Clinical characteristics of patients with familial idiopathic pulmonary fibrosis (f-IPF)

The phenotyping data from IIP patients and their relatives were collected via the European IPF Registry (eurIPFreg) to evaluate the differences between f-IIP and s-IIP groups (with a focus on IPF as a subgroup). All patients were recruited at our UGMLC sites in Giessen and Greifenstein. The purpose was to identify all patients with familial background.

The eurIPFreg is as an Internet-based, multicenter registry interlinked with the European IPF Biobank (eurIPFbank, see also http://​www.​pulmonary-fibrosis.​net) [18]. The data protection concept was approved by local and national networks, such as the TMF (Technology, Methods, and Infrastructure for Networked Medical Research e.V.) and authorities (e.g. Hessian Data Protection Officer, Protocol Nr. 412101 from 25.08.2008). Both, eurIPFreg and eurIPFbank received positive votes from institutional review boards in Europe (e.g. Ethics Committee of Justus-Liebig-University of Giessen; 111/08). Patients were included into the registry starting November 2009 if they were at least 18 years old, had IPF (prevalent or incident cases) or other DPLDs (as comparator group), as diagnosed by the expert site, and had provided written informed consent prior to the inclusion.

The clinical data were collected at the time of enrolment (baseline) and in intervals 3 to 12 month thereafter via patient and physician baseline questionnaires, which can be retrieved upon logging in to our website (http://​www.​pulmonary-fibrosis.​net) [18].

All 28 index patients were diagnosed accordingly ATS/ERS/JRS/ALAT Guidelines 2011 [32]. The definition and evidence level of f-IIP have been established in accordance to the recent publications and guidelines [32‐36].

As first line relatives, the biological parents, children and full siblings were taken into account. The f-IPF index patients as well as their families were enrolled into the eurIPFreg. When employing wide range criteria of f-IIP incidence (evidence grade C), 60 index patients were identified. However, after refining the criteria to occurrence of either nonspecific interstitial pneumonia (NSIP) or IPF in one first grade relative (mother, father, sister, brother or child), the search narrowed down to 28 index IPF patients.

The following exclusion criteria were applied: patients under 18 years of age, missing informed consent, death of the patient, other lung diseases apart from IIP, adoption of the patient, as well as evidence level C of f-IPF.

These 28 index patients were prospectively interviewed, examined, and analyzed using the patient and physician questionnaires of the eurIPFreg, including establishment of pedigree charts. The patient questionnaire included patient’s demographics, a detailed medical history and complaints. The physician questionnaire contained data of physical examination and laboratory tests, pulmonary function, HRCT, echocardiography, 6 MWD, co-morbidities as well as other information concerning relevant patient’s diagnosis and therapy [18]. In addition, blood samples were obtained and archived in eurIPFbank.

Adult relatives of the f-IPF index patients were also recruited into the eurIPFreg, and were invited to our DPLD outpatient clinic in order to undergo examination as part of this study. On site, a detailed medical history, a physical examination, a blood collection, a whole body plethysmography, a DLCO test and a blood gas analysis were performed. In total, 52 healthy relatives were included, 26 of whom have carried out all the above-mentioned examinations. One of the 52 family members was diagnosed with an early form of IPF and was re-classified as an affected relative.

The pedigree charts were developed in collaboration with affected families during the clinical presentation of the patient either in our outpatient clinic or via telephone interview. All family members from all known generations were taken into account.

The following data were collected by each individual subject: family relationship, date of birth, date or at least age at death, cause of death, complaints and all known illnesses. In particular, we were looking for the presence of lung diseases. Individuals with known or suspected IIP were a subject to a more extensive survey: the explorative diagnostics with evaluation of the entity and stage of pulmonary fibrosis, age at first manifestation and diagnosis, smoking status, exposure to occupational and environmental risks. In 24 DPLD patients, included in the pedigree charts, only medical history data could be collected. The pedigree charts were generated with the program GenoPro, version 2.5.4.1.

The data of all s-IIP patients (incl. IPF) recruited at the same sites as the familial form were extracted from the eurIPFreg, applying inclusion and exclusion criteria mentioned above. The following comparison groups were formed: f-IPF vs. s-IPF (n = 147), f-IPF vs. s-NSIP (n = 17), f-IPF vs. unclassifiable s-IIP (n = 69), f-IPF vs. s-COP (n = 42).

The comparison of change over time in forced vital capacity (FVC) was performed between the f-IPF index patients and s-IPF patients. The main requirement for statistical analysis here was the presence of at least five FVC measurements. The required conditions were met by 21 index patients and 54 s-IPF patients. The last recorded FVC value was the last measurement before lost to follow-up, death of the patient, start on antifibrotics, or lung transplantation.

The statistical analyses were carried out in cooperation with the Department of Medical Statistics of the Justus Liebig University Giessen. For the data evaluation, we used software “R Statistics” (R version 3.1.3). The individual groups were independent of each other. The p-value for statistical significance was set as p ≤ 0.05. For distribution tests, we performed the Shapiro-Wilk test as well as the Anderson-Darling test. For normally distributed variables, an independent two-tailed t-test or Mann-Whitney test were conducted, both with data adjustment. For nominal scaled parameters, the Chi-square test and the Fischer test were performed.

Of the 28 identified index patients, all had the diagnosis IPF with following levels of diagnostic confidence: 24 definite, three probable, one possible. Together with the 51 affected family members (then total of 79 patients), these patients were diagnosed to have IPF (56%), nonspecific interstitial pneumonia (NSIP, 3%), unclassifiable IIP (u-IIP, 1%) and other DPLD (40%). The IPF diagnosis was done in accordance with an Official ATS/ERS/JRS/ALAT Clinical Practice Guideline 2011 [32]. In those cases in whom we were able to collect the clinical, radiological and pathological information and in whom the IPF diagnosis criteria of the official ATS/ERS/JRS/ALAT Clinical Practice Guideline were met, the diagnosis of IPF could be settled / confirmed. We did not receive the relevant information for all relatives of the index patients. In some cases, especially in case the relative died earlier, we were not able to retrieve reliable further information. In that case a diagnosis of DPLD was considered without any further differentiation. Evidence level A for f-IPF was met by 14 families. The remaining 14 families corresponded to the evidence level B. Compared to them, the s-IIP group (286 patients) showed following distribution of diagnoses: IPF - 51.2%, NSIP - 5.9%, uIIP - 24%, cryptogenic organizing pneumonia (COP) - 14.6%, other DPLD - 4%.

In total, 314 IIP patients (both, familial and sporadic forms) were recruited at our two UGMLC sites. The 28 f-IPF index patients represented 8.92% of all IIP cases. The confidence levels of IPF diagnosis were defined accordingly ATS/ERS/JRS/ALAT Guideline 2011 and are shown in Table 1 [32].

When relating the diagnosis of all familial cases to the sporadic comparator group, percentage values from 1.4% (u-IIP) up to 23% (IPF) were encountered (see Table 2).

All IPF index patients underwent high-resolution computed tomography (HRCT) for diagnosis prior to this study as a part of diagnostic routine. In all cases, the radiological findings were consistent with a UIP, of them definite UIP in 66.7% and possible UIP in 33.3% of the cases. In total, 14 index patients additionally underwent a lung biopsy for histopathological diagnosis; a video-assisted thoracoscopic surgery (VATS) was performed in 11 patients; in all of the 14 cases UIP pattern was seen.

Twenty-one index patients received a bronchoscopy with a bronchoalveolar lavage (BALF); alveolar macrophages were present in 74.8 ± 37.47%, neutrophilic granulocytes in 10 ± 31.81%, eosinophilic granulocytes in 6.54 ± 0.71% and lymphocytes in 8.2 ± 4.95% of the cases (mean values ± SD). In 25 cases autoantibodies (anti-nuclear antibodies = ANA, extractable nuclear antibodies = ENA, rheumatoid factor = RF, CCP, ANCA and basal membrane antibodies) were tested and not found to be abnormal; in general, prior to IPF diagnosis, connective tissue diseases-associated ILD were excluded.

Of the 147 sporadic IPF patients in 97% of the cases function tests were performed, 98% underwent HRCT, 76% performed 6 MWD, 33% of the diagnoses were additionally confirmed by histology. Bronchoscopy was performed in 73% of the patients, echocardiography in 63% of patients.

Both groups showed no difference in demographics (61 vs. 79% males) and smoking history. The f-IPF group differed by an earlier age at the onset of the disease (55.4 vs. 63.2 years; p < 0.001) and broader age range (36–72 years). With an average of 67.0 years, f-IPF index patients died earlier than sporadic cases (71.8 years), although the result was not statistically significant (p = 0.059). The results are shown in Table 3.

The leading onset symptom of f-IPF index patients was dyspnea (79%), similar to that in the s-IPF group. Interestingly, f-IPF patients reported more frequently on persistent, dry cough without expectoration, as compared to s-IPF group (58 vs. 11%, respectively; p < 0.001). During physical examination, “velcro-like” crackles were found in 96% of f-IPF patients, nail widening in 74% and finger clubbing in 13% of all cases. Oxygen supplementation (LTOT) hat 4.76% of the f-IPF patients with mean flow of 2.5 l, as compared to 22.13% of the s-IPF group (mean flow 3.01 l).

On average, f-IPF index patients showed a milder extent of functional impairment at the time point of diagnosis vs. the s-IPF group (VC p = 0.027, FVC p = 0.011, RV/TLC p = 0.026, DLCO p = 0.006, pO2 p = 0.015). The results are shown in Table 4.

The most common co-morbidity in the f-IPF group was pulmonary hypertension (26%); obstructive sleep apnea was seen in 7% of the cases, and pulmonary embolism in 7% of the patients. The percentage of patients ultimately treated with antifibrotic drugs was similar in both groups (p = 0.208). Lung transplantation was performed in 15% of f-IPF patients, as compared to 5% in s-IPF patients (p = 0.094). The average age of death in f-IPF group was 67 years, as compared to 71.8 years in s-IPF group (p = 0.059).

The decline in FVC was compared between the f-IPF index patients and the s-IPF patients. The statistic requirement for this analysis was the presence of at least five FVC-values, reflecting decline over time. FVC levels measured after therapy start with antifibrotics or after lung transplantation were excluded. These criteria were met by 21 index f-IPF patients and 54 s-IPF patients. The last recorded FVC value was the last measurement before lost to follow-up, death of the patient, start on antifibrotics or lung transplantation.

The s-IPF group (n = 54) lost 0.014% (95% CI: 0.009–0.019%) of the FVC on average per day (relative decline). Corresponding to this, the average decline in 6 months was 2.48% (95% CI: 1.568–3.382%), or 4.96% per year. The mean value of the starting point of the FVC in s-IPF group was 62% predicted. However, possibly due to the broad scattering of data, the difference in FVC decline between the groups did not reach statistical significance (p = 0.12). In a logarithmic analysis of FVC decline, the starting value in the f-IPF group was 4.30 (95% CI: 4.19–4.41). In contrast, the logarithmic starting point of FVC for sporadic IPF was 4.18 (95% CI: 4.11% - 4.25). The interpolated slope of the FVC for f-IPF was − 0.00028 (95% CI: − 0.00018 – − 0.00039), and for s-IPF group was − 0.00014 (95%-CI: − 0.00009 – − 0.00019). The data are shown in Fig. 1.

In general, lung cancer occurred in nine families, in one of the families in two members. Occurrence of multiple malignancies of different organs was found in two families. In six families, patients were rather young when receiving first diagnosis of lung cancer (32, 38, 39, 48, 50 and 50 years, respectively). In one family, it was noticed that all four affected IIP patients died within a very short time after diagnosis due to disease progression. Furthermore, it was observed in five pedigree charts that the younger generation became ill over a decade earlier than the generation of the parents (phenomenon of anticipation).

Within the families, altogether 79 cases of DPLD including the 28 index patients, as well as 51 further patients with DPLD were detected (44 IPF patients, two NSIP, one unclassifiable IIP, and 32 other DPLD). In fourteen families, two members were affected; another seven families had three affected members per family, six families had four DPLD cases, and one family had six relatives with DPLD.

Pedigree charts could be completed for 25 of the 28 f-IPF patients. In four pedigree charts, a father to son inheritance was observed. In two pedigree charts, the disease appeared to originate from both parental sites. In contrast, in 16 families (ten times inheritance mother to child and six times inheritance father to child), the disease only seemed to originate from one parent. Of these, in one family, children of the same mother (but different father) were affected. Only one generation was affected in seven families (three times exclusively brothers, four times mixed genders).

In the following, four pedigree charts are presented as example. The index patient is marked with an arrow. All members suffering from DPLD are marked yellow with a blue border. Patients suffering from another lung disease are marked with a blue border. Family members without DPLD are marked with a magenta border. The relationships in pedigree charts always refer to the index patient.

This family tree comprised four generations with 13 individuals. The grade of evidence of f-IPF was level A, because the index patient (III: 3) had at least one immediate relative diagnosed with IPF or NSIP. In this family, the brother (III: 5) fulfilled this condition. The disease seemed to have its origin on the maternal side. In the mother (II: 3) no diagnosed pulmonary fibrosis could be detected. However, shortly before her death she began suffering from progressive dyspnea and cyanosis without history of nicotine consumption. The aunt of the index patient suffered from pulmonary fibrosis. In the third generation, all three members were affected by an IPF. Inheritance pattern here seems to be autosomal dominant, since all children were affected, despite the fact that they have different biological fathers. The data are shown in Fig. 2.

This family included six generations with 39 members. The index patient is in position III: 4. There was an evidence level A (index patient III: 4 and sister III: 5). The cousin was also affected by diffuse parenchymal lung disease (III: 7). He was accidentally diagnosed with lung fibrosis because of auscultation of velcro-crackles in the routine checkup. His daughter (IV: 2) also did not have any complaints at the time of first diagnosis of pulmonary fibrosis. In total, ten other blood relatives of the index patient suffered from some form of lung disease. In particular, childhood pneumonia was very common (IV: 8, V: 3, V: 4). Since only one generation was affected with IPF, inheritance mode is challenging to interpret in this example. The results are shown in Fig. 3.

The pedigree chart 3 shown here comprised six generations with 42 family members and fulfilled the evidence level B (IV: six and III: 4). Four members of the family were known to be affected by pulmonary fibrosis (III: 4, III: 7, IV: 6, IV: 10). The index patient (IV: 6) had died of a rapidly progressive lung fibrosis with subsequent respiratory failure at the age of 71 years. More family members had died due to pulmonary fibrosis: his father (III: four at the age of 64), his uncle (III: seven at the age of 64) and a cousin (IV: 10 at the age of 71). The index patient had no siblings. The aunt (III: 5) had died at the age of 80 and had no pulmonary fibrosis during her lifetime. Other children of the DPLD-affected uncle (IV: 8, IV: 12) died early at the age of 43 and 49, respectively, without having developed (or diagnosed) pulmonary fibrosis during their lifetime. The most likely inheritance seems to be an autosomal dominant inheritance, as both father and uncle seem to pass on the gene Fig. 4.

The pedigree chart four observed five generations consisting of 53 individuals. Evidence level B was met by the index patient (III: 5) and her father (II: 6). Three generations were affected by pulmonary fibrosis (I, II and III). The sister (III: 4), the daughter (IV: 9) and two nieces (IV 14 and 15) of the index patient had shown no evidence of diffuse parenchymal lung disease at the time of the study. In the oldest generation (I), the grandmother (I: 4) died at the age of 86. In the following generation (II), the father and the uncle died earlier than the grandmother. The father was diagnosed with IPF at the age of 79 years and so the uncle (II: 7) was diagnosed with IPF with 67 years. At age 53, the index patient (III: 5) was diagnosed much earlier than relatives in the previous generations. Due to the rapid progression of the disease, the index patient received a lung transplant at age 64. Up to now, no cases of pulmonary fibrosis have been observed in generations IV and V. This family tree demonstrates the phenomenon of anticipation. The younger the generation, the sooner the disease was diagnosed, but also the sooner the patients died or become lung transplant Fig. 5.

The youngest generation was defined as the youngest family member affected by DPLD. Upstream of this generation was the parents’ generation, the grandparents’ generation and so on. The youngest generation in our cohort affected 49 members (40%). In this generation, the age at first diagnosis was 58 ± 11 years (mean ± SD values), which was significantly lower as compared to the year of first diagnosis in the parents’ generation (66 ± 13 years, p = 0.013). The affected family members of youngest generation died on average at the age of 66 years. In the parental generation, 25 persons were diagnosed with DPLD (38%), and they died on average at the age of 71 years. The grandparent generation was first diagnosed at the age of 72 and died on average at age 79 (see Table 5).