Comparative studies of frequency, incidence and prevalence
Registries of the epidemiology of different ILDs have been compiled in several countries. However, they remain scarce due to the difficulties that arise in obtaining a specific diagnosis as many of these conditions are rare. Many of the available data come from prospective registries of data reported by respiratory physicians, for example in Flanders (Belgium), Germany, Italy, Spain and Greece (table 2). These registries have the disadvantage that the registered populations may not be representative of the true populations of patients. They also do not necessarily represent the true overall incidence and prevalence. However, they do allow comparison of the relative frequencies of the different ILDs. The data show that the most frequent ILDs are IPF and sarcoidosis, which together comprise about 50%. The data also show some interesting differences between countries, such as lower proportions of IPF in Flanders, of sarcoidosis in Spain, of ILD associated with CTD in Germany, and of extrinsic allergic alveolitis (hypersensitivity pneumonitis) (EAA) in the Italian, Spanish and Greek registries.
In addition, in Denmark there is a population-based registry, encompassing the entire population and comparing the periods 1995–2000 and 2001–2005, which undoubtedly provides more complete data. This registry shows a lower frequency of IPF but a higher frequency of ‘atypical’, nonspecific fibrosis in the second period, which can, at least partly, be attributed to the new classification of the IIPs with a more specific and restricted definition of IPF.
|Sarcoidosis||112 (31)||69 (26)||83 (35)||344 (30)||87 (12)||76 (15)||330 (34)||60 (23)|
|IPF/IIP||62 (17)||50 (19)||76 (32)||417 (37)||287 (39)||215 (42)||234 (24)||66 (25)|
|COP/BOOP||10 (2.3)||9 (3.4)||16 (6.8)||57 (5.0)||38 (5.1)||53 (10)||51 (5.3)||18 (7.0)|
|(C)EP||9 (2.2)||7 (2.7)||27 (2.3)||21 (2.2)||7 (2.7)|
|CTD||27 (7.5)||19 (7.2)||5 (2.1)||69 (9.3)||51 (19)||120 (12)||30 (12)|
|Vasculitis#||5 (1.4)||4 (1.5)||2 (0.8)||25 (2.2)||14 (1.5)||6 (2.3)|
|EG/HX||13 (3.6)||7 (2.7)||73 (7.2)||6 (0.8)||15 (3)||37 (3.8)||7 (2.7)|
|EAA||47 (13)||32 (12)||25 (11)||50 (4.3)||38 (5.1)||34 (7)||25 (2.6)||7 (2.7)|
|Drug¶||12 (3.3)||12 (5)||6 (2.6)||21 (1.8)||21 (4)||17 (1.8)||4 (1.5)|
|Pneumoconiosis+||19 (5.0)||18 (6.8)||6 (2.6)||55 (7.4)||20 (2.0)||8 (3.1)|
|33 (9.1)||27 (10)||12 (5.1)||69 (9.3)||82 (8.5)||40 (15)|
|Others||13 (3.8)||10 (3.8)||124 (11)||76 (10)||9 (2)||15 (1.5)||6 (2.3)|
Table 2 – Comparison of the distribution of interstitial lung diseases (ILDs) in respiratory physicians’ prospective registries. Data are presented as n (%), unless otherwise stated. RENIA: Registry of Interstitial Pneumopathies of Andalusia; SEPAR: Sociedad Española de Neumología y Cirugía Torácica; IPF: idiopathic pulmonary fibrosis; IIP: idiopathic interstitial pneumonia; COP: cryptogenic organising pneumonia; BOOP: bronchiolitis obliterans organising pneumonia (not necessarily cryptogenic); (C)EP: (chronic) eosinophilic pneumonia; CTD: connective tissue disease; EG/HX: eosinophilic granuloma/histiocytosis X; EAA: extrinsic allergic alveolitis (hypersensitivity pneumonitis). #: Goodpasture’s, granulomatosis with polyangiitis (Wegener’s), Churg–Strauss, etc.; ¶: radiation was also included in the Italian and SEPAR registries; +: coal worker’s pneumoconiosis was excluded in the Flemish, Italian and SEPAR registries.
Incidence and prevalence of specific subgroups of ILD
The most important ILDs are sarcoidosis, IPF (previously called cryptogenic fibrosing alveolitis, mainly in the UK), EAA, ILD as a feature of CTD, drug-induced ILD and pneumoconiosis (for further information on the latter, see chapter 24 ). These are discussed below in more detail.
In the UK, general practice data have suggested an incidence of approximately 3 cases of sarcoidosis per 100 000 people per year (assuming a mean disease duration of 2 years). In another UK study, a similar incidence of 5 cases per 100 000 people per year was found, which remained stable over time. The incidence appears to be higher in some countries (e.g. Scandinavia), in some population groups (black Americans) and in some families. In Spain, an incidence of only 1.2 cases per 100 000 inhabitants per year has been reported and similar low numbers have also been found in Portugal. International registries have, furthermore, demonstrated differences in the typical clinical presentation of sarcoidosis between various countries; for example, the benign subacute form known as Löfgren syndrome (hilar lymphadenopathy, fever, arthritis and painful cutaneous lesions called erythema nodosum) is particularly frequent in Scandinavia.
|Incidence per 100 000 people per year|
|Nonspecific pulmonary fibrosis||7.49||9.76|
|CTD (and pulmonary haemorrhagic
Table 3 – Incidence of interstitial lung diseases in a population-based registry in Denmark (total population: 5.4 million). IPF: idiopathic pulmonary fibrosis; CTD: connective tissue disease.
Idiopathic pulmonary fibrosis
Interpretation of the incidence and prevalence data needs to take into account the current definitions and classifications of the ATS/ERS consensus reports. In a UK study published in 1999, the estimated prevalence of IPF was 15–18 cases per 100 000 people; a median survival time from diagnosis of approximately 3 years thus corresponds to the estimated incidence of about 5 cases per 100 000 people per year obtained from several studies in the UK. While incidence and mortality data suggest that the frequency of IPF is increasing in the UK, a decrease was found in the Danish registry between 1995 and 2005.
Extrinsic allergic alveolitis (hypersensitivity pneumonitis)
A very large number of causes of EAA have been reported, such as EAA in farmers, pigeon breeders and budgerigar fanciers, and EAA due to repeated exposures to isocyanates, fungi, mollusc shells, etc. There are marked variations in the prevalence of the specific disease types between countries, which is not only due to differences in the diagnostic criteria but also to the possible presence of industrial manufacturing plants, and to differences in local seasonal climate, geographic conditions and altitude. Smoking history is also important as patients with EAA are less likely to have been smokers than the general population. In a general-population cohort study based on a UK primary care database, a stable incidence of ~0.9 cases per 100 000 people per year (i.e. about 600 new cases each year) was found between 1991 and 2003.
The two most extensively studied types of EAA are farmer’s lung and pigeon breeder’s lung. Among Swedish farmers, the incidence of EAA is ~20 cases per 100 000 people per year. The reported prevalence across several countries varies between 4–170 per 1000 farmers, depending on local conditions and the diagnostic criteria used, while the frequency of hospital admission due to farmer’s lung has been estimated in Finland and Sweden to be ~3–5 per 10 000 farmers per year. In pigeon fanciers, the prevalence of clinical disease has, in the past, been estimated at ~1 case per 1000 breeders, but more recently, a prevalence of more than 10% was reported in those with regular high exposure.
Interstitial lung disease associated with connective tissue disease
It is difficult to provide accurate data on the prevalence of ILD in CTD because this depends very much on the diagnostic methods used. Rheumatoid arthritis is estimated to occur in 2% of the population and evidence of ILD can be found using chest radiography and lung function testing in up to 20% of these patients. The prevalence of systemic lupus erythematosus is 40 cases per 100 000 people, with clinically relevant ILD in an estimated 10% of cases. The prevalence of progressive systemic sclerosis is 10 cases per 100 000 population, with pulmonary fibrosis found at autopsy in up to 75% of patients and lung function impairment seen in up to 90%.
Drug-induced lung diseases
Drug-induced lung diseases account for ~1.5–5% of all ILD (table 2), but these percentages probably underestimate the real frequency. More than 300 drugs are recognised as causing respiratory disease, particularly ILDs. Specifically, drugs such as amiodarone, bleomycin, methotrexate and nitrofurantoin, as well as radiation therapy, can all cause pulmonary fibrosis. Registration of, and information about, drug-related lung diseases has been coordinated by the Pneumotox group in Dijon, France ( www. pneumotox.com ) for the past two decades.
Inorganic dust-induced occupational lung fibrosis
Pneumoconioses, such as silicosis, coal worker’s pneumoconiosis and asbestosis are discussed in chapter 24 .
Age and sex distributions
ILD, particularly IPF and drug-induced fibrosis, occur mostly in older subjects, while sarcoidosis shows a predominance in young adults of both sexes, and in women over 50 years of age. The rarer pulmonary Langerhans’ cell histiocytosis is notable for its typical occurrence in young cigarette smokers.
While IPF, EAA and inorganic dust-induced occupational ILD are more frequent in males, sarcoidosis and ILD associated with CTD are more frequent in women.
Annual mortality and survival
Extensive mortality data for the majority of European countries is available from the World Health Organization (WHO) World and Europe Mortality Databases. There are clearly differences between countries, which are partly real and partly due to differences in diagnostic and therapeutic strategies. The highest mortality rates due to ILD, of more than 2.5 per 100 000 people, are recorded in the UK, Ireland, Scandinavia, the Netherlands and Spain (figure 1).
Figure 2 shows that, among the most important ILDs, ‘chronic ILD’ (International Classification of Disease (ICD)-10 code J84, which includes IPF and other forms of fibrosing alveolitis) has the highest mortality rate, followed by ILD associated with CTD (ICD-10 code M32–M36). The mortality rates of sarcoidosis (ICD-10 code D86) and particularly EAA (ICD-10 code J67) are much lower. For sarcoidosis, the age-standardised mortality rate (per 100 000 people) in many countries is less than 0.15, but it is more than 0.30 in Denmark and Ireland. For chronic ILD, the mortality rate in most countries is less than 2 per 100 000 people, but it is at least 4 per 100 000 people in the UK, Ireland and Malta. For EAA, the mortality rate is below 0.05 per 100 000 people in the majority of the countries. For ILD associated with CTD, the mortality rate is generally below 0.6 per 100 000, but it is higher than 0.8 per 100 000 in Denmark and Norway.
Although the WHO mortality data may be more complete than clinical registries of incidence or prevalence, they should be interpreted cautiously. Firstly, the relation of mortality data to incidence or prevalence rates varies considerably between disease conditions. About 50–70% of patients with IPF will die from this disease and thus the mortality rate should be about 50–70% of the incidence rate, but only about 5% of sarcoidosis patients will die of the disease and thus the mortality rate is only 5% of the incidence. Furthermore, for systemic diseases, such as sarcoidosis and CTD, the WHO mortality data do not distinguish whether patients had related ILD or, if present, whether the ILD contributed to death; thus the mortality rates in figure 2 only partly reflect deaths due to ILD. The WHO ICD classification may also be misleading due to peculiarities of the ICD codes, and owing to the fact that the definition of the codes changed from ICD-8 to ICD-9 and ICD-10, especially for IPF. In the figures presented in this chapter, the ICD-10 code J84 (chronic ILD) is used. This is broader than just IPF. In ICD-9, a specific disease code for IPF/cryptogenic fibrosing alveolitis (code 516.3) was introduced for the first time. Studies of the use of code 516.3 in death certificates and hospital admissions data in the UK, suggest that most patients coded as having IPF (using IDC-9 codes) do, indeed, have this disease, but that about half of the individuals known to have IPF are not coded correctly and many receive the less precise code of ‘post- inflammatory fibrosis’ (code 515). Consequently, countries with higher frequencies of post-inflammatory fibrosis tend to have lower frequencies of IPF and vice versa. The ICD-10 code J84 (chronic ILD) includes IPF as well as other chronic ILDs (such as other IIPs, ‘atypical’ nonspecific lung fibrosis, post-infectious ILD, drug-induced ILD) without subdivision into these entities.