Solid Organ Allograft Survival Improvement in the United States: The Long-Term Does Not Mirror the Dramatic Short-Term Success


Corresponding author: Herwig-Ulf Meier-Kriesche,


Organ survival in the short-term period post-transplant has improved dramatically over the past few decades. Whether this has translated to a long-term survival benefit remains unclear. This study quantifies the progression of nonrenal solid organ transplant outcomes from 1989 to 2009 in liver, lung, heart, intestine and pancreas transplants. Long-term graft survival was analyzed using data on adult solid organ transplant recipients from the UNOS/SRTR database and is reported as organ half-life and yearly attrition rates. Liver, lung, heart, intestine and pancreas half-lives have improved from 5.8 to 8.5, 1.7 to 5.2, 8.8 to 11, 2.1 to 3.6 and 10.5 to 16.7 years, respectively. Long-term attrition rates have not shown the same consistent improvement, with the yearly attrition rate 5–10 years post-transplant for liver, lung, heart and pancreas changing from 4.7 to 4.3, 10.9 to 10.1, 6.4 to 5.1 and 3.3 to 4.2, respectively. Attrition rates for intestine and pancreas transplantation alone display more variability due to smaller sample size but exhibit similar trends of improved first-year attrition and relatively stagnant long-term attrition rates. With first-year survival and attrition rates almost at a pinnacle, further progress in long-term survival will come from targeting endpoints beyond first-year rejection and survival rates.


United Network for Organ Sharing


Scientific Registry of Transplant Recipients


simultaneous pancreas–kidney


pancreas after kidney


pancreas transplant alone


Outcomes of solid organ transplantation have improved significantly during the last two decades (1). Improved immunosuppressive regimens have reduced acute rejection rates, and graft loss resulting from acute rejection has also declined (2). In parallel, diagnosis and treatment of infectious complications as well as prophylaxis regimens have allowed transplant recipients to tolerate more intense immunosuppressive therapy (3). At the turn of the century, it was reported that not only short-term survival was advancing for kidney transplant recipients, but that long-term survival was improving as well (4). This would have had far reaching consequences as it suggested that targeting short-term endpoints would be followed by long-term improvements, thus implying that the ongoing strategies were successful and that acute rejection and short-term graft survival were useful surrogate endpoints for studies. Unfortunately, it later became clear that long-term survival rates had in fact changed very little despite dramatic short-term improvements, (2) and that reduced acute rejection rates were not automatically followed by better graft survival (5).

Recently it has been confirmed that kidney graft attrition rates in the first year and in subsequent years are two separate processes that will not necessarily mirror each other's evolution (6). These data highlight a dilemma in kidney transplantation where short-term survival is greater than 90% and can hardly be improved further, yet the constant long-term attritions make substantial improvements in long-term graft survival an elusive goal. Other solid organ transplants have seen similarly dramatic improvements in short-term graft survival but whether long-term graft survival is also unlinked in other solid organ transplants is unknown. The current study investigates half-lives by era, and short- and long-term graft attrition rates in heart, liver, lung, intestinal, pancreatic and simultaneous pancreas–kidney (SPK) transplantation.

Material and Methods


We examined data from the Scientific Registry of Transplant Recipients (SRTR) for transplants that occurred between January 1, 1989 and November 1, 2009 in recipients 18 years or older. Multiorgan transplants were excluded from the analysis except in the case of intestine and SPK or pancreas after kidney (PAK).

Outcome measures

We analyzed long-term graft survival by estimating survival half-lives (2) and we analyzed short- and long-term graft attrition rates stratified by year of transplant (6). Graft loss was defined as either loss of the organ or patient death with a functioning organ.

Unadjusted graft survival rates: Unadjusted graft survival rates were derived from Kaplan–Meier estimates. Percentages of graft survival were displayed with the respective standard errors as a measure of variability.


Univariate half-lives were determined as median half-lives, i.e. the intersection point of the Kaplan–Meier curve with the 50% survival threshold. Actual half-lives are defined as those instances where all patients reached the 50% mark, actuarial half-lives for those instances when only a proportion of patients had reached the 50% mark, and projected half-lives when none of the patients had reached the 50% mark. Actual and actuarial half-lives are grouped together in the tables and figures, and the projected half-lives are shown separately in the supplementary tables. Projections were obtained by forecasting the Kaplan–Meier curves from a point of stable attrition, which was fairly consistently located between 3 and 8 years of survival yielding a period of 5 years from which the forecasts were based off of. For intestine transplants, the point of stable attritions was located between 1.5 and 3.5 years due to the shorter graft life as compared to other organs analyzed. Forecasted projections were carried out using ordinary least squares point estimates. In order to show the degree of concordance or discordance of the actuarial versus forecasted half-lives we show a certain overlap between the two methods at time points where both actuarial data are still available but we also forecasted.

Attrition rates

Annualized attrition rates (annualized graft loss rates) were calculated by first acquiring actual 1-year, 3-year, 5-year and 10-year Kaplan–Meier estimated survival rates. The total number of patients surviving during the time period was subtracted from the number of patients originally entering the cohort and divided then by the original number entering the cohort to obtain an absolute failure percentage. The percentage of absolute failures was then divided by the total number of years in the follow-up interval to obtain annualized failure rate.

Statistical models

Outcomes were measured by Kaplan–Meier models. Half-lives were calculated based on actual and projected follow-up where applicable. Half-lives based on actual versus projected follow-up are displayed distinctly in the results. Projected half-lives were utilized using the univariate Kaplan–Meier model for allograft failure. All analyses were conducted using SAS (v.9.2, Cary, NC).



Between the years of 1989 and 2009, we analyzed deceased donor adult organ transplants consisting of liver (N = 79 695), lung (N = 18 326), heart (N = 38 795), intestine (N = 758), pancreas transplant alone (PTA) (1554), PAK (N = 3939), pancreas from SPK (N = 16 307), kidney from SPK (N = 16 307).

Graft survival

Figure 1 shows overall graft survival of standard criteria deceased donor transplants for liver, lung, heart and intestine between 1989 and 2009 and the respective median half-lives based on where the survival curve crosses the 50% survival line. Table 1 displays unadjusted graft survival rates for all organs with standard errors.

Figure 1.

Kaplan–Meier cumulative graft failure for liver, lung, heart and intestine transplants from transplant years 1989 to 2009.

Table 1.  Unadjusted graft survival rates by year and organ with standard errors in subscript
 HeartLiverLungIntestinePancreas (SPK)
1 year3 years5 years10 years1 year3 years5 years10 years1 year3 years5 years10 years1 year3 years5 years10 years1 year3 years5 years10 years
198982.30.974. 51.12.81
199182.20.874. 33.327.2233.327.
1996 84.80.776.80.969.
1997 84.50.877.50.972.


Figure 2 shows actual/actuarial, and projected half-lives of liver, lung, heart and intestine. The actual or actuarial half-lives by year for all organ types are shown in Table 2, with the projected half-lives marked as ‘forecasted’ in the second shaded line for liver, lung, heart, intestine, pancreas and SPK transplants. The overlap between the actuarial half-lives and projected half-life represents instances where the half-life was projected from the actuarial data but forecast was generated in parallel.

Figure 2.

Actuarial (black solid line) and projected (gray dotted line) graft failure half-lives of liver, lung, heart and intestine.

Table 2.  Kaplan–Meier estimates of cumulative graft half-lives by transplant year for liver, lung, heart, intestine, pancreas transplant alone (PTA), pancreas after kidney (PAK), pancreas from simultaneous pancreas/kidney (SPK) and kidney from SPK
Organ type19891990199119921993199419951996199719981999200020012002200320042005
Liver (N = 79 694)5.86.988.38.5109.99.910.210.5       
Liver forecasted
Lung (N = 18 326)  
Lung forecasted        4.444.
Heart (N = 38 795)8.89.498.89.410.19.710.110.410.4       
Heart forecasted       1010.510.41111.111.711.512.210.311
Intestine (N = 758)  
Intestine forecasted   
Pancreas (PTA) (N = 1554) 
Pancreas (forecasted) 
Pancreas (PKA) (N = 3939)  
Pancreas (forecasted)   
Pancreas/SPK (N = 16 307)10.56.988.679.16109.2711.310.110.711       
Pancreas/SPK (forecasted)      10.710.311.611.29.8511.410.714.812.314.216.7
Kidney/SPK (N = 16 307)       
Kidney/SPK (forecasted)      10.910.510.811.410.813.712.213.411.114.321.3

Of 79 694 liver transplants, the half-life in 1989 was 5.8 years. This increased to 10 years by 1994, but has not improved much more, with the more recent projections ranging between 9.9 and 11 years up through the year 2004. In 2005, the projected half-life is 8.5 years.

Lung transplants (N = 18 326) initially started with a half-life of 1.7 years in 1989, climbed to 4.3 years in 1993 and are projected to be 5.2 years in 2005.

The heart transplant population (N = 38 795) half-life increased a total of 2.2 years from 1989 (8.8 years) to 2005 (11 years).

Intestine transplant (N = 758) actuarial half-life was 2.1 years in 1991 but has fluctuated between 0.5 and 4.8 years. It is forecast to be 3.6 years in 2005.

Figure 3 illustrates pancreatic and kidney from SPK transplant half-lives.

Figure 3.

Actuarial (black solid line) and projected (gray dotted line) graft failure half-lives of pancreas transplant alone (PTA), pancreas after kidney (PAK), pancreas from SPK (P-SPK) and kidney from SPK (K-SPK).

Pancreas transplants alone (N = 1554) started with a half-life of 0.9 years in 1989, and have increased to actuarial half-lives ranging between 3 and 5.8 from 2000 to 2003. The forecast half-life is 5.6 in 2005. PAK (N = 3939) actuarial half-life has increased from 1.5 years in 1989 to 6.9 years in 2001, and is forecast to be 6.9 years in 2005.

Pancreas from SPK (N = 16 307) started with a half-life of 10.5 years in 1989 and 11 years in 1998. Forecasted data show an increase to 14.8 years in 2002 and 16.7 years in 2005. Kidney transplant half-lives from SPKs shows a very similar trend to pancreas, with the initial half-life in 1989 being 9.6 and forecast half-lives showing an increase to 21.3 years in 2005.

Graft attrition rates

Figures 4 and 5 show the 0–1, 1–3, 3–5 and 5–10 year attrition rates for each organ. Table 3 displays the graft attrition rates by transplant year for liver, lung, heart, intestine, pancreas and kidney from SPKs.

Figure 4.

Cumulative graft failure yearly attrition rates of all transplant types for liver, lung, heart and intestine. Attrition for 0–1 year post-transplant shown in yellow, 1–3 year post-transplant in red, 3–5 year post-transplant in green and 5–10 year post-transplant in purple.

Figure 5.

Cumulative graft failure yearly attrition rates of pancreas after kidney (PAK), pancreas alone (PTA), pancreas from SPK (P-SPK) and kidney from SPK (K-SPK). Attrition for 0–1 year post-transplant shown in yellow, 1–3 year post-transplant in red, 3–5 year post-transplant in green and 5–10 year post-transplant in purple.

Table 3a.  Kaplan–Meier estimates of cumulative graft attrition rate by transplant year for liver, lung, heart, intestine, pancreas transplant alone (PTA) and pancreas after kidney (PAK)
 0–1 year35.332.329.227.225.622.822.323.320.819.920.419.319.817.918.617.819.217.415.015.9
 1–3 years5.  
 3–5 years4.    
 5–10 years4.         
 0–1 year42.228.432.830.624.625.
 1–3 years13.513.813.512.010.711.812.112.511.  
 3–5 years15.812.113.612.912.210.612.89.910.911.311.010.511.39.99.511.0    
 5–10 years10.910.210.410.19.810.511.         
 0–1 year17.015.517.117.216.414.415.
 1–3 years4.  
 3–5 years5.    
 5–10 years6.         
 0–1 year
 1–3 years 25.033.322.225.033.3  
 3–5 years    
 5–10 years  20.012.0         
Pancreas (PTA)
 0–1 year53.655.648.627.655.432.038.931.430.919.717.525.523.
 1–3 years20.825.020.112.612.57.513.624.710.911.710.  
 3–5 years21.4    
 5–10 years15.         
Pancreas (PAK)
 0–1 year46.748.751.444.449.129.629.932.427.227.820.427.018.122.722.421.723.522.018.418.8
 1–3 years18.810.526.520.017.914.510.610.413.  
 3–5 years15.020.08.316.    
 5–10 years10.         
Table 3b.  Kaplan–Meier estimates of cumulative graft attrition rate by transplant year for pancreas from simultaneous pancreas/kidney (SPK) and kidney from SPK
Pancreas (SPK)
 0–1 year23.530.619.320.922.220.117.916.615.817.417.717.215.815.314.815.513.616.614.413.7
 1–3 years5.  
 3–5 years5.    
 5–10 years3.         
Kidney (SPK)
 0–1 year14.922.614.616.014.914.510.910.
 1–3 years5.  
 3–5 years6.    
 5–10 years5.         

The 0- to 1-year attrition rates for all organs have shown improvement (yellow lines). Liver and lungs first-year attrition rates progressively improved since 1989. In contrast, first-year heart attrition rates were lower than other organs in 1989 at 17.7, but have also shown the least improvement over time (12.2 in 2008). Intestinal transplant attrition rates, while variable, have improved also in the first year from 69.2 in 1993 to 23.3 in 2008. The 1–3, 3–5 and 5–10 year attrition rates have remained relatively stagnant for liver, lung, heart and intestine.

Pancreas transplants (Figure 5) all exhibited the most attrition rate improvement in the 0- to 1-year rate, with modest improvement in subsequent years. Kidney from SPK also showed a similar trend with attrition rates improving from 14.9 to 7.2 in the first year, then smaller improvements in the 5–10 attrition rate (5.2–4.7).


Advances in surgical techniques and immunosuppression have vastly improved transplantation throughout the past several decades, but higher risk individuals are also being transplanted. Despite this, solid organ transplant survival has improved during the last two decades, as reported by UNOS data analysis of 1, 5 and 10-year survival rates (1). Consequently median half-lives have also improved for all organs (Figure 2), (Figure 3). Half-lives are particularly helpful in communicating prognoses to prospective transplant recipients. It is well recognized that first-year graft attrition is highest, and consequently, projections of half-lives have to commence from a time point when attrition rates have stabilized usually after the first year (6). Therefore, graft attrition is at least a two-phase process.

Despite the progress in organ transplantation, our data show that long-term graft attrition has not significantly changed during the past two decades. The progressive improvement in graft survival is exclusively driven by first-year graft survival improvements. While graft attrition in the first-year post-transplant has improved dramatically over time for most organs, long-term attrition is disappointingly stable. While 1-year graft survival rates are now exceeding 80% for most organs, it is becoming clear that further significant improvements in survival have to come from reducing the long-term attrition rates. It is not surprising that dramatic improvements have been made in first-year graft survival because this is the time frame clinical studies have always targeted. Pivotal trials for immunosuppression approval in the United States are typically based on 1-year endpoints. Most recently, a combined endpoint of acute rejection within the first year, graft loss, patient death or loss to follow-up is typically required for transplant drug approval. Similarly, most other research trials in organ transplantation have focused on early survival and early endpoints; hence it comes as no surprise that short-term survival has improved dramatically, but long-term attrition has disappointingly not changed.

These data highlight the importance of examining graft attrition rates by time period after transplantation. When interpreting cumulative data such as graft survival or median half-lives, overall survival seems to improve; however, only when considering attrition rates by time period post-transplant does it become clear that there is a significant problem.

It is interesting to note the differences between transplant organs. Liver transplants follow the same trend as kidney transplants (6), with a dramatic improvement in first-year graft attrition during the past two decades, but with unchanged 1–3, 3–5 and 5–10 year attrition rates. In comparison, heart transplant first-year attrition rate improvements have not been as dramatic, and the long-term attrition rates are also stagnant. Consequently, heart transplant half-lives have not improved as much as the liver transplant half-lives. Both lung and intestinal transplantation have seen dramatic improvements in first-year attrition rates, but also for these organs, long-term attrition rates are fairly constant since 1989.

Each transplant organ has different reasons for failure but the similarity of dissociation between short and long-term attrition hints toward some possible similarities. Improvement of acute rejection rates over the last 20 years has resulted from more potent immunosuppression but is now overshadowed with more infectious and malignancy-related complications along with other direct toxicities. For example, with liver transplantation, increased immunosuppression decreases acute rejection rates in hepatitis C patients, but it may also increase the risk for hepatitis C recurrence (7). In heart transplantation, higher steroid and CNI exposure will decrease rejection rates but may contribute to coronary artery disease driving graft loss in the long term. One common problem to all organ transplants is the nephrotoxicity of calcineurin inhibitors (8). While they are very efficacious in preventing acute rejection they are also nephrotoxic and it is well known that the outcome of any nonrenal organ transplant recipients who end up on dialysis is dramatically worse (8,9). Therefore the nephrotoxicity of the current and most commonly used immunosuppressive protocols is the ‘Achilles’ heel’ of transplantation, and contributes to the dichotomous separation of short and long-term survival. Calcineurin inhibitors also have several other toxicities including diabetogenicity, dyslipidemia, hypertension and others that can advance cardiovascular and renal disease among transplant patients. Similarly, steroids have a vast array of toxicities particularly counterproductive to the cardiovascular health of transplant patients. Clearly in all organ transplants, the ideal balance between sufficient immunosuppression and toxicities is difficult to achieve. Immunosuppressive regimens are developed for populations rather than individuals. Unfortunately, markers to adequately assess an individual patient's need for immunosuppression have been elusive. When the overall immunosuppressive burden of a population is increased in the attempt to lower acute rejection rates, immunosuppression will be still be inappropriately increased in those patients who would not have rejected even with the previous regimen. For example, one may consider a hypothetical recipient population with an initial rejection rate of 40%, and immunosuppressive therapy is then intensified to decrease rejection rates to 20%. This move will benefit possibly those 20% of patients who now will not experience rejection, yet the 60% of patients who would not have had rejection on the previous regimen are now by definition over immunosuppressed. Identifying these individuals who may have better survival outcomes with less immunosuppression and hence less long-term side effects remains evasive. The converse is true for reductions in overall immunosuppression for populations when trying to deal with toxicities. Until there is an adequate assessment for the individual need for immunosuppression at any given time post-transplant, it will be difficult to achieve substantial improvements in long-term attrition rates. Gene chip arrays are yielding the promise of detecting rejection episodes possibly early enough to increase immunosuppression before any irreversible damage to the organ occurs (10). One could then empirically come to an optimal level of immunosuppression, yet these data are still preliminary and, although promising, any new markers will have to be tested in well-controlled randomized trials before they can be safely used in the clinical setting.

In summary, the long-term attrition rates for all organ transplants have not significantly changed during the past 20 years. Half-life improvement has come from dramatic improvements in short-term attrition rates. As short-term survival is excellent at this point, significant increments in half-life can only come from reductions in the long-term organ attrition rate. The transplant community and the regulatory authorities will have to shift their attention from short-term endpoints to long-term goals in both patient care and research to achieve this goal.


We would like to extend our appreciation to the Central Florida Kidney Center, Inc. for supporting this work through the endowment of the Eminent Scholar Chair in Nephrology and Hypertension. We would also like to thank Melissa Smiles for editing and proof reading the manuscript. The authors do not have any conflicts of interest or disclosures with regards to the data presented in this manuscript.


The authors of this manuscript have no conflicts of interest to disclose as described by the American Journal of Transplantation.