Bronchiectasis (BE) are the end result of many diseases. The etiological spectrum and the percentage of idiopathic BEs vary depending on the origin of the patient and the diagnostic protocol.1–11 Determining the cause of the process using an extensive protocol of diagnostic tests is justified, since identifying the etiology will affect treatment and prognosis.12–14 In a review1 of 35 European studies, 8 Asian studies, 6 from Oceania, 5 from South America, 1 North American, and 1 African study, the most frequently identified causes were: post-infective: 29.9%; immunodeficiencies: 5%; chronic obstructive pulmonary disease (COPD): 3.9% (although this was an exclusion criterion in many studies)1–4; connective tissue diseases (CTD): 3.8%; allergic bronchopulmonary aspergillosis (ABPA): 2.6%; ciliary dysfunction (CD): 2.5%; and asthma: 1.4%. In total, 44.8% were classified as idiopathic. Lonni et al.9 have published the largest cohort to date, with 1258 European patients. In Spain, no data focusing exclusively on the etiology of BE have been published, although some etiological data have been reported in other studies.15–20 The Spanish Historical BE Registry of the Spanish Society of Respiratory Disease (RHEBQ-SEPAR) was created in 2002 to establish standardized criteria for the diagnosis and treatment of patients with BE, and to determine the characteristics of this disease in Spain. This was an electronic database, accessible from the SEPAR website (www.separ.es/bronquiectasias), in which all respiratory medicine experts were invited to participate. The aim of this study was to evaluate BE etiology in a large patient cohort (included in the RHEBQ-SEPAR registry), the largest published to date, comprising a reasonably homogeneous sample, representative of the patients seen by pulmonologists. Another aim was to analyze the relationship between the different etiologies and demographic, clinical and severity factors, and the differences between patients with idiopathic BE compared to those with post-infective BE, and those with other etiologies.

This study evaluated the etiology of BE, determined using a comprehensive standardized protocol of diagnostic tests, in the largest series of BE patients published to date. Although we cannot definitively establish the prevalence of the different etiologies, as this was a voluntary registry in which not all possible patients were included, the series is reasonably representative of Spanish BE patients, due to the participation of respiratory medicine units from different autonomous communities throughout the state.

BE etiology could be identified in a large proportion of patients enrolled in the RHEBQ-SEPAR registry (75.8%), as a wide range of etiological tests was recommended,12 and the patients were followed by respiratory medicine experts who generally follow the SEPAR recommendations. Cohorts published to date have shown widely differing results (26%–90%) due to differences in study populations, criteria for defining etiologies, diagnostic studies performed, and setting in which the patient was seen.1–11 Idiopathic BE was more common among study patients seen in general clinics1,10,11,26 compared to those seen in pulmonology or specific units,1,2,4,7,9,15,27 as is the case for patients enrolled in the RHEBQ-SEPAR registry. In a review of studies published in all countries, Gao et al.1 found a rate of 44.8% of idiopathic BE, and 41.1% when only European studies were analyzed.

In our study, the most common known causes were post-infective (30%), followed by CF, PID, COPD, asthma, CD, and CTD, in line with other studies.1,4,8,9,12,15,26,27 Of the 35 European studies reviewed by Gao et al.1 (6364 patients), the mean rate of post-infective BE was 31.2%; PID: 5.3%; COPD: 5.2%; CTD: 4%; ABPA: 2.8%; CD: 2.7%; asthma: 7%; and inflammatory bowel disease: 1%. In Spain,15–20 a study in 839 patients from 7 specific BE units that also used the NSBQ diagnostic protocol,12 aimed at constructing and validating the FACED score system15 found that the cause in 37% was idiopathic, in 29.3% post-infective, in 8.3% immunodeficiencies, in 3.5% COPD, in 3.2% CD, and in 2.1% CTD. However, a study in the US7 of 106 patients identified causal factors in more than 90% of cases, the most common being immune disorders and CTD, but this is probably because it was conducted in a highly specialized center that used in-depth immunological evaluations.7 In contrast, the RHEBQ-SEPAR registry was created by respiratory medicine experts, and in our setting, diseases such as CTD, and in some cases immunodeficiencies, atopic asthma and ABPA, are seen by other specialists, so this may have led to a lower rate of some of these diseases in our series.

In our series, infections, primarily tuberculosis, continue to be the most common cause of BE. However, the possibility remains that some cases of post-infective BE might be associated with an undetected immune or genetic alteration28 that would make the patient more susceptible to infection, particularly those with concomitant rhinosinusitis.2,4 In a systematic review1 in adults, tuberculosis was also the predominant etiology for post-infective BE, and was more common in Asia than in Europe.1,6,10,11,29 In children, pneumonia was the main cause.30 In developing countries1,6,29 or in indigenous communities,5,31 infections are also more prevalent, due to poorer health and social welfare conditions.

CF also accounts for a high percentage of BE, since the RHEBQ-SEPAR included all causes of BE and the numbers of adult patients with CF seen by pulmonologists in CF units is growing,32 most of whom participated in the registry.

Rates of PID were also high (greater than those published in Europe1,9 but lower than in the U.S.7), since detailed studies for ruling out or supporting this diagnosis were recommended12 before BE was classified as idiopathic.

In our series, BE was caused by COPD in 7.8% of patients. The earlier etiological studies did not include COPD,2,4 but since the publication of the studies by Patel et al.33 and Martinez-García et al.34,35 which showed that more than 50% of patients with moderate-severe COPD had BE, reported rates are increasing (11%–15%).8,9,27,36 As expected, patients with COPD were mainly men, with a history of smoking, older age, greater frequencies of CBI due to PA, and poorer lung function.

Clinical differences were found depending on etiology, with similar results to those published in other series.9,15,27 Patients with CD had the highest rates of CBI caused by different microorganisms and asthmatics had the lowest rates.1,4,9 CD and PID patients had higher rates of rhinosinusitis, in line with other studies.1,2

In our series, only 27% of patients had FEV1<50%, and in the multivariate analysis, the etiologies most often associated with poorer lung function were post-infective BE, BE caused by asthma, and particularly, BE caused by COPD. Patients with PID and idiopathic BE were those who had better lung function. Lonni et al.9 evaluated the relationship between different levels of severity and different BE etiologies, and did not find any important differences, except for a higher rate of severe forms among BE caused by COPD and a lower rate among idiopathic BEs, coinciding with our findings. The association between COPD and BE produces a more severe clinical phenotype with a poorer prognosis,9,34–36 although there is some debate as to whether the relationship is one of causality37 or comorbidity.38

When idiopathic BEs are compared to post-infective BEs, the idiopathic forms were diagnosed earlier, had more symptoms of rhinosinusitis, were predominantly bilateral, had better lung function, and a lower rate of CBIs caused by PA, in line with previous studies.1,4 In post-infective BE, the delayed diagnosis may be associated with a variable time between the initial infection that caused BE and the subsequent superinfection that would lead to the development of new symptoms.4,8,12

When idiopathic BE is compared to NCFBE of any known cause, we found that it presents as a special phenotype, characterized by lower smoking rates, higher BMI, better lung function, fewer CBIs caused by PA, and a clear predominance of women, confirming the findings of other published series in populations with fewer patient numbers.9,27 Several studies have been published recently36,39 that have tried to define specific BE phenotypes associated with prognosis, looking at factors such as age, exacerbations, or colonization by microorganisms, as well as the role of the different etiologies. Martínez-García et al.39 observed that the phenotypes associated with a better prognosis included a greater number of patients with PID, CD or idiopathic BEs than those associated with poorer prognosis, most of whom were COPD patients. Aliberti et al.36 also found that patients with idiopathic BE were also more inclined to present less severe phenotypes with less CBI caused by PA, while COPD phenotypes were more severe and more frequently involved CBI caused by PA. This also supports our results.

This study is the first analysis of the RHEBQ-SEPAR registry that explores the causes of BE in the largest series published to date. BE etiology was identified in a large proportion of patients (75.8%), the most common causes being post-infective (30%), followed by CF, PID, COPD, asthma, CD, and CTD. Different clinical phenotypes associated with etiology can be identified. Patients with idiopathic BE had less severe disease, higher BMI, better lung function, less frequent CBI caused by PA, and a clear predominance of women. Longitudinal studies are needed to confirm the relationship of BE etiology with clinical manifestations and progress of this disease.

Casilda Olveira, Alicia Padilla and Miguel Ángel Martínez-García have contributed to the conception and design of the study, data acquisition, analysis and interpretation, and to the preparation of the article, and are guarantors of the manuscript. The other authors contributed to data acquisition and critical review of the article.

The authors declare no conflicts of interest.

The authors thank Francisco Rivas (Research Department, Agencia Sanitaria Costa del Sol, Marbella, Malaga) for his contribution to the statistical analysis of the data.