Background
Chronic obstructive pulmonary disease (COPD) is an inflammatory lung disease characterised by persistent airflow obstruction, which usually progresses [
1]. Immune activation may be one of the factors perpetuating inflammation in COPD [
2]. The immune response seen in COPD incorporates cells from innate and adaptive immune systems [
3]; an essential component of adaptive immunity is the production of antibodies by mature B lymphocytes. Antibodies are immunoglobulins which are composed of two polypeptide heavy chains and two light chains. During antibody production there is an excess of free light chains (FLCs) produced which are secreted into the circulation before undergoing renal clearance [
4]. There are two free light chain isotypes: kappa (κ) and lambda (λ), which can be measured by a highly sensitive assay [
5]. High FLC levels occur in several autoimmune and inflammatory conditions, thus suggesting they may be a biomarker of adaptive immune activation [
4]. Raised polyclonal FLCs have also been reported in a number of respiratory conditions where adaptive immunity may be important, including asthma [
6] and COPD [
7].
In addition to being a marker of immune activation, FLCs could have a direct pathogenic role in COPD. They inhibit neutrophil apoptosis [
8], inhibit neutrophil migration in vitro [
9], and are elevated in both murine models of emphysema and serum from patients with COPD [
7]. In addition FLCs bind to human neutrophils, activating them to produce IL8 in vitro; specific FLC antagonism inhibited this binding capability and reduced pulmonary neutrophilia in smoke exposed mice [
7].
The primary aim of this study was to investigate the clinical utility of FLCs as a biomarker in patients with alpha-1-antitrypsin deficiency (A1ATD). We hypothesised that FLCs would be static in stable disease, relate to disease severity, distinguish clinically relevant subgroups, relate to factors which could stimulate the adaptive immune system and associate with longitudinal outcomes, such as mortality. We then sought to replicate these associations in “usual” (non A1ATD) COPD. Furthermore we hypothesised that levels would be similar in usual COPD to A1ATD, since pulmonary immune activation has recently been shown to be similar between these groups [
10].
Discussion
Our main objective was to investigate the utility of measuring polyclonal FLCs as a clinical biomarker in severe A1ATD and usual COPD. Key properties of a clinically useful biomarker are that it is reproducible in stable disease, relates to disease severity and relates to outcome. Our results demonstrate that cFLCs meet many of these criteria, notably being associated with subsequent mortality in both our cohorts. No significant difference was seen in cFLCs taken from patients with stable disease at different time points, suggesting that cFLCs are reproducible in stable disease. We did not see a strong relationship between cFLC levels and disease severity, although there was a difference observed between patients with and without chronic bronchitis, which is recognised to be a clinically relevant subgroup within airways disease [
20].
A role for the adaptive immune system in perpetuation of inflammation in COPD has been proposed, since accumulation of B cells in large and small airways associates with worsening disease severity [
21]. FLCs, produced as a by-product of immunoglobulin synthesis by mature B cells, could be a useful marker of adaptive immune system activity [
4]. The prevalence of other autoimmune diseases was low in our A1ATD cohort, and no relationships were seen between cFLC levels and autoimmune disease burden. However, prior studies suggest that cFLCs change during periods of disease ‘activity’ (e.g. in rheumatoid arthritis [
22] and systemic lupus erythmatosus [
23]) such that presence of well controlled (inactive) autoimmune conditions might explain the lack of association observed. Furthermore, many important questions regarding the role B cells play in the development of COPD remain unanswered. For example, which antigens drive the B cell response? Is the response specific to the lung or not? If it were lung specific, then this might account for the lack of relationship to co-morbid systemic diseases linked to immune activation. Commonly hypothesised antigen sources are microbes colonising the airways, smoke constituents and breakdown products of the extracellular matrix [
24]. In the A1ATD cohort we found that chronically colonised patients had significantly higher cFLC levels, supporting the hypothesis that colonisation may be an important driving force behind adaptive immune activation.
Another theory is that infection or colonisation with bacteria leads to a breakdown in self – tolerance promoting an immune reacton to self-antigens. This concept is well established in a number of autoimmune diseases [
25] and there is some evidence supporting an autoimmune element to COPD [
2]. The difference in cFLC observed between usual COPD and A1ATD imply that this is a more important pathogenic theme in usual COPD, although this does not exclude immune activation contributing to the disease process in A1ATD. This result is contrary to the recent report of equivalent levels of lymphoid follicles in lung tissue from a small cohort of A1ATD patients with very severe lung disease, compared to usual COPD [
10]. It is possible that immune activation represents a feature of advanced disease in both conditions, as most of our patients had severe disease, thus further studies are indicated.
Mechanistically cFLCs have biological properties that could potentially damage lung tissue through interaction with neutrophils [
7‐
9]. We have shown previously that migratory accuracy of neutrophils is lower in COPD than A1ATD [
26]; it is possible that cFLCs might be partly responsible since cFLCs were significantly higher in COPD in our study. Furthermore several case reports detail nodular and cystic lung disease associated with cFLC overproduction in light chain deposition disease (LCDD) [
27‐
30], which is characterised by the deposition of non-amyloid κ or λ light chains, and presents with progressive cystic lung disease ultimately leading to respiratory failure necessitating lung transplantation [
29]. It is possible to therapeutically antagonise cFLCs using the compound F991 in animals [
6,
7,
31]. Thus it remains important to clarify whether the pro-inflammatory effects of cFLCs play a role in COPD, and thus represent a suitable drug target.
Our results also demonstrated that patients with chronic bronchitis had significantly higher cFLCs, suggestive of a greater adaptive immune response in these individuals. Chronic bronchitis is associated with more rapid FEV
1 decline [
32], increased exacerbation frequency [
33] and a greater risk of mortality [
34]. However, the difference between cFLCs in patients with chronic bronchitis compared to those without was relatively small, thus the result must be interpreted with caution in terms of clinical significance. The sex difference in cFLCs in usual COPD was unexpected. There is known to be a male predominance in haematological conditions associated with monoclonal overproduction of FLCs such as monoclonal gammopathy of unknown significance (MGUS) and multiple myeloma [
35] however in our study, any patients with an abnormal κ/λ were exluded from the analysis, hence undetected gammopathies would not have influenced our results.
Finally, we have shown that raised circulating cFLC levels are a predictor of mortality, independent of age and severity of renal impairment. Several studies have shown a link between immune system activity, inflammation and risk of death: an increase in polyclonal cFLCs predicted mortality in the general population [
17] and cFLC >65 mg/L were a risk factor for death within 100 days [
19]. The association between inflammation and cardiovascular death is well reported [
36], and 41 % of the deaths in those with cFLC > 65 mg/L were from cardiovascular disease [
17]. A recent systematic review supported the concept that the relationship between cardiovascular disease and COPD goes beyond common aetiological factors such as smoking [
37]; cFLCs could partly explain this.
Our study is limited by the relative lack of colonisation data in A1ATD and absence of this information in usual COPD. Our A1ATD cohort is similar in disease severity to the American A1ATD registry [
38], thus results are likely to be generalisable to other A1ATD populations. However the usual COPD group generally had severe COPD, and exhibited high prevalence of emphysema, thus our results may be less generalisable to milder usual COPD cohorts. The severity of their disease is highlighted by the significant differences relative to A1ATD, who in many cases appear less unwell. This is in part due to the inclusion of family screened, non-index cases in A1ATD, but not usual COPD. It would also be of interest to measure cFLCs in an adequate number of exacerbations of COPD, to see if flare ups of disease relate to cFLC levels, as they do in some autoimmune diseases. Sample collection for this is now ongoing.
Acknowledgments
The authors would like to thank all staff involved in collection of patient data at Queen Elizabeth and Heartlands hospitals, Anne Burmeister, Lakhvir Assi and all staff at the Binding site (Birmingham, UK) for performing the cFLC assays.
Some data has previously been presented in abstract form: Thorax (2013) 68: A34, The Lancet (2014) 383:S28.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JAH and AMT were involved in the acquisition of the data. JAH performed data analysis and drafted the manuscript. AMT and RAS supervised the clinical studies and helped to draft the manuscript. All authors read and approved the final manuscript.