- Email: [email protected]
Eight-year results of cyclosporine-treated patients with cardiac transplants
J
THORAC CARDIOVASC SURG
1990;99:500-9
Eight-year results of cyclosporine-treated patients with cardiac transplants To elucidate tbe long-term effects of cyclosporine, we retrospecth'ely studied 310 comecutive patients who have undergone cardiac trllmlplantation at our institution since December 1980 and in whom immunosuppression bas been maintained with cyclosporine. The ages of recipients ranged from 1 month to 64 years and of donors from 1 month to 48 years. The actuarial survival rates for cyclosporine-treated patients were 80.7% at 1 year and 59.7% at 5 years and were significantly greater than those for previous patients not treated with cyclosporine (p < 0.005). Their actuarial prevalence of rejection was 60.0 % at 1 month and 86.9 % at 1 year; 206 patients are living. The actuarial prevalence of lymphoma development was 4.6 % at 5 years but bas been significantly lower with tbe current immunosuppression protocol of lower doses of cyclosporine, and OKT3 in place of rabbit anti-thymocyte globulin (p < 0.005). Infection remains the most common cause of death. Recipients less than 50 years of age had a significantly higher actuarial survival than older recipients (p < 0.01). Male and female recipients had similar overall prevalence of survival and rejection, but men died of graft atherosclerosis significantly more frequently (p < 0.005). Rehabilitation bas been successful in 85 % of patients surviving 1 year after trllmlplantation. Of those surviving 1 year, 96.5 % were in New York Heart Association class I. Thus the results of orthotopic cardiac trllmlplantation have improved since tbe introduction of cyclosporine and have allowed measured liberalization of the criteria for recipient selection.
Mark T. Grattan, MD, Carlos E. Moreno-Cabral, MD, Vaughn A. Starnes, MD, Phillip E. Oyer, MD, PhD, Edward B. Stinson, MD, and Norman E. Shumway, MD, PhD, Stanford. Calif.
Cardiac transplantation is now a well-accepted therapy for end-stage heart disease. Nevertheless, the indications for transplantation and the criteria for recipient and donor selection continue to evolve. In December 1980 the Stanford University Department of Cardiovascular Surgery embarked on a clinical trial with cyclosporine as an integral part of immunosuppressive therapy for recipients of cardiac allografts. Previous studies have documented the efficacy of cyclosporine for cardiac transplantation in the experimental setting. The purpose of this study was to
From the Department of Cardiovascular Surgery, Stanford University Medical Center. Stanford, Calif. Presented in part at the meeting of the American Heart Association, New Orleans, La., Nov. 13-16, 1989. Received for publication March 8,1989. Accepted for publication July 20, 1989. Address for reprints: Mark T. Grattan, MD, 8010 Frost St., Suite 501, San Diego, CA 92123.
12/1/15519
500
analyze our long-term results of cardiac transplantation since the introduction of cyclosporine. This analysis has allowed us to modify our therapy when indicated and has provided a benchmark for judging future results. Patients and methods Patient selection. The cyclosporine group was composed of 310 patients who underwent orthotopic cardiac transplantation after the introduction of cyclosporine as a component of the immunosuppression regimen at Stanford University Medical Center in December 1980. The cyclosporine group was further subdivided for purposes of analysis into subgroups defined by the pretransplant diagnosis, age, and sex of the recipient and by age of the donor. These subgroups are described in detail in the Results section. Matching of donors and recipients was done to ensure compatibility of body weight (taking into account calculated pulmonary vascular resistance of the recipient) and matching ABO blood groups. In recipients with a high degree of positive crossmatches with a random panel of antigens, a negative lymphocyte crossmatch with the potential donor was performed before transplantation. No prospective matching on the basis of human leukocyte antigen phenotypes was performed.
Volume 99 Number 3 March 1990
Ninety-nine patients underwent orthotopic cardiac transplantation at Stanford from April 29, 1976, to December 1980. Cyclosporine was not in clinical use at Stanford during that period. For purposes of comparison in some of the subsequent analyses, this group is referred to as the precyclosporine group. Patients who received cardiac allografts before April 29, 1976, were excluded from analysis because of the large time difference and marked differences in immunosuppressive regimens. In addition, patients who received combined heart-lung transplants were excluded from all analyses in this study. Immunosuppression. The treatment protocols in effect when these patients received transplants varied over the period of the analysis. Patients were treated with different doses of cyclosporine, azathioprine, prednisone, anti-thymocyte globulin (ATG) of rabbit or horse origin, and OKT3. In some cases the patients did not receive one of the medications. Medications were withheld for reasons of impaired kidney function (cyclosporine), impaired liver function (azathioprine), leukopenia (azathioprine), or time of clinical availability (OKT3 and ATG). Induction cyclosporine doses were much higher in the early 19805 than presently administered doses (18 rug/kg/day in 1980 versus 4 to 8 rng/kg/day). Endomyocardial biopsy specimens were obtained to monitor graft rejection. Early postoperative acute rejection episodes classified as moderate or severe' were treated with methylprednisolone, 15 rug/kg/day, intravenously for 3 days, occasionally combined with ATG or with OKT3. Moderate or severe rejection appearing later in the patient's course was usually treated initially with high-dose oral prednisone (1.5 mg/kg/day) for 3 days, with rapid tapering to maintenance doses. Episodes that were not treated (mild acute rejection, demonstrating a perivascular infiltrate without myocyte necrosis) were not counted as rejection episodes for the purposes of this study. Infection surveillance. Potential recipients and donors were screened for the presence of anti-cytomegalovirus (CMV) immunoglobulin M and immunoglobulin G antibodies, and for serum antitoxoplasma titers. More recently, potential recipients and donors have been screened for the presence of hepatitis surface antigen and for anti-human immunodeficiency virus antibodies. After transplantation the inpatient recipient was monitored weekly for titers of anti-CMV antibodies. Surveillance cultures of the throat, buffy coat, and urine also were obtained each week. Outpatients had these studies performed weekly for the first 12 weeks, monthly for 6 months, and then yearly. Additional specimens were sent for special studies for Legionella, protozoa, viruses, fungi, or Nocardia organisms. as warranted clinically. Outpatient follow-up studies. At each outpatient clinic visit the following information was recorded: I. Hypertension---severe enough to warrant therapy (not diuretics only) 2. Rehabilitation-defined as the return to employment or. if retired, to an activity or pursuit that brought the patient significant daily satisfaction 3. New York Heart Association (NYHA) classification-at each annual visit, classification of each patient as class I, II, III, or IV based on symptoms 4. Immunosuppression-recording of doses of cyclosporine, azathioprine, and prednisone at each visit 5. Renal function over time-annual recording of serum creatinine, blood urea nitrogen, and creatinine clearance 6. Neoplasms-routine screening for neoplasms at each postoperative visit
Cyclosporine and cardiac transplants 5 0 1
Table I. Cause of braindeath in donors (1976 to 1988) Head trauma Vehicular Gunshotwound Other Cerebrovascular accident Anoxia Other Total
On site
Distant
117 26 32 45
164 57 38 43
9
10
5 234
4 316
For purposes of comparison, our earliest cyclosporine-treated subgroup (which received prophylactic rabbit ATG, cyclosporine at 18 rng/kg/day initially, and prednisone) was contrasted with our most recent treatment subgroup (which receives prophylactic OKT3, cyclosporine at 4 to 8 rug/kg/day initially, prednisone, and azathioprine). This was done to determine if the rates of development of lymphoma, hypertension. and changes in renal function were indeed improving with our current immunosuppression protocol. Cost. Hospitalization costs were determined and are expressed in terms of 1985 dollars. In addition, the mean number of transvenous endomyocardial biopsies incurred in the hospital and the mean length of stay in the hospital were tabulated. For purposes of comparison the first 60 patients in the series were compared with the latest 60 patients to determine what trends, if any, existed. Angiographic analysis. Routine coronary arteriography was done at yearly intervals to monitor for graft coronary atherosclerosis. The coronary arteries were evaluated by one experienced arteriographer. Arterial lesions were classified as mild «30% luminal stenosis), moderate (31% to 69% luminal stenosis), and severe (>69% stenosis). Stenosis in anyone of the three major coronary arteries or their major branches was considered in this classification system. Autopsy data. Autopsy data were available on all but 18 patients in the cyclosporine group. Estimation of the degree of obstruction in each of the major coronary arteries was made by comparison with a segment of the patient's own coronary artery. The patients so studied were then grouped according to those with one or more major coronary arteries with 50% or greater obstruction and those with no such obstruction. Statistical analysis. Time zero in all instances refers to the time of transplantation. Actuarial data were compiled with the Cutler-Ederer method.I and statistical significance between groups was assessed by the Cox-Mantel test'; p values less than 0.05 were considered significant. Absolute data with regard to numbers of infection episodes per unit of time ("linear" or "crude" rates) were quantified in monthly intervals as episodes per 100 patient-days to account for the difference in numbers of patients in the two groups. Comparisons were then performed between groups in each monthly interval by use of a z statistic for rates and proportions. Differences between patient characteristics (Table I) were analyzed by the unpaired Student t tests for parametric data and contingency table analysis for nonparametric, nominal data; p values less than 0.05 were considered significant."
The Journal of
5 0 2 Grattan et al.
Thoracic and Cardiovascular Surgery
120
-r-----------------------,
100
80
Number of Recipients at Age
60
40
20
o 0-9
10-19
20-29
30-39
40-49
50-59
60-69
Recipient Age (Years) Fig. I. Cardiac transplant recipients ranged in age from 3 weeks to 64 years. The greatest number of patients were ages 40 to 49 years at the time of transplantation.
10
-r-----------------------,
8
6
Drug Doses (mg/kg/day)
Cyclosporine 4
Azathioprine
2
Prednisone
o~~:$=tc:$::==*===*===*===:$===~ 24 60 72 84 a 12 36 48 Postoperative Month Fig. 2. Doses over time of the three primary long-term immunosuppressants are shown. Cyclosporine doses have been decreased to limit the development of nephrotoxicity and possibly lymphoma.
Results Pretransplant status of cyclosporine-treated recipients. Three hundred ten patients have undergone orthotopie cardiac transplantation at Stanford University Medical Center from December 1980 to the present. Fig.
1 shows the age distribution of the recipients at the time of transplantation. The largest number of patients were in the 40- to 49-year-old age range. There were 82 patients 50 years of age or older at the time of transplantation and 177 patients 20 to 49 years of age.
Volume 99 Number 3
Cyclosporine and cardiac transplants
March 1990
The most common diagnosis was idiopathic cardiomyopathy (present in 129 patients [41.6%1 before transplant). The second most common pretransplant diagnosis was coronary artery disease (present in 118 patients [38.1%]). The other pretransplant diagnoses were substantially less common: Viral cardiomyopathy occurred in 23 patients (7.4%), congenital heart disease in 10 patients (3.2%), peripartum cardiomyopathy in eight patients (2.6%), and miscellaneous diagnoses in 22 patients (9.1%). The mean waiting pericx:l from the time of acceptance of a patient as an active candidate and the actual transplantation date was 44.9 ± 57.5 (SO*) days. The range was 0 to 291 days. In our earlier cases, donors were transported to our institution where the donor cardiectomy was performed. Our present methcx:l of myocardial protection, however, has allowed us to perform distant procurement of the majority of donor organs. Thus over the last 2 years 88% of grafts have been obtained by distant procurement, and only 12% were obtained at our institution. The mean age of the donors of organs to cyclosporinetreated patients was 23 ± 7.8 (SO) years. The age range of donors' was 3 weeks to 48 years. Twenty donors were over the age of 35 years. One hundred eighty-one donors were between the ages of 19 and 35 years (see following comparative data). Causes of donor deaths. The causes of death of all donors evaluated since 1976 are shown in Table I. Vehicular accidents with head trauma remain the single most important cause of death of organ donors for our patient population. Follow-up data of recipients Hypertension. Data are available for 216 patients at 1 year of follow-up. At this time 157 (72.7%) had hypertension severe enough to warrant use of medications for blocx:l pressure control that were more potent than diuretics alone. Fifty-nine patients (27.3%) did not require antihypertensive medications at 1 year after transplantation. Hypertension occurred in 51.5% of patients in our most recent subgroup at 1 year (low-dose cyclosporine, prophylactic OKT3, azathioprine, and prednisone) and is significantly less frequent than in the earlier subgroups of cyclosporine-treated recipients (p < 0.05). Rehabilitation. At 1 year 181 of 213 patients (85%) who completed a questionnaire were completely rehabilitated, and 32 of the 213 patients (15%) were not. NYHA class. Data are available for 199 patients at 1 year after transplantation. One hundred ninety-two pa'SD = Standard deviation.
50 3
100..,..--------------,
80
60
Creatinine Clearance (ml/min)
40
20
O-+-....--..,....--.-~r--.--r_o__,---..__r--...--r--l
o
12
24
36
48
60
72
84
Postoperative Month
Fig. 3. Mean creatinine clearance declined in the first several postoperative months but then remained relatively stable over the ensuing follow-up period.
tients were in NYHA class I (96.5%), six patients were in class II (3.0%), none was in class III, and one patient was in NYHA class IV (0.5%). Immunosuppression. Data regarding the trends of dosing of the three long-term immunosuppressant medications for all patients are depicted in Fig. 2. (As described in the Patients and methcx:ls section, lower doses of cyclosporine are now being used from the time of transplantation.) Azathioprine dosing has remained essentially the same over the 8-year pericx:l. Biopsy specimens. The first 60 patients underwent an average of 5.8 ± 3.5 (SO) biopsy procedures in their initial hospitalization; the last 60 patients had an average of 2.7 ± 1.8 (SO) biopsy procedures. This change correlated with the change in the length of stay in the hospital. Length ofstay. The first 60 patients in this series stayed in the hospital an average of 41 ± 19.4 (SO) days after transplantation. The last 60 patients stayed an average of 23.9 ± 14.6 (SO) days. Renalfunction. Serum creatinine, blocx:l urea nitrogen, and creatinine clearance generally deteriorated during the initial 2 to 3 months after transplantation and then stabilized, as shown in Fig. 3. At 1 year the average creatinine clearance for all cyclosporine-treated patients was 64.9 ± 27.4 (SO) ml/rnin/rn. The earliest subgroup of cyclosporine-treated patients (cyclosporine, 18 rug/kg/day) had significantly greater impairment of renal function than had the patients with the current protocol (cyclosporine,4 to 8 rug/kg/day initially). At 1 year average creatinine clearance and serum creatinine for the high-dose cyclosporine group were 47.6 ± 18.8 (SO) ml/rnin/rn and z.Z ± 0.6(SO)mg/dl,
The Journal of Thoracic and Cardiovascular
5 0 4 Grattan et at.
Surgery
100
Nocardia Protozoa Fungi
80
60
Percent of Patients Free from Infection
Bacteria 40
Viruses 20
O+--....--r--,---r----,r----r--r--....-
o
234
5
6
7
8
Postoperative Year
Fig. 4. Infection with a variety of organisms that necessitated treatment was a problem for most patients. The actuarial prevalence of developing a treatable infection with the major classes of pathogens is shown in this figure, expressed as the percent of all patients free of the infection over time. Most of the viral infections were oral mucocutaneous Herpes simplex type 1 infections, but there were also many cytomegalovirus infections (91 patientsj.s
2.00
Rejection Episodes Per 100 Patient-Days 1.00
0.00 1
2
3
4
5
6
7
8
9 10 11 1218243036424854606672
Month Post-Transplant
Fig. S. The crude rejection prevalence expressed as rejection episodes per 100 patient-days in certain posttransplant months is shown in this figure (note nonlinear time scale, x axis). This figure demonstrates that the number of rejection episodes is much greater in the early posttransplant period than subsequently, raising the question of immunologic "tolerance" of the allograft.
respectively. Values at 1 year for the current, low-dose cyclosporine subgroup were 65.2 ± 22.3 (SD) mljrninj m and 1.5 ± 0.6 (SD) mgjdl, p < 0.05. Costs. The average cost of a cardiac transplantation (hospital charges) in 1980was $83,432 ± $31,721 (1980 dollars). By comparison, the average inhospital cost for our most recent 60 patients with transplants has been
$104,102 ± $89,592 (SD)(1985 dollars). When the rate of inflation over the intervening 5 years is considered, the inhospital charges for transplantation have not changed significantly. Infection rates. Infections severeenough to necessitate treatment were also tabulated and are shown in Fig. 4. The case attack rates of bacterial infections at 1 and 5
Volume 99
Cyclosporine and cardiac transplants 5 0 5
Number 3 March 1990
100
-~~f==;::=:;:=+=+=:::;=~=~1
80
Graft Atherosclerosis 60
Percent of Patients Free from Event 40
20 Rejection
O+--.,.---,----,---,---r---,.----,r--r-'
o
2
3
4
6
5
7
8
Postoperative Year
Fig. 6. Actuariallycompiled data regardingthe prevalence of development of lymphoma, graft atherosclerosis, and rejection are expressed as the percentof patients free of each complication, as a function of time after transplantation.
years were 39.9% (± 2.9%) and 50.9% (± 3.6%), respectively.For fungal infections the rates at 1 and 5 years were 12.9%(± 2.0%) and 19.1% (±2.9%). For viral infections these rates were 56.8% (± 3.0%) and 65.3% (± 3.2%). For protozoal infections the rates were 9.5% ( ± 1.8%)and 12.3% ( ± 2.2%). For nocardial infections the rates were 5.5% (± 1.4%) and 7.6% (± 2.0%) for 1 and 5 years, respectively. Older recipients (50 years of age or older) had a significantly higher rate of bacterial infections (p < 0.05), fungal infections (p < 0.05), and nocardial infections (p < 0.05) compared with younger recipients (aged 20 to 49 years). There were no differences in the rates of infection with viruses or protozoa. Rejection rates. The crude rejection rates (calculated in monthly intervals) for cyclosporine-treated patients at 1 month and 1 year, respectively, were 2.38 (±0.16) and 0.14 (± 0.04) events per I()() patient-days. These rates are depicted in Fig. 5. The actuarial rejection prevalence is shown in Fig. 6. At 1 month this prevalence for the cyclosporine group was 60.0% (± 2.8%). At 1 year the actuarial rejection prevalence for this group was 86.9% (±2.1%). Older recipients (2:50 years) had fewer rejection episodes (actuarial model) than younger recipients «50 years), but the difference was not significant (p < 0.10). There was no significant difference between the actuarial rejection rates of recipients who received grafts from older donors (donor age >35 years) and those who received grafts from younger donors (donor age 20 to 35 years). Similarly, male (n = 237) and female (n = 73)
recipients showed no significant difference in actuarial rejection rates. Finally, recipients treated with high-dose cyclosporine had no significant difference in actuarial rejection rates from the current recipients of low-dose cyclosporine. Graft atherosclerosis. Angiographic and autopsy data were used to determine actuarial graft atherosclerosis rates. At 1 year significant angiographic or autopsyproven lesions had developed in 6.2% ± 1.6% of cyclosporine-treated patients. By 5 years such lesions had developedin24% ± 3.9% (Fig. 6). There were no significant differences in the rates of development of graft atherosclerosis in recipients of grafts from older donors (>35 years) and recipients of grafts from younger donors (19 to 35 years). Similarly, angiographic or autopsy-proven graft atherosclerosis developed in patients in the coronary artery disease subgroup at a rate that was not significantly different from that in cyclosporine-treated patients who did not have pretransplant coronary artery disease. (For more comparative data see section on causes of death of recipients.) Neoplasms. Eleven patients had lymphoma during their postoperative course, an average of 245 days after transplantation (range, 67 to 853 days). As shown in Fig. 6, the actuarial risk of developing lymphoma at 1 and 5 years was 2.9% (± 1.0%, SEM*) and 4.6% (± 1.4%, SEM). Six of the 11 patients have died, three as a direct result of the lymphoma, two each of infection, and one of graft atherosclerosis. Twelve patients had other forms of *SEM
= Standard
error of the mean.
The Journal of
506 Grattan et al.
Thoracic and Cardiovascular Surgery
100
-r-------------------, ..........
'1-"
80
Cyclosporine-treated patients (n=310) Pre-cyclosporine patients (n=99)
---0--
"1
"."t -!-. <,
60
Percent of Patients Surviving
"f. l
40
···L.
20
o-t--,--r-..,..--,I"""""-.----.-.----r-.---r----,-..,....J o
2
3
4
5
6
7
8
9
10
11
12
Postoperative Year Fig. 7. Patients treated with cyc/osporine had a significantly higher survival rate than had patients not treated with cyclosporine (p < 0.005). 100 .......- - - - - - - - - - - - - - - - - - - - ,
80
Percent of Patients Surviving
Recipients 20-50 years old
60
40 Recipients> 50 years old 20
O+---,..---r----,--r--,.---.--.........----,----l
o
2
3
4
5
6
8
Postoperative Year Fig. 8. Cyclosporine-treated patients 50 years of age or older (n = 82) had a significantly lower survival rate than had cyclosporine-treated patients 20 to 49 years of age (n = 177, P < 0.01).
neoplasm: six had carcinoma of the skin, three had adenocarcinoma of the lung, and there was one case each of testicular carcinoma, hepatic carcinoma, and renal cell carcinoma. Patients in our current immunosuppression protocol have had a significantly lower actuarial rate of lymphoma development over 2 years 1.3% ± 1.2%, SEM) (p < 0.005) than the earliest subgroup of cyclosporinetreated recipients. Survival rates: Cyclosporine versus precyclosporine groups. The actuarial survival rate for the entire cyclo-
sporine group of patients was significantly better (p < 0.005) than that of the precyclosporine group of patients (Fig. 7). The rates for the cyclosporine group at 1 and 5 years after transplantation were 80.7% (± 2.28%) and 59.7% (± 3.5%). For the precyclosporine group of patients these rates were 62.6% (±4.9%) and 36.4% (±4.8%), respectively, at 1 and 5 years. Survival rates: Cyclosporine-treated subgroups. The actuarial survival rates for two major subgroups of the cyclosporine-treated patients (those with idiopathic cardiomyopathy and coronary artery disease) were also an-
Volume 99
Cyclosporine and cardiac transplants
Number 3
50 7
March 1990
Pulmonary Embolus CVA Pulmonary Hypertension •
Non-Lymphoid CA
r:a
Deaths in Patients 20-50 Years Old Deaths in Patients over 50 Years Old
Lymphoid CA Graft Failure Other Graft Athero Rejection Infection
o
10
20
30
40
Numbers of Deaths
Fig. 9. Causes of death for all adult patients are shown; with subdivisions in each category for deaths in patients 20 to 49 years old and in patients 50 years or older.
alyzed. At I and 5 years these rates were 79.4% ( ± 3.7%) and 55.4% (± 5.6%) for the coronary artery disease group. The rates were 82.1% (±3.5%) and 59.4% ( ± 5.4%) for the idiopathic cardiomyopathy group. These rates were not significantly different from one another or from the cyclosporine group as a whole. Survival data for adult patients with transplants who were treated with cyclosporine were also analyzed according to the age of the patient at transplantation. The older group consisted of cyclosporine-treated patients who underwent transplantation when they were 50 years of age or older (n = 82). The younger group included all cyclosporine-treated patients between the ages of 20 and 50 years at the time of transplantation (n = 177). As shown in Fig. 8, the older group had significantly lower survival rates (p < 0.01). Thus at I and 5 years after transplantation the survival rates for the older group were 76.2% (± 4.9%) and 43.8% (± 7.9%), whereas for the younger group these rates were 82.6% (± 2.8%) and 62.0% (±4.3%). Survival rates for male and female cyclosporine-treated recipients were not significantly different. Similarly, there were no significant differences in actuarial survival rates between recipients of grafts from older donors (>35 years) and recipients of grafts from younger adult donors (20 to 35 years). Causes of death of recipients (cyclosporine-treated group). Of 310 patients from the cyclosporine group 206 are alive. The proximate cause of death in the 104 patients who have died was infection in 35.6%, rejection in 16.3%, graft atherosclerosis in 14.4%, nonspecific graft failure in 5.8%, lymphoid malignancy in 4.8%, nonlymphoid ma-
lignancy in 4.8%, pulmonary hypertension in 3.8%, cerebrovascular disease in 3.8%, pulmonary embolus in 1.9%, and other causes in 8.6% (Fig. 9). Death rates from rejection. Death rates from rejection for all cyclosporine-treated patients were 4.9% (± 1.3%) and 7.0% (± 1.8%) at I and 5 years after transplantation, respectively. Female recipients died of rejection more frequently than did men, but the difference was not significant (p < 0.10). N either age of the recipients nor age of the donors significantly affected death rates from rejection. Death rates from infection. Death rates from infection for all cyclosporine-treated patients were 8.4% (± 1.6%) and 15.1% ( ± 2.7%) at I and 5 years after transplantation, respectively. Female recipients died of infection more frequently than did male recipients, but this difference was not significant (p < 0.10). Neither age of the recipient nor age of the donor significantly affected actuarial death rates from infection. Death rates from graft atherosclerosis. Death rates from graft atherosclerosis for all cyclosporine-treated patients were 0.4% (± 0.4%) and 9.8% ( ± 2.8%) at I and 5 years after transplantation, respectively. Male recipients died of graft atherosclerosis significantly more frequently than did female recipients (p < 0.05). Similarly, recipients of grafts from older donors (> 35 years) of graft atherosclerosis died more frequently than did recipients of grafts from younger donors (20 to 35 years), but the difference was not statistically significant (p < 0.10). Older recipients (>50 years), however, did not have a different rate of death from graft atherosclerosis than younger adult recipients (aged 20 to 49 years).
The Journal of Thoracic and Cardiovascular
5 0 8 Grattan et al.
Finally, there were no significant differences in rates of death from graft atherosclerosis between recipients in the coronary artery disease subgroup before transplantation and those who did not have coronary artery disease as the reason for transplantation. Discussion This article summarizes the 8-year experience at Stanford University Medical Center with cyclosporine in the management of the immunosuppression of our patients.' After its introduction the survival rates at our institution significantly improved. As a result of these improved survival rates, cardiac transplantation has been offered as standard therapy for appropriate patients with end-stage cardiac dysfunction. In 1987, 109 centers in the United States performed 1436 cardiac transplantations. In the early years of our experience we restricted the age of participants so that neither recipients over the age of 50 years nor donors over the age of 35 years were considered. As seen in Fig. 1, however, we have substantially liberalized our age criteria for recipients. We also have implanted grafts from donors as old as 48 years of age with no discernible difference in recipient survival, rejection, or graft atherosclerosis. Therefore we have been able to offer the therapeutic option of orthotopic cardiac transplantation to a larger group of patients than in the earlier years of our experience. Despite these liberalized criteria, the survival rates of recipients have actually improved. Several additional points merit further emphasis. Survival rates. In this study the survival rate of recipients ages 20 to 49 years was significantly better than that for patients 50 years of age and older. This contrasts with the published reports of other transplant groups, but these studies have followed smaller groups of patients for much shorter periods of time (1 to 2 years).6. 7 The reasons for lower survival rates in older recipients may be explained by greater susceptibility in the older patients to infections with bacterial, fungal, and nocardial organisms, although the actuarially determined rates of death by infection for these two groups do not differ significantly. Rejection. Rejection episodes occurred in 60% of patients by 1 month despite their suppressed immune status. It remained the second most common cause of death in our patients (16%). The difference between this experience and that of the Minnesota group, which reported only two rejection episodes in 57 consecutive patients over 2 years, is difficult to explain without a careful review of the exact pathologic criteria used to diagnose rejection
episodes.f It is also of some interest that there was no difference between the actuarial rates of rejection in the early cyclosporine protocol with high-dose initial cyclosporine
Surgery
(18 rug/kg/day) and prophylactic rabbit ATG and in the later protocol of lower initial doses of cyclosporine (8 mg/kg/day) and prophylactic OKT3. Hypertension and renal dysfunction. Hypertension that was severe enough to require medications more potent that diuretics occurred in 73% of patients overall but in only 52% of patients who were treated with the current low-dose cyclosporine protocol. Renal function was deleteriously affected by immunosuppressants (principally cyclosporine) in all patients; however, the early recipients of the high-dose cyclosporine protocol had significantly more impairment of renal function than had the recipients of current low-dose cyclosporine protocol. Thus overall trends in renal function for all patients may obscure these deleterious effects of the higher doses of cyclosporine in the early subgroup of recipients. Graft atherosclerosis. It is of interest that the pretransplant diagnosis of coronary artery disease did not predispose patients to the development of graft atherosclerosis. Recipients of grafts from older organ donors (>35 years) died of graft atherosclerosis more frequently, but this trend was not statistically significant (p < 0.10). One risk factor for graft atherosclerosis that we have definitivelydetermined is postoperative CMV infection in a cardiac transplant recipient. 8 A recent review of clinical and laboratory variables did not identify any other striking correlates for the development of graft athero-
sclerosis." Neoplasms. A concern arising from our early experience with cyclosporine (when we were using doses approximately twice as high) was the development of B-cell lymphomas.'? The cyclosporine doses were reduced in later patients in an effort to curtail the development of lymphoma and to minimize nephrotoxicity. Indeed, our most recent subgroup of patients (treated with prophylactic OKT3, initial cyclosporine doses of 4 to 8 mg/kg/ day, prednisone, and azathioprine) has had a significantly less frequent development of lymphoma over the first 2 years that this protocol has been in use. Because most lymphomas have developed early in our patients (usually within the first year after transplant), there is reason to believe that the problem of lymphoma development will occur less frequently as a complication, even with longer follow-up. Rehabilitation. One of the most gratifying findings of this study is that surviving patients have generally enjoyed a good quality of life~ %5% have returned to employment or an activity of choice. Indeed, 96.5% of patients were classified in NYHA functional class I at I or more years after transplantation. Thus not only are most cardiac transplant recipients living longer, but equally important, they have achieved genuine rehabilitation.
Volume 99 Number 3
Cyc/osporine and cardiac transplants 5 0 9
March 1990
REFERENCES J. Billingham ME. The postsurgical heart: the pathology of cardiac transplantation. Am J Cardiovasc Pathol 1988; 1:319-34. 2. Cutler SJ, Ederer F. Maximum utilization of the life table in analyzing survival. J Chronic Dis 1958;8:699-712. 3. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: John Wiley & Sons, 1980. 4. Glantz SA. Primer of biostatistics. 2nd ed. New York: MeGraw-Hili Co, 1987. 5. Reitz BA, Bieber CP, Raney AA, et al. Orthotopic heart and combined heart and lung transplantation with cyclosporine-A immune suppression. Transplant Proc 1981; 13:393-6. 6. Miller LW, Vitale-Noedel N, Pennington DG, McBride L, Kanter KR. Heart transplantation in patients over age fifty-five years. J Heart Transplant 1988;7:254-7.
7. Olivari MT, Antolick A, Kaye MP, Jamieson SW, Ring WS. Heart transplantation in elderly patients. J Heart Transplant 1988;7:258-64. 8. Grattan MT, Moreno-Cabral CE, Starnes VA, Oyer PE, Stinson EB, Shumway NE. Cytomegalovirus is associated with cardiac allograft atherosclerosis and death. JAMA (In press.) 9. Gao SZ, Schroeder JS, Alderman EL, et al. Clinical and laboratory correlates of accelerated coronary artery disease in the cardiac transplant patient. Circulation 1987;76:V5661.
10. Brumbaugh J, Baldwin JC, Stinson EB, et al. Quantitative analysis of immunosuppression in cyclosporine-treated heart transplant patients with lymphoma. Heart Transplant 1985;4:307-1 J.
THORAC CARDIOVASC SURG
1990;99:500-9
Eight-year results of cyclosporine-treated patients with cardiac transplants To elucidate tbe long-term effects of cyclosporine, we retrospecth'ely studied 310 comecutive patients who have undergone cardiac trllmlplantation at our institution since December 1980 and in whom immunosuppression bas been maintained with cyclosporine. The ages of recipients ranged from 1 month to 64 years and of donors from 1 month to 48 years. The actuarial survival rates for cyclosporine-treated patients were 80.7% at 1 year and 59.7% at 5 years and were significantly greater than those for previous patients not treated with cyclosporine (p < 0.005). Their actuarial prevalence of rejection was 60.0 % at 1 month and 86.9 % at 1 year; 206 patients are living. The actuarial prevalence of lymphoma development was 4.6 % at 5 years but bas been significantly lower with tbe current immunosuppression protocol of lower doses of cyclosporine, and OKT3 in place of rabbit anti-thymocyte globulin (p < 0.005). Infection remains the most common cause of death. Recipients less than 50 years of age had a significantly higher actuarial survival than older recipients (p < 0.01). Male and female recipients had similar overall prevalence of survival and rejection, but men died of graft atherosclerosis significantly more frequently (p < 0.005). Rehabilitation bas been successful in 85 % of patients surviving 1 year after trllmlplantation. Of those surviving 1 year, 96.5 % were in New York Heart Association class I. Thus the results of orthotopic cardiac trllmlplantation have improved since tbe introduction of cyclosporine and have allowed measured liberalization of the criteria for recipient selection.
Mark T. Grattan, MD, Carlos E. Moreno-Cabral, MD, Vaughn A. Starnes, MD, Phillip E. Oyer, MD, PhD, Edward B. Stinson, MD, and Norman E. Shumway, MD, PhD, Stanford. Calif.
Cardiac transplantation is now a well-accepted therapy for end-stage heart disease. Nevertheless, the indications for transplantation and the criteria for recipient and donor selection continue to evolve. In December 1980 the Stanford University Department of Cardiovascular Surgery embarked on a clinical trial with cyclosporine as an integral part of immunosuppressive therapy for recipients of cardiac allografts. Previous studies have documented the efficacy of cyclosporine for cardiac transplantation in the experimental setting. The purpose of this study was to
From the Department of Cardiovascular Surgery, Stanford University Medical Center. Stanford, Calif. Presented in part at the meeting of the American Heart Association, New Orleans, La., Nov. 13-16, 1989. Received for publication March 8,1989. Accepted for publication July 20, 1989. Address for reprints: Mark T. Grattan, MD, 8010 Frost St., Suite 501, San Diego, CA 92123.
12/1/15519
500
analyze our long-term results of cardiac transplantation since the introduction of cyclosporine. This analysis has allowed us to modify our therapy when indicated and has provided a benchmark for judging future results. Patients and methods Patient selection. The cyclosporine group was composed of 310 patients who underwent orthotopic cardiac transplantation after the introduction of cyclosporine as a component of the immunosuppression regimen at Stanford University Medical Center in December 1980. The cyclosporine group was further subdivided for purposes of analysis into subgroups defined by the pretransplant diagnosis, age, and sex of the recipient and by age of the donor. These subgroups are described in detail in the Results section. Matching of donors and recipients was done to ensure compatibility of body weight (taking into account calculated pulmonary vascular resistance of the recipient) and matching ABO blood groups. In recipients with a high degree of positive crossmatches with a random panel of antigens, a negative lymphocyte crossmatch with the potential donor was performed before transplantation. No prospective matching on the basis of human leukocyte antigen phenotypes was performed.
Volume 99 Number 3 March 1990
Ninety-nine patients underwent orthotopic cardiac transplantation at Stanford from April 29, 1976, to December 1980. Cyclosporine was not in clinical use at Stanford during that period. For purposes of comparison in some of the subsequent analyses, this group is referred to as the precyclosporine group. Patients who received cardiac allografts before April 29, 1976, were excluded from analysis because of the large time difference and marked differences in immunosuppressive regimens. In addition, patients who received combined heart-lung transplants were excluded from all analyses in this study. Immunosuppression. The treatment protocols in effect when these patients received transplants varied over the period of the analysis. Patients were treated with different doses of cyclosporine, azathioprine, prednisone, anti-thymocyte globulin (ATG) of rabbit or horse origin, and OKT3. In some cases the patients did not receive one of the medications. Medications were withheld for reasons of impaired kidney function (cyclosporine), impaired liver function (azathioprine), leukopenia (azathioprine), or time of clinical availability (OKT3 and ATG). Induction cyclosporine doses were much higher in the early 19805 than presently administered doses (18 rug/kg/day in 1980 versus 4 to 8 rng/kg/day). Endomyocardial biopsy specimens were obtained to monitor graft rejection. Early postoperative acute rejection episodes classified as moderate or severe' were treated with methylprednisolone, 15 rug/kg/day, intravenously for 3 days, occasionally combined with ATG or with OKT3. Moderate or severe rejection appearing later in the patient's course was usually treated initially with high-dose oral prednisone (1.5 mg/kg/day) for 3 days, with rapid tapering to maintenance doses. Episodes that were not treated (mild acute rejection, demonstrating a perivascular infiltrate without myocyte necrosis) were not counted as rejection episodes for the purposes of this study. Infection surveillance. Potential recipients and donors were screened for the presence of anti-cytomegalovirus (CMV) immunoglobulin M and immunoglobulin G antibodies, and for serum antitoxoplasma titers. More recently, potential recipients and donors have been screened for the presence of hepatitis surface antigen and for anti-human immunodeficiency virus antibodies. After transplantation the inpatient recipient was monitored weekly for titers of anti-CMV antibodies. Surveillance cultures of the throat, buffy coat, and urine also were obtained each week. Outpatients had these studies performed weekly for the first 12 weeks, monthly for 6 months, and then yearly. Additional specimens were sent for special studies for Legionella, protozoa, viruses, fungi, or Nocardia organisms. as warranted clinically. Outpatient follow-up studies. At each outpatient clinic visit the following information was recorded: I. Hypertension---severe enough to warrant therapy (not diuretics only) 2. Rehabilitation-defined as the return to employment or. if retired, to an activity or pursuit that brought the patient significant daily satisfaction 3. New York Heart Association (NYHA) classification-at each annual visit, classification of each patient as class I, II, III, or IV based on symptoms 4. Immunosuppression-recording of doses of cyclosporine, azathioprine, and prednisone at each visit 5. Renal function over time-annual recording of serum creatinine, blood urea nitrogen, and creatinine clearance 6. Neoplasms-routine screening for neoplasms at each postoperative visit
Cyclosporine and cardiac transplants 5 0 1
Table I. Cause of braindeath in donors (1976 to 1988) Head trauma Vehicular Gunshotwound Other Cerebrovascular accident Anoxia Other Total
On site
Distant
117 26 32 45
164 57 38 43
9
10
5 234
4 316
For purposes of comparison, our earliest cyclosporine-treated subgroup (which received prophylactic rabbit ATG, cyclosporine at 18 rng/kg/day initially, and prednisone) was contrasted with our most recent treatment subgroup (which receives prophylactic OKT3, cyclosporine at 4 to 8 rug/kg/day initially, prednisone, and azathioprine). This was done to determine if the rates of development of lymphoma, hypertension. and changes in renal function were indeed improving with our current immunosuppression protocol. Cost. Hospitalization costs were determined and are expressed in terms of 1985 dollars. In addition, the mean number of transvenous endomyocardial biopsies incurred in the hospital and the mean length of stay in the hospital were tabulated. For purposes of comparison the first 60 patients in the series were compared with the latest 60 patients to determine what trends, if any, existed. Angiographic analysis. Routine coronary arteriography was done at yearly intervals to monitor for graft coronary atherosclerosis. The coronary arteries were evaluated by one experienced arteriographer. Arterial lesions were classified as mild «30% luminal stenosis), moderate (31% to 69% luminal stenosis), and severe (>69% stenosis). Stenosis in anyone of the three major coronary arteries or their major branches was considered in this classification system. Autopsy data. Autopsy data were available on all but 18 patients in the cyclosporine group. Estimation of the degree of obstruction in each of the major coronary arteries was made by comparison with a segment of the patient's own coronary artery. The patients so studied were then grouped according to those with one or more major coronary arteries with 50% or greater obstruction and those with no such obstruction. Statistical analysis. Time zero in all instances refers to the time of transplantation. Actuarial data were compiled with the Cutler-Ederer method.I and statistical significance between groups was assessed by the Cox-Mantel test'; p values less than 0.05 were considered significant. Absolute data with regard to numbers of infection episodes per unit of time ("linear" or "crude" rates) were quantified in monthly intervals as episodes per 100 patient-days to account for the difference in numbers of patients in the two groups. Comparisons were then performed between groups in each monthly interval by use of a z statistic for rates and proportions. Differences between patient characteristics (Table I) were analyzed by the unpaired Student t tests for parametric data and contingency table analysis for nonparametric, nominal data; p values less than 0.05 were considered significant."
The Journal of
5 0 2 Grattan et al.
Thoracic and Cardiovascular Surgery
120
-r-----------------------,
100
80
Number of Recipients at Age
60
40
20
o 0-9
10-19
20-29
30-39
40-49
50-59
60-69
Recipient Age (Years) Fig. I. Cardiac transplant recipients ranged in age from 3 weeks to 64 years. The greatest number of patients were ages 40 to 49 years at the time of transplantation.
10
-r-----------------------,
8
6
Drug Doses (mg/kg/day)
Cyclosporine 4
Azathioprine
2
Prednisone
o~~:$=tc:$::==*===*===*===:$===~ 24 60 72 84 a 12 36 48 Postoperative Month Fig. 2. Doses over time of the three primary long-term immunosuppressants are shown. Cyclosporine doses have been decreased to limit the development of nephrotoxicity and possibly lymphoma.
Results Pretransplant status of cyclosporine-treated recipients. Three hundred ten patients have undergone orthotopie cardiac transplantation at Stanford University Medical Center from December 1980 to the present. Fig.
1 shows the age distribution of the recipients at the time of transplantation. The largest number of patients were in the 40- to 49-year-old age range. There were 82 patients 50 years of age or older at the time of transplantation and 177 patients 20 to 49 years of age.
Volume 99 Number 3
Cyclosporine and cardiac transplants
March 1990
The most common diagnosis was idiopathic cardiomyopathy (present in 129 patients [41.6%1 before transplant). The second most common pretransplant diagnosis was coronary artery disease (present in 118 patients [38.1%]). The other pretransplant diagnoses were substantially less common: Viral cardiomyopathy occurred in 23 patients (7.4%), congenital heart disease in 10 patients (3.2%), peripartum cardiomyopathy in eight patients (2.6%), and miscellaneous diagnoses in 22 patients (9.1%). The mean waiting pericx:l from the time of acceptance of a patient as an active candidate and the actual transplantation date was 44.9 ± 57.5 (SO*) days. The range was 0 to 291 days. In our earlier cases, donors were transported to our institution where the donor cardiectomy was performed. Our present methcx:l of myocardial protection, however, has allowed us to perform distant procurement of the majority of donor organs. Thus over the last 2 years 88% of grafts have been obtained by distant procurement, and only 12% were obtained at our institution. The mean age of the donors of organs to cyclosporinetreated patients was 23 ± 7.8 (SO) years. The age range of donors' was 3 weeks to 48 years. Twenty donors were over the age of 35 years. One hundred eighty-one donors were between the ages of 19 and 35 years (see following comparative data). Causes of donor deaths. The causes of death of all donors evaluated since 1976 are shown in Table I. Vehicular accidents with head trauma remain the single most important cause of death of organ donors for our patient population. Follow-up data of recipients Hypertension. Data are available for 216 patients at 1 year of follow-up. At this time 157 (72.7%) had hypertension severe enough to warrant use of medications for blocx:l pressure control that were more potent than diuretics alone. Fifty-nine patients (27.3%) did not require antihypertensive medications at 1 year after transplantation. Hypertension occurred in 51.5% of patients in our most recent subgroup at 1 year (low-dose cyclosporine, prophylactic OKT3, azathioprine, and prednisone) and is significantly less frequent than in the earlier subgroups of cyclosporine-treated recipients (p < 0.05). Rehabilitation. At 1 year 181 of 213 patients (85%) who completed a questionnaire were completely rehabilitated, and 32 of the 213 patients (15%) were not. NYHA class. Data are available for 199 patients at 1 year after transplantation. One hundred ninety-two pa'SD = Standard deviation.
50 3
100..,..--------------,
80
60
Creatinine Clearance (ml/min)
40
20
O-+-....--..,....--.-~r--.--r_o__,---..__r--...--r--l
o
12
24
36
48
60
72
84
Postoperative Month
Fig. 3. Mean creatinine clearance declined in the first several postoperative months but then remained relatively stable over the ensuing follow-up period.
tients were in NYHA class I (96.5%), six patients were in class II (3.0%), none was in class III, and one patient was in NYHA class IV (0.5%). Immunosuppression. Data regarding the trends of dosing of the three long-term immunosuppressant medications for all patients are depicted in Fig. 2. (As described in the Patients and methcx:ls section, lower doses of cyclosporine are now being used from the time of transplantation.) Azathioprine dosing has remained essentially the same over the 8-year pericx:l. Biopsy specimens. The first 60 patients underwent an average of 5.8 ± 3.5 (SO) biopsy procedures in their initial hospitalization; the last 60 patients had an average of 2.7 ± 1.8 (SO) biopsy procedures. This change correlated with the change in the length of stay in the hospital. Length ofstay. The first 60 patients in this series stayed in the hospital an average of 41 ± 19.4 (SO) days after transplantation. The last 60 patients stayed an average of 23.9 ± 14.6 (SO) days. Renalfunction. Serum creatinine, blocx:l urea nitrogen, and creatinine clearance generally deteriorated during the initial 2 to 3 months after transplantation and then stabilized, as shown in Fig. 3. At 1 year the average creatinine clearance for all cyclosporine-treated patients was 64.9 ± 27.4 (SO) ml/rnin/rn. The earliest subgroup of cyclosporine-treated patients (cyclosporine, 18 rug/kg/day) had significantly greater impairment of renal function than had the patients with the current protocol (cyclosporine,4 to 8 rug/kg/day initially). At 1 year average creatinine clearance and serum creatinine for the high-dose cyclosporine group were 47.6 ± 18.8 (SO) ml/rnin/rn and z.Z ± 0.6(SO)mg/dl,
The Journal of Thoracic and Cardiovascular
5 0 4 Grattan et at.
Surgery
100
Nocardia Protozoa Fungi
80
60
Percent of Patients Free from Infection
Bacteria 40
Viruses 20
O+--....--r--,---r----,r----r--r--....-
o
234
5
6
7
8
Postoperative Year
Fig. 4. Infection with a variety of organisms that necessitated treatment was a problem for most patients. The actuarial prevalence of developing a treatable infection with the major classes of pathogens is shown in this figure, expressed as the percent of all patients free of the infection over time. Most of the viral infections were oral mucocutaneous Herpes simplex type 1 infections, but there were also many cytomegalovirus infections (91 patientsj.s
2.00
Rejection Episodes Per 100 Patient-Days 1.00
0.00 1
2
3
4
5
6
7
8
9 10 11 1218243036424854606672
Month Post-Transplant
Fig. S. The crude rejection prevalence expressed as rejection episodes per 100 patient-days in certain posttransplant months is shown in this figure (note nonlinear time scale, x axis). This figure demonstrates that the number of rejection episodes is much greater in the early posttransplant period than subsequently, raising the question of immunologic "tolerance" of the allograft.
respectively. Values at 1 year for the current, low-dose cyclosporine subgroup were 65.2 ± 22.3 (SD) mljrninj m and 1.5 ± 0.6 (SD) mgjdl, p < 0.05. Costs. The average cost of a cardiac transplantation (hospital charges) in 1980was $83,432 ± $31,721 (1980 dollars). By comparison, the average inhospital cost for our most recent 60 patients with transplants has been
$104,102 ± $89,592 (SD)(1985 dollars). When the rate of inflation over the intervening 5 years is considered, the inhospital charges for transplantation have not changed significantly. Infection rates. Infections severeenough to necessitate treatment were also tabulated and are shown in Fig. 4. The case attack rates of bacterial infections at 1 and 5
Volume 99
Cyclosporine and cardiac transplants 5 0 5
Number 3 March 1990
100
-~~f==;::=:;:=+=+=:::;=~=~1
80
Graft Atherosclerosis 60
Percent of Patients Free from Event 40
20 Rejection
O+--.,.---,----,---,---r---,.----,r--r-'
o
2
3
4
6
5
7
8
Postoperative Year
Fig. 6. Actuariallycompiled data regardingthe prevalence of development of lymphoma, graft atherosclerosis, and rejection are expressed as the percentof patients free of each complication, as a function of time after transplantation.
years were 39.9% (± 2.9%) and 50.9% (± 3.6%), respectively.For fungal infections the rates at 1 and 5 years were 12.9%(± 2.0%) and 19.1% (±2.9%). For viral infections these rates were 56.8% (± 3.0%) and 65.3% (± 3.2%). For protozoal infections the rates were 9.5% ( ± 1.8%)and 12.3% ( ± 2.2%). For nocardial infections the rates were 5.5% (± 1.4%) and 7.6% (± 2.0%) for 1 and 5 years, respectively. Older recipients (50 years of age or older) had a significantly higher rate of bacterial infections (p < 0.05), fungal infections (p < 0.05), and nocardial infections (p < 0.05) compared with younger recipients (aged 20 to 49 years). There were no differences in the rates of infection with viruses or protozoa. Rejection rates. The crude rejection rates (calculated in monthly intervals) for cyclosporine-treated patients at 1 month and 1 year, respectively, were 2.38 (±0.16) and 0.14 (± 0.04) events per I()() patient-days. These rates are depicted in Fig. 5. The actuarial rejection prevalence is shown in Fig. 6. At 1 month this prevalence for the cyclosporine group was 60.0% (± 2.8%). At 1 year the actuarial rejection prevalence for this group was 86.9% (±2.1%). Older recipients (2:50 years) had fewer rejection episodes (actuarial model) than younger recipients «50 years), but the difference was not significant (p < 0.10). There was no significant difference between the actuarial rejection rates of recipients who received grafts from older donors (donor age >35 years) and those who received grafts from younger donors (donor age 20 to 35 years). Similarly, male (n = 237) and female (n = 73)
recipients showed no significant difference in actuarial rejection rates. Finally, recipients treated with high-dose cyclosporine had no significant difference in actuarial rejection rates from the current recipients of low-dose cyclosporine. Graft atherosclerosis. Angiographic and autopsy data were used to determine actuarial graft atherosclerosis rates. At 1 year significant angiographic or autopsyproven lesions had developed in 6.2% ± 1.6% of cyclosporine-treated patients. By 5 years such lesions had developedin24% ± 3.9% (Fig. 6). There were no significant differences in the rates of development of graft atherosclerosis in recipients of grafts from older donors (>35 years) and recipients of grafts from younger donors (19 to 35 years). Similarly, angiographic or autopsy-proven graft atherosclerosis developed in patients in the coronary artery disease subgroup at a rate that was not significantly different from that in cyclosporine-treated patients who did not have pretransplant coronary artery disease. (For more comparative data see section on causes of death of recipients.) Neoplasms. Eleven patients had lymphoma during their postoperative course, an average of 245 days after transplantation (range, 67 to 853 days). As shown in Fig. 6, the actuarial risk of developing lymphoma at 1 and 5 years was 2.9% (± 1.0%, SEM*) and 4.6% (± 1.4%, SEM). Six of the 11 patients have died, three as a direct result of the lymphoma, two each of infection, and one of graft atherosclerosis. Twelve patients had other forms of *SEM
= Standard
error of the mean.
The Journal of
506 Grattan et al.
Thoracic and Cardiovascular Surgery
100
-r-------------------, ..........
'1-"
80
Cyclosporine-treated patients (n=310) Pre-cyclosporine patients (n=99)
---0--
"1
"."t -!-. <,
60
Percent of Patients Surviving
"f. l
40
···L.
20
o-t--,--r-..,..--,I"""""-.----.-.----r-.---r----,-..,....J o
2
3
4
5
6
7
8
9
10
11
12
Postoperative Year Fig. 7. Patients treated with cyc/osporine had a significantly higher survival rate than had patients not treated with cyclosporine (p < 0.005). 100 .......- - - - - - - - - - - - - - - - - - - - ,
80
Percent of Patients Surviving
Recipients 20-50 years old
60
40 Recipients> 50 years old 20
O+---,..---r----,--r--,.---.--.........----,----l
o
2
3
4
5
6
8
Postoperative Year Fig. 8. Cyclosporine-treated patients 50 years of age or older (n = 82) had a significantly lower survival rate than had cyclosporine-treated patients 20 to 49 years of age (n = 177, P < 0.01).
neoplasm: six had carcinoma of the skin, three had adenocarcinoma of the lung, and there was one case each of testicular carcinoma, hepatic carcinoma, and renal cell carcinoma. Patients in our current immunosuppression protocol have had a significantly lower actuarial rate of lymphoma development over 2 years 1.3% ± 1.2%, SEM) (p < 0.005) than the earliest subgroup of cyclosporinetreated recipients. Survival rates: Cyclosporine versus precyclosporine groups. The actuarial survival rate for the entire cyclo-
sporine group of patients was significantly better (p < 0.005) than that of the precyclosporine group of patients (Fig. 7). The rates for the cyclosporine group at 1 and 5 years after transplantation were 80.7% (± 2.28%) and 59.7% (± 3.5%). For the precyclosporine group of patients these rates were 62.6% (±4.9%) and 36.4% (±4.8%), respectively, at 1 and 5 years. Survival rates: Cyclosporine-treated subgroups. The actuarial survival rates for two major subgroups of the cyclosporine-treated patients (those with idiopathic cardiomyopathy and coronary artery disease) were also an-
Volume 99
Cyclosporine and cardiac transplants
Number 3
50 7
March 1990
Pulmonary Embolus CVA Pulmonary Hypertension •
Non-Lymphoid CA
r:a
Deaths in Patients 20-50 Years Old Deaths in Patients over 50 Years Old
Lymphoid CA Graft Failure Other Graft Athero Rejection Infection
o
10
20
30
40
Numbers of Deaths
Fig. 9. Causes of death for all adult patients are shown; with subdivisions in each category for deaths in patients 20 to 49 years old and in patients 50 years or older.
alyzed. At I and 5 years these rates were 79.4% ( ± 3.7%) and 55.4% (± 5.6%) for the coronary artery disease group. The rates were 82.1% (±3.5%) and 59.4% ( ± 5.4%) for the idiopathic cardiomyopathy group. These rates were not significantly different from one another or from the cyclosporine group as a whole. Survival data for adult patients with transplants who were treated with cyclosporine were also analyzed according to the age of the patient at transplantation. The older group consisted of cyclosporine-treated patients who underwent transplantation when they were 50 years of age or older (n = 82). The younger group included all cyclosporine-treated patients between the ages of 20 and 50 years at the time of transplantation (n = 177). As shown in Fig. 8, the older group had significantly lower survival rates (p < 0.01). Thus at I and 5 years after transplantation the survival rates for the older group were 76.2% (± 4.9%) and 43.8% (± 7.9%), whereas for the younger group these rates were 82.6% (± 2.8%) and 62.0% (±4.3%). Survival rates for male and female cyclosporine-treated recipients were not significantly different. Similarly, there were no significant differences in actuarial survival rates between recipients of grafts from older donors (>35 years) and recipients of grafts from younger adult donors (20 to 35 years). Causes of death of recipients (cyclosporine-treated group). Of 310 patients from the cyclosporine group 206 are alive. The proximate cause of death in the 104 patients who have died was infection in 35.6%, rejection in 16.3%, graft atherosclerosis in 14.4%, nonspecific graft failure in 5.8%, lymphoid malignancy in 4.8%, nonlymphoid ma-
lignancy in 4.8%, pulmonary hypertension in 3.8%, cerebrovascular disease in 3.8%, pulmonary embolus in 1.9%, and other causes in 8.6% (Fig. 9). Death rates from rejection. Death rates from rejection for all cyclosporine-treated patients were 4.9% (± 1.3%) and 7.0% (± 1.8%) at I and 5 years after transplantation, respectively. Female recipients died of rejection more frequently than did men, but the difference was not significant (p < 0.10). N either age of the recipients nor age of the donors significantly affected death rates from rejection. Death rates from infection. Death rates from infection for all cyclosporine-treated patients were 8.4% (± 1.6%) and 15.1% ( ± 2.7%) at I and 5 years after transplantation, respectively. Female recipients died of infection more frequently than did male recipients, but this difference was not significant (p < 0.10). Neither age of the recipient nor age of the donor significantly affected actuarial death rates from infection. Death rates from graft atherosclerosis. Death rates from graft atherosclerosis for all cyclosporine-treated patients were 0.4% (± 0.4%) and 9.8% ( ± 2.8%) at I and 5 years after transplantation, respectively. Male recipients died of graft atherosclerosis significantly more frequently than did female recipients (p < 0.05). Similarly, recipients of grafts from older donors (> 35 years) of graft atherosclerosis died more frequently than did recipients of grafts from younger donors (20 to 35 years), but the difference was not statistically significant (p < 0.10). Older recipients (>50 years), however, did not have a different rate of death from graft atherosclerosis than younger adult recipients (aged 20 to 49 years).
The Journal of Thoracic and Cardiovascular
5 0 8 Grattan et al.
Finally, there were no significant differences in rates of death from graft atherosclerosis between recipients in the coronary artery disease subgroup before transplantation and those who did not have coronary artery disease as the reason for transplantation. Discussion This article summarizes the 8-year experience at Stanford University Medical Center with cyclosporine in the management of the immunosuppression of our patients.' After its introduction the survival rates at our institution significantly improved. As a result of these improved survival rates, cardiac transplantation has been offered as standard therapy for appropriate patients with end-stage cardiac dysfunction. In 1987, 109 centers in the United States performed 1436 cardiac transplantations. In the early years of our experience we restricted the age of participants so that neither recipients over the age of 50 years nor donors over the age of 35 years were considered. As seen in Fig. 1, however, we have substantially liberalized our age criteria for recipients. We also have implanted grafts from donors as old as 48 years of age with no discernible difference in recipient survival, rejection, or graft atherosclerosis. Therefore we have been able to offer the therapeutic option of orthotopic cardiac transplantation to a larger group of patients than in the earlier years of our experience. Despite these liberalized criteria, the survival rates of recipients have actually improved. Several additional points merit further emphasis. Survival rates. In this study the survival rate of recipients ages 20 to 49 years was significantly better than that for patients 50 years of age and older. This contrasts with the published reports of other transplant groups, but these studies have followed smaller groups of patients for much shorter periods of time (1 to 2 years).6. 7 The reasons for lower survival rates in older recipients may be explained by greater susceptibility in the older patients to infections with bacterial, fungal, and nocardial organisms, although the actuarially determined rates of death by infection for these two groups do not differ significantly. Rejection. Rejection episodes occurred in 60% of patients by 1 month despite their suppressed immune status. It remained the second most common cause of death in our patients (16%). The difference between this experience and that of the Minnesota group, which reported only two rejection episodes in 57 consecutive patients over 2 years, is difficult to explain without a careful review of the exact pathologic criteria used to diagnose rejection
episodes.f It is also of some interest that there was no difference between the actuarial rates of rejection in the early cyclosporine protocol with high-dose initial cyclosporine
Surgery
(18 rug/kg/day) and prophylactic rabbit ATG and in the later protocol of lower initial doses of cyclosporine (8 mg/kg/day) and prophylactic OKT3. Hypertension and renal dysfunction. Hypertension that was severe enough to require medications more potent that diuretics occurred in 73% of patients overall but in only 52% of patients who were treated with the current low-dose cyclosporine protocol. Renal function was deleteriously affected by immunosuppressants (principally cyclosporine) in all patients; however, the early recipients of the high-dose cyclosporine protocol had significantly more impairment of renal function than had the recipients of current low-dose cyclosporine protocol. Thus overall trends in renal function for all patients may obscure these deleterious effects of the higher doses of cyclosporine in the early subgroup of recipients. Graft atherosclerosis. It is of interest that the pretransplant diagnosis of coronary artery disease did not predispose patients to the development of graft atherosclerosis. Recipients of grafts from older organ donors (>35 years) died of graft atherosclerosis more frequently, but this trend was not statistically significant (p < 0.10). One risk factor for graft atherosclerosis that we have definitivelydetermined is postoperative CMV infection in a cardiac transplant recipient. 8 A recent review of clinical and laboratory variables did not identify any other striking correlates for the development of graft athero-
sclerosis." Neoplasms. A concern arising from our early experience with cyclosporine (when we were using doses approximately twice as high) was the development of B-cell lymphomas.'? The cyclosporine doses were reduced in later patients in an effort to curtail the development of lymphoma and to minimize nephrotoxicity. Indeed, our most recent subgroup of patients (treated with prophylactic OKT3, initial cyclosporine doses of 4 to 8 mg/kg/ day, prednisone, and azathioprine) has had a significantly less frequent development of lymphoma over the first 2 years that this protocol has been in use. Because most lymphomas have developed early in our patients (usually within the first year after transplant), there is reason to believe that the problem of lymphoma development will occur less frequently as a complication, even with longer follow-up. Rehabilitation. One of the most gratifying findings of this study is that surviving patients have generally enjoyed a good quality of life~ %5% have returned to employment or an activity of choice. Indeed, 96.5% of patients were classified in NYHA functional class I at I or more years after transplantation. Thus not only are most cardiac transplant recipients living longer, but equally important, they have achieved genuine rehabilitation.
Volume 99 Number 3
Cyc/osporine and cardiac transplants 5 0 9
March 1990
REFERENCES J. Billingham ME. The postsurgical heart: the pathology of cardiac transplantation. Am J Cardiovasc Pathol 1988; 1:319-34. 2. Cutler SJ, Ederer F. Maximum utilization of the life table in analyzing survival. J Chronic Dis 1958;8:699-712. 3. Kalbfleisch JD, Prentice RL. The statistical analysis of failure time data. New York: John Wiley & Sons, 1980. 4. Glantz SA. Primer of biostatistics. 2nd ed. New York: MeGraw-Hili Co, 1987. 5. Reitz BA, Bieber CP, Raney AA, et al. Orthotopic heart and combined heart and lung transplantation with cyclosporine-A immune suppression. Transplant Proc 1981; 13:393-6. 6. Miller LW, Vitale-Noedel N, Pennington DG, McBride L, Kanter KR. Heart transplantation in patients over age fifty-five years. J Heart Transplant 1988;7:254-7.
7. Olivari MT, Antolick A, Kaye MP, Jamieson SW, Ring WS. Heart transplantation in elderly patients. J Heart Transplant 1988;7:258-64. 8. Grattan MT, Moreno-Cabral CE, Starnes VA, Oyer PE, Stinson EB, Shumway NE. Cytomegalovirus is associated with cardiac allograft atherosclerosis and death. JAMA (In press.) 9. Gao SZ, Schroeder JS, Alderman EL, et al. Clinical and laboratory correlates of accelerated coronary artery disease in the cardiac transplant patient. Circulation 1987;76:V5661.
10. Brumbaugh J, Baldwin JC, Stinson EB, et al. Quantitative analysis of immunosuppression in cyclosporine-treated heart transplant patients with lymphoma. Heart Transplant 1985;4:307-1 J.