children in which 7% were found to be homozygous for CFTR mutations.148 Some,147 149 150 154 but not all,155 studies of small numbers of adults with disseminated idiopathic bronchiectasis have found an increased frequency of CFTR gene mutations and polymorphisms compared with that expected for the local population, the significance of which is not yet understood.

Which connective tissue disorders are associated with bronchiectasis?

Bronchiectasis has been identified in many case series of patients with connective tissue diseases, the subject of a comprehensive recent review,169 which span the period of time before, during and after the emergence of HRCT as the definitive diagnostic modality. While all studies can be criticised for lacking rigorous control populations, possible bias related to patient selection or failure to exclude other potential causes of bronchiectasis, a clearer picture is emerging, particularly for the association with rheumatoid arthritis (RA).

Studies screening on the basis of symptoms of lower respiratory tract infection in large cohorts of patients with RA have found an incidence of bronchiectasis of 3.2%,171 5.2%172 and 2.9%160 (all greater than expected contemporaneous population frequency), two of these studies being particularly powerful in assessing patients at first presentation with RA.160 171 Several secondary care RA populations have been subjected to HRCT studies. When airway involvement was suspected, bronchiectasis was seen in 30%166 and 22%,168 perhaps reflecting selection bias, but three studies assessing unselected secondary care populations have also found HRCT evidence of bronchiectasis in 30%163 167 and 41%.164 Asymptomatic bronchiectasis may be identified in 4–8% of patients with RA using HRCT.163 173 An association between RA and bronchiectasis seems likely. While some authors consider that RA-related bronchiectasis may precede the onset of joint symptoms,159 170 these studies can be criticised for not rigorously assessing other possible causes of bronchiectasis, in particular early childhood infections which can be identified in all162 or some165 patients with RA and bronchiectasis. The prognosis of patients with both RA and bronchiectasis was significantly worse than either condition alone in one study158 but not another.159

Evidence that other connective tissue disorders are associated with bronchiectasis is weaker as they have been subjected to fewer studies with usually small numbers of selected patients without control populations. A single study of patients with systemic sclerosis found bronchiectasis in 59%.161 Bronchiectasis in patients with systemic lupus erythematosus, ankylosing spondylitis, relapsing polychondritis, Marfan’s syndrome and Ehlers–Danlos syndrome has been noted.169 174 175

Inflammatory bowel diseases

The association with ulcerative colitis is well established, although there are also reports of associations with Crohn’s disease and coeliac disease. This topic is the subject of comprehensive reviews.177 178 The most well-recognised presentation is seen in patients with severe colitis who eventually come to total colectomy and then develop abrupt onset of cough with purulent sputum soon afterwards. Characteristically, the sputum is very purulent and culture negative for bacterial species. Other patients with one condition develop the other several years later and, in these cases, flare-up of one condition may or may not be associated with the flare-up of the other.178 Crohn's disease is a much less well-recognised association of bronchiectasis but, again, onset of cough and sputum has been associated with bowel resection. Coeliac disease is the most tenuous of all,176 but there is T cell infiltration in both and, because coeliac disease is often subclinical, it may deserve further research.

Disorders of ciliary function

Patients with PCD have a congenital abnormality of ciliary function. Patients may present at an early age with one or more of the problems listed in box 1. There is nearly always a history of symptoms of neonatal respiratory distress if the condition is not diagnosed and, if they present as adults, they will usually have established bronchiectasis55 as well as problems at other sites bearing cilia—for example, deafness due to recurrent middle ear problems, chronic rhinosinusitis and male infertility (note that infertility in males is by no means invariable). There is also a risk of female subfertility including ectopic pregnancy, but this is probably quite small.

Case series indicate that upper respiratory tract symptoms (regular nasal discharge, anosmia, sinusitis, hearing impairment or chronic otitis media) are almost universal in patients with PCD.55 180 181 In children with PCD, rhinitis/rhinorrhoea often begins at birth and otitis media was seen in 100% in one cohort.55

Studies examining the effects of PCD on lung function have found worse forced expiratory volume in 1 s (FEV₁) and forced vital capacity (FVC) in patients identified as adults compared with children.55 179 Lung function impairment may be severe enough to cause reparatory failure and need for transplantation.55 One study found lung function stabilised with a programme of regular follow-up, intensive physiotherapy and antibiotic treatment,179 and another found a slight increase in FVC in patients on regular prophylactic antibiotics.182

Patients or their families may wish to contact the Primary Ciliary Dyskinesia Family Support Group (for contact details see online Appendix).

Good practice point

• Patients with bronchiectasis due to PCD should be seen in secondary care at least four times each year with measurements of lung function.

Is α₁-antitrypsin deficiency a cause of bronchiectasis?

The role of α₁-antitrypsin (AAT) deficiency in the aetiology of bronchiectasis has been the subject of debate and contention over the years. An association was postulated after a number of case reports linked severe (Pi ZZ phenotype) AAT deficiency to bronchiectasis in individual or small numbers of cases.183 190 192 198–200 Many of these reports mention other possible causes of bronchiectasis in patients’ histories,184 185 and exclusion of specific conditions such as immune deficiency and CF is variable.

The large BTS case series of patients with severe (PiZZ) or moderately severe (PiSZ) AAT deficiency in the 1970s and a more recent American series, while finding a high frequency of symptoms compatible with ‘chronic bronchitis’ as defined by Medical Research Council/American Thoracic Society (MRC/ATS) criteria that might in theory be caused by bronchiectasis, found no radiological evidence of bronchiectasis on chest x-rays.194–196 Bronchograms/HRCT scans were not performed. HRCT has been used in series looking at small numbers of patients identified from databases of AAT-deficient patients, with two studies showing a low incidence of bronchiectasis and other possible underlying causes identified.189 191 Bronchiectasis was seen in 43%187 and 95%202 in other radiological series of AAT-deficient patients and in 15% of a series of 42 post-mortem examinations.197

Case–control studies have found no link with AAT deficiency in a study of 35 patients,201 and no difference in the frequency of the Pi phenotypes in 202 patients with bronchiectasis compared with a blood donor control population.186 One series of 60 patients found an over-representation of patients with PiMZ bronchiectasis compared with controls, but many of these patients had other identifiable causes of lung damage.188 In addition, it is apparent that most patients with severe AAT deficiency remain asymptomatic throughout life if they do not smoke.193

Uncontrolled case series suggest a higher incidence of bronchiectasis in AAT than that seen in controlled studies; further work is needed in this area.

Which children should be investigated for bronchiectasis?

Chronic productive cough, especially between viral colds, is probably the most important single indication to consider in children, and a chronic productive or moist cough every day for 8 weeks211 (but not a child who has an intermittent cough with periods of complete remission within the 8 weeks) should be investigated for possible bronchiectasis.24 212 216 Consideration should also be given to investigating the child with a prolonged acute cough (3–8 weeks) in whom the symptom is becoming more frequent or intense.211 Young children may not expectorate sputum. In a hospital setting there is excellent agreement between the description of a wet cough by parents and doctors and the finding of increased endobronchial secretions at bronchoscopy.209

Symptoms attributed to childhood asthma that are atypical or which respond poorly to conventional treatment may in fact be related to bronchiectasis (the reason for referral in 49% of one recent UK series of children with bronchiectasis23). In particular, only the most typical case of ‘cough variant asthma’ should not be investigated further. Localised chronic bronchial obstruction—including in particular an organic foreign body which is either left-sided, presents with consolidation or where there was delay in bronchoscopic removal—may lead to bronchiectasis.210 Severe gastro-oesophageal reflux, dyscoordinate swallowing, laryngeal cleft or late presentation of H-type fistula and oesophageal motility disorders (including after repair of oesophageal atresia) should all be considered risk factors for the development of bronchiectasis.

Microbiological factors may alert the clinician to bronchiectasis in a child. A positive sputum culture for an unusual bacterial organism can indicate an underlying disorder associated with bronchiectasis, in particular Staphylococcus aureus, Pseudomonas aeruginosa and non-tuberculous mycobacteria (CF or primary cilia dyskinesia55) or Burkholderia cepacia complex (chronic granulomatous disease or CF). Some organisms have a propensity to cause long-term sequelae; for example, particular serotypes of adenovirus (7, 14, 21)217 or Bordetella pertussis and any episode of severe pneumonia (whatever the cause) should prompt further investigation, particularly if there is incomplete resolution of symptoms or persistent physical signs.53 213 Recurrent (two or more) episodes of consolidation and either localised or multifocal212 213 persistent and unexplained chest radiographic abnormalities (suggestive of airway disease or a focal abnormality23) 12 weeks beyond the initial illness should also raise suspicion. These include infiltrates, parenchymal densities or atelectasis.212 218 However, one paper218 suggested that the exception is after respiratory syncytial virus infection where persistent atelectasis is not a risk factor for subsequent bronchiectasis.

Bronchiectasis should be considered if there are any features suggestive of CF (diarrhoea, failure to thrive, rectal prolapse, electrolyte disorder), PCD (neonatal onset of symptoms including rhinitis or respiratory distress without an obvious cause, mirror image arrangement, severe chronic serous otitis media, especially with chronic otorrhea after tympanostomy tube insertion)214 or a systemic immunodeficiency215 (severe, persistent or recurrent infections and infections with unusual organisms or children known to be HIV positive81). Even minor immunodeficiency such as functional antibody deficiency may be associated with bronchiectasis.23 A positive family history of severe or unexplained respiratory disease may also be relevant.

Which adults should be investigated for bronchiectasis?

In a clinical setting, identifying cases of bronchiectasis is dependent on eliciting symptoms compatible with the diagnosis. While clearcut in some cases, in others this is more difficult as the most common symptom of bronchiectasis—a cough productive of mucoid or purulent sputum—is also a common presentation of other respiratory diseases of adulthood, in particular COPD. Chronic or recurrent sputum production is a common presentation in primary care, and a study suggested that a practice of 10 000 patients can expect around two patients/week to consult with symptoms of persistent lower respiratory tract infection despite antibiotic therapy, 38% of whom have already identified bronchiectasis.222 In two studies of patients from primary care labelled COPD, bronchiectasis was identified in 29%225 and 50%219 and bronchiectasis was found in 68%223 and 70%220 of cohorts referred to secondary care with chronic sputum production.

Studies of patients with confirmed bronchiectasis identify a spectrum of symptoms with some patients who expectorate large volumes of offensive purulent sputum on a daily basis at one extreme52 53 and with young age at presentation46 and onset of symptoms22 221 in some. Even within these studies there is a wide spectrum in all the patient characteristics, and studies of uncharacterised patients with chronic sputum production show that clinical detection of bronchiectasis is difficult. Measures of sputum volume, sputum purulence, duration of sputum production, age at onset of symptoms, frequency of exacerbations and smoking history have low sensitivity and specificity for identifying bronchiectasis.219 223 225 Sputum from patients with bronchiectasis in a stable state while on average is more discoloured may often be indistinguishable from that of COPD.224 225 Single studies have shown than continuous rather than intermittent purulent sputum production223 and prolonged recovery from exacerbations219 are predictors of bronchiectasis. A proportion of patients labelled as COPD will be non-smokers with bronchiectasis but, even so, 81% of patients with bronchiectasis identified in a primary care study of COPD were also smokers.225 Bronchiectasis may less commonly lie behind symptoms such as a persistent non-productive cough and isolated haemoptysis in the absence of sputum production.52 53 The isolation of P aeruginosa in the sputum of patients with COPD is uncommon.226

What are the symptoms and signs of bronchiectasis in children?

When reviewing the literature on the clinical features of bronchiectasis in children, it is apparent that many papers are either old and relate to children with long-established disease20 53 213 229 or refer to indigenous populations living in poverty,42 212 228 so reducing the relevance to modern practice in the UK, and the aim must be to identify the symptoms and signs of very early disease. Recent studies found that the diagnosis may be delayed for a median of 323 or 3.6216 years after symptom onset.

While it is possible that a child with established bronchiectasis may have no habitual symptoms,227 a chronic moist cough that is productive of sputum (sometimes fetid42 53) is a frequent symptom.20 42 53 212 213 216 227 228 232 Wheeze42 212 213 227 232 and haemoptysis20 213 227 228 230 may also be described. Haemoptysis was reported largely in older studies.20 53 227 Exertional breathlessness is described in many case series.42 53 212 213 228 Children may also present with failure to thrive or malnutrition,53 212 213 although in the UK they are likely to be of normal weight and height.231 Less frequently described as presenting features are fever,228 chest pain228 and recurrent lower respiratory tract infection.

Interpretation of the literature with regard to physical findings is also difficult in terms of relevance to the current population. While classical signs of advanced established disease (digital clubbing, cyanosis, chest deformity, hyperinflation, altered posture) have been reported in recent non-UK series19 42 212 as well as many older series,20 53 213 227 228 a single recent UK study found clubbing to be absent and chest deformity uncommon in non-CF bronchiectasis.231 A significant sign is the presence of persistent inspiratory crackles on auscultation of the lungs,20 42 53 213 especially in the absence of a viral cold. Wheeze, upper airway disease and mediastinal shift may also be seen.20 42 213 227

Children may experience worsening of symptoms at times of increased levels of infection (an infective exacerbation). Clinical features of an infective exacerbation are described in more detail in Section 5.

What symptoms and signs should be assessed in an adult with bronchiectasis?

Symptoms and signs in patients with bronchiectasis may relate directly to bronchiectatic airways or may be secondary to an underlying cause.

Good practice point

• Impact of symptoms on daily life should be assessed.

Why should the underlying cause of bronchiectasis be established?

Some investigations into the underlying cause are appropriate for all patients in whom bronchiectasis is confirmed, whereas others can be used in selected patients in whom a particular aetiology is suspected and the use of investigations will be determined in part by clinical judgement. Case series have demonstrated that investigations into the underlying cause change management and identify previously unrecognised conditions that, while sometimes rare, have important treatment and/or prognostic implications (for instance, immune defects and CF in adults and PCD in children).22 23 54 232

What blood tests should be performed?

These are listed in below. Some tests should be performed in all patients who have bronchiectasis whereas others can be requested more selectively.

Blood inflammatory markers (neutrophil count, erythrocyte sedimentation rate, C-reactive protein) can be used as indirect markers of disease activity and as a signal of the severity of an exacerbation; very high values may indicate concomitant pneumonia. Values in stable state have been shown to correlate both with extent of bronchiectasis and quality of life. Similarly, patients with bronchiectasis usually have high levels of major immunoglobulin classes; in one series, 83% IgG, IgA, IgM or a combination was raised by >2SD above the mean, reflecting chronic bronchial infection. Conversely, common variable immunodeficiency (hypogammaglobulinaemia) is a relatively common cause of bronchiectasis, which is important to identify when present because immunoglobulin replacement will prevent (or slow) disease progression.63–65 Measurement of IgG subclasses are not recommended for several reasons: studies have shown that low subclass levels do not necessarily make patients susceptible to infection; they are difficult tests to perform in the laboratory and values do not always add up to total IgG; results are not consistent and may recover spontaneously. Specific (functional) antibodies are recommended instead (for detailed information see section on immunology).

Peripheral blood eosinophilia, high serum IgE and a positive RAST (specific IgE) test to Aspergillus characterises ABPA. Aspergillus precipitins are positive in a proportion of cases,22 54 123 but multiple precipitin lines suggests an aspergilloma. Rheumatoid factor is non-specific, but high values with evidence of joint disease do characterise a group of patients in whom small airways disease is prominent and immunosuppression, even in the presence of active infection, should be considered.

Good practice points

• The following tests should be performed in all patients:

  full blood count and white cell differential

  erythrocyte sedimentation rate or C-reactive protein;

  routine biochemistry.

• If clinically relevant, the following should be performed:

  rheumatoid factor, antinuclear factor and ANCA;

  functional antibody assessment (see immunology section)

What immunological tests should be done on all patients?

Antibody deficiency constitutes a significant underlying cause of bronchiectasis. Other immune deficiency disorders (primary and secondary) also play a role, although on a less frequent basis. Different approaches are therefore needed as to the definition of possible underlying antibody deficiency and to other immune deficits in the patient with bronchiectasis.

In cases where a diagnosis of bronchiectasis is established, a process of screening (universal or targeted) for antibody deficiency is warranted and was suggested as long ago as 1973.253 This would be aimed at detection, principally, of primary antibody deficiency but will also detect the majority of cases with secondary antibody defects which are complicated by structural lung damage. In either case, antibody deficiency may be a recent development or be of long standing and may be manifest overtly or in a clinically subtle manner. It is recommended that serum measurement of the major immunoglobulin isotypes is undertaken in all new (and previously uninvestigated) patients with bronchiectasis. In addition, assessment of the adequacy of antibody responses to specific antigen challenge (natural or artificial) should be assessed, either universally across all patients or by a targeted approach with investigation in selected patients. Targeted screening could, for instance, apply only to cases where more common underlying causes have already been excluded (bronchiectasis of undefined aetiology) as the second component of a biphasic investigation protocol or to cases where other indicators of potential antibody deficiency (eg, coexistent otitis media)89 are present. Local operational factors will determine whether a universal or a targeted approach to specific antibody measurement is preferable. At present there is no clinical, economic or quality of life evidence that a universal approach is superior or justified.83 89

Current processes for evaluation of antibody deficiency in the routine setting of bronchiectasis should encompass:

1. Screening measurement of serum IgG, IgA and IgM levels with electrophoresis in all patients.63 65 115 Routine assessment of serum IgE and of IgG subclass levels are not justified as screening investigations for exclusion of antibody deficiency.90 111 112 252

2. Universal or targeted assessment of baseline specific antibody responses to peptide and polysaccharide antigens. The former may include tetanus toxoid. For the latter, the capsular polysaccharides of S pneumoniae and H influenzae type b are currently most practical.

3. Where specific antibody levels are absent or subnormal, adequacy of the humoral response to challenge should be assessed by immunisation with appropriate antigens, as guided by measured baseline antibody levels, and remeasurement of relevant levels around 21 days after immunisation. Adequate evaluation of humoral immunity requires assessment of baseline antibody levels, with or without post-immunisation responses as appropriate, to both unconjugated and peptide-conjugated polysaccharide vaccines.254

Interpretation of test results should take into account recent infection history, previous immunisations, defined local normal ranges, the limitations of derived normal ranges for some of these tests and the lack of response (natural or immunisation) to specific antigens which may occur in some apparently normal individuals. Assessment of natural or immunisation response to unconjugated polysaccharide vaccines is not useful in infants <2 years and is difficult in children <5 years.

Among the most common abnormalities which will be identified by a universal screening process will be elevated serum immunoglobulin levels, principally of IgG and IgA classes.71 73 83 255 These abnormalities are generally reactive and are reflective of repeated inflammation at the bronchial mucosal surface. Normal or elevated total immunoglobulin levels can, however, mask significant defects in specific antibody production,256 which stresses the importance of assessing the functional integrity of the humoral response at the level of specific antigens as well as measuring total immunoglobulin levels.

Second-line investigation is required (1) where screening identifies a humoral immune deficit; (2) in some cases, possibly, to ascertain that reactive features do not mask a significant underlying defect; and (3) where screening for antibody deficiency has shown no significant abnormality but where there are features or indicators which result in a persisting suspicion of immune deficiency. Interpretation of abnormal problematic screening test results and planning of further investigations, where required, should ideally be undertaken with specialist input and advice from the immunology department.101 117 The general approach to investigating bronchiectasis in combination with immune deficiency is outlined in figure 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Bronchiectasis immune deficiency investigation protocol.

Contact information about local/regional immunology services which can appropriately input into the diagnosis of patients with bronchiectasis who may have underlying immune deficiency and in the further management of such patients can be obtained from the UK Primary Immunodeficiency Network (contact details in online Appendix).

What are the second-line immunological investigations and when should they be performed?

Although, in general, normality of primary screening tests occurs very rarely in the presence of a significant immune deficiency disorder, further testing of immune competence beyond those humoral ‘screening’ investigations already outlined may be indicated in some patients with bronchiectasis. For instance, interstitial pneumonitis due to Pneumocystis carinii or cytomegalovirus may complicate T cell dysfunction and staphylococcal or Aspergillus infection primarily suggests a neutrophil defect. Detailed investigation for possible immune deficits beyond simple relatively easily-defined antibody defects should be undertaken only in conjunction with specialist clinical immunology input and advice using agreed protocols. Investigations should only employ validated techniques and should be undertaken in a rational sequential fashion on the basis of presenting features.257–259 Investigations should be performed by diagnostic laboratories which are externally accredited by appropriate bodies and a definitive diagnosis of immune deficiency should be based on established and accepted criteria,102 260 although uniform guidelines on the diagnosis of some conditions have yet to be adequately defined.

Further investigation is principally directed towards identifying or refining defects in host defences, individually or in combination, and principally to test the following compartments:

1. T cell (enumeration, phenotype, in vitro and in vivo activation, proliferative capacity, cytokine production).

2. B cell/immunoglobulin (cell enumeration, phenotype, proliferative capacity, immunoglobulin quantification and functional responses).

3. Phagocyte (enumeration, adhesive capacity, chemotaxis/migration, phagocytosis, oxidative burst, killing and degradation).

4. Complement (functional assessment of pathways, activation/breakdown products, individual component assessment).

5. More rarely, other immune components (eg, natural killer cells).

When should patients have gastrointestinal investigations?

Gastrointestinal investigations should be performed at the discretion of the clinician. There will be a lower threshold for these investigations in children in whom there is a higher incidence of structural abnormalities or aspiration presenting as bronchiectasis than in adult patients.232 In adults a high incidence of bronchiectasis associated with gastric aspiration has been identified in lung transplant patients.38 261 The investigations chosen will normally include one or more of 24 h oesophageal pH monitoring, barium studies, videofluoroscopy or the identification of foam-laden macrophages on bronchoscopic samples. Identifying aspiration in the context of bronchiectasis can direct management (intensive acid suppression and fundoplication when it is felt appropriate).

When should patients have investigations to exclude CF?

As CF is associated with a more rapid progression and considerably greater mortality than non-CF bronchiectasis, it is important to identify these cases. Unless a confident alternative cause can be identified, all children presenting with bronchiectasis will need investigations to exclude CF and the minimum should be two measurements of sweat chloride and CFTR mutation analysis.

In adults, clinical judgement will be needed to decide who should have investigations. Screening all patients with bronchiectasis using sweat and genetic testing, although performed in some centres, is time-consuming for patients and staff and has not been shown to be cost-effective. Consideration should be given to factors that increase the probability of underlying CF. Two studies of patients with undifferentiated bronchiectasis found that those identified with CF were not aged >56 years22 152 and 85% were <40 years.152 Rarely has first presentation of CF in the eighth decade with respiratory symptoms been described.266 A cohort of patients with CF presenting as adults (including those with non-respiratory symptoms) had a mean age at presentation of 32 years.157 Other features seen in adults found to have CF include clinical features of malabsorption,152 a history of male infertility,22 152 childhood steatorrhoea,152 the isolation of S aureus in sputum22 152 263 and upper lobe bronchiectasis on HRCT scanning.22 A previous negative sweat test does not exclude CF.152 Some will have a history of symptoms since childhood152 or sinusitis,157 although these are also common features in patients with bronchiectasis in general.22

Guidelines for accurate sweat testing have previously been established with the recommendation that sweat chloride rather than sweat conductivity should be measured.265 The sweat test is useful in adults as well as children,264 although mean values are lower in atypical CF patients who present in adulthood157 and may fall within the normal or intermediate range.157 The greater the number of CFTR mutations screened for, the greater the number of patients with CF will be identified.152 Advice from the local clinical genetics department may be necessary. The role of nasal potential difference measurements in routine clinical practice has yet to be defined.262

When should patients have tests of ciliary function? What are the best tests to identify ciliary defects?

Tests of ciliary function can be divided into those that are indirect and may be used to screen patients (saccharin test and nasal nitric oxide (NO) measurements) and those that definitively assess function and structure (ciliary beat frequency/pattern tests and electron microscopy studies). The saccharin test is cheap and can be performed everywhere. Unfortunately it is only reliable if performed exactly as set out in box 2, so may not lend itself to occasional practice. The test is not suitable for small children who will not sit still for an hour. In patients with PCD, nasal NO and, to a lesser extent, bronchial NO is very low and in centres with access to this test it can be used to screen for PCD (nasal NO <100 parts per billion indicates need to test ciliary function). A consensus guideline for the measurement of nasal NO has been published.267 Ciliary function tests can only be performed in specialist centres (see online Appendix for centre and referral details).

In children, PCD will be considered if any of the features outlined in box 1 are present. A ciliary abnormality is unlikely in adults if the history does not go back to childhood, there is no history of chronic otitis media or upper respiratory tract symptoms,55 or if there has been a prolonged spell of several years in which the patient has been asymptomatic.

Good practice point

• For children (particularly if very young), direct referral to a specialist centre may be preferred to performing screening tests suboptimally

What are the indications for bronchoscopy?

Bronchoscopy can identify and be used to remove foreign bodies in the endobronchial tree and can show anatomical abnormalities of the bronchi. With the increasing resolution and availability of HRCT scanning, the place of bronchoscopy in the investigation of patients is unclear. In studies of a non-UK population of children, bronchoscopic abnormalities corresponded to HRCT changes although bronchoscopy did allow those abnormalities to be characterised and simple mucosal inflammation to be distinguished from bronchomalacia.212 276 A recent case report showed that a foreign body seen at bronchoscopy was not evident in the single affected bronchiectatic lobe on HRCT.21 No studies of adult patients were identified.

Bronchoscopy may be used to characterise pathogens in the lower respiratory tract. In children and adults with bronchiectasis there is limited published information on its role. Only one study of in stable state was identified271 in which bronchoscopy did not show any advantage over sputum culture at identifying lower respiratory tract pathogens. Another study in children with bronchiectasis related to HIV who had an acute respiratory illness showed that bronchoalveolar lavage (BAL) had a high yield of clinically relevant pathogens that required specific antibiotic treatment.81 In adults with stable symptoms, BAL and protected specimen brush are both sensitive in detecting lower respiratory tract organisms268–270 which were identified in 57–88% of patients. This process is moderately invasive and quantitative microbiological analysis was used. When comparing BAL with expectorated sputum culture, BAL has only slightly greater sensitivity for detecting lower respiratory tract pathogens and a lower incidence of contamination by nasal and oropharyngeal flora.269 272 275 342 When atypical mycobacterial infection is suspected on HRCT scanning as a cause of bronchiectasis, bronchoscopy with bronchial washings significantly increases the yield of mycobacterial cultures above that of sputum culture alone.273 274 Bronchoscopic lung biopsy can detect granulomas in the context of atypical mycobacterial infection.274

The identification of lipid-laden macrophages on cytology may indicate gastric aspiration.232

What are the important modalities for imaging bronchiectasis?

Chest radiography and HRCT scanning are the two most frequently used imaging tests for the diagnosis of bronchiectasis. Other imaging tests either lack the spatial resolution to demonstrate bronchial abnormalities (eg, MRI) or rely on indirect signs of disease (eg, radionuclide studies). HRCT is the mainstay for the identification of bronchiectasis. Developments in multidetector CT technology and image processing software have improved the speed of acquisition of data and depiction of airways abnormalities, respectively, but a standard HRCT examination remains sufficient for the specific task of demonstrating bronchiectasis.

What is the role of a chest x-ray?

Despite the deficiencies of the chest x-ray (notably its insensitivity for the diagnosis of early bronchiectasis and poor observer agreement), it is the usually the first imaging test used for the investigation of a patient with suspected bronchiectasis. Digital acquisition devices are capable of producing x-rays with improved visualisation of, for example, bronchiectatic airways behind the heart, with the added potential of radiation dose reduction.279 However, unless disease is severe, the radiographic signs of bronchiectasis are usually inconspicuous. Some individuals fulfilling the usual criteria for COPD will have radiographically cryptic bronchiectasis, as judged by CT criteria.225 For this reason, an apparently normal chest x-ray cannot be taken to rule out bronchiectasis. Nevertheless, some studies have suggested that the chest x-ray is rarely absolutely normal in the face of what has been imprecisely termed ‘clinically relevant’ bronchiectasis.277 278

In terms of specificity, the chest x-ray of a patient with COPD showing bronchial wall thickening (tramline and ring shadows) and large volume lungs may be erroneously interpreted as indicating bronchiectasis. The same non-specific features suggestive of bronchiectasis are also frequently encountered in subjects with asthma141 and children with acute lower respiratory infection.280

There is no good evidence to support the routine use of chest radiography in monitoring patients with bronchiectasis with no change in symptoms. Furthermore, there is very poor correlation between infective exacerbations in individuals with bronchiectasis and convincing radiographic changes.281

What is the role of HRCT?

An HRCT examination is unlikely to change the diagnosis in a patient with clinically and radiographically advanced bronchiectasis. However, numerous studies and clinical experience attest to the improved performance of HRCT over chest radiography in detecting early or mild bronchiectasis. Nevertheless, because HRCT cannot be regarded as the final arbiter for the presence or absence of ‘clinically significant’ bronchiectasis, a dogmatic recommendation that all patients with suspected bronchiectasis should be subjected to an HRCT examination is unwarranted. In this context, an appreciation of reasons for both the under- and over-calling of bronchiectasis HRCT is important.285

In the absence of a true in vivo gold standard, the accuracy of HRCT in confirming or excluding bronchiectasis is difficult to ascertain. Evidence from early studies which compared CT (using widely different technical parameters, particularly section collimation) with bronchography do not allow any definite conclusions to be drawn.282 283 While the HRCT features of frank bronchiectasis can be regarded as reasonably specific, the difficulties in definitively identifying early bronchiectasis may be considerable. The earliest abnormalities in the evolution of bronchiectasis are mild and merge with normality (the ‘hinterland of normal’). Further confounders are the effects of age and cigarette smoking, both of which conspire to cause bronchial abnormalities which may be marked enough to fulfil the HRCT criteria for bronchiectasis.286 287 Given that one of the earliest signs of suppurative lung disease is bronchial wall thickening, which is a non-specific sign of airways disease associated with the least good observer agreement284 and is encountered in other conditions such as asthma137 288 and in cigarette smokers,287 it is not surprising that the HRCT diagnosis of mild or early bronchiectasis may be contentious.

What is an optimum HRCT protocol for defining bronchiectasis?

The full potential of CT for the detection of bronchiectasis was only realised with the advent of HRCT—that is, in investigations which used a standardised protocol of 1.5 mm thick sections at 10 mm intervals.294 310 The simplest and lowest radiation dose HRCT technique remains narrow collimation (1 mm) sections obtained at 10 mm intervals from lung apex to base with the patient in a supine position, breath-holding at maximum inspiration. Images are reconstructed using a high spatial frequency reconstruction algorithm. With such a protocol, clinically important bronchiectasis is unlikely to be missed in the gaps between the thin sections. However, with the increasing use of multidetector CT scanners which allow an almost infinite variety of section thickness and interspacing, volumetric thin section CT (ie, no gaps between slices) is becoming the norm291 297 despite the extra radiation incurred.

Early studies of the detection rate of bronchiectasis with volumetric CT, previously referred to as spiral CT technique (3 mm collimation, pitch of 1.6; 24 s breath hold) showed that it was superior to a standard HRCT protocol (1.5 mm collimation at 10 mm intervals),295 but the radiation burden to patients using this protocol was over three times greater than that of conventional HRCT. This basic caveat applies to the latest generation of multidetector CT scanners although, as the number of detectors increases (at the time of writing, 64 channels), the effective radiation dose diminishes. To date, few studies have directly compared the diagnostic yield of conventional HRCT with that of volumetric HRCT and, at the same time, provided a meaningful comparison of the radiation dose of the two protocols.

There are several potential benefits of volumetric HRCT. In the context of large airways disease, one of the most frequently cited is the ability to perform three-dimensional reconstructions which can provide ‘virtual bronchoscopy’ or ‘poor man's bronchography’.300 However, although useful for the depiction of major airways, such renditions are prone to artefacts and do not confer any significant diagnostic advantage over conventional transaxial and orthogonal reconstructions for the diagnosis of bronchiectasis. A more tangible benefit is the reconstruction of images in the coronal plane which, in one study, compared with transaxial sections alone improved both the detection of bronchiectasis and observer agreement.298 Many variations in the way volumetrically acquired data is reconstructed are possible (eg, a ‘paddle-wheel’ display),301 but none has been shown to be significantly superior to transaxial displays. A further advantage of volumetric HRCT in the few individuals in whom a follow-up HRCT scan is deemed useful is the ability to compare exactly comparable sections—that is, taken at precisely the same level—so that the difficulty of making a judgement about progression of disease on conventional interspaced HRCT scans is overcome. A disadvantage of volumetric HRCT, apart from its increased radiation dose, is the greater degree of image degradation by motion artefact compared with conventional HRCT302 (particularly if the latter is performed with electrocardiographic gating303). Whether this has an important impact on the detection of early bronchiectasis has not yet been investigated.

Although the radiation dose from a standard HRCT scan is relatively low (effective radiation dose in the region of 0.9 mSv), a volumetric HRCT scan can be 1.5–6 times this dose; the most important determinants of the increased dose inherent in volumetric scanning are the mA setting (which can be adjusted according to the weight of the patient) and the type of CT scanner in terms of number of detectors (a four-channel machine delivers a greater effective radiation dose to the patient than a 64-channel scanner). Recent studies have confirmed that it is possible to reduce the mA by one-half to two-thirds of the manufacturers’ usual recommendations without any deleterious effect on the diagnostic quality of the images of volumetric293 299 or conventional289 HRCT. Reducing the kilovoltage, a manoeuvre particularly recommended in paediatric practice, also reduces the effective dose to the patient; for example, reducing the kVp from 120 to 80 kVp at a constant mA reduces the dose by approximately half.304 305 However, reducing the kVp is more likely to degrade image quality than decreasing mA, and a reduction below 100 kVp may result in unacceptable beam-hardening artefact.306

The small risk of cancer induction in patients undergoing CT scanning should not be forgotten. Although uncertainty exists about the risks at exposure levels normally encountered in diagnostic radiology, the best estimate currently in use for the general population is a 5% risk per Sievert for cancer mortality (recommendations of the International Commission on Radiological Protection, 1990). The effective dose of 6 mSv, for example, for an unmodified volumetric HRCT thus corresponds to a nominal cancer fatality risk of approximately 3 per 10 000 patients.

Additional HRCT sections taken at end-expiration may reveal features of small airways disease,292 but the identification of this feature is not needed to make the diagnosis of bronchiectasis. Sections obtained at end-expiration have been advocated to differentiate cystic bronchiectasis from other cystic lung diseases296; bronchiectatic airways usually decrease in size on expiratory scans in contrast to other cystic lesions. However, it has been reported that most cystic lesions in the lungs (bronchiectatic or otherwise) decrease in size, making these additional images of doubtful discriminatory value.307

Variations in window settings have a marked effect on the apparent thickness of bronchial walls. Narrow window settings will also alter the apparent bronchial diameter, unless the measurement of the diameter is made between points in the ‘centre’ of the bronchial walls.290 In the context of suspected bronchiectasis, a window level of between −400 and −950 Hounsfield Units (HU) and a width of between 1000 and 1600 HU have been widely recommended.308 310 A more liberal recommendation about the appropriate window level for the accurate evaluation of bronchial wall thickness has been reported in a study by Bankier et al that correlated thin-section CT with planimetric measurements of inflation-fixed lungs.309 For the accurate estimation of bronchial wall thickness the authors suggest that, irrespective of the chosen window width, the window centre should be between −250 and −700 HU, and that within this range bronchial wall thickness is not appreciably affected. Window width should be >1000 HU (a narrower window width will cause a spurious appearance of bronchial wall thickening); the suggested window width range lies between 1000 and 1400 HU.

Because of the ways in which various factors of the CT scanning protocol can alter the appearance—and even the apparent dimensions—of the bronchi, it is important that the CT technique is standardised and quality control is in place. Because of the many factors discussed, it is impossible to be prescriptive about the ‘optimal’ protocol for a CT scan tailored to detect bronchiectasis, but a summary of the two extremes—a typical protocol for a conventional HRCT using a single detector machine versus that for a volumetric HRCT using a 64-channel multidetector CT—are given below.

Good practice points

• Volumetric HRCT is superior to standard HRCT for the detection of bronchiectasis but delivers a higher radiation dose. Standard HRCT will be adequate in most instances.

• End-expiratory sections are not necessary for the detection of bronchiectasis or differentiating bronchiectasis from cystic lung diseases.

• The window centre should be between −250 and −700 HU and width between 1000 and 1400 HU.

What are the HRCT features of bronchiectasis?

Identification of dilation of the airways is a prerequisite for the HRCT diagnosis of bronchiectasis. The characteristics of bronchiectatic airways on the CT scan were first described by Naidich et al313 and there have been minor refinements subsequently.294 310 The relative size of a bronchus to its immediately adjacent pulmonary artery has been the most widely used criterion for the detection of abnormal dilation. In normal individuals the overall diameter of a bronchus is approximately the same, at any given level, as that of its accompanying pulmonary artery. The mean±SD ratio of the diameter of the bronchus (internal lumen) to the diameter of the pulmonary artery in normal individuals at sea level has been estimated to be 0.62±0.13.314 In healthy individuals minor discrepancies in the bronchoarterial diameter ratio may be seen,141 and these are more frequent with increasing age and in cigarette smokers.286 Thus, bronchial dilation in isolation in the absence of other signs cannot be regarded as a wholly specific sign of bronchiectasis. When airways lie parallel to the plane of section, abnormal dilation is recognised by a lack of normal tapering, producing a tramline (cylindrical) or flared appearance. Cylindrical bronchiectasis is by far the most common morphological pattern of bronchiectasis identified on CT.312 The usefulness of categorising bronchiectasis into cylindrical, varicose or cystic subtypes is limited, but cystic bronchiectasis usually denotes longstanding and more severe disease.

Bronchial wall thickening is a usual but inconstant feature of bronchiectasis. Problems with this variable feature have been widely debated and the definition of what constitutes abnormal bronchial wall thickening remains unresolved.223 Minor to mild degrees of bronchial wall thickening are seen in normal subjects, those with asthma, individuals with lower respiratory tract viral infections and asymptomatic smokers.315 There is no simple and robust criterion for the identification of abnormal bronchial wall thickening: Diederich et al defined abnormal bronchial wall thickening as being present if the internal diameter of the bronchus was <80% of the external diameter.311 While this sign was associated with good interobserver agreement, it cannot be applied when there is significant bronchial dilation (ie, in bronchiectasis).

Secretions within bronchiectatic airways will generally be easily recognisable as such. The larger plugged bronchi will be visible as lobulated or branching opacities. Such airways are usually seen in the presence of non-fluid filled obviously bronchiectatic airways. Mucus plugging of the smaller peripheral and centrilobular airways produces V- and Y-shaped opacities, the so-called ‘tree-in-bud’ pattern.316–318

In many patients with bronchiectasis, areas of decreased attenuation of the lung parenchyma can be identified; this mosaic attenuation pattern reflects coexisting constrictive obliterative bronchiolitis.308 Sections taken at end-expiration enhance the feature of decreased attenuation, the extent of which correlates with functional indices of airways obstruction.285 This finding is most prevalent in lobes with severe bronchiectasis but may be seen in some lobes in which there are no CT features of bronchiectasis.

A subtle degree of volume loss is a relatively early sign of bronchiectasis and is readily evident in the lower lobes on CT scanning; there is crowding of the airways and posterior displacement of the oblique fissure. CT scans will also clearly demonstrate complete collapse of lobes, although the diagnosis of bronchiectasis in acutely collapsed or consolidated lobes may be uncertain because of the reversibility of bronchial dilation in these situations.

Can HRCT identify features of specific causes?

An underlying cause for bronchiectasis is found in less than half of patients,324 and HRCT features alone do not usually allow a confident distinction between cases of idiopathic bronchiectasis versus known causes or associations of bronchiectasis. However, in some cases the pattern, distribution of bronchiectasis and associated CT features may be sufficiently characteristic for a specific underlying cause to be suggested; for example, the bronchiectasis of ABPA is typically upper zone and central in distribution, with more normal distal bronchi.125 319 325

A tendency to certain distributions has been described in groups of patients with specific conditions; for example, a lower and middle lobe distribution of cylindrical bronchiectasis with particularly marked bronchial wall thickening is reported to be typical of patients with hypogammaglobulinaemia.320 A predilection for the middle lobe has been reported in patients with immotile cilia syndrome326 and an upper lobe distribution of cylindrical bronchiectasis in patients with CF.327 328 Tracheobronchomegaly (Mounier–Kuhn syndrome) may be readily identified because of the marked dilation of the major airways and the grape-like bronchiectasis.329 In patients with bronchiectasis due to opportunist mycobacterial infection, particularly M avium intracellulare complex, there is often a suggestive triad of mild bronchiectasis concentrated in the right middle lobe and lingula, a tree-in-bud pattern and randomly scattered nodules 1–2 cm in diameter (the latter may show cavitation).322 323 Nevertheless, most studies that have sought to determine whether observers can reliably distinguish between idiopathic bronchiectasis and bronchiectasis of known cause have concluded that, although several CT features occur more frequently in certain groups of patients with an identifiable underlying cause, the CT features evaluated cannot be regarded as pathognomonic.312 319 321

How are HRCT changes related to lung function?

There is a relationship between the extent and severity of bronchiectasis depicted on HRCT and measures of airflow limitation, but the strength of correlation varies widely between studies.216 223 231 284 292 488 The disparate results reflect different study methodologies and patient selection and close analysis of the individual studies is inappropriate here, but some factors that need to be borne in mind when interpreting these studies include: (1) the type of HRCT scoring system used (eg, summative scoring of various morphological abnormalities as in the Bhalla system332 versus the grading of individual features such as bronchial wall thickening or mosaic attenuation pattern); (2) the population studied, which may potentially include presymptomatic individuals (eg, children with CF) versus older patients with advanced bronchiectasis; (3) ‘noise’ introduced in both the scoring of HRCT scans and the performance of lung function tests; (4) appropriate data analysis (many different individual morphological features may affect pulmonary function—there is the temptation to rely on univariate analysis to identify structure–function relationships but the application of multivariate techniques is invaluable in confirming the independence of correlations shown by univariate analysis); (5) the images obtained from HRCT may be mistakenly assumed to mirror microscopic as opposed to macroscopic abnormalities. As an example, areas of decreased attenuation in patients with bronchiectasis, which reflect coexistent constrictive bronchiolitis, may be misinterpreted as ‘emphysema’. In a study of patients with bronchiectasis it was reported that the widespread areas of decreased attenuation on HRCT scans were caused by emphysema, accounting for the functional gas trapping.330 However, the ‘emphysema’ seen in that study was not associated with decreased gas diffusing capacity, the functional correlate of emphysema.

Despite these caveats, most of the studies cited above have confirmed the expected link between the extent of bronchiectasis and indices of airflow limitation. In a study by Lynch et al, patients with a cystic pattern of bronchiectasis, taken to reflect more advanced disease than cylindrical or varicose patterns, had greater depression of FEV₁.331 An almost invariable finding in studies that have quantified individual abnormalities is the correlation between both bronchial wall thickness and areas of decreased attenuation (representing small airways obliteration) and airflow limitation.216 231 284

How often should radiological investigations be repeated?

Fluctuations in the pulmonary function of individuals with bronchiectasis reflect variations in a number of different morphological abnormalities including the degree of bronchial wall thickness and the volume of retained secretions in small and large airways; such changes are not necessarily evident on serial chest x-rays and even semi-objective scoring of the various expected radiographic abnormalities do not correlate well with clinical or functional features of an infective exacerbation. Nevertheless, despite the lack of evidence of their usefulness, chest x-rays tend to be repeated at follow-up. The argument that chest x-rays may reveal unexpected complications (eg, pneumothorax) is not compelling.

Over and above considerations of needless exposure to ionising radiation, there is no convincing case that repeated HRCT scans are useful in the management of patients with bronchiectasis. In anecdotal cases, HRCT scanning may provide an explanation for an otherwise unexplained step down in pulmonary function test results, but this is unusual. There may be a role for serial CT scanning to chart the evolution of bronchiectasis and to provide an explanation for fluctuations in pulmonary function tests over time. There have been few longitudinal studies in bronchiectasis evaluating the relationship between variations in pulmonary function indices and changes on CT, except in patients with CF.334 335 In adult idiopathic bronchiectasis, fluctuations in pulmonary function tests over time are most closely parallelled by changes in the extent of mucus plugging on HRCT whereas, on the baseline HRCT scan, bronchial wall thickening and the extent of mosaicism were the most important determinants of airflow limitation.333 At the present time, because of radiation dose considerations and the unknown reliability of CT scans for the detection of serial change, CT scanning should not routinely be used to monitor patients with bronchiectasis. One possible exception is patients with humoral immune deficiency in which progression of bronchiectasis may occur silently.116

Radiology in children

Justification for the need to expose a child to ionising radiation, however small the dose, should be a major consideration, particularly as children are approximately 10 times as sensitive to the effects of radiation as adults. Tailoring dose parameters to a ‘low as reasonably achievable’ CT examination should be vigorously pursued, with particular reference to reducing the mA based on the child’s weight. The use of volumetric HRCT scanning using multidetector scanners allows exactly anatomically comparable sections to be compared when evaluating sequential HRCT, which may be an important advantage in the context of clinical trials (see next section), but again the technical parameters need to be appropriately adjusted to reduce the radiation dose as far as possible. By using 1 mA/kg (in individuals <50 kg) and 120 kVp, it is possible to achieve acceptable image quality with a substantial reduction in dose; between 1.3 mSv (20 kg) and 2.5 mSv (60 kg) using a four-channel multidetector CT (see also section ‘What is an optimum CT protocol for defining bronchiectasis?’).

Aside from radiation, suboptimal image quality because of movement artefact is a particular problem in paediatric practice. Image degradation may be severe enough to render the HRCT non-diagnostic so that a repeat examination, with its attendant radiation, is required. Sedation of some sort is occasionally necessary for a restless child, but with multidetector CT it is reported that sedation is needed in less than 5% of cases.489

What scoring systems should be used for research?

A scoring system designed to quantify structural abnormalities seen on HRCT in patients with CF was first reported in 1991.332 Subsequently, various scoring systems (all based on patients with CF) have been proposed, all of which are variations on the theme of visual grading of the HRCT signs associated with bronchiectasis.336–341 No scoring system has been developed in non-CF bronchiectasis to date. No single system has been shown to be obviously superior to its competitors. With most of these systems, a semi-quantitative score is assigned to bronchiectasis (extent and severity), peribronchial thickening, mucus plugging, bullae, emphysema (this controversial abnormality is not included in more recent systems), air trapping on expiratory CT, areas of collapse or consolidation and thickening of interlobular septa.

Which organisms are isolated from the lower respiratory tract in bronchiectasis?

Studies examining the bacteriology of bronchiectasis are summarised in table AIV (Appendix 2). While many investigations into bronchiectasis contain microbiological information, only a few offer a comprehensive cross-sectional analysis of bacterial isolation. Methodology differs between studies, some using sputum culture, others BAL. The populations studied also vary. In children the predominant pathogen isolated is H influenzae19 23 271 276 with other organisms such as S pneumoniae, S aureus, M catarrhalis and P aeruginosa found much less frequently. One such study276 of 33 patients included five who had chronic suppurative lung disease without bronchiectasis on HRCT scanning but uniquely screened for mycobacteria on BAL, finding M tuberculosis in one of 28 patients with bronchiectasis.

In adults, H influenzae is the most frequently isolated pathogen, being found in up to 35% of patients.22 46 145 268 270 276 342–344 However, there is a significantly higher isolation rate of P aeruginosa than in children, this organism being isolated in 5–31% of patients. As antibiotic sensitivity patterns are considerably different for these two organisms, sputum culture results can have a direct bearing on the likely response to treatment. A significant but variable isolation rate of pathogens such as S pneumoniae, S aureus and M catarrhalis is also seen. Aspergillus species are found in a small number of patients. Studies have indicated that isolation or chronic colonisation with S aureus is associated with an increased incidence of CF and ABPA.22 263

How and when should standard microbiology be performed? At what interval should it be repeated?

The Health Protection Agency provides detailed guidance on all aspects of the collection, transport, laboratory processing, culture and antibiotic sensitivity testing of respiratory tract specimens. This information is available from the HPA website (http://www.hpa-standardmethods.org.uk/documents/bsop/pdf/bsop57.pdf.346 The recommendations below are based on the guidance from this agency. Culture of a fresh specimen of expectorated sputum is non-invasive, simple and effective in isolating pathogens from the lower respiratory tract and can be used in both adults and children,19 342 although a pharyngeal swab after coughing may be necessary in very young children and preferred in some older children. Bacteriological yield is related to purulence (ie, increased at times of an infective exacerbation) and should be obtained before the commencement of antibiotic therapy. In a single study in adult patients with bronchiectasis, a sputum specimen induced by nebulisation of hypertonic saline increased the yield of pathogenic organisms.342 Invasive investigation using bronchoscopic sampling is not necessary in routine practice although may sometimes be indicated. The simplest time to collect a specimen is at a visit to primary or secondary care, although precautions should be taken to avoid coughing in close proximity to other patients. As H influenzae and S pneumoniae may die if a specimen is not processed within 3 h, every effort should be made to ensure rapid transport of specimens to the microbiology laboratory. For patients in remote areas, rapid transport may not be possible. Published and unpublished data suggest that sputum specimens from patients with bronchiectasis may be posted to a laboratory and, if processed within 24 h, do not suffer any loss of yield.345 Laboratory culture using standard media (chocolate and blood agars) as well as supplementary media (CLED, mannitol and Sabouraud agars) is recommended to ensure isolation of likely target organisms.

A single sputum specimen may grow more than one pathogen and multiple different pathogens may be isolated with repeat testing over time.22 46 232 This may allow the distinction between intermittent isolation and chronic colonisation with an organism. Definitions of chronic colonisation differ between studies. In children, isolation of the same organism on three occasions at least 1 month apart over 1 year was used in two studies.19 231 In adults, definitions have included at least three isolates of an organism over a period of at least 3 months263 344; and at least two isolates 3 months apart over 1 year.22

Good practice points

• Sputum specimens may be obtained by deep coughing, physiotherapy or aerosol inhalation. Pharyngeal swab or bronchoscopic sampling is an alternative in children.

• For patients without prior positive culture results, three sputum specimens on different days may increase the yield of lower respiratory tract isolates.

• Specimens taken at the time of an infective exacerbation should be obtained prior to the commencement of antibiotic treatment.

When should specimens be sent for mycobacterial culture?

Routine culture of sputum for mycobacteria is not necessary but should be considered under specific circumstances when infection or relapse of M tuberculosis or opportunist mycobacterial infection is suspected. Indications for mycobacterial culture are indicated below.

Good practice points

Sputum (three early morning samples on successive days346) should be sent for mycobacterial microscopy and culture when there is: [D]

• unexplained systemic symptoms (eg, fevers, sweats, weight loss);

• a new infiltrate or cavity on the chest x-ray which does not clear with regular antibiotics;

• unexplained deterioration in clinical status not responding to usual treatment;

• middle-aged or elderly women with chronic cough and chest x-ray suggesting possible bronchiectasis (M avium complex);

• adults with bronchiectasis due to PCD.

Which lung function tests should be performed in children?

Only a few studies have specifically assessed measures of lung function in a representative population of children with bronchiectasis. Routine measurement of lung function in children with non-CF bronchiectasis may be achieved successfully in those who have reached school age. In those with an abnormality, the most frequent finding is airflow obstruction with reduced FEV₁, reduced forced expiratory flow at 25–75% FVC (FEF_25–75) and increased residual volume (RV) and/or ratio of RV to total lung capacity (TLC)19 231; FVC tends to be within normal limits or slightly reduced.19 20 69 231 In one study, spirometry identified normal values in 30%, an obstructive defect in 48% and restrictive pattern in 22%.69 FEV₁ and FEF_25–75 were negatively correlated with extent of disease as assessed by HRCT in one study,216 but in others there was no relationship.69 231 No studies examining lung function in preschool children were identified. Childhood resection of one or two lobes does not preclude achieving lung volumes within the normal range as an adult.348

The prevalence of non-specific bronchial hyper-responsiveness to histamine or methacholine is not clear but it can be seen in a subgroup of children with non-CF bronchiectasis.347 A single uncontrolled study identified bronchodilator responsiveness (>9% increase in FEV₁ after inhaled salbutamol) in 31%.

Which lung function tests should be performed in adults?

Many studies have recorded lung function in adults with bronchiectasis, although not all have excluded results from patients who had lung resection and some series included significant numbers of smokers. The most common pattern on spirometry is airflow obstruction which was seen in up to 80% of patients,221 350 351 354 357–359 although mixed obstructive/restrictive, restrictive or normal values may be seen.350 FEF_25–75 is often reduced and the RV/TLC ratio increased, but it is usual for the FVC and TLC to be normal or low.22 245 350 351 355 359 Gas transfer factor may be normal or reduced with the lowest measurements seen in more advanced disease,349 350 355 359 but transfer coefficient is usually normal.245 350 Reduced FEV₁ is correlated with breathlessness as assessed by the MRC dyspnoea score and extent of disease on HRCT.284 Colonisation with P aeruginosa is associated with worse lung function.344 FEV₁ <30% predicted is associated with increased mortality in patients with bronchiectasis who also have rheumatoid arthritis.158 Peak expiratory flow (PEF) monitoring shows a mean maximum percentage diurnal change of 8.6%.356

Studies of patients with immunodeficiency (hypogammaglobulinaemia) have shown that those receiving adequate immunoglobulin replacement therapy have better lung function than those who do not.360 361

Non-specific bronchial hyper-responsiveness to challenge with methacholine and/or histamine was demonstrated in 33%,358 69%349 and 72%350 in three studies. Assessment of reversibility of airflow obstruction after bronchodilators has been reported in a number of studies with very variable results and different testing and reporting methods. Uncontrolled studies have shown that 5%,353 12%,352 39%252 and 47%350 of patients have a significant response to salbutamol or fenoterol and 12% to ipratropium.352 Only one placebo-controlled study was identified which showed a significant improvement in lung function after placebo but a greater response to salbutamol. FEV₁ increased by a mean of 10.1%, FVC by 8.3% and PEF by 17.1%356 over the placebo response. None of the studies measured change in breathlessness in response to bronchodilators.

Is there a role for exercise testing in bronchiectasis?

Children with bronchiectasis can complete detailed exercise testing using an exercise bicycle231 or treadmill364 and this can identify functional limitation that would not be predicted by lung function testing or HRCT scanning.231 No study has examined the role of exercise testing in treatment or follow-up.

Exercise capacity in adults with bronchiectasis has been assessed using the 6 min walking test,363 incremental shuttle walking test242 365 366 and cycle ergometry.355 362 Maximum work rate measured by cycle ergometry correlated well with breathlessness score, FEV₁ and HRCT score in one study.355 The shuttle walking test detected an improvement in exercise performance after inspiratory muscle training365 and pulmonary rehabilitation.366

Can lung function tests be used to assess response to antibiotic treatment?

Many studies have used measures of lung function to assess efficacy of antibiotic treatment, mostly in relation to oral antibiotics, and information regarding intravenous and nebulised therapy is limited. Studies of short-term oral antibiotic use are nearly all in adult patients. Changes in lung function are variable with PEF, FEV₁ and FVC usually improving237 367 368 370 but sometimes remaining unchanged.370 Functional residual capacity (FRC) and TLC may increase.353 Patients may have significant improvement in other parameters such as sputum volume and purulence without any improvement in lung function. A single study in children observed reduced sputum volume with no change in FEV₁.347 Two studies of lung function response to long-term oral antibiotics were identified (both showing beneficial changes in sputum scores), one showing an improvement in FEV₁, FVC, FRC and TLC353 and the other showing no change in lung volumes other than a fall in RV.244 Carbon monoxide transfer factor and transfer coefficient do not change after short-term353 or long-term244 353 oral antibiotic treatment.

One study examining changes in lung function after intravenous antibiotic therapy found an increase in FEV₁ but no change in FVC,371 while increases in PEF, FEV₁ and FVC were seen in another.369 Lung function testing is an important aspect of assessing tolerability of nebulised antibiotics (Appendix 1) and can detect improvement after nebulised antibiotics.369 Patients with PCD show stabilisation of lung function when managed aggressively with repeated spirometry and antibiotics.179

The child with bronchiectasis

The therapeutic goals for treatment of bronchiectasis in children are to control symptoms, prevent progressive lung damage and to facilitate normal growth and development.

Central to paediatric care of bronchiectasis is the identification of any underlying cause (immunodeficiency, foreign body, aspiration, atypical CF, ciliary dyskinesia) and disease-specific therapy (immunoglobulin replacement therapy, non-oral feeding, surgery or referral to other specialists376). The importance of this approach has been highlighted in a recent study232 in which the identification of a cause led to specific management change in 56% of the cases assessed. Immunodeficiency and aspiration accounted for 52% of the cases and treatment of these underlying causes is likely to prevent further progression of disease.

The adult with bronchiectasis

The treatment aims in adult care are to control symptoms and thus enhance quality of life, reduce exacerbations and maintain pulmonary function. There is clear evidence that patients with bronchiectasis who have more frequent exacerbations have worse quality of life.251 The reduction of exacerbations should therefore be the aim of chronic management and provide a goal for future research. Lung function does decline gradually in patients with bronchiectasis351 359 372 374 but, in the current era, the decline is not rapid so only very large well-conducted studies will show significant treatment effects.241 Furthermore, there is a lack of clarity in the modern antibiotic era as to whether or not bronchiectasis is associated with decreased survival. A recent Finnish study suggested that survival is worse than asthma but not as bad as COPD in adult patients followed after their first admission to hospital; however, an Australian study suggests that patients with bronchiectasis have no change in survival compared with the general population.373 375 It is important that treatments are carefully evaluated and, in particular, that treatments routinely used in CF are not simply translated into use in non-CF bronchiectasis. In CF, the benefits of treatments in terms of lung function and survival may well outweigh the burden and side effects but, in non-CF bronchiectasis where progression and survival is less of an issue, the evaluation of new interventions against other end points associated with exacerbations and quality of life are required.

Education

Disease monitoring

As the aims of treatment are to maintain lung function and to improve quality of life by decreasing exacerbations, the aim of a follow-up visit is to ensure that the treatment plan is achieving these goals.

Role of primary care

Role of nurses

Multidisciplinary teamworking

Good practice point

• Patients with bronchiectasis should as a minimum be referred to a chest physician, physiotherapist and respiratory nurse with expertise in the condition.

Which patients should be taught airway clearance techniques?

There is no published evidence to indicate which patients should be taught airway clearance techniques. However, it is widely believed that a routine airway clearance regimen is an important component of the management of individuals who have a chronic productive cough and/or evidence of mucus plugging on HRCT scanning in order to enhance mucociliary clearance and reduce cough frequency. Due to the lack of evidence, it is impossible to say whether individuals with a non-productive cough may still benefit from seeing a physiotherapist. Although this group of patients will not need to carry out a routine airway clearance regimen, expert opinion advocates teaching an airway clearance technique to be used during infective exacerbations.

Which airway clearance technique(s) should be taught?

There are a number of airway clearance techniques that can be used in the management of patients with bronchiectasis. However, it is beyond the scope of this guideline to describe each airway clearance technique and the reader is referred to detailed descriptions of the techniques for further information.386 A survey of the current physiotherapy management of bronchiectasis in the UK found that 91% of senior physiotherapists taught the active cycle of breathing techniques routinely.384 Other techniques such as positive expiratory pressure (PEP), oscillating positive expiratory pressure, autogenic drainage and intermittent positive pressure breathing were used much less frequently.384 Most respondents also included ambulation, exercise and education on the use of inhaled therapy in the management of this patient group.384 Treatment choice appeared to be influenced as much by clinical experience as by research, reflecting the limited evidence base available to physiotherapists in this area. Certainly, 87% of respondents highlighted a need for further research regarding the physiotherapy management of patients with bronchiectasis.384

There is a considerable amount of evidence on the use of airway clearance techniques in bronchiectasis associated with CF. Extrapolation of findings is inevitable, but should be done with caution. These guidelines will focus on the evidence base for each airway clearance technique in the management of non-CF bronchiectasis. A small study (n=8) demonstrated an increase in sputum yield for chest physiotherapy compared with no physiotherapy in non-CF bronchiectasis.385 However, this was a short-term study and measured only sputum yield during and 30 min after the treatment period. There have been no long-term studies comparing any form of chest physiotherapy with no physiotherapy.

There are a wide variety of airway clearance techniques.

Active cycle of breathing techniques

The active cycle of breathing techniques is the most commonly used airway clearance technique in the UK.384 It can be (and commonly is) used in conjunction with manual techniques (eg, chest clapping and shaking) and postural drainage. When measuring sputum weight expectorated, the active cycle of breathing techniques (plus postural drainage and manual techniques) has been shown to be more effective than the test of incremental respiratory endurance.378 In addition, the active cycle of breathing techniques (plus postural drainage) is as effective as oscillating PEP (plus postural drainage and the forced expiration technique).379 382

Manual techniques

Manual techniques are used by physiotherapists to enhance the patient’s own efforts at airway clearance. They are most typically used in the UK in conjunction with the active cycle of breathing techniques. In the population with non-CF bronchiectasis, there is no evidence to show whether manual techniques provide additional benefit in the clearance of secretions over and above that of the active cycle of breathing techniques alone. Patients should, where possible, be encouraged to be independent with their chosen airway clearance technique. It has been shown that chest percussion, when used with postural drainage, does not adversely affect oxygen saturation or heart rate in non-CF bronchiectasis.238

Positive expiratory pressure (PEP)

PEP can be used as a treatment in its own right or as an adjunct to the active cycle of breathing techniques or autogenic drainage. There is currently no published evidence on the use of PEP in patients with non-CF bronchiectasis.

Oscillating PEP

Autogenic drainage

A pilot study (n=13) compared the effects of a single session of autogenic drainage versus a control (no chest physiotherapy).388 The outcome measures used were sputum weight and a measure of airway resistance called the interrupter technique (Rint). Significantly more sputum was produced during the autogenic drainage session compared with control. However, no changes in Rint were found following autogenic drainage compared with control. The absence of a significant change in Rint following autogenic drainage may be because Rint is not sensitive enough to detect changes in the airways of adults with bronchiectasis. Further research is required to assess the effectiveness of autogenic drainage in this population, and also to establish whether the interrupter technique is a valid outcome measure for use in adults with bronchiectasis.

Test of incremental respiratory endurance/resistive inspiratory manoeuvres

The test of incremental respiratory endurance is primarily used for inspiratory muscle training. However, it has been proposed as a method of airway clearance in bronchiectasis.378 A randomised crossover study was carried out comparing a single session of the active cycle of breathing techniques (incorporating postural drainage and vibrations) with a single session of resistive inspiratory manoeuvres in 20 patients with stable bronchiectasis. Sputum weight expectorated during and 30 min following the active cycle of breathing techniques was significantly greater than with the test of incremental respiratory endurance.

High-frequency chest wall oscillation

High-frequency chest wall oscillation is the application of positive pressure air pulses to the chest wall usually by means of an inflatable vest. There are few published studies available to evaluate the use of high-frequency chest wall oscillation in this population.

Research recommendation

• Further research is needed to investigate the efficacy of all the airway clearance techniques in non-CF bronchiectasis, particularly PEP, RC-Cornet, autogenic drainage and high-frequency chest wall oscillation.

Are adjuncts to airway clearance techniques useful?

There are a number of adjuncts that may be used in order to enhance the effectiveness of a chosen airway clearance technique.

How often should patients carry out airway clearance techniques? How long should an airway clearance session last?

Evidence for the frequency and duration of airway clearance techniques is not clear in non-CF bronchiectasis. However, it seems reasonable to relate frequency and duration of treatment to sputum volume, lifestyle and diurnal variation of the patient's sputum production. It is important that the airway clearance regimen is effective without unduly compromising the patient’s lifestyle.

Good practice points

• The duration and frequency of the airway clearance technique should be specific to the needs of the individual. This may alter with periods of infective exacerbation.

• Airway clearance therapy should be for 20–30 min once or twice daily.

How soon should the patient be reviewed after the initial assessment?

Initial assessment of an individual with non-CF bronchiectasis may take up to 1 h, with instruction in an appropriate airway clearance technique included. A review of the individual’s ability to effectively carry out the designated technique should be undertaken within 3 months of this initial appointment.396 At this review the optimal frequency and duration of any airway clearance regimen to optimise patient benefit and satisfaction can be discussed. Follow-up is at the discretion of the clinician based on efficacy of the demonstrated technique, understanding and disease severity. A patient should also be made aware of other airway clearance options. Some patients like to choose among different treatments. This gives the patient an element of control which may increase adherence to treatment.

What is the role of exercise?

Reduced exercise tolerance may be a problem for individuals with non-CF bronchiectasis; those with reduced exercise capacity and expiratory flow limitation have higher MRC dyspnoea scores.355 There is very little research on the effects of physical exercise in patients with non-CF bronchiectasis. A Cochrane review undertaken in 2003 concluded from the data available (two abstracts) that inspiratory muscle training improved endurance exercise and health-related quality of life.397 A further study investigated the effects of an 8-week high-intensity pulmonary rehabilitation (PR) programme and inspiratory muscle training (IMT) on patients with stable bronchiectasis.398 Thirty-two patients were randomly allocated to one of three groups: PR + sham IMT (PR-SHAM), PR + IMT (PR-IMT) or control. PR-SHAM and PR-IMT resulted in significant increases in the incremental shuttle walking test and in endurance exercise capacity compared with control. There were no statistically significant differences in the improvements in exercise between the two PR groups. Significant improvements in inspiratory muscle strength were observed in both the PR-SHAM and PR-IMT groups. There was no significant difference in the magnitude of the increase in inspiratory muscle strength between the two PR groups. Three months after the training programme the improvement in exercise capacity was maintained in the PR-IMT group but not in the PR-SHAM group. This indicates that PR is effective in improving exercise tolerance in subjects with bronchiectasis, but there is no additional short-term advantage of simultaneous IMT. However, IMT may be important in the maintenance of the training effects.

Are mucolytics and hyperosmolar agents of benefit in the long term to patients with bronchiectasis?

Bronchiectasis is characterised by hypersecretion and retention of mucus due to impaired mucociliary clearance.402 406 Children with bronchiectasis frequently have difficulty in expectorating sputum, especially during an infective exacerbation. There are no studies evaluating the use of mucolytics in children. Various agents have been tried in adults to reduce the mucus production and/or to increase mucus clearance with variable results.241 399 400 405 407 409–412

Most of the agents that have been used attack the physical properties of the mucus. Hyperosmolar inhalation has been shown to improve airway clearance in all the major chronic diseases characterised by sputum retention. Hypertonic saline inhalation and inhaled dry powder mannitol are known to accelerate tracheobronchial clearance, probably by inducing a liquid flux into the airway surface. Hypertonic saline may provide a useful adjunct to physiotherapy.390 Although short-term studies of mannitol indicate an improvement in mucociliary clearance,403 404 there are as yet no definitive clinical studies to confirm its use in children or adults with bronchiectasis.

Recombinant human DNase (rhDNase, dornase α, Pulmozyme) breaks down the DNA released at the site of infection by the neutrophils. DNA causes the sputum to become thick and tenacious; rhDNase makes the sputum less viscid and therefore easier to expectorate. This has been shown in a number of studies to be beneficial in CF. The potential for rhDNase to favourably influence symptoms in bronchiectasis has not been tested in children,401 but it has been well studied in adults and there is no evidence of benefit.241 399 Indeed, there is evidence of worsening lung function with rhDNase use in bronchiectasis.

There is a possibility of some action of carbocysteine in bronchiectasis with significant reduction in air trapping in a small trial408; however, there are insufficient data to support its clinical use. Bromhexine has been studied in acute exacerbations407 as an adjunct to antibiotic therapy and showed additional benefit in lung function and sputum. The Cochrane database suggests that bromhexine is the only mucolytic so far shown to be beneficial in the treatment of bronchiectasis exacerbations,401 407 but it is not widely available and not in the BNF.

Research recommendations

• Use of carbocysteine in bronchiectasis should be the subject of a randomised control trial to establish its clinical efficacy.

• Mannitol should be investigated further in a randomised controlled trial.

Are bronchodilators of use in bronchiectasis?

Are inhaled corticosteroids a useful treatment for bronchiectasis?

The safety and efficacy of inhaled steroids in the treatment of airway inflammation in asthma is well established. Inhaled steroids have a wide range of anti-inflammatory properties, especially in the context of chronic inflammation which plays a significant role in the pathophysiology of bronchiectasis.492

A randomised controlled crossover study of beclometasone versus placebo in adults with bronchiectasis419 showed an 18% reduction in sputum production but small changes in FEV₁ and PEF which, although statistically significant, were of doubtful clinical significance. A small 4-week study of inhaled fluticasone421 revealed a reduction in sputum inflammatory cells without significant changes in lung function. A larger study of 86 patients randomised to fluticasone or placebo for 12 months418 found no change in exacerbation frequency or lung function overall but suggested that sputum volume may be improved and that patients with Pseudomonas may gain more benefit. A further study suggested improvements in quality of life associated with inhaled corticosteroid treatment.420

There are no studies in children.

Are oral steroids indicated in the treatment of bronchiectasis?

In bronchiectasis there is no evidence for or against the use of oral steroids. A wider group of patients with obvious asthma (which may coexist with bronchiectasis) will primarily be managed with inhaled corticosteroids; a small percentage with ‘difficult’ asthma may require a maintenance dose of oral steroids.493

Leukotriene receptor antagonists and other anti-inflammatory agents

Leukotriene receptor antagonists (LRAs) have been identified as useful drugs in bronchial asthma. Leukotrienes in particular are potent bronchoconstrictor agents and act as a chemoattractant for eosinophils and also have a role in neutrophil-mediated inflammation. LRAs inhibit specific receptors for leukotrienes in the bronchiolar tissues so they may reduce bronchoconstriction, oedema, mucus secretion and eosinophil or neutrophil-mediated airway damage. Despite the theory that LRAs may be of benefit, there are no randomised controlled trials of their use in bronchiectasis in either children or adults.494

Other anti-inflammatory agents that have been studied in very small studies include nedocromil422 and indomethacin,423 neither of which showed any improvement in symptoms or lung function.

Definition of an exacerbation requiring antibiotic therapy

Adults and children

There are no randomised placebo-controlled studies evaluating the efficacy of antibiotics in exacerbations in adults or children although numerous studies (table AV, Appendix 2) indicate that antibiotics can improve symptoms and hasten recovery. Antibiotics are recommended for exacerbations that present with an acute deterioration (usually over several days) with worsening local symptoms (cough, increased sputum volume or change of viscosity, increased sputum purulence with or without increasing wheeze, breathlessness, haemoptysis) and/or systemic upset (figure 2). The goals for successful treatment are shown in figure 3.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Definition of an exacerbation needing antibiotic therapy.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Definition of successful treatment of an exacerbation. P, purulent; MP, mucopurulent; M, mucoid.

Daily symptoms of cough and sputum production are frequent and patients with more severe bronchiectasis often expectorate mucopurulent or purulent sputum and culture respiratory pathogens when apparently clinically stable. This is more common in adults. The presence of mucopurulent or purulent sputum alone or the isolation of a pathogen alone is not necessarily an indication for antibiotic treatment.

Good practice points

• The presence of mucopurulent or purulent sputum alone or the isolation of a pathogen alone is not necessarily indications for antibiotic treatment, particularly in adults.

• Antibiotics should be given for exacerbations that present with an acute deterioration with worsening symptoms (cough, increased sputum volume or change of viscosity, increased sputum purulence with or without increasing wheeze, breathlessness, haemoptysis) and/or systemic upset.

Managing patients with exacerbations

Good practice points

• Patients with an infective exacerbation of bronchiectasis should be assessed for the need for inpatient or outpatient treatment.

• Patients treated for exacerbations should have appropriate assessments according to treatment setting.

Which antibiotic regimen is recommended for exacerbations?

Good practice points

• Previous sputum bacteriology results can be useful in deciding which antibiotic to use. Table AI highlights the first-line and alternative treatments for the common bacterial pathogens implicated in exacerbations of bronchiectasis (see BNF for Children for dosage; use doses for severe infection495).

• Where possible, sputum (spontaneous or induced) or a cough swab should be obtained for culture prior to commencing antibiotics.

• Empirical antibiotics can then be started while awaiting sputum microbiology.

• In general, antibiotic courses for 14 days are standard. If there is no previous bacteriology, the first-line treatment is amoxicillin for 14 days or clarithromycin for 14 days in patients who are allergic to penicillin (see BNF for Children for dosage; use doses for severe infection495).

• Children not responding to empirical antibiotic courses should have an organism identified by cough swab or later by induced sputum/bronchoalveolar lavage.

• Intravenous antibiotics should be considered when patients are particularly unwell, have resistant organisms or have failed to respond to oral therapy (this is most likely to apply to patients with P aeruginosa).

When are combination (dual) antibiotic regimes required?

Do long-term oral antibiotics influence long-term outcome?

Good practice points

• Consider long-term antibiotics in children with frequent symptoms or severe disease.

• Assess rate of progression of bronchiectasis on long-term antibiotics (repeating HRCT 24–48 months later).

• Where possible, long-term antibiotic regimens should be determined by respiratory microbiology.

• Long-term use of oral quinolones should be avoided.

Do long-term nebulised antibiotics influence long-term outcome?

Research recommendation

• Randomised controlled trials are needed to investigate the efficacy of nebulised antibiotics in patients with bronchiectasis chronically colonised with P aeruginosa and other organisms.

Good practice points

• Long-term nebulised antibiotics are for children with frequent recurrent exacerbations (or deteriorating bronchiectasis) despite long-term oral antibiotics or if oral antibiotic therapy is not appropriate.

• If a child with chronic P aeruginosa meets the criteria for long-term antibiotics, these should be considered; regimens are shown in table AII.

• The choice of antibiotic should be guided by the antibiotic sensitivity results.

Are rotational antibiotics recommended?

Good practice point

• Rotating courses of antibiotics are not recommended routinely.

Should an attempt be made to eradicate organisms from the lower respiratory tract?

Adults and children

Some organisms, if able to colonise the lower respiratory tract in patients with bronchiectasis, can be difficult to treat. P aeruginosa often requires expensive and potentially toxic intravenous antibiotic therapy that is inconvenient and time-consuming for the patient; MRSA may be resistant to multiple antibiotics and has implications for infection control in the management of a patient with bronchiectasis in hospital. Colonisation with P aeruginosa is also associated with worse symptoms515 and quality of life scores233 and may lead to accelerated decline in FEV₁.233 483 While there are no studies to guide practice following the first isolate of either organism, an attempt to eradicate seems pragmatic. Figures 4 and 5 outline strategies derived for the eradication of P aeruginosa from the sputum of patients with CF and individual clinicians will decide which and how aggressive a strategy to employ in the setting of non-CF bronchiectasis.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Eradication algorithm for Pseudomonas aeruginosa in adults. Attempt to eradicate with a 2-week course of oral ciprofloxacin (step 1). If step 1 fails, further regimens may be considered (step 2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Eradication algorithm for Pseudomonas aeruginosa in children. Use doses according to the Children’s BNF. Use doses for severe infection.444 Attempt to eradicate with a 2-week course of oral ciprofloxacin (step 1). If step 1 fails, further regimens may be considered (step 2).

Good practice points

• In patients who have P aeruginosa isolated for the first time, an attempt should be made to eradicate using 14 courses of oral ciprofloxacin (figures 4 and 5).

• Failure to eradicate P aeruginosa with oral treatment may lead to consideration of intravenous and/or nebulised eradication therapy although there is currently insufficient evidence to recommend this (figures 4 and 5).

• For patients in whom MRSA is isolated in the sputum, an attempt to eradicate the organism should be made with drug(s), dose and duration guided by local microbiological advice.

When should opportunist mycobacteria be treated?

This is discussed in the BTS and ATS guidelines.516 517

What is the impact of long-term antibiotics on antibiotic resistance?

Good practice point

• Long-term antibiotics may result in antibiotic resistance in individual patients and alternative antibiotics should be chosen depending on sensitivity results.

Is there clinical relevance of in vitro antibiotic resistance patterns?

Is there a role for surgery in the management of patients with bronchiectasis?

Numerous studies have reported lung resection surgery for bronchiectasis in both children and adults with the principal aim of improving symptoms. Indications for surgery in the literature have included failure of medical/conservative therapy,460 464–467 518–521 recurrent respiratory infections and persistent sputum production,456 457 460 520–522 haemoptysis,460 464–467 518 523 524 chronic cough459 460 464 520 524 and persistent lung abscess.465–467 518 521

While lung resection surgery has not been subjected to randomised controlled trials, certain principles that seem to predict success emerge from the literature. First, the resection of bronchiectatic lung should be complete,458 461 464 465 518 520 525 526 so diseased lung must be localised463 465 518 and at least 10 segments should remain after surgery.527 There is no available information on contraindications, although it is unlikely that there are any absolute contraindications. Factors that appear to have an adverse outcome in some studies and so might be seen as relative contraindications include non-cylindrical disease,462 520 Pseudomonas grown on sputum culture,526 residual disease after resection and non-localised disease.458 461 463

Perioperative morbidity and mortality has been reported in many case series. The precise definition of morbidity varies, as does the reported incidence with complication rates ranging from 0%462 464 521 to >20%.518 520 523 The higher rates are in the older series and a morbidity rate of 10–19% is accepted by most authors.456 465 526 528 Perioperative mortality is low with all series reporting rates <5%460 465 518 519 521 and many with figures of 0%.458 466 467 520 524 526

With regard to long-term outcome, studies have reported many different variables with most concentrating on symptomatic improvement and subsequent admission rates. Improvement rates of 50–80%456 459 519–521 528 and >90%461 525 have been reported, although these studies do not formally control for a comparison group receiving best medical therapy.

Recommendation (3468–473)

• Bronchial artery embolisation and/or surgery is first-line therapy for the management of massive haemoptysis. [D]

Recommendations (3474–477)

• NIV can improve quality of life in some patients with chronic respiratory failure due to bronchiectasis. [D]

• Evidence for survival benefit is lacking, although for some patients are successfully treated with NIV for significant lengths of time which may reduce hospitalisations. [D]

Is there a role for lung transplantation in advanced bronchiectasis?

Lung transplantation is available for end-stage cardiopulmonary disease in children and adults, although there is a paucity of literature in bronchiectasis specifically. A general guideline is to refer patients for an evaluation for lung transplantation if the FEV₁ is <30% or if there is a rapid progressive respiratory deterioration despite optimal medical management.530 531 In patients with poor lung function, the following additional factors should lower the threshold for considering referral for transplantation assessment: massive haemoptysis, severe secondary pulmonary hypertension, ICU admissions or respiratory failure (particularly if requiring NIV). It should be noted that antibody deficiency is not an absolute contraindication to transplantation. The age of the patient predicts outcome following transplantation and discussion with a transplant centre is the best way to assess whether patients aged >60 years should be referred.

When should oxygen therapy be used?

Guidance on the appropriate use of oxygen in acute and chronic settings can be found in the BTS guideline for emergency oxygen use in adult patients.532