INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: bronchoalveolar lavage; lung transplantation; CD4-CD8 ratio; bronchiolitis obliterans; reference values

Lung transplantation (LTX) is the ultimate and often last therapeutic option for several end-stage lung diseases. The outcome after LTX is mainly determined by the development of chronic graft failure: obliterative bronchiolitis (OB). The latter may even develop during the first months posttransplantation and is the main cause of death after the first 6 mo posttransplantation. Furthermore, OB strongly limits the personally gained quality of life by LTX (1, 2). Early identification of patients "at risk" for development of OB is very important, because intervention by modulating the immunosuppressive regimens may prevent or slow down the "natural course" of the developing chronic graft failure. In contrast, once OB has developed only retransplantation is therapeutically available (3).

Besides transbronchial biopsies and lung function testing, monitoring patients after LTX by bronchoalveolar lavage fluid (BALF) analysis is increasingly used to obtain more insight into the pathogenesis of chronic allograft dysfunction and to diagnose OB (4). Recently we have shown that the differential cell count and cytokine profile of BALF can identify patients "at risk" for developing OB without having abnormal lung function or histopathological findings at that time (6).

To define an abnormal BALF profile, information on the "normal" BALF profile found in patients without further pathological airway conditions after LTX is needed. BALF results from patients after LTX are incomparable to BALF results from healthy subjects possibly due to the immunological allograft response and the immunosuppressive therapeutic regimens (9). Furthermore, BALF profiles are likely to undergo a dynamic change the first years after LTX (12). When reviewing the literature on BALF analysis in "good outcome" LTX groups, a large variability between studies is found with respect to the procedures used for BALF sampling, timing of the BAL procedure (days after LTX), and quantified cellular characteristics of BALF. This indicates the difficulty in defining "normal values," even though all patients in the reviewed studies are being described as "healthy" or "having no complications" (Table 1) (4, 5, 8, 10, 11, 16).

The aim of the present study was to define the "normal" cellular profile of BALF of patients after LTX without any signs or symptoms of accompanying airway pathology. Therefore, in a prospective cohort of LTX patients, we set out to assess BALF characteristics in a well-defined subpopulation of "good outcome" LTX patients in whom BAL was performed at fixed intervals during the first 2 yr after LTX.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients and Study Design

A prospective cohort study analysis was performed on BALF samples from all patients transplanted in our center between November 1990 and December 1998. BALF samples were excluded for analysis when acute rejection; OB (histological and/or functional decline within 2 yr after the last BAL); bacterial respiratory tract infections (also clinical diagnosis of purulent bronchitis made by bronchoscopy); or fungal, cytomegalovirus (CMV), or other viral infections were present at the time of the BAL procedure. Single, bilateral, and heart-lung TX were performed according to established techniques (22).

Diagnostic Protocol and Follow-up

Graft function was assessed by pulmonary function testing and routinely performed transbronchial biopsies before discharge and every 6 mo after LTX (6). Acute allograft rejection was diagnosed clinically or histopathologically as defined by Yousem and coworkers (23). Diagnosis of OB was based on a classification and grading by the ISHLT (23). Diagnosis of bacterial or fungal infections was based on positive BALF cultures. Active CMV infection was assessed by CMV serology and CMV antigenemia testing as described before (6, 24, 25).

Therapeutic Protocol

Immunosuppression included up to five gifts of antithymocyte globulins (rATG, Merieux, dosed at 3 mg/kg) in the first 10 d after transplantation. The maintenance immunosuppressive regimen consisted of cyclosporin (aimed at serum levels of 400 ng/ml post-LTX, tapered down in 3 wk to 150 ng/ml and maintained at this level for all time periods), azathioprine (1-3 mg/kg/d), and prednisolone (0.1-0.2 mg/kg/d). The newer immunosuppressive drugs tacrolimus and mycophenolate mofetil were not used in the first 7 yr of our study. All patients received acyclovir and cotrimoxazole prophylaxis (6).

Bronchoalveolar Lavage and Cell Isolation

BAL and bronchoscopy were routinely performed after the first month and every 6 mo after LTX and additionally in case of clinical indication. This protocol was approved by the Medical Ethics Committee. Two aliquots of 20 ml and three aliquots of 50 ml prewarmed phosphate-buffered saline (PBS) were instilled. The first 20 ml portion was investigated for viruses, bacteria, and fungi, the second 20 ml portion was isolated as bronchial fraction (BF), and the three 50 ml fractions were pooled for the alveolar fraction (AF) (6, 26, 27).

The BF and AF were further processed for leukocyte differentiation as described earlier (6) and FACS. The differentiation was determined by counting 200 cells on two slides each.

Fluorescence-activated Cell Sorter Analysis

To quantify percentage and total number of lymphocyte subtypes direct immunofluorescence in the AF was performed with monoclonal antibodies (MAb) phycoerythrin and fluorescein isothiocyanate conjugated. The following MAbs were used: CD3/CD4, CD3/CD8 (IQ-Products), CD4/CD25, HLADR/CD4, HLADR/CD8 (Beckton Dickinson), and CD45RO/CD4, CD45RO/CD8 (Dako, Denmark). Briefly, one million cells of the AF were incubated for 30 min at 4° C with 10 µl of MAb followed by centrifugation at 590 × g for 5 min at 10° C and washing two times with 0.5% (wt/vol) bovine serum albumin (BSA) in PBS. Labeled cells were analyzed using a fluorescence-activated cell sorter-Calibur (Beckton Dickinson).

Statistical Analysis

Data were analyzed using SPSS/PC⁺ software (SPSS Benelux b.v., Gorinchem, The Netherlands). Cell counts and lymphocyte subtypes were compared using the nonparametric Wilcoxon signed rank test for longitudinal analysis. The differences shown in the cross-sectional data are calculated on longitudinal analysis. p Values < 0.05 were considered as significant. All results are presented as median and range, unless otherwise stated.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchoalveolar Lavage Selection

BALF samples (501) were collected between 1990 and 1998. A total of 201 samples were excluded for further analysis because of the death of the respective individual or development of OB within 2 yr of the last BAL. Of the remaining 300 samples, 67 were excluded because of signs or symptoms of infectious complications (bacterial, viral, or fungal) and 64 samples were excluded thereafter because of acute rejection at time of BAL. The remaining 169 BALF samples were included for analysis. The BALF samples were distributed over four time periods (< 75, 76-200, 201-400, and 401-800 d after LTX) and cross-sectionally analyzed for cellular characteristics (Table 2). The BALF samples included in the first time period after LTX (< 75 d) were also longitudinally analyzed (Table 3). The longitudinal samples were distributed over three time periods (< 75, 76-400, and 401-800 d after LTX). Of these 169 BALF samples, all alveolar fractions (AF) and 139 bronchial fractions (BF) were adequate to analyze. In 30 BF samples, no recovery of the instilled fluid was obtained.

Patient Characteristics

The 169 BALF samples were obtained from 63 patients (32 female, 31 male) with a mean age of 43.4 yr (SD ± 10.4 yr) at time of transplantation. Ten of these patients received single LTX, 1 heart-lung TX, and 52 bilateral LTX. Six patients had lung fibrosis as pretransplant diagnosis, 21 ₁-antitrypsin (₁-AT) deficiency (all without augmentation treatment), 11 chronic obstructive pulmonary disease (COPD), 5 bronchiectasis, 5 primary pulmonary hypertension, 5 secondary pulmonary hypertension, and 10 cystic fibrosis pre-LTX.

Recovery

Recovery of both the bronchial and alveolar fraction percentages was comparable between all time periods, although the percentage recovered fluid was much larger in the alveolar fraction: 66% (2-87) when compared with the bronchial fraction: 25% (1-100) (Table 2).

Cellular Content and Differentiation

In the longitudinal data analysis, total cell count (AF) decreased from 234 × 10³ (70-610 × 10³) cells/ml in the first BAL within 75 d after LTX to 120 × 10³ (30-470 × 10³) cells/ml during the second year posttransplantation (p < 0.001) (Figure 1). Total cell counts had the same profile in the cross-sectional data: a decreasing total cell count from 234 × 10³ (70- 610 × 10³) cells/ml (< 75 days) to 103 × 10³ (10-840 × 10³) cells/ml during the second year post-LTX (Figure 2).

View larger version (32K): [in this window] [in a new window]   Figure 1.   Longitudinal profile of the total cell count (alveolar fraction) of 19 patients with good outcome after LTX.

View larger version (22K): [in this window] [in a new window]   Figure 2.   Cross-sectional course of the total cell count of the alveolar fraction of 63 patients with good outcome after LTX.

Overall, cell differentiation in the alveolar fraction was characterized by a predominance of alveolar macrophages: 94% (70-100) and low numbers of lymphocytes: 2% (0-24) and neutrophils: 3% (0-28). In the longitudinal analysis (Table 3) as well as in the cross-sectional data (Table 2), a slight decrease in lymphocyte percentage was shown: 5% (0-23) < 75 d to 1.5% (1) 401-800 d post-LTX (p < 0.05). No difference was found in the cell differentiation of the sequential BF (Table 2).

Lymphocyte Subtype Analysis (AF Only)

CD3⁺ A slight, but significant decrease in total CD3⁺ cell count and percentage CD3⁺ lymphocytes was noticed in both the longitudinal (Table 3) and the cross-sectional analysis (Table 4).

CD4⁺/CD8⁺ The CD4/CD8 ratio for the longitudinal follow-up increased significantly from 0.32 (0.11-0.63) just posttransplantation to 0.74 (0.21-4.27) the second year after LTX (p < 0.01) (Figure 3). These data are supported by the cross-sectional data in which a rise from 0.32 (0.11-0.46) just posttransplantation to 0.62 (0.16-4.27) the second year after LTX occurred (Figure 4).

View larger version (21K): [in this window] [in a new window]   Figure 3.   Longitudinal profile of the CD4/CD8 ratio (alveolar fraction).

View larger version (22K): [in this window] [in a new window]   Figure 4.   Cross-sectional course of the CD4/CD8 ratio (alveolar fraction).

View larger version (34K): [in this window] [in a new window]   Figure 5.   Cross-sectional absolute (A, C ) and percentage (B, D) cell counts for CD4+ and CD8+ cells showing the much more pronounced decrease in CD8+ cells after LTX.

Activation markers Both CD4⁺HLADR and CD8⁺HLADR expression decreased in the first months after LTX and remained stable after the first 75 d (Table 4). CD4⁺CD45Ro remained unaffected after LTX and CD8⁺CD45Ro showed a slight decrease (Table 4). Furthermore, CD4⁺CD25⁺ T cells showed a significant decrease from 46% (0-68) < 75 d after LTX to 21%, 16.5%, and 24% at the subsequent periods after LTX (p < 0.01) (Table 4).

Subgroup Analysis of the ₁-AT-Deficient Recipients

Because a large number (n = 21) of the included patients are ₁-AT deficient (a subgroup that is showing a lower incidence of OB [1]) we performed the same analysis on these patients. The results were the same as in the whole group. Total cell counts decreased from 235 × 10³ (range 320) in the first time period (< 75 d after LTX) to 100 × 10³ (620) in the last time period (400-800 d after LTX), CD3⁺ from 87% (12) to 83% (43), CD4⁺ from 21% (15) to 30% (50), CD8⁺ from 67% (17) to 37% (43), CD4⁺CD25⁺ from 49.5% (7) to 19% (41), and finally the CD4/CD8 ratio from 0.41 (0.23) to 0.80 (4.11).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study is the first to assess the normal longitudinal profile of cellular BALF characteristics after LTX. These data were determined in patients who can be considered as "normals," thus without any complications or signs or symptoms of additional airway pathology after LTX. We show that there is an important physiological posttransplantation change of the cellular BALF profile. Total cell counts, CD3⁺, CD4⁺, CD8⁺, and CD4⁺/CD25⁺ lymphocyte counts and CD4/CD8 ratio all undergo a dynamic change during the first 2 yr after LTX.

Cellular and immunological contents of the bronchial fraction of BALF have been shown to discriminate patients at risk for developing OB very early after LTX. Therefore, we analyzed both the bronchial fraction, which is assumed to represent the proximal airways, and the alveolar fraction, reflecting peripheral airways (6, 26). Cell differential counts were found to be comparable in both fractions in our "good outcome" patients (Table 2). Thus, there is no advantage of BF over AF in this respect or vice versa.

Previous studies have published data on cellular and immunological characteristics of BALF in "good outcome" LTX groups in relation to the pathogenesis and diagnosis of OB and acute rejection episodes (Table 1) (4, 5, 8, 10, 11, 16). However, our review of these studies shows that there are large differences in the total and differential cell counts between these studies, even though the "good outcome" groups were always defined as stable, without infection and rejection. Possible explanations for these discrepant findings may be found in the difference in timing in performing the BAL procedure after LTX, as well as different immunosuppressive regimens and laboratory protocols for BALF analysis used (Table 1). It has been suggested that the total cell count remains high for years. A high total cell count has indeed been described by several authors the first weeks post-LTX only, as has been confirmed by our results (9, 12). Our study now shows that when a "good outcome" group is defined, there is a definite decline in total cell count 1 yr post-LTX.

We analyzed 169 BALF samples (Figure 1) from 63 patients after LTX in whom no signs of OB occurred within 2 yr after the last BAL and no signs of acute rejection; acute bronchitis; or bacterial, fungal, or viral infections were present at the time of the BAL procedure. Thus, without reasonable doubt, we selected and investigated the BALF samples that can provide as clearly as possible insight in the normal course of alveolar cell influx in LTX patients. We therefore propose using these data for further evaluation of BALF cellular profiles in individuals post-LTX.

One may wonder whether the BALF cellular characteristics in "good outcome" patients after LTX are in the normal range, thus comparable to BALF cellular characteristics in healthy nonsmoking subjects. When comparing the BALF cellular profile in healthy nonsmoking subjects previously studied in our laboratory with our "good outcome" patients 401-800 d after LTX, it appears that cell counts in this LTX group are reasonably within the normal range: total cell counts LTX: median 103 × 10³/ml (range 10-840) versus healthy 100-150 × 10³/ml; neutrophil counts LTX: 1% (0-24) versus healthy 0-2%. The percentage alveolar macrophages seems to be higher in our subjects (85-92% in healthy subjects) and the lymphocyte percentage is lower compared with normal subjects (7-12% in healthy subjects) (27, 28, 29). Marked differences are found in the lymphocyte subtypes in BALF between our "good outcome" LTX subjects (401-800 d) when compared with reference values for nontransplanted, nonsmoking, healthy subjects. First, the percentage T cells (CD3⁺) is lower in the healthy subjects (50-73%) when compared with our LTX population (median 83%, range 51-99). Furthermore, there is a much lower CD4/ CD8 ratio in the LTX group: 0.62 (0.16-4.27) in our LTX group the second year after transplantation versus 1.5-2.0 in normal subjects. Two years after LTX, the total CD4⁺ count of the "good outcome" LTX patients reaches the normal range for healthy persons, whereas the total CD8⁺ count remains high above the normal values: 1.55 × 10⁴ cells/ml (LTX) versus 0.5 × 10⁴ cells/ml (healthy persons) (28, 29). This high percentage CD8⁺ may indicate ongoing allograft response. Our findings are also supported by Farver and coworkers who recently showed a high percentage of CD8⁺ T cells in subjects with LTX rejection and nonrejection when compared with nonsmoking healthy subjects. The corresponding percentage CD4⁺ cells was lower for the LTX patients when compared with the healthy subjects (30).

A longitudinal change in the BALF cellular profile is to be expected after LTX because of the abrupt start of allograft response, with responding humoral and cellular host defense, the initiated aggressive immunosuppressive therapy, and the knowledge that approximately 3 mo after LTX the donor cells are replaced by recipient cells (9). A longitudinal course has also been suggested by two earlier studies performing serial BALF analysis after LTX in unselected patients (12, 13). Defining the normal course in BALF cellular characteristics after LTX in a longitudinal set-up has not been determined so far in a properly selected group of patients post-LTX without any other pulmonary complications. Our results show a constantly increasing CD4/CD8 ratio over the first 2 yr post-LTX. This increasing CD4/CD8 ratio is due to a mild decline in total CD4⁺ cell count and a much more pronounced decline of the total CD8⁺ cell count, thus providing a rising percentage CD4⁺ and decreasing percentage CD8⁺ lymphocytes. These results are not compatible with current available literature on sequentially performed analysis of CD4/CD8 ratios the first year after LTX. The three published studies (with 14 patients evaluated in total) suggest a post-LTX decrease in CD4/CD8 ratio (13, 14, 18). In contrast, in our well-defined population, we observed an increasing percentage in CD4⁺ cells resulting in a rising CD4/CD8 ratio. This observation is supported by the study of Crim and coworkers, who also noticed a rise in the percentage CD4⁺ cells and CD4/ CD8 ratio after LTX in 10 "good outcome" patients (15).

The dynamic change in, for example, CD8⁺ counts after LTX has important bearings on the outcome of previously performed studies using flow cytometric data, as is illustrated by the studies performed by Ward and Snell (10, 31). They found no difference in flow cytometric data between OB and stable LTX patients, yet with a large difference in the time post-LTX at which the BAL procedures were performed: 855 ± 212 d and 340 ± 323 d, respectively, that is, in our study CD8⁺ counts declined for this difference in time from 4.71 × 10⁴ cells/ml to 1.55 × 10⁴ cells/ml. If Ward and Snell had used "time post-LTX" matched groups, they probably would have found a difference in the flow cytometric data between OB and their stable LTX patients (10, 31).

Our findings of a dynamic change in CD4⁺ and CD8⁺ cell count and thus the CD4/CD8 ratio after LTX can be explained by the direct and ongoing allograft response and the immunosuppressive treatment given. The direct allograft response ensures a high total CD8⁺ cell count the first weeks after transplantation, and this decreases due to the strong immunosuppressive blockage of cyclosporin or tacrolimus on the CD4⁺ (Th1) cells. The graft recipient CD4⁺ cells are directly activated due to direct recognition of cell surface markers of donor APCs. The CD4⁺ cells are preferentially activated into Th1 cells, which induce a delayed type hypersensitivity reaction resulting in acute allograft rejection (32, 33).

We also investigated lymphocyte activation markers. We found a high percentage of CD4⁺CD25⁺ lymphocytes ( chain of the interleukin-2 receptor) just post-LTX with a decreasing percentage over the following 2 yr, both resembling the results found by Crim and coworkers (15). The high percentage just post-LTX indicates very aggressive reactivity to donor cells in the early postoperative phase. The remaining higher level of CD4⁺CD25⁺ cells when compared with healthy non-LTX subjects probably reflects the ongoing allograft response (10, 15, 32). The number of CD4⁺HLADR and CD8⁺HLADR cells started at a highly activated level, thereafter decreasing and leveling off during the first year after LTX to remain stable. Our findings 2 yr after LTX are comparable to HLA-DR findings of previously reported flow cytometric data (10). When comparing these data with HLA-DR BALF values of healthy non-LTX volunteers, the HLA-DR expression remains at a much higher level for LTX patients, most likely due to the chronic allograft response (32).

From our study we conclude that a decrease in total cell count, lymphocytes, and CD3⁺ lymphocytes and an increase in the CD4/CD8 ratio after LTX represent the natural course of cells in BALF in patients without pathological airway conditions after LTX. We were able to define control values for the BALF cellular profile in patients with no complications after LTX. We suggest using these control values as a tool for diagnosing patients with pulmonary complications after LTX and for the follow-up treatment regimens. It appears that dynamic post-LTX profiles have to be taken into account when research is performed or clinical interpretations are made on BALF cellular characteristics after LTX.