Volume 24, Issue 8 , Pages 968-982, August 2005
Registry of the International Society for Heart and Lung Transplantation: Eighth Official Pediatric Report—2005
Article Outline
The eighth official pediatric report of the International Society for Heart and Lung Transplantation continues to expand the cumulative international experience in pediatric thoracic organ transplantation. The trend of decreasing early mortality continues to be evident in the current review, although we are noting decreasing incremental improvements as the early mortality improves. Late mortality continues as an ever-present threat, however. Comparing the different eras from the late 1980s to the early 2000s, the slope for late mortality seems unchanged. There appear to be differences in late mortality based on the age of the recipient, but within a specific age group there is minimal difference in mortality by era of transplant. Clearly, the challenges ahead continue to lie in the realm of access to transplantation through improved donor resources and innovative means of dealing with late graft loss, which occurs largely through the processes of rejection.
The current report evaluates the effect of some of the variations in immunosuppressive therapy currently used and whether these variations significantly influence late mortality due to chronic rejection, coronary artery vasculopathy or bronchiolitis obliterans. We have dissected further the role of induction therapy and variations on the basis of maintenance immunosuppression. As the number of long-term survivors has increased, we have increased our analysis of risk factors for 5-year survival conditional on survival to the first year. Furthermore, morbidity analysis has been extended to include patients up to 7 years post-transplant. The overall numbers of new pediatric heart and lung transplant procedures has remained constant over the last 5 years. The indications have remained fairly stable, as have the early results of transplantation. In the lung transplant portion of this report, the number and survival of living donor transplants are compared with non-living donor lung transplants for the first time.
Our analysis of late survival, by definition, reflects a prior state of the art. However, surrogates for late survival, such as early rejection, can be used to evaluate current approaches and project an impact on late survival. In concert with this approach, we have added new figures demonstrating predicted survival based on varying recipient and donor characteristics. These predictive figures graphically display how variations in conditions at the time of transplant can have a legacy in late survival. The Registry data continue to document that late survivors enjoy excellent functional status with an apparent plateau in the percent of individuals suffering from morbidity, except for the relentless but slow progression in renal disease and graft vasculopathy. All figures and tables from this report, and a more comprehensive set of Registry slides, are available at www.ishlt.org/registry/.
Statistical methods
Survival rates were calculated using the Kaplan-Meier method and compared using the log rank test. Multivariate analyses were done using Cox proportional hazards regression analysis. The results of multivariate analyses are reported as relative risks (RRs). RR significantly greater than 1 indicate an increased risk with the event. RR of less than 1.0 indicate that the event was less likely to occur when the variable was present. Based on the RRs and baseline survivals obtained by the Cox proportional hazards model, predicted survival rates were computed for specified patient/donor/transplant profiles. Non-specified variables were assumed to be at the median, a position that would have a neutral effect on outcome.
Heart transplantation
Transplant Volumes and Indications
The age distribution of pediatric heart recipients is shown in Figure 1. For transplants performed between January 1996 and June 2004, the first year of life shows the largest number of recipients, compared with any other single year. There was an increasing frequency of transplant recipients in the early teenage years. Over the last 5 years, the number of recipients remained stable at approximately 360 per year, with the breakdown in the age groups also remaining fairly constant during this period (Figure 2). For recipients <1 year of age, a congenital diagnosis was most likely to lead to transplantation (Figure 3). The percentage of recipients requiring a re-transplantation within the first year remained exceedingly low. Compared with the earlier period (1988 to 1995), the overall percentage of patients with a congenital diagnosis has decreased. However, a congenital diagnosis again seems to have accounted for an increasing percentage of recipients in recent years. The diagnosis leading to heart transplantation for recipients between 11 and 17 years of age has also been stable (www.ishlt.org/registry/). However, the number of patients requiring re-transplantation has increased from 3% to 6%.
Figure 1.
Age distribution of pediatric heart recipients (transplants: January 1996 to June 2004).
Figure 2.
Age distribution of pediatric heart recipients by year of transplant.
Figure 3.
Diagnosis in pediatric heart transplant recipients (age: <1 year).
Immunosuppression
For the recent era (January 2000 to June 2004) maintenance immunosuppression is shown for patients evaluated at 1 year and 5 years after transplant (Figure 4). Virtually all patients are on either cyclosporine or tacrolimus as a calcineurin inhibitor, with tacrolimus now becoming slightly more common than cyclosporine at 1-year follow-up. This is the first time we have seen a larger number of recipients on tacrolimus than cyclosporine. Only a small percentage of patients are on rapamycin at either 1 or 5 years after transplant. Anti-proliferative agents are used in about 75% of recipients at 1 and 5 years after transplant, but the current experience shows that mycophenolate mofetil (MMF) is now more commonly used than azathioprine at both 1 and 5 years. Finally, prednisone as a component of long-term immunosuppression has undergone decreased utilization. Approximately 60% of patients are on prednisone at 1 year and only 40% at 5-year follow-up.
Figure 4.
Pediatric heart recipients’ maintenance immunosuppression at time of follow-up. Note: Different patients analyzed at Year 1 and Year 5 (follow-up: January 2000 to June 2004).
Just under 50% of all pediatric recipients between January 2001 and June 2004 received induction immunosuppression (Figure 5). The polyclonal preparations were most common, followed by interleukin-2 receptor antibodies. OKT3 use has diminished to nearly negligible levels. The effects of induction use and maintenance immunosuppression on the incidence of rejection in the first year are displayed in Figure 6. Overall, the percent of patients treated for rejection within the first year was lower for patients receiving tacrolimus as compared with cyclosporine, regardless of induction therapy. For both cyclosporine and tacrolimus there was a trend toward decreasing rejection in patients receiving induction immunosuppression, but these differences were not statistically significant. Similarly, there was a reduction in rejection for females when tacrolimus was the primary calcineurin inhibitor. For males the differences were not significant.
Figure 5.
Pediatric heart recipients’ induction immunosuppression (follow-up: January 2001 to June 2004).
Figure 6.
Percentage of pediatric heart transplant recipients treated for rejection in the first year stratified by maintenance immunosuppression and induction (transplants: January 2000 to June 2003).
When combinations of maintenance immunosuppression were compared there was a trend for a decreasing percentage of tacrolimus-treated patients who were treated for rejection within the first year compared with those treated with cyclosporine (Figure 7). There also appeared to be a trend toward decreased rejection within the first year for patients treated with cyclosporine and MMF vs cyclosporine and azathioprine, but the results were not statistically significant. However, in patients receiving tacrolimus, the combination with MMF trended toward a non-significant increase in rejection when compared with tacrolimus plus azathioprine. For female recipients, tacrolimus plus MMF was more likely to be associated with rejection in the first year than tacrolimus plus azathioprine.
Figure 7.
Percentage of pediatric heart transplant recipients treated for rejection in the first year stratified by maintenance immunosuppression (transplants: January 2000 to June 2003).
Outcomes
SurvivalSurvival rates computed with the Kaplan-Meier method for pediatric heart transplant recipients over the entire study period from 1982 to 2003 are shown in Figure 8. The inexorable decline in survival continued throughout the 18 years of follow-up. Despite the greater early mortality seen in infant recipients, the graft half-life for the infant age group still cannot be estimated out to 14 years, as >50% of the recipients are still alive. The slope of decay for the infant survivors appears shallower than the other age groups, but clearly there is still annual attrition that affects all age groups. In Figure 9 the conditional survival (i.e., subsequent survival for recipients who survived at least 1 year after transplant) for the period from 1982 through June 2003 is shown for the different age recipients. When looking at conditional survival, both the infants (age <1 year) and children (age 1 to 10 years) show a significantly lower risk of late mortality than adolescents (11 to 17 years). The slope for the infant age group again appears less steep than for the older pediatric recipients. In fact, at 14 years after transplant, >70% of infant recipients who survived at least 1 year after transplantation are still alive. The conditional half-life for all childhood-age recipients was 17.5 years. The half-life for survival conditional on survival to 5 years was not computable for any age group (www.ishlt.org/registry/). Over 80% of infant recipients who were alive at 5 years remained alive at 14 years after transplant. Importantly, the excellent late survival shown on these conditional slides only is applicable to transplants performed during the early pediatric years.
Figure 8.
Pediatric heart transplant Kaplan-Meier survival (1/1982 to 6/2003).
Figure 9.
Pediatric heart transplant conditional Kaplan-Meier survival (1/1982 to 6/2003).
When all recipients were evaluated over the entire experience from 1982 through June 2003, the gradual improvement with era of transplant is primarily a result of improved early post-transplant survival. In contrast, the curves for late survival appear to follow a similar slope regardless of era. In fact, as reported in a previous Registry analysis,1 the most dramatic reductions in early mortality occurred in the infant age group, a trend that has continued into the most recent era. Otherwise, for adolescents in particular, there has been little ongoing improvement in early mortality in the most recent era (www.ishlt.org/registry/).
Figure 10 shows comparisons of calcineurin inhibitors, tacrolimus and cyclosporine with regard to survival. In this analysis, late survival rates were conditional on survival to 14 days to eliminate reporting biases of immunosuppression due to early death. There was no difference in long-term survival in patients treated with tacrolimus or cyclosporine as the primary calcineurin inhibitor used at discharge. Although Figure 7, Figure 8 appear to indicate a reduction in acute rejection within the first year when tacrolimus was the primary calcineurin inhibitor, this reduction in rejection incidence did not appear to translate into a reduction in mortality. These data are somewhat surprising, because this and earlier Pediatric Registry reports have shown an association between rejection within the first year and risk of late death.1, 2 We will continue to evaluate the influence of immunosuppressive drugs and drug combinations on these important parameters. Ultimately, survival could be a powerful arbiter in drug selection, but at present the data are inconclusive, at least in terms of late survival.
Figure 10.
Pediatric heart transplant Kaplan-Meier survival stratified by tacrolimus vs cyclosporine use at discharge (transplants: January 1998 to June 2003) conditional on survival to 14 days.
Risk Factors for Mortality
The risk factors for 1-year mortality from multivariate analysis for the era 1995 to June 2003 are shown in Table 1. The diagnosis leading to transplant continues to be the greatest risk factor, but depends on whether or not the patient was on extracorporeal membrane oxygenation (ECMO). Those patients with a congenital diagnosis who required ECMO clearly had the highest risk, with RR of 1-year mortality of 4.16, compared with patients not on ECMO. Patients on ECMO without a congenital diagnosis also experienced an increased risk of mortality when compared with the non-ECMO reference group, but to a lesser degree (RR = 1.9). The year of transplant and whether the patient was in a hospital and on a ventilator at the time of transplant were also significant factors for 1-year mortality.
Table 1. Pediatric Heart Transplants (1/1995 to 6/2003) Risk Factors for 1-Year Mortality
| Variable | N | Relative risk | p-value | 95% confidence interval |
|---|---|---|---|---|
| Congenital diagnosis, on ECMO | 69 | 4.16 | <0.0001 | 2.66 to 6.51 |
| Congenital diagnosis, no ECMO | 974 | 2.19 | <0.0001 | 1.74 to 2.77 |
| ECMO, diagnosis other than congenital | 68 | 1.9 | 0.0211 | 1.10 to 3.28 |
| Year of transplant: 1995 vs 1998 | 362 | 1.9 | 0.001 | 1.30 to 2.77 |
| Year of transplant: 1996 vs 1998 | 342 | 1.62 | 0.018 | 1.09 to 2.42 |
| Hospitalized (including ICU) | 2,132 | 1.55 | 0.0007 | 1.20 to 2.00 |
| Year of transplant: 1997 vs 1998 | 366 | 1.5 | 0.0477 | 1.00 to 2.23 |
| On ventilator | 448 | 1.35 | 0.0239 | 1.04 to 1.76 |
| Female recipient | 1,300 | 1.22 | 0.0409 | 1.01 to 1.48 |
The continuous factors associated with risk of 1-year mortality include donor age, weight ratio, serum creatinine and bilirubin. The relationship between donor age and RR of 1-year mortality is displayed in Figure 11 for the recent era. The relationship between risk and age is complex, with a nadir for the RR occurring for donor age in the teenage years. There may be confounding variables that explain the risk with the younger age donors, because the youngest recipients would most likely have a congenital diagnosis and would be at greater risk for 1-year mortality (Table 1). Younger donors are also more likely to be undersized relative to recipients. The donor/recipient weight ratio also shows an inverted parabolic-shaped curve with an increased risk at the greatest weight ratio and also at the lowest weight ratio (www.ishlt.org/registry/). The ideal donor weight/recipient weight ratio appears to be just over 1.5. This is significant in that young recipients, often with a congenital diagnosis, would likely be in the highest weight ratio group due to the use of older donors.
Figure 11.
Pediatric heart transplant (1/1995 to 6/2003) risk factors for 1-year mortality donor age.
Pre-transplant creatinine and bilirubin were also significantly associated with 1-year mortality (www.ishlt.org/registry/). Both of these markers can presumably identify patients with end-organ damage, due to either their congenital heart disease or cardiomyopathic disease, and these factors place them at greater risk for 1-year mortality after transplantation.
As reported in prior Registry analyses, a number of factors are not associated with 1-year mortality, including intravenous inotropes, a recent history of infection in the recipient, and even age alone. Surprisingly, pulmonary artery pressure and pulmonary vascular resistance were not associated with 1-year mortality, as observed in earlier reports. Perhaps physicians’ assessments are beginning to eliminate patients with some of the previously noted risk factors. Interestingly, panel-reactive antibody (PRA) also was not associated with a significant increase in risk for 1-year mortality. Donor issues that were unrelated to 1-year mortality include clinical infection, cause of death and gender. Interestingly, graft ischemic time was also not associated with 1-year mortality.
There was a highly significant difference between late survival, which depended on whether or not rejection occurred in the first year (Figure 12), in a univariate analysis. However, the difference in survival based on early rejection was only about 6% by 5 years after transplant. Because over half of the recipients are reported to have rejection within the first year, there appears to be a sub-set of these patients that go on to have increased risk of late poor graft outcome. We have also previously reported a negative impact on long-term survival when prednisone was continued throughout the first year. The same observations were made in the prior Registry report (www.ishlt.org/registry/).
Figure 12.
Pediatric heart transplant Kaplan-Meier survival based on rejection within the first year (transplants: January 1998 to June 2002).
Risk factors for 5-year mortality from multivariate analysis are shown in Table 2. Recipient need for dialysis or ECMO, congenital diagnosis and re-transplant status were all associated with increased risk of 5-year mortality. However, the number of patients either on dialysis or ECMO was quite low, and probably the most clinically significant risk factor for 5-year mortality is having the pre-transplant diagnosis of congenital disease. The factors not significantly associated with 5-year mortality, including a history of recent infection, elevated bilirubin, pulmonary pressures, hospitalization at the time of transplant, donor cause of death, ischemia time, HLA mismatch and transplant center volume, were not related to 5-year mortality. The continuous factors associated with 5-year mortality include patient weight and, to a lesser extent, pre-transplant creatinine (www.ishlt.org/registry/).
Table 2. Pediatric Heart Transplants (1/1995 to 6/1999) Risk Factors for 5-Year Mortality
| Variable | N | Relative risk | p-value | 95% confidence interval |
|---|---|---|---|---|
| Recipient on dialysis | 17 | 2.64 | 0.0038 | 1.37 to 5.09 |
| Congenital diagnosis, on ECMO | 24 | 2.53 | 0.0039 | 1.35 to 4.77 |
| Re-transplant | 63 | 1.81 | 0.0110 | 1.15 to 2.86 |
| Congenital diagnosis, age = 0, PGE | 124 | 1.72 | 0.0095 | 1.14 to 2.59 |
| Congenital diagnosis, no ECMO | 522 | 1.69 | <0.0001 | 1.34 to 2.13 |
| Female donor | 667 | 1.28 | 0.0158 | 1.05 to 1.55 |
| Intravenous inotropes | 633 | 1.26 | 0.0366 | 1.01 to 1.57 |
Because much of the mortality occurs within the first post-transplant year, it is worthwhile to look at risk factors for 5-year mortality conditional upon survival to the first year (Table 3). These data may identify an immunologically at-risk group for late graft failure. Requiring ECMO for non-congenital heart disease, re-transplant and congenital diagnosis are factors that continue to be associated with increased risk. However, a history of treatment for rejection carries the most significant increase in RR and affects a large number of patients. Interestingly, female recipients appear to be at somewhat higher relative risk for conditional 5-year mortality. The continuous factors associated with conditional 5-year mortality include recipient weight and ischemic time. In contrast to the risk curve for recipient weight for simple 5-year mortality, which has a complex parabolic shape, there is now a fairly linear relationship between patient weight and RR for conditional 5-year mortality (Figure 13). The relationship between recipient weight and late mortality likely reflects the data for 5-year conditional survival, indicating that the youngest, and presumably smallest, patients are at lower risk for late mortality.
Table 3. Pediatric Heart Transplants (1/1995 to 6/1999) Risk Factors for 5-Year Mortality Conditional on 1-Year Survival
| Variable | N | Relative risk | p-value | 95% confidence interval |
|---|---|---|---|---|
| ECMO, diagnosis other than congenital | 22 | 2.94 | 0.0126 | 1.26 to 6.86 |
| Re-transplant | 50 | 2.41 | 0.004 | 1.32 to 4.37 |
| Congenital diagnosis, age = 0, PGE | 93 | 2.28 | 0.0198 | 1.14 to 4.55 |
| Treated for rejection (after transplant hospitalization) | 328 | 2.19 | <0.0001 | 1.57 to 3.03 |
| Female recipient | 536 | 1.45 | 0.0282 | 1.04 to 2.01 |
Figure 13.
Pediatric heart transplant (1/1995 to 6/1999) risk factors for 5-year mortality conditional on 1-year survival recipient weight.
When early mortality is removed from consideration, ischemic time seems to be inversely related to relative risk of 5-year mortality; that is, the longer the ischemic time the lower the risk of late mortality (Figure 14). These data are somewhat surprising in some respects because ischemic time has been linked to risk of early graft failure. The reason for this apparently paradoxical finding is unclear, and we will track this observation in future pediatric reports.
Figure 14.
Pediatric heart transplant (1/1995 to 6/1999) risk factors for 5-year mortality conditional on 1-year survival ischemia time.
Issues around the time of transplant that affect early mortality, such as hospitalization, bilirubin, creatinine ventilator and pulmonary resistance, were not significant in the analysis of conditional 5-year mortality. Perhaps surprisingly, PRA was not associated with conditional 5-year mortality, but the number of patients with high levels was small. The donor factors not related to 5-year mortality include cause of death and history of infection at the time of donation. Transplant factors, including year of transplant, donor/recipient weight ratio and transplant center, were not related to late conditional mortality.
Cause of Death
The cause of death for pediatric heart transplant recipients between January 1992 and June 2004 is shown in Table 4. As seen in previous reports, early mortality was primarily related to graft failure, some cases of which may have included patients with rejection that was coded as “graft failure.” In the period between 30 days and 1 year after transplant, acute rejection dominates the cause of death, followed by infectious causes. There are still a number of graft failures that may be related to rejection. Between 1 and 3 years after transplant, acute rejection and graft failure were the dominant causes, whereas risk of infectious causes decreased. However, there now appears to be a risk due to coronary artery vasculopathy, which accounted for almost 40% of the deaths between 3 and 5 years after transplant and about 30% beyond 5 years. Nevertheless, acute rejection and graft failure continue to play a role in late mortality in all eras. The number of late graft failures may reflect a combination of both coronary vasculopathy and rejection. Lymphoma became more prominent with time and explains about 8% of the deaths at >5 years; fortunately, this number was not increased compared with previous Registry reports.
Table 4. Pediatric Heart Transplant Recipients: Cause of Death (Deaths: January 1992 to June 2004)
| Cause of death | 0 to 30 days (N = 335) | 31 days to 1 year (N = 281) | >1 year to 3 years (N = 198) | >3 years to 5 years (N = 139) | >5 years (N = 252) |
|---|---|---|---|---|---|
| Coronary artery vasculopathy | 3 (0.9%) | 26(9.3%) | 36(18.2%) | 52(37.4%) | 72(28.6%) |
| Acute rejection | 27(8.1%) | 76(27.0%) | 54(27.3%) | 18(12.9%) | 32(12.7%) |
| Lymphoma | 0 | 6(2.1%) | 10(5.1%) | 3(2.2%) | 21(8.3%) |
| Malignancy, other | 0 | 4(1.4%) | 2(1.0%) | 1(0.7%) | 9(3.6%) |
| CMV | 1(0.3%) | 7(2.5%) | 1(0.5%) | 0 | 0 |
| Infection, non-CMV | 47(14.0%) | 46(16.4%) | 17(8.6%) | 6(4.3%) | 16(6.3%) |
| Primary failure | 58(17.3%) | 11(3.9%) | 6(3.0%) | 8(5.8%) | 12(4.8%) |
| Graft failure | 79(23.6%) | 31(11.0%) | 35(17.7%) | 27(19.4%) | 49(19.4%) |
| Technical | 21(6.3%) | 2(0.7%) | 2(1.0%) | 1(0.7%) | 1(0.4%) |
| Other | 15(4.5%) | 16(5.7%) | 16(8.1%) | 14(10.1%) | 24(9.5%) |
| Multiple organ failure | 36(10.7%) | 29(10.3%) | 3(1.5%) | 2(1.4%) | 6(2.4%) |
| Renal failure | 1(0.3%) | 4(1.4%) | 1(0.5%) | 0 | 0 |
| Pulmonary | 24(7.2%) | 16(5.7%) | 9(4.5%) | 6(4.3%) | 7(2.8%) |
| Cerebrovascular | 23(6.9%) | 7(2.5%) | 6(3.0%) | 1(0.7%) | 3(1.2%) |
Specific Complications
The data on morbidity at 1 and 5 years after transplant are similar to those seen in earlier Registry reports. The incidence of hypertension increased from about 40% of patients at 1 year after transplant to 60% by 5 years. Also, we have previously shown that some of the increased risk for hypertension was associated with the use of prednisone in the maintenance immunosuppressive regimen.
We have focused our analysis this year on the cumulative prevalence of specific morbidities in survivors at 7 years after transplant (Table 5). Hypertension was present in about 67% of recipients and significant renal dysfunction increased to 11.4%, which is essentially the same as that reported in the previous analysis. Twenty-five percent of patients have hyperlipidemia, but only 4% were reported to have diabetes. Coronary artery vasculopathy was reported in 15% of patients within 7 years, which is much lower than that observed in adult heart transplant recipients.3 There was a steady decrease in freedom from coronary artery vasculopathy (CAV), with about 75% of patients CAV-free at 7 years after transplant (Figure 15).
Table 5. Post-Heart Transplant Morbidity for Pediatrics Cumulative Prevalence in Survivors Within 7 Years Post-transplant (Follow-up: April 1994 to June 2004)
| Outcome | Within 7 years | Total number with known response |
|---|---|---|
| Hypertension | 65.8% | (N = 357) |
| Renal dysfunction | ||
| Total | 11.4% | (N = 370) |
| Abnormal creatinine <2.5 mg/dl | 7.8% | |
| Creatinine >2.5 mg/dl | 2.2% | |
| Long-term dialysis | 0.5% | |
| Renal transplant | 0.8% | |
| Hyperlipidemia | 26.3% | (N = 388) |
| Diabetes | 3.9% | (N = 358) |
| Coronary artery vasculopathy | 15.2% | (N = 217) |
Figure 15.
Freedom from coronary artery vasculopathy for pediatric heart recipients (follow-up: April 1994 to June 2004).
There was no significant difference in the incidence of vasculopathy between patients who received induction therapy and those who did not (Figure 16). However, there may be differences between specific induction agents and their effects on late development of vasculopathy. These data can be analyzed further in future reports as the number of patients followed increases. Age of the recipient has been significantly related to freedom from coronary vasculopathy (Figure 17). The infant-age recipients appear to have the lowest risk of vasculopathy and the slope of the curve appears to be more gradual than seen in the older children, and certainly the adolescents. It is of interest how these curves show similar trends to those seen in overall survival based on age group.
Figure 16.
Freedom from coronary artery vasculopathy for pediatric heart recipients (follow-up: April 1994 to June 2004) stratified by induction.
Figure 17.
Freedom from coronary artery vasculopathy for pediatric heart recipients (follow-up: April 1994 to June 2004) stratified by age group.
Figure 18 depicts survival after diagnosis of CAV. Graft survival was approximately 50% at 2 years after diagnosis of CAV. There was a very sharp decrease in the graft survival, most of which occurred by 6 months to 1 year after diagnosis of CAV. These data for graft survival are notably poorer than the reported patient-survival curves in the seventh official pediatric report, apparently indicating the increasing effect of re-transplantation. There was no significant difference in graft survival as a function of age of recipient once the diagnosis of vasculopathy was established.
Figure 18.
Graft survival after diagnosis of coronary artery vasculopathy for pediatric heart recipients (follow-up: April 1994 to June 2004) stratified by age group.
Rejection during the first year has been strongly associated with a decrease in late survival. As shown in Table 6 rejection during the first year is associated significantly with a risk of vasculopathy diagnosed between 1 and 3 years after transplant. In this report we extended the analysis to also include patients reported with CAV between 3 and 5 years post-transplant. The number of patients with late follow-up was much smaller, but there was no significant difference in the incidence of vasculopathy based on a history of rejection in the first year. If these observations are borne out in subsequent analyses, they could indicate that early rejection is followed by the development of early vasculopathy. However, early rejection may not increase the risk of late-onset vasculopathy, which may be much more closely associated with late rejection occurring beyond the first year.
Table 6. Pediatric Heart Recipients Relationship of Rejection and Coronary Artery Vasculopathy (Follow-up: April 1994 to June 2004)
| Rejection During Year 1 | Reported CAV Between Years 1 and 3 Post-transplanta | Reported CAV Between Years 3 and 5 Post-transplantb | ||||
|---|---|---|---|---|---|---|
| Yes | No | All | Yes | No | All | |
| Yes | 36(8.1%) | 408(91.9%) | 444(100%) | 12(6.5%) | 174(93.5%) | 186(100%) |
| No | 13(3.0%) | 420(97.0%) | 433(100%) | 9(4.6%) | 188(95.4%) | 197(100%) |
a Difference: p = 0.001 |
b p = 0.4185. |
Renal dysfunction is a late morbidity that bears close scrutiny, because many of the immunosuppressive agents have long-term nephrotoxicity. Although some degree of renal dysfunction is common, the freedom from severe renal dysfunction remains >90% at 7 years after transplant (www.ishlt.org/registry/). As in many of the other Kaplan-Meier analyses, there has been a steady decrease in freedom from the event, but to date the rate of decay for renal dysfunction is less than for vasculopathy or other direct cardiac graft-related issues. As our follow-up is extended, the decrease in freedom from severe renal dysfunction will be monitored closely and may be a prime area to evaluate in terms of immunosuppressive therapy.
Despite the risks of morbid events, the functional status of surviving pediatric heart transplant recipients continues to be excellent at 7 years after transplant (Figure 19). As long as 7 years after transplant, >90% of the pediatric recipients have physiologic rehabilitation and continue to function without activity limitations. The frequency of significant medical interventions, as reflected by the number of hospitalizations, also seems to decrease with increasing time since transplant (www.ishlt.org/registry/). For patients with the longest follow-up, between 6 and 7 years, >75% do not require hospitalization during late follow-up. Thus, the overall quality of survival continues to appear to be quite good.
Figure 19.
Pediatric heart recipient functional status of surviving recipients (follow-up: April 1994 to June 2004).
Predicted Pediatric Heart Transplant Survival
New to the Registry report this year, we have developed projected survival curves based on hypothetical patient profiles. The intent of these curves is to graphically display how some of the identified risk factors can affect mortality in both the short and long term. As examples, Figure 20 shows the predicted survival for 2 hypothetical patients. These cases are similar in that they are infants at 3 months of age and similar weight. Case 1 is a patient with a normal creatinine level who is not hospitalized at the time of transplant; the predicted 1-year survival for this patient would be approximately 90%. Case 2 is a similarly aged youngster with congenital heart disease, but with a creatinine of 1.5 mg/dl and on ECMO and a ventilator at the time of transplant. This patient has an estimated 1-year survival of approximately 75%, a rather striking difference from Case 1. Figure 21 shows the predicted 1-year survival for Case 3, a 6-year-old, non-hospitalized recipient with a congenital diagnosis and normal creatinine, with a predicted 1-year survival of about 90%. Figure 22 shows the predicted survival curve for a hypothetical adolescent with cardiomyopathy (Case 4) compared with a similarly aged patient with a congenital diagnosis and renal dysfunction (Case 5). The figure shows the prominent impact of congenital heart disease on survival, even for adolescent-aged recipients.
Figure 20.
Pediatric heart transplant (1/1995 to 6/2003) 1-year predicted survival—hypothetical Cases 1 and 2.
Figure 21.
Pediatric heart transplant (1/1995 to 6/2003) 1-year predicted survival—hypothetical Case 3.
Figure 22.
Pediatric heart transplant (1/1995 to 6/2003) 1-year predicted survival—hypothetical Cases 4 and 5.
Long-term survival can also be influenced by prominent recipient factors (Figure 23). Figure 23 shows 3 patients who all have a congenital diagnosis leading to transplant, but differ primarily with regard to recipient age. Case 6, who is 3 months of age, has an estimated 5-year survival of >70% based on our experience with patients transplanted between 1995 and 1999. The infant survival compares favorably to the 6-year-old recipient (Case 7), whose survival tracks very closely to the 14-year-old recipient (Case 8). Both of these older age categories have predicted 5-year survivals of closer to 60%. Clearly, a congenital cardiac diagnosis leading to transplant has a strong negative impact on survival, but it is greatest in the older age recipient as compared with the infant.
Figure 23.
Pediatric heart transplant (1/1995 to 6/1999) 5-year predicted survival—hypothetical Cases 1, 2 and 3.
The impact of early events can also be reflected in late predicted survival for those surviving the first post-transplant year. For example, Figure 24 shows a hypothetical case of a 3-month-old child with a congenital diagnosis who underwent transplant but did not have rejection in the first year, compared with a similar child undergoing a transplant who was treated for rejection in the first year. The increasingly negative effect on late survival for those patients treated for early rejection can be readily seen. The power of these predicted survival curves is in the ability to combine multiple pre- and post-transplant recipient- and donor-related factors, and then use this information to generate a predicted survival curve. We anticipate that this approach may ultimately lead to protocols designed to minimize the impact of negative variables on long-term survival.
Figure 24.
Pediatric heart transplant (1/1995 to 6/1999) 5-year predicted survival conditional on survival to 1 year—hypothetical Case 1.
In summary, the eighth official pediatric report documents a stable volume of new pediatric heart transplant procedures but an increasing number of long-term survivors. The quality of survival is excellent as demonstrated by the physiologic rehabilitation and decreasing incidence of hospitalization. However, there are still issues related to late morbidity, late rejection and graft vasculopathy. We have continued to evaluate the data to look for risk factors associated with early and late mortality and morbidity. This report provides a further analysis of induction therapy and maintenance immunosuppression.
The predicted survival curves are new to this pediatric report and provide another way of using the current data to establish ways to help both patients’ families and health-care providers better understand the present impact of transplantation and to look for new ways to increase survival in the future. We continue to extend the period of follow-up as these children age, and future Registry reports will track late morbidity and mortality while also evaluating early outcomes of patients transplanted in a more recent era. Obviously, for pediatric recipients, late outcome becomes paramount to both families and care providers, because there is a decades-long survival need. Fortunately, the current data suggest that the patients’ half-life for survival once they have lived through the first year will likely be decades-long in the current era. The data from the current report indicate an ongoing concern that adolescent recipients seem to fare less well in the long term. On the other hand, the notable observation of an inverse relationship between ischemic time and late graft survival suggests new insights into the graft-recipient interaction, which is so critical to the late survival of all recipients. Predictive curves may allow us to model some of these factors, which could lead to changes in management and possible improvement in late adolescent survival.
Lung transplantation
Transplant Volume and Indications
The number of pediatric lung transplant recipients has been fairly constant over the last 5 years (Figure 25). The breakdown of the age of the recipient has also been fairly stable, with about 70% of the recipients in the teenage age bracket (Figure 26). This year, we have also added whether the donor type was deceased or living into the analysis; in the period between January 1996 and June 2004, a small number of donors were living and this trend will be tracked in future reports. Perhaps not all living donor transplant procedures were reported to the Registry because none are indicated in the infant age bracket. It is possible that not all centers have been reporting living donor procedures and we will evaluate this further in future reports.
Figure 25.
Age distribution of pediatric lung recipients by year of transplant.
Figure 26.
Age distribution by donor type for pediatric lung recipients (transplants: January 1996 to June 2004).
Table 7 shows the indications for pediatric lung transplantation for 2 age brackets: infants (<1 year of age) and childhood-age recipients (between 1 and 10 years of age). As reported previously, congenital heart disease is the leading diagnosis for lung transplantation for infants, but cystic fibrosis becomes the most common during the childhood years, followed by primary pulmonary hypertension and congenital heart disease. For the overall experience with adolescent recipients, cystic fibrosis accounts for 72% (Figure 27). The recent trend has been one of increasing frequency for cystic fibrosis and decreasing frequency for primary pulmonary hypertension. For this overall experience the third most common diagnosis leading to lung transplantation is re-transplantation, which is only 6% of the total.
Table 7. Pediatric Lung Transplantation: Indications (Transplants: January 1991 to June 2004)
| Diagnosis | Age: <1 year | Age: 1 to 10 years | ||
|---|---|---|---|---|
| Cystic fibrosis | 72 | 36.5% | ||
| Primary pulmonary hypertension | 7 | 13.7% | 26 | 13.2% |
| Congenital heart disease | 24 | 47.1% | 21 | 10.7% |
| Idiopathic pulmonary fibrosis | 12 | 6.1% | ||
| Pulmonary vascular disease | 7 | 13.7% | 6 | 3.0% |
| Re-transplant: non-obliterative bronchiolitis | 3 | 5.9% | 8 | 4.1% |
| Re-transplant: obliterative bronchiolitis | 11 | 5.6% | ||
| Interstitial pneumonitis | 3 | 5.9% | 7 | 3.6% |
| Obliterative bronchiolitis (not re-transplant) | 9 | 4.6% | ||
| Bronchiectasis | 2 | 1.0% | ||
| COPD/emphysema | 2 | 1.0% | ||
| Other | 7 | 13.7% | 21 | 10.7% |
Figure 27.
Diagnosis in pediatric lung recipients age 11 to 17 years.
Immunosuppression
The maintenance immunosuppression at 1 and 5 years after transplant follow-up is shown in Figure 28. What is clear is that tacrolimus has largely replaced cyclosporine as the preferred early calcineurin inhibitor, with up to about 75% of recently transplanted patients on tacrolimus at 1-year follow-up. The 5-year follow-up still shows a preponderance of patients on tacrolimus but a larger number on cyclosporine, probably reflecting their primary immunosuppression at the time of transplant. Previous Registry reports1 demonstrated that, until very recently, cyclosporine was more commonly used than tacrolimus, even at the 1-year follow-up. Thus, there seems a clear trend toward decreasing cyclosporine use both early and late after transplant. The use of rapamycin continues to be infrequent in pediatric lung transplantation. For the first time in the Registry reports of pediatric lungs, mycophenolate mofetil (MMF) is now more commonly used as an anti-proliferative agent than azathioprine. Almost 50% of patients are on MMF at both 1 and 5 years after transplant in this most recent era. Finally, virtually all patients at both 1 and 5 years after transplant remain on prednisone as part of a triple-drug immunosuppressive regimen.
Figure 28.
Pediatric lung recipient maintenance immunosuppression at time of follow-up. Note: Different patients analyzed at Year 1 and Year 5 (follow-up: January 2001 and June 2004).
Previous Registry reports1 have shown an increased utilization of induction therapy. This year’s report indicates that almost 50% of patients receive some form of induction immunosuppression at the time of transplant (Figure 29). The use of the interleukin-2R antagonists was about twice as likely as the polyclonal anti-T-cell preparations, and these classes of agents now account for virtually all induction use.
Figure 29.
Pediatric lung recipient induction immunosuppression for follow-up between January 2001 and June 2004.
Outcomes
SurvivalThe survival of pediatric lung transplant recipients for 3 different eras (1988 to 1992, 1993 to 1997, 1998 to 2003) is shown in Figure 30. There were no statistically significant differences between the 3 eras. The unconditional half-life has improved from 2.6 years in 1988 to 4.7 years in the period between 1998 and 2003. However, the half-life for survival, conditional on surviving to the first year post-transplant, was 8.2 years in the early era and 6 years in the most recent era. Thus, there may have been some improvements in early survival but these have not been matched by comparable improvements in late survival. Perhaps patient selection and technique may be confounding variables contributing to an apparent lack of improvement by era.
Figure 30.
Pediatric lung transplant Kaplan-Meier survival by era (transplants: January 1988 to June 2003).
There was also no significant difference in survival by age of the recipient (Figure 31), as has been the case in prior Registry reports. For patients who survive the first year after transplant the conditional half-life is approximately 7.5 years. The current Registry report includes an analysis of survival after transplant for pediatric recipients aged 11 to 17 years, according to donor type. These data are for the entire experience from January 1990 to June 2003; however, most of these living donor transplant procedures were performed after 1995. There is no difference in survival of the recipient based on whether the donor was living or deceased (Figure 32). The separation at the end of the curve reflects the small number of patients before 1995.
Figure 31.
Pediatric lung transplant conditional Kaplan-Meier survival by age group (transplants: January 1990 to June 2003).
Figure 32.
Pediatric lung transplant Kaplan-Meier survival by donor type for age 11 to 17 years (transplants: January 1990 to June 2003).
Survival after pediatric lung transplantation varies by procedure type (Figure 33). The number of single-lung procedures is about 10% of the number of bilateral procedures performed. Clearly, patients receiving a single-lung procedure fare less well than those receiving a bilateral double-lung procedure. The previous Registry report1 showed that much of this difference was seen in patients with the diagnosis of primary pulmonary hypertension, but apparently it applies to other diagnoses as well.
Figure 33.
Pediatric lung transplant Kaplan-Meier survival by procedure type (transplants: January 1990 to June 2003).
Induction immunosuppression therapy is used in about half of the pediatric lung recipients, and its use did not correlate with a difference in survival (Figure 34). The divergence at the 3-year time-point is likely artifactual due to the small numbers available for review at 3 years; at up to 2 years the curves are virtually superimposable.
Figure 34.
Pediatric lung transplant Kaplan-Meier survival stratified by induction use (transplants: January 2001 to June 2004).
Causes of Death
Table 8 shows the breakdown for cause of death vs period of time post-transplant. Donor issues and surgical complications are the predominant causes of graft loss in the first 30 days. Between 1 month and 1 year after transplant, infection is the leading cause of graft loss, accounting for almost 50% of deaths. Between 1 and 3 years after transplant, infection remains important, but now the leading cause is bronchiolitis, which maintains a constant risk at about 40% of deaths, even beyond 5 years after transplant. Lymphoma is an infrequent cause of death throughout the entire period of follow-up. In contrast to the pediatric heart transplant experience, acute rejection is rarely listed as a cause of death for pediatric lung recipients.
Table 8. Pediatric Lung Transplant Recipients: Cause of Death (Deaths: January 1992 to June 2004)
| Cause of death | 0 to 30 days (N = 60) | 31 days to 1 year (N = 89) | >1 year to 3 years (N = 99) | >3 years to 5 years (N = 46) | >5 years (N = 26) |
|---|---|---|---|---|---|
| Bronchiolitis | 6(6.7%) | 34(34.3%) | 20(43.5%) | 12(46.2%) | |
| Acute rejection | 1(1.7%) | 3(3.4%) | 1(2.2%) | ||
| Lymphoma | 3(3.4%) | 4(4.0%) | 1(2.2%) | 3(11.5%) | |
| CMV | 7(7.9%) | ||||
| Infection, non-CMV | 7(11.7%) | 34(38.2%) | 20(20.2%) | 12(26.1%) | 2(7.7%) |
| Graft failure | 22(36.7%) | 16(18.0%) | 17(17.2%) | 4(8.7%) | 4(15.4%) |
| Cardiovascular | 8(13.3%) | 2(2.2%) | 2(2.0%) | ||
| Technical | 9(15.0%) | 1(1.1%) | |||
| Multiple-organ failure | 2(3.3%) | 9(10.1%) | 9(9.1%) | 3(6.5%) | 2(7.7%) |
| Other | 11(18.3%) | 8(9.0%) | 13(13.1%) | 4(8.7%) | 3(11.5%) |
Specific Complications
The cumulative prevalence of morbidity factors within 5 years after pediatric lung transplantation is shown in Table 9. By 5 years post-transplant, approximately 72% of patients have hypertension, 30% have diabetes, and 22% have renal dysfunction. However, severe renal dysfunction (creatinine >2.5 mg/dl, dialysis or transplant) occurs in <10%.
Table 9. Post-Lung Transplant Morbidity for Pediatrics Cumulative Prevalence in Survivors Within 5 Years Post-transplant (Follow-up: April 1994 to June 2004)
| Outcome | Within 5 years | Total number with known response |
|---|---|---|
| Hypertension | 71.8% | N = 85 |
| Renal dysfunction | ||
| Total | 22.1% | N = 86 |
| Abnormal creatinine <2.5 mg/dl | 12.8% | |
| Creatinine >2.5 mg/dl | 5.8% | |
| Long-term dialysis | 2.3% | |
| Renal transplant | 1.2% | |
| Hyperlipidemia | 4.5% | N = 88 |
| Diabetes | 30.2% | N = 86 |
| Bronchiolitis obliterans | 30.6% | N = 62 |
There is a steady decrease in the proportion of recipients who are free from bronchiolitis obliterans (Figure 35), and about 50% of surviving patients have been diagnosed with bronchiolitis obliterans by 5 years after transplant. The accumulating incidence of bronchiolitis obliterans is analogous to the accumulating incidence of coronary artery vasculopathy in pediatric heart recipients. However, bronchiolitis obliterans is about twice as likely to occur as coronary artery vasculopathy. The use of induction immunosuppressive therapy has not influenced the percent of patients suffering from bronchiolitis obliterans, at least out to 3 years of follow-up (Figure 36).
Figure 35.
Freedom from bronchiolitis obliterans for pediatric lung recipients (follow-up: April 1994 to June 2004).
Figure 36.
Freedom from bronchiolitis obliterans for pediatric lung recipients by induction use (transplants: April 1994 to June 2004).
Functional Status
The functional status of surviving pediatric lung recipients is excellent, with no activity limitations in the vast majority out to 5 years of follow-up (Figure 37). As shown in previous reports, hospitalization is more common in pediatric lung recipients than in heart recipients, with about 50% requiring hospitalization through 5 years of follow-up.
Figure 37.
Pediatric lung recipients’ functional status (follow-up: April 1994 to June 2004).
In summary, the overall experience for pediatric lung transplantation shows a stable number of recipients and little change in the short- and long-term outcomes with era. We also could not demonstrate any benefit based on age of the recipient, nor could we demonstrate a benefit based on whether the donor was living or deceased. Single-lung transplantation is at a disadvantage compared with double-lung procedures. However, for patients who survive the first year post-transplant, the 7-year conditional survival is good, and the data indicate most patients are without activity limitation. Changes in immunosuppression were documented and include more frequent use of tacrolimus and MMF in recent years. However, at this time-point, there has been no significant change in survival of the recipients. Morbidity remains a prominent issue, primarily in the form of hypertension, diabetes and bronchiolitis obliterans in late survivors. However, the development of diabetes in adolescents is clearly related to the diagnosis of cystic fibrosis leading to transplant.2 Late death and late morbidity, in the form of bronchiolitis obliterans, still reflect an ongoing problem, with control of immune activation and the balance between the risk of infection and the risks of chronic rejection in the pediatric lung recipient.
The authors thank Dawn Schmeck for help with manuscript preparation.
References
- . Registry for the International Society for Heart and Lung Transplantation (seventh official pediatric report—2004) . J Heart Lung Transplant . 2004;23:933–947
- The Registry of the International Society for Heart and Lung Transplantation (sixth official pediatric report—2003) . J Heart Lung Transplant . 2003;22:636–652
- . The Registry of the International Society for Heart and Lung Transplantation (twenty-first official adult heart transplant report—2004) . J Heart Lung Transplant . 2004;23:796–803 July;
PII: S1053-2498(05)00344-X
doi:10.1016/j.healun.2005.05.020
© 2005 International Society for Heart and Lung Transplantation. Published by Elsevier Inc. All rights reserved.
Volume 24, Issue 8 , Pages 968-982, August 2005
