Bronchiolitis obliterans syndrome-free survival after lung transplantation: An International Society for Heart and Lung Transplantation Thoracic Transplant Registry analysis

Abstract

Background: Lung transplant (LTx) recipients have low long-term survival and a high incidence of bronchiolitis obliterans syndrome (BOS). However, few long-term, multicenter, and precise estimates of BOS-free survival (a composite outcome of death or BOS) incidence exist.

Methods: This retrospective cohort study of primary LTx recipients (1994-2011) reported to the International Society of Heart and Lung Transplantation Thoracic Transplant Registry assessed outcomes through 2012. For the composite primary outcome of BOS-free survival, we used Kaplan-Meier survival and Cox proportional hazards regression, censoring for loss to follow-up, end of study, and re-LTx. Although standard Thoracic Transplant Registry analyses censor at the last consecutive annual complete BOS status report, our analyses allowed for partially missing BOS data.

Results: Due to BOS reporting standards, 99.1% of the cohort received LTx in North America. During 79,896 person-years of follow-up, single LTx (6,599 of 15,268 [43%]) and bilateral LTx (8,699 of 15,268 [57%]) recipients had a median BOS-free survival of 3.16 years (95% confidence interval [CI], 2.99-3.30 years) and 3.58 years (95% CI, 3.53-3.72 years), respectively. Almost 90% of the single and bilateral LTx recipients developed the composite outcome within 10 years of transplantation. Standard Registry analyses "overestimated" median BOS-free survival by 0.42 years and "underestimated" the median survival after BOS by about a half-year for both single and bilateral LTx (p < 0.05).

Conclusions: Most LTx recipients die or develop BOS within 4 years, and very few remain alive and free from BOS at 10 years post-LTx. Less inclusive Thoracic Transplant Registry analytic methods tend to overestimate BOS-free survival. The Registry would benefit from improved international reporting of BOS and other chronic lung allograft dysfunction (CLAD) events.

Keywords: BOS-free survival; ISHLT Registry; bronchiolitis obliterans syndrome; chronic lung allograft dysfunction; lung transplant.

Figure 1:. Derivation of study cohort from the ISHLT Thoracic Transplant Registry

Flowchart of lung transplant recipients and application of eligibility criteria that resulted in the final study cohort that we stratified by transplant procedure type. BOS assessments “not reported” refers to sites or transplant collectives that do not report BOS information as part of their standard data reporting. *LTx= lung transplantation; Tx=transplantation; HLTx=heart-lung transplantation; BOS=bronchiolitis obliterans syndrome.

Figure 2:. BOS-free survival after lung transplantation, overall and stratified by transplant procedure type

Kaplan-Meier survival methods estimated the proportion in whom bronchiolitis obliterans syndrome (BOS) and death had not occurred at follow-up after lung transplantation (LTx), stratified by the two BOS status determination Scenarios [Scenario 1 (orange) versus Scenario 2 (green); see Table 1 for Scenario definitions] that we used to determine the presence of absence of BOS. Numbers below X-axis time points represent those at risk. Cohorts consisted of primary LTx reported from 1/1/1994 to 12/31/2011, with outcomes assessed through 12/31/2012. Analyses censored for retransplant, end of study follow-up, and loss to follow-up (i.e., used the mid-point of the reporting period last reported as alive). Figure 2A shows unadjusted BOS-free survival in the entire eligible study cohort (Single LTx and Bilateral LTx combined), while Figures 2B and 2C show estimates stratified by transplant procedure type (Single LTx, Figure 2B; Bilateral LTx, Figure 2C). Single LTx and bilateral LTx recipients had a median time to BOS or death of 3.2 for Scenario 1 to 3.3 years for Scenario 2, and 3.6 years for Scenario 1 to 4.0 years for Scenario 2 respectively. For each Scenario, bilateral LTx had greater BOS-free survival than single LTx (log rank p < 0.001).

Figure 3:. Survival after lung transplantation, stratified by transplant procedure type

Kaplan-Meier survival for single and bilateral lung transplant (LTx). Cohorts consisted of primary LTx reported from 1/1/1994 to 12/31/2011, with outcomes assessed through 12/31/2012. Analyses censored for retransplant, end of study follow-up, and loss to follow-up (i.e., used the mid-point of the reporting period last reported as alive), but not censored for BOS. Bilateral LTx (blue) had better unadjusted survival than single LTx (red), and the discrepancy became larger during follow-up (log-rank test p < 0.001; unadjusted hazard ratio for death in single LTx versus bilateral LTx 1.56 [95% CI 1.49–1.63]). Thin lines surrounding each curve represent the upper and lower 95% CI limits.

Figure 4:. Freedom from BOS after lung transplantation, overall and stratified by transplant procedure type.

Kaplan-Meier survival methods estimated the proportion in whom bronchiolitis obliterans syndrome (BOS) had not occurred at follow-up after lung transplantation (LTx), stratified by the two BOS status determination Scenarios [Scenario 1 (orange) versus Scenario 2 (green); see Table 1 for Scenario definitions] used for determining the BOS status (present or absent). Numbers below X-axis time points represent those at risk. Cohorts consisted of primary LTx reported from 1/1/1994 to 12/31/2011, with outcomes assessed through 12/31/2012. Analyses censored for death, retransplant, end of study follow-up, and loss to follow-up (i.e., used the mid-point of the reporting period last reported as alive). Figure 4A shows estimates of overall freedom from BOS in the entire eligible study cohort (Single LTx and Bilateral LTx combined), while Figures 4B and 4C show estimates stratified by transplant procedure type (Single LTx; Figure 4B; Bilateral LTx, Figure 4C). For each Scenario, bilateral LTx had greater freedom from BOS than single LTx (log rank p < 0.001). For each Scenario, single LTx recipients had a shorter median time to BOS than bilateral LTx recipients, with the median time to BOS ranging from 0.14 years (Scenario 1) to 0.78 years (Scenario 2).