Why has it been so difficult to prove the efficacy of alpha-1-antitrypsin replacement therapy? Insights from the study of disease pathogenesis

Abstract

Alpha-1-antitrypsin is the most abundant circulating protease inhibitor. It is mainly produced by the liver and secreted into the circulation where it acts to prevent excessive proteolytic damage in the lungs by the enzyme neutrophil elastase. The most common severe deficiency allele is the Z mutation, which causes the protein to self-associate into ordered polymers. These polymers accumulate within hepatocytes to cause liver damage. The resulting lack of circulating α₁-antitrypsin predisposes the Z homozygote to proteolytic lung damage and emphysema. Other pathways may also contribute to the development of lung disease. In particular, polymers of Z α₁-antitrypsin can form within the lung where they act as a pro-inflammatory stimulus that may exacerbate protease-mediated lung damage. Researchers recognized in the 1980s that plasma α₁-antitrypsin levels could be restored by intravenous infusions of purified human protein. Alpha-1-antitrypsin replacement therapy was introduced in 1987 but subsequent clinical trials have produced conflicting results, and to date there remains no widely accepted clinical evidence of the efficacy of α₁-antitrypsin replacement therapy. This review addresses our current understanding of disease pathogenesis in α₁-antitrypsin deficiency and questions why this treatment in isolation may not be effective. In particular it discusses the possible role of α₁-antitrypsin polymers in exacerbating intrapulmonary inflammation and attenuating the efficacy of α₁-antitrypsin replacement therapy.

Keywords:

• α₁-antitrypsin deficiency
• emphysema
• augmentation therapy

Introduction

Alpha-1-antitrypsin deficiency was first described in 1963 by Laurell and Eriksson¹ who noted the absence of the α₁ band on serum protein electrophoresis in 5 out of 1500 samples. Three of these individuals had developed early onset emphysema. The association with liver disease was reported 6 years later by Sharp and colleagues² and more recently α₁-antitrypsin deficiency has been associated with the development of bronchiectasis,³ asthma,⁴ vasculitis,⁵ and panniculitis.⁶

Alpha-1-antitrypsin is a member of the serine protease inhibitor or serpin superfamily of proteins. It is secreted mainly by hepatocytes but also by lung and gut epithelial cells,^7–9 neutrophils¹⁰ and alveolar macrophages,¹¹ and is present in the plasma at a concentration of 1.5–3.5 g/L (when measured by an immunodiffusion method). Circulating plasma α₁-antitrypsin is a 394-amino-acid, 52 kDa, acute-phase glycoprotein that acts to inhibit the proteolytic enzyme neutrophil elastase. This enzyme is released at sites of inflammation and, if unregulated, causes proteolytic damage to connective tissue. This is particularly important in the lung as it is persistently exposed to inhaled pro-inflammatory stimuli.

Most individuals carry the normal “M” allele of α₁-antitrypsin. However more than 100 different alleles have been identified to date, of which over 30 affect either the amount or the function of the molecule in vivo. Most α₁-antitrypsin variants are named according to their migration during isoelectric focusing, variants A–L running faster and N–Z slower than the normal “M” protein. The most common severe deficiency mutant is the “Z” (Glu³⁴²Lys) allele which is thought to have originated in northern Europe (where the prevalence is 2%–4%)^12,13 and is also seen as the predominant severe mutation in North America, Australia, and New Zealand.¹⁴ Other important disease-causing alleles include Siiyama (Ser⁵³Phe), prevalent in Japan,¹⁵ and Mmalton (Δ⁵²Phe) which is the most common rare deficiency allele seen in Sardinia¹⁶ and which has been reported sporadically in the UK and Canada. Table 1 details the characteristics of recognized deficiency and null¹⁷ alleles.

Table 1 Pathogenic alleles that cause α₁-antitrypsin deficiency

Variant	Mutation	Molecular basis of disease	Clinical features	Epidemiology
Deficiency alleles
I^98	Arg^39Cys	Protein misfolding; able to form heteropolymers. Reduced serum protein	No clear disease association	Disease only reported in compound heterozygotes
King’s^Citation99	Hisp^334Asp	Rapid polymerization in hepatocyte endoplasmic reticulum, delayed secretion	Neonatal jaundice. Presumed high risk of emphysema in homozygote/compound heterozygote	Case report
Mheerlen^Citation100	Pro^369Leu	Retained in the endoplasmic reticulum, none secreted	High risk of emphysema in homozygotes/compound heterozygotes. Unknown liver disease risk	Case report
Mmalton^Citation101	Δ^52Phe (M2 variant)	Intracellular degradation and polymerization; low serum concentration	Well established association with liver disease and emphysema in homozygotes	Most common rare deficiency allele in Sardinia;^Citation16 seen sporadically in the UK and Canada
Mmineral springs^Citation102	Gly^67Glu	Abnormal post-translational biosynthesis but no polymerization; low serum concentration	Emphysema in homozygotes	Unusual as described in a Afro-Caribbean individual in the United States
Mnichinan^Citation103	Δ^52Phe and Gly^148 Arg	Intracellular polymerization in hepatocytes and plasma deficiency	Risk of liver disease and emphysema	Case report (Japanese family with consanguineous origin)
Mpalermo^Citation104	Δ^51Phe	Serum deficiency	High risk of emphysema in homozygotes	Case report
Mprocida^Citation105	Leu^41Pro	Unstable protein structure leading to intracellular degradation; reduced catalytic activity of circulating protein	High risk of emphysema in homozygotes	Case report
Mvall d’hebron^Citation106 (=Mwurzburg)^Citation100	Pro^369Ser	Retained in the endoplasmic reticulum, none secreted	Presumed risk of emphysema in homozygotes/compound heterozygotes; 50% normal serum α1-antitrypsin level in M/vall d’hebron (/wurzburg) heterozygotes	Case reports from Spain and Germany
Mvarallo^Citation107	Δ^41–^51, replaced with 22 bp sequence creating stop codon at 70–71	Unknown intracellular defect	Presumed risk of emphysema in homozygotes/compound heterozygotes; 50% normal serum α1-antitrypsin level in M/Mvarallo heterozygote	Case report
Pittsburgh^Citation108	Met^358 Arg	Function altered to an antithrombin	Fatal bleeding disorder	Case report
Plowell (=QO Cardiff) and Pduarte^Citation109^–^Citation111	Asp^256Val (M1 and M4 alleles respectively)	Intracellular degradation and plasma deficiency	Increased risk of emphysema in Z/QO compound heterozygotes	Case report
S	Glu^264Val	Protein misfolding and reduced secretion; able to form heteropolymers with Z α1-antitrypsin	Emphysema seen in SZ heterozygotes but less severe than in ZZ.^Citation112 Cirrhosis reported in SZ heterozygotes^Citation113	Most common deficiency variant. Carrier frequency: 1:5 Northern Europe 1:30 USA 1:23 Australian Caucasian 1:26 New Zealand Caucasian Rare/non-existent in Asia, Africa and Australian Aboriginals^Citation114
Siiyama^Citation115	Ser^53Phe	Intracellular degradation and polymerization; low serum concentration	Liver disease and emphysema in homozygotes	Rare, but most common deficiency allele in Japan
Wbethesda^Citation116	Ala^336Thr	Intracellular degradation, serum levels 50% normal	Risk of liver disease and emphysema in compound heterozygotes	Case report
Ybarcelona^Citation117	Asp^256Val and Pro^391His	Unknown intracellular defect; very low serum protein	Severe emphysema reported in homozygote	Case report
Z	Glu^342 Lys	Intracellular degradation and polymerization; low serum concentration	Homozygotes: well established association with liver disease and emphysema. MZ heterozygotes may be more susceptible to airflow obstruction^Citation118 and chronic liver disease^Citation119	Commonest severe deficiency variant. Carrier frequency: 1:27 Northern Europe 1:83 USA 1:75 Australian Caucasian 1:46 New Zealand Caucasian Not seen in China, Japan, Korea, Malaysia, Northern and Western Africa^Citation111
Zausburg (=Ztun)^Citation120,Citation121	Glu^342 Lys (M2 variant)	Intracellular degradation and polymerization; low serum concentration	Liver disease and emphysema in homozygotes/compound heterozygotes	Case report
Zwrexham^Citation122	Ser^−19 Leu and Glu^342 Lys (Z mutation)	Poor expression, low serum concentration	Emphysema reported in compound Z/Zwrexham compound heterozygotes. Unclear whether Ser^−19 Leu would cause disease in absence of Z mutation	Case report
Null (QO) alleles
QO Bellingham^Citation123	Lys^217 stop codon	No detectable α1-antitrypsin mRNA	High risk of emphysema in homozygotes/compound heterozygotes	Case report
QO Bolton^Citation124	Δ1bpPro^362 causing stop codon at 373	Truncated protein; intracellular degradation and no secreted protein	High risk of emphysema in homozygotes/compound heterozygotes	Case report
QO Cairo^Citation125	Lys^259 stop codon	Unknown intracellular defect	High risk of emphysema in homozygotes/compound heterozygotes	Case report
QO Clayton^Citation126	Pro^362 insC causing stop codon at 376	Truncated protein; intracellular degradation and no secreted protein	High risk of emphysema in homozygotes/compound heterozygotes	Case report
QO Devon (=QO Newport)^Citation122	Gly^115Ser and Glu^342 Lys (Z mutation)	Intracellular degradation and polymerization; reduced serum concentration	Risk of emphysema and liver disease in compound heterozygotes. Unclear whether Gly^115Ser would cause disease in absence of Z mutation	Case report
QO Granite Falls^Citation127	Δ1bpTyr^160 causing stop codon	No detectable α1-antitrypsin mRNA	Severe emphysema reported in Z compound heterozygote	Case report
QO Hong Kong^Citation128	Δ2bpLeu^318 causing stop codon at 334	Truncated protein; intracellular aggregation (no polymerization), degradation and no secreted protein	High risk of emphysema in homozygotes/compound heterozygotes	Case reports (individuals of Chinese descent)
QO Isola di Procida^Citation129	Δ17 Kb inc. exons II–V	No detectable α1-antitrypsin mRNA	Emphysema reported in Mprocida compound heterozygote	Case report
QO Lisbon^Citation104	Thr^68Ile	Truncated protein; not secreted	High risk of emphysema in homozygotes. 50% normal serum α1-antitrypsin in M/QO Lisbon heterozygotes	Case report
QO Ludwisghafen^Citation130	Ile^92 Asn	Disruption of tertiary structure; intracellular degradation and no detectable serum protein	High risk of emphysema in homozygotes/compound heterozygotes	Case report
QO Mattawa (M1allele)^131/QO Ourém (M3 allele)^132	Leu^353Phe causing stop codon at 376	Truncated protein; misfolding and reduced serum levels	Emphysema reported in homozygotes	Case reports
QO Riedenburg^Citation133	Whole gene deletion	No gene expression	High risk of emphysema in homozygotes/compound heterozygotes	Case report
QO Saarbueken^Citation104	^1158dupC causing stop codon at 376	Truncated protein; not secreted	High risk of emphysema in homozygotes. 50% normal serum α1-antitrypsin in M/QO Saarbueken heterozygotes	Case report
QO Trastevere^Citation134	Try^194 stop codon	Reduced mRNA, degradation of truncated protein; not secreted	Emphysema reported in compound heterozygote	Case report
QO West^Citation135	ΔGly^164 Lys^191	Aberrant mRNA splicing, intracellular degradation and no detectable serum protein	Emphysema reported in compound heterozygote	Case report

Individuals with abnormal alleles have a plasma deficiency due to a lack of secretion of α₁-antitrypsin from hepatocytes.¹⁸ The severity of the plasma deficiency is predictable with the S allele reducing plasma levels to 60% of normal and the Z allele to 10% of normal; thus an SZ compound heterozygote has plasma levels that are 40% of normal. The resulting lack of elastase inhibition contributes to tissue destruction and panlobular emphysema, particularly in the inflamed lungs of smokers who have a significantly reduced life expectancy when compared with never smokers.¹⁹ Indeed α₁-antitrypsin deficiency is the only known genetic cause of emphysema and is found in 1%–2% of all cases of chronic obstructive pulmonary disease (COPD).²⁰

Alpha-1-antitrypsin augmentation therapy was developed to replace the deficient circulating protein and so ameliorate the progression of the associated emphysema.²¹ It is now widely used in many countries for individuals with severe deficiency of circulating α₁-antitrypsin. However no randomized controlled studies have convincingly shown it to be an effective strategy in slowing the progression of lung disease or reducing mortality.²² This may be due to a lack of suitably powered studies although the pathogenesis of this condition is complex which may mean that this approach has only limited efficacy. This review considers the factors that may mitigate against the complete or partial effectiveness of α₁-antitrypsin augmentation therapy.

Pathogenesis of disease: α1-antitrypsin function, processing and polymerization

Normal “M” α₁-antitrypsin is secreted from the liver and acts as a very effective protease inhibitor. It binds neutrophil elastase via a methionine residue at position 358, on the reactive center loop of the protein (Figure 1). After binding, the enzyme is translocated from one end of the protein to the other in association with insertion of the reactive loop into β-sheet A. This forms a covalently linked complex of enzyme and inhibitor that is cleared from the circulation.

Figure 1 (A) Inhibition of neutrophil elastase by α₁-antitrypsin. After docking (left) the neutrophil elastase (grey) is inactivated by movement from the upper to the lower pole of the protein (right). This is associated with insertion of the reactive loop (red) as an extra strand into β-sheet A (green). Reproduced from Lomas et al¹³⁶ with permission. (B) The structure of α₁-antitrypsin is centered on β-sheet A (green) and the mobile reactive center loop (red). Polymer formation results from the Z variant of α₁-antitrypsin (Glu342 Lys at P₁₇; arrowed) or mutations in the shutter domain (blue circle) that open β-sheet A to favor partial loop insertion (step 1) and the formation of an unstable intermediate (M*). The patent β-sheet A can either accept the loop of another molecule (step 2) to form a dimer (D), which then extends into polymers (P). A small proportion of the unstable serpin molecules can accept their own loop (step 3) to form an inactive, thermostable, latent conformation (L). The individual molecules of α₁-antitrypsin within the polymer are colored red, yellow, and blue. Reproduced from Gooptu et al²⁹ with permission.

The Z (Glu342 Lys), Siiyama, and probably many other alleles result in a delay in folding in the secretory pathway of hepatocytes.^23,24 Much of the protein fails to fold and is “timed out” by the folding sensor manosidase I.²⁵ This material is a substrate for endoplasmic reticulum (ER) associated degradation and is destroyed by the proteasome.^26–28 A proportion is folded correctly, trafficked through the endoplasmic reticulum and Golgi apparatus and secreted into the circulation. The remainder is folded to a near-native conformation but does not achieve the native state. The Glu342 Lys mutation results in an expansion of β-sheet A and perturbation of the F helix to form an intermediate that we have termed (M*).²⁹ The reactive center loop of a second α₁-antitrypsin molecule can then bind to this intermediate, forming a dimer that extends to form a polymer.¹⁸ The resultant polymer has an ordered, repeating structure but no anti-neutrophil elastase activity, because the reactive loop that is central to the polymerization process is buried in β-sheet A of another molecule (Figure 1). These polymers are sequestered within the ER of hepatocytes where they form diastase resistant, periodic acid-Schiff stain positive inclusions, which are associated with liver disease (Figure 2).¹⁸ The Z α₁-antitrypsin that is correctly folded and trafficked through the secretory pathways still has the propensity to form polymers. These have been identified in the lung³⁰ and in biopsies from the skin³¹ and kidneys³² from Z α₁-antitrypsin homozygotes.

Figure 2 (A) Electron microscopy (×20,000) of a hepatocyte from a Z homozygote showing a massive inclusion (arrowed) in the endoplasmic reticulum. Reproduced from Lomas et al¹⁸ with permission. (B) The intra-hepatic polymers of mutant Z α₁-antitrypsin have the appearance of beads on a string on electron microscopy. Reproduced from Lomas et al¹³⁷ with permission.

Study of the consequences of abnormally folded protein accumulation in hepatocytes has helped shed light on the mechanism of this gain-of-function toxicity that causes α₁-antitrypsin deficiency associated liver disease. The presence of polymers of α₁-antitrypsin within the ER causes ER stress. This is defined as a state in which unfolded proteins accumulate and aggregate within the ER, perturbing normal ER function. Misfolded protein accumulation triggers the unfolded protein response (UPR); resident ER chaperones involved in protein folding are upregulated and translation of the abnormal protein is downregulated to restore homeostasis.³³ An excess of protein traffic, for example following viral infection of a cell, triggers the ER overload response (EOR), resulting in calcium-dependent NF-κB activation.³⁴

It is striking that polymers of α₁-antitrypsin do not activate the UPR in the absence of a second “hit” such as heat or accumulation of other misfolded proteins; this is likely to be due to the ordered nature of the polymers.³⁵ They do however activate the EOR.^36,37 This results in inflammatory mediator production,³⁸ relative resistance to cell death,^36,39 and an increase in cell proliferation. Though the precise mechanism by which this causes liver disease has not been elucidated, it follows that abnormal cell survival in the setting of a pro-inflammatory environment may lead to both hepatitis and neoplasia.

Clinical features of α1-antitrypsin deficiency related liver disease

Intracellular polymers form inclusions that are associated with neonatal hepatitis, cirrhosis, and hepatocellular carcinoma.^40,41 Cholestatic jaundice affects one in ten neonates with α₁-antitrypsin deficiency, 15% of whom develop juvenile cirrhosis. Liver disease can also become clinically relevant later in life with almost 50% of patients over the age of 50 having histological features consistent with cirrhosis in an autopsy series.⁴¹ Vaccination against hepatitis A and B is recommended as viral hepatitis may predispose to development of chronic liver disease.⁴² It is recommended that alcohol consumption does not exceed 60 g/day but there is currently no proven association between alcohol excess and the development of liver disease in α₁-antitrypsin deficiency.⁴³ Weight control is advisable in view of an association between obesity and cirrhosis in these individuals.⁴⁴ Screening for cirrhosis and hepatocellular carcinoma (HCC) with liver ultrasound is warranted; some clinicians may choose to use serum alpha-fetoprotein as an additional screening tool for HCC. Treatment of hepatic failure follows that for any other condition; liver transplantation is an option and accounted for 1.1% of adult liver transplants in the US in the period 1995–2004. Liver transplantation cures the circulating deficiency of α₁-antitrypsin and 5-year survival rates are excellent at around 83%.⁴⁵

Alpha-1-antitrypsin deficiency and lung disease

Alpha-1-antitrypsin deficiency is classically associated with early-onset, lower zone emphysema. Smoking individuals with severe circulating deficiency of α₁-antitrypsin tend to develop clinical disease in the third or fourth decade with the most common reported symptoms being breathlessness, cough, and wheeze.⁴⁶ Those with a smoking history develop more severe disease at an earlier age than could be explained by smoking alone. There is often a significant delay from symptom onset to diagnosis;⁴⁷ a 2003 survey revealed a time lapse of more than 5 years between first symptom and diagnosis in a cohort of 1851 individuals with α₁-antitrypsin deficiency.⁴⁸ A chest radiograph typically shows hyperinflation and a paucity of vascular markings in the lower zone; pulmonary function testing reveals evidence of airflow obstruction, gas trapping, and impaired gas transfer. The natural history of α₁-antitrypsin deficiency-associated emphysema is highly variable but disease is often progressive, with faster lung function decline seen in ongoing smokers,⁴⁹ those with recurrent exacerbations,⁵⁰ and those with environmental dust exposure.⁵¹

It is apparent from the study of the rare null mutations of α₁-antitrypsin that the lack of circulating protein (ie, loss-of-function) plays a vital role in the development of emphysema. Indeed, null α₁-antitrypsin homozygotes have particularly severe disease.⁵² There is now increasing evidence that α₁-antitrypsin polymers are also important in the development of the panacinar emphysema that is classically seen in this condition. Alpha-1-antitrypsin enters the lung from the circulation by passive diffusion⁵³ and is also produced locally in alveolar macrophages and bronchial and alveolar epithelial cells.^7,8,11 In individuals homozygous for the Z allele, all α₁-antitrypsin has the propensity to form polymers regardless of its source. Indeed, polymers have been detected in bronchoalveolar lavage fluid^54,55 and explanted lung sections from Z homozygotes, where they are seen both around capillaries (consistent with circulating polymers) and epithelial cells (suggesting local synthesis). This local production of polymers is exacerbated by the presence of cigarette smoke.⁵⁶ The most compelling evidence for pulmonary α₁-antitrypsin production of polymers comes from a bronchoalveolar lavage sample from an α₁-antitrypsin ZZ homozygote following liver transplantation which contained polymers; because wildtype M antitrypsin is produced from the transplanted liver this confirms that Z α₁-antitrypsin is produced locally and then forms polymers.⁵⁵

In stark contrast to the anti-inflammatory properties of monomeric M α₁-antitrypsin, Z α₁-antitrypsin polymers are pro-inflammatory, acting as neutrophil chemoattractants. Polymers, largely located in the interstitium, attract neutrophils as they migrate from capillary to alveolus in response to inflammatory mediators induced by cigarette smoke.⁵⁶ Here they cause neutrophil degranulation and release of elastases and other degradative enzymes.⁵⁷ Additionally, the lungs of Z α₁-antitrypsin homozygotes contain increased levels of chemotactic cytokines including interleukin-8 (IL-8) and leukotriene B4 (LTB₄), compared with controls.⁵⁸ This may be a response of alveolar macrophages to uninhibited neutrophil elastase⁵⁹ but may also reflect stress signaling pathways, perhaps including the UPR, induced by intracellular polymers in epithelial cells and alveolar macrophages.

The pro-inflammatory environment in the lung is further amplified by a number of other mechanisms: (i) monomeric Z α₁-antitrypsin is ten-fold less efficient than wild type protein at inhibiting neutrophil elastase,⁶⁰ (ii) oxidation of α₁-antitrypsin by superoxide radicals may not only reduce its efficacy further,⁶¹ it may also stimulate release of IL-8 and monocyte chemoattractant protein-1 from epithelial cells,⁶² (iii) the reduction in intracellular α₁-antitrypsin may lead to loss of inhibition of caspase-3, leading to uncontrolled cellular apoptosis.⁶³ This is in contrast to the relative paucity of apoptosis seen in the liver and perhaps reflects the activation of the UPR in response to the second hit, which is the inflammatory milieu of the lung in Z α₁-antitrypsin homozygotes.

Taken together, a model for the pathogenesis of lung disease can be proposed in the Z α₁-antitrypsin homozygote involving both loss-of-function and gain-of-function components. The lack of functional α₁-antitrypsin (due to reduced secretion, reduced antiprotease activity, and the polymerization of Z α₁-antitrypsin) creates a pro-inflammatory and proteolytic environment. This is exacerbated by the presence of interstitial polymers and cytokines, both of which are chemotactic for neutrophils, and can be further driven by oxidation of α₁-antitrypsin due to cigarette smoke. Finally, a lack of intracellular α₁-antitrypsin prevents inhibition of apoptotic pathways. The resulting inflammation, proteolysis, and cell death lead to the development of panacinar emphysema, the hallmark of α₁-antitrypsin deficiency-related lung disease.

Systemic diseases associated with α1-antitrypsin deficiency

Alpha-1-antitrypsin deficiency is associated with other inflammatory conditions including panniculitis and anti-neutrophil cytoplasmic antibody (ANCA) positive vasculitis. The lesions of panniculitis contain neutrophils that co-localize with polymers, though normal skin of the Z α₁-antitrypsin homozygote also contains polymers.³¹ It remains to be seen whether polymers are simply an incidental finding, or whether they play a role in the pathogenesis of the disease. The repeated observation that intravenous α₁-antitrypsin replacement therapy can have an effect on flare frequency and severity of panniculitis⁶⁴ points to loss-of-function being of greater significance in this manifestation of disease.

Case–control studies have confirmed a link between Z α₁-antitrypsin deficiency and ANCA-associated vasculitis, and polymers can also be seen in the renal biopsies of affected individuals.³² In addition to the low absolute levels of circulating α₁-antitrypsin, it is postulated that polymers may become trapped within glomeruli, promoting neutrophil degranulation and inflammation as they do within the interstitium of the lung.³²

Current treatment strategies

The management of patients with α₁-antitrypsin deficiency related emphysema closely resembles that of patients with “usual” smoking related COPD. Smoking cessation is crucial to prevent progression of disease with the rate of forced expiratory volume in 1 second (FEV₁) decline being significantly less in those who quit successfully.⁴⁹ Other potentially preventative strategies include minimization of respiratory irritant exposure and pneumococcal and influenza vaccination.⁴³ Inhaled corticosteroids are commonly used and theoretically should reduce neutrophilic airway inflammation; small studies have suggested a possible benefit in terms of lung function in some patients^65,66 and inhaled bronchodilators may result in symptomatic benefit despite little objective evidence of a bronchodilator response. Oxygen used in accordance with national guidelines and pulmonary rehabilitation programs may also be useful.

Lung volume reduction surgery (LVRS) is uncommonly used in individuals with α₁-antitrypsin deficiency. Individuals with emphysema of any cause with predominantly basal disease have a worse outcome than those with upper lobe disease.⁶⁷ Alpha-1-antitrypsin deficient patients undergoing LVRS achieve initial improvements in FEV₁ and exercise tolerance comparable with non-deficient patients but the benefits are shorter lived.⁶⁸ It has been suggested that this is a consequence of disease distribution rather than α₁-antitrypsin deficiency per se⁶⁸ but currently the American Thoracic Society/European Respiratory Society guidelines do not recommend this procedure pending the emergence of further evidence.⁴³ There are case reports of successful use of less invasive procedures including endobronchial valve placement⁶⁹ though experience remains limited.

Lung transplantation is also an option for patients with end-stage emphysema. Because COPD often develops at an early age in α₁-antitrypsin-deficient individuals, they tend to be good candidates. Alpha-1-antitrypsin deficiency accounted for 3.2% of lung transplants (and 10% of transplants for emphysema) in the International Society for Heart and Lung Transplantation Registry in 2009,⁷⁰ with a 1-year survival rate of 86% and a 3-year survival of 69% for transplants performed between 2006–2010. These figures are comparable with the general COPD population, 84% of whom are alive at 1 year and 67% at 3 years posttransplant.

Alpha-1 antitrypsin replacement therapy: current practice

The concept of using purified human α₁-antitrypsin as intravenous replacement therapy was first described in 1981 by Gadek and colleagues who demonstrated normalization of serum α₁-antitrypsin levels and establishment of anti-elastase activity within the lower respiratory tract with weekly infusions.²¹ Alpha-1 antitrypsin replacement therapy was subsequently approved by the United States Food and Drug Administration (FDA) in 1987 based on the demonstration that normal serum levels could be achieved with regular treatment.⁷¹ It is also available in Canada and many European countries and replacement therapy is recommended in the joint American Thoracic Society/European Thoracic Society statement on the management of α₁-antitrypsin deficiency.⁴³ Many other countries including the UK, New Zealand, and Australia await proof of clinical benefit before licensing. It is not indicated in individuals who have partial α₁-antitrypsin deficiency (MZ heterozygotes) and in α₁-antitrypsin deficiency-associated liver disease.

Replacement therapy is given as weekly intravenous infusions at a dose of 60 mg/kg. This is based on evidence that with such dosing, trough α₁-antitrypsin levels can be kept above the “protective” threshold of 80 mg/dL. This threshold is based on the observation that patients with heterozygous phenotypes whose levels of α₁-antitrypsin exceed this level do not usually develop lung disease.⁷¹ Four options are available for replacement therapy in the US: Prolastin^® (Talecris Biotherapeutics, Research Triangle Park, NC), Aralast^® (Baxter Healthcare, Deerfield, IL), Zemaira^® (CSL Behring, King of Prussia, PA), and Glassia^® (Baxter, Deerfield, IL), with a fifth, Trypsone^® (Grifols, Barcelona, Spain), available in Spain. All consist of purified human α₁-antitrypsin. The individual characteristics of each have been thoroughly reviewed elsewhere⁷² but they all appear to be equally effective with subtle differences in storage, preparation, infusion rate, and cost. Glassia^®, licensed in October 2009 by the FDA, is the only ready-to-use formulation but requires a slower infusion rate of 0.04 mL/kg/minute, half that of the other preparations. The cost of replacement is dependent on body mass but may be upwards of US$100,000 per annum – and is a lifelong treatment.⁷³

Intravenous α₁-antitrypsin is generally considered to be well tolerated and safe. It is contraindicated in patients with IgA deficiency as purified protein may contain small amounts of IgA, prompting a severe hypersensitivity reaction. As it is purified from human plasma there is a theoretical risk of transmission of infectious agents including viruses and prions, but such transmission has never been reported. The commonest reported side effects from the original preparation (Prolastin^®) are dizziness, fainting, and dyspnea.⁷⁴

Alpha-1-antitrypsin replacement therapy: evidence of efficacy

Evidence of a biochemical effect of weekly intravenous antitrypsin was quick to emerge. Serum α₁-antitrypsin levels can be maintained above the postulated protective level and anti-elastase activity can concurrently be detected in bronchoalveolar lavage fluid.⁷¹ Infused α₁-antitrypsin is functionally active and antibodies do not develop following repeated infusion.⁷⁵ There are inherent difficulties in attempting to establish the clinical efficacy of α₁-antitrypsin replacement therapy. Though the gene frequency makes α₁-antitrypsin deficiency as common as cystic fibrosis, the heterogeneous nature of the disease means that only around 5%–10% of patients have been diagnosed.⁷⁶ This means that the pool of patients on whom to undertake trials is relatively small and inevitably results in the majority of evidence coming from larger observational studies rather than the few randomized controlled trials, each of which contains a small number of participants.⁷⁷ Additionally, the evolution of emphysema is slow, meaning that protracted studies are necessary to show differences in lung function decline and mortality.⁷⁸ Indeed, Schluchter et al estimated that to detect a 40% reduction in mortality in 5 years, 684 α₁-antitrypsin deficient individuals with an FEV₁ of 35%–49% predicted would need to be recruited over a 2-year period.⁷⁹ Furthermore enthusiasm from pharmaceutical companies to fund such trials may be lacking with their products already freely available in many countries.

Non-randomized trials

It was not until 1997 that the first observational study with concurrent controls addressing clinical efficacy of α₁-antitrypsin replacement therapy was published. Seersholm and colleagues demonstrated a significantly slower rate of FEV₁ decline in a cohort of 198 German α₁-antitrypsin deficient patients on replacement therapy when compared with 97 Danish controls, particularly in those with an FEV₁ of 31%–65% predicted.⁸⁰ In 1998 the National Heart Lung and Blood Institute α₁-antitrypsin deficiency registry study group reported on 927 patients enrolled in their registry. They showed no overall difference in lung function decline between treated and untreated groups. However there was a significant benefit of treatment seen in those with moderate impairment of lung function (FEV₁ 30%–64% predicted); conversely those with an FEV₁ > 80% predicted had a faster annual decline in FEV₁ whilst on treatment than untreated individuals (P = 0.09). They also reported overall mortality figures in favor of α₁-antitrypsin replacement, though no difference in those with an FEV₁ > 50% predicted.⁸¹ Wencker et al took a different approach in 2001, comparing decline in lung function in 96 patients before and after initiation of augmentation therapy. These authors observed an overall benefit of treatment, but in contrast to previous observational studies, they demonstrated no significant difference in those with FEV₁ in the range 30%–65%. The most significant benefit was seen in the small subgroup with mildly impaired (> 65% predicted at enrolment) but rapidly declining FEV₁, with annual FEV₁ loss reducing from 255.7 mL/year to 45.8 mL/year on augmentation therapy (P = 0.0016).⁸² Recently Tonelli et al studied 164 individuals from the Alpha-1 Foundation DNA and Tissue Bank, finding a benefit of treatment in ex-smokers with an FEV₁ of less than 50% predicted. It was again noted that those with better lung function (FEV₁ > 60% predicted) did worse on augmentation therapy as measured by FEV₁ decline.⁸³ This finding may in part be explained by a tendency for those with good but rapidly declining lung function to be started on augmentation therapy. The authors of these observational studies all acknowledge the limitations of their findings. Treatment decisions were made by patients’ physicians and the baseline demographics between groups were unmatched; in particular the baseline FEV₁ was lower in the treatment cohort in all studies.

Randomized controlled trials

Alpha-1-antitrypsin replacement therapy had been available in the United States for 12 years before the first of two randomized controlled trials (RCTs) assessing its efficacy was published. In 1999 Dirksen et al⁸⁴ reported on a cohort of 58 ex-smokers with severe plasma α₁-antitrypsin deficiency and an FEV₁ of 30%–80% predicted. The active group was treated with 250 mg/kg iv α₁-antitrypsin at 4-weekly intervals for at least 3 years. This dose had previously been shown to provide protective serum α₁-antitrypsin levels for an average of 25 days of the 28-day interval.⁸⁵ The study’s primary outcome measure was FEV₁; there was no statistically significant difference between groups but a trend toward faster decline of FEV₁ was seen in the treated group. In contrast the authors were able to demonstrate a trend towards slower loss of lung density (measured by computed tomography [CT]) in the treated group, albeit with substantial deterioration in both groups. The study was not designed to address mortality.

In 2009 the results of the “exacerbations and computed tomography scan as lung end-points” (EXACTLE trial),⁸⁶ which used CT-assessed lung density as its primary outcome measure, were published. Participants had severe α₁-antitrypsin deficiency, were ex- or never smokers and members of the treatment group were given 60 mg/kg weekly iv α₁-antitrypsin replacement (the accepted dosing regimen) for a minimum of 2 years. The authors again reported loss of lung density in both groups with a difference of borderline significance in favor of the treatment group over the entire treatment period in one of four analysis methods. No difference was seen in lung function parameters or exacerbation rate between groups, though a post hoc analysis suggested that exacerbation severity may be milder in the treatment group. A third RCT which plans to address the question of mortality, albeit only as a secondary endpoint, is ongoing.⁸⁷

A meta-analysis of some of the early trials comprising 1509 patients concluded that FEV₁ decline was slower in treated versus untreated individuals.^{80–82,84,88} This was particularly evident in those with an FEV₁ of 30%–65% predicted, who experienced a 26% reduction in the rate of decline in FEV₁ as a result of replacement therapy.⁸⁹ The randomized trials were the subject of two meta-analyses that reached quite different conclusions. A Cochrane review concluded that a lack of evidence regarding mortality combined with conflicting evidence for treatment efficacy from FEV₁ and CT data meant that replacement therapy cannot be recommended at present and that further studies should be large enough to detect an effect on mortality.²² A response from the expert community raised concerns over the methodology of this review, particularly regarding differences in drug and dosing regimens, and dropout rates between studies. The authors also questioned whether observational data should have been included in a meta-analysis of this rare condition.⁷⁷

Stockley and colleagues re-analyzed the 1999 RCT data and, combined with the EXACTLE data, demonstrated a significant reduction in lung density decline in the treated versus untreated group (P = 0.006). They agreed that further studies are warranted, but that there is sufficient evidence of treatment efficacy that in future trials an iv placebo group is not ethically warranted.⁹⁰

Exacerbations and airway inflammation

One study asking patients to self-report symptoms after initiation of replacement therapy found that 56 of 89 patients reported a definite decrease in exacerbation number from an average of 3–5 per year to 0–1 per year.⁹¹ Others have noted ongoing exacerbations despite treatment⁹² and the EXACTLE trial did not find a reduction in exacerbation frequency on treatment. Augmentation therapy does however seem to have an effect upon markers of airway inflammation. Stockley et al noted an increase in sputum α₁-antitrypsin to the levels seen in nondeficient individuals with intravenous replacement; this was associated with a decrease in sputum LTB₄ but there was no significant change in markers of neutrophil number and activation: IL-8 and myeloperoxidase (MPO). They suggest that this may reflect decreased LTB₄ secretion by macrophages due to a reduction in uninhibited elastase.⁹³

Table 2 summarizes the trial data discussed above. The very nature of observational studies along with their conflicting results must limit their role in confirming treatment efficacy. RCTs to date have been small scale. They suggest a lack of treatment benefit as measured by lung function, but raise the possibility of a positive impact on the rate of lung density loss as measured by CT, a parameter thought to be a more accurate measure of disease progression than lung function.⁹⁴ The effect of augmentation therapy on mortality has not been adequately addressed. As outlined by Schluchter and colleagues⁷⁹ the scale of trial required would be likely to require international collaboration. With disparate current clinical practice and a lack of agreement on the ethical design of future trials to assess disease progression, it seems this question will remain unanswered.

Table 2 Studies of α₁-antitrypsin replacement therapy

Author(s)	Study type	Primary (secondary) outcome measure(s)	Salient results
Non-randomized studies
Seersholm et al^Citation80	Observational study with concurrent controls	FEV1 decline	– Slower FEV1 decline in treated group Most marked benefit in those with FEV1 31%–65% predicted
NHLBI study^Citation81	Observational study with concurrent controls	FEV1 decline; Mortality	– No overall difference in FEV1 decline Slower FEV1 decline in those with FEV1 30%–64% predicted Faster decline in those with FEV1 > 75% predicted – Decreased mortality in treated group
Wencker et al^Citation82	Observational study of patients pre- and during treatment	FEV1 decline	– Overall slower FEV1 decline on treatment Most significant benefit in those with FEV1 > 65% no significant difference in those with FEV1 30%–65%
Tonelli et al^Citation83	Observational study with concurrent controls	FEV1 decline (mortality)	– Slower decline in treated group if FEV1< 50% predicted and ex-smoker. Faster decline in treated group if FEV1 > 60%. – No difference in mortality
RCTs
Dirksen et al^Citation84	Double-blind placebo controlled RCT	FEV1/DLCO decline, (CT densitometry)	– No difference in FEV1 or DLCO – Trend towards slower annual loss of lung density in treated group
Dirksen et al^Citation86	Double-blind placebo controlled RCT	CT densitometry (lung function, health status, exacerbations)	– Trend towards reduction in lung density loss in treated group – No difference in FEV1 or DLCO – No difference in exacerbation rate but exacerbations less severe in treated group
Meta-analyses
Gøtzsche and Johansen^Citation22	Review of RCTs	Mortality, FEV1 decline	– Reduction in lung density loss in treated group over whole trial period – Trend towards faster FEV1 decline in treated group (P = 0.06)
Stockley et al^Citation90	Integrated analysis of RCTs	CT densitometry, FEV1 decline	– Reduction in lung density loss in treated group (P = 0.006) – Trend towards faster FEV 1 decline in treated group (P = 0.321)
Studies on exacerbations
Lieberman^Citation91	Patient survey	Exacerbation frequency	– Reduction in exacerbation frequency from 3–5 per annum to 0–1 per annum following initiation of treatment
Stockley et al^Citation93	Clinical study using BAL from patients on α1-antitrypsin replacement	Sputum inflammatory markers	– Sputum LTB4 significantly reduced following weekly 60 mg/kg α1-antitrypsin replacement. – Non-significant reductions in IL-8 and MPO

Abbreviations: BAL, bronchoalveolar lavage fluid; D_LCO, diffusing capacity of the lung for carbon monoxide; CT, computed tomography; FEV_1, forced expiratory volume in 1 second; IL-8, interleukin 8; LTB₄, leukotriene B4; MPO, myeloperoxidase; NHLBI, national heart lung and blood institute; RCTs, randomized controlled trials.

Assessing replacement therapy in the context of disease pathogenesis

Practical problems with clinical assessment of intravenous α₁-antitrypsin replacement have been outlined above. Putting this treatment in the context of disease pathogenesis may provide more understanding of the questionable efficacy demonstrated to date. Weekly iv α₁-antitrypsin provides protective levels of circulating protein as measured by a threshold over which individuals do not develop lung disease. However α₁-antitrypsin is an acute phase protein with levels increasing by up to 130% in response to stress and peaking at up to 6 g/L;⁹⁵ perhaps the ability to mount such a response may be of critical importance rather than the baseline level. Alpha-1-antitrypsin secretion may also be under the control of growth hormone working synergistically with other pituitary hormones, which leads to variable hepatic serum and mRNA levels.⁹⁶ Replacement therapy cannot address this; it can be envisaged that a patient would be particularly vulnerable shortly before the next infusion is due, especially the significant number of individuals who receive bi-weekly or monthly treatment.⁸¹ Based on knowledge of disease pathogenesis it follows that α₁-antitrypsin replacement therapy should address some of the drivers of lung disease: (i) it can effectively inhibit neutrophil elastase where the inefficient Z α₁-antitrypsin monomers and inactive polymers would otherwise leave uninhibited elastase activity; (ii) the resulting lack of free elastase reduces LTB₄ release from alveolar macrophages, thereby reducing the chemotactic signaling to neutrophils.⁹³ Despite this, levels of MPO are unchanged in those on replacement suggesting that equal numbers of neutrophils are present (though they will have their elastase inhibited).

However replacement therapy will not affect (i) the formation of Z α₁-antitrypsin polymers that become lodged in the lung interstitium and act as neutrophil chemoattractants; (ii) intracellular Z α₁-antitrypsin polymers in epithelial cells and alveolar macrophages that continue to induce stress signaling and release of pro-inflammatory cytokines.

We can postulate that a patient on augmentation therapy may have less tissue destruction as a result of better regulation of neutrophil elastase activity, but will continue to have chronic inflammation in response to intrapulmonary polymers. In the small minority of patients who are α₁-antitrypsin_null homozygotes then lung damage entirely reflects the absence of neutrophil elastase inhibition. This group is most likely to benefit from replacement therapy. However in the majority of patients who have polymerogenic mutations it may be that the benefit of replacement therapy is attenuated by the ongoing inflammatory response to polymers.

Clinical and laboratory studies designed specifically to address this question would be of enormous value. A trial comparing treatment efficacy in individuals homozygous for null α₁-antitrypsin alleles with those with polymerogenic alleles may shed light on the impact of α₁-antitrypsin polymers in attenuating the beneficial effects of augmentation therapy. This would provide crucial information on the relative role of α₁-antitrypsin polymers in the pathogenesis of emphysema. Additionally if small molecules that block α₁-antitrypsin polymerization⁹⁷ can be further developed, these can be tested in vivo to determine whether the absence of intrapulmonary polymers slows the rate of development of emphysema.

Conclusion

The current literature provides good evidence of the safety and biochemical effect of intravenous α₁-antitrypsin replacement but there is no widely accepted proof that it affects disease progression or mortality. Further large scale RCTs would help clarify the impact of long-term intravenous α₁-antitrypsin replacement on loss of lung density, which is emerging as an important surrogate marker of disease. Combined results from such trials may confirm CT densitometry as the best marker of disease severity and progression in α₁-antitrypsin deficiency as well as adding to a body of evidence that could be used to determine any effect on mortality.

The evolving knowledge of the molecular basis of disease requires us to re-evaluate the utility of replacement therapy in individuals with α₁-antitrypsin deficiency. A logical approach is to turn away from simply replacing deficient protein and toward tackling the key feature of this disease: the tendency of mutant α₁-antitrypsin to polymerize.¹⁸ Strategies that result in the secretion of normally folded, functional protein would impact on both intracellular stress signaling and the systemic inflammation caused by polymers and in turn should prevent the development of both liver and lung disease, although whether this will become a reality remains to be seen.

Acknowledgments

JAD is an MRC Clinical Training Fellow. DAL is supported by the Medical Research Council (UK), the Engineering and Physical Sciences Research Council, GlaxoSmithKline, the Alpha-1 Foundation, and Papworth NHS Trust.

Disclosure

The authors report no conflicts of interest in this work.