Disease Modification in Emphysema Related to Alpha-1 Antitrypsin Deficiency

ABSTRACT

Alpha-1 antitrypsin deficiency (AATD) is associated with premature onset of emphysema resulting from low serum A₁-PI levels. The only available pharmacological treatment affecting the underlying cause of AATD is A₁-PI therapy. AATD-related emphysema is considered a good model to study disease-modifying effects of treatment as the causative process has been identified. Disease modification is a sustained improvement in disease state following therapeutic intervention that persists when therapy is discontinued. Appropriate trial design and the use of valid study endpoints are key to illustrating disease modification, particularly in clinical trials of rare diseases where it can be difficult to recruit sufficient numbers of patients. Delayed-start trials are advantageous ethically as all patients ultimately receive active treatment and imaging techniques have proven promising as valid study endpoints. Specifically, computed tomography (CT) measured lung density has been used to monitor emphysema and is considered a more sensitive outcome than pulmonary function tests to monitor disease progression. This review will discuss the importance of clinical endpoints and trial design to determine disease modification and will review the evidence for disease modification in AATD-related emphysema. Until recently, clinical studies have not shown a significant effect of A₁-PI therapy, possibly due to insufficient numbers of patients, short duration of clinical trials and lack of appropriate trial design. A recently completed randomised trial and open-label extension study followed a larger study population for a longer duration and incorporated a delayed-start design. The results demonstrated clinical efficacy of A₁-PI therapy and indicate that treatment is disease-modifying.

KEYWORDS:

• Disease modification
• alpha-1 antitrypsin deficiency
• emphysema
• A₁-PI therapy
• computed tomography

Introduction

Alpha-1 antitrypsin deficiency (AATD) is a co-dominant, hereditary condition caused by mutation of the SERPINA1 gene which encodes alpha-1 antitrypsin (also known as alpha₁-proteinase inhibitor [A₁-PI]). It is known to affect around 3.4 million individuals worldwide, although many of them are not yet diagnosed (1). Severe AATD is characterised by low levels of A₁-PI in blood and tissues and can lead to liver disease (cirrhosis or hepatocellular carcinoma), caused by aggregation of misfolded A₁-PI as polymers in hepatocytes, and/or chronically progressing lung disease, including emphysema, as a result of uncontrolled degradation of lung tissue by proteases, mainly neutrophil elastase (NE) (2, 3). Severe AATD presents with pulmonary symptoms, often including coughing, wheezing, shortness of breath and recurrent respiratory tract infections that are similar to those observed in typical chronic obstructive pulmonary disease (COPD) (4–6). Whilst the pathomechanism of AATD-related COPD is distinct from COPD per se, recent evidence suggests that in both emphysema is a major contributor (7, 8). As the causative process has been identified, AATD-associated emphysema is considered a good model to study a possible disease-modifying effect of therapy on the course of the disease. Disease modification can be broadly defined as the sustained change in a disease state that occurs following therapeutic intervention (9), whereby the progression of emphysema is slowed. None of the currently approved pharmacological therapies for COPD has unequivocally demonstrated this effect.

The main target of A₁-PI replacement therapy is to recover the normal physiological balance between proteases and antiproteases, and interrupt the cascade that perpetuates active pulmonary disease. In recent years, clinical trials have sought to provide evidence for clinical efficacy of A₁-PI replacement therapy in patients with emphysema due to AATD. Following several inconclusive trials, the recent publication representing the differently designed and largest clinical trial completed to date provides clinical data on efficacy of A₁-PI therapy and, importantly, on its disease-modifying effect (10). This review will discuss the importance of clinical endpoints and trial design to determine disease modification and will review the evidence for disease modification in AATD. We will demonstrate that careful, unorthodox trial design is of key importance when studying relatively rare disorders, such as AATD, and slowly progressing diseases, such as emphysema, especially in the absence of sensitive and easily measurable biomarkers.

Disease modification in AATD-related emphysema

Disease modification can be described as an improvement or stabilisation of a disease state resulting from a reduction in the rate of disease progression that occurs following therapeutic intervention, which may persist after the intervention is discontinued (9). As such, disease-modifying therapies should impact on the pathological and pathophysiological mechanisms underlying the disease rather than the clinical or symptomatic effects of the disease (11, 12). Although disease state can be influenced by therapeutic interventions in several ways, current consensus is that disease-modifying treatments are those that alter the rate of disease progression (Figure 1). Other treatments may provide symptomatic improvements that have either a sustained or temporary effect (Figure 1). Whilst symptomatic treatments may ‘turn back the clock’ and provide a sustained improvement in disease symptoms which may persist after treatment has ceased, these are not considered to be truly disease-modifying (9).

Figure 1. Impact of treatments on disease progression (9). This figure shows the hypothetical effects of symptomatic and disease-modifying treatments on disease progression in early-start and delayed-start trial models. Disease-modifying treatments provide a sustained alteration in disease progression. Reprinted from COPD: Journal of Chronic Obstructive Pulmonary Disease. Halpin and Tashkin. Defining Disease Modification in Chronic Obstructive Pulmonary Disease. COPD 2009: 6: 211–213 with permission of the publisher Taylor & Francis Ltd (www.tandfonline.com).

In AATD-related emphysema, as in typical COPD, symptomatic treatments such as bronchodilators or corticosteroids do not affect the underlying cause of the disease and the progressive destruction of lung tissue (3, 13). Only lifestyle modifications, such as smoking cessation, have demonstrated some effect on disease progression; however, it is not entirely clear to what extent and after what period the deteriorating effect of smoking ceases (14). The only available treatment for the underlying cause of AATD is treatment with A₁-PI, which was demonstrated by early biochemical studies to be able to partially restore A₁-PI levels in the serum (to above the 11-µM protective threshold) and importantly in the epithelial lining fluid of the lungs (15, 16). Consequently, it has been expected that therapy with A₁-PI might slow the rate of lung tissue destruction and reduce the frequency and severity of lung infections (15, 17, 18). Accordingly, retrospective analysis of data from the US AATD registry demonstrated significantly lower relative risk of death in A₁-PI treated subjects (RR = 0.64, p = 0.02) and slower rate of lung function loss (assessed by forced expiratory volume in 1 second [FEV₁]) in a subgroup with baseline FEV₁of 31–49% predicted (19). Within this analysis, study design was the major confounding factor. The observational registry study was not a prospective, placebo-controlled, randomised study and therefore it was susceptible to numerous limitations resulting from differences in socioeconomic factors, health delivery systems and varying length of treatment. Despite these limitations, a similar conclusion regarding the effect of A₁-PI replacement therapy on FEV₁ decline rate was provided by a smaller observational European study (20).

Subsequent randomised clinical studies, Danish-Dutch (21) and EXACTLE (22), also failed to demonstrate clinical efficacy of A₁-PI in AATD-related emphysema or provide evidence of disease modification. Both demonstrated a clear favorable trend on the lung density loss assessed by computed tomography (CT) densitometry, but no benefit on less sensitive endpoints, such as FEV₁, was observed. Additionally, both trials were not designed to address mortality. It was generally considered that both studies were underpowered with cohorts of 58 and 77 patients, respectively, and a much larger group of patients (150–500) would be necessary to demonstrate efficacy of A₁-PI therapy within the classic double-arm, randomised design. This requirement would be difficult to fulfill considering the relatively low prevalence of AATD and the low number of potentially available patients with AATD-related emphysema. As in other orphan disorders, obtaining the optimal sample size in AATD clinical trials proves difficult. Hence, careful trial design and selection of clinical endpoints seems to be of crucial importance as demonstrated by recently published data (10).

Trial design in disease modification studies

In order to demonstrate a disease-modifying effect in AATD and other chronic and slowly progressing diseases, the design of the clinical trial is important. Standard trial models, such as parallel group randomised controlled trials where patients receive either placebo or active treatment, are characterised by minimal bias and therefore considered the gold standard. However, in addition to ethical issues around running a long-duration placebo-controlled trial, particularly in rare disorders such as AATD, they do not provide clear distinction between temporary symptomatic and disease-modifying effects.

Studies of disease modification performed in other fields have led to a greater understanding of the concept of disease modification and consequently improvements in trial design. Disease modification has been most extensively studied in neurological diseases such as Alzheimer's disease, Parkinson's disease and multiple sclerosis which, like AATD, are chronic, slowly progressing diseases. Some of these studies utilised a withdrawal trial design where patients are randomised to either placebo or active treatments and subsequently all patients are switched to placebo (23, 24). For example, in one study patients with Parkinson's disease were followed for 3 months during washin and for 2 months during washout (withdrawal period) (25). However, evaluation of disease-modifying effects was confounded by short-term symptomatic effects and the short duration of washout was insufficient to demonstrate if any effects of treatment were sustained and disease-modifying. This is an important issue to be considered as the majority of disease modification studies follow an add-on design, i.e., concurrent symptomatic optimal therapy is not modified.

More recently, delayed-start trial designs have been utilised to demonstrate disease modification, which have two phases. In the first phase patients are randomised to receive either placebo or active treatment, as in classical placebo-controlled, parallel group study design, but differences in study endpoints between the two groups at the end of this phase could be due to symptomatic effects or short-term benefits and not due to disease modification. In phase 2, both groups receive active treatment; if a difference in a study endpoint remains at the end of this stage a treatment may be considered to be disease-modifying (23, 24). This trial design has been used effectively to study the effect of rasagiline, a monoamine oxidase type-B inhibitor, in the treatment of Parkinson's disease (26).

Expert consensus and guidance provided by the US Food and Drug Administration (FDA) suggests that trial designs such as withdrawal or delayed-start represent appropriate means to illustrate sustained effects on disease progression (27–29). Ethically, delayed-start trial designs may be more appropriate and within such trials disease modification is shown if patients initially on placebo and switched to active treatment fail to ‘catch up’ with patients who received active treatment for the entire duration (Figure 1).

Endpoints in disease modification studies

In addition to careful clinical trial design, key considerations for disease modification studies include the use of reliable biomarkers that correlate with clinical and/or surrogate endpoints, and the use of highly sensitive and reliable endpoints.

Guidance on the use of biomarkers and study endpoints has been provided by the FDA (30). To date numerous biomarkers are under development for use in cardiovascular disease, infectious diseases, cancer, neuropsychiatric disease, and autoimmune and inflammatory disease (30). In asthma, β-adrenergic receptor polymorphisms have been shown to predict short-term patient outcome (30) and in COPD, elastin degradation products are one of the many potential effective biomarkers currently under investigation (31). As demonstrated in the ECLIPSE trial numerous study outcome measures, including pulmonary function tests (32, 33), biomarkers (34, 35), health outcomes (32, 36) and exacerbations (32, 37) have been evaluated as markers in COPD. However, a key challenge in COPD is the identification of markers that can accurately measure disease progression, are responsive to treatment and are superior to the gold standard, FEV₁ (32, 38). The utilisation of imaging techniques as valid study endpoints has shown promising results. In particular, serial CT scanning to monitor quantifiable changes in lung density has been used to monitor disease progression in emphysema (30, 35).

Likewise, the study of disease modification in AATD-related emphysema is confounded by the lack of a standard definition and agreement of appropriate markers by which modification should be assessed. As in COPD, neither lung function (FEV₁) nor biochemical biomarkers (elastin degradation) provide adequate sensitivity to conclusively demonstrate the effect of A₁-PI therapy, or lack of it, within the standard double arm randomised trial. Furthermore, these outcomes along with CT lung density have been underpowered in previous clinical studies to show a treatment effect of A₁-PI therapy (17, 21, 22).

Problems with traditional outcome measures

FEV₁ has traditionally been used in patients with COPD to monitor expiratory flow limitation as a surrogate marker for disease progression. However, our understanding of the pathogenesis of COPD is evolving and evidence indicates that several pathological processes, including narrowing of small airways, bronchiolar fibrosis, and destruction of alveolar tissue, contribute to the clinical picture of COPD (39). As FEV₁ is a surrogate marker, it may not be the most appropriate measure to monitor disease progression. This may also explain why changes in FEV₁ have been shown to be poorly correlated with other outcome measures such as dyspnoea, exercise capacity, exacerbation and mortality (32). Furthermore, as changes in FEV₁ reflecting lung tissue destruction or remodelling typically occur slowly over time, and given the intra-subject variability of FEV₁ measures (40), a large and lengthy clinical trial would be required to demonstrate modest effects of treatment on FEV₁ (39,40). Demonstrating a therapeutic reduction in FEV₁ decline of around 20 mL/year would take several years and require the enrollment of at least 1,000 subjects (39). Consequently, in clinical studies to date only smoking cessation has been shown to have a significant impact on FEV₁ decline and alter disease progression (39, 40); whilst treatment with corticosteroids or bronchodilators have been shown to provide improvements in FEV₁ which may last for several years if continuously applied, they do not alter the progression of disease and decline in lung function in the long term (39, 41). Following initial trials conducted in the 1990s, it was generally accepted that FEV₁, then the gold standard endpoint in monitoring COPD progression, is not suitable in AATD research and no study on augmentation therapy could be adequately powered to demonstrate a significant treatment effect. A meta-analysis of five clinical studies and 1,509 AATD patients demonstrated a significant reduction in the decline in lung function with A₁-PI therapy. However, an important bias of this analysis was that only a single study containing 58 subjects was designed as a placebo-controlled trial (42).

Given the limitations of FEV₁ for monitoring COPD, several alternatives, including physiological parameters and patient-centric outcome measures, have been suggested as means to monitor COPD progression. Several surrogate outcome measures have been suggested for COPD as shown in Table 1 (9, 41, 43, 44); these include dyspnoea, health-related quality of life (HRQL), exercise capacity, exacerbations and mortality. Improvements in dyspnoea have been observed for several long-lasting bronchodilators, long-acting antimuscarinic agents and inhaled corticosteroids. Significant HRQL improvements have been observed with numerous treatment options, although these changes have not been clinically meaningful (41). Similarly, several treatments have been shown to improve exercise capacity, as measured by the shuttle walk or cycling test, and reduce the rate of common COPD exacerbations (41). Mortality rate was measured during both the TORCH (45) and UPLIFT (46) studies whereby treatment with salmeterol and fluticasone combined or tiotropium, respectively, reduced mortality rates in COPD patients although these changes did not reach significance (41, 45, 46). The body mass index, airflow obstruction, dyspnoea and exercise (BODE) index combines several outcome measures to provide a multi-dimensional tool to monitor disease progression in COPD (47). A recent study by de Torres et al. (48) illustrates that the BODE index is a useful tool in COPD that is more sensitive at predicting survival than ABCD GOLD categories or FEV₁ alone (48,49). In patients with severe emphysema, lung volume reduction surgery resulted in a decrease in the BODE index that correlated with increased survival and reduced mortality (50). In studies of AATD patients, utilisation of alternative pulmonary function measures, mortality outcomes, HRQL assessments and exacerbations has proven to be insufficient to demonstrate significant effects of A₁-PI treatment (10, 19, 22, 51). As with FEV₁, this is due to poor sensitivity of these measures to detect changes in AATD-related emphysema and the need for studies to be sufficiently powered, i.e., with larger patient populations and/or longer duration.

Table 1. Summary of outcome measures used to monitor COPD and emphysema progression (9, 39, 41, 43, 44, 69).

Outcome measure	Method/measurement	Rationale for use	Advantages	Disadvantages
Physiological parameters
FEV1	Spirometry	The traditional hallmark of COPD was respiratory flow limitation that declines over time	Repeatable measureAccessible/non-invasive	FEV1 changes slowly over time – not easily modified by interventionsIntra-subject variabilityPoor correlation with other endpoints (e.g., mortality, exacerbations)
Bronchoalveolar lavage (BAL)/bronchial biopsy	Measurement of inflammatory changes	Increase in inflammatory cells (e.g., neutrophils, macrophage and CD4/CD8 T lymphocytes) has been demonstrated as disease progresses	Accurate marker of disease progression	InvasiveLack of standardisation (BAL fluid)
Induced sputum	CytokinesChemokines	The presence of cytokines/chemokines is a marker of inflammation	Easy procedureWell-tolerated	Long-term reproducibility unknownPredominantly samples from large but not small airwaysProcedure can induce inflammatory responseSolubilisation procedure can interfere with other cytokine/chemokine radioimmune assays
Exhaled air	NOCOVolatile hydrocarbons (alkanes, pentanes, ethane)	Presence of inhaled gases such as NO, CO and hydrocarbons is a marker of inflammation	Readily accessibleRepeatable	Levels of NO/CO may be only slightly elevated in COPD Long-term reproducibility unknown. Lack of standardisationLittle known regarding correlation with long-term disease progression
Patient-centric outcomes
Dyspnoea	Baseline dyspnoea indexTransition dyspnoea index	Dyspnoea is a clinically important and recognised measure of respiratory and activity limitation	Simple to performRepeatableCorrelates to lung function and mortality	Dyspnoea worsens over time and changes due to intervention can be maskedStandardisation and validity have been questioned
HRQL	SGRQ	HRQL has been shown to worsen over time and show improvements with treatment	Simple to monitor	Reflects effects/symptoms of the disease rather than the disease itselfDepends on accurate patient reporting
Exercise capacity	6-minute walk testCycloergometry testConstant work rate submaximal exercise test	Activity levels have been shown to worsen over time with COPD progression and have shown variable improvements with treatment	Constant work rate submaximal exercise test is repeatable	Measures such as 6-minute walk test show poor repeatability due to learning effect and effort differences
Exacerbations	Patient and/or physician reports of frequency and/or severity	Exacerbations have been shown to worsen over time and show improvements with treatments	Easy to measure	Frequency can vary over time and with different disease stagesDepends on accurate patient reporting
Imaging techniques
CT	Lung densitometry	Accepted by the FDA as a highly sensitive measure for monitoring emphysema progression	ReproducibleHighly sensitiveCorrelates to other surrogate markers (e.g., FEV1, exacerbation, mortality)	ExpenseNot routine in hospital/clinicsEthical discussion regarding exposure to radiation
Mortality
Mortality rate	Physician report	Ultimate measure of disease progression	Useful for studies in disease populations	Not suitable for monitoring disease progression in individual patientsCo-morbidities may affect mortality rate

Outcome measures used in several studies to monitor disease progression in COPD and emphysema together with their advantages and disadvantages for use are shown.

CO, carbon monoxide; COPD, chronic obstructive pulmonary disease; CT, computed tomography; FEV₁, forced expiratory volume in 1 second; HRQL, health-related quality of life; NO, nitric oxide; SGRQ, St George's respiratory questionnaire.

Computed tomography

Whilst surrogate markers such as dyspnoea, HRQL and exercise capacity may be beneficial in monitoring disease progression in COPD, they can be affected by external influences such as co-morbidities and may not accurately represent pathological changes which may be ongoing in the lung. Instead alternative measures that can recognise the two major disease processes which affect disease progression, namely airway disease and alveolar disease, are more suitable to gain an insight into the disease-modifying behavior of a therapy.

Only highly sensitive, objectively assessed measures that allow a clear distinction between airway and alveolar disease might provide a satisfactory tool to assess therapeutic effect and provide evidence of disease modification. Traditional lung function measures are indirect by nature, e.g., FEV₁, forced vital capacity (FVC) and diffusing capacity of carbon monoxide (DLCO) assess the changes that result from emphysema progression but do not represent the changes in lung structure per se. Therefore these measures are not sensitive enough to detect emphysema progression over short periods of time, i.e., a couple of years. Imaging techniques, such as CT lung densitometry, can not only help to distinguish emphysema from other airway disease but can be used to quantify the severity of disease and may also be used to characterise the distribution of emphysema in the lungs. This methodology best reflects the pathological changes that occur in the lung with AATD-related emphysema. However, CT densitometry is not able to distinguish between lobular forms of emphysema, e.g., centrilobular and panlobular emphysema, and so for diseases such as COPD in general and cystic fibrosis, other CT-based methodologies are used, such as pattern recognition-based analysis and measuring bronchial wall thickening.

Compared to traditional lung function measures CT lung densitometry is a more reliable, reproducible and 2.5-fold more sensitive measure to monitor emphysema progression (51, 52). CT lung densitometry has been shown to correlate with other outcome measures including FEV₁, health status and exercise capacity (53, 54). Consequently CT lung densitometry has been accepted by the FDA as a clinically meaningful tool for assessing the impact of treatment on emphysema disease progression in AATD (55) and a 2003 statement from the ATS and ERS highlighted the need for clinical trials utilising CT (2).

In a clinical trial setting, CT methodology is standardised to ensure that data can be collected and analysed consistently between different centers. Despite this, in routine clinical practice there remains a requirement for standardisation, including alignment of measured parameters, CT scanner setup and software algorithms, and there is a lack of evidence to guide the use of different methodological approaches (52). Widespread use of CT lung densitometry is also limited by expense, availability, and the need for specialised software, e.g., for densitometry analyses, and experienced personnel to analyse the data (56, 57). Moreover, although CT has been shown to correlate with other clinical endpoints (53, 54), a clear correlation with mortality is lacking; presently, only a single study has suggested that CT may better predict mortality in AATD patients than lung function parameters (58). The lack of data demonstrating function or survival benefit arising from therapy has hindered widespread acceptance of CT as an outcome measure, and patients and health authorities frequently prefer cheaper, easily accessible outcomes with clearer physical effects, such as improvements in FEV₁.

Clinical studies of disease modification in AATD-related emphysema

Limitations of randomised placebo-controlled trial design

In order to demonstrate a disease-modifying effect in AATD-related emphysema, appropriate trial design and choice of study endpoints is critical. Lung function measures, i.e., FEV₁, have inherent methodological issues that limit their ability to demonstrate clinical efficacy of A₁-PI therapy: lung function generally declines slowly in emphysema and consequently demonstrating disease-modifying effects in a standard two-arm trial setup would necessitate unfeasibly long follow-up and large patient cohorts (21). To overcome this issue, the two relatively small randomised placebo-controlled trials were the first to utilise CT lung density as a more sensitive endpoint to assess the effect of A₁-PI therapy on emphysema progression (21, 22). First, in a study of Danish and Dutch ex-smokers with AATD the efficacy of A₁-PI therapy was assessed using CT scan measurements taken at forced residual capacity. Whilst the study demonstrated a protective effect of A₁-PI against loss of lung density, this did not reach statistical significance (p = 0.07) (21). Subsequently, the EXAcerbations and Computed Tomography scan as Lung End-points (EXACTLE) study followed 77 patients randomised 1:1 to A₁-PI or placebo over 2–2.5 years (22). The efficacy of A₁-PI therapy in AATD patients was assessed using CT measurements taken at total lung capacity (TLC) and the number of exacerbations. Similar to earlier studies a trend towards a beneficial effect of A₁-PI treatment compared to placebo was observed with CT densitometry measurements (p-values ranged between 0.049 and 0.084); however, no significant differences in the frequency of lung exacerbations were observed between the groups (A₁-PI = 2.55 ± 2.14; Placebo = 2.19 ± 1.33, p = 0.265) although the severity of exacerbations was reduced (22, 52). Although results from the two studies did not reach significance when analysed alone, the trials were of comparable design and a combined analysis of results from these two studies suggested that A₁-PI therapy may significantly reduce the decline in lung density in patients with AATD (p = 0.006) (54).

It is generally acknowledged that both trials were not adequately powered to demonstrate a treatment effect as they were not conducted over a sufficient period of time, and did not include enough patients (21). Consequently, disease-modifying effects of A₁-PI therapy could not be ascertained. However, it is important to note that the designs of the Danish-Dutch and EXACTLE studies of disease modification in AATD did not include delayed or withdrawal periods and therefore may not have been optimal (21, 22). Furthermore, the findings from the Danish-Dutch and EXACTLE studies highlight that standard trial designs are often inadequate to demonstrate disease modification. From these studies it was concluded that in order to determine whether A₁-PI therapy demonstrates a significant protective effect, patients would need to be followed for a longer duration ≥4 years and a greater number of patients (150–500) would need to be included.

When planning clinical trials in orphan diseases a major difficulty is recruiting sufficient numbers of patients to provide adequate statistical power for the study. In rare diseases such as AATD, utilising a delayed-start trial design, where patients receive the same active treatment but at different starting times, may overcome the need for a large patient population. Also, all patients ultimately receive active treatment, and so ethically it may be more appropriate than withdrawal or washout trial designs and trials where patients spend extended time on placebo. The accuracy of this approach was clearly demonstrated by a recent study by Chapman and colleagues. They not only enrolled a larger study population that was assessed over 4 years but also included both an Early-Start and Delayed-Start group (Figure 2) that permitted disease-modifying effects to be assessed (10).

Figure 2. RAPID and RAPID Extension trial design. This figure shows the design of the RAPID trial; subjects in RAPID received A₁-PI or placebo for 2 years. During the RAPID Extension trial all non-US subjects received A₁-PI for a further 2 years.

Evidence from the delayed-start trial

The Randomised, placebo-controlled trial of augmentation therapy in Alpha-1 Proteinase Inhibitor Deficiency (RAPID) and the RAPID Extension trial represent the largest completed clinical study to date of A₁-PI therapy in AATD patients and the only trial specifically designed to look at the disease-modifying effect of treatment (10). RAPID investigated A₁-PI therapy in 180 patients with AATD; 93 received A₁-PI and 87 received placebo (10). The primary endpoint was the change in CT lung density followed over 24 months (Figure 2). Following completion of the main two-arm, placebo-controlled RAPID trial, non-US subjects were invited to join RAPID Extension – a 2-year, open, non-controlled study in which all patients received A₁-PI therapy (Figure 2) from which interim data has recently been published (10). Due to the delayed-start trial design the potential disease-modifying effects of A₁-PI therapy could be investigated. Patients on active therapy who received A₁-PI substitution for 4 years throughout the RAPID and RAPID Extension trial were designated as the Early-Start cohort whilst those who initially received placebo and switched to active treatment during the RAPID Extension trial were termed the Delayed-Start cohort.

During the RAPID trial a significant difference in the annual rate of decline in lung density (0.74 g/L; p = 0.033, two-sided test, Figure 3) was observed between A₁-PI and placebo (−1.45 and −2.19 g/L/year, respectively). This corresponds to a reduction of 34% in the annual rate of lung density decline when compared with placebo. Interim analysis of the RAPID Extension trial showed that switching patients from placebo to A₁-PI resulted in a reduction in the slope of decline in lung density (from −2.19 g/L/year with placebo to −1.31 g/L/year with A₁-PI). The rate of decline observed with A₁-PI was distinctly slower than that observed during the placebo treatment phase in the RAPID trial and between 24 and 48 months the rate of annual decline was similar between the Early-Start and Delayed-Start cohorts. However, the lung density loss that occurred in the Delayed-Start cohort prior to A₁-PI therapy (dashed line) was not regained upon initiation of A₁-PI therapy (Figure 3, lower solid line).

Figure 3. Disease modification in the RAPID and RAPID Extension trials (10). Graph showing the annualised rate of decline in physiologically adjusted P15 (g/L) at TLC over 48 months (ITT population). Reprinted from The Lancet, Vol 386. Chapman, KR et al., Intravenous augmentation treatment and lung density in severe α1 antitrypsin deficiency (RAPID): a randomised, double-blind, placebo-controlled trial, pp. 360–368. Copyright (2015), with permission from Elsevier.

Analysis of both trials demonstrates that A₁-PI therapy is able to slow the progression of AATD-related emphysema and the rate of lung density loss by 50% compared with placebo. The findings from the RAPID trials strengthen those from the EXACTLE trial and the Danish-Dutch trial, which suggested a trend towards a protective effect of A₁PI but did not reach statistical significance. Furthermore treatment with A₁-PI therapy may delay the time to terminal lung function by several years (10), and this association with a survival benefit may also improve the acceptance of CT as a valid outcome. Due to the careful implementation of the appropriate delayed-start study design the combined RAPID and RAPID Extension trial data are the first to show a disease-modifying effect of specific treatment in pulmonary emphysema, i.e., A₁-PI therapy in AATD. Whilst the RAPID trial and RAPID Extension trial indicate that A₁-PI therapy is effective irrespective of the time of intervention, lung tissue lost during periods of no treatment cannot be regained. This is consistent with the impact of delayed-start treatment in that the clinical outcome measure in the Delayed-Start group never fully catches up with the Early-Start group (Figure 1).

Results from the RAPID and RAPID Extension trial highlight the need for more timely diagnosis of AATD to enable patients to receive appropriate treatment. As AATD-related emphysema is a progressive lung disease, earlier diagnosis may facilitate implementation of lifestyle changes and specific treatment options to slow lung density decline in affected patients. However, AATD-related emphysema may be difficult to differentiate from other COPD phenotypes solely on the basis of clinical symptoms, and on average the delay between symptom onset and a correct diagnosis of AATD is 5–7 years (6, 59). Given that lung tissue lost cannot be regained there is a need for both accurate and early diagnosis to allow the potential for early commencement of A₁-PI treatment to prevent the loss of lung tissue. Likewise, it opens new perspectives for patients with non-AATD-related emphysema by demonstrating that with better understanding of disease mechanisms more effective, targeted therapies might be able to prevent/slow disease progression.

Future perspectives in AATD

Beyond the use of A₁-PI in AATD-related emphysema, there is growing interest in disease modification of AATD that has future potential and requires acknowledgement. Experimental in vitro and in vivo AAT gene transfer with viral vectors has resulted in enhanced A₁-PI serum levels and a promising safety profile. However, human clinical trials with adeno-associated viral vectors did not achieve a protective level of A₁-PI in serum (60). Retinoic acid and other retinoids activate nuclear hormone receptors, Retinoic Acid Receptor (RAR) and Retinoid X Receptor (RXR), to activate and/or repress gene expression to modulate structure and function of tissues (61). Additionally, retinoids regulate growth factor receptors associated with alveolar septation and epithelial repair (61). As such, there is interest in the potential for retinoids to restore lung function and slow the progression of emphysema; however, studies of retinoids in humans (62) have thus far failed to replicate findings in animals (63, 64). Also, severe AATD may present as liver disease (cirrhosis), and there is growing interest in disease modification that protects proteostasis mechanisms. Potential molecules include carbamazepine, which has shown promise in mouse models (65), and is currently being investigated in a phase II clinical trial for its use in AATD-related liver disease (66). An alternative approach is to use antisense oligonucleotides to silence the mutated AAT gene and reduce the hepatic production of the mutant Z-A₁-PI protein and polymer formation. This may potentially prevent liver damage or allow the liver to repair. This approach has proven effective in animals (67) and is currently under investigation in a phase I clinical trial (68).

Conclusions

When considering chronic progressive diseases simply treating the associated symptoms is not sufficient to provide long-term improvements on disease progression. Disease-modifying treatments that affect the underlying cause of the disease can slow the progression of disease. Crucial to the success of clinical trials of potential disease-modifying therapies is careful trial design incorporating the use of reliable and sensitive study endpoints and/or markers of disease progression.

An important factor for consideration is that clinical outcome measures never truly catch up if disease-modifying treatment is delayed compared with initiated early, which emphasises the importance of early treatment. In the context of AATD-related emphysema appropriate and timely treatment is critical to prevent loss of lung tissue, particularly as the lung tissue lost prior to treatment cannot be regained. A₁-PI therapy is the only pharmacological treatment that currently addresses the underlying cause of disease in AATD and has been demonstrated to be disease-modifying, i.e., able to reduce the decline in lung tissue loss and delay the time to terminal respiratory failure and lung transplantation. The mechanism of action of A₁-PI is consistent with a disease-modifying effect; A₁-PI therapy preserves lung function by protecting the lung parenchymal tissue and air exchange membranes (alveoli) from degradation by NE.

Well-designed clinical trials are needed to conclusively prove the efficacy of disease-modifying therapies and careful study design in the RAPID trial has led to disease-modifying therapy recently being proven for AATD-related emphysema. Future clinical trials should incorporate careful trial design, such as delayed-start trial design, along with the use of highly sensitive outcome measures and/or reliable biomarkers that correlate with clinical outcomes and other surrogate endpoints.

Declaration of interest

Dr. Chorostowska-Wynimko has received grants, personal fees and non-financial support from Grifols, grants from Baxalta US Inc, and personal fees from CSL Behring. Editorial assistance was provided by Meridian HealthComms (Plumley, UK), funded by CSL Behring.