You are here

  • Home
  • Online First
  • Optimising bronchoalveolar lavage: lessons from alpha-1 antitrypsin deficiency

Article Text

Article menu

Download PDFPDF

Chronic obstructive pulmonary disease
Original research
Optimising bronchoalveolar lavage: lessons from alpha-1 antitrypsin deficiency
  1. Malcolm Herron1,2,
  2. Suzanne Roche1,2,
  3. Daniel D Fraughen1,
  4. Ronan C Heeney1,
  5. Lasya Kanchi1,
  6. Emma J Leacy1,
  7. Michelle Casey1,2,
  8. Cedric Gunaratnam1,2,
  9. Tomás P Carroll1,
  10. Mark P Murphy1,
  11. Noel G McElvaney1,2
  1. 1Royal College of Surgeons in Ireland School of Medicine, Dublin, Ireland
  2. 2Beaumont Hospital, Dublin, Ireland
  1. Correspondence to Dr Malcolm Herron; malcolmherron{at}gmail.com

Abstract

Background Bronchoalveolar lavage (BAL) is essential in determining the efficacy of novel therapies in alpha-1 antitrypsin deficiency (AATD). These require initial proof-of-concept demonstration that treatment administration increases alpha-1 antitrypsin (AAT) levels and/or anti-neutrophil elastase inhibitory capacity (ANEC) in the lung. Early-phase studies often encounter high interindividual variability of BAL results, primarily stemming from the inherent dilution characteristics of returned BAL fluid. A BAL protocol that minimises this variability is needed for reliable comparison of biochemical endpoints in the lung.

Methods The study population included 21 severe AATD (ZZ), 22 moderate AATD (MZ) and 23 non-AATD (MM) individuals, further categorised as healthy, unobstructed current smokers or patients with chronic obstructive pulmonary disease (COPD). An additional six ZZ individuals were receiving intravenous alpha-1 augmentation therapy. We compared common BAL correction methods—albumin, total protein and epithelial lining fluid (ELF) volume measured by urea—when reporting early-phase biochemical endpoints, AAT and ANEC.

Results BAL performed with a paediatric bronchoscope (PB) improved alveolar sampling compared with a traditional adult bronchoscope. Both uncorrected and ELF-corrected BAL demonstrated high interindividual variability regardless of lung health status. BAL total protein correction minimised interindividual variability, producing significant differences in AAT and ANEC between all genotypes, the strongest relationship with plasma AAT levels (r2=0.83), greatest inter-lobar concordance in AAT levels (r2=0.76) and strong correlation between BAL AAT and ANEC (r2=0.88).

Conclusions By capitalising on the marked consistency in AAT levels between AAT genotypes, and the close relationship between plasma and lung AAT levels, we demonstrate reliable alveolar sampling that aligns closely with plasma.

  • Alpha1 Antitrypsin Deficiency
  • Bronchoscopy
  • COPD Pathology

Data availability statement

Data are available upon reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/thorax-2024-221797

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Bronchoalveolar lavage (BAL) is essential in evaluating the biochemical efficacy of novel therapies in alpha-1 antitrypsin deficiency (AATD).

    • However, comparing BAL findings among individuals is challenging due to the inherent dilution characteristics of returned BAL fluid.

    • Consequently, many early-phase studies of novel therapeutics encounter considerable variability of BAL outcomes between individuals, complicating the comparison of placebo and treatment groups.

    WHAT THIS STUDY ADDS

    • In this study, we outline a BAL protocol designed to enhance precision and reliability when ascertaining and reporting BAL biochemical endpoints.

    • We detail a BAL protocol aimed at improving alveolar sampling, minimising interindividual variability to enable the detection of incremental changes in vital biochemical endpoints and establishing a robust correlation with plasma results.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Accurately and reliably portraying the lung with BAL is vital in understanding disease mechanisms, identifying potential therapeutic targets, evaluating biomarkers in the lung, and in early-phase clinical trials that rely on BAL for proof of concept.

    Background

    Bronchoalveolar lavage (BAL) has been used as a research tool for five decades.1–4 It offers a relatively non-invasive means of evaluating the epithelial lining fluid (ELF) of the lower airways, giving unique insights into disease pathogenesis, therapeutic interventions and response to therapies. Classically, a bronchoscope is passed into the lung until its circumference seals an airway, usually at the third to fourth order bronchus, ensuring instilled saline will only pass into the distal airways and alveoli, a position termed a BAL wedge. The returned fluid on suctioning contains a mixture of instillate and ELF, including ELF proteins and cells. This gives insight into the type of inflammatory cells, their levels of activation and the inflammatory cytokines and protease anti-protease interactions in the bronchoalveolar compartment.

    A major obstacle in BAL research is that returned BAL fluid does not directly represent ELF, rather a varying dilution of ELF, and while plasma or serum is directly comparable between patients, the comparison of BAL is more complex. This has led to variable reporting of results using different methods of BAL correction, including correcting for protein and albumin, and calculation of ELF volume most commonly by a urea method, but earlier attempts used potassium, methylene blue and calcium.5–9 The determined ELF volume is then used to estimate the concentration of ELF contents. Various factors can affect the volume of returned BAL fluid, and therefore the presumed dilution of ELF, such as the lobe sampled, the position of the patient, body mass index (BMI), gender, and significantly for this work, disease-related loss of lung elasticity which promotes collapse of the distal airway under suction.10 11 Even when fluid is recovered, the variable return volume is assumed to alter the dilution of sampled ELF, therefore complicating the comparison between healthy and chronic obstructive pulmonary disease (COPD) individuals.

    BAL has been instrumental in providing evidence of biochemical efficacy for therapeutic interventions. This is best exemplified in alpha-1 antitrypsin deficiency (AATD). AATD is an autosomal co-dominant disorder resulting from mutations in the SERPINA1 gene. Alpha-1 antitrypsin (AAT) is a key protease inhibitor and acute phase anti-inflammatory glycoprotein. It is predominantly produced in the liver where it enters the circulation and subsequently the lungs. Most people carry two copies of the wild-type M allele of SERPINA1 (the Pi*MM genotype) and have normal circulating AAT levels, and therefore normal lung AAT levels. Homozygosity for the pathogenic Pi*Z allele (Glu342Lys, rs28929474) results in severe AAT deficiency in the circulation and therefore the lungs. This results in insufficient anti-protease protection and the unopposed action of serine proteases such as neutrophil elastase (NE).12 13 These proteases can destroy structural proteins of the lung matrix and when introduced into the lung cause emphysema in animal models.14 15 Z homozygous individuals are at significantly increased risk of developing early-onset emphysema, particularly if they smoke. Research in AATD has led to the only specific therapy for COPD to date whereby once-weekly intravenous infusion of AAT purified from the plasma of MM individuals has been shown to restore levels of AAT and anti-neutrophil elastase inhibitory capacity (ANEC) above a protective level in the circulation and lung.16 The initial studies of this therapeutic approach relied heavily on BAL to show that administering intravenous AAT augmented the levels of AAT in the lung, as measured by AAT in ELF, quantified by the urea method and that this exogenous AAT increased ELF ANEC.16 These proof-of-concept data were sufficient to gain regulatory approval in 1987, and augmentation therapy subsequently achieved clinical validation by demonstrating preservation of lung density, improved symptoms, reduced spirometric decline and improved survival, highlighting the importance of demonstrating early biochemical efficacy.17–19 Since the original description of intravenous AAT therapy, a large number of other treatments have been evaluated for AATD. These include gene therapy, gene editing, intravenous administration of recombinant forms of AAT, aerosolisation of plasma-purified and recombinant AAT, and a variety of oral and aerosolised anti-NE compounds.20–23 All of these therapies require, as an initial proof of concept, demonstration that administration increases AAT and/or ANEC in the lung, which is best demonstrated by BAL.

    Variable BAL protocols and correction of results have led to significant disparities in lung lavage results in therapeutic trials. This variability in AAT levels, NE activity and ANEC has made many of these early phase studies difficult to interpret and compare.24–27 The need for standardisation and optimising the reliability of BAL methods is vital in determining the mechanisms of disease pathogenesis, identifying potential therapeutic targets, evaluating biomarkers in the lung, and in these early phase studies that often rely on relatively small patient numbers to determine a proof of concept.

    In this manuscript, we describe a BAL protocol that addresses a significant challenge in BAL research: achieving reliable comparative results between individuals. This enables precise comparisons between healthy individuals and those with lung disease, as well as assessment of therapeutic effects on the lungs. Our method is exemplified in AATD and general COPD, but the principles are broadly applicable.

    Methods

    Demographics

    Patients referred to the Irish National Centre of Expertise for AATD were evaluated in this study. The participants included 21 individuals with severe AATD (ZZ), 22 with moderate AATD (MZ) and 23 controls with normal AAT levels (MM). Additionally, six ZZ patients were receiving intravenous alpha-1 augmentation therapy. These 72 individuals were further categorised as healthy, unobstructed current smokers or patients with COPD (table 1). Lung health status was determined using clinical evaluation, CT of thorax and pulmonary function testing.

    View this table:
    Table 1

    Demographic details of the study cohorts

    Chemicals and reagents

    BAL proteins were measured with Pierce BCA Protein Assay Kit. ELF volumes were determined from BAL and time-matched plasma urea concentrations measured with Invitrogen Urea Nitrogen (BUN) Colorimetric Detection Kit.6 Secretory leucocyte protease inhibitor (SLPI), AAT, albumin and surfactant D were measured with ELISA (R&D Systems). ANEC was measured by fluorescence resonance energy transfer assay using purified NE (Elastin Products, TS563) and specific NE substrate, Abz-APEEIMRRQ-EDDnp (Peptide Institute, 3230-V). Samples were run in duplicate in all experiments.

    Bronchoscopy procedure

    Bronchoscopy and BAL were performed by a single operator with the patient in a supine position. The anterior segment of the right upper lobe was wedged at the sixth to eighth order bronchus with a 3.8 mm diameter paediatric bronchoscope (PB, Olympus BF-3C160). In the case of lower lobe samples, the anterior segment of the left lower lobe was used. 20 mL of 0.9% NaCl was instilled into the airway and returned under gentle suction to ensure airway patency and to remove excessive mucus and airway debris. Following this, 150 mL of 0.9% NaCl was instilled in aliquots of 50 mL and returned sequentially into a syringe via manual suction, while clamping wall suction. The fluid was returned almost immediately following instillation. The extracted fluid from 150 mL instillation was combined, the return volume was recorded and immediately placed on ice. A subset of the study population underwent sampling with a wider diameter traditional adult bronchoscope (AB, Olympus BF-H190) wedged at the third or fourth order bronchus. All participants were offered clinical review at 3 months following bronchoscopy and an outcome questionnaire was performed targeted at patient safety and tolerability.

    BAL processing

    BAL fluid underwent centrifugation for 10 min at 459×g and 4°C to generate supernatant and a cell pellet. The cell pellet was resuspended in 1 mL Dulbecco’s phosphate buffered saline, the cells counted with a hemocytometer, and the viability determined with 0.4% trypan blue solution. A modified Giemsa-Wright (Diff-Quik, Sigma) slide was prepared to establish a differential cell count by centrifugation of 50 000 cells in 100 µL 0.9% NaCl for 5 min at 28×g. Aliquots containing 106 cells were stored in 1 mL freezing medium (50% RPMI, 40% FBS, 10% DMSO) at −80°C. BAL supernatant was stored at −80°C. BAL processing in each case was completed within 2 hours of bronchoscopy.

    Blood processing

    Phlebotomy was performed within 30 min of BAL. Then, 9.6 mL of whole blood in lithium heparin tubes underwent centrifugation for 10 min at 541.5×g to generate plasma. The plasma was aliquoted and stored at −80°C.

    Bronchoscope comparison

    Thirty-nine paediatric bronchoscope (PB; Olympus BF-3C160—outer diameter 3.8 mm) samples were compared with 14 adult bronchoscope (AB; Olympus BF-H190—outer diameter 5.5 mm) samples. Comparisons were made in healthy MM, healthy MZ and MM patients with COPD. Bronchoscope comparison used neutrophil proportion, SLPI as an airway marker and surfactant D as an alveolar marker.28–33

    Statistical analysis

    Results are expressed as mean±SD, unless otherwise specified. The data were analysed using GraphPad Prism V.10 for MacOS. For comparison of two paired groups, a paired t-test was used, with a Wilcoxon test used in non-Gaussian populations. An unpaired t-test was used in the comparison of unpaired groups, with a Mann-Whitney test used in non-Gaussian populations. Comparison of three or more unmatched groups was by analysis of variance (ANOVA). Correlations were achieved by Spearman correlation. A value of p<0.05 was considered statistically significant.

    Results

    BAL performed with a PB improves alveolar sampling

    The importance of BAL wedge position in the airway was determined by comparison of returned fluid from the distal BAL wedge obtained by a narrow 3.8 mm PB, and the proximal BAL wedge permitted with a wider 5.5 mm AB. All subsequent BAL in the study was performed with PB following early demonstration that PB BAL improved alveolar sampling and minimised excessive larger airway sampling. Thiry-nine PB BAL samples were compared with 14 AB BAL samples (table 2). PB BAL sampled more ELF (1.1 mL vs 0.43 mL, p<0.01) and contained more surfactant D (1.2 µg/mg protein vs 0.72 µg/mg protein, p=0.01), less SLPI (0.44 µg/mg protein vs 0.85 µg/mg protein, p=0.04) and lower neutrophil proportions (1.2% vs 4.3%, p=0.02) than AB BAL, confirming a greater degree of alveolar sampling by PB. Lower ELF sampling by AB BAL created erroneous overestimation of its contents when ELF correction was applied (table 2: AB surfactant D 782 nM vs PB 401 nM, p=0.16).

    View this table:
    Table 2

    Comparison of BAL performed with a paediatric (PB) and a traditional adult bronchoscope (AB)

    The influence of COPD and AATD on BAL return

    The proportion of returned BAL fluid was significantly lower in COPD than healthy individuals (figure 1A, 34% vs 55%, p<0.0001). The presence of COPD rather than AATD dictated lower BAL return, as no differences were observed between unobstructed MM, MZ and ZZ individuals (figure 1B). Furthermore, FEV1 (% predicted) was the single strongest determinant of BAL return (figure 1C, r2=0.34).

    Figure 1
    Figure 1

    (A) Comparison of BAL return in healthy and COPD individuals regardless of AAT genotype (mean 55% vs 34%, p<0.0001), N=59 (table 1: healthy 22, COPD 37). (B) Comparison of BAL return in healthy and COPD (C) individuals by AAT genotype, N=59. (C) Correlation between FEV1 (% predicted) and BAL return (%) (N=59, Spearman r2=0.34, p<0.0001). AAT, alpha-1 antitrypsin; BAL, bronchoalveolar lavage; FEV1, forced expiratory volume in 1 s.

    Comparison of common BAL correction methods

    The ideal BAL correction method uses a BAL component that is constant between all of the individuals being studied. BAL albumin concentrations, ELF volumes recovered and protein concentrations were not significantly different between genotypes and health states (figure 2A–C, ANOVA, p=0.38, 0.55, 0.78, respectively). However, albumin’s contribution to total protein content of BAL is not constant, and its variability within and between groups makes it a less useful correction method. As a result, AAT levels corrected using a urea ELF method show greater agreement with those corrected by total protein (figure 2D, r2=0.79) than by albumin (figure 2D, r2=0.57).

    Figure 2
    Figure 2

    (A) BAL albumin concentrations in healthy, smoking (S) and chronic obstructive pulmonary disease (C) individuals for each AAT genotype. (B) BAL epithelial lining fluid (ELF) volumes recovered. (C) BAL total protein concentrations. Mean and SEM displayed. Comparisons made with analysis of variance. (D) Correlation between ELF AAT and BAL AAT corrected for albumin (Spearman r2=0.57, p<0.0001), and ELF AAT correlated with BAL AAT corrected for total protein (Spearman r2=0.79, p<0.0001). AAT, alpha-1 antitrypsin; BAL, bronchoalveolar lavage.

    Reporting of BAL AAT levels

    Plasma AAT levels were significantly different between all genotypes (figure 3A). The expected trend in AAT levels is observed in uncorrected and ELF-corrected BAL (figure 3B,C); however, significant differences between all genotypes were not observed due to variability within groups. This variability resulted in several ZZ individuals having higher ELF AAT levels than MZ individuals, which is inconsistent with their respective plasma levels (figure 3A,C). Protein-corrected BAL AAT (figure 3D) minimised interindividual variability, producing significant differences between all genotypes and a pattern that closely resembles plasma AAT levels. The ability of protein correction to reduce interindividual variance is observed in obstructed and unobstructed individuals, confirming its utility regardless of BAL return or lung health status.

    Figure 3
    Figure 3

    (A) Plasma AAT levels in non-AATD (MM), moderate AATD (MZ), severe AATD (ZZ) and ZZ AATD 48 hours after IV-AAT infusion (Aug), divided into healthy, smoking (S) and chronic obstructive pulmonary disease (C) groups. (B) Uncorrected BAL AAT levels. (C) ELF AAT levels. (D) Protein-corrected BAL AAT levels. All panels display PB BAL results in the same 72 individuals (table 1), unless otherwise specified. Individual values and mean displayed. Comparisons made with analysis of variance. AATD, alpha-1 antitrypsin deficiency; AAT, alpha-1 antitrypsin; BAL, bronchoalveolar lavage; ELF, epithelial lining fluid; PB, paediatric bronchoscope.

    BAL AAT levels are dictated by plasma AAT levels

    An increase in plasma AAT levels predictably increases AAT levels in the lung. This relationship is strongest when BAL AAT levels are corrected for protein (figure 4C, r2=0.83), thus enabling the most accurate prediction of BAL AAT levels from more readily accessible plasma levels. Urea ELF correction does not improve on the correlation observed in uncorrected BAL (figure 4A,B, r2=0.67 vs 0.72).

    Figure 4
    Figure 4

    (A) Correlation between plasma AAT levels and uncorrected BAL AAT levels (Spearman r2=0.72, p<0.0001). (B) Correlation between plasma AAT levels and ELF AAT levels (Spearman r2=0.67, p<0.0001). (C) Correlation between plasma AAT levels and protein-corrected BAL AAT levels (Spearman r2=0.83, p<0.0001). All panels display PB BAL results in the same 72 individuals (table 1). AAT, alpha-1 antitrypsin; BAL, bronchoalveolar lavage; ELF, epithelial lining fluid; PB, paediatric bronchoscope.

    Comparison of AAT levels between lobes

    Twenty-three patients (8 MM, 6 MZ, 8 ZZ, 1 augmentation ZZ) underwent paired BAL sampling in the right upper lobe (UL) and the left lower lobe (LL). Greater inter-lobar concordance in AAT levels was observed with protein correction than ELF correction (figure 5A,B, r2=0.76 vs 0.41). BAL AAT levels corrected for protein were the same in the UL and LL (figure 5B, Wilcoxon p=0.96), whereas ELF correction demonstrated a trend towards higher AAT levels in the lower lobes (figure 5A, Wilcoxon p=0.15). Consequently, combining protein-corrected samples from both lobes improved the comparison of lung AAT levels between genotypes (figures 3D and 5D). Whereas, combining ELF-corrected inter-lobar samples increased variability, resulting in a loss of significance between augmented and baseline ZZ individuals (figures 3C and 5C).

    Figure 5
    Figure 5

    (A) Correlation between ELF AAT levels in the upper and lower lobes in 23 paired samples (Spearman r2=0.41, p=0.001; Wilcoxon, p=0.15). (B) Correlation between protein-corrected BAL AAT levels in the upper and lower lobes in 23 paired samples (Spearman r2=0.76, p<0.0001; Wilcoxon, p=0.96). (C) Comparison of ELF-corrected BAL AAT levels between genotypes and health states in upper lobe (black dots) and lower lobe (grey dots) samples. (D) Comparison of protein-corrected BAL AAT levels between genotypes and health states in upper lobe (black dots) and lower lobe (grey dots) samples. AAT, alpha-1 antitrypsin; BAL, bronchoalveolar lavage; ELF, epithelial lining fluid.

    Reporting of BAL anti-elastase capacity

    To further explore the benefit of protein correction, ANEC, a functional readout of BAL protease inhibition, was determined. BAL ANEC is a vital endpoint in trials of novel therapeutics in AATD, demonstrating not only the presence of the therapeutic in the lung but also its ability to potentially impact disease pathogenesis. AAT is the dominant contributor to ANEC in the lung with mean BAL AAT levels of 209 (MM), 75 (MZ), 8.8 (ZZ) and 55 pmol/mg protein (augmentation ZZ) (figure 6A), and mean BAL ANEC levels 184 (MM), 81 (MZ), 8.1 (ZZ) and 48 pmol/mg protein (augmentation ZZ) (figure 6B). BAL ANEC strongly correlates with BAL AAT levels (figure 6C, r2=0.88).

    Figure 6
    Figure 6

    (A) BAL AAT levels in healthy, smoking (S) and chronic obstructive pulmonary disease (C) patients for each AAT genotype. (B) BAL anti-neutrophil elastase capacity (ANEC). Comparisons made with analysis of variance. (C) Correlation between BAL AAT corrected for protein and BAL ANEC corrected for protein (Spearman, r2=0.88). All panels display PB BAL results from the same 72 individuals (table 1). AAT, alpha-1 antitrypsin; BAL, bronchoalveolar lavage; PB, paediatric bronchoscope.

    Patient safety and tolerability

    There were no intra-procedural complications. A telephone questionnaire assessing post-procedural complications, patient safety, tolerability and satisfaction was performed in all participants, with 89% response rate. The most commonly reported symptoms were transient sore throat (11%), chest discomfort (5%), cough (5%), dyspnoea (3%), fever (3%) and transient hoarseness (1%). Four patients required a course of outpatient oral antibiotics within 2 weeks after bronchoscopy (3 ZZ COPD, 1 MM COPD) with subsequent resolution of their symptoms. Four patients reported symptomatic improvement after the procedure (2 MZ COPD, 2 ZZ COPD). Of these participants, 84% indicated that they would be willing to undergo bronchoscopy and BAL again solely for research purposes.

    Discussion

    In this study, we optimised several key parameters to enhance the precision of BAL. By capitalising on the close relationship between plasma and lung AAT levels, we demonstrate reliable lung sampling that aligns closely with plasma. Additionally, the marked consistency in plasma AAT levels between AAT genotypes allows demonstration of a BAL method that can detect incremental changes in the lung. As a result, the BAL concepts outlined in this study have broad potential applications.

    First, we have demonstrated that PB BAL improves alveolar sampling by showing it decreases neutrophil proportions, decreases levels of the predominantly larger airway anti-protease SLPI and increases levels of alveolar surfactant D (table 2).28–33 A similar airway sampling effect has previously been shown resulting in the first BAL aliquot being labelled an airway sample and thus discarded, whereas we advocate for distal sampling with a narrow bronchoscope and pooling of all aliquots.28 34 The alveolar sampling achieved by PB is essential to show the potential disease-modifying effect of novel therapeutics at the site of alveolar and terminal bronchiole destruction in COPD, as demonstrated by BAL ANEC. PB allows a BAL wedge at the sixth to eighth order bronchus, rather than at the third or fourth order permitted by a larger diameter traditional AB. Traversing multiple airway bifurcations with PB avoids an exponential increase in larger airway sampling that occurs with a wider AB.

    Second, we have demonstrated the impact that improved alveolar sampling has on BAL results. The superior distal position of PB in the airway and proximity to the alveoli maximises the surface area that is lavaged, yielding greater ELF volume recovery (table 2, PB 1.1 vs AB 0.4 mL ELF volume, p<0.01). In contrast, AB proximal position in the airway and greater distance from the alveoli leads to excessive large airway sampling and reduced alveolar sampling. The resultant low ELF recovery by AB creates erroneous overestimation of BAL components when ELF correction is applied, leading to higher ELF surfactant in AB than PB (table 2, AB surfactant D 782 vs PB 401 ELF nM, p=0.16). In contrast, surfactant levels were higher in PB than AB BAL when uncorrected (PB 3.9 vs AB 3.0 nM, p=0.22) and significantly higher when protein corrected (PB 1.2 vs AB 0.72 µg/mg protein, p=0.01), highlighting that urea ELF correction does not accurately represent the lung. Thus, multiple institutions performing BAL with different diameter bronchoscopes in different order bronchi will create interindividual variance which is amplified when ELF correction is applied.

    Third, correction of BAL results is essential to reduce interindividual variability (figure 3). High variability of uncorrected BAL AAT levels was observed within groups of healthy and obstructed individuals (figure 3B) and occurred independently of diminished BAL return (figure 2). Consequently, uncorrected BAL produced the expected trend in AAT levels between AAT genotypes, but significance was not reached between genotypes, creating the need for either a larger number of patients, or a correction method that minimises the variability between patients (figure 3B). This is best achieved by protein correction as we demonstrate here (figure 3D). Despite PB sampling, ELF correction does not adequately reduce the SD within groups, rendering BAL AAT levels not significantly different between genotypes with significantly different plasma levels (figure 3A,C). In studies aiming to increase AAT levels and ANEC in the lung from the ZZ range towards the MZ or even MM range, this type of discrimination is extremely important. The superior ability of protein correction to discriminate incremental changes in AAT directly relates to its close correlation with plasma (figure 4C, r2=0.83). The ELF AAT correlation with plasma in this study (figure 4B, r2=0.67), and in the original description of IV-AAT replacement therapy (r2=0.7), does not improve on the correlation observed in uncorrected BAL (figure 4A, r2=0.72).16 Lung levels of AAT are dictated by circulation levels which in turn are determined by liver production of AAT or exogenous intravenous AAT. Therefore, the BAL correction method that ascertains the strongest plasma relationship most accurately represents the lung.

    Lastly, the inter-lobar agreement in protein-corrected AAT levels confirms that the current practice of performing BAL in multiple lobes and combining the returned fluid is unnecessary in AATD (figure 5B,D).4 6 24 If comparisons between lobes are deemed necessary, we recommend protein correction. Applying ELF correction to returned fluid from multiple lobes further increases the variability of results due to a weaker inter-lobar ELF AAT correlation, resulting in loss of significance between augmented and baseline ZZ individuals (figure 5A,C). The observed trend in greater ELF AAT levels in the lower lobes (figure 5A) is an artefact of reduced ELF volume recovery in lower lobe BAL samples (LL 0.70 vs UL 0.94 mL, p=0.04), akin to the artefactual overestimation of BAL components that occurs in AB versus PB samples. The inter-lobar agreement in protein-corrected AAT levels is of special interest in severe AATD given the heterogeneity of the lung disease, which is classically regarded as lower lobe predominant.

    The complexities of accurately and reliably evaluating the lung with BAL have led to study findings that are often difficult to interpret and compare. The original description of IV-AAT therapy reported a wide range of ELF ANEC, with three pretreatment ZZ individuals’ ELF ANEC exceeding the lowest MM individual.16 With our method, ZZ and MM groups are very significantly different (figure 6B). More recent studies evaluating novel therapeutics have failed to show significant post-treatment rises in ELF AAT owing to variability between patients, or have reported large ranges in pretreatment ZZ ELF AAT (59.8–966.8 nM).24 35 This 16-fold range in ZZ ELF AAT exceeds the 10-fold difference that is observed between the plasma AAT levels of MM and ZZ individuals (figure 3A, 30.1 vs 2.8 µM). Another recent study evaluated post-treatment rises in ELF AAT following inhalation of AAT. BAL was performed pretreatment and post-treatment in patients receiving 80 mg daily, 160 mg daily or placebo. A 1.9 µM confidence interval (95% CI −1032 to 884 nM) in ELF AAT levels was observed following administration of placebo. Any method that exhibits such variability in patients receiving placebo is not an ideal starting point for comparing treatments. In this study, the 884 nM rise in ELF AAT after treatment with placebo exceeded the rise observed in some of the treatment cohort (95% CI 841 to 3800 nM). A 70-fold range in ELF AAT concentrations was observed after administration of a 160 mg dose (95% CI 1 to 71 µM).27 Given these data, it is clear that the current BAL approach is inadequate for accurately portraying the lung, making it challenging to assess the potential of new treatments.

    BAL is essential in the evaluation of the anti-protease effect of novel lung-directed therapies in AATD. Given the current investigational landscape for AATD includes many novel therapeutics, a BAL protocol that is easily reproducible, ensures alveolar sampling and can demonstrate incremental changes in AAT and ANEC with minimal variability in small numbers of patients is of significant value. Protein-corrected PB BAL offers the most reliable means to assay biochemical endpoints in the lung.

    Data availability statement

    Data are available upon reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved by Beaumont Hospital Ethics Committee reference number 18/52. Participants gave informed consent to participate in the study before taking part.

    Acknowledgments

    The authors wish to thank the patients at the Irish Centre for Genetic Lung Disease, Dublin, Ireland without whom the study would not be possible. We also wish to acknowledge the contributions of Kristina Joksaite, Louise Clarke, Siobhan Murray, Geraldine Harte, Marie Dowling, Justyna Junger, Rosie Carroll, Jean Kirwan, Linda Phoenix, Geraldine Kelly, Anne Marie O’Dowd, Domonique Proceviat and Ammar Janjuaa.

    References

    1. ↵
      1. Reynolds HY,
      2. Newball HH
      . Analysis of proteins and respiratory cells obtained from human lungs by bronchial lavage. J Lab Clin Med 1974;84:559–73.
      OpenUrlPubMedWeb of Science
    2. ↵
      1. Lynn WS,
      2. Kylstra JA,
      3. Sahu SC, et al
      . Investigations of black bronchoalveolar human lavage fluid. Chest 1977;72:483–8. doi:10.1378/chest.72.4.483
      OpenUrlCrossRefPubMed
    3. ↵
      1. Crystal RG,
      2. Reynolds HY,
      3. Kalica AR
      . Bronchoalveolar lavage. The report of an international conference. Chest 1986;90:122–31. doi:10.1378/chest.90.1.122
      OpenUrlCrossRefPubMed
    4. ↵
      1. Reynolds HY
      . Bronchoalveolar lavage. Am Rev Respir Dis 1987;135:250–63. doi:10.1164/arrd.1987.135.1.250
      OpenUrlPubMedWeb of Science
    5. ↵
      1. Baughman RP,
      2. Bosken CH,
      3. Loudon RG, et al
      . Quantitation of bronchoalveolar lavage with methylene blue. Am Rev Respir Dis 1983;128:266–70. doi:10.1164/arrd.1983.128.2.266
      OpenUrlPubMedWeb of Science
    6. ↵
      1. Rennard SI,
      2. Basset G,
      3. Lecossier D, et al
      . Estimation of volume of epithelial lining fluid recovered by lavage using urea as marker of dilution. J Appl Physiol 1986;60:532–8. doi:10.1152/jappl.1986.60.2.532
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Bell DY,
      2. Haseman JA,
      3. Spock A, et al
      . Plasma proteins of the bronchoalveolar surface of the lungs of smokers and nonsmokers. Am Rev Respir Dis 1981;124:72–9. doi:10.1164/arrd.1981.124.1.72
      OpenUrlPubMedWeb of Science
    8. ↵
      1. Davis GS,
      2. Giancola MS,
      3. Costanza MC, et al
      . Analyses of sequential bronchoalveolar lavage samples from healthy human volunteers. Am Rev Respir Dis 1982;126:611–6. doi:10.1164/arrd.1982.126.4.611
      OpenUrlPubMedWeb of Science
    9. ↵
      1. G TBC
      . Proteins in Bronchoalveolar Lavage Fluid. Am Rev Respir Dis 1990;141:S183–8. doi:10.1164/ajrccm/141.5_Pt_2.S183
      OpenUrlCrossRef
    10. ↵
      1. Shikano K,
      2. Abe M,
      3. Shiko Y, et al
      . What are the factors affecting the recovery rate of bronchoalveolar lavage fluid? Clin Respir J 2022;16:142–51. doi:10.1111/crj.13462
      OpenUrlCrossRefPubMed
    11. ↵
      1. Koda K,
      2. Hozumi H,
      3. Yasui H, et al
      . Predictors for bronchoalveolar lavage recovery failure in diffuse parenchymal lung disease. Sci Rep 2021;11:1682. doi:10.1038/s41598-021-81313-5
    12. ↵
      1. Janoff A
      . Elastases and emphysema. Current assessment of the protease-antiprotease hypothesis. Am Rev Respir Dis 1985;132:417–33. doi:10.1164/arrd.1985.132.2.417
      OpenUrlPubMedWeb of Science
    13. ↵
      1. Snider GL,
      2. Ciccolella DE,
      3. Morris SM, et al
      . Putative role of neutrophil elastase in the pathogenesis of emphysema. Ann N Y Acad Sci 1991;624:45–59. doi:10.1111/j.1749-6632.1991.tb17005.x
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Snider GL,
      2. Lucey EC,
      3. Christensen TG, et al
      . Emphysema and bronchial secretory cell metaplasia induced in hamsters by human neutrophil products. Am Rev Respir Dis 1984;129:155–60. doi:10.1164/arrd.1984.129.1.155
      OpenUrlPubMedWeb of Science
    15. ↵
      1. Senior RM,
      2. Tegner H,
      3. Kuhn C, et al
      . The induction of pulmonary emphysema with human leukocyte elastase. Am Rev Respir Dis 1977;116:469–75. doi:10.1164/arrd.1977.116.3.469
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Wewers MD,
      2. Casolaro MA,
      3. Sellers SE, et al
      . Replacement therapy for alpha 1-antitrypsin deficiency associated with emphysema. N Engl J Med 1987;316:1055–62. doi:10.1056/NEJM198704233161704
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Ellis PR,
      2. Holm KE,
      3. Choate R, et al
      . Quality of Life and Mortality Outcomes for Augmentation Naïve and Augmented Patients with Severe Alpha-1 Antitrypsin Deficiency. Chronic Obstr Pulm Dis 2023;10:139–47. doi:10.15326/jcopdf.2022.0339
      OpenUrlPubMed
    18. ↵
      1. McElvaney NG,
      2. Burdon J,
      3. Holmes M, et al
      . Long-term efficacy and safety of α1 proteinase inhibitor treatment for emphysema caused by severe α1 antitrypsin deficiency: an open-label extension trial (RAPID-OLE). Lancet Respir Med 2017;5:51–60. doi:10.1016/S2213-2600(16)30430-1
      OpenUrlCrossRefPubMed
    19. ↵
      1. Fraughen DD,
      2. Ghosh AJ,
      3. Hobbs BD, et al
      . Augmentation Therapy for Severe Alpha-1 Antitrypsin Deficiency Improves Survival and Is Decoupled from Spirometric Decline-A Multinational Registry Analysis. Am J Respir Crit Care Med 2023;208:964–74. doi:10.1164/rccm.202305-0863OC
      OpenUrlCrossRefPubMed
    20. ↵
      1. Garver RI,
      2. Chytil A,
      3. Courtney M, et al
      . Clonal gene therapy: transplanted mouse fibroblast clones express human alpha 1-antitrypsin gene in vivo. Science 1987;237:762–4. doi:10.1126/science.3497452
      OpenUrlAbstract/FREE Full Text
    21. ↵
      1. Rosenfeld MA,
      2. Siegfried W,
      3. Yoshimura K, et al
      . Adenovirus-mediated transfer of a recombinant alpha 1-antitrypsin gene to the lung epithelium in vivo. Science 1991;252:431–4. doi:10.1126/science.2017680
      OpenUrlAbstract/FREE Full Text
    22. ↵
      1. Hubbard RC,
      2. Crystal RG
      . Strategies for aerosol therapy of alpha 1-antitrypsin deficiency by the aerosol route. Lung 1990;168 Suppl:565–78. doi:10.1007/BF02718179
      OpenUrlCrossRefPubMedWeb of Science
    23. ↵
      1. Strnad P,
      2. Mandorfer M,
      3. Choudhury G, et al
      . Fazirsiran for Liver Disease Associated with Alpha1-Antitrypsin Deficiency. N Engl J Med 2022;387:514–24. doi:10.1056/NEJMoa2205416
      OpenUrlCrossRefPubMed
    24. ↵
      1. Stoller JK,
      2. Rouhani F,
      3. Brantly M, et al
      . Biochemical efficacy and safety of a new pooled human plasma alpha(1)-antitrypsin, Respitin. Chest 2002;122:66–74. doi:10.1378/chest.122.1.66
      OpenUrlCrossRefPubMedWeb of Science
    25. ↵
      1. Carp H,
      2. Miller F,
      3. Hoidal JR, et al
      . Potential mechanism of emphysema: alpha 1-proteinase inhibitor recovered from lungs of cigarette smokers contains oxidized methionine and has decreased elastase inhibitory capacity. Proc Natl Acad Sci USA 1982;79:2041–5. doi:10.1073/pnas.79.6.2041
      OpenUrlAbstract/FREE Full Text
    26. ↵
      1. Walsh RL,
      2. Dillon TJ,
      3. McLennan G
      . A rapid procedure for the measurement of elastase inhibitory function of plasma and bronchoalveolar lavage fluid. Clin Biochem 1992;25:263–70. doi:10.1016/0009-9120(92)80031-b
      OpenUrlCrossRefPubMed
    27. ↵
      Phase II, double-blind, placebo-controlled study to explore the ELF and plasma concentration as well as safety of inhaled alpha-1 antitrypsin in alpha-1 antitrypsin deficiency subjects. Clinicaltrials.gov. NCT02001688. 2020.
    28. ↵
      1. Betsuyaku T,
      2. Takeyabu K,
      3. Tanino M, et al
      . Role of secretory leukocyte protease inhibitor in the development of subclinical emphysema. Eur Respir J 2002;19:1051–7. doi:10.1183/09031936.02.00253202
      OpenUrlPubMedWeb of Science
    29. ↵
      1. De Water R,
      2. Willems LN,
      3. Van Muijen GN, et al
      . Ultrastructural localization of bronchial antileukoprotease in central and peripheral human airways by a gold-labeling technique using monoclonal antibodies. Am Rev Respir Dis 1986;133:882–90.
      OpenUrlPubMedWeb of Science
    30. ↵
      1. Fryksmark U,
      2. Ohlsson K,
      3. Polling A, et al
      . Distribution of antileukoprotease in upper respiratory mucosa. Ann Otol Rhinol Laryngol 1982;91:268–71. doi:10.1177/000348948209100308
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. Mooren HW,
      2. Kramps JA,
      3. Franken C, et al
      . Localisation of a low-molecular-weight bronchial protease inhibitor in the peripheral human lung. Thorax 1983;38:180–3. doi:10.1136/thx.38.3.180
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Madsen J,
      2. Kliem A,
      3. Tornoe I, et al
      . Localization of lung surfactant protein D on mucosal surfaces in human tissues. J Immunol 2000;164:5866–70. doi:10.4049/jimmunol.164.11.5866
      OpenUrlAbstract/FREE Full Text
    33. ↵
      1. Stahlman MT,
      2. Gray ME,
      3. Hull WM, et al
      . Immunolocalization of surfactant protein-D (SP-D) in human fetal, newborn, and adult tissues. J Histochem Cytochem 2002;50:651–60. doi:10.1177/002215540205000506
      OpenUrlCrossRefPubMedWeb of Science
    34. ↵
      1. Heron M,
      2. Grutters JC,
      3. ten Dam-Molenkamp KM, et al
      . Bronchoalveolar lavage cell pattern from healthy human lung. Clin Exp Immunol 2012;167:523–31. doi:10.1111/j.1365-2249.2011.04529.x
      OpenUrlCrossRefPubMed
    35. ↵
      1. Stocks JM,
      2. Brantly M,
      3. Pollock D, et al
      . Multi-center study: the biochemical efficacy, safety and tolerability of a new alpha1-proteinase inhibitor, Zemaira. COPD 2006;3:17–23. doi:10.1080/15412550500493220
      OpenUrlCrossRef

    Footnotes

    • Contributors MH, TPC, SR, DDF, MPM, MC, CG and NMcE conceived and initiated the study. MH, TPC, MPM and NMcE designed the study. RCH and TPC confirmed AAT status in the participants. MH performed the bronchoscopy, bronchoalveolar lavage and phlebotomy. MH and MPM processed the bronchoalveolar lavage. MH, EL, LK and MPM processed the phlebotomy samples. MH, RCH, LK, EL and MPM designed and performed experiments. SR, DDF, MH, MC, CG and NMcE performed clinical follow up in the participants. SR, DDF, RCH, LK, EL and MH performed follow up questionnaires in the participants. MH drafted and wrote the manuscript. MH, SR TPC, MPM and NMcE edited and revised the manuscript. MH, SR, DDF, RCH, LK, EL, MC, CG, TPC, MPM and NMcE read the final version of the manuscript and interpreted the results. MH, the corresponding author, is the guarantor of the manuscript content and attests that all listed authors meet authorship criteria.

    • Funding This study was funded by Alpha-1 Foundation, PROMARIC-NSRP; North-South Research Programme; Higher Education Authority (HEA); Department of Further and Higher Education, Research, Innovation, and Science (DFHERIS) and the Shared Island Fund.

    • Competing interests NMcE: In the past 36 months, unrestricted research grant: Grifols, CSL Behring research grant in alpha-1 antitrypsin deficiency; advisory board: Intellia, Vertex, Inhibrx, Takeda, Dicerna, Centessa in the area of alpha-1 antitrypsin deficiency; speaker honorarium: CS Behring ERS; patent for development of resistant form of AAT in CHO cells.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Author note Artificial intelligence was not used at any stage of the study.

    Linked Articles