Phenotyping Before Starting Treatment in COPD?

ABSTRACT

Chronic Obstructive Pulmonary Disease (COPD) is a heterogeneous and complex disease with great morbidity and mortality. Despite the new developments in the managements of COPD, it was recognized that not all patients benefit from the available medications. Therefore, efforts to identify subgroups or phenotypes had been made in order to predict who will respond to a class of drugs for COPD. This review will discuss phenotypes, endotypes, and subgroups such as the frequent exacerbator, the one with systemic inflammation, the fast decliner, ACOS, and the one with co-morbidities and their impact on therapy. It became apparent, that the “inflammatory” phenotypes: frequent exacerbator, chronic bronchitic, and those with a number of co-morbidities need inhaled corticosteroids; in contrast, the emphysematous type with dyspnea and lung hyperinflation, the fast decliner, need dual bronchodilation (deflators). However, larger, well designed studies clustering COPD patients are needed, in order to identify the important subgroups and thus, to lead to personalize management in COPD.

KEYWORDS:

• Chronic bronchitis
• co-morbidities
• emphysema
• endotype
• exacerbators
• genotype
• inflammation

Introduction

Chronic Obstructive Pulmonary Disease (COPD) is defined as a common preventable and treatable disease characterized by persistent airflow limitation and associated with an enhanced chronic inflammatory response in the airways and lung to noxious particles or gases (1,2).

Recently, the significant heterogeneity of clinical presentation of the disease and the complexity of the physiobiology of COPD had been evaluated (3,4). This had led to the recognition of the importance of phenotyping the various expressions of the disease in order to better manage COPD with a final goal to personalize the treatment of each individual patient (5–7).

It is a great improvement that we have moved from the airflow limitation (FEV¹) into severity scales with symptoms, exacerbations in the management of the disease (8–10). The recent, updated Global Initiative for Chronic Obstructive Lung Disease (GOLD) guidelines had introduced the A,B,C,D, system to verify the severity of the disease. This scale includes measures of symptoms such as dyspnea using the Medical Research Council (MRC) score or COPD Assessment Test (CAT) and the number of exacerbations (1).

Although this system had a number of limitations, it was the first system to introduce symptoms and exacerbations to determine the severity of COPD (1).

In addition, GOLD proposed to use this severity index (A,B,C,D) in order to administer the various modes of treatments available for COPD (1).

However, detailed analysis of the various clinical characteristics (phenotypes) of the patients is missing in the GOLD document (11).

It becomes evident that not all kind of treatments are suitable for all the COPD patients, and phenotyping, subgrouping, and/or endotyping would be essential in order to tailor the management of an individual patient with COPD. It would be of great clinical value to be able to predict who will respond to specific class of drugs, leading to personalized medicine (12–14).

This brief review will present first the definition(s) of phenotype, subgroup, and endotype, and then will discuss the various phenotypes of COPD and their significance on therapy.

Phentotype(s) (definition)

The following definitions are needed in order to facilitate the understanding of the pathogenesis of diseases and their management in general.

Genotype

Genotype is the genetic constitutional of an individual organism, the inherited map that carries the genetic code (DNA). However, not all organisms with same genotype look or act the same way, because of epigenetic influences such as, environmental factors. The final, end result of both genetic and environmental factors is called phenotype.

Thus, the definition of phenotype could be: The set of observable characteristics of an individual resulting from the interaction of its genotype with the environment (15).

Finally, terms such as subgroup, endotype, and subtype have been used to define distinct patho-physiological mechanism. Thus, endotype differs from phenotype (the observable characteristics) since it is referring to a subgroup with a distinct pathobiological mechanism. It is common finding in COPD it's presentation with a number of phenotypes that could be subdivided into a number of endotypes. Finally, the above terms are used by many researchers indistinguishably, and this causes a great confusion in the medical literature (5,16,17).

Phenotype(s) primarily applies in diseases with a strong epigenetic pathobiological mechanism(s) such as, COPD. It is obvious from the definition that this disease is the result of an environmental insult(s) (cigarette smoke, biomass fuels) that interacts with the genotype of a “susceptible” individual (18–20).

Finally, strategies to detect susceptibility loci of complex diseases in relationship with specific phenotypes had been described by Rice et al. (21).

Genetic factors in COPD: A brief summary

The pathways from genotype to phenotype in COPD are summarized in Figure 1. This figure illustrates that if the genome is genetically severely altered, as in a1-antithrypsin deficiency, the disease appears without a significant involvement of the environment (Figure 1A). However, it is well known that environmental insults such as smoking and biofuel smoke exposure can promote the earlier development of the disease in this condition (22). In addition, an interaction between genetics and the environment is supposed to enhance the development of the disease in susceptible individuals (Figure 1B) (23).

Figure 1. From Genotype to Phenotype in COPD. The most common pathway is the B where environmental insults such as cigarette smoking, Air pollution, and Biomass Fuel are affecting the genome of susceptible individuals (For more details, see the text). A = α1-antitrypsin deficiency; B = environmental insults: smoking, air pollution, biomass fuel; C = nutrition; B,C = EPIGENETICS.

Proposed pathogenetic mechanisms in COPD include the protease-antiprotease imbalance, response to oxidative stress, cell death, and inflammation (24–26). Family-based and single-gene studies have discovered genes and loci that are associated with COPD susceptibility, while recent genome-wide association studies (GWASs) have discovered novel candidate genetic pathways. The phenotype of a1-antithrypsin deficiency, caused by mutations in SERPINA1, accounts only for the 1–2% of the total COPD population; however, it is the only established genetically driven cause of COPD that has a potential intervention so far (27). Inducible heme oxygenase (HO-1) is a cytoprotective enzyme that plays a critical role in lung defense against inflammatory and oxidant-induced cellular and tissue injury in COPD (28). Polymorphisms of the HO-1 promoter associated with reduced HO-1 expression have been linked with increased susceptibility to smoking-induced emphysema (29,30). Other candidate genes for COPD recently identified by GWASs include CHRNA3/5 (cholinergic nicotine receptor alpha 3/5), IREB2 (iron regulatory binding protein 2), HHIP (hedgehog-interacting protein), FAM13A (family with sequence similarity 13, member A), and AGER (advanced glycosylation end product–specific receptor), (31–34). Although their pathological roles are still largely unknown, potential liaisons with COPD susceptibility and associations with emphysema severity (35), the chronic bronchitis subtype (36), lung function (37), COPD exacerbations (38), and pathogenesis of pulmonary hypertension (39) have been proposed and replicated in multiple populations.

The most common pathway to develop COPD is the epigenetic one, when environmental factors affecting the genome of susceptible individuals (Figure 1B) (39–42).

The basic epigenetic mechanisms are DNA methylation, which suppresses gene transcription, and various post-translational modifications of core histones (histone acetylation, methylation, ubiquitination, and phosphorylation), that may result in either activation or repression of genes (40–42). Several studies have proposed that oxidative damage related to smoking is associated with epigenetic modifications that may mediate COPD development and progression by modulating gene expression of proinflammatory cytokines and inflammatory signaling pathways (43,44), as well as by inducing cell apoptosis (45). Moreover, evidence shows that oxidative damage contributes to genetic instability of specific non-coding DNA sites (microsatellite sequences, telomeres, promoters, and sites of methylation) adjacent to or included in genes implicated in the pathogenesis of COPD or supposed to be associated with certain phenotypic expressions of the disease (46,47).

However, in some cases, there is a protective role of the environment that can prevent the development of the disease, such as nutrition (Figure 1C). This may explain the fact that not all smokers develop COPD (23).

Since the insults of the environment could hit different loci of the genome of a susceptible individual, a number of phenotypes of COPD could be identified. This may be the answer to such heterogeneity seen in COPD. This also could explain the fact that clinical phenotypes can overlap in the same patient (several pathogenetic pathways) (48).

In addition, since the interaction between the environment and the genome has a long life duration, it raises the question if a certain phenotype is constant over time.

Phenotyping COPD

As noticed above, a phenotype of a heterogeneous disease defines a subgroup of patients with similar observable characteristics. In such complex disease as COPD that lasts the whole life, this exercise is quite difficult. It is obvious that clustering patients with clinical or epidemiological criteria, including imaging (HRCT) (49–52), is quite different than clustering by pathophysiological or pathobiological criteria, and it is extremely complex if we use a large number of criteria. Due to these limitations, the medical literature is lacking sufficient number of valid studies on the issue of the phenotypes of COPD (5,16,17,49–57).

Historically, two classical phenotypes of COPD had been described: the Chronic Bronchitic and the Emphysematous one. Using pathophysiological criteria, these phenotypes were named also as “Blue Bloaters” and “Pink Puffers” (58). Recently, a number of distinct phenotypes had been described such as, the frequent exacerbator, the fast decliner (fast drop of FEV¹ over the years), the phenotype with systemic inflammation, the one with a number of severe co-morbidities such as cardiovascular or metabolic ones (59), and the one of significant hyperinflation.

To complicate even further the issue of phenotyping COPD patients, we acknowledge that any one individual may manifest multiple phenotypes, etiologically different. Although there are significant difficulties and confusions in this area of research, the current scientific efforts are focusing to identify biomarkers that can describe a similar underlying mechanism(s) and thus, define better a COPD phenotype or endotype. Therefore, the efforts are to describe disease attributes different between COPD patients that are related to meaningful clinical outcomes, such as, symptoms, exacerbations response to treatment disease progression, or death (3,12,49–59).

The major goal of defining phenotypes in COPD is to identify the individuals that could respond to specific mode of treatment, because it is well known, that not all patients are affected by all available medications. This may lead to personalized treatment with consequence the better manager of the disease, better quality of life for the patient, and reduction in the cost of therapy.

Finally, the overlap syndrome between COPD and asthma (ACOS) has been considered as specific phenotype of COPD or asthma. As we believe that this syndrome is very rate and could be relevant only to the smoking Asthmatic, we are not going to discuss it in detail (60,61). Table 1 lists the most common phenotypes of COPD, although there is no consensus on the number or the type of phenotypes in the medical literature (16). However, the future of COPD is phenotyping the disease (60–67).

Table 1. COPD phenotypes.

Classical:	Chronic Bronchitic
[Clinico-histological	Emphysematous
Criteria]	Mixed
Historical (pathophysiological criteria)	Blue Bloater – Pink Puffer
One Criteria	Frequent Exacerbator
	Inflammatory phenotype
	Fast decliner (FEV^1)
	Co-morbidities
	Current smokers
Multi Criteria	•Chronic bronchitis with frequent exacerbations
	•Emphysema with frequent exacerbations
	•Younger with severe disease, few co-morbidities, poor nutritional status, and poor health status
	•Older with moderate disease, obesity, cardiovascular and metabolic co-morbidities
	•ACOS: Asthma-COPD – Overlap Syndrome

Classical COPD phenotypes

The classical division of COPD patients into blue bloater and Pink Puffers had been described in many textbooks as well as the clinical definition of Chronic Bronchitis or the histological one of Emphysema. In brief:

Chronic bronchitis

Cough and sputum production for at least 3 months in each of two consecutive years.

Emphysema

Destruction of the gas-exchanging surfaces of the lung (alveoli), beyond the terminal bronchioles.

Since both characteristics can coexist in many patients, the term COPD had been introduced. To describe the pathogenetic mechanisms, the first attempt of phenotyping COPD was to use the terms “pink puffer” and “blue bloater.” In brief, the pink puffer is a patient with emphysema as primary pathology: destruction of the airways distal to terminal bronchiole, and gradual damage of the pulmonary capillary bed. This leads to less ventilation–perfusion mismatch than the blue bloaters, and thus, to relatively “normal” blood gases due to hyperventilation (puffers). These persons usually develop muscle wasting and weight loss, although they have less hypoxemia than the blue bloaters.

The Blue Bloater

The Blue Bloater has primary pathology as that of chronic bronchitis. (Mucus-producing glands hypertrophy, goblet cell metaplasia, chronic inflammation of the bronchial tree). The ventilation–perfusion mismatch is increased, leading to hypoxemia, and hypercapnia occurs due to altered pattern of breathing. They present cyanosis in the face and lips, thus the “Blue” (58).

New COPD phenotypes

Briefly we are going to present the criteria used to define the new COPD phenotypes.

Frequent exacerbator

Phenotype is defined if the patient has two or more exacerbations per year and implies worse prognosis. This phenotype was detected from clinical records or asking the patient the right question. According to Hurst et al., this is a quite stable phenotype over the years (60). However, this statement was challenged by Donaldon et al. analyzing further the findings of the ECLIPSE study (68).

The fast decliner

Phenotype is defined if the patient shows greater than the average fall in Forced Expiratory Volume in 1 second (FEV¹). This needs at least 3-year measurements of FEV¹ to conclude the fast decliner, and this is a significant limitation, in detecting this phenotype (69,70). However, if HRCT and lung function tests were used, it had been shown that this phenotype reflects Emphysema (71–72).

Inflammatory phenotype

Inflammatory phenotype is described in patient with persistent elevation of serum inflammatory markers, such as C-reactive protein or other proinflammatory cytokines (73–76).

Current smoker phenotype

Some researchers consider the current smoking COPD patients as a district phenotype with specific psychosomatic behavior, worse prognosis, and poor adherence and response to treatment (77).

The systemic or co-morbidities phenotype

This refers to the patient with a number of cardiovascular, metabolic co-morbidities. However, no specific criteria had been reported, as far as the number or the severity of those co-morbidities is concerned. Epidemiological studies had shown that COPD patients commonly are suffering from other chronic diseases such as the metabolic syndrome, arterial hypertension, ischemic heart diseases, diabetics, osteoporosis, and psychological disorders (anxiety/depression).The theory of “spill over” had been proposed as the possible pathobiological mechanism(s). According to the “spill over,” the prime site of the “burning” and the inflammation is the lungs, but the products of this inflammation are spilled over via the circulation to various sensitive organs such as the brain, the liver, the muscles, and the kidneys. (78–81). It is obvious that this pathogenetic mechanism is present only in a specific phenotype or subgroup of COPD, and this is not the case in all COPD patients. Finally, a number of other phenotypes had been proposed such as the “eosinophilic” one. However, eosinophilia had been seen mainly during the exacerbations in some patients (82–85). In addition, an effort to identify clinical phenotypes by cluster analysis had been reported (86).

Phenotypes in European guidelines for COPD

A recent review had shown that the identification of patient subtypes had been included in various European Guidelines for COPD (11,59). In addition, it was noticed that there was a great variability between those Guidelines as far as the criteria used to identify the phenotypes are concerned. However, the classical ones Emphysema, chronic Bronchitis, Pink Puffer, Blue Bloater, those with dyspnea, the frequent exacerbators, and the overlap syndrome ACOS were presented in most of the guidelines (11,59).

Pharmacotherapy of COPD by clinical phenotypes

The current management of COPD has pharmacotherapy and non-pharmacotherapy measures. The pharmacotherapy of COPD includes primarily b2 agonists, anti-muscarinics, inhaled or per os steroids, theophyline, PDE4 inhibitors, mucolytics, and macrolides.

Most of the recent guidelines made an effort to guide the therapy of COPD according to the phenotype of the patients, and this is considered as significant improvement in the management of COPD (11). The GOLD guidelines propose for patient group A: SAMA or SABA, for B: LAMA or LABA for C: ICS + LABA or LAMA and for D: ICS + LABA + LAMA as first-choice treatments in stable COPD patients. However, number of discrepancies can occur using this scale, for example, (1) a patient can be there by either an FEV¹ <50% or (2) exacerbations + dyspnea. The first case could be a fast decliner with hyperinflation and the second having an inflammatory phenotype, thus, two different patients. Therefore, the new refine system uses both GOLD 1,2,3,4 Scale based on FEV¹ and the A,B,C,D based on symptoms and exacerbations (1).

It was noticed, however, that there are extremely heterogeneous recommendations in the various European guidelines, as far as, the choice of the first line treatment choice according to phenotyping (11,59). However, the Spanish guidelines for the treatment of COPD had used the combination of emphysema, chronic bronchitis, exacerbations and the overlap COPD-asthma and proposed four phenotypes to facilitate the mode of treatment (87). No exacerbator, ACOS, exacerbator with emphysema, and exacerbator with chronic bronchitis were the four clinical phenotypes of the Spanish effort, with the following recommendations being summarized in Table 2. This is a practical guideline to help clinicians choose the most appropriate treatment for their COPD patients (87,88).

Table 2. The Spanish approach of pharmacological treatment guided by clinical phenotypes.

	Bronchodilators	ICS	Mucolytics	PDE4 inhibitors	Macrolides
No Exacerbator	✓
ACOS	✓	✓
Exacerbator with emphysema	✓	✓	✓
Exacerbator with chronic bronchitis	✓	✓	✓	✓	✓

ACOS = Asthma-COPD overlap Syndrome; ICS = Inhaled Corticosteroids; PDE4 = Phosphodiesterase 4.

Modified from Ref. (11).

Bronchodilators have been the therapeutic mainstay for patients with COPD and in many trials have been shown their efficiency for reduction of exacerbation frequency. Recent meta-analyses have shown that all LAMAs were equally effective in preventing moderate-to-severe exacerbations (89). A meta-analysis of LABAs showed a decrease of COPD exacerbations that was close to 20%, although differences between them were reported (90). LABA/LAMA has shown better effects on lung function and exacerbations than mono-component long-acting bronchodilator therapies (91–93).

LABA/ICS has been shown to decrease rate of COPD exacerbations and improve lung function and health-related quality of life (94). The network meta-analysis compared four different classes of long-acting inhalers (LABA, LAMA, ICS, and ICS/LABA) and has shown improvement of quality of life and lung function by combination drugs (LABA and ICS) and least on ICS alone at 6 and at 12 months (94). Overall, LAMA and LABA inhalers had similar effects, particularly at 12 months.

Recently, the withdrawal of the inhaled corticosteroids was tested and showed no significant effect on the number of exacerbations (95). However, the decline in FEV¹ was greater in patients without steroids than in those who continued having inhaled steroids (95).

In addition, the effectiveness of dual LABA + LAMA versus Salmeterol + Fluticasone on exacerbations had been tested and shown superior (96). Finally, the combination of fluticasone + Vilanterol reported very effective in clinical practice in COPD (97).

PDE-4 inhibitors may play an important role in preventing COPD exacerbations in patients with chronic bronchitis COPD phenotype (98). Therapy with a PDE-4 inhibitor was associated with a reduced number of COPD exacerbations (odds ratio (OR) 0.77; 95% confidence interval (CI) 0.71–0.83, high-quality evidence). For every 100 people treated with PDE-4 inhibitors, six more remained exacerbation-free during the study period compared with placebo (number needed to treat for an additional beneficial effect (NNTB) 20; 95% CI 16–27) (98,99).

Macrolide antibiotics have been demonstrated to have anti-inflammatory and immunomodulating properties and have been used for the reduction of COPD exacerbations (100). A large, prospective, placebo-controlled, randomized trial on the use of azithromycin for prevention of COPD exacerbations showed that macrolide was associated with a significant decrease in exacerbation frequency (101). Use of prophylactic antibiotic in the long term should be mindful of the balance between benefits to individual patients and the potential harms caused by antibiotic overuse (102,103).

Mucolytics have been used to reduce sputum viscosity and facilitate expectoration, and they also have anti-oxidative properties. The most recent systematic review of clinical effectiveness of mucolytics included 34 studies and has shown a small, but statistically significant, reduction in exacerbations in treated COPD patients (104).

Reviewing the literature (105–107) for the topic of phenotyping of COPD and their treatment, we could only recommend the following: The phenotypes of Chronic Bronchitis, the frequent exacerbator, the one with a number of co-morbidities, and the systemic inflammatory phenotype are probably due to a “spill over” an enhanced or “abnormal” inflammatory pathobiogenetic mechanism, and thus, inhaled corticosteroids should be included in their management. PDE4 inhibitors, macrolides, and mucolytics could be added in specific subgroups of the above (1,11). In contrast, the emphysematous, the fast decliner, those with significant hyperinflation and severe dyspnea, the basis of their treatment should be long-acting bronchodilators (deflators). In severe cases of these phenotypes, dual bronchodilation with LAMA + LABA should be used (108).

Table 3 summarizes the above in a practical way for the clinician!

Table 3. Simple clinical approach to treatment using clinical phenotypes in stable COPD.

	Phenotype	Pharmacotherapy
GROUP A	Chronic bronchitis	ICS
	Inflammatory phenotype	LAMA, LABA
	Frequent exacerbator	±PDF4, inhibitors
	Systemic manifestation	±Macrolides
	With co-morbidities
GROUP B	Emphysema	Dual Bronchodilation with LAMA + LABA
	Fast decliner
	Significant lung hyperinflation
	±Frequent exacerbators

In conclusion, this review noticed the lack of scientifically robust studies clustering COPD patients according to specific pathogenetic mechanism(s) (endotypes) or specific clinical characteristics (phenotypes). Thus, recommendations of modes of treatment based on phenotyping of patients of COPD are at their infancy (109,110).

Therefore, large, well-designed studies are needed in order to identify more accurately the phenotypes, or endotypes or subgroups of this heterogeneous and complex disease, leading to improvement in personalized medicine.

Declaration of interest

None.