Treatment of lung disease in alpha-1 antitrypsin deficiency: a systematic review

Background Alpha-1 antitrypsin deficiency (AATD) is a rare genetic condition predisposing individuals to chronic obstructive pulmonary disease (COPD). The treatment is generally extrapolated from COPD unrelated to AATD; however, most COPD trials exclude AATD patients; thus, this study sought to systematically review AATD-specific literature to assist evidence-based patient management. Methods Standard review methodology was used with meta-analysis and narrative synthesis (PROSPERO-CRD42015019354). Eligible studies were those of any treatment used in severe AATD. Randomized controlled trials (RCTs) were the primary focus; however, case series and uncontrolled studies were eligible. All studies had ≥10 participants receiving treatment or usual care, with baseline and follow-up data (>3 months). Risk of bias was assessed appropriately according to study methodology. Results In all, 7,296 studies were retrieved from searches; 52 trials with 5,632 participants met the inclusion criteria, of which 26 studies involved alpha-1 antitrypsin augmentation and 17 concerned surgical treatments (largely transplantation). Studies were grouped into four management themes: COPD medical, COPD surgical, AATD specific, and other treatments. Computed tomography (CT) density, forced expiratory volume in 1 s, diffusing capacity of the lungs for carbon monoxide, health status, and exacerbation rates were frequently used as outcomes. Meta-analyses were only possible for RCTs of intravenous augmentation, which slowed progression of emphysema measured by CT density change, 0.79 g/L/year versus placebo (P=0.002), and associated with a small increase in exacerbations 0.29/year (P=0.02). Mortality following lung transplant was comparable between AATD- and non-AATD-related COPD. Surgical reduction of lung volume demonstrated inferior outcomes compared with non-AATD-related emphysema. Conclusion Intravenous augmentation remains the only disease-specific therapy in AATD and there is evidence that this slows decline in emphysema determined by CT density. There is paucity of data around other treatments in AATD. Treatments for usual COPD may not be as efficacious in AATD, and further studies may be required for this disease group.


Introduction
Alpha-1 antitrypsin deficiency (AATD) is an inherited condition that predisposes patients to chronic obstructive pulmonary disease (COPD) 1 albeit with considerable variability of clinical phenotype. 2 The spirometric diagnosis of AATD-related COPD and "usual" (unrelated to AATD) COPD is identical, but it manifests at a younger age. Not all subjects develop pulmonary disease and those who do vary in presentation and subsequent decline. 3 The first subjects observed with low alpha-1 antitrypsin (AAT) levels were young smokers with basal panacinar emphysema, 4

Methods
The systematic review protocol is registered with PROSPERO (CRD42015019354). Standard systematic review methodology was used, aimed at minimizing bias, with reference to the Cochrane Handbook for Systematic Reviews of Interventions. 24

Search strategy
The following databases were searched by using no date or language restrictions (Supplementary materials): MEDLINE, MEDLINE In Process and EMBASE (via Ovid), Cochrane Library (Wiley) CENTRAL, CDSR, HTA, EED, and DARE. In addition, Conference Proceedings Citation Index via Web of Science and British Library's ZETOC were searched for conference proceedings and abstracts. ClinicalTrials.gov and World Health Organization International Clinical Trials Registry Platform were searched for ongoing trials. References of included studies and reviews were checked.
Study selection criteria Figure 1 shows the selection criteria. Systematic reviews and primary study designs assessing treatment were eligible provided they included data on $10 participants with AATD. Nonsystematic reviews and preclinical studies were excluded. Studies comprising mixed populations were permitted if AATD data were available separately. Any intervention for AATD-related lung disease was eligible for inclusion. The only restriction placed on outcome measures was of reporting .3 months after initiation of therapy, thus limiting to the most clinically relevant studies for this chronic disease. Studies which, for instance, measured solely AAT concentration before and after augmentation dosing were not included, and this criterion ensured such studies were not selected.

Data extraction and assessment of bias
Initial search yields (titles and abstracts) were screened for relevance by two reviewers independently, resolving any disagreement by discussion. Relevant articles were obtained and assessed against the full selection criteria, with translation of non-English language articles if required. Full-text articles were dually reviewed as before, with data extraction and bias assessment completed by one reviewer and checked by another. Missing data were requested from study authors if necessary (Supplementary materials). Reasons for excluding studies were documented and summarized in the PRISMA flow diagram (Figure 2). 25

evidence synthesis
Most data were suitable only for narrative synthesis, due to heterogeneity of study design and outcomes. Studies were divided into subgroups to reflect four "themes": Treatment of lung disease in alpha-1 antitrypsin deficiency AATD-specific, surgical, and medical treatments used in usual COPD, and other treatments. Data from three RCTs of AAT augmentation were meta-analyzed, the study design being sufficiently similar to allow this, [26][27][28] including changes in lung density (measured by quantitative CT scan at least annually), FEV 1 % predicted of normal, DLCO, QoL (St Georges Respiratory Questionnaire [SGRQ]), and annualized exacerbation rate (measured at least 6 monthly). Assessment of clinical and methodological heterogeneity was used to determine a fixed or random effect model; fixed effects using mean difference was appropriate in all but DLCO where a standardized mean difference was used. Differences between baseline and follow-up were synthesized and annualized; I 2 and τ 2 statistics are reported where appropriate.

Results
Following the removal of duplicates, 7,296 records were reviewed with 52 trials, comprising 5,632 participants' data included in the final analysis ( Figure 2). The two systematic reviews identified were briefly described in the "Introduction" section and justified this new review; hence, they will not be discussed further. The complete study characteristics of the remaining studies are shown in the Supplementary materials.

AATD-specific treatment: augmentation
There were 26 eligible studies using a number of different product brands including Prolastin and Prolastin-C, Aralast, Zemaira, Trypsone, Respitin, and Glassia. Dosing regimes ranged from weekly to monthly with most using 60 mg/kg/week.

RCTs
Three international multisite RCTs were identified, [26][27][28] including a total of 320 participants (58 26 -180 28 per study), conducted over 2-3 years, with optional subsequent openlabel elements in two studies. 27,28 All studies used CT density change as an outcome, two utilizing it as an experimental measure 26,27 and one powering for CT density as the primary outcome (RAPID trial 28 ). Other outcome measures included QoL, spirometry, and gas transfer and COPD exacerbation rates. Two studies used a standard dose (60 mg/kg body weight) intravenous infusion weekly, 27,28 while the third used 250 mg/kg infused every 4 weeks. 26 Mortality data were only reported by RAPID 28 with one death on treatment and three on placebo. Adverse events (AEs; treatment related or not) were similar in the two studies that reported them; 27,28 one did not report AEs. 26 The remaining outcomes were possible to meta-analyze and are summarized in Figures 3 and 4. Heterogenicity was not detected (all I 2 =0%). In all studies, volume-corrected CT scans at total lung capacity were performed, this being the preferred and most validated scanning method, 27,29-31 though three scanning methods were reported in RAPID. Lung density was analyzed by regression in the study by Dirksen et al 26 and RAPID 28 with four analysis methods in EXACTLE. 27 EXACTLE's "method 1" 27 was used in this meta-analysis as it utilizes the same regression technique as other studies; sensitivity analyses using other EXACTLE methods did not alter the outcome or significance. Annual deterioration in lung density was less on augmentation; difference 0.79 g/L/year (95% confidence interval [CI] 0.29-1.29; P=0.002, Figure 3), demonstrating a slower rate of emphysema progression. There was no significant difference in annual FEV 1 % predicted decline on treatment (0.56% predicted/ year [1.14-0.29; P=0.20]). A small nonsignificant difference of -0.11 (-0.33-0.11; P=0.34) in DLCO was observed.
Annual exacerbation rate was not reported in the earliest study, 26 but there was a small, statistically significant increase in annual exacerbations on treatment (0.29/year [0.02-0.54; P=0.02], Figure 3) when meta-analyzing the other studies. 27,28 Small and nonsignificant changes in health status were observed in both groups, 27,28 demonstrating greater worsening in SGRQ on placebo 0.83 (-3.55 to 1.89; P=0.55, Figure 4).

Observational controlled studies
There were six eligible controlled observational studies, comprising 2,610 participants. AEs and reasons for starting/ stopping therapy were reported by one registry; severe events occurred at a rate of 9.5% (69/720 infusions). 32  25 Articles were excluded where there were insufficient data in the public domain for the study to be assessed against the inclusion criteria including data from ClinicalTrials.gov. Abbreviation: COPD, chronic obstructive pulmonary disease.

1299
Treatment of lung disease in alpha-1 antitrypsin deficiency The largest observational study analyzed data from 1,129 patients in the US AATD registry split into three groups "always receiving" (n=390), "partly receiving" (n=357), or "never receiving" (n=382) augmentation. 33 Dosing was not standardized with only 51.3% being dosed weekly throughout the study. A survival analysis was conducted, but excluded 81 subjects (55 deaths) due to missing data, such that results could have been biased. Overall mortality was 18.1% (n=204); it was significantly higher for subjects who never received augmentation therapy (as opposed to sometimes or always) when FEV 1 ,50% predicted (P,0.001). Mortality rates were low for other subjects and did not differ between augmentation therapy groups. FEV 1 decline was calculated using a slope equation in 927 patients with n=202 excluded due to insufficient data; patients receiving augmentation with mean FEV 1 values of 35%-49% predicted had a slower rate of FEV 1 decline (-73.7±6.8 vs -93.2±8.9; P=0.03), though this was not seen in the whole group.
Three other studies investigated the effect of AAT augmentation on FEV 1 decline. [34][35][36] Seersholm et al 34 undertook a nonrandomized surveillance study in two cohorts. A statistically significant difference in FEV 1 annual decline was observed (-53 [48][49][50][51][52][53][54][55][56][57][58] vs -75  mL/year in treatment vs placebo; P=0.02). 34 The other two studies concurred with this result. Wencker et al conducted a prepost study of augmentation, using inclusion criteria of $2 lung function measurements prior to augmentation and two following commencement of therapy within a minimum period of 12 months. 36 35 Again, FEV 1 decline was worse in untreated patients (+10.61±21.4 vs -36.96±12.1 mL/year; P=0.05). All three studies stratified patients to groups by their FEV 1 at presentation -FEV 1 ,30%, 30%-65%, and .65% predicted. [34][35][36] Patients with FEV 1 ,30% were consistently observed not to benefit from augmentation in terms of FEV 1 decline. Two of the three studies showed those with an FEV 1 .65% to have statistically significant reductions in FEV 1 decline when on augmentation (-122.5±108.4  35,36 Treatment effect was restricted to patients with FEV 1 .30% and ,65% (-62 [57][58][59][60][61][62][63][64][65][66][67] vs -83  mL/year; P=0.04) in one study. 34 When FEV 1 at commencement of therapy was used to group patients, a statistically significant decrease of rate of decline during treatment was seen if FEV 1 ,30% (53.4±45.3 to 22.1±16.0 mL/year; P,0.0001). 36  The most recent study by Barros-Tizon et al 37 was a retrospective medical records review of 127 participants evaluating the effect of augmentation on exacerbation rate. Seventy-five patients had $1 exacerbation during the 18 months follow-up required prior to commencement of augmentation. Dosing regimens were not standardized with an average dose of 60.7±3.8 mg/kg/week (Table S1) different from other studies, 33 and many patients had missing data. Controlled studies of differing dose and/or drug manufacturer Five RCTs 38-42 with 176 participants investigated differences between dosing regimens of a product 38 or made comparisons between different products. [39][40][41][42] Campos et al 38 used a double-blind crossover study of 60 versus 120 mg/kg/week of Prolastin-C to investigate safety and pharmacokinetics in 30 patients. They reported no increase in AEs, a higher trough serum AAT concentration, and that the higher dose was tolerated well by patients. Four double-blind RCTs [39][40][41][42] completed investigating pharmacokinetic equivalence between weekly dosing of the new investigational product (Respitin, 42 Zemaira, 41 Prolastin-C, 39 and Glassia 40 ) and control (60 mg/kg/week Prolastin). All the studies had an optional open-label continuation study for safety data collection (ranging from 8 weeks 39 to 2 years 41 ) and reported no significant difference in AEs between the investigational product and control. Pharmacokinetic equivalence was reported in all studies with none demonstrating superiority over Prolastin. Stoller et al, 42 and Sandhaus et al 40 reported no significant differences in lung function parameters between the investigation groups or over the duration of the studies. One death was reported (respiratory arrest related to COPD) in the study by Stocks et al, 41 which was in the control arm and considered unrelated to study medication.

Observational uncontrolled studies
Twelve uncontrolled observational studies using intravenous AAT augmentation in 2,526 patients were included, which assessed safety and tolerability, [43][44][45] pharmacokinetics, [46][47][48] pulmonary neutrophilic inflammation, 49 longitudinal change in lung function, [50][51][52] and clinical characteristics or recipients. 53,54 AAT augmentation was reported as safe and well tolerated (one study drug-related AE out of 555 doses) 45 with 26 AEs probably/possibly related to the study drug, 43,45,47 largely comprising symptoms commonly observed in infusions of other protein-based products. 52 Other adverse reactions reported 43,49 were one widespread skin reaction, 49 headache and increased shortness of breath, and hospital

1301
Treatment of lung disease in alpha-1 antitrypsin deficiency treatment for fever, hypotension, and hypoxemia. 43 In all, 61 of 2,526 patients died with none attributed to AAT augmentation. 46,51,52 Augmentation was again well tolerated and deemed safe in a small study of 60 mg/kg/week versus 240 mg/kg/month. 50 Biochemical efficacy at the higher dose given monthly did not demonstrate a protective AAT level in 3 of 16 patients. 50 A second study investigating the pharmacokinetics of an alternative dose (Prolastin, 120 mg/kg every 2 weeks) was completed in 23 participants; 46  Similarly, effective concentrations of AAT were reported by two short-term studies 43,44 and one 3-year study. 51 A study investigating 18-fluorodeoxyglucose (18FDG) positron emission tomography-CT as a novel noninvasive biomarker 49 had a subgroup of 10 patients with severe AATD. The study demonstrated that 12 weekly infusions of AAT augmentation therapy had no effect on circulating neutrophil 18FDG uptake and activity in the lung.
Pulmonary function was reported in five studies with no significant changes in rates of decline or spirometry being observed. 46,47,[50][51][52] An observational registry study where longitudinal FEV 1 follow-up was available in 287 patients over 37.8±18.9 months showed significant differences in decline in FEV 1 ,30% predicted and FEV 1 30%-65% groups (-35.6±21.3 vs -64.0±26.4 mL; P=0.0008). 52 AlphaNet in USA is a not-for-profit health management company, which undertook a prospective study involving 922 members on augmentation, using monthly telephone interviews to collect data including exacerbation history and QoL (SGRQ) at baseline, months 6, and 12. 53 Totally, 2,268 exacerbations (mean 2.45±1.3/subject/year) were observed, mostly graded to be of moderate severity. SGRQ did not change by the minimal clinically important difference (MCID) during the 12-month follow-up in any age group, though differences were seen according to age. The same cohort was used 54 to describe exacerbations in more detail. Mean duration was 17.4±11.4 days by symptom-based classification, with a trend for increasing frequency and duration as disease severity increased; Global initiative for chronic Obstructive Lung Disease (GOLD) stage I and II patients had shorter, fewer exacerbations than GOLD stage III and IV (both P,0.05). Annual hospitalization for exacerbations was decreased in 4/14 patients in one study where study entry annual rates were used to assess augmentation; however, the study was insufficiently powered to establish the effectiveness in modifying disease. 43

COPD medical management
There were only two eligible studies in AATD using traditional COPD medical management strategies. 17,55 The first reported good uptake and efficacy of influenza vaccination in the AlphaNet cohort (n=939), 55 with 766 (81.6%) patients receiving vaccination, who then had fewer unscheduled outpatient department visits and critical care admissions (P=0.04 and 0.01, respectively). No statistically significant differences in exacerbation frequency or respiratory outcomes were observed. Another observational study by AlphaNet investigated the use of a multimodal self-management program with a combination of directed patient self-education, organized supervision, and health care provider education (Alpha-1 Disease Management and Prevention Program) in 1,028 participants. 17 This 2-year study used data from the first "observational" year as comparator to a second year when the management plan was delivered. A total of 905 participants completed the 2 years demonstrating low attrition. There was strong evidence to support improved compliance in some medications (long-acting beta agonist [LABA] P,0.001, theophylline P=0.01, and systemic steroids P=0.02) and supplementary oxygen (P,0.01) along with reductions in annual exacerbation frequency and duration (P,0.001 and P=0.04, respectively). There were no significant changes to health status.

COPD surgical management Lung volume reduction (LvR)
Six studies investigated the use of Lung Volume Reduction Surgery (LVRS) in AATD. Five studies (n=71 patients) used an open surgical technique, [56][57][58][59][60] and all demonstrated improvements in either physiological measurements or dyspnea. Benefits were inferior and shorter in duration than usual COPD patients in all studies. One small RCT randomized participants to LVRS (n=10) or medical treatment (n=6); higher 2-year mortality (20% vs 0%) occurred in the surgical group, albeit alongside improvements in SGRQ. 59 There was one published study using endobronchial valves, which demonstrated their safety in AATD patients with significant benefits in mean FEV 1 at 6 months, 1, and 2 years (P=0.0022, 0.0067, and 0.033, respectively). 61 The generalizability of this study is not evident as this cohort included strict inclusion criteria submit your manuscript | www.dovepress.com Dovepress Dovepress 1302 edgar et al including severe heterogeneous emphysema demonstrated by CT scan and scintigraphy, residual volume $140%, FEV 1 15%-45%, and optimal lobe selection. This resulted in fewer than half of the referrals meeting these criteria. 61

Other management
A double-blind, parallel-group RCT of Palovarotene to reduce inflammation and promote structural repair in the lungs failed to demonstrate significant differences in change from baseline for CT density, FEV 1 , or gas transfer measures and exacerbation. 73

Unreported trials
A study of an inhaled AAT augmentation therapy (NCT01217671) is listed as completed, but limited data are currently available. 74,75 One registered study (NCT00242385) 76 compared two augmentation products in a double-blind, crossover RCT. Only one outcome (AEs) was eligible for inclusion as follow-up for all other outcomes were limited to 35 days; similar safety profile and no serious AEs were seen, but no statistics were available. A second study compared two Prolastin-C levels at 60 and 120 mg/kg to assess safety and pharmacokinetics (NCT01213043). 77 No statistical analysis was available but 30 participants completed the study with similar treatment-emergent AEs and safety profile to other studies; no serious AEs occurred. There is currently one study (NCT01983241) 78 investigating 60 and 120 mg/kg weekly administrations of Prolastin-C against placebo to determine safety and efficacy using change from baseline in CT lung density as the primary outcome. 79

Risk of bias
Risk of bias for the meta-analyzed RCTs is summarized in Figure 5A. Risk of bias in RAPID 28 was low; however, in the two earlier augmentation studies, this was more difficult to assess and generally unclear due to poor reporting, specifically of details such as allocation concealment. The other six RCTs demonstrated low or unclear bias, mostly due to lack of detail in the manuscripts, commonly around blinding. Figure 5B summarizes the risks in the remaining studies. The mixture of RCTs and cohort and observational studies (both prospective and retrospective) led to an overall moderate risk of bias especially in the nonblinded studies. In the controlled or quasi-controlled studies, there was often insufficient reporting of the blinding process and allocation concealment, except in more recent studies, and a trend toward selective outcome reporting. 18,36,38 Within the uncontrolled retrospective studies, selection bias was the main issue, with patients being taken from self-selecting groups or registries. Transplant studies from registries suffered from high attrition rates, which may have introduced survivor bias. [64][65][66][67][68][69]

Discussion
The available evidence for the management of AATD was largely centered on augmentation, which seemed to benefit emphysema, but with significant deficits in the evidence base for outcomes more typically assessed in COPD studies, such as exacerbation frequency and QoL. Although it is not unreasonable to think that strategies used to treat COPD unrelated to AATD could be extrapolated to AATD patients, there was almost no evidence proving this.

Augmentation
Augmentation was demonstrated to be safe and well tolerated in numerous studies using different products. Investigations into dosing regimes demonstrated few benefits from .60 mg/ kg/week 38 (the "standard" dose), and pharmacokinetic equivalence was observed between different manufacturers' products. 40,41 The meta-analysis confirmed that augmentation therapy is able to slow down the progression of severity of emphysema when measured by CT density change compared with placebo (P=0.002). By combining data from up to 320 patients included in three RCTs, it adds to an increasing body of literature demonstrating the usefulness of CT densitometry as a surrogate measure of emphysema and outcome in both AATD and usual COPD. 20,80,81 However, the MCID has not yet been established for change in CT density, which would be helpful for interpretation. Meta-analysis also revealed a small but significant increase in annual exacerbation rate on augmentation. This is somewhat counterintuitive and contradicted by studies included in narrative synthesis, which either showed a benefit or no significant difference in exacerbation rate or severity. 17,37,54 Increased contact with health care professionals, simply due to attending for infusions, could lead to increased reporting of exacerbations, but this is the opposite result from published observational studies. These findings require caution in their interpretation as none of the studies were powered to detect change in exacerbation rates as an outcome. Studies in usual COPD appropriately powered to detect change in exacerbation frequency often require several thousand patients. 82 With advances in technology and a reduction in the cost of digital patient symptom diaries, there are an increasing number of studies using them to collect contemporaneous symptom data, including a recent inhaled augmentation therapy (NCT01217671) trial. 75 This has been beneficial in usual COPD trials 83,84 as it avoids recall bias and gives accurate data on severity and duration; hence, the results will be of interest. The Spanish study that focused on exacerbations had issues with missing data and also used variable treatment regimens, such that bias and lack of comparability to RCTs are an issue. 37 There are several other reasons why the exacerbations result should be viewed with caution; first, prestudy exacerbation frequency was not reported in the three RCTs. Prior exacerbation rate predicts future exacerbation rate; 85,86 hence, if a difference was present pretreatment, it could account for the observation posttreatment. Second, power calculations were not based on detecting change in exacerbation frequency. Nevertheless, this area will be important for further study because faster decline in lung function and significantly increased health care costs occur due to frequent exacerbations in usual COPD, 87,88 an effect that might be more marked in AATD due to higher inflammatory burden during events. 89 Most observational controlled studies demonstrated some differences in FEV 1 decline with recipients of AAT augmentation benefiting over those not receiving therapy. 34,35,52 However, such study designs have weaknesses, often relating to differences between treated and untreated groups. For instance, sex, FEV 1 , and follow-up time differences were seen between treated (German) and untreated (Dutch) patients, 34 and age, FEV 1 , symptoms, inhaled therapies, and oxygen use differed in another study 35 indicating that biases are likely present, which could have influenced results. In most cases, attempts were made to adjust for major differences statistically when comparing groups, and subgroup analyses suggested that benefits did not accrue in patients with FEV 1 ,30% predicted. However, since it has been demonstrated that gas transfer declines more rapidly at this stage, 86 it may simply be that different outcome measures were needed to demonstrate any effect. In one study, there was evidence that rapidly declining patients benefited more; 52 however, there were only 11 such patients, and the definitions of slow and rapid decline utilized are not widely adopted in practice.

COPD medical management
There was limited evidence about the efficacy of COPD treatments in AATD, and no trials reported the effects of typical treatments such as inhaled bronchodilators (long acting muscarinic antagonist, LABA, etc), steroid combinations (LABA/inhaled corticosteroids [ICS]), or PR. The available evidence was of poor quality, being prone to acquisition and potentially other biases, due to its retrospective, observational design. However, the results for influenza vaccination and self-management were at least suggestive of clinical benefit to AATD patients, thus consistent with the usual COPD literature. 22 There is no biological reason to suppose that such interventions would differ in efficacy between usual COPD and AATD.
However, data from the uncontrolled studies 54 and sputum work 89 suggest that exacerbations are longer and more proinflammatory in AATD, such that treatments targeting their rate (eg, LABA/ICS) might be more beneficial in AATD. Recently, an unfavorable muscle response to exercise has been shown in a small number of AATD patients, compared with usual COPD, 90 suggesting that evidence about PR might not be wholly generalizable to AATD either. This is backed up by a study of a 5-week PR program pretransplantation, where AATD patients were observed to have a smaller improvement in 6-min walking test distance compared with usual COPD (47.9 vs 60.6 m). 91 However, the PR course was short, and the study was not eligible for inclusion in this review due to short follow-up duration; furthermore, no difference was seen after statistical adjustment for confounders. Hence, uncertainty exists about the value of COPD treatments, and further work is needed.

COPD surgical management
AATD patients receiving LVRS demonstrated benefits, albeit inferior when compared with usual COPD. In the single study, compared to medical care (as in the National Emphysema Treatment Trial study), there were more deaths in surgically treated patients, even though their QoL improved. However, underpowering meant there were no meaningful statistics and the results should be viewed with caution. Higher mortality could reflect differences in the distribution of emphysema in AATD, making LVRS technically more difficult because access to the lung bases is needed. 92 One small study demonstrated safety and benefits lasting up to 2 years from endobronchial valves in carefully selected patients -further investigation to reproduce these findings is warranted; however, it is encouraging and suggests that volume reduction might yet be a viable strategy. There is no evidence for endobronchial coiling in AATD yet, though suggestion of benefit in more homogeneous disease 93 implies they could be valuable.
When compared to usual COPD, AATD lung transplantation recipients had improved survival, but this was potentially confounded by differences in age and comorbidity. 68 Earlier onset of severe disease is well documented, and therefore, a survival benefit posttransplant may not be surprising. Two studies comparing AATD patients receiving transplant or not 71,72 had conflicting results with respect to survival benefit, despite selection criteria for transplantation being similar in both countries (Sweden and UK). It is possible that more stringent matching procedures in the UK cohort 72 explain the difference. Nevertheless, significant health status benefits occurred after transplant indicating that it is appropriate when QoL is poor. However, unlike pulmonary fibrosis where there is unified evidence of survival benefit (in part due to the poor prognosis of disease), 94 uncertainty regarding survival benefits should be discussed with patients and is not only unique to AATD but also apparent in usual COPD and may represent the heterogeneity of disease. 94 The single study that demonstrated decline in AATD transplant recipients included only five patients and a time frame (2 years) for the calculation of FEV 1 decline, which is too short to be accurate; hence, the results may not be reliable. 62 Increased posttransplant gastrointestinal complications requiring laparotomy were demonstrated in one study 63 with those patients having a longer intensive care unit stay. There was no statistically significant effect on mortality or duration of mechanical ventilation observed, International Journal of COPD 2017:12 submit your manuscript | www.dovepress.com

1305
Treatment of lung disease in alpha-1 antitrypsin deficiency and the authors note that this was a small single-center retrospective study.

Conclusion
There is good evidence from this systematic review that intravenous augmentation therapy slows decline in emphysema determined by CT density. This supports its use as a surrogate end point and demonstrates that augmentation remains the primary disease-specific therapy. There is paucity of data around other treatments in AATD including potential bias in the selection and reporting of clinical trials. As augmentation therapy is an expensive treatment, a full economic review is needed, and further work on optimizing patient selection for therapy could help rationalize treatment in the UK. All treatments for usual COPD may not be as efficacious in AATD due to important differences in disease process, and studies in specific treatments such as PR need to be appropriately powered to this disease group.