Cardiovascular outcomes after kidney–pancreas and kidney–alone transplantation1

Kidney International, Vol. 60 (2001), pp. 1964–1971

Cardiovascular outcomes after kidney–pancreas and kidney–alone transplantation1 ENNIO LA ROCCA, PAOLO FIORINA, VALERIO DI CARLO, ETTORE ASTORRI, CLAUDIO ROSSETTI, GIOVANNI LUCIGNANI, FERRUCCIO FAZIO, DANIELA GIUDICI, MARCO CRISTALLO, GIUSEPPI BIANCHI, GUIDO POZZA, and ANTONIO SECCHI Departments of Internal Medicine, General Surgery, Nuclear Medicine, Anesthesiology, and Nephrology, San Raffaele Scientific Institute, and Internal Medicine Chair, Universita´ Vita e Salute, San Raffaele Scientific Institute, Milan; and Cardiology Chair, Universita´ di Parma, Parma, Italy

Cardiovascular outcomes after kidney–pancreas and kidney– alone transplantation. Background. This study retrospectively assessed, with an intention-to-treat analysis, the effect of kidney–pancreas transplantation (KP) on survival and cardiovascular outcome in type 1 diabetic uremic patients. Methods. A total of 351 uremic type 1 diabetic patients were enrolled on a waiting list for KP: 130 underwent KP transplantation, 25 underwent kidney transplantation alone (KA), whereas 196 patients remained on dialysis (WL). The three populations had similar cardiovascular conditions. Actuarial survival rates and causes of death were recorded over a period of seven years. Finally, 23 KP and 13 KA patients underwent left radionuclide ventriculography, during a follow-up of four years. Results. In the entire group of 351 patients the seven-year survival rate was 77.4% for KP, 56.0% for KA and 39.6% for WL (KP vs. WL, P ⫽ 0.01). Cardiovascular death rate was 7.6% in KP, 20.0% in KA and 16.1% in WL (KP versus WL, P ⫽ 0.03; KP vs. KA, P ⫽ 0.16). In the subsample studied with radionuclide ventriculography, left ventricular ejection fraction improved in KP, but did not in KA, with significant differences between groups at two and four years. At four years only the KP patients presented normal values of diastolic parameters, including the peak filling rate, time-to-peak filling rate, and peak filling rate/peak ejection rate ratio. Glycated hemoglobin was negatively associated with the ejection fraction, peak filling rate and peak filling rate/peak ejection rate ratio, and positively associated with the time-to-peak filling rate. Conclusions. Normalization of blood glucose metabolism and improvement of blood pressure control obtained with KP transplant is associated with positive effects on survival, cardiovascular death rate, and left ventricular function.

1

See Editorial by Langone and Helderman, p. 2035.

Key words: pancreas transplantation, kidney transplantation, dialysis, cardiovascular disease, survival, neoplasm, radionuclide ventriculography, left ventricular function, glycometabolic control. Received for publication October 16, 2000 and in revised form May 21, 2001 Accepted for publication June 4, 2001

 2001 by the International Society of Nephrology

Some editorials have recently addressed the question of whether the risks associated with pancreas/kidney– pancreas transplantation, that is, major surgery and immunosuppression [1–3], are counterbalanced by the positive effects on the acute and chronic complications of diabetes [4]. Furthermore, these procedures seem to be associated with a higher incidence of cancer and infection than in the general population [5, 6]. Positive effects of transplant, such as an improvement in peripheral neuropathy [7], the quality of life, stabilization of retinopathy [8], and a decrease in glomerular volume of the transplanted kidney, already have been shown [9]. However, except for the results of Gaber et al, who reported a significant improvement of cardiac function by echocardiography in pancreas/kidney transplanted patients up to five years, no other studies aimed at evaluating the effects of pancreas transplantation on cardiovascular outcome [10, 11]. It is noteworthy that diabetic patients are at increased risk for cardiovascular morbidity and mortality [12–19], with an estimated cardiovascular death rate ranging from 8.9% to 14.9% [12]. Some investigators suggest that this is related to diffuse peripheral coronary atherosclerosis, although the question concerning the existence of a specific diabetic cardiomyopathy is still under discussion [18]. Left ventricular mass has been reported to be higher in diabetic patients than in controls [14, 15, 19]. Diastolic filling is frequently impaired [14] and hypertension, particularly in diabetic-uremic patients, is still a major problem [3]. The aim of our study was to retrospectively assess, with an intention-to-treat analysis, the effect of good glycometabolic control obtained through kidney–pancreas transplantation on actuarial survival, cardiovascular events, causes of death, and left ventricular systolic/diastolic function in diabetic type 1 uremic patients.

1964

La Rocca et al: Kidney–pancreas transplant outcomes

1965

Table 1. Characteristics of patients when enrolled (no statistical difference was observed among the three for each of the parameters considered)

Age years Duration of dialysis months Duration of diabetes years Glycated hemoglobin % Creatinine mg/dL Total cholesterol mg/dL Triglycerides mg/dL Patients on anti-hypertensive therapy % Coronary artery disease % Smoking habits %

Kidney–pancreas transplanted patients (N ⫽ 130)

Kidney–alone transplanted patients (N ⫽ 25)

Dialysed type 1 diabetic patients (N ⫽ 196)

44.0 ⫾ 1.2 35.3 ⫾ 3.3 24.5 ⫾ 0.9 11.0 ⫾ 1.8 8.80 ⫾ 0.49 223.4 ⫾ 10.6 180.1 ⫾ 10.1 99 17.5 21.8

46.9 ⫾ 1.5 36.7 ⫾ 3.1 25.0 ⫾ 0.7 11.1 ⫾ 2.0 8.84 ⫾ 0.63 208.9 ⫾ 7.9 183.5 ⫾ 17.5 100 17.5 20.5

46.5 ⫾ 0.6 38.0 ⫾ 4.1 30.1 ⫾ 1.1 11.5 ⫾ 2.2 8.14 ⫾ 0.33 216.7 ⫾ 6.3 197.3 ⫾ 15.4 98 13.0 28.3

METHODS Patients Between January 1984 and January 1998, 351 type 1 diabetic uremic patients were enrolled on our waiting list for kidney–pancreas transplantation; 130 of them underwent kidney–pancreas transplantation (KP; 105 whole pancreas and 25 segmental pancreas). Patients who received only a renal transplantation (25 patients) due to macroscopic damage at the time of organ harvest, constituted the kidney alone (KA) group. One hundred and ninety-six patients remained on dialysis and on waiting list (WL) for immunological reasons such as low HLA matching and/or antilymphocyte antibodies. The clinical characteristics of the three patient populations were similar when enrolled on the waiting list (Table 1), exclusion criteria being previous strokes, major amputations and severe dilated cardiomyopathy. Coronary artery disease was defined on the basis of resting electrocardiogram (ECG), thallium-201 myocardial perfusion scintigraphy, and coronarography in patients positive to thallium-201 scintigraphy. None of the patients presented a pathological ejection fraction, and none of them presented characteristics compatible with any of New York Heart Association classes for heart failure. Particularly as regards to cardiovascular condition, no differences such as previous myocardial infarction, lipid status, and smoking habit were evident (Table 1). Transplantation Organs for transplantation were obtained from cadaveric donors through Nord Italia Transplant. Pancreas transplantation was performed as previously described [20–22]. All the transplanted patients received the following immunosuppressive treatment: anti-thymoglobulins (IMTIX), cyclosporine 6 mg kg/day, azathioprine 1 mg kg/day, and prednisone 10 mg/day. Episodes of renal rejection were treated with pulses of 500 mg of methylprednisolone. In cases of “steroid-resistant” rejection, OKT3 or a course of IMTIX was used. Kidney– pancreas patients were insulin-independent, whereas

kidney–alone patients were on conventional subcutaneous insulin therapy (average 48 IU/day). Cyclosporine levels were within the therapeutic range. Clinical follow-up Actuarial survival of patients and grafts was calculated up to seven years after transplantation. A complete clinical and instrumental assessment aimed at evaluating cardiovascular conditions (physical examination, ECG, carotid and leg arteries’ Doppler sonography, chest x-ray and blood pressure values) was performed yearly in all the transplanted patients. Cardiovascular events (episodes of angina pectoris with ECG modifications, acute myocardial infarction, acute heart failure) were regularly recorded in all patients. Blood pressure was measured with sphygmomanometer with subjects in the sitting position, and the average of the last two measurements was recorded. Hypertension was considered present if any of the following conditions were met: systolic blood pressure of 140 mm Hg or more, diastolic blood pressure of 90 mm Hg or more, or reported use of a medication for hypertension. The category of medication assessed in this study were angiotensin-converting enzyme (ACE) inhibitors, beta-blockers, and calcium-channel antagonists [23]. Blood pressure was recorded in each outpatient control at approximately every three to six months. Pancreas and renal function (glycated hemoglobin and serum creatinine) were tested at enrollment and every six months thereafter. Radionuclide left ventriculography From January 1993 to January 1998 all transplanted patients underwent radionuclide left ventriculography studies. Only patients with at least four years follow-up, who were not receiving beta blockers and digitalis, and who did not develop a myocardial infarction, were considered for this study (23 KP and 13 KA transplanted patients). During this time period 45 patients underwent transplantation. Four of them died (2 KP and 2 KA), 4 of them developed cardiovascular complications (2 acute

1966

La Rocca et al: Kidney–pancreas transplant outcomes

Table 2. Pre-transplant characteristics of patients who underwent radionuclide assessment during the follow-up

Age years Duration of dialysis months Ejection fraction % Diastolic dysfunctiona Heart rate bpm Ischemic ECG Ischemic myocardial scintigraphy Hypertensive patients Total cholesterol mg/dL Triglycerides mg/dL

Kidney–pancreas transplanted patients (N ⫽ 23)

Kidney–alone transplanted patients (N ⫽ 13)

38.0 ⫾ 1.5 26.2 ⫾ 4.3 67.0 ⫾ 4.2 69.5% 76.3 ⫾ 2.6 1/23 1/23 100% 233.0 ⫾ 15.7 185.4 ⫾ 14.7

42.6 ⫾ 2.5 35.0 ⫾ 4.9 63.8 ⫾ 6.3 84.6% 82.6 ⫾ 3.8 2/13 2/13 100% 212.5 ⫾ 10.7 180.8 ⫾ 29.3

Data were expressed as mean ⫾ standard error (no significant differences were found). a Assessed with Doppler echocardiography evaluation

myocardial infarction in the KA group and 2 acute heart failure, 1 in KP and 1 in KA), while one patient required beta-blocker therapy. Thirty-six patients were eligible for the study. The two groups were comparable before transplant for age, duration of dialysis, total cholesterol, triglyceride levels, ejection fraction and diastolic dysfunction (Table 2). On the day of the test, after a 15 minute rest in recumbent position, patients underwent radionuclide left ventriculography as previously reported [13]. Statistical methods The primary end point was death of patients. Actuarial survival rate at seven years was calculated for patients and grafted organs. The secondary endpoint was cardiovascular outcome. The baseline data in the two groups of patients were compared by Student t test for unpaired data, chi-square test for categorical variables and if not normal distribution was present Mann-Whitney U test was used. Differences in survival were assessed for statistical significance by Kaplan-Meier survival analysis. The influence of different pre-transplant parameters on survival was assessed with Cox regression analysis, and the frequency of cardiovascular events in the three groups was compared by ANOVA. The Spearman rank order test was used to assess correlation between parameters. Data were expressed as means ⫾ standard error. RESULTS Patients, graft survival and causes of death At seven years of follow-up, the patient actuarial survival rate was higher in the kidney–pancreas than in the kidney–alone and dialyzed groups (Fig. 1). Cox regression analysis showed no correlations between survival and duration of type 1 diabetes (d.f. ⫽ 1, P ⫽ 0.24), duration of dialysis (d.f. ⫽ 1, P ⫽ 0.78), cold ischemia time of the kidney (d.f. ⫽ 1, P ⫽ 0.5), while a significant value was present for pre-transplant age (d.f. ⫽ 1, P ⫽

0.007). Cardiovascular death rate (acute myocardial infarction, acute heart failure, lethal arrhythmias) was 7.6% in KP, 20.0% in KA patients and 16.1% in dialyzed patients (KP vs. WL, d.f. ⫽ 1, P ⫽ 0.03; KP vs. KA, d.f. ⫽ 1, P ⫽ 0.16). Neoplasms were lethal in 4.6% of the kidney–pancreas and in 4.0% of the kidney–alone group (NS). Infections were lethal in 0% of kidney– pancreas patients, 4.0% of kidney–alone and 4.2% of dialyzed patients (NS). Actuarial kidney graft survival at one, four and seven years was 93.4%, 87.2%, and 85.2%, respectively, in the kidney–pancreas group, and 95%, 85%, and 70%, respectively, in the kidney-alone group. Actuarial whole pancreas survival at one, four, and seven years was 76.7%, 70.6%, and 56.6%, respectively. Creatinine levels were similar in the two groups of transplanted patients, (KP vs. KA, 1.3 vs. 1.4 mg/dL, 1.7 vs 1.8 mg/dL, and 1.8 vs 1.9 mg/dL at one, four and seven years, respectively). Glycated hemoglobin was lower in KP than in KA group (KP vs. KA, 6.2 vs. 7.9%, 6.3 vs. 8.8%, and 6.2 vs. 8.2% at one, four, and seven years, respectively; all P ⬍ 0.01). The enrollment in the two groups was equal over time and the accrual rate is shown in Figure 2 (KP ⫽ 8.6 transplanted patients per years; KA ⫽ 1.6 transplanted patients per years). Cardiovascular outcome in transplanted patients A lower rate of hypertensive patients at one year—but not at two years—was observed in kidney–pancreas than in kidney–alone transplanted patients (Table 3). Moreover, in the kidney-alone a higher rate of acute heart failure and myocardial infarction was observed as compared to the kidney–pancreas group (Table 3). No differences were shown for episodes of angina pectoris (Table 3). Left ventricular function Ejection fraction, assessed in 23 kidney–pancreas and in 13 kidney–alone was similar at three months after transplantation (Table 4). An improvement was shown in kidney–pancreas after two and four years of follow-up, whereas in the kidney–alone group the ejection fraction remained stable (Table 4). Left ventricular diastolic function assessed at four years was normal in the kidney– pancreas but impaired in the kidney–alone patients (Table 5). In particular, the peak filling rate, peak ejection rate and peak filling rate/peak ejection rate ratio were normal in the kidney–pancreas group and impaired in the kidney–alone group (Table 5). Time-to-peak filling rate was significantly higher in the kidney–alone group than in kidney–pancreas. Heart rate and arterial blood pressure values were similar in the groups at the time of diastolic assessment. Total cholesterol, creatinine and cyclosporine were similar in the groups studied, whereas kidney-pancreas transplanted patients showed lower triglicerydes and glycated hemoglobin (Table 5). The hypertension rate was lower in KP patients, although this

1967

La Rocca et al: Kidney–pancreas transplant outcomes

Fig. 1. Patient actuarial survival rate in 130 kidney–pancreas (KP; 䉬), in 25 kidney–alone (KA; 䊏), and in 196 dialyzed type 1 diabetic (WL; 䉱) patients. Kidney–pancreas versus dialyzed, P ⫽ 0.01, d.f. ⫽ 5.8. Patients included in the analysis each year for the two groups of transplanted patients are: year 1, 113 KP and 23 KA; year 2, 94 KP and 21 KA; year 3, 85 KP and 18 KA; year 4, 73 KP and 16 KA; year 5, 62 KP and 12 KA; year 6, 47 KP and 9 KA; year 7, 36 KP and 6 KA.

Table 4. Evaluation of ejection fraction and glycated hemoglobin during the follow-up in 23 kidney–pancreas and 13 kidney–alone transplanted patients Kidney–pancreas transplanted patients (N ⫽ 23)

3 1 2 4

Fig. 2. Distribution by years of kidney-alone (KA; ) and kidney– pancreas (KP; 䊏) transplantation at Hospital San Raffaele.

Table 3. Rate of hypertensive patients and cardiovascular events after kidney–pancreas or kidney–alone in uremic type 1 diabetic patients

Hypertensive patients at 1 year Hypertensive patients at 2 years Acute myocardial infarction Acute heart failure Angina pectoris NS is not significant.

Kidney–pancreas transplanted patients (N ⫽ 130)

Kidney–alone transplanted patients (N ⫽ 25)

P value

50%

80%

⫽ 0.01

64%

84%

⫽ 0.08

3% (N ⫽ 4/130) 3% (N ⫽ 4/130) 2.3% (N ⫽ 3/130)

20% (N ⫽ 5/25) 20% (N ⫽ 5/25) 8% (N ⫽ 2/25)

⫽ 0.01 ⫽ 0.01 NS

months year years years

Kidney–alone transplanted patients (N ⫽ 13)

Glycated hemoglobin %

Ejection fraction %

Glycated hemoglobin %

Ejection fraction %

5.9 ⫾ 0.1c 5.9 ⫾ 0.9c 6.0 ⫾ 0.2c 6.0 ⫾ 0.1c

67.0 ⫾ 0.8c 70.1 ⫾ 1.0a 71.6 ⫾ 1.1ab 76.6 ⫾ 1.1ab

8.5 ⫾ 0.3c 9.0 ⫾ 1.3c 9.1 ⫾ 0.4c 8.6 ⫾ 0.4c

63.6 ⫾ 2.7 65.6 ⫾ 2.6 65.9 ⫾ 3.1b 65.6 ⫾ 1.4b

a Ejection fraction at 3 months was statistically different versus 1, 2 and 4 years (P ⬍ 0.001) b Ejection fraction was statistically different between kidney–pancreas and kidney–alone at 2 years (P ⫽ 0.04) and 4 years (P ⫽ 0.0001) c Glycated hemoglobin was statistically lower in kidney–pancreas than in kidney–alone groups during the whole follow-up (P ⬍ 0.001)

was not significant. In the kidney–alone group, diastolic abnormalities were positively associated with a higher rate of acute heart failure episodes (KP ⫽ 0 vs. KA ⫽ 4, ␹2 ⫽ 3.6, d.f. ⫽ 1, P ⫽ 0.02). Glycated hemoglobin, evaluated in both groups of transplanted patients was negatively associated with ejection fraction at two and four years (Fig. 3). Moreover, in transplanted patients, glycated hemoglobin was negatively associated with the peak filling rate and the peak filling rate/peak ejection rate ratio, while it was positively associated with the time-to-peak filling rate (Fig. 4). Subanalysis of kidney-alone group Kidney–alone group (25 patients) was constituted of patients who received only the kidney due to macroscopic damage of the pancreas at harvesting. Survival, car-

1968

La Rocca et al: Kidney–pancreas transplant outcomes

Table 5. Left ventricular function and metabolic parameters, evaluated at 4 years in 23 kidney–pancreas transplanted patients and in 13 kidney–alone transplanted patients Kidney–pancreas Kidney–alone transplanted transplanted patients patients (N ⫽ 23) (N ⫽ 13) Ejection fraction % Peak ejection rate (EDV/sec) Peak filling rate (EDV/sec) Peak filling rate/peak ejection rate Time to peak filling rate msec Heart rate bpm Blood pressure mm Hg Systolic Diastolic Triglycerides mg/dL Cholesterol mg/dL Creatinine mg/dL Glycated hemoglobin % Hypertensive patients

76.6 ⫾ 1.1a 3.94 ⫾ 0.11a 4.33 ⫾ 0.19a 1.11 ⫾ 0.05b 153.1 ⫾ 6.1b 76.1 ⫾ 2.4

65.6 ⫾ 1.4a 3.41 ⫾ 0.14a 3.11 ⫾ 0.18a 0.91 ⫾ 0.03b 181.9 ⫾ 14.7b 73.5 ⫾ 3.7

135.3 ⫾ 4.5 81.1 ⫾ 2.0 122.5 ⫾ 10.9b 224.8 ⫾ 10.3 1.38 ⫾ 0.09 6.0 ⫾ 0.1c 39.1%

138.2 ⫾ 3.6 85.2 ⫾ 3.1 169.1 ⫾ 19.6b 235.3 ⫾ 16.3 1.65 ⫾ 0.21 8.6 ⫾ 0.4c 69.2%

EDV is end diastolic volume. Kidney–pancreas vs kidney–alone: a P ⬍ 0.01; b P ⬍ 0.05; c P ⬍ 0.001

Fig. 4. Correlation between left ventricular diastolic parameters and glycated hemoglobin at four years. (A) Glycated hemoglobin, evaluated in the entire group of transplanted patients, was negatively related to peak filling rate (r ⫽ ⫺0.5, P ⫽ 0.003) and (B) to the peak filling rate/ peak ejection rate (r ⫽ ⫺0.4, P ⫽ 0.02), while (C) it was positively related to time to the peak filling rate (r ⫽ 0.4, P ⫽ 0.02).

diovascular outcomes and cardiovascular events were similar between kidney–alone and kidney–pancreas failed (Table 6). Kidney–pancreas with early failure of the pancreas showed follow-up characteristics similar to kidney alone rather than to kidney–pancreas group (Fig. 5). Fig. 3. Correlation between ejection fraction and glycated hemoglobin at two (A) and four years (B). Glycated hemoglobin, evaluated in the whole group of transplanted patients, was negatively associated to ejection fraction at 2 (r ⫽ ⫺0.3, P ⫽ 0.03) and 4 years (r ⫽ ⫺0.5, P ⫽ 0.008).

DISCUSSION Our study demonstrates a favorable role of pancreas transplantation toward slowing the progression of some of the features of macrovascular disease in uremic dia-

1969

La Rocca et al: Kidney–pancreas transplant outcomes Table 6. Cardiovascular outcomes in the three groups of patients Cardiovascular death rate

KP KA KP failed

N

N of deaths

119 25 11

9/119 (7.5%) 5/25 (20.0%) 1/11 (9.0%)

Acute myocardial infarction

Acute pulmonary edema

Angina pectoris

N of events 2/119 (1.6%) 4/25 (16.0%) 2/11 (18.1%)

2/119 (1.6%) 5/25 (20.0%) 3/11 (27.2%)

2/119 (1.6%) 2/25 (8.0%) 1/11 (9.0%)

Abbreviations are: KP, kidney–pancreas transplanted patients; KA, kidney–alone transplanted patients; KP failed, kidney–pancreas transplanted patients with pancreas early thrombosis.

Fig. 5. Comparison of the actuarial survival (A) and glycated hemoglobin (B) between the three groups, kidney–pancreas (KP; 䉬), kidney– alone (KA; 䊏) and KP failed (KP failed; 䉱) (KP different vs. other 2 groups).

betic patients, mainly blood pressure control, heart events and cardiac performance. Even if our approach to defining coronary artery disease understimates “silent” ischemic disease, given that angiographic disease can be found in asymptomatic patients with normal echoes and normal noninvasive tests. Clinical characteristics, cardiovascular risk factors and the hypertension rates were similar in the three groups of patients upon entrance into the study, and, therefore, it can be concluded that the different outcome was the consequence of the transplant. Our analysis was confirmed by the outcomes observed in the subgroup of kidney–pancreas failure, which showed that the survival, cardiovascular death rate and events were closer to the kidney–alone group than the kidney–pancreas transplanted patients. Furthermore, it is unlikely that our results can be ex-

plained on the basis of a different transplant accrual rate, given that the enrollment in the two groups was equal over time and that the accrual rate does not explain the results (Fig. 2). In our study, the improvement of hypertension was associated with kidney–pancreas transplant and with the consequent glycometabolic control. This improvement was no longer significant at two years, probably due to the long-term effects of cyclosporine administration to other influencing factors [23, 24]. It was already observed that kidney–pancreas transplantation exerts a positive influence on cardiorespiratory reflexes [25] and on systolic function [10]. It is noteworthy that diastolic filling is frequently impaired in type 1 diabetic patients [14]. In agreement with Gaber’s findings of a significant improvement of cardiac shortening fraction from baseline to 12 months, a statistical increase of ejection fraction during the follow-up was evident in the kidney–pancreas transplanted patients in our series. A stabilization of systolic function was observed for diabetic kidney–alone recipients, but throughout the follow-up no statistical differences were observed in term of ejection fraction. The early improvement of systolic function shown by Gaber in kidney–alone recipients could be underestimated in our patients, since no radionuclide left ventriculography was performed six months after transplantation. Furthermore, the improvement of diastolic function observed in terms of early/active peak velocity and early/ active integral ratio up to five years by the Memphis group, and its association with glycometabolic control, have been confirmed by our findings [11]. Our study showed a normal peak filling rate and peak filling rate/peak ejection rate ratio in patients after kidney–pancreas transplant, but not in the kidney–alone group. The data observed in the kidney–alone group are similar to those observed in uncomplicated type 1 diabetic patients with 15 years of disease, as reported in a previous study using the same methods [14]. Striking differences in volume distribution between the two groups, as a consequence of bicarbonate loss due to bladder diversion of the transplanted pancreas, can be excluded because of appropriate bicarbonate administration in these patients. The close relationship between metabolic control and

1970

La Rocca et al: Kidney–pancreas transplant outcomes

cardiac function is confirmed by the correlation between glycated hemoglobin and cardiac indices when the transplanted patients’ data are pooled together. The peak filling rate/peak ejection rate ratio is close to 1 in healthy subjects, athletes and compensated type 1 diabetic patients; on the contrary it is reduced in coronary and hypertensive patients [14]. The diastolic index seems to be more reliable in identifying an underestimated cardiac dysfunction that could not be detected with the assessment of ejection fraction alone. It is well known that metabolic alterations lead first to diastolic disfunction than to systolic disfunction. More than 30% of patients with diabetic chronic heart failure presented a normal systolic function, with an isolated diastolic impairment [14]. Impaired diastolic filling and, in particular, the isolated presence of abnormal ventricular filling is clearly shown by the lower peak filling rate and peak filling rate/ peak ejection rate ratio in the kidney–alone group. This might explain why these patients are prone to acute heart failure with pulmonary edema. Our findings show that by resolving their diabetic condition it is possible to partly reverse diabetic cardiomyopathy in these patients. Normalization of glycometabolic control could directly affect myocardial metabolism, hypertension, progression of coronary artery disease and diabetic neuropathy. Whether these results are the consequence of tight glycemic control or a direct effect of insulin/C-peptide secretion at the cardiovascular level remains to be clarified. In particular, it is interesting to note that the systemic drainage of the pancreatic graft with peripheral hyperinsulinemia could expose the patients to the risk of hyperinsulinemia [26, 27]. Portal drainage with first pass of insulin through the liver could lead to more physiologic insulin levels with substantial reduction of dyslipidemia and probably of coronary artery disease, as shown by Gaber’s group [26, 27]. The lower incidence of cardiovascular events found in the kidney–pancreas patients compared with the kidney–alone patients is consistent with the observation of diastolic function improvement in the same population, thus confirming the role played by glucose metabolism on cardiovascular performance. The stabilization of cardiac function observed in the kidney–alone group is most probably related to the control of fluid volume and the partial reduction of ventricular afterload, which accompanies successful renal transplantation [28]. Finally, the rate of lethal neoplasm observed in our study is similar to that reported for uremic patients [28], showing that kidney–pancreas and kidney–alone transplantation does not enhance neoplasm death rate when compared to dialyzed patients. In conclusion, kidney–pancreas transplanted patients showed a better survival than patients transplanted with the kidney–alone and than patients on a waiting list. This was associated with a low prevalence of cardiovascular events in those patients receiving a kidney–pancreas trans-

plant. This low incidence of cardiovascular events clearly was associated with an enhanced ejection fraction, better blood pressure control and normalization of diastolic filling. Therefore, we conclude that the combination of pancreas transplantation with kidney transplantation improves survival and cardiovascular outcome in uremic diabetic patients. ACKNOWLEDGMENT This work was partially supported by the Italian Ministry of Health (Ricerca corrente). Reprint requests to Antonio Secchi, M.D., Medicina I, San Rafaele Scientific Institute, Via Olgettina 60, 20132 Milano, Italy. E-mail: [email protected]

REFERENCES 1. Ryan EA: Pancreas transplant: For whom? Lancet 351:1072–1073, 1998 2. Luzi L: Pancreas transplantation and diabetic complications. N Engl J Med 339:115–117, 1998 3. Remuzzi G, Ruggenenti P, Maurer SM: Pancreas and kidney/ pancreas transplants: Experimental medicine or real improvement? Lancet 343:27–31, 1994 4. Landgraf R: Impact of pancreas transplantation on diabetic secondary complications and quality of life. Diabetologia 39:1415– 1424, 1996 5. Martinenghi S, Dell’Antonio G, Di Carlo V, Pozza G, et al: Cancer arising after pancreas and/or kidney transplantation in a series of 99 diabetic patients. Diabetes Care 20:272–275, 1996 6. London NJ, Farmery SM, Will EJ, et al: Risk of neoplasia in renal transplant patients. Lancet 346:403–406, 1995 7. Martinenghi S, Comi G, Galardi G, Di Carlo V, et al: Amelioration of nerve conduction velocity following simultaneous kidney/ pancreas transplantation is due to the glycaemic control provided by the pancreas. Diabetologia 40:1110–1112, 1997 8. Ramsay CR, Goetz FC, Sutherland DER, et al: Progression of diabetic retinopathy after pancreas transplantation for insulindependent diabetes mellitus. N Engl J Med 318:208–214, 1988 9. Fioretto P, Steffes MW, Sutherland DER, et al: Reversal of lesions of diabetic nephropathy after pancreas transplantation. N Engl J Med 339:69–75, 1998 10. Gaber AO, el-Gebely S, Sugathan P, et al: Early improvement in cardiac function occours for pancreas–kidney but not diabetic kidney-alone transplant recipients. Transplantation 59:1105–1112, 1995 11. Gaber AO, Wicks MN, Hathaway DK, Burlew BS: Sustained improvements in cardiac geometry and function following kidneypancreas transplantation. Cell Transplant 9:913–918, 2000 12. Balkau B, Pyorala M, Shipley M, et al: Non cardiovascular disease mortality and diabetes mellitus. Lancet 350:1680, 1998 13. Shapiro LM, Leatherdale BA, Mackinnon J, Fletcher RF: Left ventricular function in diabetes mellitus. II. Relation between clinical features and left ventricular function. Br Heart J 45:129–132, 1981 14. Astorri E, Fiorina P, Gavaruzzi G, et al: Left ventricular function in insulin-dependent and in non-insulin-dependent diabetic patients: radionuclide assessment. Cardiology 88:152–155, 1997 15. Shapiro LM: Echocardiographic features of impaired ventricular function in diabetes mellitus. Br Heart J 47:439–444, 1982 16. Garcia MJ, Mc Namara PM, Gordon J, Kannel WB: Morbidity and mortality in diabetics in the Framingham population sixteen year follow-up study. Diabetes 23:105–111, 1974 17. Page MMcB: Watkins PJ: The heart in diabetes: Autonomic neuropathy and cardiomyopathy. J Clin Endocrinol Metab 6:377–388, 1997 18. Starling MR: Does a clinically definable diabetic cardiomyopathy exist? J Am Coll Cardiol 15:1518–1520, 1990

La Rocca et al: Kidney–pancreas transplant outcomes 19. Galderisi M, Anderson KM, Wilson PWF, Levy D: Echocardiographic evidence for the existence of a distinct diabetic cardiomyopathy (The Framingham Heart Study). Am J Cardiol 68:85–89, 1991 20. Secchi A, Di Carlo V, Pozza G: Pancreas and islet transplantation current progresses, problems and perspectives. Horm Metab Res 29:1–8, 1997 21. Pozza G, Bosi E, Secchi A, et al: Metabolic control of Type 1 (insulin-dependent) diabetes after pancreas transplantation. Br Med J 291:510–513, 1985 22. Prieta M, Sutherland DER, Goete FC, et al: Pancreas transplant results according to the technique of duct management: Bladder versus enteric drainage. Surgery 84:633–639, 1987 23. Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: The sixth report of the Joint National Committee on Prevention, Detection, Evalua-

24. 25.

26. 27. 28.

1971

tion, and Treatment of High Blood Pressure. Arch Intern Med 157: 2431–2446, 1997 Curtis JJ, Luke RG, Jones P, et al: Hypertension after successful renal transplantation. Am J Med 85:134–138, 1988 Navarro X, Kennedy WR, Loewenson RB, Sutherland DER: Influence of pancreas transplantation on cardiorespiratory reflexes, nerve conduction, and mortality in diabetes mellitus. Diabetes 39: 802–806, 1990 Stratta RJ, Gaber AO, Shokouh-Amiri MH, et al: A prospective comparison of systemic-bladder versus portal-enteric drainage in vascularized pancreas transplantation. Surgery 127:217–226, 2000 Stratta RJ, Gaber AO, Shokouh-Amiri MH, et al: Experience with portal-enteric pancreas transplant at the University of Tennessee–Memphis. Clin Transplant 84:239–253, 1998 Maisonneuve P, Agodoa L, Gellert R, et al: Cancer in patients on dialysis for end-stage renal disease: An international collaborative study. Lancet 354:93–99, 1999