Abstract
Alpha-1-antitrypsin (α1-AT) deficiency is mainly evaluated in the diagnostic process of chronic obstructive pulmonary disease (COPD). Around 95% of individuals with severe α1-AT deficiency carry the PI*ZZ genotype. Little is known about the epidemiology of the remaining deficient α1-AT variants, which are called ‘rare’ due to their low prevalence. The retrospective revision of 3511 α1-AT deficiency determinations performed in Barcelona from 1998 to 2010 detected 1.6% of cases with rare α1-AT alleles, a rate similar to those reported in other European studies. Among these variants, PI*I and PI*Mmalton represented 54% of cases. Hence, the so-called ‘rare’ α1-AT alleles may not be rare as has been assumed. It would be of interest to implement simple allele-specific molecular biology methods to study the most prevalent rare variants in each region. Augmentation therapy is recommended in patients with emphysema and PI*ZZ genotype, but there is little evidence regarding the implications of rare variants on therapy.
Keywords
Introduction
Alpha-1-antitrypsin (α1-AT) deficiency, one of the most common inherited disorders worldwide (25 cases per 10,000 White individuals), is mainly evaluated in the differential diagnosis of chronic obstructive pulmonary disease (COPD) or chronic liver disease, the two major conditions associated with this deficiency in adults [Janciauskiene et al. 2011; Stoller and Abboussan, 2005]. α1-AT is a 394-amino-acid plasma protein, which, despite its small size, shows surprisingly high polymorphism, with more than 120 allelic variants, at least 60 of them associated with deficient α1-AT levels [Janciauskiene et al. 2011]. The most common normal α1-AT allele is the M allele and the most frequent deficient variants are Z (G/A, Glu342Lys, in exon V) and S (A/T, Glu264Val, in exon III). More than 95% of individuals with severe deficiency carry the PI*ZZ genotype [Janciauskiene et al. 2011; Stoller and Abboussan, 2005]. Little is known about the epidemiology of the remaining deficient α1-AT variants, some of which are associated with a significant reduction or absence of plasma α1-AT (e.g. null alleles which do not express the protein). Owing to their low frequency, these non-S and non-Z deficient variants are called ‘rare’ and most of them are difficult or even impossible to characterize using the isoelectrofocusing (IEF) method for α1-AT phenotyping or the usual allele-specific genotyping assay included in large-scale screening programs for detecting P*IS and PI*Z alleles [Molina et al. 2011]. This may have contributed to misclassification of many of these cases, which would misrepresent their true frequency. Most of these alleles can only be detected by molecular biology techniques, such as genome sequencing, which is not available in all routine laboratories [Miravitlles et al. 2010].
The only specific treatment for patients with this disease is weekly intravenous infusion of human α1-AT obtained from blood donors [Tonelli and Brandly, 2010]. The indication for treatment includes identification of the deficient PI*ZZ or null genotypes [Tirado-Conde et al. 2008]. Identification of rare variants in patients with emphysema will help to establish the need for augmentation therapy in these patients. The aim of this study was to establish the prevalence and type of rare α1-AT alleles in a Spanish population by retrospective review of all α1-AT deficiency determinations performed in the clinical laboratory of Hospital Universitari Vall d’Hebron (Barcelona) over the last 12 years.
Methods
We performed a retrospective revision of the results obtained in all α1-AT deficiency studies, carried out in our laboratory by IEF phenotyping and/or genome sequencing of α1-AT exons II to V in samples with low plasma levels of α1-AT. Our laboratory in Hospital Vall d’Hebron was the central laboratory of the Spanish α1-AT Deficiency Registry during the study period [Miravitlles et al. 1998; Lara et al. 2011] and samples were referred from all over the country for determination of plasma levels, phenotyping, and genotyping, and in particular, for studying all discordant α1-AT deficiency cases (low serum concentrations, but no S or Z alleles identified) by sequencing the entire coding region of the α1-AT gene.
The methods for phenotyping in serum and for dried blood spot sample processing has been extensively described [Costa et al. 2000]. The protocol for genome sequencing has also been described in detail [Jardí et al. 1998, 2000]. We considered rare deficient variants, other than S or Z, to be those associated with serum levels below the normal range (i.e. 120 mg/dl), either in heterozygote or in homozygote combination. Results are presented as the percentage of rare variants over the total of samples analyzed in the period of study.
All studies performed in samples referred from outside the hospital catchment area included AAT sequencing and genotyping, because all corresponded to samples with discordant results from previous analyses performed in the hospitals where the patients consulted (Rest of Spain column in Table 1). The studies corresponding to our hospital samples (Barcelona column in Table 1) were processed under the following algorithm: AAT levels were determined, and in cases with AAT <120 mg/dl, phenotyping by IEF was performed. If AAT values were in accordance with the range for the observed phenotype, laboratory results were considered definitive. However, if AAT levels did not correspond to the phenotype and an EDTA blood sample was available, AAT genotype was determined by II to V exon sequencing. When no sample was available, a new sample was requested from the attending physician. In three patients with severe AAT deficiency (AAT <30 mg/dL), this additional sample could not be obtained and therefore, genotyping was not performed (Not genotyped in Table 1).
Alpha-1-antitrypsin (α1-AT) variants observed in the current study, comparison of rare alleles with those reported in other studies.
COPD, chronic obstructive pulmonary disease; DBS, dried blood spot, RFLP, random fragment length polymorphism; Seq. (II–V), sequencing of α1-AT exons II to V.
‘not genotyped’: no EDTA blood sample available.
Ferrarotti et al. [2005]; bZorzetto et al. [2008]; cDenden et al. [2009]; dBals et al. [2007]; eFerrarotti et al. [2008].
AATD studies with discordant AAT values/phenotyping from the rest of Spain
Results
A total of 3511 α1-AT deficiency studies in samples with serum α1-AT levels <120 mg/dl were performed in our clinical laboratory from 1998 to 2010. One hundred and eight studies corresponded to samples referred from outside the hospital catchment area. We found rare α1-AT alleles in 56 (1.6%) cases from our hospital and in 21 (19.4%) samples from the rest of Spain. These rare alleles include the PI*Ybarcelona variant (Pro391His) in seven individuals from two families [Miravitlles et al. 2003] and a severe α1-AT deficiency (<30 mg/dl) in nine cases from two other families. There were no mutations in the DNA coding sequence in six of these cases, and an EDTA blood sample was not available for genotyping in the remaining three cases. Among the α1-AT null alleles, we detected PI* Q0clayton and Q0bellingham in six individuals from two families [Rodriguez-Frias et al. 2011]. There was a high presence of the deficient alleles, PI*I (Arg39Ser) and PI*Mmalton (Phe52 deletion), in 26 (34%) and 15 cases (20%), respectively. In Table 1, the results of our review (Spain current study column) are compared with the results obtained in five different studies performed in Italy [Ferrarotti et al. 2005], Switzerland [Zorzetto et al. 2008], Tunisia [Denden et al. 2009], Germany [Bals et al. 2007], and Italy from 1996 to 2004 [Ferrarotti et al. 2008].
Discussion
During our years of activity as the central laboratory for the Spanish Registry, we detected a significant percentage of rare α1-AT alleles in our population (1.6%). The prevalence of these alleles falls in the middle of the range (0.6–4.2%) observed in previous studies carried out in nearby regions [Denden et al. 2009; Ferrarotti et al. 2005, 2008; Zorzetto et al. 2008; Bals et al. 2007], and is particularly similar to that reported in Italy (1.2%). These rare alleles include the variants PI*Ybarcelona [Jardí et al. 1998; Miravitlles et al. 2003] and Mvall d’hebron [Jardi et al. 2000], which seem to be autochthonous to our area, and others that have been previously described in other geographic areas [Rodriguez-Frias et al. 2011]. There was a remarkably high frequency of the deficient alleles, PI*I and PI*Mmalton, which, taken together, accounted for 54% of all α1-AT rare alleles in our region. Interestingly, in a recent study performed in Ireland [Carroll et al. 2011], in which the total frequency of rare AAT variants was 0.7%, PI*I variants accounted for 90% of the total. This observation, together with the similar frequency of PI*S (0.05) in Spain and Ireland, suggests ethnic relationships between the two countries [Carroll et al. 2011].
In a study performed in Tunisia, rare alleles were detected in 4.2% of all cases included, whereas only one case of the major PI*Z allele (as heterozygous PI*MZ) was observed [Denden et al. 2009]. Rare alleles are very common in Central and Southern Italy, where the PI*Mmalton variant (particularly in Sardinia) and the PI*Mprocida (Leu41Pro) variant are more prevalent than PI*Z [Ferrarotti et al. 2005]. In Switzerland, a high prevalence (2.8%) of rare α1-AT alleles has been reported in patients with α1-AT deficiency; PI*Mwurtzburg (Pro369Ser) is the most common, present in 13% of patients with rare alleles [Zorzetto et al. 2008]. In this last study, 20% of rare α1-AT genotypes corresponded to four previously uncharacterized variants, such as allele Lys259Ile in four cases (10% of rare alleles). α1-AT deficiency detection programs performed with dried blood spot (DBS) technology are based on detection of the major P*IS and PI*Z alleles [Molina et al. 2011]. However, they can be interpreted as indirect evidence of the prevalence of rare α1-AT alleles. In this sense, two DBS studies performed in Italy [Ferrarotti et al. 2008] and Germany [Bals et al. 2007] reported significant percentages of non-S and non-Z α1-AT-deficient alleles (4% and 0.6%, respectively). In relation to these last data, in a DBS screening program developed by the Spanish α1-AT Deficiency Registry in 2005, 3.7% of α1-AT-deficient samples did not show the PI*S or PI*Z alleles [de la Roza et al. 2005], a percentage similar to that observed in Italy [Ferrarotti et al. 2008]. These data suggest that rare variants may be more common in areas where the main deficient variants (e.g. PI*Z) are less prevalent [Stolk et al. 2006], such as Japan, where the most common variant is deficient PI*Siiyama (Ser53Phe), present in most α1-AT-deficient patients [Seyama et al. 1995]. The low prevalence of rare variants recently reported in Ireland (0.7%), where the PI*S and PI*Z major AAT variants showed frequencies of 0.05 and 0.02, respectively [Carroll et al. 2011], similar to the highest found in European countries, seems to reinforce this hypothesis of an inverse relationship between the more frequent PI*S/PI*Z variants and rare AAT variants.
With regard to ‘the most frequent among the low frequency’ α1-AT alleles, PI*I and PI*Mmalton were the most prevalent in our study. The PI*I allele was the most widespread (34% of our Spanish population, compared to 28% in Switzerland and 5% in Italy), and PI*Mmalton was also very common (20% in our study, 60% in Tunisia, 8% in Switzerland, and 35% in Italy). It must be kept in mind that the PI*Malton and PI*I phenotypes are both difficult to interpret without molecular biology studies [Jardí et al. 1997]. The phenotype of the PI*Mmalton allele results in an IEF pattern identical to that seen in the normal PiM variant, and the phenotype of PI*I can be easily confused with the IEF pattern of the PI*F allele (Arg223Cys) because they share the same amino acid substitution (Arg to Cys). The methodological difficulties associated with routine α1-AT deficiency assays are a common problem in the study of rare alleles [Rodriguez-Frias et al. 2011].
Owing to the considerable diversity and prevalence of α1-AT-deficient variants, there is some speculation that their biological selection may be a protective mechanism against infectious agents in the pre-antibiotic era because of their ability to amplify the inflammatory response [Lomas, 2006]. Thus, in regions where the majority of α1-AT-deficient variants (such PI*Z) are extremely uncommon, rare α1-AT alleles may have been selected to compensate for the absence of these major alleles and fulfill this suggested protective role.
The only specific treatment available for patients with pulmonary emphysema associated with α1-AT deficiency is intravenous augmentation therapy with a preparation of the human protein administered in weekly infusions [Tonelli and Brantly, 2010]. Scientific societies have developed strict criteria for initiating this type of therapy, which include demonstration of a deficient genotype [Tirado-Conde et al. 2008]. Determination of the frequency and clinical impact of rare deficient phenotypes will help to establish the requirements for augmentation therapy in emphysema patients carrying these variants.
In conclusion, although the prevalence of rare variants reported in the various related studies are not directly comparable because of the different clinical and methodological characteristics of the research performed, our results suggest that these so-called ‘rare’ α1-AT alleles are not as rare as has been assumed. For this reason, and taking into account the limitations of routine IEF assay for their study [Rodriguez-Frias et al. 2011], it would be of interest to implement simple allele-specific molecular biology methods, such as real-time PCR melting curves, currently used in α1-AT screening programs, to study the most prevalent rare variants in each region. This could be done with PI*Mmalton in Mediterranean countries or with PI*I in the Spanish and Irish populations. The results presented here indicate the strong need to develop standardized studies to obtain conclusive data on the prevalence of rare AAT variants.
Footnotes
Acknowledgements
The authors thank the members of the Spanish α1-AT Deficiency Registry for providing the external cases included in this study.
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
The authors declare no conflicts of interest in preparing this article.