Author + information
- Received April 16, 1999
- Revision received June 28, 2000
- Accepted September 7, 2000
- Published online January 1, 2001.
- Neda F Mulla, MD, FACC∗,* (, )
- Joyce K Johnston, RN†,
- Laura Vander Dussen, RN†,
- W.Lawrence Beeson, MSPH∗,
- Richard E Chinnock, MD∗,
- Leonard L Bailey, MD, FACC‡ and
- Ranae L Larsen, MD, FACC∗
- ↵*Reprint requests and correspondence: Dr. Neda F. Mulla, 11234 Anderson St., Room 4433, Loma Linda, California 92354
The study objectives were to determine posttransplant coronary artery disease (TxCAD) incidence, predisposing factors and optimal timing for retransplantation (re-Tx) in pediatric heart transplantation (Tx) recipients.
The TxCAD limits long-term survival following heart Tx, with re-Tx being the primary therapy. Information on risk factors and timing of listing for re-Tx is limited in children.
The records of children who survived >1 year post-Tx at Loma Linda University were reviewed. Nonimmune and immune risk factors were analyzed.
TxCAD was documented in 24 of 210 children. Freedom from TxCAD was 92 ± 2% and 75 ± 5% at 5 and 10 years’ post-Tx, respectively. The TxCAD diagnosis was established at autopsy in 10 asymptomatic patients who died suddenly within nine months following the most recent negative angiograms. The remaining 14 children had angiographic diagnoses of TxCAD and had symptoms and/or graft dysfunction (n = 10) or positive stress studies (n = 4). Three of 14 died within three months after the diagnosis was made. Eleven patients underwent re-Tx within seven months of diagnosis; nine survived. Univariate and multivariate analyses showed that only late rejection (>1 year posttransplant) frequency (p = 0.025) and severity (hemodynamically compromising) (p < 0.01) were independent predictors of TxCAD development. Freedom from TxCAD after severe late rejection was 78 ± 8% one year postevent and 55 ± 10% by two years.
Late rejection is an independent predictor of TxCAD. Patients suffering severe late rejection develop angiographically apparent TxCAD rapidly and must be monitored aggressively. Both TxCAD mortality and morbidity occur early; therefore, we recommend immediate listing for re-Tx upon diagnosis.
Posttransplant coronary artery disease (TxCAD) is one of the leading factors limiting long-term graft and patient survival following heart transplantation (Tx) (1,2). Therapeutic options are primarily limited to retransplantation (re-Tx) (3). Information regarding the course of TxCAD once diagnosed—and therefore timing of re-Tx—is largely lacking in the pediatric population. The perceived TxCAD incidence in our pediatric recipients was low (4%) at five years’ post-Tx as reported by Bailey et al. in 1995 (4). However, with continued follow-up of those patients, a higher incidence of TxCAD was noted, prompting reevaluation of the extent and scope of the disease. The objectives of this study were to determine the incidence, predisposing risk factors and clinical course of TxCAD for pediatric heart transplant recipients and to identify optimal timing for re-Tx.
The medical records of children (<18 years) transplanted at Loma Linda University who had survived more than one year after Tx were retrospectively reviewed. Patients were excluded if follow-up data were not available. The study period encompassed November 1985 through April 1998. All patients had selective coronary angiography using Judkins catheters appropriate for size and standard techniques. Nitroglycerin was not routinely used and was only given if coronary artery spasm was suspected. Serial cine films were reviewed by a panel of pediatric cardiologists, cardiothoracic surgeons and an adult cardiologist and were compared for local or diffuse vessel irregularities, loss of third-order branching, vessel pruning and presence of myocardial bridging. Angiographically significant TxCAD was defined as the presence of >50% luminal reduction whether local or diffuse. Pathologic examination was performed in all cases but two. On pathological examination, TxCAD was considered moderate if intimal proliferation caused >50% luminal reduction and severe if there was >75% luminal loss in any coronary artery branch.
Nonimmune risk factors studied included age at Tx, gender, pre-Tx diagnosis, length of follow-up, graft ischemic time, cytomegalovirus (CMV) infection (defined as new onset of clinical symptoms of CMV accompanied by seroconversion), number of serious infections, lipid profile and donor/recipient weight ratio, age difference, gender and race mismatch. Immune risk factors studied were level of immunosuppression (i.e., monotherapy vs. double immunosuppression) and rejection history.
Rejection was defined as any event requiring augmentation of immunosuppression. Events precipitating such augmentation included one or more of the following: 1) endomyocardial biopsy grade ≥III; 2) acute echocardiographic changes of rejection; and/or 3) clinical symptoms and signs of rejection. Echocardiographic changes of rejection as previously described by this institution are increased myocardial wall thickness with resulting increased left ventricular mass and decreased volume, depressed systolic function and new onset of mitral regurgitation (5). Asymptomatic patients with biopsy grade ≥III on routine annual biopsy were treated for rejection.
Rejection history was expressed as linearized rejection rates (rejections per patient-month) for the following post-Tx intervals: first six months, >6 to 12 months and late rejection frequency (frequency of rejections >1 year post-Tx per patient-month). Previously undiagnosed rejection detected only upon postmortem examination was not classified as a rejection event and was excluded from statistical analysis. Severe late rejection was defined as rejection episodes that took place >1 year post-Tx and were characterized by hemodynamic compromise accompanied by moderate to severe echocardiographic evidence of graft dysfunction. Hemodynamic compromise was defined if the child 1) required inotropic support for maintenance of cardiac output, 2) had congestive heart failure, or 3) required mechanical ventilation for pulmonary edema.
Data collected to identify the clinical course of TxCAD included interval from diagnosis to re-Tx or death, any intervening symptoms or cardiac events and outcome following re-Tx.
Immunosuppression and rejection treatment protocols
Cyclosporine and methylprednisolone boluses were used for induction as well as cytolytic antibody therapy if the Tx took place beyond the neonatal period. A combination of cyclosporine and azathioprine (1 mg/kg) was used as maintenance therapy. Target cyclosporine levels were initially 250 to 300 ng/ml, gradually lowered to 100 to 150 ng/ml by the end of the first year post-Tx. Fluorescence polarization immunoassay (FPIA) was the preferred cyclosporine assay method. Steroids were not used as routine immunosuppression. In our early experience, azathioprine was discontinued at the end of the first year if the recipient was a neonate (≤30 days of age) at Tx and was free of significant episodes of rejection. However, all but 16 children were eventually restarted on azathioprine because of rejection or renal insufficiency. Patients were not routinely placed on calcium channel blockers or angiotensin-converting enzyme inhibitors.
All rejection episodes were treated with intravenous (IV) methylprednisolone (20 mg/kg/dose) given twice daily for eight doses. Recurring or persistent rejection was reversed by use of rabbit-derived thymogobulin or equine-derived antithymocyte globulin antibody therapy and repeat IV steroids. Recalcitrant rejection was treated with a variable combination of methotrexate, conversion to tacrolimus or total lymphoid irradiation.
While awaiting re-Tx, candidates were pretreated with high-dose (2 gm/kg) IV immunoglobulin (IVIG) every four weeks for three doses if there was significant elevation of percent reactive antibody titers. The induction protocol was not different from primary Tx. For maintenance therapy, cyclosporine levels were maintained at 200 to 250 ng/ml, and either methotrexate or mycophenolate mofetil was used in place of azathioprine.
Rejection surveillance protocol
Rejection surveillance was achieved using clinical evaluation by the transplant physicians, echocardiography and endomyocardial biopsy. The frequency of Tx clinic visits was twice per week for the first six weeks, then less often if free of rejection. Echocardiography frequency was three times per week for the first two weeks’ post-Tx, then twice a week for the next four weeks, decreasing to once a month for one year and finally once every three months thereafter. Endomyocardial biopsy schedule was according to age at Tx. Recipients below two years at Tx had annual biopsies only. Recipients between two and eight years at Tx had biopsy at 1 month, 3 months and 12 months’ post-Tx and annually thereafter. Recipients ≥9 years at primary Tx or following re-Tx were routinely biopsied at 1 month, 2 months, 3 months, 6 months, 12 months and annually thereafter.
Coronary artery disease surveillance protocol
Before 1994, coronary angiography was performed every two years and was the only testing modality for TxCAD. This was believed to be adequate given the perceived low incidence of TxCAD at that time. However, the diagnosis of new cases of TxCAD in 1994 and 1995 heightened awareness of the problem. After 1994, the protocol was changed to annual coronary angiography, annual dobutamine stress echocardiography (DSE) for children >3 years old and annual exercise electrocardiogram (ECG) for children >6 years. The DSE protocol used in this population is detailed in a separate article and has a reported sensitivity of 72% and specificity of 80% when compared to angiography (6). Patients with suspected mild abnormalities on coronary angiography or positive DSE underwent biannual coronary angiography. After a severe late rejection episode, children were monitored with coronary angiography two months after the event; if negative, angiography was repeated at six months and then at increasing intervals thereafter.
Univariate analysis was performed with independent Student t-test for parametric continuous variables and with chi-square test for nominal data. Nonparametric data were expressed as median and range and analyzed using the Mann-Whitney U-test. A multiple logistic regression model was used to test variables that reached statistical significance with univariate analysis. Kaplan-Meier survival analysis was used to estimate freedom from TxCAD. The variability of these estimates is indicated by ± 1 SEM (standard error of the mean).
During the study period, 313 children were transplanted; 234 survived >1 year. Twenty-four of the 234 survivors were excluded for lack of follow-up data; the remaining 210 Tx recipients constituted the study group. Median age at Tx was 57 days (range, 1 day to 15.5 years) and mean follow-up period was 69 ± 28 months, with 110 children having >5 years’ follow-up. Of the 210 patients, 24 had TxCAD documented by coronary angiography and/or pathological examination. Twelve of 14 children diagnosed with angiographically significant TxCAD (>50% luminal reduction) were confirmed to have moderate to severe TxCAD by pathological evaluation. Two families declined the performance of an autopsy.
Incidence of TxCAD
Freedom from significant TxCAD at 3, 5, and 10 years’ post-Tx was 97 ± 1%, 92 ± 2%, and 75 ± 5%, respectively; milder disease may have been missed (Fig. 1). TxCAD alone accounted for 17% of all late deaths (>1 year post-Tx); an additional 14% had a combination of cellular rejection and TxCAD (total of 31%).
Clinical status and outcome for TxCAD patients
Of the 24 patients with TxCAD, 10 died suddenly with no diagnosis of TxCAD until autopsy. The clinical information on the 10 patients with TxCAD who died suddenly is summarized in Table 1. Nine of these patients (90%) had negative angiograms only three to nine months before their deaths. None had symptoms indicative of myocardial ischemia before death. Autopsy examination showed moderate rejection in 6 of 10 patients (60%) that may have contributed to their death.
The remaining 14 children had angiographic diagnosis of TxCAD. Three of 14 children died and 11 underwent re-Tx. The clinical information on the 14 patients with TxCAD who were diagnosed with angiography is summarized in Table 2. Coronary angiography was performed because of symptoms in three patients (bradycardia, syncope and congestive heart failure). Indications for coronary angiography in the remaining patients included routine annual coronary angiogram (n = 4), follow-up following severe late rejection (n = 5) and positive exercise ECG (n = 1) and positive DSE (n = 1).
Three patients died suddenly 10 days to 3 months after the angiographic diagnosis was made. Ten of 14 patients had systolic ± diastolic graft dysfunction and/or were symptomatic. Four asymptomatic patients free of graft dysfunction had evidence of ischemia on performing DSE (n = 2), exercise stress echocardiography (n = 1) and exercise ECG (n = 1). Hemodynamic data at the time of TxCAD diagnosis were normal except for three patients who had elevated (>15 mm Hg) left ventricular end-diastolic pressures and low cardiac output (<3.0 liter/min/m2). Overall, 42% of TxCAD patients were asymptomatic but died suddenly within nine months of the most recent angiogram, 42% had symptoms and/or graft dysfunction at diagnosis, and 16% were asymptomatic but had positive stress studies.
Ten of the children underwent re-Tx between five days and seven months following diagnosis (median, 3 months). One of our earliest cases diagnosed with TxCAD did not undergo re-Tx until two years after diagnosis when progressive graft dysfunction and worsening congestive heart failure prompted listing for re-Tx. The child who waited seven months before re-Tx had progressive congestive heart failure symptoms, graft dysfunction and eventually hemodynamic instability requiring inotropic support.
Analysis for nonimmune predisposing risk factors
Univariate analysis of gender, pre-Tx diagnosis, length of follow-up, graft ischemic time, number of serious infections, lipid profile, CMV infection, donor/recipient weight ratio, gender mismatch, and age difference did not show any significant predisposing factors for the development of TxCAD (Table 3). Statistical significance was reached (p = 0.034) for donor/recipient race mismatch, which was more prevalent in TxCAD-free patients (60% vs. 38%). There was no statistical difference (p = 0.07) in the incidence of TxCAD between neonatal recipients and older children (Fig. 2). A comparison of patients with and without angiographic TxCAD diagnosis in regards to age at Tx, length of follow-up, year of Tx, type of CAD and incidence of rejection yields no significant differences.
Analysis for immune risk factors
Univariate comparison of monotherapy versus double immunosuppression was not statistically significant for development of TxCAD. Univariate analysis for rejection (Table 4)shows that rejection frequency in the first six months’ post-Tx did not reach statistical significance. However, rejection frequency in the latter half of the first post-Tx year and late rejection frequency were significant at p = 0.01 and p = 0.0001, respectively. Overall, 67% of all patients were free of rejection after the first post-Tx year, whereas 33% had one or more episodes of late rejection. A history of severe late rejection episodes was present in 48% of TxCAD patients compared to only 10% of patients without TxCAD (p < 0.0001).
Multiple logistic regression shows that late rejection frequency (p = 0.025) and severe late rejection episodes (p < 0.01) were independent predictors of TxCAD. The odds ratio and 95% confidence interval for development of TxCAD in the presence of ≥1 late rejections were 3.95 (1.11, 14.00). The odds ratio (OR) and 95% confidence interval (CI) for development of TxCAD after episode(s) of severe late rejection were 4.19 (1.40, 12.54). The development of TxCAD in the subgroup that suffered episodes of severe late rejection was rapid (Fig. 3)with freedom from TxCAD being 78 ± 8% one year following the event and 55 ± 10% by two years. Freedom from TxCAD after the first late rejection (regardless of severity) (Fig. 4)was 94 ± 3% at one year, 73 ± 6% at three years and 62 ± 8% by five years.
Eleven patients have undergone re-Tx for TxCAD. Early in our experience, one patient died one month after re-Tx owing to rejection. Another patient developed aplastic anemia in the first month following re-Tx and died. Nine have survived, with a median follow-up period of 37 months (range, 17 to 63 months). The median number of rejections and infections was 1 (range 0 to 4) and 1 (range 0 to 5), respectively.
Reports of TxCAD incidence in pediatric transplant recipients have varied widely by source and time of publication. Earlier reports indicated a high incidence of 27% to 35% over mean follow-up periods of three years (1,7,8). Later studies reported a lower incidence of 2% to 8% at two to four years’ post-Tx (4,9,10). A recent study reported a 19% incidence among 68 pediatric recipients who survived >5 years (11). Most likely, the proportion of children with TxCAD is relative to follow-up duration. In this study the incidence of TxCAD was 8% and 25% by 5 and 10 years’ post-Tx. This is a higher incidence than previously perceived, probably due to the longer follow-up period (mean, 5.8 years) in this study. The contribution of TxCAD to late deaths has been high. Pahl et al. (10)reported 37% of all late deaths to be due to TxCAD. In the present study, TxCAD contributed to 31% of all deaths occurring more than one year post-Tx.
The characteristic pathology of TxCAD involves diffuse concentric intimal thickening, which makes coronary angiography an insensitive tool because uniform luminal reduction may not be apparent until it becomes severe. The present study concurs with prior reports of angiographic insensitivity (9,10,12,13).
Predisposing risk factors
Nonimmune risk factors
Pediatric transplant literature has not shown any correlation of TxCAD with nonimmune risk factors. The power of this study, at current sample size, is <50% for each of the risk factors examined. In this study, the association of donor/recipient race mismatch with TxCAD (less mismatch in TxCAD patients) is unexplained.
Immune risk factors
The present study supports the earlier findings by Addonizio et al. (7), who described a greater number of late rejections in nine children with late graft loss. Pathological examination of eight of these grafts showed rejection only in two and a combination of rejection plus coronary artery disease in six. Noncompliance was believed to have an important role in late graft loss (14).
Only late rejection history predicted TxCAD in the present study. Both the frequency and the severity of late rejections were significant. Frequent late rejections (regardless of severity) andsevere late rejection episodes (even if no other rejections existed) were independent predictors of TxCAD development. Noncompliance may contribute to episodes of significant late rejection; however, documentation of noncompliance has been very difficult. If indeed late acute rejection episodes relate to periods of subtherapeutic cyclosporine levels, then the importance of dual immunoregulation cannot be overstressed.
The time interval between severe late rejection episodes and the ability to detect TxCAD angiographically is relatively brief. Freedom from angiographic coronary artery disease was only 55% by two years following severe late rejection. Rigorous surveillance is needed to avoid sudden death from TxCAD, which has occurred as early as two months following the rejection event in those children.
Optimal timing for retransplantation
Transplant coronary artery disease mortality is high, with many of the deaths occurring suddenly. The multicenter study compiled by Pahl and colleagues (10)reports only 9 of 58 TxCAD patients surviving, including 5 who underwent re-Tx. Many of the deaths were sudden and unexpected (10). In the present study, 13 of 24 (54%) TxCAD patients died suddenly, and 2 died following re-Tx. Overall, only 9 of 24 (37%) patients with TxCAD survived, all after re-Tx. Sudden death occurred within a relatively short time (3 to 9 months) for those with negative angiograms and 10 days to 3 months for those with positive coronary angiography. It remains to be seen whether the sudden death incidence will be decreased with heightened surveillance protocols and identification of high-risk children. TxCAD morbidity, manifesting as bradycardia, syncope, graft dysfunction and congestive heart failure, is also high.
Re-Tx is the only documented therapeutic option of long-term benefit in children (3,15). In the face of high TxCAD mortality/morbidity and lengthy waiting periods for donor organs, we recommend immediate listing for re-Tx upon establishing a diagnosis of TxCAD. There is no advantage to waiting until a child shows evidence of graft dysfunction or symptoms of myocardial ischemia secondary to TxCAD. In the absence of angina, symptomatic monitoring of ischemia is not reliable; hence, the first manifestation may be sudden death.
The outcome following re-Tx has been encouraging thus far. This is consistent with reports of improved outcome of elective re-Tx for CAD compared with that for acute re-Tx required for graft failure/rejection (16–18). These studies report no difference in rejection and infection rates when compared to primary Tx. The duration of follow-up for the present cohort of re-Txs is inadequate to evaluate long-term mortality and morbidity.
The question of whether increased immunosuppression decreases the incidence of TxCAD was addressed in a study by Addonizio et al. (9), who reported a drop in TxCAD incidence from 18% to 2% by going from double to triple immunosuppression (both groups were on maintenance steroid therapy). However, reports on steroid therapy remain divided. Braunlin and colleagues (8)reported a 35% incidence of TxCAD in children receiving triple immunosuppression. Other investigators found that a higher prednisone dose was substantially deleterious, and some adult centers are now moving away from maintenance steroid therapy (19,20). The effect of steroid therapy on TxCAD could not be tested in the present study as steroids are primarily used for acute rejection therapy and not for routine maintenance therapy and hence cannot be statistically separated from the rejection event.
The perceived benefits of routine triple immunosuppression must be weighed against the risks of infection and neoplasia. Increasing immunosuppression for all recipients would mean that the 67% of children in the present study who never experienced late rejection would be exposed to unnecessary growth impairment, infections, and neoplastic risks.
Finally, late rejection is an independent predictor of TxCAD. The most rapid angiographic development of TxCAD will follow episodes of severe (hemodynamically compromising) late rejection. The morbidity and mortality of TxCAD are high and occur early. Successful re-Tx has improved TxCAD survival. We therefore recommend rigorous screening for TxCAD especially after episodes of severe late rejection and immediate listing for re-Tx upon diagnosis. Further studies centering on CAD prevention are needed.
- dobutamine stress echocardiography
- posttransplant coronary artery disease
- Received April 16, 1999.
- Revision received June 28, 2000.
- Accepted September 7, 2000.
- American College of Cardiology
- Larsen R.L.,
- Applegate P.M.,
- Dyar D.A.,
- et al.
- Addonizio L.J.,
- Hsu D.T.,
- Douglas J.F.,
- et al.
- Johnson D.E.,
- Aldermann E.L.,
- Schroeder J.S.,
- et al.
- St. Goar F.G.,
- Pinto F.J.,
- Aldermann E.L.,
- et al.
- Douglas J.F.,
- Hsu D.T.,
- Addonizio L.J.
- Mehra M.R.,
- Ventura H.O.,
- Jain S.P.,
- et al.
- Kobashigawa J.A.,
- Laks H.,
- Drinkwater D.C.,
- et al.