Evaluation of high-flux hemodiafiltration efficiency using an on-line urea monitor.

On-line urea monitoring of the effluent dialysate offers a real-time assessment of dialysis efficiency and metabolic/nutritional characteristics of hemodialysis patients. Quantitative parameters were evaluated by dialysate urea kinetic modeling (DUKM) with an on-line urea sensor in 23 patients treated by high-flux hemodiafiltration (HDF) (215 sessions of 210 to 240 minutes with a mean blood flow rate of 367 +/- 44 mL/min). Overall, the mean effective Kt/V was 1.52 +/- 0.29, the urea mass removed (22.8 +/- 5.5 g/session or 814 +/- 198 mmol/session), the solute removal index (SRI) 73% +/- 6.1%, and the mean normalized protein catabolic rate (nPCR), 1.15 +/- 0.31 g/kg/day. Blood urea kinetic modeling (BUKM), based on pre- and postsession urea concentrations, using equations from Daugirdas and Garred to calculate equilibrated Kt/V and nPCR, respectively, were in good agreement with DUKM, the differences observed appearing not clinically relevant. The variability of evaluated parameters was verified over consecutive sessions for a mean period of 3 weeks in the entire group. Mean variation in Kt/V was 8%; in urea mass removal, 18%; and in nPCR, 18%. When assessed over 1 week in a subgroup of 13 patients, Kt/V and PCR remained relatively stable, and urea mass removal per- and postsession declined from 23.5 +/- 8.0 g (840 +/- 285 mmol) to 18.7 +/- 6.3 g (667 +/- 225 mmol) from the first to the third session of the week, most likely as a consequence of interdialytic intervals. Predialysis urea concentrations followed the same trend. In the current study, DUKM with on-line urea sensor has confirmed that HDF is a highly efficient renal replacement modality; the variability observed in quantitative parameters supports a need for frequent adequacy monitoring. On-line urea monitoring of effluent dialysate is a simple device that provides the opportunity to tailor treatment to patient needs.

Platelet-activating factor production during hemodialysis: effect of BN 52021.

Platelet activating factor (PAF) production and platelet-lipoxygenase activity were studied during hemodialysis (HD) with cuprophane membranes. Six patients were treated with first-use dialyzers (FU), and 6 patients with reused dialyzers (RU). In a random and double-blind design, 2 HD were performed for each patient, with or without BN 52021 pretreatment, a selective PAF antagonist. Platelet and leukocyte counts were performed before pretreatment and 30 min before HD starting (T-30), at the beginning of HD (T0) and after 15 and 30 min of HD (T15, T30). PAF production was analyzed by direct phase HPLC. To determine platelet-lipoxygenase activity, 12-HETE was detected by reverse phase high performance liquid chromatography (HPLC) after blood stimulation by the ionophore A23187. In the FU group, PAF and 12-HETE were produced during the first 30 min of HD. After BN 52021 pretreatment, PAF production was suppressed and platelet-lipoxygenase activity reduced. In the RU group, neither PAF nor 12-HETE production occurred, and BN 52021 had no effect. We conclude that PAF, which was involved in both platelet and leukocyte activation that occurred during hemodialysis, can be considered as a bio-incompatibility marker.

Plasma acetate levels during hemodialysis.

Before dialysis, acetate levels in hemodialyzed patients (0.27--1.1 mmol/1) were more dispersed than in normal subjects (0.20--0.65 mmol/l) and the mean value of plasma acetate was slightly higher (0.52 mmol/l versus 0.31 mmol/l). Though dialysis conditions were almost identical, the acetate kinetics during hemodialysis were very different: in most subjects, plasma acetate concentrations reached a "plateau" (mean value 5.6 mmol/l) whereas in others a continuous rise was observed, suggesting that with patients having chronic renal failure there were important individual or occasional differences in the ability to metabolize acetate. The acetate loads per minute (or mass transfers) were calculated from the blood compartment with plasma values (plasma flow and concentrations), rather than from the dialysate and using the combined calculations (plasma and whole blood values). The results ranged between 2.4 and 4.1 mmol/min. A very important and rapid fall in arterial acetate concentrations occurs in the first 20 min after the end of the dialysis and proves the rapid turnover of the acetate in man.

MRI abnormalities in tardive dyskinesia.

Most investigators studying tardive dyskinesia (TD) hypothesize that the condition is due to a neurochemical abnormality of the striatum. Recently, numerous CT studies have been done to verify brain abnormalities in patients with TD; the findings have, however, been conflicting. The present study was designed to detect possible neuropathological abnormalities in the basal ganglia in a young sample of schizophrenic patients with TD as compared with schizophrenic patients without TD and normal controls. Magnetic resonance imaging (MRI) was used to measure the volumes of the caudate, putamen, globus pallidus, lateral ventricle, and intracranium. The volumes of the caudate nuclei of the patients with TD were significantly smaller than the volumes of the caudate nuclei of the patients without TD and normal controls. This abnormality in the caudate may be related to some previous conditions, which may prove a substrate that is necessary for TD to establish itself in association with neuroleptic use. Further studies are necessary to confirm our findings and to determine the pathophysiologic nature of these structural alterations and the role played by neuroleptics, whether primary or secondary.

Aluminium in continuous ambulatory peritoneal dialysis and post dilutional hemofiltration.

Aluminium (Al) has been shown to be a constant contaminant of sterile apyrogenic solutions used in continuous ambulatory peritoneal dialysis (CAPD) or post dilutional hemofiltration (HF). The Al content of these solutions varied widely from 1 to 50 micrograms/l. The clinical significance of such an Al contamination was difficult to ascertain. However, the available information suggests little clinical toxicity from these low Al levels, except in those patients who ingest large amounts of aluminium hydroxide as a phosphate binder. In view of the well-known consequences of long-term Al exposure of dialysis patients, it is suggested that Al concentration in CAPD solutions and HF substitution fluids should not exceed 10 micrograms/l.

The effect of an angiotensin antagonist (saralasin) on arterial pressure and plasma aldosterone in hemodialysis-resistant hypertensive patients.

The effect of an angiotensin II antagonist (saralasin) on arterial pressure, plasma renin activity (PRA) and plasma aldosterone concentration (PAC) was assessed in seven dialysis-resistant hypertensive patients. During saralasin infusion performed before hemodialysis, mean arterial pressure fell by 8 to 18.3% in six out of the seven subjects; arterial pressure was normalized in three of them. After hemodialysis (6 subjects), a normal arterial pressure was achieved in five patients. One patient was resistant to saralasin before and after dialysis. A negative correlation (r = 0.62) was obtained between pre-infusion PRA and the change in mean arterial pressure induced by saralasin. Post-infusion PRA increased in saralasin responsive patients, the change in PRA being correlated (r = 0.82) with the pre-infusion PRA. Plasma aldosterone concentration was variably affected by saralasin; a negative correlation between pre-infusion PAC and the absolute change in PAC during saralasin was obtained (r = 0.72). The role of angiotensin II in the maintenance of a high arterial pressure in chronic dialysis patients was demonstrated. In saralasin-resistant patients, more vigorous ultrafiltration is proposed.

Mycobacterium haemophilum and mycobacterium xenopi associated infection in a renal transplant patient.

The case is presented of a renal-transplant patient in Europe with a Mycobacterium haemophilum infection in association with M. xenopi infection. Clinical signs suggested the diagnosis of mycobacteriosis, which was confirmed by a skin biopsy. Despite antitubercular treatment which rapidly eliminated M. xenopi, the patient's condition did not improve until M. haemophilum was identified. Minimal inhibitory concentrations of various antimicrobial compounds showed a lack of efficacy of isoniazid, and rifampin had no clinical effect. The patient recovered only after careful surgical drainage of the lesions and the administration of minocycline. The pathogenesis of such mycobacterioses is discussed, with focus on the immunodepressive status which in our patient may have been partially induced by a cytomegalovirus reinfection.

Amyloidosis-related cauda equina compression in long-term hemodialysis patients. Three case reports.

STUDY DESIGN: These case reports illustrate the neurologic manifestations due to beta 2 microglobulin amyloid deposition at the lumbar spine level in long-term hemodialysis patients. OBJECTIVE: Radiologic investigations suggested the amyloid origin of extradural soft tissue deposition, which was confirmed by histologic examination after surgical excision. SUMMARY OF BACKGROUND DATA: Although cervical myelopathy is a recently recognized complication of long-term dialysis-related beta 2 microglobulin amyloidosis, neurologic manifestations due to amyloid deposition at the lumbar spine level have rarely been reported. METHODS: Three case reports of cauda equina compression in long-term hemodialysis patients are presented. Follow-up radiography, computed tomography, and magnetic resonance imaging were performed and patients underwent surgical decompression of the thecal sac. RESULTS: In two patients, the compression resulted from the development of a destructive spondylarthropathy, and from the infiltration of extradural spaces and ligaments by an abnormal soft tissue. The third patient had lumbar spinal stenosis due to multiple disc protrusion and to hypertrophy of facet joints and ligamentum flavum. Multilevel laminectomies enabled excision of an abnormal fibrous tissue responsible for the thecal sac compression. Histologic examination of the excised fibrous tissues disclosed amyloid deposits in intervertebral discs, apophysial joints, and ligaments. CONCLUSIONS: In long-term hemodialysis patients, cauda equina compression may develop as the consequence of beta 2 microglobulin amyloid deposition in lumbar intervertebral discs, facet joints, and ligaments. Magnetic resonance imaging is well suited to show the extent of the compression and supports the argument for the amyloid origin of extradural soft tissue.

Protein catabolic rate determination from a single measurement of dialyzed urea.

Protein catabolic rate (PCR, in g protein/kg/day) for anuric patients can be accurately determined without blood sampling by equating urea generation over 7 days to the urea dialyzed in the three dialyses of this period as measured by partial dialysate collection (PDC) or with a urea monitor. The feasibility of determining the week's dialyzed urea from measurement of urea dialyzed in a single session, obviating the need to monitor three consecutive dialyses, was examined in a steady-state simulation of 540 anuric patients spanning the full range of dialysis parameters. It was found that the first, midweek, and last dialyses account for nearly constant fractions (37.9, 32.1, and 30.0%, respectively) of the week's urea removal, leading to equations of the form: PCR = CU/BW + 0.17 where U is the grams of urea dialyzed in the first, midweek, or final dialysis of the week, C = 2.45, 2.89, or 3.10, respectively, and BW is the patient's dry weight in kilograms. These equations were tested on 1312 weeks of PDC data gathered in 42 dialysis patients. Using the midweek U resulted in a mean absolute error in PCR < 0.05 g/kg/day when compared to PCR determined using all three of the week's U values.

On-line dialysis quantification in acutely ill patients: preliminary clinical experience with a multipurpose urea sensor monitoring device.

Direct dialysis quantification offers several advantages compared with conventional blood urea kinetic modeling, and monitoring urea concentration in the effluent dialysate with an on-line urea sensor is a practical approach. Such a monitoring device seems desirable in the short-term dialysis setting to optimize and personalize both renal replacement therapy and nutritional support of acutely ill patients. We designed a urea monitoring device consisting of a urea sensor, a multichannel hydraulic circuit, and a PC microcomputer. The sensor determines urea from catalysis of its hydrolysis by urease in liquid solution during neutral conditions. Hydrolysis of urea produces NH4+, and creates an electrical potential difference between two electrodes. Each concentration determination of urea is the average value of 10 measurements; samples are diverted and measured every 7 min. Laboratory calibration of the urea sensor has demonstrated linearity over the range 2-35 mmol/L. Urea monitoring was performed throughout the treatment course, either on the effluent dialysate or ultrafiltrate in seven acutely ill patients treated by either hemofiltration (n=5) or hemodiafiltration (n=2). The slope of the concentration of urea in the effluent over time was used to calculate an index of the dialysis dose delivered (Kt/V), urea mass removal, and protein catabolic rate. In addition, samples of the effluent were drawn every 21 min, and sent to the central laboratory for measurement of urea concentrations using an autoanalyzer. Kt/V values also were calculated with Garred's equation using pre and post session concentrations of urea in blood. Concentrations of urea in the effluent determined by the urea sensor were found to be very close to those obtained from the central laboratory over a wide range of values (3 to 42 mmol/L). In addition, Kt/V values for both hemofiltration and hemodiafiltration, when calculated with concentrations of urea in the effluent obtained by the urea sensor, did not significantly differ from Kt/V values obtained from the laboratory concentrations of urea in the effluent. On-line urea sensor monitoring of the effluent suppresses the cumbersome task of total effluent collection, and the complexity of urea kinetic analysis. The multipurpose prototype described here represents a new, simple, and direct assessment of dialysis dose and protein nutritional status of acutely ill patients, and is suitable for various modalities.

Dialysis solutions buffered with lactate or bicarbonate: in vitro comparison of two dialysis solutions on human peritoneal cell growth from ESRD and non-ESRD patients.

In order to evaluate the injury to the mesothelial cell layer during long-term peritoneal dialysis (PD), a dialysis solution (solution A), buffered with bicarbonate, stabilized with 10 mmol/L glycylglycine, and sterilized by filtration (0.22 micron double filtration, pH = 7.4), was compared to traditional heat sterilized lactate solution (solution B) on human mesothelial cell cultures. The respective effects of both solutions were evaluated on first passage cells by 3H thymidine incorporation after 72-, 96-, 120-, and 144-h contact. Mesothelial cells to be cultured were obtained from the omentum biopsies of 7 end-stage renal disease (ESRD) patients (during first peritoneal catheter placement) and from 7 non-ESRD patients undergoing abdominal surgery. Solution A (diluted 1/5) induced a time-dependent stimulation of growth in 6 cases of ESRD patient cell cultures, and inhibition occurred only in 1 case. Stimulation was also observed in 5 non-ESRD patient cell cultures, and no effect occurred in 2 cases. Solution B inhibited growth in all the cultures except in 1 case of an ESRD patient in which no effect was observed. This study shows that solution A induced mesothelial cell proliferation, while an inhibitory effect of solution B was observed. No significant differences were observed between the sensitivity of mesothelial cells from ESRD and non-ESRD patients. Further analysis will be carried out to identify precisely the cause of the differences observed: buffer or glycylglycine by themselves and/or glucose by-products.

[Renal response to acute expansion of the extracellular volume in renal transplant patients]. / Réponse rénale à l'expansion aiguë du volume extracellulaire chez les transplantés rénaux.

The response to acute isotonic saline loading (1 800 ml in 3 hours) was assessed in 12 normotensive and 11 hypertensive renal transplant recipients. Both groups had similar renal function, daily urinary excretion of sodium and doses of steroids. The natriuretic response to saline was not affected in hypertensive transplants and changes in blood pressure, renin and aldosterone were identical in both groups. Similar correlations between pre-saline fractional excretion of sodium (FENa+) and the FENa+ obtained during saline were obtained in normotensive and hypertensive patients. These results demonstrate that in renal transplant recipients of young donors no exaggeration of the response to saline occurs in hypertensive subjects thus suggesting that normal kidneys may carry their characteristics when transplanted in a new environment. The role of renal denervation remains unclear.

[Hypertension and stenosis of the graft artery: effects of conversion enzyme inhibition]. / Hypertension artérielle et sténose de l'artère du greffon: effets de l'inhibition de l'enzyme de conversion.

The use of angiotensin converting enzyme inhibitors may lead to reversible renal insufficiency in transplant patients with transplant renal artery stenosis (TRAS). We assessed acute effects of captopril (50 mg, p. os) in 7 cadaver kidney recipients (mean age: 35.6 +/- 4 yrs) with TRAS, 9 to 46 mo after transplantation. All patients were treated by prednisolone and azathioprine. After captopril administration, mean arterial pressure decreased from 127 +/- 6 to 119 +/- 7 mmHg, effective renal plasma flow from 152 +/- 19 to 118 +/- 19 ml/min/1.73 m2, glomerular filtration rate from 59 +/- 8 to 39 +/- 10 ml/min/1.73 m2 and filtration fraction from 0.39 +/- 0.02 to 0.32 +/- 0.07. Among the 7 patients, 2 developed immediate and transient anuria; 4 presented a net decrease of GFR, only one had stable GFR. This patient was chronically treated by captopril; as BP was not controlled, furosemide (40 mg p. os) was added. Serum creatinine increased from 180 to 250 mumol/l. Percutaneous angioplasty was done without decrease in BP; however, treatment by captopril and furosemide could be reinstitued without deterioration in renal function. We conclude that: acute renal failure in kidney graft recipients with TRAS is frequent, but not mandatory; sodium depletion induced by diuretics enhances the fall in GFR; acute effect of captopril must be assessed in patients with TRAS before the use of this product as long term antihypertensive treatment.

[Effect of inhibition of angiotensin converting enzyme on hypertension following kidney transplantation]. / Effets de l'inhibition de l'enzyme de conversion de l'angiotensine dans l'hypertension artérielle après transplantation rénale.

The activation of the renin angiotensin system is thought to be an important factor contributing to hypertension following kidney transplantation (TX). We studied 21 hypertensive renal transplant recipients, without evidence of acute graft rejection or transplant artery stenosis, 6 to 60 months post-TX. The acute responses of mean arterial pressure (MAP) and renal hemodynamics (ERPF: effective renal plasma flow, 131I-Hippuran clearance) and function (GFR: glomerular filtration rate, creatinine clearance; UNaV: urinary sodium excretion rate) to converting enzyme inhibition (CEI) by captopril were assessed. CEI induced a decrease in MAP (118 +/- 2 to 110 +/- 2 mmHg), renal resistance (RR: 0.27 +/- 0.02 to 0.21 +/- 0.01) and filtration fraction (FF: 0.31 +/- 0.02 to 0.23 +/- 0.01). ERPF (307 +/- 24 to 333 +/- 18 ml/min/1.73 m2) and GFR (88 +/- 5 to 78 +/- 5 ml/min/1.73 m2) were not significantly changed. UNaV increased by 53 +/- 24 mumol/min. Changes in MAP (r = -0.66), ERPF (r = 0.74) and FF (r = -0.88) were significantly correlated with the log of control plasma renin activity (PRA). In 10 patients with an increase of ERPF (range: + 30 to + 70%) and no change in GFR, the activated renin system could originate from the recipient's own kidneys. In the remaining 11 patients, CEI was associated with no increase in ERPF (change: + 2 to - 27%) and a fall in GFR, a response suggesting a possible intrarenal vascular damage. These results indicate that RAS participates in the regulation of systemic and renal vascular tone, with a possible predominant effect on efferent glomerular arteriole.

[Arterial hypertension and maintenance hemodialysis: effects of specific inhibition of angiotensin II by saralasin acetate]. / Hypertension artérielle et hémodialyse de suppléance. Effets de l'inhibition spécifique de l'angiotensine II par l'acétate de saralasine.

The "effective" contribution of angiotensin II in blood pressure regulation was investigated in 6 patients on maintenance hemodialysis who were hypertensive at the time of the study (MAP 133 +/- 5 mmHg). Saralasin, a specific angiotensin II inhibitor, was infused at 0.5 and 2.5 microgram/kg/mn three hours before andone hour after hemodialysis. Before hemodialysis, a mean arterial pressure decrease of 13.2 to 19 p. 100 was obtained in 5 patients, arterial pressure being normalized in three of them. After hemodialysis, saralasin induced a normalization of arterial pressure in these 5 subjects. One patient, who was resistant to the saralasin infusion before and after the hemodialysis procedure, can be considered as purely volume-dependent. The renin-angiotensin system is probably one of the primary determinant of dialysis-resistant hypertension. However, a negative response to saralasin should encourage to control hypertension by more vigorous ultrafiltration during dialysis.

Mitochondrial myopathy with respiratory muscle involvement: a case report.

A case of a 10-year-old patient with a benign congenital myopathy, suddenly aggravated because of an accentuated deficit in respiratory muscles is reported. The institution of assisted respiration at night allowed the patient to return to her daily activities. Examination of muscular biopsy with ultra-microscope permitted the diagnosis of mitochondrial myopathy.

[Increased muscular irritability syndrome: treatment with nifedipine. Report of a case]. / Síndrome de irritabilidade muscular aumentada: tratamento com nifedipina. Registro de um caso.

In 1980 Alberca et al. described a patient with a syndrome of increased muscle irritability, who presented ondulating muscle rolling movements and electrically silent cramps, myoedema and muscle reactions to mechanical stimulation similar to myotonic response, suggesting a disfunction at myofibrillar level. We saw a similar case, of a male patient, 21 years of age, who complained of cramps of severe intensity for the past four years. These cramps were painful in the upper and lower limbs and impaired his locomotion; they were electrically silent. At percussion the patient showed severe idiomuscular contraction, with a period of increased relaxation, similar to a myotonic reaction and also, prolonged myoedema and rolling muscle contractions. Electromyography was normal, as were histochemical and electron microscopy studies. We carried out a therapeutic trial with niphedipine (a calcium antagonist), on the assumption that the patient showed a disturbance of the myofibrillar function--even though physiopathogenesis of the hyperirritability muscle syndrome was not yet clearly defined--and with a basis on the importance of the intracytoplasmatic level of Ca++ free in the muscle contraction mechanism, not only as the initiating factor of the contractile process, but also as a quantitative controller of the mechanic tension development through regulation of the amount of ATP metabolized during muscle activity. Administration of the drug in a dose of 40 mg daily, per os, brought a remission of the symptoms after two weeks, and the patient could walk normally again. On the introduction of a placebo, on two different opportunities, there occurred a recrudescence of the symptoms after about one week's time.(ABSTRACT TRUNCATED AT 250 WORDS)

[Coronary ischemia after kidney transplantation in patients over 60 years of age]. / Ischémie coronarienne chez les transplantés rénaux de plus de 60 ans.

Many elderly patients (aged over 60) are waiting for renal allograft but are considered a risk population for myocardial infarction or sudden cardiac death. Between 1987 and January 1st, 1990, 22 elderly patients (16 men and 6 women; mean age 63.4 +/- 0.7 years; range 60-69 years) underwent cadaveric kidney transplantation. Twenty patients had been on haemodialysis for 4.8 +/- 1 years (range: 1-16.3), one had been on peritoneal dialysis for 5 years and one had developed chronic rejection of a previous cadaveric allograft. Among these 22 patients, 6 (27 percent) developed antibodies directed against the random lymphocyte panel and one had a second transplant. Cardiovascular risk factors were systematically estimated, including medical history, cervical arterial Doppler, echocardiography and stress thallium testing (STT). Coronary angiography was performed if there was evidence of myocardial ischaemia. STT did not show any coronary disease in 16 patients and was not maximal (less than 60 percent of maximal theoretical stress) in 4 patients. In 2 patients, STT showed myocardial ischaemia: one, with a history of angina, underwent coronary angiography and angioplasty; the other had silent ischaemia and refused coronary angiography. Two patients died of myocardial infarction. One of them was the patient with silent ischaemia and positive STT test. The patient with angioplasty had a successful transplantation 20 months later, without any cardiovascular complication. After 9 to 32 months of follow-up, 19 patients have functioning allografts (mean serum creatinine level 151 +/- 6 mumol/l). One patient lost his kidney from urinary fistula. Eight patients (36 percent) developed acute rejection and all responded to intravenous corticosteroids. No immunological transplant loss was observed. After 18 months, the graft actuarial survival rate was 85 percent. We conclude that elderly patients free of clinical or silent myocardial ischaemia are good transplant recipients. The cardiovascular risk could be prevented by using STT during the pretransplant screening. If the STT is not maximal or shows coronary ischaemia, coronary angiography is mandatory.

[Functional reversibility of chronic cyclosporine nephrotoxicity in patients after kidney transplantation]. / Réversibilité fonctionnelle de la néphrotoxicité chronique de la ciclosporine chez le transplanté rénal.

The long term use of cyclosporin in renal transplant recipients may be complicated by chronic nephrotoxicity, evidenced by renal functional deterioration and suggestive histological lesions. In 11 renal transplant recipients treated with cyclosporin since 28 +/- 5.8 months, we reduced (n = 6) or stopped (n = 5) this drug after chronic nephrotoxicity was diagnosed. Five months later, we conducted hemodynamic studies and observed significant increases in renal plasma flow (I131 hippuran clearance from 239.5 +/- 106 to 327 +/- 50 ml/min/1.73 m2) and glomerular filtration rate (DTPA-TC clearance from 43 +/- 15 to 67 +/- 10 ml/min/1.73 m2) and a decrease in renal vascular resistances. We suggest that cyclosporin-associated chronic nephrotoxicity is accompanied by some degree of reversible vasoconstriction, or that histological lesions, particularly cyclosporin arteriolopathy, can disappear after cyclosporin withdrawal.

[Prognostic factors of success in renal transplantation]. / Facteurs pronostiques du succès de la transplantation rénale.

From January 1984 to January 1989, 139 kidneys were retrieved from 74 brain dead donors in our institution. The transplantation was performed either locally (79), or in an other French institution (40). The five year actuarial survival rate, for the 139 kidneys retrieved in Montpellier, was 65 percent. Many factors about the donor, the retrieval and the recipient, which may affect the graft survival, were entered in a Cox multivariate analysis. The minimal follow up duration was 18 months. The risk factors studied included: donor parameters (age, sex, cause of death, haemodynamic parameters and renal function); retrieval parameters (kidney alone or multiorgan harvesting, discoloration and renal perfusion quality); organ characteristics (multiple arteries and cold ischemia time); recipients parameters (age, sex, prior transplantation, local transplantation or not, and HLA matching). A first multivariate analysis included only pretransplant risk factors. The risk factors for graft loss, as identified by the Cox model, were in the order: donor's age (P = 0.03), arterial pressure (P = 0.01), prior transplantation of the recipient (P = 0.01) and kidney discoloration quality during the retrieval (P = 0.008). Early post transplant parameters were included within this Cox model (poor early renal function, need for dialysis, serum creatinine level at one week). The need for dialysis therefore was identified as the main predictive value (P = 0.002). The 4 other risk factors, selectioned in the first model, always remained significant.