Since the first American Thoracic Society statement regarding the diagnosis and management of severe alpha-1 antitrypsin (AAT) deficiency in 1989 (1) and the initial Canadian Thoracic Society standards statement in 1992 (2) (which was updated in 2001) (3), significant advances in understanding the cell and molecular biology of AAT and the diagnosis, natural history, and treatment of individuals with AAT deficiency have occurred. These new developments, including completion of several large, longitudinal studies in both Europe and North America and a small randomized controlled trial of augmentation therapy, have provided important new insights that have impacted the clinical management of individuals with severe deficiency of AAT.

In the context of these new developments, a need was felt to reexamine recommendations for optimal management of AAT deficiency, to synthesize current knowledge of diagnosis and management for practicing clinicians, and to identify key remaining questions in need of further investigation. With these purposes in mind, a Task Force to develop a new standards document regarding the diagnosis and management of individuals with severe AAT deficiency was formed in 1998 under the auspices of the American Thoracic Society and the European Respiratory Society, with additional sponsorship and support by the Alpha-1 Foundation, the American College of Chest Physicians, and the American Association for Respiratory Care. Under a contractual arrangement, the Veterans Administration Technology Assessment Program, Office of Patient Care Services, Veterans Health Administration provided education regarding preparing an evidence-based document and support in conducting literature searches.

In keeping with current standards for developing evidence-based recommendations for optimal care, the current Task Force has undertaken a systematic review of current literature regarding AAT deficiency. Every effort was made to identify the scientific evidence for positions taken and to identify where there was little or no evidence. In the absence of ratable evidence, consensus among members of the Task Force determined the recommendation.

This summary document briefly describes the organization and preparation of the Task Force's report and provides an executive summary of key clinical recommendations. The three following sections are the full systematic reviews prepared by the three individual writing groups that comprised the AAT Deficiency Task Force.

The goal of the AAT Deficiency Task Force was to prepare and present for the medical and interested lay communities the reasoned, current views of a large international group of experts regarding the current diagnosis and management of individuals with AAT deficiency, using a systematic review and the evidence-based approach. The Task Force undertook to evaluate the full clinical and management dimensions of this multisystem illness, including lung, liver, and other organ manifestations. Also, issues relating to the ethical, legal, social, psychological, and economic implications of genetic testing for AAT deficiency were addressed.

A planning group was assembled in the Fall of 1997, when sponsorship and funding by the major sponsors—the American Thoracic Society, the European Respiratory Society, and the Alpha-1 Foundation—was finalized. Additional support from the Alpha-1 Foundation, the American College of Chest Physicians, and the American Association for Respiratory Care allowed the Planning Committee to assemble the full membership of the Task Force and to proceed.

As presented in Figure 1, the AAT Deficiency Task Force consisted of an Executive Committee, three individual Writing Groups comprising international experts, and a Steering Committee (composed of the Executive Committee and the Chairs of each of the three Writing Groups). Preparation of the systematic review was aided by members of the Health Care Technology Assessment Program of the Department of Veterans Affairs, who provided ongoing input and guidance to the project regarding literature searches and evidence-based medicine methods. Administrative assistance was provided by the American Thoracic Society.

The membership of the Task Force was fully constituted by September 1998, at which point Writing Groups began to review literature and to draft documents for subsequent review by the Steering Committee. The Steering Committee conducted a number of conference calls and five face-to-face meetings between Fall 1998 and Fall 2001 to review the evolving documents. Individual Writing Group documents were finalized by Fall 2001 for final editing by the Executive Committee and subsequent submission to the sponsoring organizations. Reviews were received in June 2002 and the revised document was resubmitted in Fall 2002 for final approval. Approval was granted by the American Thoracic Society in December 2002, when an additional review of salient literature led to a final update of the document.

While the Executive Committee has attempted to minimize overlap between the three documents, the Task Force's stated goal of preparing three individual documents, each complete and with its own emphasis, references, and supportive tables and figures, will inevitably lead to some overlap.

Finally, in the context that research is ongoing and that current understanding of AAT deficiency and optimal management is evolving, the Task Force recognizes the need for periodic review and updating of management recommendations.

Plasma AAT levels are usually determined by rocket immunoelectrophoresis, radial immunodiffusion, or, more recently, by nephelometry. Commercially available standards, especially those used for radial immunodiffusion, tend to overestimate the AAT concentration by as much as 35–40% (45). To discriminate between historic values obtained using the nonpurified standard and those obtained with the pure standard developed by the U.S. National Institutes of Health, the former are expressed as milligrams per deciliter (mg/dl), and the latter in micromolar units (μmol/L or μM). The two units are, however, often used interchangeably in many continental European countries, irrespective of the standard used. Moreover, nephelometry may also overestimate AAT levels, because of interference with lipids or hemoglobin. Also to be considered is that AAT is an acute-phase reactant, and inflammatory conditions may augment the steady state plasma AAT levels in Z heterozygotes. It should be noted that a “protective” threshold level of 11 μmol/L previously maintained corresponds to 80 mg/dl if measured by radial immunodiffusion and to 50 mg/dl if measured by nephelometry (see Table 4)

The most widely used method for identifying AAT variants is their separation based on the isoelectric point by means of thin-layer isoelectric focusing (IEF). This technique, commonly referred to as “phenotyping,” requires skill and experience and should be performed in reference laboratories. The IEF specificity may be further enhanced by coupling it with an immunoblot or by using an immobilized pH gradient IEF gel (46). Phenotyping may be performed on serum or plasma samples. Some laboratories perform IEF on “dried blot spot” samples, using a blood drop absorbed on special paper, allowing for easier transport of samples. This method is suitable for screening purposes, but the identification of a deficient variant should be confirmed on serum or plasma samples.

Diagnosis at a molecular level (“genotyping”) is performed on genomic DNA, extracted from circulating mononuclear blood cells. Known mutations may be detected by allele-specific amplification or analysis. Lack of recognition of a known mutation may imply the presence of a new variant. In this case, a gene scan should be performed by means of direct sequencing, or denaturing gradient gel electrophoresis (47). Molecular level diagnosis has been made easier by the commercial availability of test kits capable of detecting S and Z alleles in whole blood or mouthwash samples. However, available kits will miss null alleles and plasma levels of AAT may also be necessary.

PI*ZZ deficiency is inherited as an autosomal codominant gene. The risk of a homozygous offspring is 1 in 4 for each birth if both parents are carriers of the Z allele. If one parent is PI*ZZ and the other heterozygous, then all children are either carriers or affected (PI*ZZ).

There is no routinely available method developed for the prenatal diagnosis of the condition. Amniocentesis or chorionic villus sampling (48, 49) provides the material on which genetic testing can be performed. Requests for prenatal diagnosis may be based on a history of perinatal liver disease in a previous sibling, in which case the risk of developing liver disease may rise substantially (9, 50). Several techniques of prenatal gene identification have been reported and are available in limited cases. All require techniques of DNA amplification and use specific probes to provide adequately specific diagnostics (51, 52). However, financial and practical considerations limit their usefulness.

Postnatal detection of AAT deficiency depends on a high level of suspicion. The technology for rapid screening techniques is available and utilizes DNA amplification from heel blood samples (53). Postnatal detection may occur in the setting of neonatal hepatitis or a strong family history. Otherwise, most cases remain undetected unless emphysema, liver disease, or rare complications develop. Although AAT deficiency is one of the most common codominant disorders to affect Caucasians, routine screening is not performed.

It is recognized that nonsmoking individuals with the homozygous Z phenotype have a remarkably delayed onset of symptoms and some may have an almost normal life span (54). Thus, the exact prevalence of AAT deficiency in most populations remains unknown and many afflicted individuals remain undiagnosed. Large-scale screening programs of the newborn or adult populations in the United States and Europe (except for Sweden) have not been undertaken because of cost and issues of personal vulnerability related to the presence of an inherited abnormality. Because there is presently no cure for the disease, subjects and their families must withstand the emotional stress of living with this knowledge at a time when they may be totally asymptomatic. Also, identification of the abnormality could compromise an individual's status with respect to insurability and employment. Because avoidance of smoking and exposure to hazardous respiratory environments may benefit the prognosis of individuals who have AAT deficiency, there is some medical justification for early detection. However, in recognition of the positive and negative factors related to detection, informed consent for performing the diagnostic tests should be obtained from the subject by the attending physician after a thorough explanation of the issues involved.

Three categories of genetic testing have been specified (see the Genetics, Psychosocial, Ethics, and Economic Issues section). The first type is labeled “diagnostic” testing and refers to the testing of individuals with symptoms and/or signs consistent with an AAT deficiency-related disease.

The second type of testing is labeled “predispositional” testing, which refers to identifying asymptomatic individuals who may be at high risk of having AAT deficiency. The third type of testing is labeled “screening,” which refers to programs designed to search in populations for persons possessing certain inherited predispositions to disease. The hallmark of screening is that there should be no previous suspicion that any given individual has the condition being tested. Specific recommendations and associated recommendation grades for testing in specific groups of individuals are given in Table 8 of the Genetics, Psychosocial, Ethics, and Economic Issues section.

Subjects with abnormal blood levels should be investigated further to provide a qualitative evaluation of their AAT disorder. Even subjects with a borderline normal AAT plasma level (12–35 μmol/L or 90–140 mg/dl) and their first-degree relatives should undergo qualitative testing, because these levels may correspond to an intermediate level phenotype (SZ, SS, MZ) and a relative with asymptomatic or misdiagnosed AAT deficiency may be uncovered within the family.

Beside occasional observation of a reduced or absent α₁-globulin band on electrophoresis, an AAT level is particularly important in patients with early-onset pulmonary emphysema with or without a history of cigarette smoking. Testing should also be performed in siblings of AAT-deficient individuals and considered for offspring and when there is familial aggregation of symptoms of shortness of breath and chronic cough. Furthermore, it is recommended that all subjects with COPD or asthma characterized by incompletely reversible airflow obstruction should be tested once for quantitative AAT determination (see Table 5)

Optimal clinical practice would indicate that full lung function testing including spirometry, static lung volumes, arterial blood gas analysis, and gas transfer should all be assessed at baseline to fully document the physiologic status of patients with obstructive pulmonary disease (77, 78).

Because AAT-deficient individuals mainly develop fixed airflow obstruction, it does not seem justified to assess variability of airflow obstruction by peak expiratory flow monitoring in most subjects with AAT deficiency. However, follow-up of patients should include spirometry at yearly intervals.

Chest roentgenography in early disease is usually normal. In advanced disease, hyperinflation and increased radiolucency of the lungs, particularly in the lower lung segments, are evident. Diaphragms are low and flat, and exaggerated verticality of the heart, increased anterior–posterior diameter of the chest, and widening of the retrosternal space are present. Of these criteria, diaphragmatic flattening in the lateral projection is probably the most specific. Vascular markings are decreased, mainly in the lower zones, in contrast to the preponderance of upper zone disease in non-AAT deficiency emphysema (63, 79–83). Concomitant enlargement of the hilar pulmonary arteries is evidence of possible pulmonary hypertension.

Computed tomography is much more sensitive than plain chest radiography or pulmonary function tests for the presence of emphysema (74, 83, 84). High spatial frequency reconstruction of images (high-resolution computed tomography [HRCT]) is more sensitive for detecting morphologic changes such as bullous disease and bronchiectasis. On HRCT, emphysema is characterized by the presence of areas of abnormally low attenuation, which can be contrasted with surrounding normal lung parenchyma if a sufficiently low window level (–600 to –800 Hounsfield units [HU]) is used. In AAT deficiency, the classic finding is panacinar emphysema in terms of uniform abnormally low attenuation of lobules, and predominant lower lobe distribution. Pulmonary vessels in the affected lung appear fewer and smaller than normal (85), but mild and even moderately severe panacinar emphysema can be subtle and difficult to detect radiographically (86). Whereas focal areas of emphysema usually lack distinct walls, bullae, by definition are sharply demarcated by a thin wall and measure 1 cm or more in diameter. Bullae are more common in usual (non-AAT deficiency) emphysema (79).

Despite the well recognized association between AAT deficiency and the early development of emphysema, only a limited number of studies have assessed the association between AAT deficiency and bronchiectasis. In Eriksson's original patients, bronchiectasis was reported in 2 of 23 patients (59). Guest and Hansell observed bronchial wall thickening and/or dilation in 7 of 17 patients (41%) with AAT deficiency (82) and King and coworkers found evidence of bronchiectasis in 6 of 14 patients (43%) with AAT deficiency (87). Cuvelier and coworkers reported that the frequency of AAT deficiency alleles is not increased among patients with bronchiectasis, whereas other studies suggest bronchiectasis is more common in Hispanic patients (88). The study by Cuvelier and coworkers suggests that bronchiectasis is more a result of emphysematous changes in the parenchyma than of AAT deficiency per se (88).

Pixel values of CT images represent tissue density, which makes it possible to calculate densitometric parameters from frequency histograms of pixels within the lung. These densitometric parameters provide a quantitative assessment of the extent of emphysema. Various methods for calculating densitometric parameters have been reported in the literature. The “density mask” method enhances areas of abnormally low attenuation, using a computer program to highlight pixels to create a density mask within any desired range. The density mask parameter is defined as the percent area of lung highlighted relative to the total lung area. Highlighting all pixels with attenuation values below –910 HU correlates with pathologic scores and is comparable to that obtained by visual assessment (89). The “percentile” method assesses the extent of emphysema by the cutoff point that defines a given percentile of the histogram (e.g., the tenth percentile is extracted from the histogram as the density value, in Hounsfield units, at which 10% of the pixels have lower densities) (90). A third possibility is to assess the extent of emphysema by “mean lung density” (91).

Ventilation–perfusion scanning may be a useful tool in detection of early changes associated with AAT deficiency, as even individuals with relatively normal lung function may have abnormal scans (63, 76, 92–95). Typically, the ventilation scan shows symmetric distribution of xenon-133 throughout all zones of the lung during the equilibrium phase followed by a symmetric delay in washout, most prominent in the lung bases and midzones. A symmetric loss of pulmonary arterial perfusion is also found, most marked in the bases.

In summary, a chest X-ray should be performed as an initial test for incidental lung lesions or possible identification of localized bullous disease. However, CT scanning, including thin slices for morphology and thick slices for densitometry, is presently the most definitive technique to detect and quantify the presence of emphysema in AAT deficiency.

The principal source of variation of lung density measurements is the level of inspiration. Ranging from full inspiration to full expiration, lung densities more than double (100–102). The volume of air in the lung can be derived from CT images, that is, CT lung density histograms (102–104). Half the lung density is composed of blood in the microvascular circulation (105). When a deep breath is taken, not only is the volume of air increased but also blood is sucked into the lung, which makes the relation between inspiratory level and lung density more complicated.

The belief of some investigators that spirometric control of ventilation during scanning improves the reproducibility and accuracy of density measurements (106–108) has been complicated by the recognition that patients with COPD are less able to successively reproduce levels of full expiration and inspiration to allow standardization (109, 110).

Analysis of repeated CT scans at various levels of inspiration has revealed the relationship within an individual between lung density and volume. On the basis of this relationship, densitometric parameters can be standardized by log-transformed lung volume, which corrects for differences in lung volume between scans and eliminates the need for spirometrically controlled CT (111). Using a broad range of pixel percentiles from the 10th to the 30th (corresponding to densities ranging from –950 to –890 HU), the annual decline was found to be 2 HU, corresponding to a loss of lung tissue of 2 g/L lung volume (111). CT quantitation of emphysema by densitometric parameters seems to be a more sensitive measure of the progress of emphysema as compared with pulmonary function tests (e.g., FEV₁) (98, 112, 113). Inspiratory CT was superior to expiratory CT for longitudinal estimation of structural abnormalities caused by aging and smoking (113) and the pixel percentile was more robust than the pixel index for monitoring the progress of emphysema (111).

In a randomized clinical trial of AAT augmentation therapy over a period of 3 years with 56 patients, the sensitivity to measure the progression of emphysema by the percentile method proved two- to threefold higher than any parameter measured by spirometry or CO-diffusing capacity. This implies that new drug trials using CT as any outcome parameters are feasible with a fivefold lower number of patients (112).

Because it has been shown that CT lung density is influenced by age, normal CT attenuation values for the lung by age should be established (101, 113–116). Furthermore, pixel attenuation values fluctuate with the position in the thorax and change with aging of the X-ray tube (117).

A problem related to the use of CT for monitoring the progress of emphysema is radiation exposure. Limiting the examination to a single slice 5 cm below the carina would markedly reduce the radiation dose and results derived from calculations based on a thin slice were similar to results based on a volume scan of the whole lung (111). Also, a reduction of the electrical current (mA) to levels 10 times below standard settings has little influence on lung density measurements (118, 119).

Because it is widely accepted that lung damage in individuals with AAT deficiency results from an imbalance between neutrophil elastase (NE) and AAT, it is reasonable to suppose that factors related to the activity of NE and/or to turnover of lung extracellular matrices could be indicators of lung destruction (120). The former (active NE and the NE-specific fibrinogen-derived Aα peptide) (121) are indirect markers, whereas the latter (desmosine/isodesmosine [DES/IDES] and elastin-derived peptides) are direct markers of elastin degradation and are therefore more closely linked to the true clinical outcome of emphysema (122). Elastin-derived peptides may be measured immunologically in plasma or urine, and have been found to be significantly increased in individuals with COPD compared with control subjects by direct measurement (123) and by radioimmunoassay (124); however, the incomplete biochemical characterization of the immunoreactive material raises concerns about its suitability for controlled clinical trials.

DES and IDES are cross-linked amino acids unique to mature elastin that are contained in elastin-derived peptides in the bloodstream and excreted in urine. Urinary DES levels are higher in smokers with a rapid decline of lung function indices than in those with slow decline (125). In addition, they are as high in subjects with AAT deficiency as in subjects with usual, AAT-replete COPD (123). Preliminary evidence that AAT augmentation therapy decreases urinary DES excretion in AAT-deficient subjects (126) prompted investigators to design clinical trials of short-term supplementation therapy with the primary end point being reduction of the urinary rate of excretion of DES/IDES. Results of one of these trials could not confirm the preliminary finding (127). Taken together, studies thus far support urinary DES/IDES as potential candidates for monitoring progression of lung disease and efficacy of therapy in AAT-deficient subjects, but further evidence is needed to demonstrate methodologic reliability and clinical relevance to progression.

In summary, there are adequate data to suggest that FEV₁ and CT densitometry are reliable methods to detect progression of emphysema in AAT deficiency. Spirometry should initially be performed on an annual basis until it is clear that progression is not rapid when a reduction in frequency of assessment can be initiated.

As discussed in the Liver and Other Diseases section, a variety of other conditions have been reported as associated with AAT deficiency. These include systemic vasculitides such as anti-neutrophil cytoplasmic antibody-positive Wegener's granulomatosis and necrotizing panniculitis (149–152), peripheral neuropathy (153), and cerebral or peripheral artery aneurysms (154). The latter report has not been substantiated by larger studies (155). Some report amelioration of signs of vasculitis with α₁ antiprotease treatment (149, 156). It is notable that results from the NHLBI Registry of 1,129 patients failed to reveal an association of death with any condition other than lung and liver disease (137).

A rare complication of PI*ZZ is the development of necrotizing panniculitis (157, 158). This condition is an inflammatory response to an unknown stimulus with typical necrotic lesions in the subcutis and the dermis, which can be extensive.

Only 2.5% of newborns diagnosed with PI*ZZ die because of acute liver failure. Patients over 50 years of age can also develop hepatocellular carcinoma and liver cirrhosis (9). However, the most impressive finding in more recent studies (10, 11) is the predominant role of cirrhosis-related mortality, especially in elderly never-smokers. Cirrhosis and its complications were the main cause of death in 12 of 41 patients and occurred in 14 of 41 patients, but in only 2 of 23 smokers as compared with 12 of 17 never-smokers (p < 0.001). Hepatocellular carcinoma occurred in five cirrhotic livers but was not seen in any noncirrhotic livers.

The prevalence of cirrhosis in AAT deficiency has been underestimated. The figure of 5% maintained in a World Health Organization report (56) now seems too low. The total lifetime risk of cirrhosis is more realistically in the range of 30–40% (9, 10).

Polymerization of AAT protein is thought to be involved in the pathogenesis of these liver diseases, but the exact mechanism of the pathogenesis is unknown.

The first body of evidence suggesting that PI*MZ deficiency carries a risk for chronic obstructive pulmonary disease was found when phenotypes were determined for patients with established respiratory disease, and PI*MZ patients were found to be more prevalent than would be expected by the gene frequency in the population (159–168). A few studies have suggested that this relationship does not exist (169–172).

Many of these initial studies were performed by identifying a PI*MZ cohort by their relationship to symptomatic PI*ZZ patients. When obstructive abnormalities were found in this patient population (173–176), these studies were criticized because the airway obstruction might represent genetic or environmental factors other than AAT deficiency. In fact, the clustering of COPD in families without AAT deficiency mandated the study of individuals from generalized populations rather than relatives of symptomatic patient populations (see Appendix 2).

The largest trial to date evaluating PI*MZ patients, conducted by Seersholm and coworkers, demonstrated that 9% of Danish hospitalizations for COPD exacerbations (n = 17,061) occurred in PI*MZ individuals (177). Compared with PI*MM patients, the PI*MZ phenotype carried a relative risk of hospitalization for COPD of 2.2 (95% confidence interval, 1.5–3.0). Furthermore, this excess risk was concentrated in 40- to 79-year-old first-degree relatives of PI*ZZ index cases, confirming the familial tendency to develop COPD in PI*MZ heterozygotes. It is noteworthy that a high prevalence of the PI*MZ phenotype was found in individuals with cor pulmonale. Further studies of familial clustering of COPD in PI*MZ individuals are included in Appendix 5.

Smaller population-based studies have been performed to determine whether PI*MZ individuals have an excess prevalence of COPD when matched for the presence of cigarette smoking. These studies have suggested similar FEV₁ values between PI*MM and PI*MZ patients (178–186), with a few exceptions (187–189).

However, subtle abnormalities of lung function have been seen in PI*MZ patients. Lung recoil pressures, mean forced expiratory flow during the middle half of the FVC (FEF_25–75%) (173, 174, 181, 190, 191), frequency dependence of total pulmonary resistance (192), frequency dependence of dynamic compliance (173), ventilation inhomogeneity (192), and arterial oxygen tension have been found to be different in PI*MZ patients than in PI*MM patients. Most authors have suggested that these minor abnormalities have no clinical importance.

Population-based autopsy studies have also been at variance. One postmortem study has shown an increased prevalence of emphysema in MZ patients, although there was no association of premorbid symptoms or premature death (193). Another showed no difference in emphysema prevalence compared with PI*MM patients (194).

To reconcile these studies and further analyze the covariate influence of cigarette smoking on airway obstruction, Silverman and coworkers (195) constructed regression equations for FEV₁ and FEF_25–75% versus cigarette pack-years for PI*MZ relatives of PI*ZZ cases to demonstrate that PI*MZ patients do have an increased risk of airflow obstruction that is intermediate between that of patients with the PI*MM and PI*ZZ phenotypes. Another population-based longitudinal study by Eriksson and coworkers (196) monitored FEV₁ serially for 6 years in a cohort of PI*MZ individuals. The FEV₁ decline among smokers, 75 ml/year, was greater than that among PI*MZ nonsmokers, 40 ml/year (p < 0.05).

Additional data from the NHLBI Lung Health Study indicate the risk of the MZ phenotype in the smoking population. Among the 5,887 male and female smokers recruited into the NHLBI Lung Health Study, 283 subjects with the fastest decline in FEV₁ (ΔFEV₁ = –154 ± 3 ml/year) and 308 who had no decline (ΔFEV₁ = +15 ± 2 ml/year) were genotyped for polymorphisms in the AAT protein, microsomal epoxide hydrolase, vitamin D-binding protein, and tumor necrosis factor genes. Rapid decline of FEV₁ was associated with the MZ genotype of the AAT gene (odds ratio [OR], 2.8; p = 0.03). This association was stronger for a combination of a family history of COPD with MZ (OR, 9.7; p = 0.03). These data suggest that the MZ genotype results in an increased rate of decline in lung function and interacts with other familial factors. The microsomal epoxide polymorphisms were, like PI*MZ AAT, associated with a more rapid decline in lung function. In this study, the AAT S and 3′ polymorphisms, vitamin D-binding protein isoforms, and tumor necrosis factor (TNF-α 200304467oc G308A and TNF-α–A252G) polymorphisms were not associated with an accelerated rate of decline in lung function (197). Overall, many studies have solidified the understanding that cigarette smoking is a risk for COPD in PI*MZ individuals (see Appendix 6).

Some PI*MZ population-based cohorts have been shown to have more breathlessness (198) and wheezing than do PI*MM patients, particularly if subjected to the stresses of environmental dust or smoking. Others have not shown aspects of these findings in large populations (178, 182, 198–201).

The influence of the work environment has been evaluated in cotton mill workers at risk for byssinosis and the PI*MZ phenotype was found to be a risk factor for symptoms by multivariate analysis, in which endotoxin levels were also measured and controlled for (202, 203). Although other studies have found no association between PI*MZ patients and byssinosis symptoms (204, 205), these studies did not control for the stronger risk factor of endotoxin exposure. Other dusty occupations associated with risk may include grain working (206) and mining (207). Current studies addressing environmental air quality as a risk factor for COPD in PI*MZ individuals are listed in Appendix 7.

A less complete database has been assembled for atopic diseases (see Appendix 8). Although the PI*MS heterozygote may have an increased risk of asthma (208, 209), PI*MZ gene prevalence may (186) or may not (201, 210–212) be higher than in the general population of patients with asthma. Other studies have reported an increase in PI*MZ prevalence in populations with aspirin-sensitive asthma (213), nasal polyposis, and sinusitis (214, 215).

Other respiratory diseases that have been reported as occurring more frequently among PI*MZ individuals include community-acquired pneumonia (216), rheumatoid interstitial lung disease (217), and Wegener's granulomatosis (10).

The influence of the PI*MZ phenotype on respiratory disease appears to be established with an increased risk of chronic obstructive pulmonary disease. From the available studies, the concept of a multifactorial genetic, smoking, and environmental interaction has been established for PI*MZ patients. The usual PI*MZ individual who smokes has mild spirometric abnormalities that manifest later in life. A more substantial risk for symptomatic COPD may occur during the stresses of environmental and occupational challenge in heavy cigarette smokers, particularly in relatives of patients with obstructive lung disease.

Studies have indicated that smoking imposes a major risk for the development of pulmonary emphysema, but in the absence of smoking the risk of developing emphysema is negligible (218–220) (see Table 8 and Appendix 9). However, the effects of smoking are profound and produce reduction in pulmonary function, which is often as marked as in PI*ZZ smokers. Although no studies have been done to assess the effects of occupational and environmental pollutants on individuals with the SZ phenotype, it is reasonable to conclude that they are as vulnerable as individuals with the PI*ZZ phenotype.

Intravenous administration of purified human AAT concentrate was shown to increase lung levels of AAT in AAT-deficient individuals in the early 1980s (238, 239). Antineutrophil elastase capacity in the lung epithelial lining fluid, obtained by bronchoalveolar lavage increased to 60–70% of normal in homozygous AAT-deficient individuals who received human plasma-derived AAT once weekly at a dose of 60 mg/kg body weight. On the basis of these studies, a purified preparation of AAT concentrate derived from fractionated plasma (Prolastin) was manufactured and shown to be biologically effective (240–242). Other preparations of pooled human plasma antiprotease are also available (240), and recombinant technology-based drugs are under development.

Results of these studies formed the basis for U.S. Food and Drug Administration approval in the United States in 1988, followed thereafter by approval by regulatory agencies in Canada, Germany, Spain, and Italy. To date, only a single randomized placebo-controlled trial has been done to determine whether augmentation therapy attenuates the development of emphysema (112). Such trials are difficult and expensive to mount because of the large number of patients and prolonged time needed to adequately power such a study. Recommendations regarding the use of augmentation therapy have been issued by the American Thoracic Society earlier (243) and by the Canadian Thoracic Society (244, 245).

Two reports have addressed the issue of clinical efficacy of augmentation therapy in concurrently controlled observational studies, a German–Danish study (246, 247) and the NHLBI Registry in the United States (137). The aim of the former study was to compare the decline in FEV₁ between Danish patients who had never received augmentation therapy and German patients treated with weekly infusions of AAT. The yearly decline in FEV₁ in the treated group (–53 ml) was significantly lower than in the untreated group (–75 ml) (p = 0.02). Notably, the two groups differed with respect to sex and initial FEV₁. In comparing the different degrees of functional impairment, a significant effect of the treatment was demonstrated only in the group of patients with an initial FEV₁ of 31–65% predicted. The authors concluded from this nonrandomized study that weekly infusion of human AAT in patients with moderately reduced lung function may slow the annual decline in FEV_1.

In the NHLBI Registry, 1,129 subjects with severe deficiency of AAT (serum levels less than 11 μM) were registered. They were monitored for 3.5 to 7 years. The 5-year mortality rate was 19%. In a multivariate analysis, the mortality rate was lower in those receiving augmentation therapy as compared with those not receiving therapy (OR, 0.79; p = 0.02). The mean FEV₁ decline was 54 ml/year and there was no overall difference between those receiving augmentation therapy and nonrecipients. However, among those in the subgroup with moderate emphysema (i.e., American Thoracic Society Stage II emphysema with FEV₁ 35–49% predicted), the rate of FEV₁ decline was significantly slower in subjects receiving augmentation therapy (p = 0.03).

These two observational studies suggest that progression of emphysema may be slowed in patients with moderate emphysema (FEV₁ 31–65% predicted). Moreover, mortality may be decreased in patients with a lower FEV₁.

To extend the assessment of augmentation therapy efficacy, a small randomized clinical trial including 26 Danish and 32 Dutch patients has been performed (112). Active treatment involved monthly infusion of 250 mg of AAT per kilogram; placebo was albumin. The 2-year study was too small to show any difference in the slope of decline of lung function, but loss of lung density by CT showed a trend in favor of the actively treated group (50% reduction in decline, p = 0.07). This study portends the value of lung density measurements by CT as an end point for efficacy.

Adverse reactions to AAT concentrate when given intravenously have been rare. Between February 1989 and December 1995, about 58,000 infusions were administered to 443 patients in one series (247). One hundred and twenty-four mild adverse reactions (fever, chills, dyspnea) were reported in 65 patients. Three patients terminated treatment with AAT concentrate because of repeated chills and fever immediately after infusion. In four patients, an anaphylactic reaction occurred, but in all cases complete recovery was obtained. An IgE-mediated anaphylactic reaction following the third intravenous infusion has been published (248). Because no viral transmission has been observed, this group does not consider immunization for hepatitis B to be essential before beginning therapy.

Aerosol application of AAT in patients with AAT deficiency increases AAT concentration and antielastase activity in the lower respiratory tract in a dose-dependent fashion (249). Deposition of radioactively labeled AAT showed more peripheral deposition in patients with mild emphysema compared with central deposition in those with advanced emphysema (250). No bronchospasm has been reported. Preliminary data suggest that once or twice daily administration of aerosolized AAT may produce sustained antielastase protection of the lungs.

A number of recombinant forms of AAT have been developed as well as recombinant secretory leukoprotease inhibitor (251). Human AAT genes have been transferred into sheep embryos, resulting in the secretion of mannose-glycosylated AAT into the milk of these animals (252).

Currently several synthetic, low molecular weight elastase inhibitors are being evaluated, but their clinical efficacy and safety have not been reported.

In the context that no randomized controlled trial has definitively demonstrated the clinical efficacy of augmentation therapy, the weight of available studies of the clinical efficacy of AAT augmentation therapy indicates a lowered overall mortality and a slower rate of FEV₁ decline in augmentation therapy recipients with FEV₁ values of 35–65% of predicted. Use of high-resolution CT scans to measure lung tissue density shows great promise for providing a practical, quantitative way to assess the efficacy of augmentation therapy in future studies.

Harvey Sharp and associates first described cirrhosis in alpha-1 antitrypsin (AAT) deficiency in 10 children from six different kindreds (1). In 1971, Sharp also described intrahepatocyte periodic acid–Schiff diastase (PAS-D)-resistant globules or inclusions, the result of polymer formation by mutant PI*Z protein in the endoplasmic reticulum (2). Individuals manifesting this intrahepatocytic aggregation are at increased risk of developing cirrhosis. This section, prepared by the Liver and Other Diseases Writing Group of the Alpha-1 Antitrypsin Deficiency Task Force, presents a systematic analysis of the strength of association among various forms of severe and intermediate AAT deficiency and liver disease in the various age groups.

The pathophysiology of liver disease in AAT deficiency is different from that of lung disease. Although a variety of theories have been proposed to explain liver injury in individuals with homozygous PI*ZZ AAT deficiency, the most widely accepted explanation, and the only one supported by significant experimental evidence, is the “accumulation theory” (reviewed in references 3 and 4). This theory states that liver injury in AAT deficiency results from accumulation of mutant, AAT Z protein molecules within the endoplasmic reticulum (ER) of hepatocytes. Large quantities of mutant AAT Z protein are synthesized in the liver of PI*ZZ individuals, but about 80–90% of the Z protein synthesized appears to be retained within hepatocytes rather than being efficiently secreted. Intracellular processes, known as the quality control apparatus, recognize that the nascent mutant AAT Z polypeptide is abnormal during biogenesis and direct it to a pathway of retention within the ER rather than allowing export from the hepatocyte. The quality control apparatus of the cell involves “molecular chaperone” proteins that are members of the heat shock protein family.

The accumulation theory of liver injury in AAT deficiency is supported by several lines of evidence. First, reports concerning various experimental mice transgenic for the human AAT Z gene have shown the accumulation of AAT Z protein within hepatocytes. Furthermore, the patterns of liver injury, including neonatal liver disease, hepatic fibrosis, and late hepatocellular carcinoma, are similar to the human disease (5–7). Although the results of the AAT Z transgenic mouse experiments have been criticized for their variability, which could result from environmental or other mouse genetic factors, they do completely rule out the possibility that the liver injury is due to low circulating levels of AAT in the serum. This is because these mice have normal levels of endogenous antielastases as directed by their intact murine antielastase genes.

Other evidence for the accumulation theory has come from investigations of the intracellular processing of AAT Z protein within the ER. Studies have shown that the majority of PI*ZZ individuals are “protected” from liver injury by efficient mechanisms for the intracellular degradation of the AAT Z protein retained within the ER. However, patients susceptible to liver injury appear to have inefficient intracellular degradation mechanisms, leading to a net increase in the ER accumulation of the retained AAT Z protein (8). Host-specific genetic or environmental differences in the hepatocellular response to AAT Z intracellular accumulation may explain the clinical observation that only a subpopulation of PI*ZZ individuals develop significant liver injury.

Interestingly, the AAT Z molecule has an increased tendency to form polymers by the so-called loop–sheet insertion mechanism (9). Molecular structural studies have suggested that a single amino acid substitution in the Z mutant and some other deficient proteins results in insertion of the loop of one Z molecule into the A sheet of another, so that long interlocking polymers form. However, it is unclear whether the ER retention signal for AAT Z occurs at a step proximal to the formation of polymers or whether polymerization itself triggers a retention signal within the cell.

The exact relationship between AAT Z polymerization and hepatocellular injury remains unclear. Although polymerization of AAT Z has been proven to occur in vivo, the exact links to the pathophysiology of the human disease remain undefined. Studies of environmental factors that could predispose AAT deficiency patients to liver injury have so far yielded conflicting results. Suggestions that breast-feeding might be protective against the development of liver disease in children have not been consistent across all reports (10, 11). Some studies have suggested an increase in the prevalence of viral hepatitis infection in heterozygous PI*MZ adults with liver disease; other studies in adults have not found such a correlation (12, 13). However, a theory supports the possibility that AAT deficiency–associated liver disease can be exacerbated by viral hepatitis, because both hepatitis B and hepatitis C viruses express proteins that are selectively retained in the ER of hepatocytes (14). It has been suggested that whatever cellular injury results from the ER retention of proteins could be compounded by the additional retention of these viral proteins. However, it is clear that nearly all PI*ZZ children and adults with liver disease are free from hepatitis B and C infections, and that PI*MZ individuals do not develop AAT deficiency–associated liver disease during their childhood years (15).

Alpha-1 antitrypsin deficiency (AAT deficiency) PI*ZZ predisposes to liver disease, often presenting as jaundice in early infancy. In countries with a high prevalence of AAT deficiency, it is a common cause of neonatal cholestasis. Alpha-1 antitrypsin deficiency is also a common metabolic disease in children with end-stage liver disease for whom liver transplantation is performed.

The association of liver disease with AAT deficiency in adults is less clear than in children. In particular, the association between intermediate AAT deficiency (PI*MZ, PI*SZ) and liver disease has been controversial. Although many studies on the association between AAT deficiency and chronic liver disease (CLD) have been reported, variations in approach and study results preclude a clear understanding. The current systematic review and metaanalysis was undertaken to analyze evidence in the literature regarding the causal link between AAT deficiency and CLD in children and adults.

AAT deficiency has been shown to be involved in immune processes. For example, heterozygotes for the Z allele have been reported to be at increased risk of developing anterior uveitis and rheumatoid arthritis as well as a variety of collagen vascular diseases (115). Overall, these associations appear weak and often controversial and seem to lack impact on clinical management of the patient. Pheno- or genotyping of AAT in such patient cohorts may be motivated from a research point of view, but is rarely justified for clinical reasons. More important are a number of case reports linking the PI*Z deficiency state to systemic vasculitis and glomerulonephritis (116–122). The results of two relatively small series reported in 1993 for the first time showed a possible relationship between the Z gene and the presence of C-ANCA (antineutrophil cytoplasm antibodies) (123, 124). Since then, numerous studies have confirmed a strong relationship between hetero- and homozygous PI*Z AAT deficiency and small vessel-necrotizing vasculitides, in particular Wegener's granulomatosis and microscopic polyangiitis (125–131). As is evident from Table 13, the link between the PI*Z allele and anti-PR-3 (antiproteinase-3 or C-ANCA) is convincing. As a rule, in each studied cohort of C-ANCA-positive patients at least one PI*Z homozygote is identified. It means that about 2% of all patients with anti-PR-3-positive multisystemic vasculitis can be expected to be PI*Z homozygotes, the remaining PI*Z individuals being heterozygotes. In none of the reports in Table 13 has the smoking behavior of subjects been included. Proteinase-3 is a neutrophil elastase-like serine protease localized in the primary granules of the neutrophil. It possesses a potent tissue-destructive capacity. In the extracellular fluid, AAT is a major physiologic inhibitor of PR-3. It has been suggested (132) that the subnormal response of plasma AAT seen in vasculitic patients enhances the risk of fatal outcome, suggesting that AAT has a crucial function as a protective protein in vasculitic syndromes. Rather than being merely an etiologic risk factor, the PI*Z variant for AAT deficiency may have an adverse accelerative effect on a vasculitic process once it starts.

Despite the firm association between Z allele–associated, C-ANCA-positive Wegener's granulomatosis, it is obvious that AAT deficiency is only a minor genetic risk factor for the development of vasculitic disease. However, the association attracts attention to the putative pathogenetic importance of a protease–antiprotease imbalance in these conditions (125). The relative lack of AAT could theoretically also promote the development of autoimmunity due to the exposure of a normally intracellular antigen, PR-3. Alternatively, the observations could be explained by a linkage disequilibrium phenomenon, in which immunologically important gene(s) tend(s) to be inherited with a particular AAT gene (130). Finally, the intrinsic properties of the mutant Z and S proteins, particularly their polymerization ability, could be pathogenetically important. Future clinical studies on PI*Z-associated vasculitis in particular and other vasculitic syndromes in general should include attempts to analyze the pathogenetic role of smoking-induced AAT deficiency in these conditions. At present, available literature does not permit any meaningful analysis of the role of smoking in nephropathy or panniculitis.

Homozygotes and heterozygotes are overrepresented among patients with anti-PR-3-positive vasculitic syndromes (Table 13), which is why diagnostic testing for AAT deficiency is indicated for this group. Considering the low incidence of vasculitic disorders in the population, it is obvious that the relative risk of a particular PI*Z heterozygote for developing a vasculitic complication is small. In heterozygotes with active disease, the AAT level is frequently normal (127), which is why quantitation alone is insufficient. IEF is mandatory to exclude heterozygosity. There is a need for large prospective collaborative studies to confirm the possible adverse prognostic effects of AAT deficiency in this setting and for randomized controlled studies focusing on the effect of augmentation therapy.

An association between the PI*S variant and C-ANCA-positive vasculitic disorders has been reported in only one study (125). In P-ANCA (anti-myeloperoxidase)-positive vasculitis, a slightly increased frequency of individuals with the PI*S allele was reported in series from the UK and Italy (130, 133). In the latter study, the MZ phenotype was more prevalent among patients than control subjects (8 versus 1.5%), but with no significant difference between C-ANCA and P-ANCA. This report also describes subnormal AAT levels in the normal PI*M phenotype during acute illness.

Various types of glomerulonephritis have been reported in AAT deficiency, predominantly in the pediatric age group (134). Specifically, there are three case reports (lacking ANCA data) of systemic vasculitis with glomerulonephritis in young adults with severe AAT deficiency (116, 118, 121). Glomerulonephritis alone in the elderly with severe PI*ZZ deficiency seems to be rare (135). In the latter case, cirrhosis and emphysema were also present. One 30-year-old man with AAT deficiency with cirrhosis in early childhood developed IgA nephropathy and hypertension in adult life. He had no lung disease (136). In 20 children with AAT deficiency–associated liver disease, glomerular changes were seen in 79% versus only 43% in age-matched control subjects with hepatic failure not related to AAT deficiency. The glomerular changes were heterogeneous, with a high proportion of mesangiocapillary glomerulonephritis (134). In one study (137), five AAT-deficient children with end-stage liver disease and preoperative evidence of kidney pathology developed severe hypertension after liver transplantation. Dramatic improvement of renal function after liver transplantation in AAT deficiency has been reported (138).

In conclusion, when evaluating nephropathy in adults with AAT deficiency, an ANCA-positive vasculitic complication is the most probable explanation, whereas in children, an association with chronic liver disease is more likely. Nephropathy should be borne in mind, particularly in children, when patients are evaluated for liver transplantation.

Biochemical data (139–141) support the concept of a systemic alteration in elastin metabolism in abdominal aortic aneurysms (AAAs). The inhibition of various proteolytic enzymes, such as neutrophil elastase by AAT, may have an important role in maintaining the integrity of connective tissue, including blood vessel walls. Furthermore, reports of familial aggregation of AAA (142, 143) and the male predominance in most cases (143) suggest a significant genetic component in the pathogenesis of some aneurysms. Against this background, several authors have tried to establish a link between AAT deficiency and the formation of AAA. Early reports were based on isolated case presentations (144). The first large study on 47 patients with AAA suggested a higher frequency of AAT deficiency in patients with AAA (145). In contrast, in a case-control autopsy-based study (146) in 31 PI*ZZ individuals, no increased frequency of AAA could be substantiated. More recently, no evidence for an association between AAT deficiency and AAA was found in larger, separate groups of patients from Pittsburgh, Pennsylvania, and from London, UK (147). There was also no evidence of association in a Swedish study of 102 consecutive patients with AAA (148). In the latter study, AAAs with a diameter of 40 mm or more were analyzed separately, but even then no increased PI*Z allele frequency could be found. Furthermore, in patients with homozygous AAT deficiency, abdominal aortic diameter did not differ from that in non-PI*Z control subjects (149). However, male individuals had significantly lower aortic stiffness values than did control subjects, a finding that might be compatible with an early vessel wall abnormality.

A German study (150) comprising 103 patients found heterozygous AAT deficiency to be of little or no importance for the development of AAA. In a later study (151), the authors found an increased prevalence of the rare phenotype PI*MV. PI*V is of unknown functional significance.

In conclusion, the evidence for a postulated link between AAT deficiency and AAA is weak. In clinical practice, screening of such patients for AAT deficiency is not warranted.

Several studies have focused on AAT deficiency as a possible risk factor for the development of intracranial aneurysms (IAs), but results are equivocal. Among 362 consecutive patients with AAT deficiency seen at the Mayo Clinic during a 10-year interval, 3 had suffered an aneurysmal subarachnoidal hemorrhage, which is considerably higher than would be expected by chance (152). Alpha 1-antitrypsin deficiency (PI*MZ, MS, and ZZ) was found to be more frequent in patients with IA than in the general population (153), but other studies have failed to find a significant excess of AAT deficiency in patients with IA (154–156). Furthermore, the pattern of inheritance seen in familial IA does not fit any single Mendelian model, suggesting a genetic heterogeneity and possibly a strong influence of environmental gene interactions (157). It appears that AAT deficiency does not constitute a major genetic risk factor in IA formation. However, a possible pathogenetic role for a disturbed protease–antiprotease balance in formation and rupture of IAs has not been excluded and further study seems warranted. Screening for AAT deficiency as a marker for asymptomatic IA cannot be recommended.

One report (158) described elevated serum leukocyte elastase levels both in patients with ruptured and unruptured aneurysms. The elevated levels (the sum of free and complex elastase) correlated with degree of elastin degradation (154) in the superficial temporal arteries, suggesting a role for elastase in systemic elastin degradation including aneurysm formation. The presence of high elastase levels in patients with unruptured IA could not be confirmed in a smaller Japanese report (156). These authors suggested that the increased elastase levels in patients with ruptured aneurysms is attributable to the leukocytosis transiently occurring after a subarachnoidal hemorrhage. The small size of the study population, sex bias (156), and different methods for elastase assay may contribute to the conflicting results.

Another group of papers focuses on the putative role of acquired AAT deficiency in the development of IA. In formation of multiple aneurysms, cigarette smoking is a definite risk factor (159). Reduced activity of AAT, because of oxidation of methionine in the reactive center, might explain the increased risk of IA rupture in smokers (160). Similarly, a reduced functional capacity of AAT may be due to low levels of antioxidants (161) and increase the risk of aneurysm bleeding.

Schievink and coworkers at the Mayo Clinic have repeatedly reported on the possible link between AAT deficiency and both intra- and extracranial arterial dissections and fibromuscular dysplasia. Their first case reports on these associations (162–164) have been extended more recently (165, 166). In four consecutive patients with subarachnoid hemorrhage due to spontaneous intracranial arterial dissection seen over 2 years, two were PI*MZ, one was PI*MS, and one was PI*MM. In another study (166), AAT phenotyping was performed in three consecutive patients (all smoking females) who underwent bypass surgery for extracranial arterial dissection of the extracranial internal carotid artery. Two had the PI*MZ phenotype. The authors suggest that this phenotype may predispose to the development of fibromuscular dysplasia. The figures are suggestive, but any causal link between these rare conditions and AAT deficiency must be considered conjectural until larger studies with adequate control subjects become available.

Necrotizing panniculitis, characterized by inflammatory and necrotizing lesions of skin and subcutaneous tissue, represents the least common but well-recognized complication of AAT deficiency. The first case (probably an MZ individual) was published in 1972 (167). Rubinstein and coworkers reported two cases in PI*ZZ individuals in 1977 (168). To date, about 40 cases (some of them reported more than once) have been reported (169). In two-thirds of the cases, severe deficiency, PI*ZZ, has been present, but PI*MZ, SS, and even MS phenotypes have been seen. Familial occurrence has been reported (170). Alpha-1 antitrypsin deficiency–associated panniculitis has sometimes (167, 168, 170) been referred to as Weber-Christian panniculitis, a condition characterized by relapses of a febrile, nonsuppurative panniculitis that may have numerous underlying causes. In contrast, AAT deficiency–associated panniculitis should be considered as a distinctive form of panniculitis, leading to spontaneous ulcerations and drainage of the lesions and having fairly distinct histopathological features.

Typically (171), the panniculitis starts with painful, hot, red, tender nodules on thighs and/or buttocks in a young adult, AAT-deficient individual (mean age, 40 years). There is an equal sex distribution of affected individuals. Subsequently, ulcerations occur with drainage of clear, yellow, oily, odorless fluid, sterile at culture. In approximately one-third of these patients, trauma may have precipitated the disease (171–173). For appropriate histopathological evaluation, deep excisional specimens, with a large amount of tissue, are required (171). Fat necrosis is frequent, as is the presence of normal-appearing fat juxtaposed to inflammatory and necrotic panniculus. Fragmentation and loss of elastic tissue in areas of inflammation are frequently observed but vasculitis, except that resulting from necrosis, is infrequent (171, 172). The prognosis in these patients is variable and partly dependent on the presence of other AAT deficiency-related complications such as cirrhosis or emphysema. The panniculitis can be lethal.

Necrotizing panniculitis in AAT deficiency is an unusual complication. In a World Health Organization report (174), a prevalence of less than 1 case per 1,000 was mentioned. If the recommendation (174) to analyze plasma AAT levels in all cases of biopsy-proven severe panniculitis is followed, particularly in factitious and necrotizing cases, this rate of prevalence may well increase considerably.

For obvious reasons, no controlled trials that provide a basis for clear treatment recommendation are available. Family screening and antismoking counseling are essential. Corticosteroids, antibiotics, or cytostatic drugs appear useless. Augmentation therapy with purified human AAT or fresh frozen plasma to restore plasma and local tissue levels of AAT appears rational, safe, and effective (175–178). Restoration of plasma AAT levels after liver transplantation led to permanent cure in one case (169). Dapsone, either alone in less severe cases, or combined with augmentation therapy, may be of additional value. Its mechanism of action in this setting is, however, unclear (175).

In a 1996 report (179) of 14 cases (8 Swedish and 6 cases reported in the literature) of systemic necrotizing vasculitis in patients with severe AAT deficiency, skin involvement was present in all 14. In this cohort, characterized by multiple organ involvement (median number of organs affected was 8) and fatal outcome, the majority of patients had biopsy-proven microscopic polyangiitis or Wegener's granulomatosis. The cutaneous abnormalities were heterogeneous: erythematous or necrotizing papules, vesicles, palpable purpura, subcutaneous nodules, erythematous plaques, and relapsing ulcerative panniculitis. Necrotizing and leukocytoclastic vasculitis were the most frequent histopathological findings. A few reports of isolated cutaneous vasculitis in PI*ZZ AAT deficiency have appeared. The first was that of a 2-year-old girl with generalized, nonpruritic eruption resistant to treatment (180). A case of persisting vasculitis in a 49-year-old PI*ZZ male responded dramatically to administration of purified AAT (181).

Studies of genetic markers in psoriasis led Beckman and coworkers (182) to phenotype AAT in 72 psoriatic patients. The frequency of the Z allele was found to be significantly (p < 0.001) increased. Despite a relatively high risk (172, 173), the association would, however, explain only a minor part of the etiology of psoriasis. Their results were confirmed in several reports from the United States (183–185). MZ psoriatics had an earlier onset and a more severe disease than did non-Z individuals (185). Electron microscopic features suggested a defective inhibition of proteolytic activity in MZ psoriasis (186).

In 1975, Doeglas and Bleumink first published findings concerning 92 patients with various types of chronic urticaria (187) in whom decreased values of AAT were found. This phenomenon was particularly present in patients with cold urticaria and acquired idiopathic angioedema. Their findings were confirmed by several groups (188–190) but criticized by others (191, 192). However, in 1985 Doeglas and coworkers definitely confirmed their early findings (193) by phenotyping 281 patients with chronic urticaria. The MZ phenotype was significantly more frequent in the urticaria group. Again, the Z allele predominated in the groups with cold urticaria and acquired angioedema. The mechanism underlying this predisposition in carriers of the Z allele is unknown.

It seems reasonable to postulate a role for a serum protease inhibitor such as AAT in the prevention of proteinase-induced damage of the pancreas. Case reports have described acute pancreatitis in individuals with AAT deficiency (194, 195) and hemorrhagic pancreatitis has occurred after endoscopic retrograde cholangiopancreatography (196), but no significant excess of deficient phenotypes could be seen in a larger study including 31 patients with acute pancreatitis (197). Neither could these authors find a link between chronic pancreatitis and AAT deficiency (197). Such an association was first postulated in a South African study including 110 patients with chronic pancreatitis (198). There was a significant excess of the phenotype MZ in these patients, mostly alcoholics, but there was no association between the MZ phenotype and idiopathic chronic pancreatitis. Conflicting reports regarding the association between AAT deficiency and pancreatitis continue to appear. For example, Seersholm and Kok-Jensen (199) have reported an association between PI*Z AAT deficiency and pancreatitis, whereas Witt and coworkers (200) have observed no association of AAT deficiency with chronic pancreatitis. In summary, the data relating exocrine pancreatic disease to AAT deficiency remain inconclusive.

Islet cell hyperplasia is a universal feature of AAT deficiency (201). It has been postulated that an accumulation of aggregated Z AAT in islet cells may provoke a sequential development of such hyperplasia from benign to malignant tumors. Against this background, it is of interest that two endocrine pancreatic tumors, one benign and one malignant, were observed in the Malmö, Sweden series of individuals with AAT deficiency (202).

Alpha-1 antitrypsin immunoreactivity was demonstrated by immunofluorescence in epithelial cells of normal human small intestine (203). A case of AAT deficiency (probably PI*Z) in an elderly man with emphysema, cirrhosis, and total intestinal mucosal atrophy was reported in 1975 (204), supporting earlier observations of AAT levels compatible with heterozygous AAT levels (no phenotyping) in children with celiac disease (205, 206). However, later and larger studies from Spain and Ireland refute an association between AAT deficiency and celiac disease (207, 208).