Abstract

Background. Haemolytic uraemic syndrome (HUS) is a rare and severe disease of various aetiologies in adults. The effect of fresh frozen plasma (FFP) infusion in adults suffering from HUS is not well defined. The aim of this retrospective study was to analyse the causes of HUS in adults admitted in a single renal intensive care unit (ICU) and to determine the life and renal prognosis factors, while most patients (78%) received FFP infusion.

Methods. We recorded clinical, biological, and histological data of 55 adults admitted in our renal ICU for HUS between 1990 and 1998, 49 of them having had a renal biopsy. By stepwise logistic regression analysis, we examined the parameters that were associated with the in‐hospital mortality and renal function at discharge.

Results. HUS complicated different diseases in 40 patients (HIV infection n=18, nephropathies n=10, allotransplantation n=7, malignant diseases n=5) and appeared as a primary in 15 patients. Factors influencing the in‐hospital mortality were positive HIV serology (odds ratio (OR) >20, P=0.0002) and requirement for haemodialysis (OR >35, P=0.004). A pre‐existing nephropathy was a bad prognosis factor for renal function (OR >99, P=0.02), while fever was associated with better renal prognosis (OR=1/10, P=0.033).

Conclusions. HUS in adults remains a severe disease, with a high mortality rate in HIV patients and in those who required haemodialysis. However, as compared with previous studies, we observed an improvement in renal outcome, particularly in patients with primary HUS, suggesting a beneficial effect of FFP infusion, at least in these forms.

Introduction

The association of microangiopathic anaemia with thrombocytopenia and acute impairment of kidney function characterizes the haemolytic uraemic syndrome (HUS). HUS is a rare disease in adults, accounting for less than 5% of the causes of acute renal failure. Compared with the usual paediatric form, HUS in adults is more heterogeneous and may also complicate several underlying diseases [1,2] such as HIV infection [3], chronic nephropathies, hypertension [4], and organ transplantation [5]. Thus, the renal and life prognosis of HUS in adults has varied widely according to the different studies [2,4,6]. In addition, the treatment of HUS remains still uncodified and the benefit of fresh frozen plasma (FFP) infusion is debated [7], while it is clearly established that FFP is effective and must be administrated in patients with the closely related thrombotic thrombocytopenic purpura syndrome (TTP) [8,9].

Considering the close relationship between HUS and TTP and the poor spontaneous prognosis of HUS in adults [1012], several groups, including ours, have used FFP in patients with HUS [2,6,13]. Before the widespread use of plasma, a mortality rate of over 50% used to be reported [2,7,13,14]. More recently, a mortality rate lower than 10% was reported [2]. The rate of chronic renal failure after HUS ranged from 40 to 60% in adults series [2,4,6]. In order to evaluate the effect of FFP in adult HUS, we performed a retrospective analysis of the clinical, biological, and pathological data of all the patients admitted in our renal intensive care unit (ICU) since 1990, when administration of FFP became a part of the treatment of HUS in our centre.

Subjects and methods

A retrospective study was performed, comprising the medical data of 55 adults admitted for HUS in the renal ICU of the Hôpital Tenon from 1990 to 1998. Two of the authors (E.R. and J.D.S.) saw all these patients. HUS was defined by the following criteria: (i) haemolytic anaemia (Hb under 12 g/dl with elevated reticulocytes over 120 000/μl, presence of schistocytes, elevated LDH over 400 UI/l, low haptoglobin serum level under 1.5 g/l), and on thrombocytopenia under 150000 platelets/μl; (ii) acute renal failure with creatinine plasma level higher than 17 mg/l and/or evidence of thrombotic microangiopathy (TMA) on renal biopsy. Patients had no other identifiable cause of anaemia or thrombocytopenia, in particular no intravascular coagulation.

Clinical parameters at admission were analysed: sex, age, and fever, mean and diastolic arterial blood pressure (MAP and DAP), diarrhoea prodroma, and neurological signs. The aetiological context and the usual treatment of patients including drugs known as responsible for HUS were analysed also. Biological data gathered at entry comprised of haemoglobin, platelets, serum lactate dehydrogenase level, white blood cells, and serum creatinine. Normal renal function at discharge was defined by serum creatinine under 17 mg/l to permit the comparison of results with previous studies [6].

In 49 cases, a renal biopsy was performed at admission or in the days following. In the other cases it could not be performed due to the denial of a patient (one case), severe uncontrolled hypertension (one case), or persistent severe thrombocytopenia under 30 000/μl (four cases). Renal lesions were evaluated according to the severity of glomerular and/or arteriolar and arterial TMA. The following criteria for thrombotic microangiopathy were considered: microthrombosis in glomerular capillaries or in renal small arteries; endothelial cell swelling; subendothelial hyaline deposits and double contour appearance of the glomerular capillary wall. Associated tubulo‐interstitial lesions (acute tubular necrosis and interstitial fibrosis) and underlying chronic nephropathy were also recorded. Renal biopsies were re‐evaluated consecutively by the same pathologist (B.M.), who was unaware of the severity of HUS and of its outcome.

All patients were treated in the renal ICU in our department. The symptomatic treatment consisted of the restoration of vital functions, the treatment of hydro‐electrolytic disorders, the transfusions of red blood cells, and the control of hypertension. Antihypertensive drugs were used to maintain the MAP under 90 mmHg. ACE inhibitors were preferred (e.g. enalapril at a dose of 20 mg/day). In addition, specific therapy was used with FFP infusion at a dose of 20 ml/kg/day every day at the beginning, and every second day in the time following, according to the evolution of haemolysis, thrombocytopenia, and decrease in serum creatinine level. When the patient was anuric and/or when he had cardiac overload or uncontrolled hypertension, he received FFP while undergoing plasma exchange (PE) treatment, each second day, in most cases alternating with haemodialysis. Exchanged volume accounted for 2 l plasma, replaced with 1200–1600 ml of FFP and 500 ml of human albumin or macromolecular solutions. In addition, PE therapy was used as the first treatment for severe HUS with neurological impairment, or as secondary treatment in the absence of improvement under FFP infusion alone. Steroids were administrated at the dose of 1 mg/kg/day of prednisolone when thrombocytopenia was lower than 30 000/μl or when the associated disease required this treatment (e.g. lupus, graft rejection.). In some cases supportive therapy finally included anti‐aggregant doses of aspirin and polyvalent immmunoglobulin infusion.

The analysis of risk factors for mortality and chronic renal insufficiency at discharge comprised defined clinical, biological, and histological data at admission, aetiological context and therapy, including requirement for dialysis during hospitalization. Statistical analysis was done as follows: a univariate logistic regression was performed using Wald test. Continuous variables were discretized with medians as cut‐off. Covariates were submitted into the multivariate logistic regression model, when associated at the univariate step with a P value <0.20. The two‐tailed value for significance was P<0.05. Results are expressed as odds ratio (OR) with their 95% confidence intervals (CI). Goodness of fit χ2 is given for each model. Mann–Whitney non‐parametric test was performed to assess the association between DAP and pre‐existent nephropathies or primary HUS. Calculations were done with Statview 5.0 Software® 1992–1998 (SAS Institute Inc., NC).

Results

Patient presentation

The main characteristics of the patient are summarized in Table 1. The AIDS patients have been described elsewhere [3]. Eight of them (44.4%) had a concomitant CMV infection. Among the patients with nephropathies, two had a membrano‐proliferative glomerulonephritis with cryoglobulinaemia complicating HCV infection treated by INF‐α. All kidney‐transplanted patients received cyclosporin, four received azatioprine, and one received mycophenolate mophetil. The liver‐transplanted patient—for HCV‐associated cirrhosis—received FK 506. The bone marrow‐transplanted patient—for acute lymphoblastic leukaemia—received cyclosporin. The malignant diseases were a prostatic carcinoma (n=2), an acute transformation of a chronic myeloid leukaemia (n=1), and a myeloma (n=1). Thus, HUS occurred on a previously normal kidney in 15 patients (27.3%). Verotoxin detection in stools was positive in three patients (one being an AIDS patient) out of four documented cases. One case of HUS occurred in postpartum and one case complicated a contraceptive oestroprogestative treatment. In the other cases, we found no evident cause for HUS: one patient had a familial Mediterranean fever without amyloid deposition on the renal biopsy and one other patient had Crohn's disease [15].

Among the 49 patients who underwent renal biopsy, the morphologic features of the TMA depended on the group studied. In the primary group, most of the patients (seven of 12, 58.3%) had a predominant glomerular form of TMA (Table 2). Interestingly, a large proportion of the biopsies showed an acute tubular necrosis associated with the TMA (five of 12, 41.6%). Moreover, in one case, tubular necrosis was observed without clear‐cut microangiopathic lesions despite the presence of defined biological criteria for HUS. In the nephropathy‐ and systemic disease‐associated HUS group, all patients had a vascular TMA (100%) and none had tubular necrosis (0%). Interstitial fibrosis was a common feature in the biopsies from two groups of nephropathy‐ and malignancy‐associated HUS. Because of the small number of the patients in each group, no statistical analysis could be performed.

Table 1. 

Patients characteristics

Sex ratio (M/F)  37/18 
Age: mean (SD) [range]  41 (14.5) [19–88] 
Fever  15 (27.3%) 
Mean arterial pressure (mmHg): 112.5 (28.1) [64–174] 
   mean (SD) [range]  
Malignant hypertension: n (%)  12 (21.8%) 
Diarrhoea  16 (29.1%) 
Central nervous impairment  25 (45.4%) 
Haemodialysis requiring patients  22 (40%) 
Temporary mechanical ventilation  15 (27.3%) 
HIV‐associated HUS  18 (32.7%) 
Nephropathy‐associated HUS  10 (18.2%) 
Allotransplantation‐associated HUS   7 (12.7%) 
Malignant disease‐associated HUS   5 (9.1%) 
‘Primary’ HUS  15 (27.3%) 
Sex ratio (M/F)  37/18 
Age: mean (SD) [range]  41 (14.5) [19–88] 
Fever  15 (27.3%) 
Mean arterial pressure (mmHg): 112.5 (28.1) [64–174] 
   mean (SD) [range]  
Malignant hypertension: n (%)  12 (21.8%) 
Diarrhoea  16 (29.1%) 
Central nervous impairment  25 (45.4%) 
Haemodialysis requiring patients  22 (40%) 
Temporary mechanical ventilation  15 (27.3%) 
HIV‐associated HUS  18 (32.7%) 
Nephropathy‐associated HUS  10 (18.2%) 
Allotransplantation‐associated HUS   7 (12.7%) 
Malignant disease‐associated HUS   5 (9.1%) 
‘Primary’ HUS  15 (27.3%) 
Table 2. 

Morphologic features of renal biopsies from 49 adults with HUS

HUS 49 biopsies TMA
 

 

 

 
Tubular necrosis Interstitial fibrosis Other features 

 
Glomerular
 
Arterial
 
Combined
 
Cortical necrosis
 

 

 

 
Primary (12) –  
HIV‐associated (16) 2a Endothelial CMV inclusions: 6; intracapillary 
          glomerular foam cells: 7 
Nephropathies or systemic disease associated (10)        
   Malignant nephrosclerosis (4) – – – –  
   Advanced sclerosing nephropathy (1) – – – – Chronic vascular lesions: 4; secondary FSGS: 2 
   Systemic sclerosis (2) – – 1a – – Mucoid intimal hyperplasia: 2; fibrinoid necrosis: 2 
   Lupus nephritis (1) – –  – – MGN with segmental and focal lesions; acute and 
          hyalinized arteriolar lesions 
   Mixed cryoglobulinaemia glomerulonephritis (2) –  – – Glomerular intracapillary thrombi: 2; IgG, IgM, κ, λ 
          deposits in some arteriolar lesions 
Transplantation (7)        
   Kidney (5)b – – – Glomerulitis: 1; acute cellular rejection: 2 
   Liver (1) – – – – Associated chronic arterial lesions and focal mesangiolysis 
   Bone marrow (1) – – – – –  
Malignancy (4)  – Associated chronic vascular lesions: 3; granulomatous 
          tubulo‐interstitial nephritis: 1 
HUS 49 biopsies TMA
 

 

 

 
Tubular necrosis Interstitial fibrosis Other features 

 
Glomerular
 
Arterial
 
Combined
 
Cortical necrosis
 

 

 

 
Primary (12) –  
HIV‐associated (16) 2a Endothelial CMV inclusions: 6; intracapillary 
          glomerular foam cells: 7 
Nephropathies or systemic disease associated (10)        
   Malignant nephrosclerosis (4) – – – –  
   Advanced sclerosing nephropathy (1) – – – – Chronic vascular lesions: 4; secondary FSGS: 2 
   Systemic sclerosis (2) – – 1a – – Mucoid intimal hyperplasia: 2; fibrinoid necrosis: 2 
   Lupus nephritis (1) – –  – – MGN with segmental and focal lesions; acute and 
          hyalinized arteriolar lesions 
   Mixed cryoglobulinaemia glomerulonephritis (2) –  – – Glomerular intracapillary thrombi: 2; IgG, IgM, κ, λ 
          deposits in some arteriolar lesions 
Transplantation (7)        
   Kidney (5)b – – – Glomerulitis: 1; acute cellular rejection: 2 
   Liver (1) – – – – Associated chronic arterial lesions and focal mesangiolysis 
   Bone marrow (1) – – – – –  
Malignancy (4)  – Associated chronic vascular lesions: 3; granulomatous 
          tubulo‐interstitial nephritis: 1 

CMV, cytomegalovirus; FSGS, focal and segmental glomerulosclerosis; MGN, membranous glomerulonephritis. aPatchy cortical necrosis except in one case of HIV‐associated HUS. bOne specimen sampling unsatisfactory.

Patient treatment

Forty‐three patients (78.2%) received FFP infusions, 26 (47.3%) of them being treated by PE. Among the 26 patients who received PE, two suffered from haemorrhagic complications, and eight patients from infectious complications, among whom two AIDS patients died. Twelve patients did not receive FFP. Two patients had rapid improvement in renal function under rehydration alone. Five patients with malignant hypertension received intensive antihypertensive treatment alone. One kidney‐transplanted patient had HUS complicating an acute vascular rejection and was treated by steroids alone. Two patients suffering from acute transformation of leukaemia were treated by specific chemotherapy and radiotherapy. Three AIDS patients could not have either FFP or PE, because of haemodynamic instability. Twenty‐seven (49.1%) patients received steroids. ACE inhibitors were administrated in 27 patients (49.1%). Sixteen patients (29.1%) received aspirin and six (10.9%) infusions of polyvalent immunoglobulins.

Prognosis

Thirteen out of the 55 patients (23.64%) died, essentially in the HIV group (eight patients corresponding to a 44.4% mortality rate in this group). Two patients (13.33%) in the primary HUS group died: a 40‐year‐old woman died under massive gastrointestinal haemorrhage after 8 days of FFP infusion without PE and an 80‐year‐old woman suffering from a verotoxin‐associated HUS died from cerebral impairment after grand mal seizures, despite 16 PEs. The three remaining patients who died were those who had prostatic carcinoma, recurrence of acute leukaemia, and systemic sclerosis with cardiovascular complications. The HIV patients died between 6 and 36 days after admission, and those with non‐HIV HUS between 4 and 78 days after admission.

Tables 3 and 4 summarize the significant parameters influencing the in‐hospital mortality in the univariate and multivariate analyses. It is important to note that the haemorrhagic complications occurred in the absence of PE therapy and independently of any treatment with aspirin, which was rather associated with a better survival in univariate analysis.

At discharge, 19 of the 42 surviving patients (46%) had a chronic renal insufficiency, requiring renal replacement therapy in seven cases. The data of the univariate and multivariate analyses are given in Tables 5 and 6, respectively. To permit a correct multivariate analysis, we used the Mann–Whitney test to prove the strong association of the DAP with the nephropathies group (where the DAP was elevated) and the primitive HUS group (where the DAP was rather low). Then the DAP was excluded and the analysis could be followed. The results concerning the prognostic factors are compared with previous studies of HUS in adult patients in Table 7.

Table 3. 

Factors influencing the in‐hospital mortality in univariate analysis (P<0.20)


 
Surviving patients
 
Deceased patients
 
Significant first step
 
Mean arterial pressure 111 (28.8) 119 (25.9) 0.20 
   (mmHg): mean (SD)    
Diastolic arterial pressure  88.6 (23.5)  98.1 (20.0) 0.12 
   (mmHg): mean (SD)    
AIDS (% of patients)  23.81%  61.54% 0.02 
Transplanted patients  16.67%   0% 0.18 
Haemodialysis requiring  28.57%  76.92% 0.0031 
   patients    
Steroids  54.76%  30.77% 0.20 
Aspirin  33.33%   7.7% 0.084 
Haemorrhagic complications   2.38%  15.38% 0.1249 
Infectious complications  19.05%  46.13% 0.0598 

 
Surviving patients
 
Deceased patients
 
Significant first step
 
Mean arterial pressure 111 (28.8) 119 (25.9) 0.20 
   (mmHg): mean (SD)    
Diastolic arterial pressure  88.6 (23.5)  98.1 (20.0) 0.12 
   (mmHg): mean (SD)    
AIDS (% of patients)  23.81%  61.54% 0.02 
Transplanted patients  16.67%   0% 0.18 
Haemodialysis requiring  28.57%  76.92% 0.0031 
   patients    
Steroids  54.76%  30.77% 0.20 
Aspirin  33.33%   7.7% 0.084 
Haemorrhagic complications   2.38%  15.38% 0.1249 
Infectious complications  19.05%  46.13% 0.0598 
Table 4. 

Factors influencing the life prognosis in multivariate analysis*


 
χ2
 
P value
 
OR
 
95% CI lower
 
95% CI upper
 
HIV positive serology 7.8 0.0002**  20.3 2.5  167.5 
Haemodialysis 8.1 0.004**  35.7 3.0  417.4 
   requiring patients      
Haemorrhagic 7.5 0.0062** 200.3 4.5 8881.7 
   complications      

 
χ2
 
P value
 
OR
 
95% CI lower
 
95% CI upper
 
HIV positive serology 7.8 0.0002**  20.3 2.5  167.5 
Haemodialysis 8.1 0.004**  35.7 3.0  417.4 
   requiring patients      
Haemorrhagic 7.5 0.0062** 200.3 4.5 8881.7 
   complications      

*Pearson (goodness of fit, P=0.92).

**Significant (P<0.05).

Table 5. 

Factors influencing renal prognosis at discharge in univariate analysis (P<0.2)


 
Normal renal function
 
Chronic renal insufficiency
 
Significance first step
 
Number of patients 23 19  
DAP mmHg: mean (SE) 72.2 (16) 101.4 (25) 0.0066 
MAP mmHg: mean (SE) 98.7 (19) 125 (31) 0.013 
Fever 52.2% 27.8% 0.20 
Diarrhoea 40.9% 17.6% 0.17 
Malignant hypertension 13.0% 42.1% 0.0426 
LDH (UI/ml): mean (SE) 2039 (1326) 1381 (1077) 0.049 
Platelets (per ml): mean (SE) 93 696 (70 923) 133 524 (98 480) 0.20 
White blood cells (per ml): mean (SE) 10 922 (8608) 7711 (4728) 0.18 
Nephropathies 8.7% 36.9% 0.055 
Primary HUS 47.8% 10.5% 0.017 

 
Normal renal function
 
Chronic renal insufficiency
 
Significance first step
 
Number of patients 23 19  
DAP mmHg: mean (SE) 72.2 (16) 101.4 (25) 0.0066 
MAP mmHg: mean (SE) 98.7 (19) 125 (31) 0.013 
Fever 52.2% 27.8% 0.20 
Diarrhoea 40.9% 17.6% 0.17 
Malignant hypertension 13.0% 42.1% 0.0426 
LDH (UI/ml): mean (SE) 2039 (1326) 1381 (1077) 0.049 
Platelets (per ml): mean (SE) 93 696 (70 923) 133 524 (98 480) 0.20 
White blood cells (per ml): mean (SE) 10 922 (8608) 7711 (4728) 0.18 
Nephropathies 8.7% 36.9% 0.055 
Primary HUS 47.8% 10.5% 0.017 
Table 6. 

Factors influencing chronic renal insufficiency at discharge in multivariate analysis*


 
χ2
 
P value
 
OR
 
95% CI lower
 
95% CI upper
 
Nephropathies 5.37 0.020** 99.6 2.03 4883.9 
LDH level 5.82 0.016** 10.7 0.89  128.5 
Fever 4.52 0.033**  0.11 0.014    0.84 
Haemodialysis 3.51 0.061 10.74 0.87  128.5 

 
χ2
 
P value
 
OR
 
95% CI lower
 
95% CI upper
 
Nephropathies 5.37 0.020** 99.6 2.03 4883.9 
LDH level 5.82 0.016** 10.7 0.89  128.5 
Fever 4.52 0.033**  0.11 0.014    0.84 
Haemodialysis 3.51 0.061 10.74 0.87  128.5 

*Pearson (goodness of fit, P=0.97).

**Significant (P<0.05).

Table 7. 

Prognosis in this study compared with eight publications on HUS in adult patients

Studies Shiepatti FCSG Hayward Matsumae Hollenbeck Lara Dundas Present study 
Year of publication 1992 1992 1996 1996 1998 1999 1999  
Type of study Retrospective, Retrospective, Retrospective, Retrospective, Retrospective, Retrospective, Prospective, Retrospective, 
    multicentric    multicentric    monocentric    monocentric    monocentric    monocentric    monocentric    monocentric 
Recruitment 1978–1988 1980–1990 1982–1994 1973–1993 1974–1995 1978–1998 1996 1990–1998 
Number of patients 43 53 52 28 45 126 (31 HUS) 22 55 
Number of renal 28 30 49 
   biopsies         
Plasma therapy (PE) 67% 50%?, % PE? ‘Most’ patients Not mentioned 65.1% PE 98.5%, 97% PE 72.7% PE 78%, 39% PE 
Chronic renal 69.8% 40%? Not mentioned 35.7% 33.3% Not mentioned Not mentioned 34.5% 
   failure rate at         
   discharge         
Mortality rate 14% 15%? 8% 10.7% 7% 10% (30‐day) 45% elderly 9.1% (23.6% 
          patients    including HIV 
           patients) 
Risk factors Old age  Not determined Not determined Low haemoglobin Clinical severity Neurological Haemodialysis; 
   for death        level and high    score >6    features    positive HIV 
        leukocyte count      serology 
        at admission    
Risk factors for Severe renal PE could improve Renal injury Severe ARF, severe PE improved renal Fever associated Not studied Pre‐existing 
   chronic renal    involvement    the outcome    at biopsy    arterial and glomerular    prognosis; histology:    with lesser risk     nephropathy, 
   failure       lesions at biopsy    not correlated with    of relapse     renal lesions 
        renal prognosis      at biopsy 
Studies Shiepatti FCSG Hayward Matsumae Hollenbeck Lara Dundas Present study 
Year of publication 1992 1992 1996 1996 1998 1999 1999  
Type of study Retrospective, Retrospective, Retrospective, Retrospective, Retrospective, Retrospective, Prospective, Retrospective, 
    multicentric    multicentric    monocentric    monocentric    monocentric    monocentric    monocentric    monocentric 
Recruitment 1978–1988 1980–1990 1982–1994 1973–1993 1974–1995 1978–1998 1996 1990–1998 
Number of patients 43 53 52 28 45 126 (31 HUS) 22 55 
Number of renal 28 30 49 
   biopsies         
Plasma therapy (PE) 67% 50%?, % PE? ‘Most’ patients Not mentioned 65.1% PE 98.5%, 97% PE 72.7% PE 78%, 39% PE 
Chronic renal 69.8% 40%? Not mentioned 35.7% 33.3% Not mentioned Not mentioned 34.5% 
   failure rate at         
   discharge         
Mortality rate 14% 15%? 8% 10.7% 7% 10% (30‐day) 45% elderly 9.1% (23.6% 
          patients    including HIV 
           patients) 
Risk factors Old age  Not determined Not determined Low haemoglobin Clinical severity Neurological Haemodialysis; 
   for death        level and high    score >6    features    positive HIV 
        leukocyte count      serology 
        at admission    
Risk factors for Severe renal PE could improve Renal injury Severe ARF, severe PE improved renal Fever associated Not studied Pre‐existing 
   chronic renal    involvement    the outcome    at biopsy    arterial and glomerular    prognosis; histology:    with lesser risk     nephropathy, 
   failure       lesions at biopsy    not correlated with    of relapse     renal lesions 
        renal prognosis      at biopsy 

Discussion

The prognosis of adult HUS under plasma therapy is not well known. The data from the literature are difficult to interpret because of the large heterogeneity of the studies concerning the definition of HUS, the diversity of its aetiologies in adults, and the heterogeneity of treatment in retrospective studies [4,6].

We report here an analysis of the different prognostic factors of HUS in adults in a retrospective monocentric study. This study is original because it concerns a recent cohort of HUS in adults for whom the treatment was homogenous, including in most of the cases the administration of FFP and for whom the diagnosis of TMA was confirmed by a renal biopsy. In this study, we showed that the life prognosis of HUS in adults remains compromised with an overall in‐hospital mortality rate of 23.64% if AIDS patients are considered, falling to 9.1% if AIDS patients are excluded. Indeed positive HIV serology represented one of the worst life prognosis factors in multivariate analysis, as well as the requirement for haemodialysis and haemorrhagic complications. AIDS‐associated HUS is less frequently observed now with the use of effective antiretroviral therapy, but may still be encountered, especially in untreated patients. We, and others, have reported previously the possible role of CMV in these patients with AIDS‐associated HUS [3]. AIDS patients with HUS had usually a very low CD4 count and a profound immunosuppression, resulting in frequent infectious complications and a poor survival. Occurrence of HUS in these patients is a dramatic and often fatal event leading rapidly to death in more than 40% of cases. However, it is important to treat these patients with FFP, since two of them recovered a normal renal function after they have received FFP. In our study, the requirement for haemodialysis is a significant risk factor for patient death. This factor, which has not been reported previously, may reflect more severe forms of HUS. In another study, Matsumae et al. [1] have already reported that the severity of the acute renal failure significantly influenced renal survival. Haemodialysis in itself is not responsible for patient deaths. Similarly, infectious complications, which were more frequently observed in patients requiring dialysis, were not a significant death risk factor.

None of the biological markers such as serum creatinine level [4,16] and anaemia at admission— which have been shown to be predictive of poor renal or life prognosis—was a significant risk factor in our study. The low mean white blood cell count that we observed, can be explained by the high proportion of AIDS patients who were frequently leukopenic. In eight cases there was no thrombocytopenia at admission (platelets between 150 000 and 366 000/μl), mainly in patients who had underlying systemic diseases. In these particular forms of non‐thrombocytopenic HUS, the TMA was, nevertheless, confirmed at renal biopsy. In these cases, either thrombocytopenia has been transient before the patient has been admitted or the bone marrow production of platelets was able to compensate for the moderately increased platelet consumption.

Concerning the prognosis factors influencing the renal function of surviving patients at discharge, our study indicates that an underlying renal disease and a requirement for haemodialysis during the hospitalization were significant independent risk factors for the patients to have chronic renal insufficiency. By univariate analysis, high MAP, high DAP, and high serum creatinine levels were also significantly associated with late chronic renal failure but interestingly, these parameters were not found as independent risk factors in the multivariate analysis. This suggests that these factors were rather related to the underlying nephropathy than to HUS itself. Conversely, a high LDH level and fever was significantly associated with a better renal prognosis as published previously [13]. These features were mainly encountered in the primary forms of HUS. We observed a low mortality rate (13.3%) and an excellent renal outcome (73.3% recovering of normal renal function) for these primary forms of HUS, as compared with a 20–30% mortality rate and a 40–60% chronic renal failure rate in previous studies when plasma was not a part of the HUS treatment [13]. The data suggest that plasma infusions improve the renal outcome in primary adult HUS. It is now admitted that plasma is the main treatment of TTP, whether plasma infusion or PEs are used [8,9,12]. Several retrospective studies in adults suggested that plasma therapy may improve renal and life prognosis of patients with HUS [2,6,13,16]. It is possible that HUS in children, which are verotoxin‐mediated in more than 80% of cases, have a limited course in parallel with the gastrointestinal infection. Once the infection is over, the HUS is over too. In agreement with this hypothesis is the lack of recurrence of HUS after transplantation in children, except in atypical or familial forms, which are not related to verotoxin‐induced endothelial lesions. In contrast, in adults, infection with verotoxin‐producing enterobacteria is not the major cause of HUS as demonstrated in our present study and the rate of recurrence of HUS after transplantation is high, reaching 60–75% as shown in a recent study that we performed in France [5]. Moreover, in adults, patients with HUS frequently have neurological signs, i.e. approximately 50% in the present study, and some of them may suffer from TTP with renal involvement, which to date is difficult to distinguish from HUS. In this setting, plasma therapy would be indicated and effective. Based on these considerations and on the results shown in the present study and by other groups [6,9,13], we think that plasma therapy should be used in adult patients with HUS, especially when a profound thrombocytopenia, a severe haemolytic anaemia and neurological signs are associated with the renal involvement. Plasma infusion (20–30 ml/kg/day) may be as effective as PEs [17] but renal failure, hypertension, and cardiac overload often make the plasma infusion hazardous, and we recommend the use of PEs in these cases.

In conclusion, our study shows the wide diversity of HUS that is encountered in adults, in contrast to what is observed in paediatric units. HIV‐related HUS has a very bad prognosis, but fortunately has almost disappeared since effective antiretroviral therapy is available in our country. The symptomatic treatment, especially the strict control of severe hypertension and the appropriate monitoring and treatment of renal failure should be combined with the use of plasma infusion or PEs until the thrombocytopenia has resolved. In non‐HIV patients, a low but significant mortality rate, around 12%, is still observed mainly due to cerebral haemorrhage. In those who survived, the renal prognosis is far better than reported previously in the patients with primary forms of HUS with only 30% having permanent renal failure. In contrast, patients with HUS complicating the course of a pre‐existing nephropathy have a poor renal prognosis; 60–70% developing chronic renal failure. Prospective studies evaluating adult HUS outcome under plasma therapy are now needed.

Correspondence and offprint requests to: Eric Rondeau, Service de Néphrologie A, Hôpital Tenon, 4 rue de la Chine, F‐75020 Paris, France. Email: rondeau@b3e.jussieu.fr

References

1
Matsumae T, Takebayashi S, Naito S. The clinico‐pathological characteristics and outcome in hemolytic‐uremic syndrome of adults.
Clin Nephrol
 
1996
;
45
:
153
–162
2
Hollenbeck M, Kutkuhn B, Aul C, Leschke M, Willers M, Grabensee B. Haemolytic‐uremic syndrome and thrombotic‐thrombocytopenic purpura in adults: clinical findings and prognosis factors for death and end‐stage renal disease.
Nephrol Dial Transplant
 
1998
;
13
:
76
–81
3
Maslo C, Peraldi MN, Desenclos JC et al. Thrombotic microangiopathy and cytomegalovirus disease in patients infected with human immunodeficiency virus.
Clin Infect Dis
 
1997
;
24
:
350
–355
4
Schieppati A, Ruggenenti P, Plata Cornejo R et al. for the Italian Registry of Haemolytic Uremic Syndrome. Renal function at hospital admission as a prognosis factor in adult hemolytic uremic syndrome.
J Am Soc Nephrol
 
1992
;
2
:
1640
–1644
5
Lahlou A, Lang P, Charpentier B et al. Hemolytic uremic syndrome. Recurrence after renal transplantation.
Medicine
 
2000
;
79
:
90
–102
6
Anonymous. French Cooperative Study Group for HUS. Adult hemolytic uremic syndrome with renal microangiopathy. Outcome according to therapeutic protocol in 53 cases.
Ann Med Intern Paris
 
1992
;
143 [Suppl. 1]
:
27
–32
7
Dundas S, Murphy J, Soutar RL, Jones GA, Hutchinson SJ, Todd WTA. Effectiveness of therapeutic plasma exchange in the 1996 Lanarkshire Escherichia coli. O157:H7 outbreak.
Lancet
 
1999
;
354
:
1327
–1330
8
Rock GA, Shumak KH, Buskard NA et al. Comparison of plasma exchange with plasma infusion in the treatment of thrombotic thrombocytopenic purpura.
N Engl J Med
 
1991
;
325
:
393
–397
9
Bell WR, Braine HG, Ness PM, Kickler TS. Improved survival in thrombotic thrombocytopenic purpura‐hemolytic‐uremic syndrome. Clinical experience in 108 patients.
N Engl J Med
 
1991
;
325
:
398
–403
10
Georges JN. How I treat patients with thrombotic thrombocytopenic purpura‐hemolytic uremic syndrome.
Blood
 
2000
;
96
:
1223
–1229
11
Remuzzi G. Hemolytic uremic syndrome: past and present.
Am J Kidney Dis
 
2000
;
36
:
LIV
–VI
12
Ruggenenti P, Remuzzi G. Pathophysioloy and management of thrombotic microangiopathies.
J Nephrol
 
1998
;
11
:
300
–310
13
Lara PN, Coe TL, Zhou H, Fernando L, Holland PV, Wun T. Improved survival with plasma exchange in patients with thrombotic thrombocytopenic purpura‐hemolytic uremic syndrome.
Am J Med
 
1999
;
107
:
573
–579
14
Morel‐Marenger L, Kanfer A, Solez K, Sraer JD, Richet G. Prognostic importance of vascular lesions in acute renal failure with microangiopathic hemolytic anemia (hemolytic‐uremic syndrome): clinicopathologic study in 20 adults.
Kidney Int
 
1979
;
15
:
548
–558
15
Peraldi MN, Akposso K, Haymann JP, Lahlou A, Sraer JD. Haemolytic‐uremic syndrome in patients with Crohn's disease.
Nephrol Dial Transplant
 
1997
;
12
:
2744
–2745
16
Pereira A, Mazzara R, Monteagudo J et al. Thrombotic thrombocytopenic purpura/hemolytic uremic syndrome: a multivariate analysis of factors predicting the responses to plasma exchange.
Ann Hematol
 
1995
;
70
:
319
–323
17
Neild GH. Hemolytic uremic syndrome/thrombotic thrombocytopenic purpura: pathophysiology and treatment.
Kidney Int
 
1998
;
53 [Suppl 64]
:
S45
–S49

Comments

0 Comments