Graft survival in primary thoracic organ transplant recipients: A special report from the International Thoracic Organ Transplant Registry of the International Society for Heart and Lung Transplantation

A detailed analysis of posttransplant patient survival has been a key component of the annual reports of the International Thoracic Organ Transplant Registry. These survival analyses—published separately for adult and pediatric heart transplant (HT) and lung transplant (LT) recipients— have all analyzed patient death as the primary outcome.^1–4 Because some transplant recipients may receive a retransplant when the transplanted organ shows signs of graft failure, retransplantation has contributed to patient survival time in these analyses. The purpose of this special Registry report is to analyze the outcome of graft survival, defined by a composite outcome of time to death or retransplantation in thoracic organ transplant recipients and thus assess how long patients survive with their primary transplant. Specifically, this report has 3 objectives: (1) to provide an estimate of graft survival (freedom from death or retransplant) in adults and children with HT or LT, (2) to describe cumulative retransplantation rates after primary transplant, and (3) to describe multivariable models of 10-year graft survival following primary transplant. Because no new patients have been added to the Registry since the 2019 report, this report presents analysis of transplant recipients previously described in the 2019 and 2021 reports^1–8 and is based on data submitted to the Registry through early 2019.

Statistical methods

Data collection, conventions, and statistical methods

This special report from the International Thoracic Organ Transplant Registry was developed using data submitted to the Registry from national and multinational transplant collectives as well as individual transplant centers. Overall, 437 adult HT centers, 306 pediatric HT centers, 232 adult LT centers, and 139 pediatric LT centers contributed data to the Registry for this special report.

This report analyzes 97,140 adult HT recipients, 11,909 pediatric HT recipients, 60,680 adult LT recipients, and 2,052 pediatric LT recipients who underwent primary transplant between January 1992 and June 2017 with follow-up data in the International Society for Heart and Lung Transplantation (ISHLT) Registry through April 19, 2019. Univariate comparisons of graft survival (time to death or retransplant) were performed using Kaplan-Meier survival curves and compared across different groups using log-rank tests. A competing risk method was used to assess cumulative incidence of retransplantation in primary organ transplant recipients, and the incidence rates were compared across different groups using the Fine and Gray tests. Multivariable Cox regression models for 10-year graft survival were developed for each organ by age group. In addition to the analyses included in this report, additional analyses are presented in the online slide sets (https://ishl-tregistries.org/registries/slides.asp). The ISHLT website also contains slide sets for previous annual reports. This report references specific online e-slides when particular data are discussed but not shown due to space limitations. The ISHLT Registry website (https://ishlt.org/research-data/registries/ttx-registry#data-fields-look-up-tables-forms) provides detailed spreadsheets of the data elements collected in the Registry.

Analytical conventions

The analyses of HT or LT recipients do not include combined heart-lung transplant data. For time-to-event rates, follow-up of recipients not experiencing the event is censored at the most recent annual follow-up. Time-to-event graphs (survival graphs) are truncated when the number of subjects at risk becomes < 10. Multivariable Cox models were limited to variables in Table 1.

Table 1.

Baseline Characteristics for Primary Transplants (January 1992 to June 2017)

Baseline characteristics	Adult heart (n = 97,140)	Pediatric heart (n = 11,909)	Adult lung (n = 60,680)	Pediatric lung (n = 2,052)	p-valuea
Geographic location, n (%)					< 0.0001
Europe	39,936 (41.1)	3,192 (26.8)	21,658 (35.7)	789 (38.5)
North America	50,356 (51.8)	8,076 (67.8)	34,424 (56.7)	1,115 (54.3)
Other	6,848 (7.0)	641 (5.4)	4,598 (7.6)	148 (7.2)
Transplant era, n (%)					< 0.0001
1992–2000	35,159 (36.2)	3,376 (28.3)	11,341 (18.7)	535 (26.1)
2001–2009	31,921 (32.9)	4,158 (34.9)	20,790 (34.3)	759 (37.0)
2010–2017	30,060 (30.9)	4,375 (36.7)	28,549 (47.0)	758 (36.9)
Recipient age (years)	54.0 (26.0–67.0)	6.7 (0.1–17.0)	55.0 (24.0–68.0)	14.0 (0.0–17.0)	N/A
Recipient male (%)	77.9	55.9	56.1	42.9	< 0.0001
Recipient BMI (kg/m2)	25.7 (19.3–33.7)	16.2 (12.1–26.3)	23.8 (17.1–31.7)	16.8 (12.9–23.6)	< 0.0001
Recipient pretransplant ECMO (%)	0.5	5.9	2.0	3.8	< 0.0001
Recipient pretransplant ventilator (%)	2.6	18.8	4.4	16.2	< 0.0001
Recipient pretransplant dialysis (%)	2.4	2.6	0.5	0.5	< 0.0001
Recipient eGFR (mL/min/1.73 m2)b	75.3 (35.2–142.6)	82.6 (27.1–159.7)	95.1 (53.9–162.2)	122.1 (60.1–214.8)	< 0.0001
Donor age (years)	33.0 (16.0–56.0)	8.1 (0.2–38.7)	38.0 (16.0–62.0)	14.0 (0.0–51.0)	< 0.0001
Ischemic time (hours)	3.0 (1.5–5.0)	3.6 (1.6–5.8)	4.9 (2.4–7.6)	5.5 (3.0–8.3)	< 0.0001
Diagnosis (heart), n (%)					N/A
Congenital heart disease	2,319 (2.5)	4,603 (39.8)	N/A	N/A
Dilated cardiomyopathy	N/A	5,494 (47.4)	N/A	N/A
Hypertrophic cardiomyopathy	2,210 (2.4)	N/A	N/A	N/A
Ischemic cardiomyopathy	37,778 (40.6)	N/A	N/A	N/A
Nonischemic cardiomyopathy	44,745 (48.1)	N/A	N/A	N/A
Restrictive cardiomyopathy	2,082 (2.2)	N/A	N/A	N/A
Valvular heart disease	3,208 (3.5)	N/A	N/A	N/A
Other	674 (0.7)	1,483 (12.8)	N/A	N/A
Diagnosis (lung)					N/A
Cystic fibrosis	N/A	N/A	9,330 (15.7)	1,201 (60.0)
Chronic obstructive pulmonary disease	N/A	N/A	22,397 (37.6)	N/A
Idiopathic pulmonary fibrosis	N/A	N/A	15,033 (25.3)	N/A
Interstitial lung disease	N/A	N/A	N/A	103 (5.2)
Obliterative bronchiolitis	N/A	N/A	N/A	103 (5.2)
Pulmonary vascular disease	N/A	N/A	N/A	336 (16.8)
Other	N/A	N/A	12,736 (21.4)	258 (12.9)

Abbreviations: BMI, body mass index; ECMO, extracorporeal membrane oxygenation; eGFR, estimated glomerular filtration rate, N/A, not applicable.

Summary statistics included transplants with nonmissing data.

Continuous factors are expressed as median (5th to 95th percentiles).

^ap-value was calculated using a chi-square (for categorical variables) or Kruskal-Wallis test (for categorical variables).

^bGFR was estimated using the Cockcroft-Gault formula for adults and modified Schwartz formula for pediatrics.

Baseline characteristics

Baseline characteristics in adult and pediatric HT recipients and in adult and pediatric LT recipients are presented in Table 1 (eSlides 4–6). The most common indications for HT were nonischemic cardiomyopathy (48%) and ischemic cardiomyopathy (41%) in adults and dilated cardiomyopathy (47%) and congenital heart disease (40%) in children. The most common indications for LT were chronic obstructive pulmonary disease (38%) and idiopathic pulmonary fibrosis (25%) in adults and cystic fibrosis (60%) and pulmonary vascular disease (17%) in children. There were some interesting differences between the cohorts. Children were much more likely to need ventilator support before transplant (19% of HT and 16% of LT recipients) compared to adults (2.6% of HT and 4.4% of LT recipients). Advanced renal failure with need for dialysis was more common in HT recipients (2.4% of adults and 2.6% of children) compared to LT recipients (0.5% of adults and 0.5% of children). The distribution of ischemic time was significantly different across the 4 groups (p < 0.0001), with lower median donor ischemic time in HT recipients (3 hours in adults and 3.6 hours in children) compared to LT recipients (4.9 hours in adults and 5.5 hours in children).

Graft survival (freedom from death/retransplantation)

We first compared graft survival (freedom from death/re-transplant) in transplant recipients (1992–2017) stratified by age group and organ. In the analysis cohort, the total number of deaths/retransplants was 48,505 in adult HT recipients, 4,435 in pediatric HT, 34,256 in adult LT, and 1,253 in pediatric LT recipients. The median graft survival was 14.2 years in pediatric HT recipients, 11.3 years in adult HT recipients, 5.8 years in adult LT recipients, and 4.6 years in pediatric LT recipients (Figure 1, eSlide 8). The median graft survival has increased by transplant era in all groups of transplant recipients. The median graft survival in adult HT recipients transplanted during 2001–2009, the most recent era for which it could be computed, was 12.1 years and is trending to be higher in those transplanted during 2010–2017 (Figure 2, eSlide 9). The median graft survival in pediatric HT recipients transplanted during 2001–2009 was 15.1 years and is also projected to be higher in those transplanted during 2010–2017 (Figure 3, eSlide 10). The median graft survival in adult LT recipients transplanted during 2010–2017 was 6.3 years (Figure 4, eSlide 11) and in pediatric LT recipients transplanted during 2010–2017 was 5.5 years (Figure 5, eSlide 12). Lower graft survival in pediatric LT recipients compared to adults may be explained by the distribution of pretransplant diagnoses and lower survival in adolescent cystic fibrosis recipients compared to their adult counterparts, as previously reported.^9,10

Fig. 1.

Graft survival (freedom from death/re-transplant) in transplant recipients (January 1992-June 2017) stratified by age group and organ.

Fig. 2.

Freedom from death/re-transplant in adult HT recipients by transplant era (1992–2000, 2001–2009, 2010–2017).

Fig. 3.

Freedom from death/re-transplant in pediatric HT recipients by transplant era (1992–2000, 2001–2009, 2010–2017).

Fig. 4.

Freedom from death/re-transplant in adult LT recipients by transplant era (1992–2000, 2001–2009, 2010–2017).

Fig. 5.

Freedom from death/re-transplant in pediatric LT recipients by transplant era (1992–2000, 2001–2009, 2010–2017).

We next compared graft survival among subgroups of adult HT recipients. A comparison of graft survival in transplant recipients stratified by region revealed that those transplanted in Europe had higher early mortality but longer median graft survival compared to those transplanted in North America and in other parts of the world (11.6, 11.1, and 11.4 years, respectively; eSlide 13). The median graft survival was longer in HT recipients aged 18–39 years compared to those aged 40–59 years and recipients aged ≥60 years (13.2, 11.8, and 9.7 years, respectively), probably due to lower burden of age-related comorbidities in younger recipients or possibly due to younger donors given to younger recipients (Figure 6, eSlide 14). Survival curves in male and female adult HT recipients were identical for the first 7 years after transplant, but long-term survival was better in women compared to men (median graft survival 11.8 vs 11.1 years; eSlide 15) as previously reported.¹¹ Finally, worse renal function at the time of HT was associated with higher early mortality and worse graft survival with at least 3-year difference in survival between recipients with estimated glomerular filtration rate (eGFR) < 30 compared to ≥30 mL/min/1.73 m² (eSlide 16).

Fig. 6.

Freedom from death/re-transplant in adult HT recipients by age group (January 1992—June 2017).

Among pediatric HT recipients, those transplanted in Europe had a similar early mortality compared to those transplanted in North America and a longer median graft survival compared to children transplanted in North America and other parts of the world (16.9, 13.5, and 11.1 years, respectively) (eSlide 17). When stratified by age, the median graft survival was highest in infants aged < 1 year, intermediate in children aged 1–10 years, and lowest in children aged 11–17 years at the time of HT (18.2, 14.5, and 12.4 years, respectively; Figure 7, eSlide 18). These differences may be due to immunologic advantage in the youngest recipients but may also be due to higher prevalence of nonadherence in teenage recipients.¹² The median graft survival was longer in male vs female children (15 and 13.5 years, respectively), opposite of that observed in adult HT transplant recipients (eSlide 19). As seen in adult HT recipients, worse renal function at transplant is also associated with shorter graft survival in pediatric HT recipients (eSlide 20).

Fig. 7.

Freedom from death/re-transplant in pediatric HT recipients by age group (January 1992-June 2017).

Among adult LT recipients, a comparison of graft survival in different regions of the world showed that while early posttransplant survival was similar, the median graft survival in those who underwent LT in Europe was longer compared to recipients in North America and in other parts of the world (6.8, 5.4, and 6.2 years, respectively; eSlide 21). Similar to the trend noted in adult HT recipients, the median graft survival was significantly longer in LT recipients who were aged 18–39 years at transplant compared to those who were aged 40–59 years and recipients who were aged ≥60 years (7.0, 6.2, and 4.8 years, respectively; Figure 8, eSlide 22). The median graft survival was longer in female LT recipients compared to male LT recipients (6.2 vs 5.6 years; eSlide 23). Finally, as noted above for adult and pediatric HT recipients, renal dysfunction at the time of transplant was associated with lower graft survival (eSlide 24).

Fig. 8.

Freedom from death/re-transplant in adult LT recipients by age group (January 1992-June 2017).

Pediatric LT recipients also showed regional differences in graft survival, with higher graft survival in children transplanted in Europe compared to those who received LT in North America and in other parts of the world (median graft survival 5.5, 4.1, and 4.1 years, respectively; eSlide 25). Stratified by age, median graft survival was higher in children aged 1–10 years, compared to infants aged < 1 year and children aged 11–17 years at the time of LT (5.6, 4.8, and 4.2 years, respectively; Figure 9, eSlide 26). Children aged 1–10 years have pulmonary vascular disease as the most common indication for LT, whereas those aged 11–17 years have cystic fibrosis as the most common indication, which is associated with worse survival, as described earlier. The median graft survival was longer in male vs female children (4.9 vs 4.2 years, respectively; eSlide 27), similar to pediatric HT recipients, and opposite of that observed in adult HT and LT recipients. Finally, pediatric LT recipients with worse renal function at the time of transplant tended to have shorter median graft survival though the difference did not reach statistical significance (eSlide 28).

Fig. 9.

Freedom from death/re-transplant in pediatric LT recipients by age group (January 1992-June 2017).

Cumulative incidence of retransplant after primary transplant

We first compared cumulative incidence of retransplantation by age group and organ. The number of retransplants in the study cohort was 1,566 in adult HT recipients, 759 in pediatric HT, 2066 in adult LT, and 234 in pediatric LT recipients. Cumulative incidence of retransplantation was highest in pediatric LT recipients (Figure 10, eSlide 30). By 10 years posttransplant, 14.4% of pediatric LT recipients, 5.9% of pediatric HT recipients, 4.3% of adult LT recipients, and 1.5% of adult HT recipients had received retransplantation. By 20 years after primary transplant, 16.8% of pediatric LT recipients, 13.7% of pediatric HT recipients, 5.3% of adult LT recipients, and 2.5% of adult HT recipients had received retransplantation (Figure 10, eSlide 30). Thus, most pediatric lung retransplants occur within the first decade after primary transplant, whereas pediatric heart retransplants are evenly distributed during the first and second decades. The cumulative rate of retransplantation in pediatric HT recipients was different by region: by 20 years posttransplant, 15.7% of pediatric HT recipients in North America, 8.8% of recipients in Europe, and 10.8% of recipients in other regions of the world had received a retransplant (Figure 11A, eSlide 31). In contrast, the cumulative incidence of retransplantation in pediatric LT recipients was similar in Europe and North America (Figure 11B, eSlide 32). A much higher retransplantation rate in pediatric HT and LT recipients compared to adult HT and LT recipients is not surprising because pediatric recipients are still very young at the time of their second transplant and are prepared by providers to expect retransplantation for chronic graft dysfunction if no contraindications exist. The cumulative incidence of retransplantation among adult HT and adult LT recipients was highest among recipients aged 18–39 years at transplant, lower for those aged 40–59 years at transplant, and least for those aged ≥60 years (Figure 12, eSlides 33, 35). This suggests that a higher burden of comorbidities with age and perhaps older age itself make retransplantation less likely. In addition, the higher rates of retransplantation among the youngest adult recipients may be due to higher immune activation in younger recipients, leading to higher rates of rejection-associated graft failure. Among pediatric HT recipients, those aged 1–10 years at first HT had the highest retransplant rate, whereas among pediatric LT recipients, those aged 11–17 years had the highest retransplant rate (eSlides 34, 36). Retransplant rate in adult HT and LT recipients was significantly different by transplant era, although the absolute difference was small (eSlides 37, 39) and has not changed over time in pediatric HT and LT recipients (eSlides 38, 40).

Fig. 10.

Cumulative incidence of re-transplantation in transplant recipients (January 1992- June 2017) by age group and organ.

Fig. 11.

Cumulative incidence of re-transplantation in pediatric HT recipients (panel A) and pediatric LT recipients (panel B) by geographic location (January 1992- June 2017).

Fig. 12.

Cumulative incidence of re-transplantation in adult HT recipients (panel A) and adult LT recipients (panel B) by recipient age at primary transplant (January 1992- June 2017).

Multivariable analysis

We next performed multivariable Cox regression analyses to evaluate risk factors for 10-year graft loss (death or retransplantation) by age group and organ, considering all variables in Table 1. Dialysis was not included to avoid collinearity as the models already included GFR. Among adult HT recipients, significant categorical variables associated with higher risk of graft loss included valvar cardiomyopathy, ischemic cardiomyopathy, congenital heart disease, or restrictive cardiomyopathy (all vs nonischemic cardiomyopathy), extracorporeal membrane oxygenation (ECMO), ventilator support, and transplant in Europe or other parts of the world (vs in North America; Figure 13, eSlide 42). More recent transplant eras and patients with hypertrophic cardiomyopathy (vs nonischemic cardiomyopathy) were at lower risk. The youngest and oldest adult HT recipients had increased risk of graft loss (U-shaped relationship) (Figure 14, eSlide 44). Other continuous factors associated with higher risk of graft loss included higher BMI, renal dysfunction (Figure 14, eSlides 45–46), older donor age, and longer ischemic time (Figure 15, eSlides 47, 48).

Fig. 13.

Statistically significant categorical risk factors for 10-year graft loss (death or re-transplant) in adult HT recipients with 95% confidence intervals (January 1992-June 2017).

Fig. 14.

Statistically significant risk factors for 10-year graft loss in adult HT recipients with 95% confidence intervals: recipient age (panel A), body mass index (panel B), and estimated glomerular filtration rate (panel C) (January 1992-June 2017).

Fig. 15.

Statistically significant risk factors for 10-year graft loss in adult HT recipients with 95% confidence intervals: donor age (panel A) and ischemic time (panel B) (January 1992-June 2017).

Among pediatric HT recipients, categorical factors associated with lower risk of 10-year graft loss in the multivariable model included more recent eras of transplant, while female sex, diagnosis of congenital heart disease (vs dilated cardiomyopathy), ventilator support, and ECMO support at transplant were associated with higher risk of 10-year graft loss (Figure 16, eSlide 50). Continuous factors for higher risk of graft loss were older recipient age, lower recipient eGFR at transplant, and older donor age (Figure 17, eSlides 52–54). Adjusted for these clinical factors, children who underwent HT in Europe were at lower risk of graft loss (vs in North America), and those who underwent HT in other parts of the world were at higher risk of graft loss (Figure 16, eSlide 50). This may be explained by regional differences in transplant population and in regional approaches to retransplant surgery that results in a lower rate of retransplant in Europe (Figure 11A, eSlide 31), delaying the occurrence of the composite outcome in the current analysis.

Fig. 16.

Statistically significant categorical risk factors for 10-year graft loss (death or re-transplant) in pediatric HT recipients with 95% confidence intervals (January 1992-June 2017).

Fig. 17.

Statistically significant risk factors for 10-year graft loss (death or re-transplant) in pediatric HT recipients with 95% confidence intervals: recipient age (panel A), recipient estimated GFR (panel B), and donor age (panel C) (January 1992-June 2017).

Among adult LT recipients, factors associated with increased risk of 10-year graft loss included diagnosis of idiopathic pulmonary fibrosis or other diagnoses (vs chronic obstructive lung disease), ECMO support, and ventilator support at transplant (Figure 18, eSlide 56). More recent transplant eras, female sex, and diagnosis of cystic fibrosis (vs chronic obstructive lung disease) were associated with a lower risk of graft loss. The youngest and oldest adult LT recipients had increased risk of graft loss (U-shaped relationship; Figure 19, eSlide 58). Other continuous variables associated with increased risk of graft loss were higher BMI, renal dysfunction, and older donor age (Figure 19, eSlides 59–61). Interestingly, shorter ischemic time was associated with increased risk of graft loss (eSlide 62), as described in previous Registry reports.

Fig. 18.

Statistically significant categorical risk factors for 10-year graft loss (death or re-transplant) in adult LT recipients with 95% confidence intervals (January 1992-June 2017).

Fig. 19.

Statistically significant risk factors for 10-year graft loss in adult LT recipients with 95% confidence intervals: recipient age (panel A), recipient BMI (panel B), estimated GFR (panel C), and donor age (panel D) (January 1992-June 2017).

Finally, among pediatric LT recipients, factors associated with lower risk of 10-year graft loss in the multivariable model included more recent eras of transplant and male sex (Figure 20, eSlide 64). Older age at transplant was associated with a higher risk of graft loss (Figure 21, eSlide 66). Adjusting for these factors, pediatric LT recipients in Europe were at lower risk of graft loss than those in North America (Figure 20, eSlide 64).

Fig. 20.

Statistically significant categorical risk factors for 10-year graft loss (death or re-transplant) in pediatric LT recipients with 95% confidence intervals (January 1992-June 2017).

Fig. 21.

Significant categorical risk factors for 10-year graft loss (death or re-transplant) in pediatric adult LT recipients with 95% confidence intervals: recipient age (January 1992-June 2017).

Conclusion

In this special Registry report, we have analyzed the composite outcome of death and retransplantation following primary thoracic transplant, evaluating how long adult and pediatric HT and LT recipients survive with the primary graft. This is an important question patients and families ask when being counseled for primary transplant. The report provides readers the tools to answer this question with some nuance, incorporating the discussion about era, risk factors, and retransplantation. We found that retransplantation is performed far more frequently in pediatric HT or LT recipients (vs adult recipients). Furthermore, among adult HT and LT recipients, retransplantation is more frequent among the youngest (aged 18–39 years) candidates. The analysis also shows that retransplantation rates in transplant recipients have remained relatively stable over time, suggesting that the improvement in patient survival over time, noted across age groups and organ are due to improvements in graft survival and advances in posttransplant management. This work complements the previous ISHLT Registry report from 2014, which focused on outcomes after thoracic retransplantation.^13,14 These advances have been incremental, however, and continuing to identify clinical opportunities to improve graft longevity remains the most important challenge for transplant providers. We believe that analyses presented here will generate discussion among the thoracic transplant community for improving outcomes in these patients.