Study objectives To present diagnosis and treatment modalities of children with interstitial lung disease associated with frequent or rare surfactant protein C gene (SFTPC) mutation.
Patients Twenty-two children with chronic lung disease associated with SFTPC mutation in a heterozygous form.
Results Mutations located in the BRICHOS domain (‘BRICHOS domain’ group) were identified in six children, whereas 16 children carried mutations located outside the BRICHOS domain (‘non-BRICHOS domain’ group). The median age of onset was 3 (0–24) months. Four patients had neonatal respiratory distress, and symptom onset was associated with acute bronchiolitis in nine patients. Cough, tachypnoea and failure to thrive were initially noticed in all the children. Physical examination at presentation revealed tachypnoea (n=22), clubbing (n=1) and crackles (n=5). Low oxygen saturation (<95%) was observed in 18 patients. The predominant findings on initial high-resolution CT (HRCT) scans were basal-predominant ground-glass opacity (n=21) and cystic spaces (n=3). Bronchoalveolar lavage fluid (BALF) cell counts showed 379±56×103 cells/ml with increased neutrophil percentage (18±4%) independent of the mutation status. The median follow-up was 3.2 (1–18.3) years. Eighteen patients were treated by monthly methylprednisolone pulses associated with oral prednisolone (n=16), hydroxychloroquine (n=11) and/or azithromycin (n=4). Fifteen patients benefited from enteral nutrition.
Conclusion Initial diagnosis is based on clinical presentation, radiological features and BALF analysis, but the definitive diagnosis requires genetic analysis. Although progressive improvement was seen in most patients, the development of new therapeutic strategies with minimal side effects is needed.
Statistics from Altmetric.com
Pulmonary surfactant is a multimolecular complex located at the air–liquid interface. It is composed of bimolecular layers of lipid mixtures, mainly of phospholipids, and 2–3% of specific proteins called surfactant proteins A, B, C and D (SP-A, SP-B, SP-C and SP-D, respectively). SP-B and SP-C are two hydrophobic proteins. They are synthesised by alveolar type II cells in a propeptide form (pro-SP-B and pro-SP-C)1 and become mature proteins after an extensive post-translational processing, and finally are secreted in the alveolar lumen within the lipid bilayer.2 SP-B deficiency has been known to be responsible for lethal neonatal respiratory distress,3 and Nogee et al first described a surfactant protein C gene (SFTPC) mutation in 2001.4 This mutation was observed in one allele of a patient suffering from chronic interstitial pneumonitis and was inherited following a dominant pattern. More recently, SFTPC mutations were reported not only in newborns with severe alveolar syndrome but also in children and even in adults with chronic interstitial lung disease (ILD).5,–,10 They always occur in one single allele. The mutation I73T (c.218 T>C) is the more prevalent mutation. Other mutations have been described, usually located in the C-terminus of pro-SP-C BRICHOS domain (F94-I197), a highly conserved domain. The BRICHOS domain is composed of approximately 100 amino acids in the proprotein and is found in several proteins associated with degenerative and proliferative diseases.11 It is thought to have a chaperone-like function and to play a critical role in pro-SP-C targeting and processing.12,–,14
What is already known on this topic
Mutations in the surfactant protein C gene (SFTPC) have been recently associated with the development of diffuse lung disease, particularly sporadic and familial interstitial lung disease.
What this study adds
Although neonatal presentation may be observed in some patients with mutations in the BRICHOS domain, surfactant protein C gene (SFTPC)-associated lung disease usually presents within the first months of life with cough, tachypnoea and failure to thrive. The association of these non-specific symptoms with bilateral ground glass on high-resolution tomography scanner indicates bronchoalveolar lavage and SFTPC screening before considering lung biopsy.
In Bullard's study of 195 children with chronic ILD of unknown aetiology, 24 (14.4%) were found to have an SFTPC mutation.8 However, recent publications have shown that, in adults, SFTPC mutations are not common in sporadic cases of ILD15 and highlighted the need to define pertinent criteria to guide genetic screening. Therefore, in this study, we aim to describe the clinical and radiological presentation as well as the histological and bronchoalveolar lavage fluid (BALF) features of SFTPC mutation-associated disorders. The endorsement of a centre of rare respiratory diseases by the French ministry of Health in 2006 led to the improvement in the diagnosis of this previously unknown disease and presented the opportunity to focus on the development of a new, efficient and well-tolerated treatment. In this context, we have reviewed the treatment modalities and outcome of these children.
Materials and methods
Twenty-two subjects in whom SFTPC mutations were identified in the molecular genetic division of the Armand Trousseau hospital, Paris, France (n=20)16 or by Larry Nogee in Johns Hopkins Hospital, Baltimore, USA (n=2) were enrolled in the study. The controls used for western blot analysis were 10 children who underwent BAL to explore isolated uveitis and did not have acute or chronic ILD. A control scan was obtained from an age-matched patient with congenital cystic adenomatoid malformation and without ILD. Four patients, members or relatives of a same family, have been reported previously.17
The following data were obtained from the medical records:
family history and clinical presentation (failure to thrive is defined as height or weight measurements falling 2 major percentile lines using standard growth charts of the Centre for Early Childhood Development and Education);18
chest radiography and high-resolution CT (HRCT) scans finding;
lung biopsy report;
therapeutic modalities: corticosteroids (methylprednisolone pulse and oral steroids), oxygen therapy, additional medication, enteral nutrition;
length of follow-up.
The study was granted a limited waiver of authorisation and was approved by our institutional review board.
Bronchoscopy and BALF analysis
Eighteen patients underwent fibreoptic bronchoscopy under sedation and topical anaesthesia or under general anaesthesia. Bronchoscopy under sedation was performed after premedication with atropine and midazolam, local anaesthesia with 2% lidocaine and administration of premixed 50% nitrous oxide and oxygen during the procedure. General anaesthesia was performed with sevofluorane associated in some cases with propofol. The bronchoscope was wedged in a segment or subsegment of the middle lobe or lingula, and BAL was performed with 5 ml/kg of 0.9% saline solution. Recovered BALF was strained through gauze, and an aliquot was returned from each segment. Aliquots were pooled and submitted for cytological, virological and bacteriological examination. The remaining BALF was centrifuged (500g×5 min) to sediment BAL cells. After determination of the protein concentration using the Bradford method with the BioRad Protein Assay Kit (BioRad, Hercules, California), the supernatant was aliquoted and stored at −20°C for later analysis. The cell pellet was washed in red blood cell lysis buffer (0.1 M, NH4Cl in 0.1 M Tris, pH 7.6, 0.1 mM EDTA) and resuspended in phosphate-buffered saline for qualitative analysis by haemocytometry.
Surfactant protein analysis in BALF
BALF samples were concentrated by lyophilisation. After lyophilisation, pellets were resuspended in sample buffer (without sodium dodecyl sulfate (SDS)). An equal amount of protein (30 μg) from each sample was fractionated by SDS–polyacrylamide gel electrophoresis (Interchim, Montluçon, France) on 16% Tris–Tricine gels. Proteins were further electrotransferred to a nitrocellulose membrane (Whatman, Dassel, Germany) and probed by immunoblotting using mature SP-C, pro-SP-C-mature, SP-B and pro SP-B antibodies (recombinant antihuman from rabbit, charge 1/5000; Seven Hills Bioreagents, Cincinatti, Ohio, USA for SP-C, proSP-C, mature SP-B, and recombinant antihuman from rabbit; Chemicon, Temecula, California, USA for proSP-B). As second antibody, we used horseradish peroxidase conjugate goat Ig-G antirabbit from donkey (GE Healthcare, Slough, UK). Bound antibodies were detected using SuperSignal West Dura Extended Duration substrate (Pierce, Rockford, Illinois) according to the manufacturer's instructions and exposed to Bio-Rad Molecular Imager ChemiDoc XRS System. The images were analysed using Quantity One software (BioRad, Hercules, California) and compared with 10 controls from children without acute or chronic ILD.
The Mann–Whitney U test (SAS software, version 9.1; SAS Institute, Cary, North Carolina) was applied to compare the two groups of patients. Data are expressed as mean±SEM. A p value equal to or less than 0.05 was considered significant.
Among the 22 patients with an SFTPC mutation, 15 had the more common mutation I73T (p.Ile73Thr), one had a mutation located in the mature peptide (c.116T>C (V39A)), and six carried a mutation in the BRICHOS domain, of proSP-C (c.325-1G>A, c.424delC, c.435G>C (Q145H), c.563T>C (L188P), c.566G>A (C189Y), and c.581T>C (L194P)).16 Thus, patients were separated into two groups, ‘BRICHOS domain’ and ‘non-BRICHOS domain,’ regarding the location of SFTPC mutations. The SFTPC mutation was familial in 15 patients and de novo in seven patients (five in the ‘non-BRICHOS’ domain group and two in the BRICHOS domain group).
Characteristics at presentation
The clinical characteristics of the patients are summarised in table 1. Symptoms usually appeared within the first month of life, but the mean age of onset was significantly higher in the ‘non-BRICHOS domain’ group compared with that of the ‘BRICHOS domain group.’ Presentation with neonatal respiratory distress was seen only in children with non-frequent mutations. Three children from the ‘non-BRICHOS domain’ group had symptom onset after 18 months. Patients presented in 41% of the cases with acute bronchiolitis, and respiratory syncitial virus (RSV) was the most commonly isolated agent in 55% of children independently of the mutation status. All children presented with cough, tachypnoea, failure to thrive and a history of gastro-oesophageal reflux disease (GERD) was noticed in eight patients. Low-oxygen saturation, meaning that oxygen saturation is <95% measured transcutaneously by pulse oximetry, was exhibited in 82% of the children. Physical examination revealed seldom crackles in the basal areas of the lungs independent of the mutation status. Chest radiographs and HRCT scan data were available in all cases (table 2). Abnormalities at the time of initial evaluation were predominantly diffuse ground-glass pattern predominant in bases (figure 1). Cystic spaces and pneumothorax were initially only observed in ‘non-BRICHOS domain’ patients.
BALF cell count was performed in 18 children, and showed a normal cell count with increased neutrophil percentage independent of mutation status (table 2). Western blot analysis of BALFs using mature SP-C, pro-SP-C, SP-B and pro-SP-B antibodies was performed in 10 cases and is summarised in table 3. Mature SP-C was detectable in all the children with SFTPC mutation but with a very weak band pattern in all the cases. An accumulation of 26 kDa pro-SP-C form was observed in 90% of the children with SFTPC mutation versus 0% in the control. All children from the ‘non-BRICHOS domain group’ had an additional 16 kDa band. The presence of SP-B was exhibited in all the patients, but interestingly, none of the 10 patients with SFTPC mutations displayed a normal profile of pro-SP-B (figure 2).
Histological analysis was performed in 13 patients (table 4). All the patients regardless of their mutation status shared common characteristics with alveolar type II cell hyperplasia and septal thickening. Inflammatory cells infiltrate within the interstitium in most of the cases in association with intra-alveolar macrophages accumulation. Mild fibrosis was noticed in the biopsies of children older than 6 months.
Treatment and outcome
Treatments and outcome are summarised in table 5. Twenty patients required supplemental oxygen at the time of diagnosis, and among them, 19 required long-term O2 therapy (>3 months). Along with the preventive measures, systemic corticosteroids with methylprednisolone pulse were the preferred therapy for 18 patients. As indicated in table 3, oral steroid treatment was administered alone or more frequently with methylprednisolone pulse. Adrenal insufficiency prophylaxis, proton-pump inhibitors, Pneumocystis carinii prophylaxis (cotrimoxazole) and calcium were provided for all the patients under steroids. Hydroxychloroquine and azithromycin were administered in association with steroids in 45% and 18% of the cases, respectively. In one patient, mycophenolate mofetil was used in association with methylprednisolone pulses. In the addition of medical treatments, 16 patients received enteral nutrition with nasogastric tube (n=4) or gastrostomy (n=12). RSV prophylaxis by paluvizumab was performed in nine children. A short tetracosactrin test (patient treated with steroids), blood cell count analysis and electroretinogram analysis (patients treated with hydroxychloroquine) were performed on a regular basis. The only side effect observed from the different treatments was temporary adrenal insufficiency in children under steroids.
The median duration of follow-up in the total population was 3.2 (1–18.3) years. During this time, 16 patients were weaned off oxygen after a mean duration of 10.9±5.9 months, whereas four children were still on oxygen after a mean of 32±14.6 months. These four patients (three from the ‘non-BRICHOS domain’ group) were treated by methylprednisolone associated with oral steroids (n=4) and azithromycin (n=3), and none of them received hydroxychloroquine. At the end of the study, only four children (all from the ‘BRICHOS domain’ group) were totally asymptomatic among the 16 who were weaned off oxygen. The other 12 showed progressive clinical improvement but remained symptomatic with moderate dyspnoea and exercise intolerance.
In this study, we have reviewed the presentation of 22 patients with SFTPC mutations. We have shown that clinical and radiological features were not very specific and independent of the mutation status. In contrast, we have shown a decrease in mature SP-C, an accumulation of 26 kDa proSP-C form in the western blot analysis of BALFs and a specific proSP-C expression protein profile in the ‘non-BRICHOS domain’ patients characterised by the presence of 16 kDa proSP-C form.
Unlike SP-B deficiency, respiratory symptoms associated with SFTPC mutations usually begin within the first months of age. In the present study, acute bronchiolitis was noticed at presentation in almost half of the cases and coincides well with a triggering role of viral infection. In vitro and in vivo studies have suggested that viral infection such as RSV may act as a trigger of the disease.19 Indeed, it has been shown that SP-C knockout mice were more susceptible to RSV infections,18 20 and Bridges et al have shown that alveolar cell lines stably transfected with mutated SFTPC result in increased susceptibility to viral-induced cell death.21 Cough, tachypnoea and failure to thrive were constant findings independent of mutation status. Failure to thrive was observed in all of our patients and was associated with abnormal digestive motricity and GERD. Interestingly, digestive manifestations often precede respiratory symptoms and may, at least in the beginning, be foremost. These gastrointestinal symptoms are occasionally mentioned in the literature.22 23 They are attributed to the respiratory illness and may potentiate respiratory symptoms. Cyanosis, retraction and tachypnoea were constant findings, and physical examination was usually normal with only occasional crackles. Radiological findings in our patients were similar to those of previously published cases with a consistent ground-glass pattern,19 22 24 and cysts are rarely observed in the first HRCT scan.
As previously reported in children younger than 2 years with ILD, we have shown in this study that the BALF cell count of children with SFTPC mutations revealed a moderate increase in total cell number and increased percentage of neutrophils. The presence of SP-C precursors in BALF of almost all of the children in whom BALF analysis was available is consistent with an aberrantly processing of SP-C but may also be explained by alveolar damage resulting from altered surfactant metabolism. Indeed, the presence of SP-C precursors was also observed in other surfactant function disorders such as SP-B or ABCA3 deficiency.25 26 While the presence of SP-C precursors is certainly not specific to SFTPC mutation-associated diseases, a 16 kDa pro-SP-C form was found in all the patients from the ‘non-BRICHOS domain’ group and coincides with disruption of the SP-C metabolism. Interestingly, although mature SP-B was observed in the BALF of all the children with SFTPC mutation, the alveolar profile of pro-SP-B was abnormal in 80% of these patients. In murine models, it has been shown that SP-C metabolism is totally dependent on SP-B processing.27 28 Indeed, all the individuals having an SP-B deficiency also have abnormal SP-C synthesis.29 Our results indicate that the presence of abnormal pro-SP-C may also interfere with SP-B homeostasis.
Lung histological analysis showed similar findings previously published in SFTPC mutation-associated disorders with thickened septa, alveolar type II cells hyperplasia and accumulation of intra-alveolar macrophages. All these features are not specific to SFTPC mutation and have also been described in ABCA3 deficiency or idiopathic ILD.26 30 31
A diagnosis algorithm is proposed on figure 3. With this report as our basis, we would strongly encourage HRCT scan in children with abnormal chest x-ray, persistent cough, tachypnoea and/or hypoxaemia associated with failure to thrive and gastrointestinal symptoms. When HRCT scan shows a bilateral ground-glass opacity pattern, BALF analysis and genetic screening for I73T common mutation are recommended. In addition to cytological and infectious information, mature SP-C analysis in BALF, which is consistently decreased in lung disorders associated with SFTPC mutation, may support SFTPC sequencing. In the absence of SFTPC mutation, a lung biopsy will be considered.
The effects of SFTPC mutations on surfactant homeostasis have not been clearly elucidated. From now, the only data available result from in vitro cell models of SFTPC deletion located in the ‘BRICHOS domain,’ and there are no data on the more common mutation I73T. It has been demonstrated that accumulation of misfolded proSP-C in the endoplasmic reticulum causes disruption of the ubiquitin/proteasome system,13 activates apoptosis and interferes with non-mutated protein synthesis. It results in the release of pro-inflammatory cytokines and enhances T cells and fibroblasts recruitment.32 These data support the use of anti-inflammatory drugs, the most recognised treatment of lung disorders associated with SFTPC mutations.30 33 However, the pharmacological mechanisms of these drugs in this context are still unknown, and it remains difficult, from the present data, to determine the optimal treatment of this disease. Corticosteroid was used in 20 of the 22 patients, and methylprednisolone pulse was the preferred form of steroids (19/22). Oral prednisolone is initially added and progressively tapered. Although a short-term benefit of methylprednisolone pulse was consistently noticed in our patients, extreme caution must be given in attributing the observed long-term improvement to steroid treatment. Moreover, it must be balanced with the known toxicity of steroids. The use of methylprednisolone pulse, although its mechanisms of action are not yet elucidated, may have helped to reduce steroid side effects that were not observed in our patients.
In the absence of studies, the choice of azithromycin and/or hydroxychloroquine remains highly dependent on the habits and experiences of the different centres. The use of azithromycin was based on its anti-inflammatory effects that have been demonstrated in panbronchiolitis34 and cystic fibrosis.35 Interestingly, macrolids are also known for their antiamyloid properties,36 and it has been demonstrated that the C-terminal domain (CTC) of the SP-C precursor prevents SP-C amyloid fibril formation.37 Thus, azithromycin treatment may limit the formation of amyloid fibril formation secondary to SFTPC mutation. Hydroxychloroquine treatment has already been reported in association with steroid7 or alone.38 In addition to having anti-inflammatory properties, hydroxychloroquine has been shown to cause inhibition of the intracellular processing of the precursor of SP-C,39 which may explain its therapeutic effects. Nutritional support was performed in almost all of the patients and certainly helped to restore height and weight development. In the present study, nine patients received RSV prophylaxis. The current report and previous in vitro studies showed that infections during the first months of life may also negatively influence the natural course of the disease.20 21 Thus, it might encourage clinicians to consider viral infection prophylaxis such as flu or RSV in the first 2 years. Although deaths and lung transplantation have been reported in other children with SFTPC mutation despite the use of similar therapies, all of our patients were alive at the end of the study. Moreover, we have shown that, unlike SP-B deficiency and regardless of the molecule used in association with corticosteroids as well as the modality of their administration, the patients improved progressively. Furthermore, in most of the cases, the children are weaned off oxygen. These findings emphasise the benefits of early diagnosis with respect to nutritional support, viral infection prophylaxis, genetic counselling and avoiding misdiagnosis. They also emphasise the need to develop new molecules and perform international randomised control studies that are more suited to the low prevalence of the disease. Physiopathological mechanism is probably different whether or not the mutation is in the BRICHOS domain. Thus, clinicians should distinguish between the two groups in future studies. These trials will provide the opportunity to achieve a strategic approach regarding the optimal combination and the timing of different treatments.
The authors thank all the members of the affected families and their referring clinicians for their collaborative participation in this study. The authors are grateful to L Nogee (Johns Hopkins University School of Medicine, Baltimore, Maryland) for his help in sequencing and helpful comments on manuscript. The authors also thank MC Miesch for technical assistance on surfactant protein western blot.
GT and RAT contributed equally to this work.
Competing interests None.
Patient consent Obtained.
Ethics approval Ethics approval was provided by the CPP Saint Antoine.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.