Familial Segregation Highlights Challenges in Rare <i>SFTPA2</i> Variant Classification

We noted with appreciation the study by Louvrier et al. investigating the molecular causes of early-onset interstitial lung disease (ILD). They identified likely pathogenic/pathogenic variants in surfactant-related and other genes involved in early-onset ILD in 62/699 patients (8.9%), of which 13 (21%) were located in the SFTPA2 gene [1]. One of the SFTPA2 variants involved the 20 base pair deletion c.635_654del (NM_001098668.4) variant, which is predicted to result in a frameshift leading to a premature termination codon, p.(Tyr212Serfs*17) and was classified by the authors as likely pathogenic. They detected this variant in a patient with pulmonary fibrosis (PF), who had disease onset at 44 years of age and no reported family history of PF or lung cancer. Interestingly, we detected the same SFTPA2 c.635_654del variant in a 71-year-old female patient with familial PF. However, we simultaneously identified the rare TERT c.2011C>T (NM_198253.2) variant in this patient, which causes a nonsynonymous amino-acid substitution in the reverse transcriptase domain, p.(Arg671Trp). This TERT variant co-segregated with disease in all five relatives with PF (Figure 1). The clinical phenotypes observed in affected individuals aligned most closely with those reported for TERT variant carriers; the age at diagnosis was relatively advanced [2] and HRCT patterns corresponded mostly to patterns regularly described in TERT variant carriers (Figure 1). Following ACMG/AMP guidelines, the TERT c.2011C>T variant was classified as likely pathogenic. Rare heterozygous variants in the SFTPA2 gene have been associated with adult-onset pulmonary fibrosis and lung adenocarcinoma [2, 3]. To date, all disease-associated variants are non-synonymous variants located in exon 6, which encodes the carbohydrate recognition domain. Carriers of SFTPA2 variants typically exhibit a non-specific interstitial pneumonia pattern on HRCT containing ground glass opacities and septal thickening, with disease onset generally occurring before the age of 60 [2]. In our PF-family, lung adenocarcinoma was not reported and the SFTPA2 c.635_654del variant did not co-segregate with PF (Figure 1). The c.635_654del variant has an allele frequency of 0.00024 in the European, non-Finnish population in the GnomAD database and since it lies within the terminal exon, the transcript is unlikely to undergo nonsense-mediated decay. Pathogenic SFTPA2 variants have been shown to cause increased protein instability, defective oligomerization, intracellular protein aggregation, and chronic endoplasmic reticulum stress [3, 4], a cascade considered to result in cell apoptosis as previously documented for SFTPC mutations [2]. SFTPA2 variants are therefore thought to exert their pathogenic effects through toxic gain-of-function mechanisms; loss-of-function has not been implicated as a disease-causing mechanism. The cellular consequences of the SFTPA2 635_654del variant remain unknown, and functional studies are needed to clarify its impact. Notably, recent findings by Desroziers and co-workers showed that co-expression of wild-type SP-A2 can mitigate the toxic effects of mutant proteins in trans [3]. Future functional analyses should ideally consider the broader haplotypic context, as additional cis or trans variants in SFTPA2 may influence its effect. In conclusion, segregation analysis in our PF-family highlights challenges in variant classification and indicates that available evidence is currently insufficient to classify the SFTPA2 c.635_654del variant as likely pathogenic. The authors have nothing to report. J.V. reports a lecture fee from Chiesi paid to her institution. J.-M.F. reports advisory board fees from Boehringer Ingelheim.