Is personalized medicine a realistic goal in idiopathic pulmonary fibrosis?

KEYWORDS:

• Idiopathic pulmonary fibrosis
• personalized medicine
• pulmospheres
• biomarkers
• T-cells
• autoimmunity
• 3D spheroids
• cellular invasion

1. Introduction

Personalized or precision medicine is based on the premise that the clinical presentation, natural history, and responsiveness to drug therapies are invariably variable across most human diseases. This variability or heterogeneity is particularly evident in complex, polygenic diseases such as idiopathic pulmonary fibrosis (IPF). Although it is in such complex diseases that personalized medicine could have the greatest impact, it is also in such diseases that it is most challenging. Although high-throughput genomics was the main impetus behind the personalized/precision medicine revolution, other factors such as environment, lifestyle, and aging may influence disease development and progression. Thus, one or more of these factors may be considered when applying personalized approaches to the prevention, diagnosis, and treatment of disease. There has been recent interest in applying personalized/precision medicine approaches, specifically to IPF [1–3].

The success of personalized medicine in IPF will be dependent on identifying clinically definable subphenotypes of the disease. These subphenotypes should be based on a distinct (or predominant) pathobiology (‘endotype’) which drives the risk of disease expression, natural course and/or progression, and responses to specific therapeutic interventions. Although personalized medicine approaches has led to targeted therapies in mongenic diseases in which the disease mutation and its effects on cellular function and pathogenicity is known [4], such approaches are being applied to more complex, polygenic diseases. For example, an asthma subgroup defined by Th2-driven inflammation (endotype) and a Th2-linked biomarker (periostin; subphenotype) demonstrated a favorable response to Th2/IL-13-targeted therapy [5]. Success in personalized medicine for IPF will, similarly, be predicated on identifying clinically definable endotype–phenotype subgroups on the basis of biomarkers that inform the underlying pathobiology.

For this discussion, we broadly define biomarkers as any measure of the underlying pathobiology that is informed by – omics approaches (e.g. genomics, transcriptomics, metabolomics, proteomics, and the microbiome), specific proteins and/or their modifications, and lung tissue-based sampling/analysis. Although each approach has its advantages and disadvantages, active research is being conducted in each of these areas, further illuminating the heterogeneity of IPF and the value of identifying endotype–phenotype subgroups.

2. Genomics

Genome-wide association studies in IPF have identified several genetic loci that seem implicate host defense, cell–cell adhesion, and DNA repair processes in disease susceptibility [6]. A common single-nucleotide polymorphism (SNP) variant located in the promoter region of the mucin gene, MUC5B, is strongly associated with IPF; patients with this variant allele may represent a pathogenically distinct endotype that confers both a significantly higher predisposition to disease, but with a significantly longer survival [7]. Gene variants may also be predictive of responsiveness to anti-fibrotic drug therapies. A SNP variant of the gene (TT genotype, rs3750920) encoding toll interacting protein, TOLLIP, was associated with significant reduction in composite endpoint risk defined by death, transplantation, hospitalization, or a decline of >10% in FVC, while a nonsignificant increase in composite endpoint risk was seen in those with the CC genotype [8]. A major challenge is to determine how SNPs in candidate genes induce biochemical and cellular functional changes that contribute to disease susceptibility or progression.

3. Circulating biomarkers

Plasma biomarkers have been shown to be useful in prognostication of patients with IPF. For example, elevated plasma levels of MMP-7, ICAM-1, and IL-8 were predictive of overall survival in an IPF cohort [9], supporting pathobiological endotypes of epithelial cell injury and inflammation. Peripheral blood mononuclear cell gene expression profiles suggestive of T-cell activation was predictive of transplant-free survival in IPF [10]. Interestingly, candidate genes identified in this profile correlated with the levels of circulating CD4⁺CD28⁺ T cells, which had previously been shown to portend a poor prognosis [11]. More recently, a peripheral blood 52-gene expression signature was found to be predictive of outcomes in patients with IPF, and this signature correlated with responses to therapy [12]. Another study correlated distinct pathologies in the IPF lung involving bronchiolisation and lymphoid aggregates with circulating levels of MMP3 and CXCL13, respectively [13]. Increased concentrations of extracellular matrix-derived protein fragments were associated with disease progression, while the rate of this increase predicts survival [14]. Oxidative stress has been implicated in the pathogenesis of IPF [15,16], and circulating levels of protein-associated tyrosine oxidation products are elevated in IPF [17]. It remains to be determined if these biomarkers can be employed to define a distinct IPF endotype, or whether these biomarkers are predictive of disease progression or responsiveness to specific drug therapies.

4. Tissue-based biomarkers

Whole lung tissue and cells from lungs of patients undergoing biopsy or lung transplant may serve to identify pathologic endotypes of disease. These may include precision-cut lung slices [18], 3D lung organoids [18,19], and pulmospheres [20]. Pulmospheres have been applied to study another proposed IPF endotype, fibroblast invasion, and this approach showed that clinical responsiveness to currently approved FDA drugs for IPF was associated with ex vivo responses of pulmospheres to the same drugs [20]. A distinct advantage of this multicellular pulmosphere model is that it incorporates many of the cell autonomous and non-cell autonomous factors that may contribute to specific pro-fibrotic cellular phenotypes. Another practical advantage is that biopsy specimens can be cryopreserved prior to assays. It remains to be determined whether transbronchial biopsies or cryobiopsies (as opposed to surgical biopsies) will provide sufficient numbers of cells to form pulmospheres. Additional studies of pulmospheres are required to explore other pro-fibrotic cellular endotypes of the disease, in addition to prospective evaluations of their predictive power.

In summary, it is becoming increasingly clear that clinical heterogeneity in IPF may be related to distinct, yet overlapping, pathobiological mechanisms. These mechanisms may represent disease endotypes such as epithelial cell dysfunction, impaired host defense, T-cell exhaustion, fibroblast activation, oxidative stress, and senescence/aging. The identification of biomarkers that are linked to specific endotypes will be critical to understanding their respective roles in disease initiation and/or progression. Molecular imaging to identify unique and distinct endotypes must be considered. Pharmacogenomic approaches may be able to aid physicians to more precisely prescribe drug therapies. As newer drug targets are identified and novel anti-fibrotic drugs developed, a critical unmet need is the discovery of biomarkers that reliably measure and predict the clinical behavior of distinct endotypes–phenotypes in IPF. Recent advances indicate that personalized medicine is indeed a realistic goal in IPF, with the hope that targeted therapies with greater efficacy and tolerability will be available to IPF patients in the not too distant future.

Declaration of interest

VJ Thannickal reports receiving research support from the U.S National Institutes of Health and the U.S Department of Veteran’s Affairs. VB Antony reports receiving research support from the U.S National Institutes of Health. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed. A reviewer on this mansucript have disclosed they are an employee of a company developing biomarkers.

Additional information

Funding

This manuscript has received support from the National Institute of Health (P01 HL114470, R01 AG046210) and by the Department of Veteran’s Affairs Merit Award (5I01BX003056).