Investigators interested in performing clinical studies of rare diseases or trials of orphan drugs are faced with challenges usually not encountered in clinical trials of larger populations. Obvious drawbacks include the small size of the trial population and the fact that patients are often geographically dispersed, thus with an inherent diversity of healthcare system. This is particularly true for disorders with geographic prevalence (Hermansky-Pudlak syndrome in Puerto Ricans) or gender predominance (lymphangioleiomyomatosis [LAM] in women of childbearing age). In these circumstances, it is almost impossible to undertake the randomized, double/blind, placebo-controlled studies that now represent the gold standard of clinical trials, on which quality of evidence judgements are made. In this regard, the choice of appropriate trial methodology and meaningful outcome parameters is a matter of intense debate in even those disease such as idiopathic pulmonary fibrosis (IPF) where there are robust end point data and the disease is sufficiently common for large clinical trials of therapy to have been undertaken. In rarer diseases, in order to prove efficacy in a study with small patient numbers, the compound under investigation needs to show a stronger treatment effect than in a study with large numbers, thus highlighting the need for more robust methods of data analysis for small samples. Likewise, drawing conclusions from trials performed in a limited number of patients may be dangerous. Good examples include several early studies investigating antiestrogen therapy – consisting of surgical castration by oophorectomy, administration of tamoxifen, progesterone, and gonadotropin-releasing hormone agonist or luteinizing hormone-releasing hormone – that have reported beneficial effects in LAM [4, 5]. Subsequent careful scrutiny of some of these studies has, however, revealed that while the treatment under investigation may have improved some aspects of the disease, for example chylothorax or chylous ascites, other affected organs including especially pulmonary involvement, were not affected and in some instances progression was the outcome. Now that lung transplantation has become an acceptable treatment option for patients with LAM, more experimental treatments must be used with caution because of the potential complications due to adverse effects that might jeopardise the eligibility for or outcome after lung transplantation. A good example is the use of surgical or medical castration that may not produce any beneficial effect on the disease course, but may exert long-term effects on bone metabolism, particularly in the postoperative period of lung transplantation.

It is estimated that 80 % of the identified rare diseases have a genetic origin. However, the importance of environment triggers, in addition to the genetic susceptibility, is becoming increasingly clear. Many diseases require this gene/environment interaction to manifest. Furthermore, different combinations of genetic susceptibility and individual trigger agents are also likely to explain diversity in terms of organ manifestations and disease severity, which, in turn, accounts for the inconsistent genotype-phenotype correlations. A good example includes chronic beryllium disease where a powerful genetic predisposition requires less antigenic stimulus but stronger environmental exposures can provoke disease in those individuals who have less strong susceptibility genotypes [6].

Rare lung diseases generally affect individuals from birth through about age 60, and are uncommon in the elderly. Some conditions display racial and ethnic prevalence. For instance, sarcoidosis varies in prevalence and severity across ethnic boundaries(3). Available measures of prevalence suggest that it is not a common disease. United Kingdom mass surveys in the 1950s and 1960s disclosed radiographic abnormalities consistent with sarcoidosis in 9 [7] to 36 [8] per 100,000 of those screened. Similar studies in Scandinavia, carried out over the same decades, revealed a combined prevalence of 28 per 100,000 examined persons [9]. In 1964, Bauer and Löfgren [10] summarized the findings from 29 surveys (in ten cases nationwide) carried out in 24 countries; the results varied widely from 0.2 per 100,000 (in Portugal, Brazil, and Uruguay) to the highest figure of 64 per 100,000 in Sweden. In general the prevalence is higher the greater the degrees of latitude from the equator for reasons that are not understood. In the USA, sarcoidosis is more common in African Americans than whites. Applying cumulative incidence estimates, the lifetime risk of sarcoidosis is 2.4 % for African Americans and 0.85 % for American Whites [11]. The wide variation in these estimates presumably reflects differences in diagnostic labelling and in the age, gender, and morbid distributions of the screened populations. In addition, specific sarcoidosis phenotypes are more prevalent in certain populations, such as uveitis and cardiac involvement in Japanese, Löfgren’s syndrome (an acute and self-limiting form of sarcoidosis characterized by bilateral hilar lymph adenopathy, erythema nodosum/arthralgia and uveitis) in Scandinavians, lupus pernio (a chronic purplish indurated lesion seen mainly on ears, cheeks, lips and nose) in Puerto Ricans. Rare lung disease may also display regional variation. This is the case of Hermansky-Pudlak syndrome, a disease characterized by oculo-cutaneous albinism, a bleeding diathesis, and diffuse lung fibrosis (prevalence of 1 in 1,800 in Puerto Rico but only isolated case reports and small clusters in the rest of the world) [12, 13], and in pulmonary alveolar microlithiasis, a disease associated with sand-like particles in the lung, which occurs predominantly in Japan and Turkey [14].

The cause and pathophysiology of rare diseases are largely unknown. Up to 2009, one or more responsible genes were identified for only 2,105 of the over 6,000 rare diseases listed on the Orphanet website (www.​orpha.​net). The environment, including exposure to microorganisms, may also play a role, particularly in the presence of a patient’s compromised immune system, making these disorders varied and complex. However, for the vast majority of these diseases, no research is being conducted into causation, which, in turn, complicates both diagnosis and research into interventions. In fact, for some entities there are no guidelines and current recommendations are based on very limited data and expert opinion.

Rare lung diseases are often difficult to diagnose because of inconsistent case definition. For instance, hepatopulmonary syndrome is a rare disorder defined by a triad of liver disease, intrapulmonary vascular dilatation, and abnormal gas exchange [15]. It may also be a rare complication of more common chronic liver disease, such as liver cirrhosis [16]. However, the definition of “abnormal gas exchange” has varied widely in the published literature [17]. Diverse diagnostic thresholds lead to variable prevalence, render it difficult to compare studies and complicate patient recruitment owing to confused selection criteria. Expert consensus statements and guidelines – not available for most rare diseases – would undoubtedly facilitate consistent disease definitions.

Pulmonary involvement from rare diseases may represent only one end of a spectrum of clinical manifestations. This is the case, for instance, of Birt-Hogg-Dubé (BHD) syndrome, an autosomal dominant disorder caused by germ line mutations in the FLCN (folliculin) gene located on chromosome 17p11.2, and characterized by skin fibrofolliculomas, multiple lung cysts, spontaneous pneumothorax, and renal cancer [18]. BHD-associated skin lesions may also include angiofibroma, which are more typically associated with tuberous sclerosis. In turn, tuberous sclerosis may manifest with pneumothorax (caused by rupture of lung cysts), and renal cysts or tumours and should therefore be considered in the differential diagnosis of BHD [19]. The diagnosis of BHD is based on both clinical features and histology. However, the wide variability of clinical expression and the sporadic (in the majority of cases) occurrence of renal cancer or pneumothorax make the diagnosis challenging.

Pulmonary alveolar proteinosis (PAP) is a rare condition characterized by the accumulation of surfactant within alveolar macrophages and alveoli. Based on recent data from basic science and translational research, PAP is now recognized as a highly heterogeneous syndrome belonging to a much larger group of disorders of surfactant production (A) and clearance (B) – collectively known as disorders of surfactant homeostasis. The former group include mutations in genes encoding surfactant proteins (SP)- B, and C [20] or proteins involved in surfactant lipid metabolism (i.e., ABCA3; [21]). Conversely, PAP syndrome belongs to the second category of disorders, and can be either idiopathic (primary PAP) or secondary (to inhalation of dust, such as silica, or underlying immunologic and hematologic diseases that alter macrophage function). Primary PAP appears to be an autoimmune disorder associated with the presence of neutralizing autoantibody directed against granulocyte macrophage colony-stimulating factor (GM-CSF; [22]). Deficient GM-CSF activity, in turn, results in defective alveolar macrophages that are unable to maintain surfactant homeostasis and display defective phagocytic and antigen-presenting capabilities [23]. Recent studies suggest that neutrophil dysfunction, contributes to the increased susceptibility to lung infections observed in PAP [24]. Data from animal models suggest that the phenotypic and immunologic abnormalities of PAP can be corrected by GM-CSF augmentation therapy, although the effectiveness of this approach is not obvious since patients have autoantibodies (whose origin remains a mystery) to GM-CSF rather than abnormal protein levels.

Despite the complexity of the different diseases that are included in the “disorders of surfactant homeostasis”, this is a wonderful, if sadly uncommon, example of the way in which basic, animal-based, research can digitate so effectively into an understanding of the basis of human disease that in turn stimulates the development of an effective, novel, treatment strategy.

For many individuals who develop a rare disease, the period between the emergence of the first symptoms and the appropriate diagnosis often involves unacceptable and highly risky delays, as well as in many instances the wrong diagnosis being made that leads to the administration of inappropriate and sometimes dangerous treatments. A survey of the delay in diagnosis for eight rare diseases in Europe has been conducted by EURORDIS (European Organization for Rare Diseases) in collaboration with 67 European rare disease organizations [1]. The main findings of this survey were that 25 % of patients had to wait between 5 and 30 years from early symptoms to confirmatory diagnosis of their disease. Before receiving a confirmatory diagnosis, 40 % of patients first received an erroneous one and others received none. Twenty-five percent of patients had to travel to a different region to obtain the final diagnosis and 2 % had to travel to a different country. The diagnosis was announced in unsatisfactory terms or conditions in 33 % of cases, and in unacceptable conditions in 12.5 % of cases. The genetic nature of the disease was not communicated to the patient or family in 25 % of cases. Intuitively, the consequences of misdiagnosis include clinical worsening of the patient’s health – even leading to the death of the patient – and loss of confidence in the healthcare system. This is utterly unacceptable – imagine if this had occurred to a member of one’s own family – and cannot continue.

Clinical features alone usually do not allow discrimination between rare and common lung diseases. In fact, the diagnostic delay of a rare disease is mainly accounted for by the fact that in early stages symptoms may be absent, masked, misunderstood or confused with other diseases [25]. This implies that the goal of primary care should be the early recognition that a rare lung disease might be present and the determination of an appropriate threshold for referral to centres with specific expertise. The need for more global, accessible educational tools is clear. “You only see what you look for and you only look for what you know”.

From a research perspective, these delays present barriers to recruitment while from a clinical perspective they contribute to patient morbidity. For instance, most patients suffer episodes of pneumothorax before being diagnosed with LAM. In addition, misdiagnosis not only leads to inappropriate – and ineffective – treatments but also to unnecessary risks (pregnancy or air travel in the case of LAM). Similarly, in patients with Hermansky-Pudlak syndrome, invasive procedures should be avoided if at all possible, because of the patient’s tendency to bleed. In this latter case, a correct diagnosis allows other family members to be screened for the syndrome, with the demonstration of absent dense bodies on whole mount electron microscopy of platelets being diagnostic. Diseases caused by a single, mutated gene – such as alpha-1 antitrypsin deficiency, surfactant protein disorders and cystic fibrosis – lend themselves to family screening. This is essential as diseases diagnosed at an early stage are more likely to be properly treated and with the knowledge that a rare disease is present in the family, other individuals are less likley to fall victim to an erroneous diagnosis.