**4.1 Who should be tested?**

196 Chronic Obstructive Pulmonary Disease – Current Concepts and Practice

demonstrated that MZ heterozygosity was associated with airflow obstruction in two large populations (Sorheim et al. 2010). In general, studies comparing COPD cases with healthy controls have found an excess of MZ individuals among COPD cases, but many studies comparing FEV1 (% predicted) in MZ and MM subjects from population-based samples have not found significant differences. There are several factors that undermine many of these conclusions. Firstly, a lack of correction for active and passive cigarette smoking exists in many studies. Secondly, the use of spirometry in defining lung disease is flawed, as subjects with normal spirometry values can have evidence of emphysema on high-resolution CT (Spaggiari et al. 2005). However, it is clear that the weight of evidence is now in favour of a risk of COPD in MZ individuals, but explanations of this risk are limited to the decreased antiprotease levels in subjects, and have not taken into account alternative mechanisms such as Z polymer generation. In the setting of disease management, an individual with smoking-related COPD who is informed of a diagnosis of MZ AATD may be further motivated towards smoking cessation (Carpenter et al.

Early studies that examined the relatively common MS phenotype found no increase in COPD risk but an increase in bronchial hyperreactivity among MS individuals (Townley et al. 1990) but this was not replicated in a larger study (Miravitlles et al. 2002). In a recent meta-analysis of case-control and cross-sectional studies examining COPD risk in MS individuals a small but significantly increased risk was found (Dahl et al. 2005). However, after correction for smoking the MS phenotype was not associated with elevated risk for COPD. Moreover, studies that measured pulmonary function did not find a difference between MS and MM individuals. Another study did attempt to address the risk of COPD in SS individuals but found no increased risk of obstructive lung disease and was limited by small sample size (McGee et al. 2010). The effect of the I and F mutations on the AAT molecule has been described earlier, but to date any mention of COPD risk associated with these mutations is limited to case reports describing compound heterozygotes (Kelly et al.

It is well established that ZZ individuals have a high risk of developing COPD. However, MZ and SZ individuals also have significantly reduced levels of AAT, and are at risk of developing COPD. Anecdotally, a significant number of MZ and SZ individuals (both smokers and non-smokers) from our AATD clinic have severe COPD at a young age. The risk of COPD in heterozygotes has traditionally been explained by the weakening of the antiprotease shield in the lung. However, we know the Z mutation confers harmful gain of function properties on the AAT protein. While the ZZ genotype is relatively well-studied, there is little information regarding MZ, SZ, and other less common genotypes and to date there have been no investigations into the functional consequences of AATD heterozygosity. This is a vital clinical and public health question, as there are predicted to be over 6.6 million MZ and 230,000 SZ individuals in the US alone (de Serres, Blanco, and Fernandez-Bustillo 2010) and the total economic costs of COPD in the US were estimated to be almost \$50 billion in 2010 (National Heart 2009). From a basic research perspective, a careful

**3.6 Other AATD phenotypes as genetic risk factors for COPD** 

2007).

1989; Baur and Bencze 1987).

**3.7 Conclusion** 

Guidelines issued by both the World Health Organisation and the American Thoracic Society/European Respiratory Society (ATS/ERS) recommend the establishment of targeted screening programmes for the detection of patients with AATD (Alpha 1 antitrypsin deficiency: memorandum from a WHO meeting 1997; American Thoracic Society/European Respiratory Society statement: standards for the diagnosis and management of individuals with alpha-1 antitrypsin deficiency. 2003). The biggest problem in the area of AATD is under-diagnosis with most cases misdiagnosed as COPD or non-responsive asthma. As a result, long delays between presentation of first symptoms and correct diagnosis are commonplace and prevent optimal management of the condition, despite education and awareness efforts (Stoller et al. 2005; Campos et al. 2005; Kohnlein, Janciauskiene, and Welte 2010). Compared to population-based studies, which are difficult and expensive to perform on a large scale, targeted detection programmes offer a much higher rate of AATD detection, are easier to perform, and are more cost-effective. However, as they target symptomatic groups the possibility of missing asymptomatic individuals remains. For this reason, comprehensive screening of family members of known AATD individuals is crucial as it offers the most realistic prospect of detecting asymptomatic relatives (Hogarth and Rachelefsky 2008). In the Irish targeted detection programme first-degree relatives of not only ZZ, but SZ and MZ individuals are recommended for testing.

Data from several countries suggests that less than 10% of individuals with severe AATD have been recognised clinically (Aboussouan and Stoller 2009), and improving detection rates is the most urgent issue in the coming years. Several barriers to testing for AATD in

#### **ATS/ERS Recommendations for Diagnostic Testing**

Adults with symptomatic emphysema or COPD

Adults with asthma with airflow obstruction that is incompletely reversible after aggressive treatment with bronchodilators

Asymptomatic individuals with persistent obstruction on pulmonary function tests with identifiable risk factors (e.g. cigarette smoking, occupational exposure)

Adults with necrotising panniculitis

Siblings of individuals with AATD

Individuals with unexplained liver disease, including neonates, children, and adults, particularly the elderly

Table 2. ATS/ERS recommendations for diagnostic testing for AATD (type A recommendations).

Alpha-1 Antitrypsin Deficiency – A Genetic Risk Factor for COPD 199

available (Miravitlles et al. 2010). The traditional gold standard for the diagnosis of AATD has been phenotype analysis by isoelectric focusing but there has been a move in the last few years to a combination of genotyping and quantification (Snyder et al. 2006). Ultimately, while there are pros and cons to both methods of sampling, the decision to collect whole

The World Health Organization has recommended that AAT levels should be measured at least once in all COPD patients and this position was supported by the American Thoracic Society (ATS) and the European Respiratory Society (ERS). The most important consideration when quantifying AAT is the fact that, as an acute phase reactant, AAT can be markedly elevated during infection and inflammation. This is especially relevant if testing COPD patients during an exacerbation. While AAT levels in ZZ individuals are so low that any increase is marginal, circulating AAT levels in heterozygotes (both MZ and SZ) can be "falsely" elevated to levels similar to those observed in MM individuals (Fig. 8). A pronounced acute phase response in many individuals is observed in the MM, MS, and MZ groups; however, the acute phase response in SS, SZ, and ZZ groups is blunted. For this reason, quantification of AAT is no substitute for phenotype or genotype analysis, which is

Fig. 8. AAT levels among the various phenotype classes identified in the Irish National

AAT levels are routinely measured by immunoassay techniques such as nephelometry and turbidimetry, or less commonly by radial immunodiffusion (RID) (Viedma et al. 1986). These methods are based on the use of a specific antibody which binds the AAT in a serum sample. Discrepancies can exist when comparing these methods for serum AAT quantification. For example, nephelometric methods can overestimate AAT concentrations due to haemoglobin or lipid interference, while RID-based methods have been shown to overestimate AAT concentrations by as much as 35 – 40% (Brantly et al. 1991) and are less precise than nephelometric methods with higher coefficients of variation (Alexander 1980). Moreover, the lower sensitivity inherent to the RID method because of the high lower limit of detection (0.33 g/L) becomes a factor when testing severely deficient ZZ individuals, as

**"Protective threshold"**

blood or DBS is often guided by cost considerations.

**4.2 Quantification of AAT** 

not influenced by the acute phase.

AATD Targeted Detection Programme (TDP).

the majority have AAT concentrations < 0.33 g/L.

the COPD population exist, including a fear of genetic discrimination, financial concerns, and privacy concerns (Stoller et al. 2007). Fears of genetic discrimination have been allayed in recent years with preventative legislation enacted in several countries, including Ireland and the US. Barriers to testing among physicians include lack of awareness and knowledge of AATD, lack of access to testing methods, and testing fatigue among physicians who do not encounter AATD initially and give up testing. An element of therapeutic nihilism can also exist, with the mistaken belief that identifying AATD in a COPD patient offers no immediate clinical benefit. Initiatives to increase detection rates might include automatic physician alerts suggesting AATD testing on pulmonary function test reports of patients with fixed airflow obstruction (Rahaghi et al. 2009), better medical and patient education in the area of AATD (Fromer 2010), and a red flag to recommend testing for AATD on laboratory reports of patient with low levels of AAT. The strategy of electronic prompting offers the greatest potential, and has been trialled in several regional hospital laboratories in Ireland to great effect.

Fig. 7. Diagnostic algorithm for testing of whole blood utilised in the Irish National AATD Targeted Detection Programme (TDP). All IEF results are correlated to the quantification of serum AAT.

The advent of finger-prick tests using dried blood spots (DBS) as a source of DNA has allowed home testing for AATD, with easier transportation of samples to the laboratory (Costa et al. 2000). This method of testing eliminates the fear of needles for the patient, and is also cheaper as the test does not require a visit to a general practitioner. Identification of a deficient variant should be confirmed with serum or plasma AAT quantification, as genotyping of DBS sample can miss rarer alleles such as Mmalton (Rodriguez-Frias et al. 2011). For this reason, finger-prick kits are used in the Irish detection programme only for screening family members of index cases who possess Z or S alleles, with whole blood preferred as this allows identification of rare AAT variants. Several laboratories have developed methods of quantifying AAT from DBS which enhance the diagnostic options available (Miravitlles et al. 2010). The traditional gold standard for the diagnosis of AATD has been phenotype analysis by isoelectric focusing but there has been a move in the last few years to a combination of genotyping and quantification (Snyder et al. 2006). Ultimately, while there are pros and cons to both methods of sampling, the decision to collect whole blood or DBS is often guided by cost considerations.

#### **4.2 Quantification of AAT**

198 Chronic Obstructive Pulmonary Disease – Current Concepts and Practice

the COPD population exist, including a fear of genetic discrimination, financial concerns, and privacy concerns (Stoller et al. 2007). Fears of genetic discrimination have been allayed in recent years with preventative legislation enacted in several countries, including Ireland and the US. Barriers to testing among physicians include lack of awareness and knowledge of AATD, lack of access to testing methods, and testing fatigue among physicians who do not encounter AATD initially and give up testing. An element of therapeutic nihilism can also exist, with the mistaken belief that identifying AATD in a COPD patient offers no immediate clinical benefit. Initiatives to increase detection rates might include automatic physician alerts suggesting AATD testing on pulmonary function test reports of patients with fixed airflow obstruction (Rahaghi et al. 2009), better medical and patient education in the area of AATD (Fromer 2010), and a red flag to recommend testing for AATD on laboratory reports of patient with low levels of AAT. The strategy of electronic prompting offers the greatest potential, and has been trialled in several regional hospital laboratories in

**Venous Blood Collection**

**AAT Concentration: Nephelometry Phenotype: Isoelectric focusing**

**variant**

Fig. 7. Diagnostic algorithm for testing of whole blood utilised in the Irish National AATD Targeted Detection Programme (TDP). All IEF results are correlated to the quantification of

The advent of finger-prick tests using dried blood spots (DBS) as a source of DNA has allowed home testing for AATD, with easier transportation of samples to the laboratory (Costa et al. 2000). This method of testing eliminates the fear of needles for the patient, and is also cheaper as the test does not require a visit to a general practitioner. Identification of a deficient variant should be confirmed with serum or plasma AAT quantification, as genotyping of DBS sample can miss rarer alleles such as Mmalton (Rodriguez-Frias et al. 2011). For this reason, finger-prick kits are used in the Irish detection programme only for screening family members of index cases who possess Z or S alleles, with whole blood preferred as this allows identification of rare AAT variants. Several laboratories have developed methods of quantifying AAT from DBS which enhance the diagnostic options

**Phenotype: MM Phenotype: known** 

**Normal Deficient**

**Rare mutation?**

**Sequence to identify rare variant**

**Phenotype: unknown variant**

Ireland to great effect.

serum AAT.

The World Health Organization has recommended that AAT levels should be measured at least once in all COPD patients and this position was supported by the American Thoracic Society (ATS) and the European Respiratory Society (ERS). The most important consideration when quantifying AAT is the fact that, as an acute phase reactant, AAT can be markedly elevated during infection and inflammation. This is especially relevant if testing COPD patients during an exacerbation. While AAT levels in ZZ individuals are so low that any increase is marginal, circulating AAT levels in heterozygotes (both MZ and SZ) can be "falsely" elevated to levels similar to those observed in MM individuals (Fig. 8). A pronounced acute phase response in many individuals is observed in the MM, MS, and MZ groups; however, the acute phase response in SS, SZ, and ZZ groups is blunted. For this reason, quantification of AAT is no substitute for phenotype or genotype analysis, which is not influenced by the acute phase.

Fig. 8. AAT levels among the various phenotype classes identified in the Irish National AATD Targeted Detection Programme (TDP).

AAT levels are routinely measured by immunoassay techniques such as nephelometry and turbidimetry, or less commonly by radial immunodiffusion (RID) (Viedma et al. 1986). These methods are based on the use of a specific antibody which binds the AAT in a serum sample. Discrepancies can exist when comparing these methods for serum AAT quantification. For example, nephelometric methods can overestimate AAT concentrations due to haemoglobin or lipid interference, while RID-based methods have been shown to overestimate AAT concentrations by as much as 35 – 40% (Brantly et al. 1991) and are less precise than nephelometric methods with higher coefficients of variation (Alexander 1980). Moreover, the lower sensitivity inherent to the RID method because of the high lower limit of detection (0.33 g/L) becomes a factor when testing severely deficient ZZ individuals, as the majority have AAT concentrations < 0.33 g/L.

Alpha-1 Antitrypsin Deficiency – A Genetic Risk Factor for COPD 201

Hydragel 18 A1AT isofocusing kit (Zerimech et al. 2008). This isoelectric focusing (IEF) method on agarose gel has an added immunofixation step which utilises a specific antibody to AAT, rendering it superior to traditional IEF techniques, and improving the resolution and reproducibility. The AAT phenotype is determined by visual inspection by at least two independent observers and by comparison with reference standards. There are some unusual cases where a phenotype analysis will not lead to correct diagnosis. M/Null, S/Null, and Z/Null individuals will appear as MM, SS, ZZ phenotypes, respectively, although genotyping will not detect Null mutations either. Another example is of a ZZ individual who has received a liver transplant (Hackbarth et al. 2010). This would result in an MM phenotype on serum analysis by IEF. The transplant recipient would have normal AAT levels as a result of the donor liver (presumably MM) but any offspring would still

**MM MZ ZZ MS**

inherit the Z mutation.

(right).

**4.4 Genotyping of AAT** 

Schmitz 1999; Bartels et al. 2009).

**ZZ MZ MM MZ MS**

study is required, and rare AAT mutations are not detected.

Fig. 9. Technique of isoelectric focusing by polyacrylamide gel electrophoresis (left)

compared to isoelectric focusing by agarose gel electrophoresis followed by immunofixation

A novel method for the simultaneous quantification of AAT and identification of the Z and S mutations in a single sample has been recently described which uses liquid chromatography/tandem mass spectrometry (Chen et al. 2011). This has the potential to combine quantification and phenotyping in a single step but a feasibility and cost analysis

Genotyping assays are typically performed using PCR-based restriction fragment length polymorphism (RFLP) analysis or by melting curve analysis on real-time PCR instruments with specific primers and probes designed for the Z and S mutations (Ferrarotti et al. 2004; Rodriguez et al. 2002). RFLP methods, although cheaper, are time-consuming and have been superseded by the faster and more efficient melting curve methods (Aslanidis, Nauck, and

The advantage of genotyping assays is that they allow the rapid screening of DNA collected by dried blood spots and are not as prone to errors in interpretation as the IEF method. The inherent limitation of the genotyping assay is that most laboratories include only primers for the deficient mutation. This means that in an assay for the Z mutation, a rare mutation will


Table 3. AAT phenotypes and serum AAT concentrations analysed as part of the Irish national targeted detection programme (TDP). Some unusually low AAT concentrations in the MM, MS, and MZ groups were measured in patients with chronic liver disease. Data presented as mean AAT (g/L) +/- standard error of the mean (SEM).

An important consideration in optimising screening programmes for AATD is what cut-off level of AAT is suitable to adopt, below which samples should be investigated further by phenotype or genotype analysis. The choice of cut-off level has obvious financial implications and can cause unnecessary anxiety in a patient if inappropriate follow-up testing is performed. Moreover, cut-off values do not apply when testing family members of known AATD individuals or when testing paediatric cases of liver disease. Phenotype or genotype analysis should be performed in these cases regardless of AAT concentration. This can depend on whether the screening programme aims to detect severely deficient individuals (for example ZZ, Z/Null, Z/Mmalton), or if the aim is to also detect MZ heterozygotes. For example, a cut-off of 1.0 g/L may confidently detect 100% of severely deficient AATD cases, but may miss some SZ cases and large percentage of MZ cases (Kok, Willems, and Drenth 2010). A Swiss-Italian study has shown that MZ individuals with a C-reactive protein (CRP) level of > 0.8 g/L had higher mean AAT concentrations than MZ individuals with lower CRP levels, reflecting the acute phase nature of AAT production (Zorzetto et al. 2008). The cautionary note in using CRP levels to correct for systemic inflammation and an acute phase response when measuring AAT is that CRP is also liver-derived. Thus, in patients with chronic liver disease the ability of the liver to produce either AAT or CRP, or both, may be severely impaired. The choice of cut-off for AAT will reflect the goals of the screening programme and depends on the cost of extra testing and budgetary constraints. If financially permitted, we would advocate the phenotypic or genotypic analysis of all COPD patients.

#### **4.3 Phenotyping of AAT**

In the manner of Laurell and Eriksson, it is possible to detect ZZ individuals by careful visual inspection of electrophoretic patterns on routine serum protein electrophoresis, particularly in ZZ subjects where the absence of the alpha-1 globulin band is so striking (Malfait, Gorus, and Sevens 1985). However, this diagnostic method is not guaranteed to detect all ZZ cases, and will not detect SZ and MZ phenotypes with any confidence (Slev et al. 2008). For this reason, the gold standard for the diagnosis of AATD is isoelectric focusing, which is based upon the isoelectric point of the AAT protein and separates the various isoforms of AAT based on their migration in a specific pH gradient. Each isoform migrates to the position within the pH gradient where the overall charge of the molecule is zero. Qualitative detection and characterisation of AAT variants is carried out in our laboratories by isoelectric focusing using the Hydrasys electrophoresis platform (Sebia) and the

**Phenotype N Mean AAT (g/L) AAT range (g/L)** 

Table 3. AAT phenotypes and serum AAT concentrations analysed as part of the Irish national targeted detection programme (TDP). Some unusually low AAT concentrations in the MM, MS, and MZ groups were measured in patients with chronic liver disease. Data

An important consideration in optimising screening programmes for AATD is what cut-off level of AAT is suitable to adopt, below which samples should be investigated further by phenotype or genotype analysis. The choice of cut-off level has obvious financial implications and can cause unnecessary anxiety in a patient if inappropriate follow-up testing is performed. Moreover, cut-off values do not apply when testing family members of known AATD individuals or when testing paediatric cases of liver disease. Phenotype or genotype analysis should be performed in these cases regardless of AAT concentration. This can depend on whether the screening programme aims to detect severely deficient individuals (for example ZZ, Z/Null, Z/Mmalton), or if the aim is to also detect MZ heterozygotes. For example, a cut-off of 1.0 g/L may confidently detect 100% of severely deficient AATD cases, but may miss some SZ cases and large percentage of MZ cases (Kok, Willems, and Drenth 2010). A Swiss-Italian study has shown that MZ individuals with a C-reactive protein (CRP) level of > 0.8 g/L had higher mean AAT concentrations than MZ individuals with lower CRP levels, reflecting the acute phase nature of AAT production (Zorzetto et al. 2008). The cautionary note in using CRP levels to correct for systemic inflammation and an acute phase response when measuring AAT is that CRP is also liver-derived. Thus, in patients with chronic liver disease the ability of the liver to produce either AAT or CRP, or both, may be severely impaired. The choice of cut-off for AAT will reflect the goals of the screening programme and depends on the cost of extra testing and budgetary constraints. If financially permitted, we

**MM** 3621 1.49 +/- 0.01 0.62 – 4.90 **MS** 568 1.21 +/- 0.02 0.62 – 3.82 **MZ** 802 0.89 +/- 0.01 0.50 – 4.08 **SS** 29 0.86 +/- 0.04 0.62 – 1.54 **SZ** 87 0.61 +/- 0.02 0.33 – 1.20 **ZZ** 90 0.24 +/- 0.01 0.11 – 0.52

presented as mean AAT (g/L) +/- standard error of the mean (SEM).

would advocate the phenotypic or genotypic analysis of all COPD patients.

In the manner of Laurell and Eriksson, it is possible to detect ZZ individuals by careful visual inspection of electrophoretic patterns on routine serum protein electrophoresis, particularly in ZZ subjects where the absence of the alpha-1 globulin band is so striking (Malfait, Gorus, and Sevens 1985). However, this diagnostic method is not guaranteed to detect all ZZ cases, and will not detect SZ and MZ phenotypes with any confidence (Slev et al. 2008). For this reason, the gold standard for the diagnosis of AATD is isoelectric focusing, which is based upon the isoelectric point of the AAT protein and separates the various isoforms of AAT based on their migration in a specific pH gradient. Each isoform migrates to the position within the pH gradient where the overall charge of the molecule is zero. Qualitative detection and characterisation of AAT variants is carried out in our laboratories by isoelectric focusing using the Hydrasys electrophoresis platform (Sebia) and the

**4.3 Phenotyping of AAT** 

Hydragel 18 A1AT isofocusing kit (Zerimech et al. 2008). This isoelectric focusing (IEF) method on agarose gel has an added immunofixation step which utilises a specific antibody to AAT, rendering it superior to traditional IEF techniques, and improving the resolution and reproducibility. The AAT phenotype is determined by visual inspection by at least two independent observers and by comparison with reference standards. There are some unusual cases where a phenotype analysis will not lead to correct diagnosis. M/Null, S/Null, and Z/Null individuals will appear as MM, SS, ZZ phenotypes, respectively, although genotyping will not detect Null mutations either. Another example is of a ZZ individual who has received a liver transplant (Hackbarth et al. 2010). This would result in an MM phenotype on serum analysis by IEF. The transplant recipient would have normal AAT levels as a result of the donor liver (presumably MM) but any offspring would still inherit the Z mutation.

Fig. 9. Technique of isoelectric focusing by polyacrylamide gel electrophoresis (left) compared to isoelectric focusing by agarose gel electrophoresis followed by immunofixation (right).

A novel method for the simultaneous quantification of AAT and identification of the Z and S mutations in a single sample has been recently described which uses liquid chromatography/tandem mass spectrometry (Chen et al. 2011). This has the potential to combine quantification and phenotyping in a single step but a feasibility and cost analysis study is required, and rare AAT mutations are not detected.
