INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The lower respiratory tract is usually kept sterile by effective local host defenses. However, in the presence of established airways disease, bacteria are often isolated from expectorated mucus even when the patient appears clinically stable (1). This presents problems in the determination of the cause of an acute exacerbation of airways disease and hence rationalization of therapy. In a recent review these problems were highlighted although it was emphasized that when the bacterial load was low, the local host defenses may be adequate (2). However, as the bacterial load increases, recruitment of the secondary host defenses would result in increased airways inflammation and neutrophil influx (2). This concept is supported by the expectation that recruitment of secondary host defenses (especially neutrophils) would lead to the development of purulent secretions as a result of the presence of myeloperoxidase (MPO) (3). In the classic study by Anthonisen and colleagues (4), antibiotic therapy only showed a significant benefit in patients with chronic obstructive pulmonary disease (COPD) who by definition had to have all presenting symptoms for their exacerbation including increased breathlessness, sputum volume, and sputum purulence. The resolution of such episodes requires a combination of effective therapy resulting in the loss of the inflammatory stimulus and an appropriate acute phase response (2). In this respect ₁-antitrypsin (AAT) may play a critical role by inactivating elastase activity released by the activated airway neutrophils, both downregulating inflammation and protecting airways tissue from enzyme-induced damage (2).

Subjects with AAT deficiency often have airways disease (5, 6) and hence will be susceptible to bacterial colonization and acute bacterial infective exacerbations. However, because of their deficiency, such subjects will not have an appropriate AAT acute phase response which may therefore influence the degree of airways damage caused by such infections. The nature of the airway inflammation and its response to bacterial exacerbations have yet to be studied in AAT deficiency.

The purposes of the present study were twofold: first, to assess the inflammatory nature of acute exacerbations of COPD in subjects with AAT deficiency and compare this with COPD patients without deficiency; second, to monitor the inflammatory process and its resolution in patients with AAT deficiency after appropriate antibacterial therapy. In particular, we wished to determine the concentration of chemoattractants interleukin-8 (IL-8) and leukotriene B₄ (LTB₄) in the bronchial secretions, the recruitment of neutrophils as reflected by sputum MPO activity, the sputum concentration of active neutrophil elastase and its natural inhibitors AAT and secretory leukoprotease inhibitor (SLPI), as well as protein leakage as a measure of airways inflammation. In addition, we wanted to assess the serum acute phase response of both C-reactive protein (CRP) and AAT.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We studied 11 patients with chronic cough and sputum production, who were known to have homozygous AAT deficiency (PiZ), and were attending a specialist clinic. None of the patients had ever received AAT replacement therapy. The patients contacted one of us (A.H.) and were seen in the clinic within 48 h of the onset of new or worsening symptoms suggestive of an acute exacerbation of their disease. The presence of an exacerbation was based on the criteria suggested by Anthonisen and colleagues (4), and included an increase in all three of the symptoms of breathlessness, sputum volume, and sputum purulence.

On the day of clinic visit, the patients collected sputum over 4 h from rising. The sample was assessed macroscopically and assigned a number by one of us (A.H.) by comparison with a sputum color chart (Bronkotest; Heredilab, Salt Lake City, UT). Values were assigned a number from 0 (water clear) to 8 (corresponding to the deepest color observed in patients with cystic fibrosis). The sputum grades zero to 2 are macroscopically mucoid in appearance (clear/white), numbers 3 to 5 are macroscopically mucopurulent (yellow/green), and 6 to 8 are macroscopically frankly purulent (dark yellow/green).

Each sample was divided into two portions. One was used to obtain a Gram stain and quantitative bacterial culture (7). The remaining portion was ultracentrifuged at 50,000 × g (4° C) for 90 min, and then the sol phase was removed and stored (70° C) until it was analyzed. A 10-ml sample of venous blood was obtained and allowed to clot and the serum was removed for analysis.

Patients were then started on 14 d antibiotics. They were commenced routinely on amoxycillin 500 mg 3 times daily unless Moraxella catarrhalis was suspected on Gram stain, and such patients were started on cefuroxime axetil 500 mg twice daily. These antibiotics were unchanged except in one patient where cefuroxime was changed to ciprofloxacin 750 mg twice daily the following day as Pseudomonas aeruginosa was cultured in addition to Moraxella catarrhalis (the dominant organism). No patient received oral steroid therapy, and all other treatment (including inhaled steroids) remained constant throughout the study. The patients returned 1, 3, 5, 7, 14, and 28 d after commencement of therapy and at these visits samples were collected and assessed as previously described except for follow-up Day 1 when serum alone was collected. All patients filled in a daily diary card on which they recorded their morning and evening peak expiratory flow rate (PEFR) (best of three), as well as their symptoms. Finally the patients indicated the day on which they felt back to their stable clinical state.

The data obtained on the day of presentation were compared with that from 11 control patients with COPD who had normal AAT. These patients were chosen from a database of 136 outpatients presenting with a similar acute exacerbation (increasing breathlessness, increasing sputum volume, and purulence). These control patients were matched with the AAT-deficient patients for stable state FEV₁ (percent predicted) to ensure a comparable degree of lung function impairment. However, complete samples for the control patients were only available for the day of presentation, Day 5, and Day 28 (stable clinical state) for comparison with the AAT-deficient patients.

Sputum Biochemistry

MPO. Sputum MPO activity was used as an assessment of neutrophil influx and measured as described previously (8). The MPO concentration was derived from the standard curve using a single preparation of lysed neutrophils (freeze/thaw). Results for individual samples were obtained from this curve by interpolation and expressed as arbitrary units/ml. The interassay coefficient of variation (standard deviation/ mean × 100) was less than 10%.

Chemoattractants. Sputum IL-8 was measured by ELISA using a commercially available kit (R&D Systems Europe Ltd., Abingdon, UK) and LTB₄ was measured by ELISA (Amersham International plc, Buckinghamshire, UK). The lower limit of detection for these assays was 0.008 nM for IL-8 and 0.17 nM for LTB₄. The interassay coefficient of variation of these two assays was less than 10% and samples spiked with pure compound resulted in greater than 85% recovery.

Elastase activity. Neutrophil elastase activity present in the samples was measured spectrophotometrically using the synthetic substrate methoxysuccinyl-ala-ala-pro-val-paranitroanilide (MeOSAAPVpNa) as described previously (9). The interassay coefficient of variation was less than 1% with a lower limit of detection of 1 nM. Samples with activity below 1 nM were considered to be zero for statistical purposes.

Sputum elastase inhibitors: AAT. AAT concentration in sputum sol phase was measured by ELISA relative to a commercially available serum standard (The Binding Site Ltd., Birmingham, UK) as described previously (10). The AAT concentration (µM) was obtained by interpolation from the standard curve (interassay coefficient of variation was less than 5%).

Sputum elastase inhibitors: secretory leukoprotease inhibitor. SLPI was measured by ELISA using a commercially available kit (R&D Systems Europe Ltd., Abingdon, UK). The lower limit of detection was 0.01 µM for SLPI. The interassay coefficient of variation was less than 10% and samples spiked with pure compound resulted in greater than 85% recovery.

Protein leakage. Albumin in sputum sol and serum was measured by radial immunodiffusion using a commercially available kit (The Binding Site Ltd., Birmingham, UK). The interassay coefficient of variation was less than 5%. The sputum/serum ratio for albumin was obtained for each patient sample and multiplied by 100 for convenience as described previously (11). This value was used as a measure of protein transudation from plasma indicating the degree of lung airway leakage.

Acute phase response. CRP and AAT were measured in serum by radial immunodiffusion using commercially available kits (The Binding Site Ltd., Birmingham, UK). The interassay coefficient of variation was less than 5% for both assays.

Statistical Analysis

Values are reported as mean (± standard error). The Friedman test was used to compare the inflammatory markers throughout the exacerbation, and, where significant, the Wilcoxon rank test for paired data was used to compare results from the start of the exacerbation to different points during the exacerbation. The Mann-Whitney U test for unpaired data was used to compare data between patients with and without AAT deficiency. A p value < 0.05 was considered to be statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Demographic details for the AAT-deficient and control patients are shown in Table 1. The AAT-deficient group were younger but both groups were receiving similar therapy and, in particular, eight of the AAT-deficient group and nine of the control group were on long-term inhaled corticosteroids. No patients were on oral steroid therapy and neither group had evidence of bronchiectasis either clinically or visible on high-resolution computed tomography of the chest. The results are shown for postbronchodilator FEV₁, ratio of forced expiratory volume in one second to vital capacity (FEV₁/VC), TLC, and gas transfer (kCO) expressed as a percentage of the value predicted for the patient's age, height, and sex (12). Neither group had significant reversibility (> 12% increase) to nebulized ₂-agonist (salbutamol 5 mg). The average rise in FEV₁ was 55.0 ml (± 10.8) for the AAT-deficient group and 65.0 ml (± 36.4) for the control group.

Bacteriology

All samples obtained at presentation had greater than 25 white blood cells (with < 10 epithelial cells) per high-power field on Gram stain. A predominant bacterial species was seen in all 11 samples from the AAT-deficient patients and these subsequently grew one or two bacterial pathogens (Table 2) with a median bacterial load of 6.6 × 10⁸ colony-forming units per milliliter (cfu/ml) (range 2.7 × 10⁷ to 5.4 × 10⁹). The sputum samples from the non-AAT-deficient patients also contained one or two bacterial organisms on quantitative culture. The predominant species included nontypeable Haemophilus influenzae (n = 6), Streptococcus pneumoniae (n = 3), and Moraxella catarrhalis (n = 2). The bacterial load was similar to that for the AAT-deficient group (median 2.3 × 10⁸ cfu/ml; range 1.2 × 10⁷ to 9.0 × 10⁹).

Sputum Analysis at Presentation

When first seen, all patients with AAT deficiency had purulent sputum (see METHODS) with a mean sputum color value of 4.6 (SE ± 0.3) which is consistent with the high sputum MPO level. In addition, the sputum concentration of both the chemoattractants IL-8 and LTB₄ were high, and free elastase activity was present in 10 of the 11 sputum samples (range, 5 to 1,631 nM). Although the sputum SLPI concentration was low, the sputum AAT concentration was relatively high for this group consistent with increased protein leakage from serum. This increased leakage was supported by the high sputum/serum albumin ratio.

The control patients without AAT deficiency showed differences compared with the patients with AAT deficiency at presentation (Table 3). Although the average sputum MPO value was similar to that seen in the AAT-deficient group (p = 0.17), the sputum elastase activity (detectable in eight of 11 patients) was lower (p = 0.02) with a range from 0 to 166 nM in the nondeficient patients. In addition, the average sputum concentrations of both IL-8 and LTB₄ were lower in the nondeficient group (p = 0.01 and p = 0.02, respectively) and the protein leakage was lower although this just failed to reach statistical significance (p = 0.06). The sputum concentrations of both SLPI and AAT were higher in the nondeficient group (p = 0.02 and p < 0.001, respectively).

The Effect of Antibiotic Therapy in Patients with AAT Deficiency

After 14 d of therapy the pathogen had been cleared in eight of the 11 AAT-deficient patients and the bacterial load was lower in the remaining three (Table 2). All patients improved clinically (as assessed by the symptom diary card) during antibiotic therapy. There was a reduction in sputum color (p < 0.01), MPO (p < 0.005), IL-8 (p < 0.01), LTB₄ (p < 0.005), and sputum elastase activity (p < 0.01) which decreased in the 10 subjects in whom it was present at the start of the exacerbation, becoming undetectable in two. The sputum AAT fell (0.11 ± 0.04 µM, p < 0.005) and the SLPI concentration rose (p < 0.01). Finally there was a reduction in protein leakage in the airway as the sputum/serum albumin ratio decreased (p < 0.01). These results are summarized in Figures 1-4.

View larger version (15K): [in this window] [in a new window]   Figure 1.   Time course of response of sputum color (open bars) and MPO (solid bars) after treatment for the acute exacerbation. The average data ± standard error are shown for samples at presentation, at different times during therapy up to Day 14, and in the stable state 2 wk after cessation of therapy. Asterisks indicate values that are significantly reduced compared with those at presentation (*p < 0.05; **p < 0.01; ***p < 0.005).

View larger version (13K): [in this window] [in a new window]   Figure 2.   Sputum chemoattractant concentrations during the acute exacerbation. The mean values for IL-8 (open bars) and LTB4 (solid bars) are shown at presentation, during the course of treatment for the exacerbation, and 2 wk later when clinically stable (± standard error). Asterisks indicate average values significantly lower than at presentation (*p < 0.05; **p < 0.01; ***p < 0.005). Values for LTB4 in the stable clinical state are significantly higher than at the end of treatment (p < 0.05).

View larger version (15K): [in this window] [in a new window]   Figure 3.   Time course of changes in elastase activity (open bars) and secretory leukoprotease inhibitor (solid bars). Asterisks indicate values for elastase or SLPI that were significantly different from presentation (*p < 0.05; **p < 0.01; ***p < 0.005).

View larger version (8K): [in this window] [in a new window]   Figure 4.   Sputum/serum albumin ratio is shown during the course of the exacerbation (± standard error). Asterisks indicate values significantly lower than at presentation (*p < 0.05; **p < 0.01).

Acute Phase Response

At presentation the serum CRP and AAT concentrations were raised to average values of 34.0 mg/L (± 18.5) and 4.8 µM (± 0.7), respectively, in patients with AAT deficiency (Figure 5). The results for CRP were similar to that for the nondeficient subjects (42.9 mg/L ± 19.3, p = 0.9). On the other hand, the control subjects had greater (p < 0.001) serum concentrations of AAT (31.3 ± 3.4 µM) at the start of the exacerbation, as would be expected.

View larger version (12K): [in this window] [in a new window]   Figure 5.   Acute phase proteins. The average value (± standard error) for CRP and 1-antitrypsin are shown during the course of the exacerbation. Asterisks indicate values that are higher than in the stable clinical state (Day 28) (*p < 0.05; **p < 0.01).

By the end of therapy (Day 14) the serum values for the AAT-deficient subjects had fallen significantly (CRP: 8.9 mg/ L ± 3.9, p = 0.02; AAT 4.1 ± 0.5 µM, p = 0.03) and remained low when the patients were reviewed on Day 28 while still clinically stable (CRP: 7.0 mg/L ± 2.9 and AAT 3.6 ± 0.5 µM). Similarly, both the serum values of CRP and AAT were reduced at this stage in the nondeficient subjects (CRP 7.0 mg/L ± 2.6 and AAT 23.8 ± 2.2 µM, both p < 0.05).

Time Course of the Response in Patients with AAT Deficiency

The patients showed a gradual improvement in symptoms and indicated a return to their stable clinical state (on the diary card), on average 13.2 (± 2.1) d after the start of therapy. The biochemical response to antibiotic therapy, however, was generally rapid with a significant decrease in sputum color number (p < 0.05) to 3.8 ± 0.2 by the third day of treatment. This was associated with a significant decrease in sputum MPO (p < 0.01), IL-8 (p < 0.005), and LTB₄ concentrations (p = 0.02) by the third day of treatment. The lowest average concentrations of MPO, IL-8, and LTB₄ were found by the fifth, seventh, and fourteenth day of therapy, respectively (these results are summarized in Figures 1 and 2). By the third day of treatment sputum elastase activity had decreased (p < 0.01) (Figure 3) and the sputum AAT concentration had fallen (p < 0.005) to 0.17 µM (± 0.06). This latter effect was related to a rapid fall in protein leakage (p = 0.02) as shown in Figure 4. However, the sputum SLPI concentration rose (p < 0.005) to 2.9 ± 1.3 µM by the third day of treatment (Figure 3).

On the fifth day, the MPO, elastase activity, IL-8, LTB₄, and protein leakage remained low in the deficient subjects (Figures 1-4). These results were similar to those observed in the nondeficient patients at this time with the exception that patients with AAT deficiency still had higher IL-8 concentration and elastase activity (AAT-deficient patients: IL-8 = 14.1 ± 2.8 nM and elastase activity = 60.9 ± 44.0 nM; nondeficient patients: IL-8 = 4.9 ± 1.9 nM and elastase activity = 2.6 ± 1.7 nM; both p < 0.005) and the sputum AAT concentration remained lower (p < 0.001) in the deficient patients compared with the control patients (0.12 µM ± 0.04 and 0.48 µM ± 0.11, respectively).

The serum acute phase response in patients with AAT deficiency was most obvious with CRP which peaked on the first day after presentation and decreased significantly (p < 0.05) by the fifth day of treatment (Figure 5). The AAT response was small but peaked at Day 3, falling thereafter (average presentation = 4.8 ± 0.7 µM; Day 3 = 5.5 ± 0.7 µM; Day 14 = 4.1 ± 0.5 µM; and Day 28 = 3.6 ± 0.5 µM).

Stable Clinical State

When the AAT-deficient patients were restudied 28 d after presentation (14 d after cessation of therapy), seven were colonized with an identifiable organism (Table 2) although the viable bacterial load in these seven subjects (median value on Day 28 = 8.4 × 10⁷; range = 5.0 × 10⁶ to 4.2 × 10⁹) was lower than that at presentation (median = 6.6 × 10⁸ organisms/ml; range = 2.7 × 10⁷ to 5.4 × 10⁹; p < 0.05).

The continued well-being of the patients was reflected in a persistently lower sputum color compared with that at presentation (p < 0.01). The average sputum MPO and IL-8 concentration remained low although the sputum LTB₄ concentration (Table 4) was higher (p = 0.02) than at the end of therapy (9.5 ± 2.6 nM) as summarized in Figure 2. Elastase activity became detectable in all but one patient by Day 28 although the average value remained low (Figure 3). Finally sputum SLPI concentration remained high (Figure 3) and the sputum AAT concentration remained low (0.10 µM ± 0.03), consistent with a persistent reduction in protein leakage (Figure 4).

The results obtained on Day 28 were not significantly different from those obtained in the AAT-deficient patients when they had been studied in the stable clinical state more than 6 wk before their exacerbation (Table 4).

All nondeficient patients improved with significant reductions in MPO (0.4 ± 0.1 arbitrary U/ml, p = 0.02), elastase activity (1.7 ± 1.1 nM, p = 0.02), LTB₄ (4.9 ± 1.3 nM, p = 0.01), sputum AAT (0.49 µM ± 0.10, p = 0.02), albumin leakage (0.6 ± 0.1%, p < 0.01), and CRP (7.0 mg/L ± 2.6, p = 0.04) by Day 28 and the SLPI concentrations rose by Day 28 (11.8 µM ± 6.5, p = 0.02). Comparison with the results obtained in the AAT-deficient subjects on Day 28 showed that some differences still existed as the MPO, LTB₄, and elastase activity were all lower in the nondeficient group (p < 0.001, p < 0.01, and p < 0.001, respectively) whereas sputum AAT was higher (p < 0.001). However, the IL-8 (8.8 nM ± 2.9), SLPI, albumin ratio, CRP, and the bacterial load (six were colonized with a median bacterial load of 5.5 × 10⁷ cfu/ml [range, 3.0 × 10⁶ to 1.7 × 10⁹ cfu/ml]) were similar to that of the AAT-deficient subjects.

PEFR

Peak flow rates were recorded throughout the exacerbation. In patients with AAT deficiency, the mean morning prebronchodilator PEFR at presentation with the exacerbation was 236.4 L/min ± 41.4, which increased (p < 0.005) by Day 14 to 261.8 L/min ± 45.4 (11.9%). The mean evening postbronchodilator PEFR at the start of the exacerbation was 260.0 L/min ± 40.5 which also increased (p < 0.005) by Day 14 to 278.6 L/min ± 40.8 (10.2%). The mean morning and evening PEFR on Day 28 were not significantly different from Day 14 (Day 28: mean morning 265.5 L/min ± 44.6; mean evening 285.5 L/min ± 43.6).

The peak flow rates in the nondeficient patients were not significantly different from the AAT-deficient group throughout the study and followed a similar pattern. The mean morning prebronchodilator PEFR at presentation with the exacerbation was 205.0 L/min ± 32.7, which also increased (p < 0.005) by Day 14 to 225.5 L/min ± 32.2 (12.1%). The mean evening postbronchodilator PEFR at the start of the exacerbation was 216.8 L/min ± 30.6 which increased (p < 0.005) by Day 14 to 236.4 L/min ± 31.7 (9.6%). The mean morning and evening PEFR on Day 28 were not significantly different from Day 14 (Day 28: mean morning 223.6 L/min ± 31.7; mean evening 237.3 L/min ± 31.0).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study all AAT-deficient patients presented with an increase in symptoms and purulent sputum production associated with the presence of a high bacterial load of a single or (in three cases) two bacterial pathogens. The exacerbations were characterized by high sputum color and MPO concentration consistent with the presence of a significant number of neutrophils, which is probably a result (at least in part) of the high concentration of the two chemoattractants IL-8 and LTB₄.

Elastase activity was present in most of the samples at presentation and may be expected to have a detrimental effect on airways host defenses (13). The enzyme activity probably reflects several factors including increased neutrophil recruitment (as indicated by MPO), reduced SLPI concentration probably as a result of suppression of inhibitor release by elastase itself (14), and low AAT concentration (see RESULTS), despite a small acute phase response and increased protein leakage into the airways.

This latter concept is supported by comparison with subjects who had normal AAT. In these patients, the severity of the exacerbation indicated by the acute phase protein CRP, influx of neutrophils (as assessed by MPO), and bacterial load was similar. However, the activity of elastase in the sputum was much lower, which probably reflects the greater influx of AAT. This inactivation of elastase would, in turn, prevent the effect on SLPI release accounting for the higher concentration of this protein which, in turn, would also facilitate elastase inactivation.

Of interest, the subjects with AAT deficiency also had higher levels of the chemoattractants LTB₄ and IL-8 than the nondeficient subjects at presentation. The exact source of these chemoattractants is uncertain because both can be derived from the activated neutrophils (15, 16). In addition, the infection could stimulate epithelial cells to release IL-8 (17). However, both patient groups had a similar influx of neutrophils (MPO) and bacterial load, suggesting this was not the case. Previous studies have suggested that macrophage release of LTB₄ is increased in AAT deficiency as a result of free elastase activity (18). In addition, elastase has been reported to increase IL-8 release from epithelial cells (19). Thus the greater elastase activity may explain the increases in both chemoattractants seen in AAT deficiency. Nevertheless it might be predicted that the greater concentration of chemoattractants should increase neutrophil recruitment (MPO) in the AAT-deficient group, and yet, this was not the case (see RESULTS). However, recent studies from our group have indicated that the relationship between both IL-8 and LTB₄ and MPO is shallow (9). The differences in chemoattractant concentrations between these two patient groups would only produce a predicted change in MPO equivalent to 0.5 arbitrary U/ml which is unlikely to be detected in the relatively small group of patients studied here.

After initiation of therapy in the AAT-deficient group, the inflammatory response reversed briskly in the group as a whole. Sputum color, MPO, and the chemoattractants all declined, as did the elastase activity. This latter effect was probably enhanced by the rise in SLPI concentration (approximately 2 µM), despite the slight decrease in average sputum AAT concentration of 0.08 µM by Day 3 (due mainly to decreased protein leakage). The sputum became sterile in the majority of subjects and the systemic acute phase response (both CRP and AAT) settled. These results provide an objective measure to support the patients' subjective feeling of improvement and a return to their normal clinical state and are consistent with a resolution of the exacerbation and the associated inflammation.

Two weeks after the cessation of therapy the sputum remained sterile in four of 11 patients and in the other patients the bacterial load was less than at presentation. Inflammation remained low as indicated by the sputum color, MPO, IL-8, elastase activity (although the majority of samples had become positive), protein leakage, and acute phase response, although the LTB₄ concentration had risen (though less than at presentation). These results are, however, consistent with the patient's usual clinical state, as confirmed by comparison with samples from the 11 patients studied at least 6 wk before the clinical exacerbation (see Table 4). In addition, these extra data confirm that the change at the start of the exacerbation was major and indicated a clear pathological process associated with their clinical deterioration.

Of interest, however, the LTB₄ concentration had risen in the whole group studied here 2 wk after treatment had ceased although the reasons are currently uncertain. It is unlikely that this change represents the early stages of relapse after antibiotic therapy because the results are similar to those obtained in the stable clinical state at least 6 wk before the exacerbation (Table 4). In previous studies in AAT deficiency, Hubbard and coworkers identified high LTB₄ concentrations in bronchoalveolar lavage (18). They argued that the source could be macrophages stimulated by neutrophil elastase as a result of the defective AAT inhibitory screen. Whether this mechanism is correct and applicable to the airways, where SLPI is the major antielastase in the stable state (20), remains to be clarified.

However, it should be noted that the elastase activity in these samples was not significantly greater than at the end of treatment (Day 14) although more samples were positive. In addition, more samples had become colonized with bacteria and this could also have an effect by stimulating LTB₄ release from airway phagocytes. Clearly this complex interrelationship requires further study.

Some differences were still observed between the AAT- deficient and nondeficient subjects by Day 28. The nondeficient patients had higher sputum AAT (as would be expected) but lower MPO, elastase activity, and LTB₄. Although the nature of the study meant that preexacerbation data were not available for all the nondeficient patients, subsequent data collected at least 2 mo after the exacerbation were unaltered (data not shown). This suggests that Day 28 data were also equivalent to the stable state in these patients and that AAT deficiency is associated not only with increased inflammation at the start of the exacerbation but also in the stable state. However, several factors including lung function (21), smoking habit (22, 23), airway bacterial load (24), and treatment such as inhaled corticosteroids can influence inflammation (8). For this reason we were careful to match both groups for all these potential confounding factors as well as the severity of their exacerbation as indicated by the objective measurement of CRP. It is likely, therefore, that the results reported here reflect, predominantly, the AAT deficiency itself and increased upper airways inflammation as an ongoing phenomenon. Clearly further studies are indicated to clarify whether this alone determines the more rapid rate of progression of disease in AAT deficiency.

Finally, it is worth commenting on the acute phase response of serum AAT in the deficient subjects. Although basal levels are low in AAT deficiency, acute phase responses can be induced by danazol (25) and tamoxifen (26). This is the first report, to our knowledge, of the acute phase response in a naturally occurring infection in AAT deficiency. Unlike CRP where the response is early and marked, and resolution is rapid, the AAT response is minimal, slower (peaking on Day 3), and falls slowly. The average increase is approximately 40% but it should be emphasized that the highest concentration (5.5 µM) is still well below normal plasma concentrations in the absence of an acute phase response or even the putative lung protective threshold of 11 µM (27). Thus, despite the increase in protein leakage into the airway, the ability of AAT to modulate the inflammatory process is still markedly impaired in AAT-deficient subjects.

In conclusion, acute exacerbations of COPD in subjects with AAT deficiency, in the presence of bacteria in secretions, are associated with marked neutrophil influx. Although the specific nature of these episodes can only remain speculative, the poor AAT acute phase response and the brisk response after commencement of antibiotic therapy suggests that such episodes are caused by the bacteria and should be promptly treated to protect the airways tissues. Whether AAT deficiency makes patients more susceptible to bacterial colonization of the airways remains to be determined. However, the current study suggests that AAT replacement may play a role in modulation of elastase activity in the airway during acute exacerbations in deficient subjects.