- Email: [email protected]
The Clinical Risk in Heart Transplantation: Is It Evaluable?
The Clinical Risk in Heart Transplantation: Is It Evaluable? G. Guzzi, M. Maiani, G. Aresu, V. Tursi, E. Spagna, and U. Livi ABSTRACT At the moment, there is no score to evaluate clinical risk in heart transplantation. There is a need for such an instrument due to the extended criteria for donations and for recipient evaluation for transplantation. We divided the 203 consecutive patients who underwent heart transplantation (HTx). Between January 1999 and December 2007 into two groups: high and low risk based on several common well-defined variables. Donors were also divided into high- and low-risk groups. We matched the four groups to obtain risk cohorts: GA (high risk), GB and GC (intermediate risk) versus GD (low risk). We analyzed the 30 day-mortality showing a significant difference between GD and the other groups (P ⫽ .05) in contrast to no significant difference in 1- and 3-year survival rates among GA, GB, GC, and GD. Although the development of a specific score for heart transplantation is desirable and would be useful, a careful, case-by-case evaluation is indispensable. LINICAL RISK is defined as the probability for a patient to be the victim of an adverse event, that is, to suffer damage or inconvenience caused, albeit unintentionally, by medical care, causing prolongation of hospitalization, worsening of health condition, or death. Any surgical procedure has intrinsic risk. It should be the lowest possible; the clinical risk in surgery must be measurable to be acceptable. In cardiac surgery, surgical risk has been evaluated by several stratification methods: Parsonnet, euroSCORE, STS.1– 4 Such scores are not applicable “as it is” in heart transplantation, because the population has some peculiarities, being highly heterogeneous in pathology, clinical status, timing of surgery, donor variables, and transplant center facilities. Well-established criteria have been adopted in the general experience of heart transplantation to consider a donor ideal, acceptable, or unfit, and also to select a recipient by assessing the potential risk of death after transplantation. There criteria have been derived from several multi-institutional, retrospective studies and validated by sophisticated statistical analyse of variables representing weighted risks, which are amplified in cases of their association.5,6 Furthermore, improved results of heart transplantation have allowed selection of patients at higher risk. In the meantime, the donor shortage has stimulated the use of suboptimal donors to meet increasing demand. This more recent widening of standard criteria requires a reevaluation of risk scoring when those adverse features are matched.
C
© 2011 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 43, 307–310 (2011)
For this purpose, we retrospectively analyzed our transplantation experience focusing on the survival effects of recipient and donor variables, seeking to eventually propose a prospective score to optimize outcomes of donor-recipient matching. MATERIALS AND METHODS The analysis was performed using data on all 203 heart transplantations performed between January 1999 and December 2007. The recipients had to have at least 1-year follow-up. The 2007 Registry of the International Society for Heart and Lung Transplantation7 has been used to find the relative risk (RR) for 1-year mortality among heart transplant recipient: Among all listed variables, we close those that were present in our patients, consequently creating a risk score. The 10 variables were: ventricular assist device, insulin dependent diabetes, history of dialysis, cerebrovascular event prior to transplantation, temporary circulatory support (ECLS/IABP), mechanical ventilation at time of transplantation, prior sternotomy, age ⬎ 55 years, bilirubin ⬎ 1.1 mg/dL, and creatinine ⬎1.3 mg/dL. The sum of all variables present in our population yielded a mean calculated RR value of 1.82. Subsequently, we divided recipients into two groups: G1R or lower risk (RR ⱕ 1.82; n ⫽ 118; mean value ⫽ 1.27) and G2R higher risk (RR ⬎ 1.82; n ⫽ 85; mean value ⫽ 2.58). From the Cardiothoracic Department, University Hospital “S. Maria della Misericordia,” Udine, Italy. Address reprint requests to Giorgio Guzzi, MD, Cardiothoracic Department, University Hospital “S. Maria della Misericordia,” Udine I-33100 Italy. E-mail: [email protected] 0041-1345/–see front matter doi:10.1016/j.transproceed.2010.09.108 307
308
GUZZI, MAIANI, ARESU ET AL
Fig 1.
Donor quality was characterized on basis of seven more common variables: ischemia time ⬎180 minutes; age ⬎30 years; serum sodium ⬎150 mMol/L; presence of coronary lesions, cardiac hypertrophy (left ventricular septum thickness ⬎1.3 cm); ejection fraction ⬍50% and multiple IV inotropes at time of retrieval. Donors were divided into two groups. They were considered to be low risk when no or only one of the risk factors was present; otherwise, high-risk donors were those with more than one risk factor. Accordingly, they were divided in G1D or optimal donors (n ⫽ 156; score ⱕ 1) versus G2D suboptimals donors (n ⫽ 47, score ⬎ 1). Analyzing high- or low-risk recipients who received a high- or low-risk donor heart four matching we obtained groups (Fig 1). Group A (GA; n ⫽ 25) the highest risk possible, was composed of the worst recipients (G2R) and donors (G2D). Group B (GB) and group C (GC) included intermediate-risk subjects, either for donor (GB) or recipient characteristics (GC). The first combination of lower-risk recipient (G1R) with higher-risk donor (G2D) included 22 patients. In contrast, the second group (n ⫽ 60) matched higher-risk recipients (G2R) with lower-risk donors (G1D). Finally, group D (GD n ⫽ 97) was the one at lower risk, including the best recipients (G1R) and donors (G1D). Tables 1 and 2 list the overall characteristics of our population divided into
the four groups. Statistical evaluation was performed using SPSS, Data Mining and Statistical Analysis Software (SPPS Inc, Chicago, Ill, USA) with the Kaplan-Meier method. Statistical significance of survival curves was assessed using the log-rank test, with a cutoff value P ⱕ .05.
RESULTS
Figure 2 shows the overall results of our analysis. The 30 day-mortality was 12% in GA, 9% in GB, 12% in GC, and, finally, 3% in GD. GD showed a statistically significant difference versus the other groups (P ⫽ .05). The most common causes of death were primary graft failure and multiple organ failure. Overall, the mean follow-up was 52 ⫾ 34 months, including 41 ⫾ 33 months among GA; 52 ⫾ 34 months for GB; 48 ⫾ 33 months for GC, and 41 ⫾ 33 months for GD. Survival rates at 1 year were 80%, 86%, 85%, and 94% for GA, GB, GC, and GD, respectively; at 3 years, they were 76%, 82%, 80%, and 90% for the same groups, respectively, there was no significant difference.
Table 1.
Patients (n) M/F (%) Age (yr), mean (median) ⫾ SD; range Sternotomy, n (%) Diabetes, n (%) Dialysis, n (%) Creatinine (mg/dL), mean (median) ⫾ SD; range Bilirubin (mg/dL), mean (median) ⫾ SD; range CV events, n (%) IABP, n (%) ECMO, n (%) VAD, n (%) Ventilator, n (%) Mean cumulative RR
GA
GB
GC
GD
25 88/12 60 (62) ⫾ 10; 14–67 14 (56%) 11 (44%) 1 (4%) 1.6 (1.5) ⫾ 0.6; 0–3.1
22 77/23 55 (57) ⫾ 10; 27–70 7 (32%) 0 0 1.5 (1.5) ⫾ 0.9; 0.4–3.3
60 85/15 58 (60) ⫾ 8; 29–69 29 (48%) 24 (25%) 3 (5%) 1.5 (1.3) ⫾ 1.0; 0–7.8
96 76/24 51 (54) ⫾ 8; 12–65 21 (22%) 0 0 1.2 (1.1) ⫾ 0.3; 0.3–1.9
1.3 (1.1) ⫾ 1.0; 0–3.5
0.1 (0.1) ⫾ 0.1; 0–0.25
1.5 (1.1) ⫾ 1.5; 0.5–9.8
1.2 (0.9) ⫾ 0.9; 0.3–5.5
2 (8%) 0 2 (8%) 3 (12%) 0 2.49
1 (5%) 0 0 0 0 1.36
5 (8%) 8 (13%) 3 (5%) 9 (15%) 5 (8%) 2.69
2 (2%) 0 0 0 0 1.25
CLINICAL RISK IN HEART TRANSPLANTATION
309 Table 2.
Patients (n) M/F (%) Age (yr), mean (median) ⫾ SD; range Age ⬎ 55 yr, n (%) Ischemia time (min), mean (median) ⫾ SD; range Ischemia time ⬎ 240 min, n (%) Inotropic support, n (%) None Low (⬍10 g/kg/min) High (any dosage ⫹ cathecol.) Serum sodium (mg/dL), mean (median) ⫾ SD; range Coronary artery disease, n (%) LVEF ⬍ 50%, n (%) Cardiac hypertrophy, n (%) Variables number, mean (median) ⫾ SD; range
GA
GB
GC
25 79/21 51 (55) ⫾ 13; 18–67
22 82/18 42 (43) ⫾ 17; 14–66
60 60/40 37 (38) ⫾ 13; 17–63
96 68/32 35 (34) ⫾ 14; 13–65
14 (56%) 225 (243) ⫾ 81; 83–395
10 (45%) 235 (258) ⫾ 74; 111–350
3 (5%) 188 (204) ⫾ 56; 87–314
0 183 (185) ⫾ 56; 62–298
15 (60%)
14 (64%)
7 (12%)
15 (16%)
3 (12%) 7 (28%) 15 (60%) 147 (146) ⫾ 10; 132–173
0 4 (18%) 18 (82%) 150 (151) ⫾ 8; 136–170
12 (20%) 30 (50%) 18 (30%) 149 (150) ⫾ 7; 132–161
19 (20%) 43 (45%) 34 (35%) 145 (145) ⫾ 8; 117–163
5 (20%) 3 (12%) 4 (16%) 3.2 (3.0) ⫾ 0.8; 2–5
3 (14%) 1 (5%) 5 (23%) 3.3 (3.0) ⫾ 0.6; 3–5
DISCUSSIONS
Multiple variables influence outcomes after heart transplantation; their association could greatly affect overall results. Our study, even considering the not so large patient cohort (above all in the higher-risk groups) seemed to demonstrate that recipient are more important than donor characteristics, with the best results obtained in the GD lowest-risk group. However, also in cases of both high-risk recipients and donors, early results seemed to be satisfactory. Furthermore, donor characteristics seemed to more important by impact early survival, while recipient ones, long-term results. Obviously, larger cohorts of patients and longer follow-ups are needed to validate these results.
Fig 2.
1 (2%) 0 3 (5%) 1.4 (1.0) ⫾ 0.7; 0–2
GD
2 (2%) 0 10 (10%) 1.5 (2.0) ⫾ 0.7; 0–2
Some ethical issues arise from our analysis in cases of elevated clinical risk. The first one regards which outcome must be considered acceptable: is it preferable for heart transplant population survival; that is, to choose the best recipient on the waiting list to match with the best available donor. Or, as our analysis seems to suggest (at least for early/midterm outcomes), giving a chance of survival to all patients affected by heart failure by performing a heart transplantation? And more, in the era of extended donor and recipient criteria, which patient should be transplanted with a nonoptimal organ? We think there is no unequivocal answer to these complex medical and ethical issues, because what is medically feasible is not always appropriate. Therefore, these open
Fig 3.
310
GUZZI, MAIANI, ARESU ET AL
questions could urge a larger discussion in the scientific community about what is better to do. Our modest contribution to this topic is that, although the development of a mathematically specific score for heart transplantation is desirable and would be useful to best guide the choices about who undergoes transplantation, our policy and experience suggest that a final decision cannot leave aside an individual and case-bycase evaluation, so to offer the best opportunity to a specific patient, inspired by the principle that we cannot add days to life, but add life to any single day. Figs. 3–5 REFERENCES
Fig 4.
Fig 5.
1. Parsonnet V, et al: A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 79:I3, 1989 2. Roques F, Nashef SAM, Michel P, et al: Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 15:816, 1999 3. STS Score. Available at http://209.220.160.181/STSWebRiskCalc261/de.aspx 4. Higgins TL, Estafanous FG, Loop FD, et al: Stratification of morbidity and mortality outcome by preoperative risk factors in coronary artery bypass patients. A clinical severity score. JAMA 267:2344, 1992 5. Bourge RC, Naftel DC, Costanzo-Nordin MR, et al: Pretransplantation risk factors for death after heart transplantation: a multiinstitutional study. The Transplant Cardiologists Research Database Group. J Heart Lung Transplant 12:549, 1993 6. Young JB, Naftel DC, Bourge RC, et al: Matching the heart donor and heart transplant recipient. Clues for successful expansion of the donor pool: a multivariable, multiinstitutional report. The Cardiac Transplant Research Database Group. J Heart Lung Transplant 13:353, 1994 (discussion 364) 7. Taylor DO, Edwards LB, Boucek MM, et al: Registry of the International Society for Heart and Lung Transplantation: Twentyfourth Official Adult Heart Transplant Report—2007. J Heart Lung Transplant 26:769, 2007
C
© 2011 by Elsevier Inc. All rights reserved. 360 Park Avenue South, New York, NY 10010-1710 Transplantation Proceedings, 43, 307–310 (2011)
For this purpose, we retrospectively analyzed our transplantation experience focusing on the survival effects of recipient and donor variables, seeking to eventually propose a prospective score to optimize outcomes of donor-recipient matching. MATERIALS AND METHODS The analysis was performed using data on all 203 heart transplantations performed between January 1999 and December 2007. The recipients had to have at least 1-year follow-up. The 2007 Registry of the International Society for Heart and Lung Transplantation7 has been used to find the relative risk (RR) for 1-year mortality among heart transplant recipient: Among all listed variables, we close those that were present in our patients, consequently creating a risk score. The 10 variables were: ventricular assist device, insulin dependent diabetes, history of dialysis, cerebrovascular event prior to transplantation, temporary circulatory support (ECLS/IABP), mechanical ventilation at time of transplantation, prior sternotomy, age ⬎ 55 years, bilirubin ⬎ 1.1 mg/dL, and creatinine ⬎1.3 mg/dL. The sum of all variables present in our population yielded a mean calculated RR value of 1.82. Subsequently, we divided recipients into two groups: G1R or lower risk (RR ⱕ 1.82; n ⫽ 118; mean value ⫽ 1.27) and G2R higher risk (RR ⬎ 1.82; n ⫽ 85; mean value ⫽ 2.58). From the Cardiothoracic Department, University Hospital “S. Maria della Misericordia,” Udine, Italy. Address reprint requests to Giorgio Guzzi, MD, Cardiothoracic Department, University Hospital “S. Maria della Misericordia,” Udine I-33100 Italy. E-mail: [email protected] 0041-1345/–see front matter doi:10.1016/j.transproceed.2010.09.108 307
308
GUZZI, MAIANI, ARESU ET AL
Fig 1.
Donor quality was characterized on basis of seven more common variables: ischemia time ⬎180 minutes; age ⬎30 years; serum sodium ⬎150 mMol/L; presence of coronary lesions, cardiac hypertrophy (left ventricular septum thickness ⬎1.3 cm); ejection fraction ⬍50% and multiple IV inotropes at time of retrieval. Donors were divided into two groups. They were considered to be low risk when no or only one of the risk factors was present; otherwise, high-risk donors were those with more than one risk factor. Accordingly, they were divided in G1D or optimal donors (n ⫽ 156; score ⱕ 1) versus G2D suboptimals donors (n ⫽ 47, score ⬎ 1). Analyzing high- or low-risk recipients who received a high- or low-risk donor heart four matching we obtained groups (Fig 1). Group A (GA; n ⫽ 25) the highest risk possible, was composed of the worst recipients (G2R) and donors (G2D). Group B (GB) and group C (GC) included intermediate-risk subjects, either for donor (GB) or recipient characteristics (GC). The first combination of lower-risk recipient (G1R) with higher-risk donor (G2D) included 22 patients. In contrast, the second group (n ⫽ 60) matched higher-risk recipients (G2R) with lower-risk donors (G1D). Finally, group D (GD n ⫽ 97) was the one at lower risk, including the best recipients (G1R) and donors (G1D). Tables 1 and 2 list the overall characteristics of our population divided into
the four groups. Statistical evaluation was performed using SPSS, Data Mining and Statistical Analysis Software (SPPS Inc, Chicago, Ill, USA) with the Kaplan-Meier method. Statistical significance of survival curves was assessed using the log-rank test, with a cutoff value P ⱕ .05.
RESULTS
Figure 2 shows the overall results of our analysis. The 30 day-mortality was 12% in GA, 9% in GB, 12% in GC, and, finally, 3% in GD. GD showed a statistically significant difference versus the other groups (P ⫽ .05). The most common causes of death were primary graft failure and multiple organ failure. Overall, the mean follow-up was 52 ⫾ 34 months, including 41 ⫾ 33 months among GA; 52 ⫾ 34 months for GB; 48 ⫾ 33 months for GC, and 41 ⫾ 33 months for GD. Survival rates at 1 year were 80%, 86%, 85%, and 94% for GA, GB, GC, and GD, respectively; at 3 years, they were 76%, 82%, 80%, and 90% for the same groups, respectively, there was no significant difference.
Table 1.
Patients (n) M/F (%) Age (yr), mean (median) ⫾ SD; range Sternotomy, n (%) Diabetes, n (%) Dialysis, n (%) Creatinine (mg/dL), mean (median) ⫾ SD; range Bilirubin (mg/dL), mean (median) ⫾ SD; range CV events, n (%) IABP, n (%) ECMO, n (%) VAD, n (%) Ventilator, n (%) Mean cumulative RR
GA
GB
GC
GD
25 88/12 60 (62) ⫾ 10; 14–67 14 (56%) 11 (44%) 1 (4%) 1.6 (1.5) ⫾ 0.6; 0–3.1
22 77/23 55 (57) ⫾ 10; 27–70 7 (32%) 0 0 1.5 (1.5) ⫾ 0.9; 0.4–3.3
60 85/15 58 (60) ⫾ 8; 29–69 29 (48%) 24 (25%) 3 (5%) 1.5 (1.3) ⫾ 1.0; 0–7.8
96 76/24 51 (54) ⫾ 8; 12–65 21 (22%) 0 0 1.2 (1.1) ⫾ 0.3; 0.3–1.9
1.3 (1.1) ⫾ 1.0; 0–3.5
0.1 (0.1) ⫾ 0.1; 0–0.25
1.5 (1.1) ⫾ 1.5; 0.5–9.8
1.2 (0.9) ⫾ 0.9; 0.3–5.5
2 (8%) 0 2 (8%) 3 (12%) 0 2.49
1 (5%) 0 0 0 0 1.36
5 (8%) 8 (13%) 3 (5%) 9 (15%) 5 (8%) 2.69
2 (2%) 0 0 0 0 1.25
CLINICAL RISK IN HEART TRANSPLANTATION
309 Table 2.
Patients (n) M/F (%) Age (yr), mean (median) ⫾ SD; range Age ⬎ 55 yr, n (%) Ischemia time (min), mean (median) ⫾ SD; range Ischemia time ⬎ 240 min, n (%) Inotropic support, n (%) None Low (⬍10 g/kg/min) High (any dosage ⫹ cathecol.) Serum sodium (mg/dL), mean (median) ⫾ SD; range Coronary artery disease, n (%) LVEF ⬍ 50%, n (%) Cardiac hypertrophy, n (%) Variables number, mean (median) ⫾ SD; range
GA
GB
GC
25 79/21 51 (55) ⫾ 13; 18–67
22 82/18 42 (43) ⫾ 17; 14–66
60 60/40 37 (38) ⫾ 13; 17–63
96 68/32 35 (34) ⫾ 14; 13–65
14 (56%) 225 (243) ⫾ 81; 83–395
10 (45%) 235 (258) ⫾ 74; 111–350
3 (5%) 188 (204) ⫾ 56; 87–314
0 183 (185) ⫾ 56; 62–298
15 (60%)
14 (64%)
7 (12%)
15 (16%)
3 (12%) 7 (28%) 15 (60%) 147 (146) ⫾ 10; 132–173
0 4 (18%) 18 (82%) 150 (151) ⫾ 8; 136–170
12 (20%) 30 (50%) 18 (30%) 149 (150) ⫾ 7; 132–161
19 (20%) 43 (45%) 34 (35%) 145 (145) ⫾ 8; 117–163
5 (20%) 3 (12%) 4 (16%) 3.2 (3.0) ⫾ 0.8; 2–5
3 (14%) 1 (5%) 5 (23%) 3.3 (3.0) ⫾ 0.6; 3–5
DISCUSSIONS
Multiple variables influence outcomes after heart transplantation; their association could greatly affect overall results. Our study, even considering the not so large patient cohort (above all in the higher-risk groups) seemed to demonstrate that recipient are more important than donor characteristics, with the best results obtained in the GD lowest-risk group. However, also in cases of both high-risk recipients and donors, early results seemed to be satisfactory. Furthermore, donor characteristics seemed to more important by impact early survival, while recipient ones, long-term results. Obviously, larger cohorts of patients and longer follow-ups are needed to validate these results.
Fig 2.
1 (2%) 0 3 (5%) 1.4 (1.0) ⫾ 0.7; 0–2
GD
2 (2%) 0 10 (10%) 1.5 (2.0) ⫾ 0.7; 0–2
Some ethical issues arise from our analysis in cases of elevated clinical risk. The first one regards which outcome must be considered acceptable: is it preferable for heart transplant population survival; that is, to choose the best recipient on the waiting list to match with the best available donor. Or, as our analysis seems to suggest (at least for early/midterm outcomes), giving a chance of survival to all patients affected by heart failure by performing a heart transplantation? And more, in the era of extended donor and recipient criteria, which patient should be transplanted with a nonoptimal organ? We think there is no unequivocal answer to these complex medical and ethical issues, because what is medically feasible is not always appropriate. Therefore, these open
Fig 3.
310
GUZZI, MAIANI, ARESU ET AL
questions could urge a larger discussion in the scientific community about what is better to do. Our modest contribution to this topic is that, although the development of a mathematically specific score for heart transplantation is desirable and would be useful to best guide the choices about who undergoes transplantation, our policy and experience suggest that a final decision cannot leave aside an individual and case-bycase evaluation, so to offer the best opportunity to a specific patient, inspired by the principle that we cannot add days to life, but add life to any single day. Figs. 3–5 REFERENCES
Fig 4.
Fig 5.
1. Parsonnet V, et al: A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 79:I3, 1989 2. Roques F, Nashef SAM, Michel P, et al: Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients. Eur J Cardiothorac Surg 15:816, 1999 3. STS Score. Available at http://209.220.160.181/STSWebRiskCalc261/de.aspx 4. Higgins TL, Estafanous FG, Loop FD, et al: Stratification of morbidity and mortality outcome by preoperative risk factors in coronary artery bypass patients. A clinical severity score. JAMA 267:2344, 1992 5. Bourge RC, Naftel DC, Costanzo-Nordin MR, et al: Pretransplantation risk factors for death after heart transplantation: a multiinstitutional study. The Transplant Cardiologists Research Database Group. J Heart Lung Transplant 12:549, 1993 6. Young JB, Naftel DC, Bourge RC, et al: Matching the heart donor and heart transplant recipient. Clues for successful expansion of the donor pool: a multivariable, multiinstitutional report. The Cardiac Transplant Research Database Group. J Heart Lung Transplant 13:353, 1994 (discussion 364) 7. Taylor DO, Edwards LB, Boucek MM, et al: Registry of the International Society for Heart and Lung Transplantation: Twentyfourth Official Adult Heart Transplant Report—2007. J Heart Lung Transplant 26:769, 2007