Critical limb ischaemia as a main cause of death in patients with end-stage renal disease: a single-centre study

Transcription

1 Nephrol Dial Transplant (2004) 19: doi: /ndt/gfh404 Advance Access publication 20 July 2004 Original Article Critical limb ischaemia as a main cause of death in patients with end-stage renal disease: a single-centre study Michael Koch 1, Rudolf Trapp 1, Wolfgang Kulas 1 and Bernd Grabensee 2 1 Centre of Nephrology Mettmann and 2 Department of Nephrology and Rheumatology, Heinrich Heine University, Du sseldorf, Germany Abstract Background. Patients with end-stage renal disease (ESRD) have a high overall mortality rate, particularly due to cardiovascular morbidity. In an era of decline in cardiovascular diseases and early cardiovascular intervention, non-cardiac diseases seem to have a larger impact on overall mortality. Methods. From 1997 to 2003, all incident haemodialysis patients in a single centre were enrolled in this prospective study. Those with clinical signs of vascular disease were examined by coronary or peripheral angiographies. Physicians took the patients medical histories, examined them and followed them up until the end of the study or death. Causes of death were defined by the physicians. Results. In all, 322 patients were enrolled in the study, 38% of whom were diabetic. At the start of dialysis treatment, 38% had coronary artery disease (CAD), defined as >50% stenosis of at least one coronary artery or as definite myocardial infarction, and 14% had critical ischaemia of at least one limb (CLI). In all patients with foot lesions, CLI was defined angiographically, as evidenced by stenosis or rarefication of distal vessels in the legs. Patients who died (n ¼ 121) [due to cardiac causes (n ¼ 25), complications of CLI (n ¼ 22), stroke (n ¼ 10), cachexia following a long-standing, non-malignant disease (n ¼ 6), malignancy (n ¼ 24), infection not related to CLI (n ¼ 18) and other causes (n ¼ 16)] were older (71±10 vs 65±13 years), more often male [74/121 (61%)] and often diabetic [56/121 (46%)]. CAD was documented in 82/121 (68%). Five-year survivals in patients with no risk and diabetes without CAD or CLI, CAD and CLI were 74%, 73%, 50% and 10%, respectively. Age, CLI and smoking habits independently increased the risk of death (hazard ratios: 1.052, and 2.292, respectively). Correspondence and offprint requests to: Michael Koch, MD, Gartenstrasse 8, Mettmann, Germany. koch@dialyse-mettmann.de Conclusions. These results indicate that CLI with associated complications is not only an indicator of high mortality in patients with ESRD, but is also one of the main causes of death. Keywords: causes of death; coronary artery disease; critical limb ischaemia; end-stage renal disease; mortality; peripheral arterial disease Introduction Haemodialysis patients have a high overall mortality rate. Depending on age, mortality rises from 518 to 924/1000 patient-years [1]. These death rates are comparable with those of colon cancer [2]. High cardiovascular morbidity is reported to be the main cause of death in >50% of these patients [1], though peripheral arterial disease (PAD) and its major complication, critical limb ischaemia (CLI), play minor roles [3]. PAD is common in patients with end-stage renal disease (ESRD), with a frequency of 24 77% [4 7]. The relative risk of death in patients with ESRD and PAD is 1.58 in comparison with ESRD patients with heart disease (1.59) [8]. One would expect that risk stratification, aggressive treatment of risk factors and the new technologies of dialysis facilities would lower the cardiovascular and overall mortalities, as reported in several studies [9,10]. Concomitantly, the incidence of cardiovascular disease seems to have decreased over recent years [11]. Nevertheless, overall mortality in ESRD dialysis patients remains high. Among reasons for this is the fact that the characteristics of dialysis populations have changed during recent years. The greatest contributors to the increase in the number of patients with ESRD have been older age and diabetes [1]. These last two characteristics are also the major risk factors of vascular morbidity and mortality in ESRD [12 14]. Data on the accuracy of the indicated causes of death in patients with ESRD are limited. The purpose of our Nephrol Dial Transplant Vol. 19 No. 10 ß ERA EDTA 2004; all rights reserved

2 2548 M. Koch et al. investigation was to determine prospectively and in a standardized way the causes of death in patients with ESRD and also whether or not clinical signs at the beginning of dialysis treatment might predict death. We are convinced that a more accurate determination of causes of death is of importance in the development of mortality-lowering interventions. Subjects and methods From 1997 to 2003, all eligible and incident haemodialysis patients in a single centre were enrolled in the study, but patients with acute renal failure were excluded. All patients were treated according to a high-quality standard protocol, which included regular quality controls along with antihypertensive therapy to lower systolic blood pressure to <140 mmhg, lipid-lowering therapy to lower low-density lipoprotein cholesterol to <130 mg/dl, optimized blood sugar control to reach HbA1c <7% and supplementation with erythropoietin and Fe. If there were no contraindications, patients received a platelet aggregation inhibitor. Concerning the dialysis, bicarbonate-buffered dialysate was used and potassium was measured for every dialysis session and haemoglobin once a week. The patients were dialysed for a total of 6 18 h/week, depending on residual renal function, and they were examined routinely twice a year. During the observation period, the patients were seen by the same medical doctors, who took their medical histories, examined the patients and followed the patients through the observation period or until their deaths. All patients were seen three times a week. The patients with angina pectoris had a coronary angiography and coronary artery disease (CAD) was diagnosed when a stenosis of >50% was observed in at least one coronary artery. If angiography was not performed, CAD was diagnosed in a patient with a history of myocardial infarction and concomitant changes in the electrocardiogram. Peripheral angiography was performed in all symptomatic patients with foot lesions, such as ulceration or gangrene. CLI was diagnosed if angiography showed one or more of stenosis, slow flow phenomenon, rarefication of one or more of the peripheral arteries and if patients had ulceration or gangrene. Causes of death were defined as follows: Cardiovascular death, if a patient was hospitalized due to myocardial Table 1. Characteristics of 322 incident ESRD patients at baseline infarction with elevated levels of creatine kinase and concomitant changes in the electrocardiogram; sudden death, if a patient died unexpectedly at home; and cardiac death, due to inoperable aortic stenosis, congestive heart failure, etc. A death due to cardiac causes was defined even in a patient who had a malignancy, if a myocardial infarction was the immediate cause of death. Deaths due to CLI were defined if patients died due to sepsis or if they were malnourished after prolonged hospitalizations and were on chronic analgesia and antibiotic therapy, but if one of these patients died due to a myocardial infarction, the death was listed as a cardiovascular death. Statistical analysis Statistical analyses were performed using the SAS system for Windows Õ, version 8.2 (SAS Institute, Cary, NC, USA). CLI and CAD definition at baseline included those patients who developed CLI or CAD 6 months after starting dialysis. Survival time was calculated from the start of dialysis treatment. We censored all cases who did not survive the first 6 months on dialysis and those who received a renal graft during the observation period. Survival and mortality were estimated by life tables and survival distribution function was assessed according to the Kaplan Meier method. To compare patients with different risks at baseline, we formed four risk groups: patients with CLI, with CAD or with diabetes and, in cases where no former risks were present, patients with no risks. This implies that, for example, patients with CLI could have had diabetes or CAD concomitantly or patients with CAD could have had diabetes. The Cox proportional hazards model was used to prove independency of covariates such as age, sex, diabetes, smoking, CLI and CAD at baseline. Results We examined 322 patients with a mean age of 67 years (range: years) at the start of dialysis treatment, of whom 57% were males and 38% were diabetics. At baseline (6 months after initiating dialysis), 38% of the patients were found to have CAD. Patients with CAD were older, were more likely to be men All patients Diabetes No diabetes CAD No CAD CLI No CLI Patients (n) Age (years) 67 ± 12 69±9 66±14 71±10 65±13 a 67±10 67±13 % Diabetes b a % Male b b % Smoker c % CLI c 25 7 a % CAD b a % Low aspirin d % Statin c b Weight (kg) 74±16 75±18 73±16 73±14 74±16 79±20 73±16 d Systolic RR (mmhg) 138±17 138±19 138±16 138±17 139±17 134±16 139±18 Diastolic RR (mmhg) 77±7 76±7 78±7 76±7 78±7 75±7 77±8 Chi-square and Student s t-test: a P<0.0001; b P<0.01; c P<0.001; d P<0.05. Systolic and diastolic RR is self declared.

3 Critical limb ischaemia in ESRD 2549 Table 2. Prevalence (at baseline) and incidence (at follow-up 6 12 and >12 months later) of cardiac and peripheral events in 322 incident ESRD patients Total Time 0 prevalence (þ6 months) Time 1 incidence (6 12 months) Time 2 incidence (>12 months) CAD MI Angioplasty CABG CLI Gangrene Angioplasty Bypass Minor amputation Major amputation MI, myocardial infarction. CABG, coronary artery bypass graft. Table 3. Causes of death in incident ESRD patients with CLI or CAD at baseline and, in those who developed CLI or CAD, during follow-up Total (n ¼ 322) CLI prevalent (n ¼ 34) and to have diabetes mellitus, were often smokers, suffered more often from CLI and were more often on aspirin and statins than patients without CAD (Table 1). On the other hand, 14% of the patients had CLI at baseline. Patients with CLI were more frequently men and diabetic, had higher body weights and were more likely to be on statins than patients without CLI (Table 1). The prevalence (at baseline) and incidence (>12 months) of vascular events are listed in Table 2. Most of the patients started dialysis treatment with a tremendous burden of vascular damage. During follow-up, only a small proportion of patients developed new vascular events (probability of denovo CLI see fig. 2), 121 patients died and 12 patients received kidney grafts. Those who did not survive 6 months (n ¼ 41) were censored. No patient was lost to follow-up. The causes of death and their incidences are listed in Table 3. The most important cause of death was cardiovascular (21%) deaths that followed myocardial infarction (n ¼ 6), sudden death (n ¼ 14) and other cardiac causes (n ¼ 5). Death following CLI was observed in 22 out of 121 patients (18%) (Table 3). All of those who died due to CLI had non-healing foot lesions; C-reactive CLI incident (n ¼ 25) CAD prevalent (n ¼ 122) CAD incident (n ¼ 46) Number of deaths 121 (38%) 22 (65%) 15 (60%) 62 (50%) 20 (43%) Cardiovascular 25 (21%) (27%) 5 (25%) Myocardial infarction Sudden death Other cardiac causes CLI 22 (18%) 13 (59%) 9 (60%) 12 (19%) 7 (35%) Cerebral vascular 10 (8%) 1 (6%) 0 5 (8%) 2 (10%) Long-standing disease 6 (5%) 1 (6%) 0 4 (10%) 1 (5%) Malignancy 24 (20%) 2 (9%) 1 (7%) 8 (13%) 2 (10%) Infection/sepsis a 18 (13%) 1 (6%) 2 (13%) 7 (6%) 2 (5%) Other causes 16 (12%) 0 1 (7%) 9 (15%) 1 (5%) a Not related to CLI. Table 4. Characteristics of 22 ESRD patients who died due to CLI Long-standing therapy with antibiotics (>30 days) n ¼ 18 Hospitalization per patient (days/last 6 months) 39±18 Chronic use of analgesic drugs n ¼ 22 Weight loss (>5 kg/last 6 months) n ¼ 15 C-reactive protein before death (mg/l) 105±34 Haemoglobin before death (g/%) 10±1.2 Angioplasty n ¼ 14 Surgical revascularization n ¼ 4 Major amputation n ¼ 13 Minor amputation n ¼ 4 Time from amputation until death in 30±42 15 patients (days) Multiresistant Staphylococcus n ¼ 6 protein was high in a majority of these patients. These 22 patients had prolonged hospitalizations, were bedridden and on chronic antibiotic therapy, colonized by multiresistant bacteria, needed daily analgesia, suffered from significant weight loss and had multiple surgical interventions with local wound debridement, necrectomy, angioplasty, bypass grafting and minor and major amputations (Table 4). Of patients with CLI and major amputations, 10 of 26

4 2550 M. Koch et al Survival Distribution Function Time STRATA: No risk Diabetes CAD CLI all pts no risk Diabetes CAD CLI Fig. 1. Survival in 322 ESRD patients with no risk, diabetes, CAD and CLI during 5 years of observation. Table 5. Stepwise proportional hazards model with age, sex diabetes, smoker, CAD and CLI as explanatory variables Variable Number 2 P-value Hazard ratio SE CLI < Age < Smoker survived. No patient was withdrawn from dialysis. Patients who died tended to be older (71±10 vs 65±13 years), male [74/121 (61%)] and diabetic [56/121 (46%)]. Most of the patients who died were known to have CAD (68%), but cardiovascular death was established in only 21%. Patients with CLI carry the highest mortality risk. According to life-table analyses, the 5 year survival rate in patients with no risk was 74%, in diabetics without CAD or CLI 73%, in CAD 50% and in CLI 10% (Figure 1). The application of the Cox proportional hazards model included significant variables, such as age, diabetes, sex, smoking habits, CAD and CLI, at baseline and it yielded an independent increase in the risk of death for each of CLI, age and smoking (hazard ratios: 4.921, and 2.292, respectively; Table 5). Discussion This study has two important findings. First, if ESRD patients start dialysis with foot lesions, such as ulceration or gangrene resulting from PAD, their risk of dying is elevated, independently of diabetes or CAD. Second, the complications of CLI are among the main causes of death in ESRD patients. All patients who died of these causes were hospitalized and most of them needed morphine for pain. The majority of patients were on prolonged antibiotic therapy, if systemic inflammation was present. Nearly all patients were hospitalized for long periods, during which they underwent angioplasty, surgical reconstruction and, as a last option, amputation. Predominantly, these patients were old, diabetic and had stable CAD. Their deaths did not follow cardiovascular events, such as myocardial infarction, etc.; these patients died due to infection, sepsis or multiorgan failure after being bedridden. Our results are in accordance with the results of Jaar et al. [15], who found that older age and preexisting PAD were independent risk factors for sepsis and death in ESRD patients. The long-term outcome in patients with PAD and ESRD is worse. In patients with infra-inguinal bypass grafting, 2 year survival was only 42% [16]. In view of this markedly increased mortality, reconstructive intervention for these patients must be considered highly palliative. In multivariate analysis, those patients in our study with CLI had the highest mortality risk. Age at the start of dialysis treatment, smoking habits and CLI independently increased the risk of death according to the Cox proportional hazards model. Surprisingly, CAD did not meet the significance level of P ¼ 0.05 for entry

5 Critical limb ischaemia in ESRD 2551 Fig. 2. Probability for de novo CLI in 322 ESRD patients, depending on time on dialysis. into the model. An explanation of this phenomenon could be that CAD was frequent, despite strict diagnostic criteria, but deaths due to CAD were relatively few. On the other hand, CLI was less frequent than CAD, but death due to CLI was relatively high. The death rate due to CLI in our study is relatively high compared with corresponding numbers in the USRDS [3] and the HEMO Study [5]. The difference may be due to differences in population sampling. For instance, the HEMO Study excluded patients with chronic inflammation (serum albumin <2.6 g/l) and patients older than 80 years. It is possible that comorbid conditions, such as PAD or CLI, are under-reported [17]. A recently published study [18] has demonstrated the importance of PAD concerning mortality rates in ESRD patients. The 5 year survival rates of coronary artery bypass-grafted patients without renal failure, with renal failure and with renal failure and PAD or diabetes were 83.5%, 78.5% and 42%, respectively. In our study, patients with CLI had a 5 year survival of only 10%, independent of concomitant diabetes or CAD. The above-mentioned study does not refer at all to the possible causes of deaths in its cohort. One could speculate that the high proportion of deaths due to CLI is a single-centre bias. In a large study in Britain of 1718 patients undergoing angioplasty due to PAD, 16/69 (23%) patients with chronic renal failure died within 30 days after vascular reconstruction; the highest mortality was found in patients undergoing emergency surgery [19]. The median survival of ESRD patients undergoing arterial reconstructive surgery was 1.72 years compared with 5.17 years in a general population [20]. Our results indicate that an early amputation of a critical ischaemic limb would probably prevent death, but our experience is that only a small proportion of patients agree to this radical procedure. Treatment options for these patients are few and further studies should focus on the question of whether or not patients have a better chance of surviving with an amputation and, if so, whether or not there is a critical time to perform the amputation. Conflict of interest statement. None declared. References 1. Herzog CA, Ma JZ, Collins AJ. Poor long-term survival after acute myocardial infarction among patients on long-term dialysis. N Engl J Med 1998; 339: Murphy TK, Calle EE, Rodriguez C, Kahn HS, Thun MJ. Body mass index and colon cancer mortality in a large prospective study. Am J Epidemiol 2000; 152: VI. Causes of death in ESRD. Am J Kidney Dis 1999; 34: S87 S94 4. O Hare AM, Hsu CY, Bacchetti P, Johansen KL. Peripheral vascular disease risk factors among patients undergoing hemodialysis. J Am Soc Nephrol 2002; 13: Eknoyan G, Beck GJ, Cheung AK et al. Effect of dialysis dose and membrane flux in maintenance hemodialysis. N Engl J Med 2002; 347: Leskinen Y, Salenius JP, Lehtimaki T, Huhtala H, Saha H. The prevalence of peripheral arterial disease and medial arterial calcification in patients with chronic renal failure: requirements for diagnostics. Am J Kidney Dis 2002; 40: Schleiffer T, Holken H, Brass H. Morbidity in 565 type 2 diabetic patients according to stage of nephropathy. J Diabetes Complic 1998; 12:

6 2552 M. Koch et al. 8. Johnson JG, Gore SM, Firth J. The effect of age, diabetes, and other comorbidity on the survival of patients on dialysis: a systematic quantitative overview of the literature. Nephrol Dial Transplant 1999; 14: Meier-Kriesche HU, Ojo AO, Port FK et al. Survival improvement among patients with end-stage renal disease: trends over time for transplant recipients and wait-listed patients. J Am Soc Nephrol 2001; 12: Port FK, Orzol SM, Held PJ, Wolfe RA. Trends in treatment and survival for hemodialysis patients in the United States. Am J Kidney Dis 1998; 32: S34 S Foley RN, Herzog CA, Collins AJ. Declining incidence of cardiovascular disease in older end-stage renal disease patients in the United States, J Am Soc Nephrol 2002; 13: 432A 12. Berthoux F, Gellert R, Jones E et al. Epidemiology and demography of treated end-stage renal failure in the elderly: from the European Renal Association (ERA EDTA) Registry. Nephrol Dial Transplant 1998; 13 [Suppl 7]: Koch M, Thomas B, Tschope W, Ritz E. Survival and predictors of death in dialysed diabetic patients. Diabetologia 1993; 36: Koch M, Kutkuhn B, Grabensee B, Ritz E. Apolipoprotein A, fibrinogen, age, and history of stroke are predictors of death in dialysed diabetic patients: a prospective study in 412 subjects. Nephrol Dial Transplant 1997; 12: Jaar BG, Hermann JA, Furth SL, Briggs W, Powe NR. Septicemia in diabetic hemodialysis patients: comparison of incidence, risk factors, and mortality with nondiabetic hemodialysis patients. Am J Kidney Dis 2000; 35: Albers M, Romiti M, Braganca-Pereira CA, Fonseca RL, Silva- Junior M. A meta-analysis of infrainguinal arterial reconstruction in patients with end-stage renal disease. Eur J Vasc Endovasc Surg 2001; 22: Longenecker JC, Coresh J, Klag MJ et al. Validation of comorbid conditions on the end-stage renal disease medical evidence report: the CHOICE study. Choices for Healthy Outcomes in Caring for ESRD. J Am Soc Nephrol 2000; 11: Dacey LJ, Liu JY, Braxton JH et al. Long-term survival of dialysis patients after coronary bypass grafting. Ann Thorac Surg 2002; 74: Gerrard DJ, Ray SA, Barrio EA et al. Effect of chronic renal failure on mortality rate following arterial reconstruction. Br J Surg 2002; 89: Reddan DN, Marcus RJ, Owen W-FJ, Szczech LA, Landwehr DM. Long-term outcomes of revascularization for peripheral vascular disease in end-stage renal disease patients. Am J Kidney Dis 2001; 38: Received for publication: Accepted in revised form: