Genetic Association Study Advances Idiopathic Pulmonary Fibrosis Pathophysiology and Health Equity

Idiopathic pulmonary fibrosis (IPF) is a devastating lung disease that, despite the advent of antifibrotic agents, urgently needs additional research to discover new pathobiology to unearth novel therapeutic pathways. Human genetics support for drug targets increases the chance that related drugs will be approved as novel therapies (1). IPF susceptibility is partially attributable to genetics, with ∼20% of all IPF cases being familial (two or more IPF cases noted in a family) and with common genetic variants explaining ∼30% or more of the risk of sporadic IPF (2). Genome-wide association studies (GWASs) offer an unbiased approach to identify population-level common genetic variants associated with IPF, and the most recent IPF GWAS reported a total of 14 genetic regions associated with IPF, including novel associations at three loci including a variant (rs78238620) near KIF15 (kinesin family member 15) (3). Although GWASs are a powerful tool for the robust association of genetic variation with disease, the majority of GWAS associations fall in noncoding regions of the genome, making translation of genetic associations to disease-causing genes and disease-relevant biologic function difficult. By contrast, studies of nonsynonymous (i.e., associated with a change in protein structure) rare genetic variants allow for the direct assignment of a genetic association with a putative causal gene. Genetic studies of familial and sporadic IPF using targeted sequencing of candidate regions as well as whole-exome sequencing have identified IPF-associated rare genetic variants in surfactant protein genes and telomere-related genes, including TERT, TERC, RTEL1, PARN, and others (4–6).

In this issue of Journal, Zhang and colleagues (pp. 56–69) report the largest-to-date IPF genetic association analysis of rare deleterious protein-coding genetic variants (7). The authors used whole-exome and whole-genome sequencing of international multiple-ancestry familial and sporadic IPF cohorts with ∼3,250 total IPF cases included in the overall meta-analysis. In addition to replicating prior rare deleterious variant IPF associations in the telomere-related genes TERT, PARN, and RTEL1, the authors for the first time reported an excess number of rare deleterious KIF15 variants in IPF cases compared with controls. The enrichment of rare deleterious KIF15 variants in IPF cases is interesting considering the above-mentioned common variant IPF GWAS locus near KIF15 (3). The IPF GWAS locus near KIF15 had yet to be convincingly mapped to a causative gene. Finding rare variants of large effect in a gene near a common variant GWAS locus is not typical and significantly advances our knowledge of the pathobiology of IPF. KIF15 is in the kinesin family and participates in the bipolar spindle assembly, which is essential to cell division (8). Zhang and colleagues performed functional studies demonstrating that decreased KIF15 protein expression reduced proliferation of lymphoblastoid cells heterozygous for the KIF15 common variant rs78238620. Furthermore, the authors analyzed publicly available human lung single-cell RNA-sequencing data to demonstrate that KIF15 gene expression is present in replicating epithelial and resident immune cells in the lungs. Together, these findings provide baseline functional evidence of common and rare genetic variants in and around KIF15 perturbing a telomerase-independent pathway germane to the pathobiology of IPF.

The manuscript by Zhang and colleagues represents a significant advance in the study of genetic susceptibility to IPF; however, it also presents an opportunity to highlight the propagation of racial and/or ethnic health disparities through biomedical research generally and complex disease genetics specifically. The authors do not refer to self-reported race and/or ethnicity in the description of their IPF cohorts and remained (perhaps appropriately) focused on genetic ancestry of individuals in the performance of the quality control and statistical analysis of genetic data. That said, only 8% of individuals overall and 3% of individuals in the replication cohort were of non-European ancestry. The underlying prevalence of IPF across racial and/or ethnic groups is difficult to assess; however, in contrast to the 8% of non-White participants in the current study, an examination of 37 worldwide IPF clinical trial and registry studies demonstrated a heterogeneous estimate of 14% (range, 3–32%) of all participants being non-White (9). Moreover, only 23 of the 37 studies reported race or ethnicity, which is important to ensure diversity. A 2006 retrospective cohort study of 2,635 patients with IPF listed for lung transplantation reported that the 18% of non-White individuals had more severe IPF at transplant listing and worse post-transplant survival, suggesting that non-White individuals are underrepresented in IPF lung transplant and the proportion of non-White individuals with IPF may be even higher than 18% (10). Despite the small percentage of individuals of non-European ancestry with IPF in the manuscript by Zhang and colleagues, the authors have disrupted a paradigm by including these individuals in the overall analysis of deleterious KIF15 variants associated with IPF. Citing possible bias due to population substructure (a legitimate concern in genetic association studies), most prior IPF genetic studies have exclusively analyzed individuals of European ancestry (2, 3, 11, 12). In addition to the social imperative to incorporate more population diversity in IPF genetics research, population differences in rare deleterious variants can augment discovery of disease-associated genes when multiple ancestry groups are analyzed (13). Thus, Zhang and colleagues may not have discovered IPF-associated rare deleterious variants in KIF15 if they had limited their study to a single ancestry group. Analyzing multiple ancestries can also augment statistical power for common genetic variant studies. For instance, Putman and colleagues studied the association of the common MUC5B promoter polymorphism (rs35705950) with subpleural interstitial lung abnormalities in the COPD Genetic Epidemiology (COPDGene) study and noted a similar point estimate for the odds ratio in non-Hispanic White and African American individuals despite a large difference in allele frequency (non-Hispanic White, 11%; African American, 2%) between the two populations (14).

We should celebrate the opportunity that Zhang and colleagues have given us to investigate KIF15 as a novel pathologic mechanism for IPF independent of telomerase pathways. At the same time, we as a pulmonary community should strive to include more diverse populations in future clinical, epidemiologic, and genetic studies of IPF to improve health disparities and simultaneously increase the statistical rigor and generalizability of our research.