INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Antiprotease defenses in the lower respiratory tract can be overwhelmed when sufficient numbers of neutrophils enter airspaces and degranulate, releasing various proteolytic enzymes such as the serine proteases neutrophil elastase (NE), cathepsin G, or proteinase 3 (1). Alpha-1-antiprotease (₁-AP; also known as ₁-antitrypsin or ₁-protease inhibitor) is a member of the serpin group of serine protease inhibitors, and it is the major antiprotease that inhibits serine proteases found in distal bronchoalveolar spaces (2). Other antiproteases such as secretory leukoprotease inhibitor (SLPI), ₁-antichymotrypsin, and elafin also contribute to lung antiprotease defenses, with SLPI being the predominant antiprotease of the more proximal conducting portion of the respiratory tract (3). The serpins, ₁-AP and ₁-antichymotrypsin, are predominantly acute-phase reactants produced by the liver and circulating in blood (1), but ₁-AP can also be produced by alveolar macrophages (4) or epithelial cells (5) and ₁-antichymotrypsin by lung epithelium (6). SLPI and elafin are secreted by epithelial cells (7, 8), and SLPI and some serpins are synthesized and secreted by neutrophils (9, 10).

An altered balance between ₁-AP and NE is associated with numerous lung pathologies. ₁-AP deficiency emphysema is associated with impaired NE inhibitory capacity in epithelial surface fluid caused by depressed levels of ₁-AP (11), and intravenous ₁-AP replacement therapy may attenuate the rate of decline in lung function (12). Progressive airflow obstruction in cystic fibrosis (CF) is associated with chronic neutrophilic airway inflammation with an excess of unopposed NE due to overwhelmed antiproteases, but ₁-AP in peripheral blood is normal or increased (13). Bacterial phagocytosis may be impaired in CF owing to cleavage of opsonins or opsonin receptors by unopposed NE (14, 15), and can be restored by neutralizing NE with ₁-AP (16). Overwhelmed antiprotease defenses have also been linked to the development or progression of non-CF bronchiectasis (17). Although neutrophil influx also occurs in inflammatory/fibrotic disorders such as idiopathic pulmonary fibrosis (IPF), lung antiprotease defenses do not appear to be overwhelmed in diffuse infiltrative pulmonary disease (18).

Neutrophil influx into the transplanted lung allograft has been associated with the development of obliterative bronchiolitis (OB) (19, 20), and persistently increased concentrations of neutrophils have been associated with increased mortality posttransplant (21). Antiprotease defenses appear to decline with the appearance of bronchiolitis obliterans syndrome (BOS) after transplant, accompanied by evidence of increased oxidant stress (20). Additionally, King and coworkers found free NE activity in 3 of 7 ₁-AP-deficient transplant recipients during episodes of lung inflammation versus 1 of 4 recipients without ₁-AP deficiency (22). Because ₁-AP deficiency may predispose the lung allograft to proteolytic injury or impair bacterial clearance when significant amounts of unopposed (uninhibited) NE is present, we determined NE activity and measured ₁-AP concentrations in bronchoalveolar lavage fluid (BALF) from a large cohort of lung transplant recipients with ₁-AP deficiency, from recipients without ₁-AP deficiency, and from normal volunteer subjects to determine the effects of time of sampling in relationship to transplantation and the presence of complications such as rejection or infection. Because the presence of enough myeloperoxidase (MPO) to convert H₂O₂ to HOCl indicates increased neutrophil influx into airspaces with release of degradative enzymes from neutrophil granules, we also measured MPO activity as an indicator of neutrophil influx into airspaces with release of granular constituents.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

BALF obtained from 339 recipients who underwent lung or heart- lung transplant at the University of Wisconsin (n = 96) and the University of Pittsburgh (n = 243) were examined. Indications for lung transplantation included emphysema, CF, pulmonary fibrosis, sarcoidosis, non-CF bronchiectasis, OB, or pulmonary vascular disease. The lung transplant recipients included 179 patients who underwent single-lung transplant (SLT), 133 who had bilateral lung transplant (BLT), and 27 with combined heart and lung transplant (HLT). A total of 2,137 bronchoalveolar lavages (BAL) were examined for the presence of unopposed NE and MPO, and 21 BALF from normal subjects were examined for comparative purposes. Some patients with CF for whom the presence of unopposed NE in BALF was previously reported (23) were included in the study group. The protocol was approved by the human subjects committees at both institutions.

Maintenance immunosuppression consisted of cyclosporine A or tacrolimus plus azathioprine and prednisone. BOS was defined as a greater than 20% decline in FEV₁ from the best posttransplant value that did not improve despite enhanced immunosuppression, and acute rejection or OB was diagnosed by means of transbronchial lung biopsy using histologic criteria of Yousem and coworkers (24).

Bronchoscopy

Most BALF were obtained for surveillance purposes to detect infection, and transbronchial biopsies were performed to detect allograft rejection or infection. Bronchoscopy with BAL and transbronchial biopsy were performed at 10 to 14 d and at 30 d after transplantation for many of the transplant recipients. Additional BAL were typically performed at 3, 6, 9, and 12 mo during the first year after transplantation and also for clinical indications at any time point. The distal bronchoscope was wedged in a segmental bronchus, and four 40-ml aliquots (University of Wisconsin) or four 50-ml aliquots (University of Pittsburgh) of sterile, nonpyrogenic, isotonic sodium chloride solution were sequentially instilled through the bronchoscope and immediately recovered by gentle suction. Although this protocol was followed for nearly all BAL, on some occasions BAL was performed with smaller volumes (100 to 120 ml total) if the patient's clinical status was tenuous. The right middle lobe or lingula were lavaged in nearly all cases. If a prominent, focal infiltrate was present on chest X-ray or computed tomographic (CT) scan, BAL and transbronchial biopsies were performed in the area of the focal radiographic abnormality. Multiple transbronchial biopsies were performed in multiple lung segments after BAL to obtain lung tissue for histopathologic examination.

BALF Analysis

BALF was processed as previously described (25). A hemocytometer was used to quantitate cells, and cytocentrifuge slides were analyzed to obtain differential cell counts. Aliquots of BALF were subjected to microbiologic analysis, including quantitative culture for aerobic bacteria, detection of cytomegalovirus (CMV) and other viruses, and fungal culture.

NE activity in BALF was determined spectrophotometrically at 410 nm using a specific substrate, MeO-Suc-Ala-Ala-Pro-Val-pNA, as previously described (26). ₁-AP was measured as previously described (27) using an ELISA. MPO enzyme activity, presumably released by activated or injured polymorphonuclear leukocytes (PMNs), was quantified in samples of cell-free BALF supernatants using a continuous, initial velocity spectrophotometric assay employing 3,3',5,5' tetramethyl benzidine (TMB) as the oxidizable substrate as previously described (26).

Statistical Analysis

All data were analyzed on electronic spreadsheets (SuperCalc4; Computer Associates, San Jose, CA and EXCEL; Microsoft, Inc., Redmond, WA) and with database statistics packages for microcomputers (Abstat 4.1; Anderson-Bell, Parker, CO and SAS System; SAS Institute, Cary, NC). Values are expressed as mean ± SEM unless otherwise stated. Independent t tests, paired tests, multivariate analysis, and multiple regression analyses were performed as appropriate. BAL findings were correlated with clinical status, biopsy findings, and microbiologic analyses of BALF.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Data for BALF cell differential counts and total neutrophil concentrations are given in Table 1. Lung allograft recipients with early posttransplant BALF analyzed (10 to 14 d posttransplant) had no evidence of infection and had transbronchial lung biopsies with acute rejection scores of A0 or A1. Data for all other BAL types were obtained from individual recipients at 2 to 12 mo posttransplant except for some BALF obtained at later time points for some individuals with OB or BOS. Analysis of BALF and transbronchial biopsies revealed only the entity listed under BAL type (Table 1). Mean neutrophil concentrations were increased minimally but significantly in recipients who had no evidence of rejection (A0) or infection compared with normal subjects, and grade A1 was no different from grade A0. Neutrophil influx was increased for increasing acute rejection grade, CMV antigen/culture positivity, or CMV pneumonia. The presence of a fungus on BALF culture, the presence of  10⁵ colony-forming units (cfu)/ml BALF of aerobic bacteria, or the histologic diagnosis of OB or a clinical course typical of BOS were also associated with considerable neutrophil influx.

In recipients with emphysema and ₁-AP deficiency who were clinically stable, showed no evidence of infection, and whose transbronchial biopsy scores were A0 or A1, we found that the mean concentration of ₁-AP in BALF was, as expected, extremely low when compared with healthy, never-smoking normal subjects (Figure 1). The mean concentration of BALF ₁-AP in recipients with CF was nearly the same as that for normal subjects. ₁-AP concentrations were somewhat lower for clinically stable transplant recipients transplanted for emphysema without ₁-AP deficiency or recipients transplanted for other indications (interstitial lung disease and pulmonary vascular disease) compared with that for normal subjects, but these differences were statistically insignificant.

View larger version (17K): [in this window] [in a new window]   Figure 1.   1-AP concentrations in BALF from normal subjects and transplant recipients with or without 1-AP deficiency. All transplant recipients had grade 0 or 1 acute rejection and no evidence of infection at the time of BAL. Concentrations of 1-AP (mean ± SEM) in BALF for normal subjects were 487 ± 83 ng/ml (range 48 to 1,311 ng/ ml, n = 21), 53 ± 17 ng/ml in recipients with 1-AP deficiency (range 9 to 311 ng/ml, n = 22), 373 ± 60 ng/ml (range 46 to 1,909, n = 55) for recipients with emphysema but without 1-AP deficiency, 362 ± 67 ng/ml (range 40 to 1,016 ng/ml, n = 19) for recipients with interstitial lung disease and pulmonary vascular disease, and 482 ± 174 ng/ml (range 48 to 2,133 ng/ml, n = 11) for recipients with CF.

Concentrations of ₁-AP were increased in BALF obtained from individuals without significant rejection (acute rejection grade A0 or A1) or infection within 10 to 14 d of lung transplantation compared with BALF obtained 4 to 6 wk posttransplant from the same individuals (Figure 2). Mean ₁-AP concentrations for recipients with ₁-AP deficiency declined from 663 ± 161 to 139 ± 47 ng/ml (paired specimens, n = 7), indicating that recipients with ₁-AP deficiency can have increased ₁-AP concentrations in airspace secretions early after allograft implantation. Early ₁-AP concentrations were significantly higher for recipients transplanted for emphysema unassociated with ₁-AP deficiency (4,708 ± 1,385 ng/ml, n = 20, p < 0.05) or CF (9,668 ± 2,226 ng/ml, n = 10, p < 0.01) and remained significantly increased at 4 to 6 wk (1,532 ± 521 ng/ml for non-₁-AP-deficiency emphysema, p < 0.05; 2,231 ± 725 ng/ml for CF, p < 0.05) when compared with ₁-AP-deficient recipients. Paired specimens from individuals who went from a stable condition to having  10⁵ cfu aerobic bacteria/ml BALF compatible with lower respiratory tract bacterial infection demonstrated that recipients with ₁-AP deficiency were able to increase ₁-AP concentrations considerably (Figure 3). In comparison, recipients without ₁-AP deficiency transplanted for emphysema were able to elevate ₁-AP in BALF to higher concentrations than recipients with ₁-AP deficiency.

View larger version (19K): [in this window] [in a new window]   Figure 2.   1-AP concentrations in paired BALF specimens at early (10 to 14 d posttransplant) and later (4 to 6 wk posttransplant) time points for transplant recipients with or without 1-AP deficiency. All transplant recipients had grade 0 or 1 acute rejection and no evidence of infection at the time of BAL.

View larger version (21K): [in this window] [in a new window]   Figure 3.   1-AP concentrations in paired specimens of BALF for transplant recipients with or without 1-AP deficiency when stable with grade 0 or 1 acute rejection score versus when  105 cfu/ml aerobic bacteria and neutrophil influx were detected in BALF. 1-AP concentrations increased from 87 ± 30 ng/ml to 4,163 ± 2,171 ng/ml in recipients with 1-AP deficiency (n = 7) and from 443 ± 116 to 14,052 ± 8,340 ng/ml in BALF from recipients with emphysema but no 1-AP deficiency (n = 20).

Unopposed NE activity was present in 171 (8%) of 2,137 BALF. Seventy-six of the 339 (22%) transplant recipients had at least one BALF specimen with unopposed NE activity. Thirty-one had unopposed NE activity in multiple BALF. However, because many of these 339 recipients were not followed prospectively beginning at the time of transplantation, a comparative analysis of the frequency with which episodes of unopposed NE activity in BALF were identified was performed only for recipients who had all BAL analyzed from the time of transplant until at least 1 yr after transplantation (n = 159). As shown in Table 2, 54 of the 159 recipients had at least one BALF with unopposed NE detected, and the highest incidence was observed in recipients with ₁-AP deficiency (10 of 17, 59%). Most episodes of unopposed NE occurred in the setting of bacterial infection ( 10⁵ cfu aerobic bacteria/ml BALF), and many of these were from individuals with established OB or BOS. However, unopposed NE was present in BALF from three recipients with ₁-AP deficiency and also in four recipients transplanted for other indications at 10 to 14 d posttransplant. These recipients lacked microbiologic evidence of infection or biopsy evidence of rejection or infection. Two other recipients had unopposed NE activity in BALF at later time points associated with A3 acute rejection.

MPO activity was detected in at least one BALF from the majority of recipients, indicating that significant neutrophil influx and degranulation often occurred. However, NE release was usually not sufficient to overwhelm ₁-AP in the airspaces when MPO activity was detected in BALF. Seven hundred forty-six of the 2,137 (35%) BALF had measurable MPO activity, indicating significant neutrophil influx and degranulation, but only 171 (8%) had measurable unopposed NE activity. MPO activity (Figure 4) correlated with neutrophils/ml BALF (r = 0.53, p < 0.0001) and with BALF NE activity. However, the correlation coefficient of only 0.53 suggests a somewhat variable degree of neutrophil degranulation and necrosis.

View larger version (20K): [in this window] [in a new window]   Figure 4.   MPO activity (nmol/ml/min) in BALF versus neutrophils × 103/ml (r = 0.53, p < 0.0001).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The human lung allograft is prone to acute, reperfusion injury at the time of implantation (28), to acute rejection which is often accompanied by neutrophil infiltration into airspaces in addition to the infiltrating lymphocytes (29), and to chronic rejection (OB) in which neutrophils also infiltrate the lung and bronchioles are destroyed (19). Additionally, the transplanted lung is especially prone to infection. Finally, bacterial infection is characterized by a large influx of neutrophils into airspaces, and pulmonary bacterial infection frequently occurs in recipients who have developed BOS or histologically proven OB (25, 30). Neutrophil influx occurred under various conditions after transplantation in our transplant recipients as compared with normal subjects, as shown in Table 1. Although a minimal increase in neutrophil concentration was found for acute rejection grade 0 or 1 compared with normal subjects, a progressive increase in neutrophil influx was observed for grade 2 through grade 3 acute rejection. Increased neutrophils were also observed with CMV infection without pneumonia, with the isolation of a fungus from BALF through fungal culture, or with CMV pneumonia when compared with recipients with grade A0 or A1 rejection only. The largest increase in neutrophil concentrations occurred early after transplant, with the isolation of significant colony-forming units of aerobic bacteria, or if OB or BOS (without evidence of coexistent infection or acute rejection) were identified.

Although neutrophil influx and degranulation can be severe enough to overwhelm antiprotease defenses in the lower respiratory tract after lung transplantation as previously reported (22, 25), we found that the antiprotease level was usually adequate to prevent the occurrence of unopposed (free) NE in respiratory secretions when neutrophil influx and the presence of MPO activity were detected. However, many recipients had at least one BAL that demonstrated unopposed NE, particularly those transplanted for emphysema associated with ₁-AP deficiency. Recipients with ₁-AP deficiency had very low concentrations of ₁-AP in BALF but could increase these concentrations in the subacute phase of reperfusion injury or in the setting of airspace infection. Because individuals with ₁-AP deficiency appear to be capable of mounting only a relatively small increase in circulating ₁-AP in response to acute stress (31), these increases in ₁-AP in BALF are likely related to increased epithelial permeability rather than to increased hepatic output of ₁-AP as an acute-phase reactant. Measurement of albumin in BALF or C-reactive protein in peripheral blood may have helped resolve this issue, but these were not measured in this study. Recipients without ₁-AP deficiency had higher concentrations of ₁-AP at early time points or in response to lower respiratory tract bacterial infection. Although the antiprotease concentration was adequate to prevent unopposed NE activity in BALF in most recipients at 10 to 14 d posttransplant, including those with ₁-AP deficiency, some recipients (3 of 17 with ₁-AP deficiency and 4 of 142 with other transplant indications) had unopposed NE at this early time point unaccompanied by significant lung allograft rejection or infection. We speculate that this finding was related to a relative persistence of neutrophils and release of NE as reperfusion injury abated and lung permeability improved such that the epithelial surface fluid was sampled when ₁-AP concentrations were not adequate to neutralize NE completely. At later time points in both ₁-AP-deficient and -AP-sufficient recipients who were clinically stable, no unopposed NE was detected, as was previously reported by King and coworkers (22).

The impact of neutrophil influx into the lung allograft and release of sufficient granule constituents such as NE is unclear, but neutrophil influx has been correlated with poor outcome or the occurrence of OB (19, 20, 21, 25). Additionally, a molar excess of NE can degrade ₁-AP, and oxidants can inactivate it (32, 33). Unopposed NE may not only degrade matrix proteins such as elastin, but may cleave complement, complement receptors, and immunoglobulins (14, 15) and thereby impair clearance of airspace bacterial infection.

Recipients with ₁-AP deficiency appear to have a higher likelihood of having episodes of unopposed NE activity in respiratory tract secretions associated with neutrophil influx, likely owing to their low ₁-AP concentrations in lower respiratory tract secretions with limited ability of other antiproteases such as SLPI to neutralize unbound NE. We speculate that neutralization of unopposed NE may promote bacterial clearance from the lung allograft and limit proteolytic damage causing bronchial injury. However, the generally episodic nature of instances in which antiprotease defenses are overwhelmed makes it difficult to study the effects of antiprotease therapeutic intervention in lung transplant recipients. We also speculate that therapeutic strategies to limit neutrophil influx both with reperfusion of the newly implanted allograft (34) and at later time points with anti-inflammatory, antimicrobial agents such as macrolides, which decrease interleukin-8 concentrations and neutrophil influx into airspaces in diffuse panbronchiolitis (35), may limit graft damage and consequent impairment of lung function, particularly the bronchiolar destruction which characterizes OB/BOS.

Although ₁-AP replacement therapy may hold therapeutic benefit for individuals with evolving pulmonary dysfunction resulting from ₁-AP deficiency (12), such therapy is not recommended posttransplant for recipients with ₁-AP deficiency, because emphysema and pulmonary dysfunction, usually associated with cigarette smoking, evolve over a prolonged period of time in individuals with ₁-AP deficiency (36), and replacement therapy is exceedingly expensive. However, our data suggest that recipients with ₁-AP deficiency are especially likely to have the antiprotease defenses of the lower respiratory tract overwhelmed when significant neutrophil influx occurs, and this may occur early after transplantation in the absence of infection or significant rejection. One recipient in our study developed persistent bacterial infection accompanied by the presence of unopposed NE in multiple sequential BALF and progressive bronchiectasis in his lung allograft during the first year after transplantation. Despite rotating antibiotics his FEV₁ gradually decreased to 30% of his best posttransplant value and he required supplemental oxygen. When ₁-AP replacement therapy was resumed, his purulent secretions gradually abated as his FEV₁ gradually climbed to exceed his former best posttransplant FEV₁. His requirement for supplemental oxygen ceased, and the large neutrophil influx in his BALF subsided as the antiprotease defenses appeared to be restored with complete neutralization of unopposed NE. This recipient continues to have stable lung function more than 5 yr after transplant (4 yr after intravenous ₁-AP replacement therapy combined with chronic azalide administration was initiated) while continued on ₁-AP replacement therapy.

In summary, mean values for neutrophils/ml BALF are considerably increased for lung transplant recipients early after transplant in the apparent absence of significant rejection or infection. Mean values for neutrophils/ml BALF remain significantly but mildly elevated at later time points with no or minimal acute rejection identified on biopsy specimens and no evidence of infection. These values increase more dramatically with increasing rejection grade, recovery of bacterial or viral pathogens, or the development of BOS. Although antiproteolytic defenses against serine proteases can prevent unopposed NE activity in lower respiratory tract secretions, even with episodes of bacterial infection, these defenses can be overwhelmed in both ₁-AP deficiency and transplant recipients without ₁-AP deficiency. Recipients with ₁-AP deficiency may, however, be more prone to episodes of unopposed NE in lower respiratory tract secretions when significant numbers of neutrophils infiltrate the allograft with reperfusion injury, acute rejection, or infection.