- Email: [email protected]
Bronchiolitis Obliterans Syndrome After Single-Lung Transplantation
Bronchiolitis Obliterans Syndrome After Single-Lung Transplantation* Impact of Time to Onset on Functional Pattern and Survival Olivier Brugie`re, MD; Fabienne Pessione, MD; Gabriel Thabut, MD; Herve´ Mal, MD; Gilles Jebrak, MD; Guy Lese`che, MD; and Michel Fournier, MD
Introduction: Among risk factors for the progression of bronchiolitis obliterans syndrome (BOS) after lung transplantation (LT), the influence of time to BOS onset is not known. The aim of the study was to assess if BOS occurring earlier after LT is associated with worse functional prognosis and worse graft survival. Method: We retrospectively compared functional outcome and survival of all single-LT (SLT) recipients who had BOS develop during follow-up in our center according to time to onset of BOS (< 3 years or > 3 years after transplantation). Results: Among the 29 SLT recipients with BOS identified during the study period, 20 patients had early-onset BOS and 9 patients had late-onset BOS. The mean decline of FEV1 over time during the first 9 months in patients with early-onset BOS was significantly greater than in patients with of late-onset BOS (p ⴝ 0.04). At last follow-up, patients with early-onset BOS had a lower mean FEV1 value (25% vs 39% of predicted, p ⴝ 0.004), a lower mean PaO2 value (54 mm Hg vs 73 mm Hg, p ⴝ 0.0005), a lower 6-min walk test distance (241 m vs 414 m, p ⴝ 0.001), a higher Medical Research Council index value (3.6 vs 1.6, p ⴝ 0.0001), and a higher percentage of oxygen dependency (90% vs 11%, p ⴝ 0.001) compared with patients with late-onset BOS. In addition, graft survival of patients with early-onset BOS was significantly lower than that of patients with late-onset BOS (log-rank test, p ⴝ 0.04). There were 18 of 20 graft failures (90%) in the early-onset BOS group, directly attributable to BOS in all cases (deaths [n ⴝ 10] or retransplantation [n ⴝ 8]). In the late-onset BOS group, graft failure occurred in four of nine patients due to death from extrapulmonary causes in three of four cases. The median duration of follow-up after occurrence of BOS was not statistically different between patients with early-onset BOS and patients with late-onset BOS (31 ⴞ 28 months and 37 ⴞ 26 months, respectively; p ⴝ not significant). Conclusion: The subgroup of patients who had BOS develop > 3 years after SLT are less likely to have worrisome functional impairment develop in long-term follow-up. Considering the balance between the advantages and risks, enhancement of immunosuppression should be regarded with more caution in this subgroup than in patients with early-onset BOS. (CHEST 2002; 121:1883–1889) Key words: bronchiolitis obliterans syndrome; lung transplantation; risk factors Abbreviations: ATG ⫽ antithymocyte globulin; BOS ⫽ bronchiolitis obliterans syndrome; LT ⫽ lung transplantation; MMF ⫽ mycophenolate mofetil; MRC ⫽ Medical Research Council; NS ⫽ not significant; SLT ⫽ single-lung transplantation
transplantation (LT) is now recognized as a L ung reasonable treatment option for patients with end-stage pulmonary disease. However, long-term *From the Service de Pneumologie et Re´animation Respiratoire (Drs. Brugie`re, Thabut, Mal, Jebrak, and Fournier), Departement de Biostatistiques (Dr. Pessione), and Service de Chirurgie Thoracique (Dr. Lese`che), Hoˆpital Beaujon, Clichy, France. Manuscript received May 15, 2001; revision accepted November 14, 2001. www.chestjournal.org
survival and lung function remain limited by the development of bronchiolitis obliterans syndrome (BOS), which has been defined as a clinical syndrome of irreversible, progressive airway obstruction in the pulmonary allograft caused by the presence of Correspondence to: Olivier Brugie`re, MD, Service de Pneumologie et Re´animation Respiratoire, Hoˆpital Beaujon, 100 bd du Ge´n. Leclerc, 92000 Clichy, France CHEST / 121 / 6 / JUNE, 2002
1883
obliterative bronchiolitis.1 BOS occurs in at least 50% of lung transplant recipients surviving at least 3 months.2 The incidence of BOS is highest during the first 2 years after transplantation, but patients remain at risk indefinitely.3 BOS is not only associated with a worse survival, but may also lead to morbidity through infections due to the need to enhance immunosuppression in an attempt to halt the progression of BOS. The influence of time of BOS onset after LT on the natural history of BOS is not well-known. Our clinical impression is that BOS occurring earlier after transplantation is associated with a worse functional prognosis and worse graft survival. To test this hypothesis, we retrospectively examined the outcome of all lung transplant recipients who had BOS develop during follow-up at our center. According to the time of onset of BOS, all patients with BOS were classified into early-onset or late-onset BOS groups. BOS patterns over time, functional status at last follow-up, graft survival, and causes of deaths were compared between the two groups and are reported.
Materials and Methods Between March 1988 and May 1997, 96 LTs were performed in our center. There were 85 single-LTs (SLTs) and 11 doubleLTs. Only SLT recipients were analyzed in order to homogenize functional data, since the policy in our center is to perform SLT in all cases of nonsuppurative diseases, and since the latter represented the vast majority of our transplant recipients during the study period (89%). The 1-year survival of this cohort of 85 SLT recipients was 69.4%. Among them, 32 patients were excluded from analysis of BOS occurrence because of deaths occurring within 3 months posttransplantation (n ⫽ 23), severe primary graft failure (n ⫽ 8), or intractable airways complications (n ⫽ 1). From the cohort of the 53 remaining patients, BOS developed in 29 patients at last follow-up in September 2000. These 29 patients form the basis of this study. The standard maintenance immunosuppressive regimen consisted of cyclosporine, azathioprine, and prednisone. Cyclosporine blood levels were kept between 200 ng/mL and 300 ng/mL within the first 3 months, and thereafter according to time after transplantation and to renal function. The usual azathioprine dose was 2 mg/kg/d. Corticosteroid therapy was instituted on the first postoperative days; methylprednisolone was administered at a dose of 500 mg preoperatively and on the first postoperative day, 30 mg/d the following days, and thereafter was replaced by oral prednisone, 0,5 mg/kg/d, which was tapered progressively to 0.2 mg/kg after 12 weeks. Between March 1988 and March 1993, induction with antithymocyte globulin (ATG) therapy was used. ATG was administered for 14 days postoperatively to the first six patients, and then for 3 days in the remaining patients. In addition, some patients with BOS were converted from cyclosporine to FK506 and/or from azathioprine to mycophenolate mofetil (MMF) as part of their maintenance immunosuppression after January 1997. All transplant recipients were routinely followed up at our center at least every 3 months. At each visit, spirometry, blood gas analysis, chest radiography, and 6-min walk tests were performed. In addition, all patients were required to monitor daily their 1884
FEV1 value with a home spirometer. Follow-up was complete until time of death or September 30, 2000, for all patients. Median ⫾ SEM duration of follow-up was 68 ⫾ 22 months (range, 36 to 108 months). Bronchoscopy with BAL and transbronchial biopsies was performed only for clinical, physiologic, or radiographic changes. An episode of acute rejection was defined either after histologic confirmation by transbronchial biopsies (grade A1 or higher) or when clinical symptoms, radiologic signs, and blood gas abnormalities improved after enhanced immunosuppression. Acute rejection episodes were diagnosed by histologic criteria in 61% of the cases (87 of 143 episodes). Acute rejection was treated with daily infusion of methylprednisolone (1 g IV) for 3 days, followed by prednisone, 1 mg/kg/d, tapered to maintenance dose over 3 weeks. Resistant acute rejection was treated with ATG, 2.5 mg/kg/d for 5 days. BOS was defined as an irreversible decline in FEV1 of at least 20% of the individual baseline values, and was graded according to the criteria of the international heart and lung transplantation definition.1 In all cases, other causes of declining graft function, such as acute rejection, infection, or an airway complication, were excluded by bronchoscopy and transbronchial biopsies. Baseline was determined as the maximum FEV1, taken as the average of two consecutive measurements obtained at least 30 days apart in the first year after transplantation. Irreversible decline was graded as follows: BOS stage 0, FEV1 at least 80% of baseline; BOS stage 1, FEV1 between 66% and 80% of baseline; BOS stage 2, FEV1 between 51% and 65% of baseline; and BOS stage 3, FEV1 between 0% and 50% of baseline. The onset of BOS was defined as the date of the initial decline in pulmonary function testing results corresponding to the diagnosis of BOS (ie, a ⱖ 20% fall in FEV1 with the time scale starting from the transplantation date). Patients were categorized as follows: (1) patients with early-onset BOS (occurrence of BOS within 3 years posttransplantation); or (2) patients with late-onset BOS (occurrence of BOS after 3 years posttransplantation). Functional outcome and survival after BOS onset were compared between patients with early-onset and late-onset BOS. Recording of FEV1 values for calculation of serial changes was made over the period of time from when there was the first evidence of irreversible decline to the last time of follow-up. Functional status of patients at last follow-up was assessed by BOS stage, FEV1 (percent of predicted value), Pao2 on room air, oxygen dependence, activities of daily living, and dyspnea index using the modified Medical Research Council (MRC) scale.4 The status of patients regarding activities of daily living was categorized as follows: (1) active, recipient is able to perform all activities of daily living; (2) some limitation, recipient has dyspnea on moderate exertion, some limitations in activities, but is still independent; and (3) disabled, recipient needs assistance to perform many of the activities of daily living. A dyspnea index using the modified MRC scale was used to allow patients to subjectively grade their degree of dyspnea. In this grading, dyspnea index ranges from grade 0 (no dyspnea except with strenuous exertion) to grade 4 (dyspnea at rest). Other factors likely to play a role in BOS occurrence were also compared between patients with early-onset and late-onset BOS.5,6 These factors included the “era” of transplantation categorized as early (1988 to 1991) or late (1992 to 1997), acute rejection episodes within 6 months after operation, number of resistant acute rejection episodes, cytomegalovirus serologic status, ATG for induction, and therapies for acute rejection and BOS. Regarding cytomegalovirus status, no cytomegalovirus antibody-negative patient received organs from antibody-positive donors because of our matching policy during the period of the study. Hence, cytomegalovirus status was as follows: recipient negative/donor negative; recipient positive/donor negative; and Clinical Investigations
recipient positive/donor positive. Concerning immunosuppression, the numbers of course of pulsed methylprednisolone (1 g administered IV daily for 3 days) and cytolytic therapy (rabbit antithymocyte globulin, 2.5 mg/kg/d for 5 days) before and after BOS onset were reviewed. In addition, the percentage of patients with BOS converted from cyclosporine to FK506 and/or from azathioprine to MMF were compared between the two groups. Statistics Data are presented as median or mean ⫾ SEM. Continuous variables were compared by Mann-Whitney U test, and categorical variable by Fisher exact test. Multivariate analysis of variance with repeated measures (general linear model) using software (STATA Statistical Software, Release 6.0; STATA Corporation; College Station, TX) was performed to compare the mean values of FEV1 and the decline of FEV1 over time during the 9 first months between patients with early-onset BOS and patients with late-onset BOS (test of Wilks). Individual rates of FEV1 decline were also estimated for each subject through least-squares regression of FEV1 (expressed as percentage of predicted value) vs time (during the first 9 months) after BOS onset. A correlation between individual rates of FEV1 decline and time to BOS onset was searched using the Pearson product-moment correlation coefficient. Graft survival was determined using the Kaplan-Meier method, and comparisons between patients with early-onset BOS and those with late-onset BOS were made using the log-rank test. Patient death or graft removal in the case of retransplantation were considered as graft failure. A two-sided p value ⬍ 0.05 was considered statistically significant.
Results Characteristics of Study Cohort The 29 patients with BOS were aged from 32 to 66 years at the time of transplantation (mean age, 51 years). There were 21 male and 8 female patients. Indications for transplantation were emphysema (n ⫽ 10), ␣1-antitrypsin deficiency (n ⫽ 7), idiopathic pulmonary fibrosis (n ⫽ 3), retransplantation for bronchiolitis obliterans (n ⫽ 4), histiocytosis X (n ⫽ 3), and other (n ⫽ 2). Of these 29 SLT patients, 20 patients had earlyonset BOS (median time to onset after transplantation of 12 ⫾ 6 months) and 9 patients had late-onset BOS (median time to onset after transplantation of 54 ⫾ 17 months). The median duration of follow-up after occurrence of BOS was not statistically different between patients with early-onset BOS and patients with late-onset BOS (31 ⫾ 28 months vs 37 ⫾ 26 months, respectively; p ⫽ not significant [NS]).
Figure 1. Evolution of FEV1 expressed as percentage of predicted values for recipients in early-onset and late-onset BOS groups during posttransplantation follow-up after BOS onset. All data are shown as mean ⫾ SEM *Significant vs late-onset BOS group value at corresponding time. pred ⫽ predicted.
was characterized by an initial sharp decline in FEV1 followed by a slower decrease after the first 9 months. Conversely, patients with late-onset BOS displayed a more insidious course with continuous decline of FEV1 before stabilization at a higher level compared to that of early-onset BOS. The mean decline of FEV1 over time during the first 9 months in patients with early-onset BOS was significantly greater than that of patients with late-onset BOS (Fig 1; p ⫽ 0.04, test of Wilks). Likewise, an inverse correlation between the rate of FEV1 decline during the first 9 months (percentage of predicted FEV1 per month) and time to BOS onset in the 29 patients was observed (Fig 2; r ⫽ ⫺ 0.4; p ⫽ 0.04). In addition, mean FEV1 values after BOS onset were significantly lower in patients with early-onset BOS
BOS Patterns: Progression and Severity According to Time to Onset We observed two distinct patterns of decline in FEV1 in patients with BOS according to time to BOS onset after transplantation (Fig 1). Early-onset BOS www.chestjournal.org
Figure 2. For all patients, the individual rate of FEV1 decline was plotted against time to BOS onset. A significant correlation was found between the two parameters (r ⫽ ⫺ 0.4; p ⫽ 0.04); see Figure 2 legend for expansion of abbreviation. CHEST / 121 / 6 / JUNE, 2002
1885
than in patients with late-onset BOS at each time point during follow-up from 6 months up to ⬎ 2 years after BOS onset (Fig 1). The functional data of the 29 patients at last follow-up according to date of BOS onset are shown in Table 1. Patients with early-onset BOS had higher mean BOS stage than those with late-onset BOS (mean BOS stage, 2.7 ⫾ 0.4 and 1.5 ⫾ 0.7, respectively; p ⫽ 0.0001). In patients with early-onset BOS, all patients were in BOS stage 3 (n ⫽ 15) or in BOS stage 2 (n ⫽ 5). In the late-onset BOS group, the vast majority of patients remained in BOS stage 1 (n ⫽ 5) or BOS stage 2 (n ⫽ 3), except one patient in BOS stage 3. Mean FEV1 value at last follow-up was significantly lower in the early-onset BOS group than that of late-onset BOS group (25 ⫾ 11% predicted and 39 ⫾ 12% predicted, respectively; p ⫽ 0.004). Mean Pao2 on room air in the early-onset BOS group was also lower than that in late-onset BOS group (54 ⫾ 12 mm Hg and 73 ⫾ 11 mm Hg, respectively; p ⫽ 0.0005). The majority of patients with early-onset BOS were disabled (17 of 20 patients), with oxygen dependency (18 of 20 patients), and had an MRC dyspnea score ⱖ 3 (15 of 20 patients). On the contrary, almost all patients with late-onset BOS remained active, with a MRC dyspnea score ⱕ 2 and without oxygen dependency (eight of nine patients). No statistically significant difference for other possible confounding variables (Table 2) as a risk factor in BOS severity or progression was found between the two groups. Since we observed a trend toward a lower mean number of acute rejection episodes in patients with late-onset BOS, we also compared the mean FEV1 values after BOS onset between the two groups through multivariate analysis of variance with repeated
measures adjusted for the number of rejections. Again, mean FEV1 values after BOS onset were significantly lower in patients with early-onset BOS than in patients with late-onset BOS at each time point during follow-up from 6 months until ⬎ 2 years after BOS onset (data not shown). Actuarial Survival and Causes of Graft Loss Actuarial graft survival curves after occurrence of the BOS in the 29 recipients with early-onset BOS or late-onset BOS are illustrated in Figure 3. Graft survival of patients with early-onset BOS was significantly lower than that of patients with late-onset BOS (Fig 3; log-rank test, p ⫽ 0.04). Causes of graft failure in the two groups are detailed in Table 3. Among the 20 patients with early-onset BOS, graft failure occurred in 18 patients, resulting in death (n ⫽ 10) or retransplantation (n ⫽ 8). In this group, all causes of death were progressive respiratory failures due to severe BOS with superimposed lung infection (n ⫽ 10) and all cases of retransplantation were for BOS stage 3 (n ⫽ 8). Among the nine patients with late-onset BOS, graft failure due to death occurred in four patients. In these cases, deaths were attributed to extrapulmonary causes (n ⫽ 3) or infection of native lung (n ⫽ 1). Discussion Early-onset and late-onset BOS had distinct course and severity in this series of SLT recipients. A worse functional and vital prognosis was identified in patients with early-onset BOS even though durations of follow-up were similar in the two groups. In patients with early-onset BOS, a more rapid decline
Table 1—Functional Status of Patients With BOS at Last Follow-up According to Time of BOS Onset BOS Status Duration of follow-up before BOS onset, mo Duration of follow-up after BOS onset, mo BOS stage FEV1, mL % predicted Pao2 on room air, mm Hg 6-min walk test distance, m Activity Active Severe limitation Disabled Dyspnea score‡ Oxygen dependency, No. (%)
Early-Onset BOS Group (n ⫽ 20)
Late-Onset BOS Group (n ⫽ 9)
p Value
12 ⫾ 6 31 ⫾ 28 2.7 ⫾ 0.4 725 ⫾ 350 25 ⫾ 11 54 ⫾ 12 241 ⫾ 89
54 ⫾ 17 37 ⫾ 26 1.5 ⫾ 0.7 1,176 ⫾ 312 39 ⫾ 12 73 ⫾ 11 414 ⫾ 106
0.0001† NS 0.0001† 0.002† 0.004† 0.0005† 0.001†
0 3 17 3.6 ⫾ 0.7 18 (90)
8 0 1 1.6 ⫾ 1 1 (11)
0.0001†
0.0001† 0.001†
*Data are presented as mean ⫾ SD or No. unless otherwise indicated. †Statistically significant (p ⬍ 0.05). ‡Modified MRC scale. 1886
Clinical Investigations
Table 2—Comparison of Potential Risk Factors for Progression and Severity of BOS Between the Early-Onset BOS Group and the Late-Onset BOS Group* Variables
Early-Onset BOS Group (n ⫽ 20)
Late-Onset BOS Group (n ⫽ 9)
50 50
66 33
NS
5 (25) 8 (40) 7 (35) 14 (70) 3.1 ⫾ 1.1 4 (20) 56 ⫾ 16
4 (44) 1 (11) 4 (44) 8 (88.9) 2.2 ⫾ 1.09 2 (22) 58 ⫾ 14
NS
NS 0.06 NS NS
5.35 ⫾ 2.5 1.3 ⫾ 1.0 4 (20)
4 ⫾ 1.8 1.3 ⫾ 0.7 3 (33)
NS NS NS
Period, % 1988–1991 1992–1997 Cytomegalovirus status Recipient-negative/donor-negative Recipient-positive/donor-negative Recipient-positive/donor-positive Patients with ATG for induction Acute rejection episodes within the first 6 mo Patients with resistant acute rejection FEV1 at BOS onset, % of predicted Treatment of acute rejection and/or BOS Pulse steroids (No.) Cytolytic agents (No. of courses) Conversion to FK506 and/or MMF (No. of patients)
p Value
*Data are presented as No. (%) or mean ⫾ SEM unless otherwise indicated.
of FEV1 was observed after BOS onset, which stabilized at a lower level than that of patients with late-onset BOS. At last follow-up, all functional parameters (mean FEV1 value, 6-min walk test distance, MRC index, oxygen dependency) were significantly lower in patients with early-onset BOS. In addition, almost all of these patients were disabled with oxygen dependency, whereas most patients with late-onset BOS remained active and oxygen independent. Apart from functional status, the vital prognosis also differed between the two groups. Actuarial graft survival was lower in patients with early-onset BOS. In addition, graft failure was directly attributable to BOS in the majority of cases with early-onset BOS. On the contrary, in patients with late-onset BOS, none of the graft failures due to death were directly attributable to BOS status. Several studies5,7,8 have directly or indirectly considered the impact of time to BOS onset on the
functional pattern of BOS and/or BOS severity, but the results are controversial. Studying the onset and progression of BOS in 15 SLT recipients, it has been reported that BOS may present either suddenly or insidiously.7 In accordance with our results, the onset of BOS in patients who presented insidiously appeared to be at a later stage in their postoperative course (mean time to onset of BOS of 30 months) than in those who manifested with sudden impairment in lung function (mean time to onset of BOS of 12 months).7 In another report8 assessing the role of cytolytic therapy in the course of airflow limitation in patients with BOS, an inverse association between the rate of decline in FEV1 and the time to BOS onset was observed. Conversely, in a study5 assessing various clinical factors as potential predictors of progression and survival after the onset of BOS, no
Table 3—Cause of Graft Failures due to Death or Retransplantation Among the 29 Patients With BOS Variables
Figure 3. Kaplan-Meier graft survival curves for the 29 patients. www.chestjournal.org
Early-onset BOS patients (n ⫽ 20) Deaths BOS stage 3 (bronchopneumonia) BOS stage 3 (bronchopneumonia, pulmonary aspergillosis) BOS stage 2 (infection of the native lung) Retransplantation BOS stage 3 Late-onset BOS patients (n ⫽ 9) Deaths Bronchogenic carcinoma Hemorrhagic stroke Ventricular fibrillation Sepsis from pulmonary infection (BOS stage 1)
CHEST / 121 / 6 / JUNE, 2002
No.
8 1 1 8
1 1 1 1
1887
significant effect of time to onset of BOS was found. Nevertheless, these authors concluded that the choice of a single “cutoff” at 2 years in this study may not have been appropriate. In our study, the choice of a cutoff at 3 years to classify patients into earlyonset and late-onset BOS groups was arbitrarily chosen from our clinical impression. This arbitrary choice could have introduced a bias in our results. However, when a cutoff at 2 years was chosen, the same significant differences in the decline of FEV1 over time and in functional status at last follow-up were observed between the two groups (data not shown). Furthermore, in our study, the fact that an inverse correlation between the rate of FEV1 decline and time to BOS onset was seen strongly suggests that our data are valid. Other factors that have been suspected or shown to be associated with progression of BOS after its onset were analyzed in this study. None of them were significantly predominant within one group. A trend toward a greater number of acute rejection episodes during the first 6 months in the early-onset BOS group was observed. Because this factor is associated with an increased risk not only of BOS occurrence,5,9 but also of progression of airflow limitation and worse prognosis after BOS onset,5 we repeated the comparison of the decline in FEV1 between the two groups after adjustment for this factor. The more rapid decline of FEV1 in patients with early-onset BOS remained constant, suggesting an independent role of time to BOS onset in progression of BOS. These observed differences were not explained by additional immunosuppressive therapy in the late-onset BOS group that could account for a better stabilization of loss of lung function. Indeed, mean numbers of pulsed steroids or courses of cytologic agents and percentage of patients converted to FK506 and/or MMF were not superior in patients with early-onset BOS. Among the nine patients with late-onset BOS, six patients underwent SLT for emphysema. Although these six patients fulfilled standard criteria for BOS, we also looked at the aspect of the native lung at last follow-up to assess the potential role of allograft dysfunction due to extrinsic restriction.10 Among five patients, no radiologic shift toward the allograft was detected, eliminating a limitation of expiratory flow rate caused by hyperinflation of the native lung. In the last patient with emphysema, a moderate radiologic shift of the native lung toward the allograft developed over time. Nevertheless, a typical BOS stage 3 with bronchiectasis and chronic bronchial suppuration was observed in this case. We thus suspect that hyperinflation of the native lung was mainly attributable to “intrinsic obstruction”10 with reduction of compliance and volume of the graft in 1888
this recipient. Hence, chronic allograft dysfunction among the six SLT recipients with emphysema and late-onset BOS does not seem to be attributable to a confounding factor such as hyperinflation of the native lung. These observations provide useful information regarding the course and perhaps the management of patients with BOS. It seems plausible that due to the more rapid progression and severity of BOS, patients with early-onset BOS should receive more aggressive immunosuppression to gain a chance to stabilize the respiratory function. Conversely, enhancement of immunosuppression should perhaps be approached with caution in patients with late-onset BOS because of the tendency of slower progression of BOS. Enhanced immunosuppression in these cases may lead to lung infections which have been reported as an independent factor of worsened survival after BOS onset.5 In an attempt to explain these two different BOS patterns we have observed, one can hypothesize that different immunologic processes may occur in early-onset and late-onset BOS. In this way, it has been reported that a donor antigenspecific class II- directed reactivity was associated with a less aggressive clinical course than in class I antigen-directed reactivity in patients with BOS.11 These distinct immunopathogenetic events may perhaps occur separately during early-onset and late-onset BOS. Nevertheless, this hypothesis remains purely speculative. In conclusion, we identified a subgroup of patients who had BOS develop after 3 years. They seem to have a lower probability of having symptomatic functional impairment develop in longterm follow-up. Considering the balance between the advantages and risks, enhanced immunosuppression should be regarded with more caution in this category of patients with BOS than in patients with early-onset BOS.
References 1 Cooper JD, Billingham M, Egan TM, et al. A working formulation for the standardization of nomenclature and for clinical staging of chronic dysfunction in lung allografts: International Society for Heart and Lung Transplantation. J Heart Lung Transplant 1993; 12:713–716 2 Sundaresan S, Trulock EP, Mohanakumar T, et al. Prevalence and outcome of bronchiolitis obliterans syndrome after lung transplantation. Ann Thorac Surg 1995; 60:1341–1346 3 Kelly K, Hertz MI. Obliterative bronchiolitis. Clin Chest Med 1997; 18:319 –338 4 Surveillance for respiratory hazards in occupational settings: official statement of the American Thoracic Society. Am Rev Respir Dis 1982; 126:952–956 Clinical Investigations
5 Heng D, Sharples LD, McNeil K, et al. Bronchiolitis obliterans syndrome: incidence, natural history, prognosis, and risk factors. J Heart Lung Transplant 1998; 17:1255–1263 6 Bohler A, Kesten S, Weder W, et al. Bronchiolitis obliterans after lung transplantation: a review. Chest 1998; 114:1411– 1426 7 Nathan SD, Ross DJ, Belman MJ, et al. Bronchiolitis obliterans in single-lung recipients. Chest 1995; 107:967–972 8 Snell GI, Esmore DS, Williams TJ. Cytolytic therapy for the bronchiolitis obliterans syndrome complicating lung transplantation. Chest 1996; 109:874 – 878
www.chestjournal.org
9 Yousem SA, Dauber JA, Keenan R, et al. Does histologic acute rejection in lung allografts predict the development of bronchiolitis obliterans? Transplantation 1991; 52:306 –309 10 Moy ML, Loring SH, Ingenito EP, et al. Causes of allograft dysfunction after single lung transplantation for emphysema: extrinsic restriction versus intrinsic obstruction. J Heart Lung Transplant 1999; 18:986 –993 11 Reinsmoen NL, Bolman RM, Savik K, et al. Are multiple immunopathogenetic events occurring during the development of obliterative bronchiolitis and acute rejection? Transplantation 1993; 55:1040 –104
CHEST / 121 / 6 / JUNE, 2002
1889
Introduction: Among risk factors for the progression of bronchiolitis obliterans syndrome (BOS) after lung transplantation (LT), the influence of time to BOS onset is not known. The aim of the study was to assess if BOS occurring earlier after LT is associated with worse functional prognosis and worse graft survival. Method: We retrospectively compared functional outcome and survival of all single-LT (SLT) recipients who had BOS develop during follow-up in our center according to time to onset of BOS (< 3 years or > 3 years after transplantation). Results: Among the 29 SLT recipients with BOS identified during the study period, 20 patients had early-onset BOS and 9 patients had late-onset BOS. The mean decline of FEV1 over time during the first 9 months in patients with early-onset BOS was significantly greater than in patients with of late-onset BOS (p ⴝ 0.04). At last follow-up, patients with early-onset BOS had a lower mean FEV1 value (25% vs 39% of predicted, p ⴝ 0.004), a lower mean PaO2 value (54 mm Hg vs 73 mm Hg, p ⴝ 0.0005), a lower 6-min walk test distance (241 m vs 414 m, p ⴝ 0.001), a higher Medical Research Council index value (3.6 vs 1.6, p ⴝ 0.0001), and a higher percentage of oxygen dependency (90% vs 11%, p ⴝ 0.001) compared with patients with late-onset BOS. In addition, graft survival of patients with early-onset BOS was significantly lower than that of patients with late-onset BOS (log-rank test, p ⴝ 0.04). There were 18 of 20 graft failures (90%) in the early-onset BOS group, directly attributable to BOS in all cases (deaths [n ⴝ 10] or retransplantation [n ⴝ 8]). In the late-onset BOS group, graft failure occurred in four of nine patients due to death from extrapulmonary causes in three of four cases. The median duration of follow-up after occurrence of BOS was not statistically different between patients with early-onset BOS and patients with late-onset BOS (31 ⴞ 28 months and 37 ⴞ 26 months, respectively; p ⴝ not significant). Conclusion: The subgroup of patients who had BOS develop > 3 years after SLT are less likely to have worrisome functional impairment develop in long-term follow-up. Considering the balance between the advantages and risks, enhancement of immunosuppression should be regarded with more caution in this subgroup than in patients with early-onset BOS. (CHEST 2002; 121:1883–1889) Key words: bronchiolitis obliterans syndrome; lung transplantation; risk factors Abbreviations: ATG ⫽ antithymocyte globulin; BOS ⫽ bronchiolitis obliterans syndrome; LT ⫽ lung transplantation; MMF ⫽ mycophenolate mofetil; MRC ⫽ Medical Research Council; NS ⫽ not significant; SLT ⫽ single-lung transplantation
transplantation (LT) is now recognized as a L ung reasonable treatment option for patients with end-stage pulmonary disease. However, long-term *From the Service de Pneumologie et Re´animation Respiratoire (Drs. Brugie`re, Thabut, Mal, Jebrak, and Fournier), Departement de Biostatistiques (Dr. Pessione), and Service de Chirurgie Thoracique (Dr. Lese`che), Hoˆpital Beaujon, Clichy, France. Manuscript received May 15, 2001; revision accepted November 14, 2001. www.chestjournal.org
survival and lung function remain limited by the development of bronchiolitis obliterans syndrome (BOS), which has been defined as a clinical syndrome of irreversible, progressive airway obstruction in the pulmonary allograft caused by the presence of Correspondence to: Olivier Brugie`re, MD, Service de Pneumologie et Re´animation Respiratoire, Hoˆpital Beaujon, 100 bd du Ge´n. Leclerc, 92000 Clichy, France CHEST / 121 / 6 / JUNE, 2002
1883
obliterative bronchiolitis.1 BOS occurs in at least 50% of lung transplant recipients surviving at least 3 months.2 The incidence of BOS is highest during the first 2 years after transplantation, but patients remain at risk indefinitely.3 BOS is not only associated with a worse survival, but may also lead to morbidity through infections due to the need to enhance immunosuppression in an attempt to halt the progression of BOS. The influence of time of BOS onset after LT on the natural history of BOS is not well-known. Our clinical impression is that BOS occurring earlier after transplantation is associated with a worse functional prognosis and worse graft survival. To test this hypothesis, we retrospectively examined the outcome of all lung transplant recipients who had BOS develop during follow-up at our center. According to the time of onset of BOS, all patients with BOS were classified into early-onset or late-onset BOS groups. BOS patterns over time, functional status at last follow-up, graft survival, and causes of deaths were compared between the two groups and are reported.
Materials and Methods Between March 1988 and May 1997, 96 LTs were performed in our center. There were 85 single-LTs (SLTs) and 11 doubleLTs. Only SLT recipients were analyzed in order to homogenize functional data, since the policy in our center is to perform SLT in all cases of nonsuppurative diseases, and since the latter represented the vast majority of our transplant recipients during the study period (89%). The 1-year survival of this cohort of 85 SLT recipients was 69.4%. Among them, 32 patients were excluded from analysis of BOS occurrence because of deaths occurring within 3 months posttransplantation (n ⫽ 23), severe primary graft failure (n ⫽ 8), or intractable airways complications (n ⫽ 1). From the cohort of the 53 remaining patients, BOS developed in 29 patients at last follow-up in September 2000. These 29 patients form the basis of this study. The standard maintenance immunosuppressive regimen consisted of cyclosporine, azathioprine, and prednisone. Cyclosporine blood levels were kept between 200 ng/mL and 300 ng/mL within the first 3 months, and thereafter according to time after transplantation and to renal function. The usual azathioprine dose was 2 mg/kg/d. Corticosteroid therapy was instituted on the first postoperative days; methylprednisolone was administered at a dose of 500 mg preoperatively and on the first postoperative day, 30 mg/d the following days, and thereafter was replaced by oral prednisone, 0,5 mg/kg/d, which was tapered progressively to 0.2 mg/kg after 12 weeks. Between March 1988 and March 1993, induction with antithymocyte globulin (ATG) therapy was used. ATG was administered for 14 days postoperatively to the first six patients, and then for 3 days in the remaining patients. In addition, some patients with BOS were converted from cyclosporine to FK506 and/or from azathioprine to mycophenolate mofetil (MMF) as part of their maintenance immunosuppression after January 1997. All transplant recipients were routinely followed up at our center at least every 3 months. At each visit, spirometry, blood gas analysis, chest radiography, and 6-min walk tests were performed. In addition, all patients were required to monitor daily their 1884
FEV1 value with a home spirometer. Follow-up was complete until time of death or September 30, 2000, for all patients. Median ⫾ SEM duration of follow-up was 68 ⫾ 22 months (range, 36 to 108 months). Bronchoscopy with BAL and transbronchial biopsies was performed only for clinical, physiologic, or radiographic changes. An episode of acute rejection was defined either after histologic confirmation by transbronchial biopsies (grade A1 or higher) or when clinical symptoms, radiologic signs, and blood gas abnormalities improved after enhanced immunosuppression. Acute rejection episodes were diagnosed by histologic criteria in 61% of the cases (87 of 143 episodes). Acute rejection was treated with daily infusion of methylprednisolone (1 g IV) for 3 days, followed by prednisone, 1 mg/kg/d, tapered to maintenance dose over 3 weeks. Resistant acute rejection was treated with ATG, 2.5 mg/kg/d for 5 days. BOS was defined as an irreversible decline in FEV1 of at least 20% of the individual baseline values, and was graded according to the criteria of the international heart and lung transplantation definition.1 In all cases, other causes of declining graft function, such as acute rejection, infection, or an airway complication, were excluded by bronchoscopy and transbronchial biopsies. Baseline was determined as the maximum FEV1, taken as the average of two consecutive measurements obtained at least 30 days apart in the first year after transplantation. Irreversible decline was graded as follows: BOS stage 0, FEV1 at least 80% of baseline; BOS stage 1, FEV1 between 66% and 80% of baseline; BOS stage 2, FEV1 between 51% and 65% of baseline; and BOS stage 3, FEV1 between 0% and 50% of baseline. The onset of BOS was defined as the date of the initial decline in pulmonary function testing results corresponding to the diagnosis of BOS (ie, a ⱖ 20% fall in FEV1 with the time scale starting from the transplantation date). Patients were categorized as follows: (1) patients with early-onset BOS (occurrence of BOS within 3 years posttransplantation); or (2) patients with late-onset BOS (occurrence of BOS after 3 years posttransplantation). Functional outcome and survival after BOS onset were compared between patients with early-onset and late-onset BOS. Recording of FEV1 values for calculation of serial changes was made over the period of time from when there was the first evidence of irreversible decline to the last time of follow-up. Functional status of patients at last follow-up was assessed by BOS stage, FEV1 (percent of predicted value), Pao2 on room air, oxygen dependence, activities of daily living, and dyspnea index using the modified Medical Research Council (MRC) scale.4 The status of patients regarding activities of daily living was categorized as follows: (1) active, recipient is able to perform all activities of daily living; (2) some limitation, recipient has dyspnea on moderate exertion, some limitations in activities, but is still independent; and (3) disabled, recipient needs assistance to perform many of the activities of daily living. A dyspnea index using the modified MRC scale was used to allow patients to subjectively grade their degree of dyspnea. In this grading, dyspnea index ranges from grade 0 (no dyspnea except with strenuous exertion) to grade 4 (dyspnea at rest). Other factors likely to play a role in BOS occurrence were also compared between patients with early-onset and late-onset BOS.5,6 These factors included the “era” of transplantation categorized as early (1988 to 1991) or late (1992 to 1997), acute rejection episodes within 6 months after operation, number of resistant acute rejection episodes, cytomegalovirus serologic status, ATG for induction, and therapies for acute rejection and BOS. Regarding cytomegalovirus status, no cytomegalovirus antibody-negative patient received organs from antibody-positive donors because of our matching policy during the period of the study. Hence, cytomegalovirus status was as follows: recipient negative/donor negative; recipient positive/donor negative; and Clinical Investigations
recipient positive/donor positive. Concerning immunosuppression, the numbers of course of pulsed methylprednisolone (1 g administered IV daily for 3 days) and cytolytic therapy (rabbit antithymocyte globulin, 2.5 mg/kg/d for 5 days) before and after BOS onset were reviewed. In addition, the percentage of patients with BOS converted from cyclosporine to FK506 and/or from azathioprine to MMF were compared between the two groups. Statistics Data are presented as median or mean ⫾ SEM. Continuous variables were compared by Mann-Whitney U test, and categorical variable by Fisher exact test. Multivariate analysis of variance with repeated measures (general linear model) using software (STATA Statistical Software, Release 6.0; STATA Corporation; College Station, TX) was performed to compare the mean values of FEV1 and the decline of FEV1 over time during the 9 first months between patients with early-onset BOS and patients with late-onset BOS (test of Wilks). Individual rates of FEV1 decline were also estimated for each subject through least-squares regression of FEV1 (expressed as percentage of predicted value) vs time (during the first 9 months) after BOS onset. A correlation between individual rates of FEV1 decline and time to BOS onset was searched using the Pearson product-moment correlation coefficient. Graft survival was determined using the Kaplan-Meier method, and comparisons between patients with early-onset BOS and those with late-onset BOS were made using the log-rank test. Patient death or graft removal in the case of retransplantation were considered as graft failure. A two-sided p value ⬍ 0.05 was considered statistically significant.
Results Characteristics of Study Cohort The 29 patients with BOS were aged from 32 to 66 years at the time of transplantation (mean age, 51 years). There were 21 male and 8 female patients. Indications for transplantation were emphysema (n ⫽ 10), ␣1-antitrypsin deficiency (n ⫽ 7), idiopathic pulmonary fibrosis (n ⫽ 3), retransplantation for bronchiolitis obliterans (n ⫽ 4), histiocytosis X (n ⫽ 3), and other (n ⫽ 2). Of these 29 SLT patients, 20 patients had earlyonset BOS (median time to onset after transplantation of 12 ⫾ 6 months) and 9 patients had late-onset BOS (median time to onset after transplantation of 54 ⫾ 17 months). The median duration of follow-up after occurrence of BOS was not statistically different between patients with early-onset BOS and patients with late-onset BOS (31 ⫾ 28 months vs 37 ⫾ 26 months, respectively; p ⫽ not significant [NS]).
Figure 1. Evolution of FEV1 expressed as percentage of predicted values for recipients in early-onset and late-onset BOS groups during posttransplantation follow-up after BOS onset. All data are shown as mean ⫾ SEM *Significant vs late-onset BOS group value at corresponding time. pred ⫽ predicted.
was characterized by an initial sharp decline in FEV1 followed by a slower decrease after the first 9 months. Conversely, patients with late-onset BOS displayed a more insidious course with continuous decline of FEV1 before stabilization at a higher level compared to that of early-onset BOS. The mean decline of FEV1 over time during the first 9 months in patients with early-onset BOS was significantly greater than that of patients with late-onset BOS (Fig 1; p ⫽ 0.04, test of Wilks). Likewise, an inverse correlation between the rate of FEV1 decline during the first 9 months (percentage of predicted FEV1 per month) and time to BOS onset in the 29 patients was observed (Fig 2; r ⫽ ⫺ 0.4; p ⫽ 0.04). In addition, mean FEV1 values after BOS onset were significantly lower in patients with early-onset BOS
BOS Patterns: Progression and Severity According to Time to Onset We observed two distinct patterns of decline in FEV1 in patients with BOS according to time to BOS onset after transplantation (Fig 1). Early-onset BOS www.chestjournal.org
Figure 2. For all patients, the individual rate of FEV1 decline was plotted against time to BOS onset. A significant correlation was found between the two parameters (r ⫽ ⫺ 0.4; p ⫽ 0.04); see Figure 2 legend for expansion of abbreviation. CHEST / 121 / 6 / JUNE, 2002
1885
than in patients with late-onset BOS at each time point during follow-up from 6 months up to ⬎ 2 years after BOS onset (Fig 1). The functional data of the 29 patients at last follow-up according to date of BOS onset are shown in Table 1. Patients with early-onset BOS had higher mean BOS stage than those with late-onset BOS (mean BOS stage, 2.7 ⫾ 0.4 and 1.5 ⫾ 0.7, respectively; p ⫽ 0.0001). In patients with early-onset BOS, all patients were in BOS stage 3 (n ⫽ 15) or in BOS stage 2 (n ⫽ 5). In the late-onset BOS group, the vast majority of patients remained in BOS stage 1 (n ⫽ 5) or BOS stage 2 (n ⫽ 3), except one patient in BOS stage 3. Mean FEV1 value at last follow-up was significantly lower in the early-onset BOS group than that of late-onset BOS group (25 ⫾ 11% predicted and 39 ⫾ 12% predicted, respectively; p ⫽ 0.004). Mean Pao2 on room air in the early-onset BOS group was also lower than that in late-onset BOS group (54 ⫾ 12 mm Hg and 73 ⫾ 11 mm Hg, respectively; p ⫽ 0.0005). The majority of patients with early-onset BOS were disabled (17 of 20 patients), with oxygen dependency (18 of 20 patients), and had an MRC dyspnea score ⱖ 3 (15 of 20 patients). On the contrary, almost all patients with late-onset BOS remained active, with a MRC dyspnea score ⱕ 2 and without oxygen dependency (eight of nine patients). No statistically significant difference for other possible confounding variables (Table 2) as a risk factor in BOS severity or progression was found between the two groups. Since we observed a trend toward a lower mean number of acute rejection episodes in patients with late-onset BOS, we also compared the mean FEV1 values after BOS onset between the two groups through multivariate analysis of variance with repeated
measures adjusted for the number of rejections. Again, mean FEV1 values after BOS onset were significantly lower in patients with early-onset BOS than in patients with late-onset BOS at each time point during follow-up from 6 months until ⬎ 2 years after BOS onset (data not shown). Actuarial Survival and Causes of Graft Loss Actuarial graft survival curves after occurrence of the BOS in the 29 recipients with early-onset BOS or late-onset BOS are illustrated in Figure 3. Graft survival of patients with early-onset BOS was significantly lower than that of patients with late-onset BOS (Fig 3; log-rank test, p ⫽ 0.04). Causes of graft failure in the two groups are detailed in Table 3. Among the 20 patients with early-onset BOS, graft failure occurred in 18 patients, resulting in death (n ⫽ 10) or retransplantation (n ⫽ 8). In this group, all causes of death were progressive respiratory failures due to severe BOS with superimposed lung infection (n ⫽ 10) and all cases of retransplantation were for BOS stage 3 (n ⫽ 8). Among the nine patients with late-onset BOS, graft failure due to death occurred in four patients. In these cases, deaths were attributed to extrapulmonary causes (n ⫽ 3) or infection of native lung (n ⫽ 1). Discussion Early-onset and late-onset BOS had distinct course and severity in this series of SLT recipients. A worse functional and vital prognosis was identified in patients with early-onset BOS even though durations of follow-up were similar in the two groups. In patients with early-onset BOS, a more rapid decline
Table 1—Functional Status of Patients With BOS at Last Follow-up According to Time of BOS Onset BOS Status Duration of follow-up before BOS onset, mo Duration of follow-up after BOS onset, mo BOS stage FEV1, mL % predicted Pao2 on room air, mm Hg 6-min walk test distance, m Activity Active Severe limitation Disabled Dyspnea score‡ Oxygen dependency, No. (%)
Early-Onset BOS Group (n ⫽ 20)
Late-Onset BOS Group (n ⫽ 9)
p Value
12 ⫾ 6 31 ⫾ 28 2.7 ⫾ 0.4 725 ⫾ 350 25 ⫾ 11 54 ⫾ 12 241 ⫾ 89
54 ⫾ 17 37 ⫾ 26 1.5 ⫾ 0.7 1,176 ⫾ 312 39 ⫾ 12 73 ⫾ 11 414 ⫾ 106
0.0001† NS 0.0001† 0.002† 0.004† 0.0005† 0.001†
0 3 17 3.6 ⫾ 0.7 18 (90)
8 0 1 1.6 ⫾ 1 1 (11)
0.0001†
0.0001† 0.001†
*Data are presented as mean ⫾ SD or No. unless otherwise indicated. †Statistically significant (p ⬍ 0.05). ‡Modified MRC scale. 1886
Clinical Investigations
Table 2—Comparison of Potential Risk Factors for Progression and Severity of BOS Between the Early-Onset BOS Group and the Late-Onset BOS Group* Variables
Early-Onset BOS Group (n ⫽ 20)
Late-Onset BOS Group (n ⫽ 9)
50 50
66 33
NS
5 (25) 8 (40) 7 (35) 14 (70) 3.1 ⫾ 1.1 4 (20) 56 ⫾ 16
4 (44) 1 (11) 4 (44) 8 (88.9) 2.2 ⫾ 1.09 2 (22) 58 ⫾ 14
NS
NS 0.06 NS NS
5.35 ⫾ 2.5 1.3 ⫾ 1.0 4 (20)
4 ⫾ 1.8 1.3 ⫾ 0.7 3 (33)
NS NS NS
Period, % 1988–1991 1992–1997 Cytomegalovirus status Recipient-negative/donor-negative Recipient-positive/donor-negative Recipient-positive/donor-positive Patients with ATG for induction Acute rejection episodes within the first 6 mo Patients with resistant acute rejection FEV1 at BOS onset, % of predicted Treatment of acute rejection and/or BOS Pulse steroids (No.) Cytolytic agents (No. of courses) Conversion to FK506 and/or MMF (No. of patients)
p Value
*Data are presented as No. (%) or mean ⫾ SEM unless otherwise indicated.
of FEV1 was observed after BOS onset, which stabilized at a lower level than that of patients with late-onset BOS. At last follow-up, all functional parameters (mean FEV1 value, 6-min walk test distance, MRC index, oxygen dependency) were significantly lower in patients with early-onset BOS. In addition, almost all of these patients were disabled with oxygen dependency, whereas most patients with late-onset BOS remained active and oxygen independent. Apart from functional status, the vital prognosis also differed between the two groups. Actuarial graft survival was lower in patients with early-onset BOS. In addition, graft failure was directly attributable to BOS in the majority of cases with early-onset BOS. On the contrary, in patients with late-onset BOS, none of the graft failures due to death were directly attributable to BOS status. Several studies5,7,8 have directly or indirectly considered the impact of time to BOS onset on the
functional pattern of BOS and/or BOS severity, but the results are controversial. Studying the onset and progression of BOS in 15 SLT recipients, it has been reported that BOS may present either suddenly or insidiously.7 In accordance with our results, the onset of BOS in patients who presented insidiously appeared to be at a later stage in their postoperative course (mean time to onset of BOS of 30 months) than in those who manifested with sudden impairment in lung function (mean time to onset of BOS of 12 months).7 In another report8 assessing the role of cytolytic therapy in the course of airflow limitation in patients with BOS, an inverse association between the rate of decline in FEV1 and the time to BOS onset was observed. Conversely, in a study5 assessing various clinical factors as potential predictors of progression and survival after the onset of BOS, no
Table 3—Cause of Graft Failures due to Death or Retransplantation Among the 29 Patients With BOS Variables
Figure 3. Kaplan-Meier graft survival curves for the 29 patients. www.chestjournal.org
Early-onset BOS patients (n ⫽ 20) Deaths BOS stage 3 (bronchopneumonia) BOS stage 3 (bronchopneumonia, pulmonary aspergillosis) BOS stage 2 (infection of the native lung) Retransplantation BOS stage 3 Late-onset BOS patients (n ⫽ 9) Deaths Bronchogenic carcinoma Hemorrhagic stroke Ventricular fibrillation Sepsis from pulmonary infection (BOS stage 1)
CHEST / 121 / 6 / JUNE, 2002
No.
8 1 1 8
1 1 1 1
1887
significant effect of time to onset of BOS was found. Nevertheless, these authors concluded that the choice of a single “cutoff” at 2 years in this study may not have been appropriate. In our study, the choice of a cutoff at 3 years to classify patients into earlyonset and late-onset BOS groups was arbitrarily chosen from our clinical impression. This arbitrary choice could have introduced a bias in our results. However, when a cutoff at 2 years was chosen, the same significant differences in the decline of FEV1 over time and in functional status at last follow-up were observed between the two groups (data not shown). Furthermore, in our study, the fact that an inverse correlation between the rate of FEV1 decline and time to BOS onset was seen strongly suggests that our data are valid. Other factors that have been suspected or shown to be associated with progression of BOS after its onset were analyzed in this study. None of them were significantly predominant within one group. A trend toward a greater number of acute rejection episodes during the first 6 months in the early-onset BOS group was observed. Because this factor is associated with an increased risk not only of BOS occurrence,5,9 but also of progression of airflow limitation and worse prognosis after BOS onset,5 we repeated the comparison of the decline in FEV1 between the two groups after adjustment for this factor. The more rapid decline of FEV1 in patients with early-onset BOS remained constant, suggesting an independent role of time to BOS onset in progression of BOS. These observed differences were not explained by additional immunosuppressive therapy in the late-onset BOS group that could account for a better stabilization of loss of lung function. Indeed, mean numbers of pulsed steroids or courses of cytologic agents and percentage of patients converted to FK506 and/or MMF were not superior in patients with early-onset BOS. Among the nine patients with late-onset BOS, six patients underwent SLT for emphysema. Although these six patients fulfilled standard criteria for BOS, we also looked at the aspect of the native lung at last follow-up to assess the potential role of allograft dysfunction due to extrinsic restriction.10 Among five patients, no radiologic shift toward the allograft was detected, eliminating a limitation of expiratory flow rate caused by hyperinflation of the native lung. In the last patient with emphysema, a moderate radiologic shift of the native lung toward the allograft developed over time. Nevertheless, a typical BOS stage 3 with bronchiectasis and chronic bronchial suppuration was observed in this case. We thus suspect that hyperinflation of the native lung was mainly attributable to “intrinsic obstruction”10 with reduction of compliance and volume of the graft in 1888
this recipient. Hence, chronic allograft dysfunction among the six SLT recipients with emphysema and late-onset BOS does not seem to be attributable to a confounding factor such as hyperinflation of the native lung. These observations provide useful information regarding the course and perhaps the management of patients with BOS. It seems plausible that due to the more rapid progression and severity of BOS, patients with early-onset BOS should receive more aggressive immunosuppression to gain a chance to stabilize the respiratory function. Conversely, enhancement of immunosuppression should perhaps be approached with caution in patients with late-onset BOS because of the tendency of slower progression of BOS. Enhanced immunosuppression in these cases may lead to lung infections which have been reported as an independent factor of worsened survival after BOS onset.5 In an attempt to explain these two different BOS patterns we have observed, one can hypothesize that different immunologic processes may occur in early-onset and late-onset BOS. In this way, it has been reported that a donor antigenspecific class II- directed reactivity was associated with a less aggressive clinical course than in class I antigen-directed reactivity in patients with BOS.11 These distinct immunopathogenetic events may perhaps occur separately during early-onset and late-onset BOS. Nevertheless, this hypothesis remains purely speculative. In conclusion, we identified a subgroup of patients who had BOS develop after 3 years. They seem to have a lower probability of having symptomatic functional impairment develop in longterm follow-up. Considering the balance between the advantages and risks, enhanced immunosuppression should be regarded with more caution in this category of patients with BOS than in patients with early-onset BOS.
References 1 Cooper JD, Billingham M, Egan TM, et al. A working formulation for the standardization of nomenclature and for clinical staging of chronic dysfunction in lung allografts: International Society for Heart and Lung Transplantation. J Heart Lung Transplant 1993; 12:713–716 2 Sundaresan S, Trulock EP, Mohanakumar T, et al. Prevalence and outcome of bronchiolitis obliterans syndrome after lung transplantation. Ann Thorac Surg 1995; 60:1341–1346 3 Kelly K, Hertz MI. Obliterative bronchiolitis. Clin Chest Med 1997; 18:319 –338 4 Surveillance for respiratory hazards in occupational settings: official statement of the American Thoracic Society. Am Rev Respir Dis 1982; 126:952–956 Clinical Investigations
5 Heng D, Sharples LD, McNeil K, et al. Bronchiolitis obliterans syndrome: incidence, natural history, prognosis, and risk factors. J Heart Lung Transplant 1998; 17:1255–1263 6 Bohler A, Kesten S, Weder W, et al. Bronchiolitis obliterans after lung transplantation: a review. Chest 1998; 114:1411– 1426 7 Nathan SD, Ross DJ, Belman MJ, et al. Bronchiolitis obliterans in single-lung recipients. Chest 1995; 107:967–972 8 Snell GI, Esmore DS, Williams TJ. Cytolytic therapy for the bronchiolitis obliterans syndrome complicating lung transplantation. Chest 1996; 109:874 – 878
www.chestjournal.org
9 Yousem SA, Dauber JA, Keenan R, et al. Does histologic acute rejection in lung allografts predict the development of bronchiolitis obliterans? Transplantation 1991; 52:306 –309 10 Moy ML, Loring SH, Ingenito EP, et al. Causes of allograft dysfunction after single lung transplantation for emphysema: extrinsic restriction versus intrinsic obstruction. J Heart Lung Transplant 1999; 18:986 –993 11 Reinsmoen NL, Bolman RM, Savik K, et al. Are multiple immunopathogenetic events occurring during the development of obliterative bronchiolitis and acute rejection? Transplantation 1993; 55:1040 –104
CHEST / 121 / 6 / JUNE, 2002
1889