The diagnosis of hypersensitivity pneumonitis (HP) is challenging depending on the degree of clinical suspicion, based on the exposure to antigens known to induce this inflammatory reaction. The appearance of a characteristic pattern observed in high resolution computed tomography (HRCT) images can raise the probability of diagnosis, and a predominance of lymphocytes in bronchoalveolar lavage (BAL) can also suggest the presence of HP.1,2 While the presence of specific immunoglobulin G antibodies in the serum (ssIgG) of patients indicates exposure to this antigen,3 a positive specific inhalation challenge (SIC) with the suspected antigen is also of diagnostic value especially to establish the etiology and confirm the diagnosis.2,4–8 In fact, the sensitivity and specificity of the SIC for the diagnosis of HP by exposure to avian and/or fungal antigens have been reported by our group to be 85% and 86% respectively,9 figures very similar to those obtained by other autors.4,8 Finally, evidence of improvement and avoidance of further exposure favors the diagnosis of HP.10

The temporal relationship between overt and/or subtle/occult presence of environmental factors and symptoms may raise the suspicion of the development of HP.11,12 Several recent studies have reported that a lack of awareness of infrequent environmental factors may explain, in part, the apparent lack of an attributable antigen in 25%–60% of patients with HP.10,13,14 Although exposure to birds is a well-known causative factor for HP, to date there are no prospective studies on HP due to contact with occult avian or fungal antigens in feathers within the home, for example in feather duvets and pillows. There have been sporadic reports of HP induced by feather bedding,2,15–17 but there are no established diagnostic criteria and the prevalence of this entity remains unknown. In addition, a recent study found that at least 43% of patients diagnosed with idiopathic pulmonary fibrosis (IPF) actually had chronic HP, and that half of the cases were due to occult exposure to feather bedding.6

The main objective of this study was to raise awareness of the diagnosis of HP due to exposure to concealed feathers at home. Diagnostic criteria are proposed. To our knowledge, this is the first report of an epidemiologic study combined with a description of the long-term outcome of a series of well-defined patients with HP caused by feather bedding.

Among the 127 patients with HP prospectively diagnosed at our center between 2004 and 2013, the condition was attributed to the use of feather bedding in 33 (26%). The condition was acute in 11 (33%) patients and chronic in 22 (67%). The clinical, demographic and diagnostic test results are shown in Table 3. There were no significant differences in any of the variables between acute and chronic patients, with the exception of age (acute patients being younger, P<.04). In 10 out of 22 (45%) chronic HP patients with a HRCT pattern of UIP, IPF was initially suspected.6

Pathological Findings

Figs. 1 and 2 show examples of pathological findings. This study used a temporal classification of acute and chronic HP.9

Fig. 1.

A (case 19): Acute: peribronchiolar chronic inflammatory infiltrate with an interstitial giant cell (arrow); B (case 9): acute: interstitial chronic inflammatory infiltrate with a poorly formed granuloma (arrow) and peribronchiolar metaplasia (arrow-head); C (case 9): chronic HP: the same case, four years later: explant specimen showing airway-centered interstitial fibrosis with an interstitial giant cell (arrow); D (case 17): chronic HP: advanced fibrosis with microscopic honeycombing.

(0.27MB).

Fig. 2.

A: Chronic HP showing granulomas (arrows) and extensive peribronchiolar metaplasia (arrow-head); B: another area showing peribronchiolar metaplasia along with fibrosis.

(0.18MB).

Follow-up

The median period of follow-up since symptom onset was 7.1 (4.5–9.5) years, and median follow-up period from the first visit was 4.4 (2.0–6.0) years. During follow-up, four patients died and six underwent lung transplantation (Table 3) and Fig. 3. The cause of death was respiratory failure in two patients (one with tuberculosis), one myocardial infarction, and one of unknown cause. The mean (SD) survival from symptom onset was 12.6 (5.8) years, with cumulative survival rates of 89.5% (95% CI: 77.8–100.0) at five years and 71% (95% CI: 52.3–89.5) at 10 years (Fig. 3). Survival analysis stratified according to the disease stage at diagnosis showed a difference in 10-year survival rate of 36% (95% CI: 8.7%–63.0%) between groups (100% in patients with acute HP vs. 64% in the chronic stage). These results remained the same after adjustment to the age of 60. Fig. 4 shows the expected and observed survival curves according to disease stage at diagnosis. Survival analysis stratified according to whether HRCT findings met the criteria for UIP22 showed no significant differences in the 10-year survival between groups, either in the entire study sample (16.3%; 95% CI: 0.0%–64.4%), or in patients with chronic HP without criteria for UIP (16.8%; 95% CI: 0.0%–74.2%) (Fig. 5).

Fig. 3.

Kaplan–Meier survival curve of the 33 patients diagnosed with HP.

(0.06MB).

Fig. 4.

Kaplan–Meier survival curves according to disease stage at diagnosis: acute vs. chronic in this series of 33 HP patients*.

(0.07MB).

Fig. 5.

Kaplan–Meier survival curves according to the presence of HRCT criteria for UIP versus non-UIP in the 22 patients with chronic HP*.

(0.06MB).

This is the first prospective study, and the largest study to date, of patients with HP attributable to exposure to concealed feathers at home, diagnosed over a 10-year period. Chronic HP was the most common form of presentation (67%). We stress that 10 (45%) patients with this chronic HP presented clinical features of UIP, and IPF was initially suspected. These results correlate with those of our previous study undertaken to re-evaluate patients previously diagnosed with IPF, which showed that 43% of patients with IPF were in fact CHP,6 and that 10/16 (63%) of these IPF patients exposed to feather bedding had CHP whereas “only” 10/30 (33%) of IPF patients not exposed had CHP.6 Both studies reinforce the need for a prompt detailed history of environmental exposure in IPF patients, which should include questions regarding exposure to feather bedding.22

The etiology of HP in the various series reported is unknown in 25%–60% of cases.10,13,14 In-depth assessment of the clinical history and the use of tests such as those used in the present study are needed to determine the etiology. Until recently, in routine clinical practice concealed feathers at home were not usually considered as a cause of HP, and so in some cases in which the etiology of HP remains unknown the disease may have been secondary to this exposure. Indeed, in the present series of 127 patients diagnosed with HP, feather exposure was identified as the cause in 26% of patients, a rate that is not so different from those reported in series without an etiological diagnosis.10,13,14

The present calculated prevalence of HP due to feather duvets and pillows is 6.2 out of 100 000 feather bedding users over a 10-year period. In fact, this prevalence is much lower than that of BBD (54.6/100 000, or roughly 5/10 000 bird breeders), although these prevalence rates differ less when considering the absolute numbers (1.9/100 000 inhabitants for BBD vs. 4.9/100 000 inhabitants for feather bedding HP).

Authors’ opinions differ as to whether the diagnostic performance of ssIgG detection and SIC have been validated. According to data from our center,3 the sensitivity and specificity of ssIgG determination by ELISA in the diagnosis of BBD are 84.2% [95% CI: 62.4–94.5] and 22% [95% CI: 12%–36.7%], respectively.3 In relation to the SIC, a recent review reported a sensitivity of 85% and specificity of 86.2% for diagnosis of HP to avian antigens and fungus,7 figures similar to those obtained by other authors.4,8,23 Thus, a positive test result may be taken to indicate a high probable diagnosis of HP.5 Nevertheless, patients with chronic fibrosis tend to have lower inflammatory cell infiltrates, indicating perhaps that the test may have less sensitivity in this patient subgroup.5 Further studies are need to determine the diagnostic yield of SIC in this fibrotic HP group.

The survival of patients with acute HP over the 10-year study period was 100% compared to only 64% in patients with chronic HP. Adjusted analyses and expected survival curves for the general population indicate that these results are unlikely to be due to the differences in age between groups. This deterioration in patients with chronic HP was observed despite the fact that they avoided further contact with concealed feathers and received treatment with oral steroids. This poor prognosis of chronic HP in relation to the overall follow-up of our series of CHP.24 and other series referred to in Appendix in Table 4, suggests that physicians should consider exposure to concealed feathers in order to allow early identification of this etiology. On the other hand, there were no differences in the survival of patients with chronic HP with a HRCT with UIP pattern and those without a UIP pattern, although the small number of patients studied may have influenced these results.

In the lung biopsy some patients clinically classified as having “acute” HP presented some degree of fibrosis, considered a pathological feature of chronic HP.9 This discordance is likely due to the arbitrary definition of acute and chronic HP, based on symptom duration. In biological terms, the apparent discordance between the clinical acute form and some pathologic characteristics of CHP is due to the fact that fibrosis is not a histologic lesion of sudden onset; rather, the histological features of inflammation and fibrosis are a continuum of lesions. The presence of clubbing in nearly half of the patients (45%) with acute HP, the presence of velcro crackles in 36%, and the observation of peripheral microcysts in 27% of these acute cases in the HRCT should be interpreted in the same way. In fact, in the initial historical description of the pathology of HP in Farmer's lung, Barrowcliff and Arblaster stated that: “Thus fibrosis, a marked feature of the chronic disease and always present to some degree in those cases described as being acute….”.25

The manufacturing process of European feather bedding companies is rigorous, with thorough washing and drying of the feathers used for comforters, blankets, pillows and duvets (personal communication with three main Spanish manufacturers). In Spain, feathers are usually purchased from goose slaughterhouses where foie gras is made, although some companies import feathers from other countries. Further studies are needed to elucidate whether the impact of feather bedding on HP development varies according to its origin.

Our study has several limitations: (1) it is a single-center study performed in one area of the city; (2) the focus of the evaluation was occult avian antigens in feather bedding from different commercial sources; (3) the possible contamination from environmental fungi during the recollection of the feathers by patients cannot be absolutely ruled out; (4) some patients were confirmed by a positive SIC result; although the SIC has been validated as a diagnostic procedure in some studies4,7,8,23 and has been routinely performed as a diagnostic tool at our center for four decades,26 it is still to be accepted as a validated test by some experts in ILD.

Although it is unnecessary to establish the diagnosis in all patients with suspected HP by histopathology,22,27 in the present series the diagnosis was in fact confirmed by characteristic histopathological features of HP in over 50% of patients (17/33). The need for evidence-based or consensus-based clinical practice guidelines for diagnosis of HP is clear, and indeed is long overdue.

In conclusion, in this prospective series of ILD patients, HP was attributed to occult avian or fungal, antigens present in feather duvets or pillows in 26% of the patients diagnosed with HP. The prognosis of chronic HP following this exposure is poor, with a 10-year survival rate after symptom onset of 64%. The diagnostic protocol used in this prospective study may help to identify occult environmental factors such as the etiology of disease in patients presenting with ILD/interstitial pneumonia of unknown origin, including those patients presenting with a UIP pattern, and may help to establish the diagnosis of chronic HP in patients initially diagnosed with IPF.

The present study has been given support by: Grant FIS PI13/01377 from the Instituto de Salud Carlos III cofinanced by the European Regional Development FUND (FEDER), Grant of Cellex Fundation and Grant of Sociedad Española de Neumología (SEPAR) 2013. The sponsors of the study played no role in the study's design, conduct, and reporting.

The authors declare no conflict of interest.