Penetrance of interstitial lung disease and lung cancer in carriers of SFTPA1 or SFTPA2 pathogenic variants

Extract

Interstitial lung diseases (ILDs) are a diverse group of rare lung disorders that affect the distal lung tissue, often resulting in inflammation, fibrosis, and in severe cases, lung failure. About 20% of ILD cases are linked to genetic mutations, primarily in telomerase- and surfactant-related genes. Specifically, heterozygous germline pathogenic or likely pathogenic variants (PGVs) in SFTPA1 and SFTPA2, which encode surfactant proteins SP-A1 and SP-A2, can result in a range of outcomes from asymptomatic cases to severe lung fibrosis and adenocarcinoma, with high mortality risks [1–5].

Shareable abstract

Variants of SFTPA1 and SFTPA2 are involved in interstitial lung disease and lung cancer. Penetrance for the first event is 50% at 60 years old, rising to 89.3% at 80. Penetrance for lung cancer reached 50% at age 80, earliest case diagnosed at age 30.

To the editors:

Interstitial lung diseases (ILDs) are a diverse group of rare lung disorders that affect the distal lung tissue, often resulting in inflammation, fibrosis, and in severe cases, lung failure. About 20% of ILD cases are linked to genetic mutations, primarily in telomerase- and surfactant-related genes. Specifically, heterozygous germline pathogenic or likely pathogenic variants (PGVs) in SFTPA1 and SFTPA2, which encode surfactant proteins SP-A1 and SP-A2, can result in a range of outcomes from asymptomatic cases to severe lung fibrosis and adenocarcinoma, with high mortality risks [1–5]. Variants are all missense variants which do not affect the expression of the protein but interfere with its secretion. The mechanism of disease probably involves loss of secretion and intracellular accumulation of the proteins. The variable penetrance may be related to the nature of the variants, other genetic factors and environmental exposure such as smoking. The study focuses on estimating the penetrance of ILD and lung cancer in carriers of these PGVs in order to guide risk assessment, genetic counselling and medical monitoring.

This study was part of the French RespiFIL network for rare lung diseases and included families with SFTPA1 or SFTPA2 missense PGVs identified at Trousseau hospital. The following PGVs were included: Asn171Lys, Val178Met, Tyr186His, Tyr186Ser, Gly203Arg, Trp211Arg, Val225Met and Gly231Glu in SFTPA1; and Asn171Ile, Val178Met, Tyr181Cys, Tyr228Cys, Trp233Arg, Trp233Leu, Trp233Cys, Cys238Ser and Arg242Gln in SFTPA2. The pathogenicity of most PGVs was confirmed through in vitro functional studies [3]. The data collected included age at diagnosis, genotype and age at last follow-up.

The study models the penetrance associated with SFTPA1 and SFTPA2 PGVs by estimating survival functions, focusing on the probability of disease onset over time. The method was developed to estimate penetrance from pedigree data [6] and it accounts for familial structures while assuming autosomal dominant inheritance (allele frequency f=0.0005). The two genes were considered as a single locus due to their proximity (50 kb) and shared phenotypes. Proband's phenotype exclusion likelihood [7] was used to adjust for ascertainment bias. The model is fitted with an R implementation of the expectation–maximization algorithm. It alternates between the E-step, which updates carrier probabilities using a sum–product algorithm, and the M-step, which refines survival estimates using a weighted Kaplan–Meier estimator. Sensitivity analysis and male/female stratification have been conducted but are not presented in this letter.

This study included 27 families with a total number of 744 individuals (of which 328 have a known clinical status: disease-free, lung cancer, ILD or both). 64 individuals had a history of ILD, 23 of lung cancer and 20 of both (ILD occurred before lung cancer in 3 of 20 cases, in the other 17 cases ILD and lung cancer were diagnosed simultaneously). 221 individuals were asymptomatic, while the clinical status was unknown in 416. The median age of living individuals in the study was 42 years, with disease onset occurring at a median age of 49 years. SFTPA1 and SFTPA2 PGVs were found in 10 and 17 index cases respectively. As for individuals (index cases and relatives combined), 22 carried a SFTPA1 PGV and 37 carried a SFTPA2 PGV; all PGV carriers are among the 328 individuals with known clinical status. No significant differences in various characteristics (male/female ratio, phenotypes, number of deaths or median age at death, study time and disease onset) were found between the SFTPA1 and SFTPA2 subgroups, supporting their treatment as a single locus. Penetrance estimates to the first event (ILD or lung cancer), ILD, and lung cancer are presented in figure 1. At age 30, the penetrance to the first event was 7%. The study highlights a high penetrance of 89.3% at age 80 for the development of either ILD or lung cancer in carriers of these PGVs. At 80, the penetrance for ILD was 90.7%, and for lung cancer it was 48.4%.

FIGURE 1.

Penetrance and 95% confidence interval for ILD, lung cancer and to the first event.

This study presents the first estimates of penetrance for ILD and lung cancer in carriers of SFTPA1/SFTPA2 PGVs, providing valuable insights into the disease risk for this genetic condition. The study was based on a large cohort of affected individuals and their relatives, showing that the risk of developing ILD or lung cancer increases with age, reaching a high penetrance of 89.3% by age 80. In particular, the penetrance for lung cancer at 80 years of age was close to 50%, which is notably higher compared to other well-known lung cancer susceptibility syndromes, such as Li–Fraumeni syndrome (TP53) [8] or EGFR-associated syndrome [9]. One of the key findings is the early onset of lung cancer, with one case diagnosed at 30 years of age – the earliest reported for SFTPA1/SFTPA2 PGVs. The study confirms that ILD generally occurs earlier than lung cancer in carriers of SFTPA1/SFTPA2 PGVs.

Penetrance of dominant diseases is known to be variable. However, great variations are described depending on the gene involved and the functional consequences of the PGVs. SFTPA1/SFTPA2 PGVs are associated with an impaired expression and secretion of the corresponding SP-A1 and SP-A2 proteins. However, the reason(s) why the age at onset of the ILD symptoms varies from a few months to more than 80 years is currently unknown. Viral infections, particularly in children, and environmental or occupational exposures and smoking in adults may play a role in triggering the disease. Unfortunately, these factors could not be retrieved for a relevant number of patients and relatives for study. The study found no significant gender differences in disease penetrance, in line with the autosomal dominant inheritance pattern, though higher smoking prevalence in men in France may influence the disease risk.

Surfactant disorders are rare causes of ILDs in adults [10]. In the context of genetic counselling, once a SFTPA1/SFTPA2 PGV is identified, patients are required to inform their relatives who may seek pre-symptomatic diagnosis. Since 2016, the national network for rare lung disease (RespiFIL, www.respifil.fr) has provided support for multidisciplinary team meetings for genetic forms of ILD in adults and in children [11, 12]. As more cases are being diagnosed, few guidelines are currently available for the management of familial pulmonary fibrosis [13, 14]. The study highlights the importance of monitoring SFTPA1 or SFTPA2 PGV carriers throughout their lifetime to assess the risk of ILD and lung cancer. We also propose clinical management strategies: i) performing the first lung computed tomography around 25 years of age; ii) a 5-year timeline between two subsequent occurrences of imaging; iii) providing information on environmental factors that could increase the risk of lung fibrosis and cancer such as tobacco smoking or occupational exposures. The results also indicate that genetic testing could be beneficial even for later-onset ILD, as the penetrance of ILD reached 50% at 60 years (range 55–65), which aligns with the criteria for genetic testing in later forms of ILD as outlined by French and European guidelines.

The study has limitations, including missing data on distant relatives, smoking and environmental factors. We assumed that no ILD or lung cancers were of sporadic occurrence, which might not be entirely accurate. Clinical, pathological and molecular characteristics were previously reported with all lung cancers being adenocarcinomas without targetable genetic alterations [15]. We suggest future improvements to the model, such as the incorporation of sporadic cases, the consideration of competitive risks to prevent death or censoring bias. We plan to include exposome data in future work in order to improve genetic counselling for variant carriers. Ongoing prospective studies, such as RaDiCo-ILD2 (FIFA project), will help validate the model and deepen our understanding of these diseases. The collection of environmental factors will be instrumental in refining the conclusion of this study.