Renal recovery from acute tubular necrosis requiring renal replacement therapy: a prospective study in critically ill patients

Transcription

1 Nephrol Dial Transplant (2006) 21: doi: /ndt/gfk069 Advance Access publication 31 January 2006 Original Article Renal recovery from acute tubular necrosis requiring renal replacement therapy: a prospective study in critically ill patients Helmut Schiffl Department of Internal Medicine, University of Munich, Munich, Germany Abstract Background. Data on the incidence of end-stage renal disease (ESRD) resulting from irreversible acute tubular necrosis (ATN) are controversial. This prospective cohort study was designed to assess the need for short- and long-term dialysis in critically ill patients with severe ATN and to define risk factors for lack of renal recovery. Methods. 433 consecutive patients with clinically diagnosed severe ATN necessitating renal replacement therapy were enrolled. Eight patients were excluded because renal biopsy revealed another cause of acute renal failure. None of the remaining 425 patients had pre-existing chronic renal insufficiency. Primary outcome criteria were recovery of renal function at discharge and ESRD status at 1 year follow-up. Results. The overall in-hospital mortality of the cohort was 47%. At discharge, 57% of the 226 surviving patients had normal renal function, 33% had mild to moderate renal failure (serum creatinine: mg/dl) and 10% had severe renal failure (serum creatinine: 3 6 mg/dl). Multivariate analysis showed that neither patient characteristics (age, gender, comorbid conditions), severity of illness (APACHE III, number of failed organs) nor mode and duration of renal replacement therapy were related to recovery of renal function. After 1 year, 76 of the surviving patients had died and in one patient chronic renal failure had progressed to ESRD. Conclusions. If critically ill patients with normal renal function prior to the renal insults survive the precipitating cause of ATN, the overwhelming majority will recover sufficient renal function. Keywords: acute tubular necrosis; dialysis; end-stage renal disease; renal recovery Correspondence and offprint requests to: Prof. Dr H. Schiffl, KfH Nierenzentrum Mu nchen Laim, Elsenheimerstr. 63, D Munich, Germany. hschiffl@hotmail.com Introduction The clinical spectrum of acute renal failure (ARF) in critically ill patients is changing. Nowadays, patients are burdened with pre-existing comorbid diseases and face extensive extra-renal organ complications concurrent with the ARF episode. This explains the high demand for renal replacement therapy (RRT) in these patients. Acute tubular necrosis (ATN) is the most common form of hospital-acquired ARF. It accounts for >90% of ARF cases in the intensive care unit and may have multiple causes. Most often, ATN arises from septic, ischaemic or toxic insults to the kidneys. Despite significant advances in critical care medicine and renal replacement techniques, the mortality rate in patients with ATN who require RRT remains50% [1]. ATN is a potentially reversible process, but patients with ATN requiring RRT often die before renal recovery as a result of the severity of the underlying illness or of lethal extra-renal complications of ATN. In the majority of patients who survive, recovery of life-sustaining renal function can be expected. In a study conducted in the 1970s, 3% of survivors of ATN remained dialysis-dependent, 10% had severe chronic renal failure (serum creatinine >3 mg/dl) and a further 25% had moderate chronic renal insufficiency (serum creatinine between 1.5 and 3.0 mg/dl) [2]. A low frequency of irreversible ATN (<10%) was typical for the time period [2,3]. However, more recently, Bhandari and Turney [4] showed that irreversible ARF (predominantly ATN) requiring chronic dialysis therapy had increased from 3.7% in 1984 to 18.2% in 1995 in surviving patients. More recent reports revealed an even higher number of patients (33 68% at discharge) whose kidneys failed to recover and who needed long-term dialysis [5 8]. The changing renal outcome may be the consequence of the dramatically increasing number of older patients with a large burden of comorbid conditions or the result of more severe forms of ARF seen nowadays or a complication of the more aggressive renal replacement therapies currently used. The long-term renal prognosis ß The Author [2006]. Published by Oxford University Press on behalf of ERA-EDTA. All rights reserved. For Permissions, please journals.permissions@oxfordjournals.org

2 Renal recovery from acute tubular necrosis 1249 of ATN is unclear, because of the paucity of long-term follow-up studies. This observational study of 425 consecutive patients with dialysis-dependent ATN was designed to define the frequency of end-stage renal disease (ESRD) due to irreversible ATN and to characterize risk factors associated with irreversible ATN. Subjects and methods Study cohort Between October 1990 and October 2001, 433 consecutive critically ill patients with presumed ATN requiring RRT were recruited at the dialysis unit of the Department of Internal Medicine, University of Munich. All adult patients with ARF were eligible for the study; however, only those patients who had a clinical diagnosis of ATN (made by the attending nephrologist) and did not participate in other studies and who consented to participate in the study were included. Furthermore, patients were excluded from the study if they had pre-renal azotaemia, obstructive nephropathy, suspected acute renal parenchymal disease (acute interstitial nephritis, acute primary or secondary glomerulonephritis, vasculitis), renal transplant or pre-existing chronic renal insufficiency. Chronic kidney disease was defined as structural abnormalities in imaging studies as well as a continuous presence of abnormal urine tests and/or repeated documentation of raised serum creatinine levels (>1.3 mg%) prior to hospitalization and/or decreased levels of kidney function [glomerular filtration rate (GFR) <90 ml/min/1.73 m 2, irrespective of age]. GFR was calculated by the Cockcroft Gault formula and/or estimated from the serum creatinine concentration using an equation derived from the Modification of Diet in Renal Disease Study. ATN was differentiated from intrinsic renal diseases by detailed and accurate history, thorough physical examination, pertinent laboratory examinations and imaging studies. A fractional excretion of sodium of >2% and the presence of granular casts on microscopic examination of the urine sediment were considered typical for ATN. Renal biopsy was performed in eight patients, because the clinical presentation was considered atypical for ATN or the course of ARF was unduly prolonged. Renal biopsy revealed cholesterol embolism in five, Goodpasture s syndrome in two and Wegener s granulomatosis in one patient. Accordingly, these eight patients were excluded from the study. Renal replacement therapy Conventional haemodialysis (IHD) was performed using volumetrically controlled ultrafiltration (MTS 2008 C; FMC, Bad Homburg v. d. Ho he, Germany) with single-use hollow fibre dialysers [both cellulosic (n ¼ 116) or synthetic membranes (n ¼ 309)], heparin anticoagulation (except for patients with bleeding diathesis), bicarbonate-based dialysate, dialysate flow rates at 500 ml/min and effective blood flow rates of ml. The dialysate sodium, potassium, calcium and bicarbonate were varied depending on haemodynamic status and pre-dialysis laboratory parameters. Continuous techniques, particularly continuous high-flux veno-venous haemodialysis or continuous arteriovenous haemofiltration (CAVH) were performed with hollow-fibre haemofilters. Patients underwent RRT for one or more of the following indications: volume overload with pulmonary oedema inadequately controlled with diuretics; hyperkalaemia refractory to conservative measures; the need for hyperalimentation in the presence of insufficient urinary output; uraemic signs or symptoms for which uraemia could not be ruled out as a precipitating cause; and/or a blood urea nitrogen concentration >100 mg/dl. IHD was performed daily, on alternate days or with a varying number of sessions per week, according to estimates of catabolism or fluid status. RRT was terminated when blood urea nitrogen or serum creatinine fell spontaneously and/or urine output increased significantly. Clinical data The nature of the study was explained in detail to the patients or their next of kin; they all consented to participate in the observational study. The following variables were recorded at the first day of initiation of RRT: age, gender, comorbidity, the presumed main cause (this categorization does not deny the fact that more than one precipitating factor is operating) of ATN (sepsis, hypotension/ischaemia, exogenous or endogenous nephrotoxins), renal laboratory parameters, type of RRT (intermittent or continuous), type of dialysis membrane (cellulosic or synthetic) and APACHE III score. The primary outcome criterion of the study was recovery of renal function at discharge, defined either as the complete return of increased serum creatinine to baseline values or as partial recovery of renal function when return of serum creatinine to baseline was not met, but the patient was no longer dialysis-dependent. Failure to recover renal function was defined as dialysis-dependence at discharge and ESRD as continued dialysis dependence after 1 year. The dialysis status after 1 year was assessed by telephone interview or postal questionnaire in all patients. Statistical analysis Data at initiation of RRT for ATN are given as means±sd or expressed as percentage of the patient cohort. Differences between patient groups (complete recovery of renal function or partial renal recovery) were analysed by the nonparametric Wilcoxon Mann Whitney test for continuous variables and by Fisher s exact test for categorical variables. Multivariable logistic regression analysis was conducted to investigate risk factors for incomplete recovery of renal function. The following variables were investigated as independent risk factors using backward elimination: age, gender, severity of illness (Apache III, number of failing organs), cause of ATN, type and duration of renal replacement and biocompatibility of the dialyser membrane. A two-sided P-value of 0.05 was considered significant. All statistical analyses were performed using SPSS (SPSS Inc., Chicago, IL, USA). Results Baseline characteristics of the study population Four hundred and twenty-five critically ill patients with ATN necessitating RRT participated in this observational study. Complete data were obtained on

3 1250 H. Schiffl Table 1. Patient characteristics at recruitment Number of patients 425 Age (years) 65±12 Gender Male (%) 66 Female (%) 34 Comorbid conditions Hypertension (%) 48 Diabetes mellitus (%) 22 COPD (%) 14 Chronic heart failure (%) 18 Presumed main cause of ATN Ischaemia (%) 60 Sepsis (%) 33 Nephrotoxins (%) 7 Setting of ATN Surgical (%) 41 Medical (%) 59 Markers of ARF Serum creatinine (mg/dl) 4.3±1.1 Blood urea nitrogen (mg/dl) 83±32 Serum potassium (mmol/l) 4.9±0.9 Urine output (l/day) 1.1±0.8 Oliguria (%) 48 Severity of illness APACHE III score 88±29 Number of failing organs 2.4±0.2 Mode of RRT IHD (%) 68 CRRT (%) 13 IHD þ CRRT (%) 19 Values are means±sd, unless indicated otherwise. CRRT ¼ Continuous renal replacement therapy COPD ¼ Chronic obstructive pulmonary disease each patient. Normal renal function was documented not only by serum creatinine values (mean: 0.8± 0.2 mg/dl) but also by estimated GFR (113±12 ml/ min/1.73 m 2 ). Table 1 shows the baseline data at the start of RRT. The patient population was characterized by a high mean age (range: years) and an excessive burden of comorbid conditions, particularly cardiovascular disorders. Surgical patients were recruited mainly from cardiovascular surgery (96 patients). In medical patients, ATN was mostly related to cardiogenic shock (102 patients), sepsis or the use of nephrotoxic drugs. ATN presented as part of the multiple organ failure syndrome (number of failing organs 1 5) in the majority of critically ill patients. The leading mode of RRT was intermittent haemodialysis performed with varying frequency. A significant proportion of the patients received combined RRT, starting either with intermittent or continuous techniques followed by the other mode of extracorporeal detoxification (Table 1). In-hospital outcome of patients with ATN requiring RRT The overall in-hospital mortality of the 425 critically ill patients with severe ATN was 47%. The duration of RRT was 7 days (range: 1 50 days), the length of stay in the hospital as calculated from the time of commencement of RRT was 21 days (range: 6 89 days). Table 2. Characteristics of surviving patients at initiation of haemodialysis, stratified according to the level of recovery of renal function Complete recovery of renal function Partial recovery of renal function Number of patients Serum creatinine (mg/dl) 0.9± ±0.6 Age (years) 63±11 59±13 Gender Male (%) Female (%) Setting Surgical (%) Medical (%) Presumed aetiology of ATN Ischaemia (%) Sepsis (%) Nephrotoxins (%) Severity of illness Apache III score 84±23 87±19 Number of failing organs 2.2± ±0.4 Oliguria (%) Mode of RRT IHD (%) CRRT (%) Synthetic membranes (%) Duration of RRT (days) Median 9 10 Range Values are means±sd, unless indicated otherwise. At discharge, recovery of renal function was complete in 57% of the surviving patients. Thirty-three per cent of these patients had mild to moderate renal failure (serum creatinine >1.3 but <3 mg/dl) and 10% of the survivors had serum creatinine values between 3 and 6 mg/dl but needed no further RRT. Prediction of partial recovery from ATN necessitating RRT There were no statistically significant differences in mean age, distribution of gender, number of patients with comorbid conditions, cause of ATN, mode of RRT or duration of RRT among patients with complete or partial recovery of renal function (Table 2). Multivariate analyses revealed that neither age, gender, comorbidity, severity of illness characterized either by APACHE III score or the number of failed organs, cause of ATN nor mode of RRT or duration of RRT were independently associated with partial recovery of renal function from ATN. One year post-discharge follow-up Follow-up data were available for 222 out of 226 patients leaving the hospital (98%). During the 12 months after discharge, another 76 patients died. The surviving patients at 1 year accounted for 35% of all 425 patients of the cohort. Only one patient discharged with chronic renal insufficiency (serum creatinine 4.2 mg/dl) had progressed to ESRD.

4 Renal recovery from acute tubular necrosis 1251 Thus, <1% of patients surviving ATN needed longterm dialysis therapy at 1 year after discharge. Discussion Recovery of renal function may be a more specific outcome measure than mortality to evaluate patients with ATN. The excessive in-hospital mortality of critically ill patients with severe ATN relates both to lethal extra-renal organ complications of ATN itself and to lethal complications of the underlying disease or comorbid conditions [1]. The data of the present investigation showed that survivors of ATN had varying levels of renal recovery. Indeed, none of the patients remained on dialysis at discharge. A substantially higher frequency of ESRD due to ATN has been reported for several [5 8] but not all recent studies [9 11]. The comparison of the various studies is difficult, because of differences in the nature of the analysis (prospective vs retrospective), number and characteristics of enrolled patients, definition of ATN, severity of renal failure due to ATN and inclusion of patients with pre-existing chronic renal insufficiency (acute or chronic). In contrast to most other studies, the design of the present study was prospective, the patient population was limited to critically ill patients with well-defined ATN, all patients were enrolled on the first day of RRT and the overwhelming majority of patients was followed up for 1 year after discharge from the hospital. By definition, only patients with documented normal renal function prior to the insult to the kidneys were eligible for inclusion in the study. The frequency of ESRD in survivors of ARF has been shown to vary according to the cause of ARF, being highest in patients with acute glomerular disease (primary or secondary glomerulonephritis, vasculitis) or cortical necrosis and lowest in ATN [3,4]. In general, the early recognition of ATN relies mostly on clinical judgement rather than histopathology. Many nephrologists consider a kidney biopsy only when pre- and post-renal causes have been excluded and both clinical setting and presentation do not support the diagnosis of ATN. By comparison with histopathology, careful analysis of the patient s record and review of the history, physical examination, urinalysis and routine laboratory parameters established the cause of ARF correctly in 75 80% of cases [12]. The clinical diagnosis was 86% sensitive for identifying acute tubulointerstitial disorders and 67% sensitive for identifying acute glomerular disorders [13]. In the Madrid ARF study [10], only 4 of 337 patients with ATN (1.2%) underwent biopsy after 3 weeks of ARF. One plausible explanation for the higher incidence of lost renal function found in most studies compared with the present study, may be the fact that patients with biopsy-proven intrinsic renal disease as cause of ARF were excluded from the present but not from other investigations. The fact that the discrepancy of the incidence of ESRD in different studies may be due to the inclusion or exclusion of patients with pre-existing renal disease seems to be of even greater importance. Patients with abnormal renal function prior to the ischaemic or nephrotoxic insult to the kidneys are at a greater risk not only for acute or chronic renal failure but also for ESRD. Hospital-acquired renal insufficiency occurred significantly more often in patients with abnormal renal function (15.7%) than in patients with normal serum creatinine levels (5.3%) at baseline in the study by Nash et al. [14]. It is noteworthy that the majority of studies demonstrating a high frequency of ESRD at discharge from hospital had included patients with pre-existing chronic renal insufficiency [5 7]. The fact that the surviving patients in the present study had no ESRD at discharge is most likely due to the fact that abnormal renal function prior to the insult was an exclusion criterion. From the data collected at recruitment in this singlecentre study, it has not been possible to identify clinical features associated independently with the risk of developing post-atn renal failure or ESRD by multivariate analysis. These findings are in concordance with analyses done by others. Bhandari and Turney [4] found that neither age, gender, ventilation nor clinical severity of ARF appeared to be risk factors for irreversible renal failure. The multiple regression analysis performed by Uehlinger et al. [15], including type of RRT, presence of pre-existing chronic renal failure, haemodynamic stability, use of vasopressor agents and therapy with aminoglycosides as independent variables, revealed that pre-existing chronic renal failure was the only independent predictor of renal recovery. In contrast, the logistic regression analysis for renal recovery conducted by Augustine et al. [16], showed that neither mode of RRT nor baseline mean arterial pressure, number of baseline vasopressors or increased vasopressor dose correlated with renal recovery. However, cardiovascular instability during dialysis was negatively correlated and urine output on day 1 and preserved urine output on day 3 were positively correlated with renal recovery. Finally, Mehta et al. [17] found that severity of illness was an important determinant of renal recovery. They observed the lowest rates of renal recovery among patients with APACHE III scores over 100 compared with patients with APACHE III scores of less than 79 or 80 to 100, respectively. Given the limited and, at least in part, controversial data, persistent post-atn renal failure appears to be unpredictable and may affect any clinical group at any age. The long-term fate of patients surviving ATN requiring RRT is largely unknown. During the first year after discharge from hospital, 34% of the surviving patients died. Although we observed progression of chronic renal failure to ESRD in one patient only, long-term deterioration of renal function may occur in a larger proportion of patients surviving severe ATN. Eggers et al. [18] reported at Renal Week 2004 that hospitalizations with codes of ARF have increased

5 1252 H. Schiffl dramatically during the past 2 years in the USA and concluded that ARF is an important precursor to ESRD. Thirteen per cent of patients with ARF proceeded to ESRD within 3 years. However, both chronic kidney disease status as well as age had a strong impact on the frequency of ESRD following ARF. Compared with 28.2% of patients with pre-existing chronic disease, only 7.6% of patients with normal renal function prior to ARF progressed to ESRD after 3 years. Within this time span, 48.5% of patients older than 85 years but only 11.1% of patients younger than 65 years had ESRD, suggesting that older patients had more often pre-existing kidney disease. The retrospective analysis of Morgera et al. [19] noted that in those patients surviving ARF 10% remained dialysisdependent and >40% had chronic renal failure after 5 years. This study was performed in critically ill patients treated in the intensive care unit and it is reasonable to assume that ATN accounted for most of the causes of ARF. These data are in line with findings by Bonomini et al. [20] that in patients with biopsyproven ATN the frequency of ESRD was 6.4% at 1 year and 11.2% after 5 years of follow-up. Taken together, acute irreversible ATN is not a relevant contributor to ESRD in patients with normal renal function prior to the kidney insults. However, irreversible ATN is much more frequent in patients with pre-existing renal functional impairment. The challenge is to predict whether or not an individual patient will recover renal function from ATN. The adage that kidneys recover if patients survive ATN remains true, at least in those without pre-existing kidney disease. Conflict of interest statement. None declared. References 1. Lameire N, van Biesen W, Vanholder R. Acute renal failure. Lancet 2005; 365: Kjellstrand CM, Ebben J, Davin T. Time of death, recovery of renal function, development of chronic renal failure and need for chronic hemodialysis in patients with acute tubular necrosis. Trans Am Soc Artif Intern Organs 1981; 27: Bonomini V, Frasca GM, Raimondi C et al. Long-term follow-up of acute renal failure. Nephrol Dial Transplant. 1994; 9 [Suppl 4]: Bhandari S, Turney JH. Survivors of acute renal failure who do not recover renal function. Q J Med 1996; 89: Chertow GM, Christiansen CL, Cleary PD, Munro C, Lazarus JM. Prognostic stratification in critically ill patients with acute renal failure requiring dialysis. Arch Intern Med 1995; 155: Cosentino F, Chaff C, Piedmonte M. Risk factors influencing survival in ICU acute renal failure. Nephrol Dial Transplant 1994; 9 [Suppl 4]: Gruberg L, Mehran R, Dangas G et al. Acute renal failure requiring dialysis after percutaneous coronary interventions. Catheter Cardiovasc Interv 2001; 52: Leacche M, Rawn JD, Mihaljevic T et al. Outcomes in patients with normal serum creatinine and with artificial renal support for acute renal failure developing after coronary artery bypass grafting. Am J Cardiol 2004; 93: Jones CH, Richardson D, Goutcher E et al. Continuous venovenous high-flux dialysis in multiorgan failure: a 5-year single-center experience. Am J Kidney Dis 1998; 31: Liano F, Pascual J. Epidemiology of acute renal failure: a prospective, multicenter, community-based study. Madrid Acute Renal Failure Study Group. Kidney Int 1996; 50: Ronco C, Bellomo R, Homel P et al. Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective randomised trial. Lancet 2000; 356: Mustonen J, Pasternack A, Helin H, Pystynen S, Tuominen T. Renal biopsy in acute renal failure. Am J Nephrol 1984; 4: Wilson DM, Turner DR, Cameron JS et al. Value of renal biopsy in acute intrinsic renal failure. Br Med J 1976; 2: Nash K, Hafeez A, Hou S. Hospital-acquired renal insufficiency. Am J Kidney Dis 2002; 39: Uehlinger DE, Jakob SM, Ferrari P et al. Comparison of continuous and intermittent renal replacement therapy for acute renal failure. Nephrol Dial Transplant 2005; 20: Augustine JJ, Sandy D, Seifert TH, Paganini EP. A randomized controlled trial comparing intermittent with continuous dialysis in patients with ARF. Am J Kidney Dis 2004; 44: Mehta RL, McDonald B, Gabbai FB et al. A randomized clinical trial of continuous versus intermittent dialysis for acute renal failure. Kidney Int 2001; 60: Eggers P, Star R, Xue J et al. ARF: an under-recognized contributor to ESRD? In: American Society of Nephrology, ed. Clinical Nephrology Conferences Syllabus: Renal Week Lippincott Williams & Wilkins, Hagerstown: 2004: Morgera S, Kraft AK, Siebert G, Luft FC, Neumayer HH. Long-term outcomes in acute renal failure patients treated with continuous renal replacement therapies. Am J Kidney Dis 2002; 40: Bonomini V, Vangelista A, Frasca G et al. Long term clinical and morphological evaluation of acute renal failure. Adv Exp Med Biol 1987; 212: Received for publication: Accepted in revised form: