- Email: [email protected]
Clinical and Immunological Evaluation of Infection in Patients on Hemodialysis
Clinical and Immunological Evaluation of Infection in Patients on Hemodialysis Yoshiaki Muraya, 7 Yoshiyuki O o z o n o , 1 Jun-ichi Kadota, ~ M a s a n o b u Miyazaki, 1 Atsuro Hashimoto, 7 Keiko lida, ~ Kaoru Kawakami, 7 Ryo Shirai, ~ M i t s u o Kaku, 2 H i r o n o b u Koga, ~ Takashi Harada, 3 Shigeru Kohno, ~* a n d Kohei Hara 7 7Second Department of Internal Medicine and 3Renal Care Unit, Nagasaki University, School of Medicine, Nagasaki, Japan 2Department of Microbiology, St. Marianna University, School of Medicine, Kawasaki, Japan
Infection is a major complication associated with increased morbidity and mortality in patients on hemodialysis. We analyzed the incidence and type of infection occurring in 4841 patients on hemodialysis between 1986 and 1993 in our hospital and 11 other hemodialysis centers. Infection was noted in 193 patients (4.98 infections/1000 patients/year). Pneumonia (n = 71 ) and bacteremia (n -- 24) were the 2 most common infections, followed by tuberculosis (n = 14), herpes zoster infections (n = 12) and infections at the vascular access site (n = 12). The most commonly isolated organism in pneumonia, bacteremia and vascular access site infections was Staphylococcus aureus. Analysis of the prognosis of patients with pneumonia showed a mortality rate of 50% in patients greater than 60 years old, which was significantly higher than that of younger patients (6.7%, P < 0.01), whereas the mortality rate in patients with bacteremia was not different between the 2 age groups (60.0% vs. 57.9%, respectively). We also analyzed changes in immunological function and nutritional status in 16 patients on hemodialysis and 21 healthy control subjects. Although the phagocytic and bactericidal activities of neutrophils and monocytes were not different between the groups, superoxide production, the percentage of natural killer cells and the degree of blastoid transformation with phytohemagglutinin stimulation were significantly lower in hemodialysis patients. Low levels of Niderman's index and serum albumin and transferrin indicated poor nutritional status in these patients. Furthermore, the degree of Niderman's index and serum albumin significantly correlated with impairment of immunological function, such as reduced blastoid transformation and the number of lymphocytes. Our results suggest that analysis of the patterns of infection in patients on hemodialysis should provide better management and that improvement of malnutrition may ameliorate impaired immunity in hemodialysis patients. J Infect Chemother 1996;2:247-253 Key words: cellular immunity, chronic renal failure, hemodialysis, infection, neutrophil function, nutritional index
INTRODUCTION
T h e n u m b e r of patients with end-stage renal failure (ESRD) has been increasing in recent years. For example, dialysis was performed in 53,017 Japanese patients in 1983, but the figure had increased by 270% at the end of 1994 to 143,709.1 T h e 3 major causes of death in patients on dialysis therapy are infection, cardiovascular and cerebrovascular diseases. Although there has been extensive d e v e l o p m e n t of c h e m o t h e r a p y against infectious organisms in the last several years, infection remains a major p r o b l e m in dialysis patients. 2 It is obvious that identification of the type of microorganisms causing such infections is important in the prevention and management of these patients. Several stud-
Received Feb. 8, 1996; accepted for publication in revised form Oct. 14, 1996. *Correspondence and requests for reprints to: Second Department of Internal Medicine, Nagasaki University, School of Medicine, 1-7-1 Sakamoto, Nagasaki 852, Japan.
1341-321X/96/O204-O247/US$3.00 9 JSC/CLJ 1996
ies examined the type of infections and the causative pathogens in patients with ESRD, 3,4 but these studies were performed 10 to 20 years ago. During this period, remarkable changes in patient populations, dialysis membranes and methods of dialysis have taken place. T h e development of infectious diseases in dialysis patients is thought to be associated with impaired immune function. Changes in chemotaxis and phagocytosis of polymorphonuclear cells 5-9 and reduced cellular immunity have been reported, although results from different laboratories have conflicted. In addition, increased susceptibility to infection in these patients may be linked to malnutrition, 1~ 12 however, the relationship between nutritional status and immunological abnormalities remains poorly understood in patients on hemodialysis. In this study, we examined the incidence of infectious diseases f r o m 1986 to 1993 in p a t i e n t s o n hemodialysis in our hospital and affiliated hospitals in Nagasaki prefecture. We also examined the pattern of immunological abnormalities, including the phagocytic and bactericidal functions ofneutrophils and monocytes,
247
J Infect Chemother 1996;2:247-253
superoxide production and chemotaxis of neutrophils, and lymphocyte subsets in patients on hemodialysis. In addition, the relationship between these immunological tests and nutritional status was analyzed.
1000, 1 = 1000 to 2000 and 2 = total lymphocytes greater than 2000. Based on the N I values, patients were classified into 3 groups, those with good ( N I less than 39), m o d e r a t e (greater than 40 but less than 49), and p o o r ( N I greater than 50) nutritional status.
PATIENTS A N D METHODS
Immunological function Patients A total of 4841 patients were studied between January 1986 and D e c e m b e r 1993 at the Second D e p a r t m e n t of Internal Medicine, Nagasaki University School of Medicine and 11 affiliated institutions. These patients were on regular intermittent hemodialysis 3 times a week throughout the year. T h e diagnosis of infection was based on the presence of leukocytosis (leukocyte count greater than 10,000), neutrophilia (percentage ofneutrophils greater than 60%), positive C R P reaction (greater than 0.2 mg/dL), and elevated m u c o p r o t e i n levels (110 mg/dL). Infectious diseases were classified based on the clinical findings and results of culture. Isolation and identification of organisms were performed using standard clinical bacteriological methods. Bacterial counts exceeding 107 in s p u t u m samples and 105 in the urine were considered significant. T h e diagnosis of tuberculosis was established by Ziehl~Neelsen staining and/or a positive culture for Mycobacterium tuberculosis. We u s e d N i d e r m a n ' s i n d e x (NI) to evaluate the nutritional status. 13 T h e index was calculated as follows: N I = 158 - (16.6 x s e r u m albumin, in mg/dL) (0.78 • skinfold thickness, in millimeters) - (0.2 x serum transferrin value, in mg/dL) - (5.8 • lymphocyte score, LS). LS was expressed on a scale of 0 to 2, with 0 representing a total lymphocyte count of less than
Heparinized blood samples were collected f r o m 16 patients on hemodialysis to evaluate their immunological and nutritional status. All subjects gave informed consent to the experiment. N o n e had local or systemic infection at the time of the study and there was no history of infection in these patients. Table 1 summarizes the clinical findings in 16 hemodialysis patients as well as of 21 age-matched healthy adults. T h e total s e r u m protein, s e r u m albumin, serum transferrin and the n u m b e r of lymphocytes were significantly lower in patients on hemodialysis, while the white blood cells count, n u m b e r of neutrophils, and serum I g G did not differ between the 2 groups. After mixing the blood sample with 3% dextran, the n e u t r o p h i l s were s e p a r a t e d b y FicollH y p a q u e density gradient centrifugation. A suspension of 3 • 106 cells/mL in H a n k s solution ( p H 7.2) was used for the chemotaxis assay as described b y W e r n e r et al. 14 A 48-pore chamber was used to evaluate Chemotaxis and the results expressed as the n u m b e r of neutrophils migrating towards N-formyl-methionyl-leucyl phenylalanine ( F M L P ) in 10 fields (magnification x 1000). T h e production of superoxide by neutrophils was also assessed in these patients. 1 x 107 neutrophils were susp e n d e d in a Krebs Ringer phosphate solution ( p H 7.4), as previously described./5 T h e neutrophils were stimulated by 100 p g / m L concanavalin A (Con A) and 5 #g/
Table 1. Clinical parameters in dialysis patients and controls. Hemodialysis patients
Control
n Age (years) Sex (M/F) Duration of HD (years) Underlying disease*
16 62.1 + 12.7 7/9 3.5 + 2.3 CGN 12, DM3, RCD1
21 64.8 + 3.1 10/11 ---
Serum t~tn! prnt~in (R/d! / albumin (g/dL) IgG (mg/dL) transferrin (mg/dL) BUN (mg/dL) creatinine (mg/dL)
B a + n.n 3.4 + 0.5 1538 + 641 211.9 + 49.3 65.3 + 12.9 8.1 + 2.2
72 + 4.1 + 1479 + 270.3 + ---
WBC (/ram 3)
4462 Z 1162
5552 _+] 922
Neutrophils (/mm 3)
2744 + 427
3292 + 560
NS
Lymphocytes (/mm 3)
1155 + 470
1660 + 579
P < 0.005
AA 0.3 274 36.7
P
NS NS
P.~ n nan! P< 0.0001 NS P< 0.0001
NS
*Underlying diseases, causative disease of end-stage renal failure; CGN, chronic glomerulonephritis; DM, non-insulin dependent diabete< meflitu<; RCD, renal cystic disease; HD, hemodialysis.
248
Infection in hemodialysis patients
m L of cytochalasin E. Superoxide production was measured by the reduction ofcytochrome C using a two-wavelength spectrophotometer. T h e results were expressed as the a m o u n t of superoxide produced (in nM) per 5 x 105 neutrophils per minute. T h e phagocytic and bactericidal activity ofneutrophils and monocytes were measured using the m e t h o d des c r i b e d b y S t e i n k a m p et al. 16 T e n m L o f 2 ' - 7 ' dichlorofluorescein diacetate (DCFDA) was added to 1.0 m L of heparinized whole blood, incubated for 10 minutes at 37~ and 20 m L of fluorescein-conjugated Staphylococcus aureus was then added to the mixture and incubated for 15 minutes at 37~ A fluorescence activating cell sorter (FACS) lysing solution was added to the specimen and left to stand for 5 minutes. After centrifugation, the sample was washed and fixed with 1% paraformaldehyde and phosphate-buffered saline. T h e phagocytic and bactericidal activities of these cells were measured after the neutrophil and monocytic cell fractions were gated by FACS scanning. To assess cellular immunity in the test group, peripheral b l o o d m o n o n u c l e a r cells w e r e s t a i n e d w i t h monoclonal antibodies against various surface markers including C D 3 (T cells), CD4 (helper T cells), CD8 (suppressor T ceils), C D 4 5 R A (2H4, naive T cells), C D 4 5 R O (memory T cells) and CD3 16/56 + (natural killer, N K cells). T h e H L A - D R antigen and the interleukin-2 receptor (Tac antigen, CD25) were also examined as activation markers. After incubation with these antibodies, the samples were washed and fixed with 0.5 % paraformaldehyde and phosphate-buffered saline. T h e lymphocyte fractions were gated and cells positive for each surface marker were analyzed by F A C S c a n (Nippon Becton Dickinson Co, Tokyo, Japan). In addition, blastoid transformation was assessed by incorporation o f 3 H - T d R after stimulation with 4 y g / m L phytohemagglutinin (PHA, Sigma Chemical Co, St. Louis, MO, USA), 4 #g/mL of ConA and 1.0 p g / m L of pokeweed mitogen (PWM, Sigma). 17 T h e 3H uptake was measured by a liquid scintillation counter, and the results were expressed as the stimulation index using the following formula; stimulation index = average counts per minute with mitogen/average counts per minute with medium alone.
Statistical
analysis
Data were expressed as the m e a n _+ SD and compared for statistical differences using the unpaired Student's r test. T h e correlation between nutritional status and i m m u n o l o g i c a l s t a t u s was e x a m i n e d u s i n g t h e Spearman's rank correlation test. A P value less than 0.05 was considered to be statistically significant. RESULTS
Clinical features of infection in patients on hemodialysis Among the 4841 patients on hemodialysis, 193 (4.98 infections per 1000 patients per year) developed infection (Table 2). T h e incidence of infection ranged from 2.7% to 6.0%. Respiratory tract infections were the most c o m m o n among these patients, followed by bacteremia, urinary and gastrointestinal tract infections (Table 3). Among the respiratory tract infections, pneumonia was the most common, but tuberculosis, including pulmonary and extrapulmonary tuberculosis, was also noted, and herpes zoster infections and infections at the site of vascular access were also frequently observed. We also analyzed the time interval between the c o m m e n c e m e n t of hemodialysis and the development of each type of infection. This analysis showed that 34.7% of pneumonias, 46.8% ofbacteremias and 57.9% of cases with tuberculosis developed within the first 6 months ofhemodialysis therapy. T h e incidence decreased from 4% to 13% per year thereafter. T h e causative pathogenic microorganisms from cases of pneumonia and bacteremia are shown inTable 4. In 67.6% of pneumonia cases, causative pathogens were confirmed with S. aureus and Enterococcusfaecalis being the 2 predominant pathogens, followed by Klebsiella p n e u m o n i a e a n d P s e u d o m o n a s aeruginosa. Also, methicillin-resistant S. aureus was isolated from one-third of patients infected with S. aureus. In patients who developed bacteremia, gram-positive bacteria were detected in 13 of 24 episodes, with S. aureus the most frequently detected organism. Gram-negative bacteria, such as P. aeruginosa and Escherichia coli also caused bacteremia, and Candida albicans was the causative p a t h o g e n o f s y s t e m i c i n f e c t i o n in 2 p a t i e n t s o n hemodialysis.
Table 2. The number and rate of infection in hemodialysis patients. Year Number of patients on hemodialysis Number ot patients developing infections Rate of infection (%/patient/year)
1986
1987
1988
1989
1990
1991
1992
1993
Total
307
431
515
628
546
703
717
994
4841
10
12
14
17
21
25
43
51
193
3.3
2.8
2.7
2.7
3.9
3.6
6.0
5.1
4.0
249
J Infect Chemother 1996;2:247-253 Table 4. Causative organisms in pneumonia and bacteremia in hemodialysis patients.
Table 3. Type of infections in hemodialysis patients. Number of patients
%*
5 71 6
5.4 76.3 6.5
3
3.2
5
5.4
atypical mycobacterial infection others
2
2.1
1
1.1
Gastrointestinal infections (n = 20)
infective colitis peritonitis esophageal candidiasis cholecystitis liver abscess others
11 3 2 2 1 1
55.0 15.0 10.0 10.0 5.0 5.0
Urinary tract infection (n = 21)
cystitis pyelonephritis renal abscess
14 5 2
66.7 23.8 9.5
Nervous system (n = 3)
meningitis encephalitis
2 1
66.7 33.3
Systemic infections (n = 26)
bacteremia fever of unknown origin
24 2
92.3 7.7
Others (n = 30)
herpes zoster infection vascular site infection others
12 12 6
40.0 40.0 20.0
Disease Respiratory tract bronchial infection infection pneumonia (n = 93) pulmonary tuberculosis tuberculous pleuritis tuberculous cervical
lymphadenitis
Total n = 193; *percentage of each infection.
We also examined the prognosis of patients who developed pneumonia or bacteremia. T h e mortality rate in patients less than 60 years of age who developed pneumonia (6.7%) was significantly lower than in patients over 60 years of age (50%, P < 0.01), and was very high among patients over 80 years of age (83.3%). In contrast, the mortality rate in patients with bacteremia was not influenced by age (60% vs. 57.9%, for patients less than vs. greater than 60 years of age, respectively).
Immunological and nutritional status in patients on hemodialysis Neutrophil and m o n o c y t e function were analyzed in h e m o d i a l y s i s p a t i e n t s ( T a b l e 5). A l t h o u g h the phagocytic, bactericidal, and chemotactic activities of neutrophils from patients on hemodialysis were similar to those of the control subjects, the p r o d u c t i o n of superoxide by neutrophils was significantly less in hemodialysis patients (P < 0.05), while the phagocytic
250
Pneumonia (n= 71)
Bacteremia (n =24)
Gram positive bacteria Streptococcus sp. S. pneumoniae S. aureus S. epidermidis E. faecalis E. faecium Others
1 (2.1)* 1 (2.1) 14 (29.1) t 3 (6.3) 8 (16.6) 0 (0) 0 (0)
1 (4.2) 0 (0) 5 (20.9) 2 (8.3) 2 (8.3) 0 (0) 3 (12.5)
Gram negative bacteria M. catarrhalis H. influenzae E. coli K. pneumoniae S. marcescens E. cloacae P aeruginosa Others
1 (2.1) 2 (4.2) 1 (2.1) 5 (10.3) 3 (6.3) 1 (2.1) 4 (8.3) 3 (6.3)
0 (0) 0 (0) 3 (12.5) 2 (8.3) 0 (0) 1 (4.2) 3 (12.5) 0 (0)
Fungus C. albicans
1 (2.1)
2 (8.3)
48 (100) 23
24 (100) 0
Total Unknown
*Numbers in parentheses represent the percentage of patients in whom the specified organism was detected; t5 cases caused by methicillin-resistant S. aureus (MRSA).
and bactericidal activities of monocytes were similar between the 2 groups. FACScan analysis of the various surface markers showed that t h e T cell subsets a n d T cell activation markers such as the D R antigen and IL-2R were not significantly different between the 2 groups (Table 6), however, the percentage of C D 16/56-positive lymphocytes, recognized as N K cells, was significantly lower in hemodialysis patients (P < 0.01). Blastoid transformation in response to P H A was also significantly reduced in hemodialysis patients (P < 0.05). T h e mean Niderman's nutritional index was higher than 50 in patients on hemodialysis (patients, 50.7 + 17.3; control, 21.7 _+ 14.5; P < 0.0001), indicating that the nutritional status in these patients was poor. T h e relationship between nutritional status and the various immunological parameters was also analyzed (Table 7). Serum albumin correlated with the degree of blastoid transformation with P H A and ConA stimulation, the n u m b e r of peripheral neutrophils and lymphocytes, and serum levels of IgG and transferrin. T h e Niderman's index also correlated with the degree of PHA-blastoid transformation, the n u m b e r of lymphocytes and serum transferrin.
Infection in hemodialysis patients Table 5. Results of immunological functions of neutrophils and monocytes.
Neutrophits phagocytosis (%) bacterial killing (%) superoxide production (nM/5 x 105 PMN/min) chemotaxis (number of PMN/OIF) Monocytes phagocytosis (%) bacterial killing (%)
Patients (n = 16)
Control (n = 21 )
P
95.1 + 4.6 92.5 + 5.1
96.3 + 3.4 94.4 _+4.6
NS NS
2.61 + 1.14
3.54 + 1.12
P< 0.05
101.9 + 11.5
108.6 + 17.5
NS
65.6 + 24.7 63.7 + 25,3
70.9 + 16.5 66.9 + 15.7
NS NS
PMN, polymorphonuclear cell; OIF, oil immersion field = ten fields o f x 1000 power fields; NS, not significant.
Table 6. Results of surface markers of peripheral blood mononuclear cells and blastoid formation. Patients (n= 16) Surface marker CD3 (%) CD3 + DR + (%) CD3 + CD25 + (%) CD3-16/56 + (%) CD4 (%) CD8 (%) CD45RA (%) CD45RO (%)
65.9 + 21.1 + 8.6 + 17.2 + 41.3 + 38.1 + 33.1 + 32.9 +
9.8 13.2 5.6 10.8 13.7 12.5 8.6 10.3
Control (n=21)
2.78 + 1.02 1.20 + 0.91 2.59 + 1.58
Niderman's index
P
60.1 + 10.2 NS 15.0 + 5.8 NS 8.6 + 4.7 NS 28.1 + 8.4 P < 0.01 41.8 + 9.7 NS 30.2 +_10.3 NS 22.6 + 7.8 NS 30.0 + 8.5 NS
Blastoid transformation Response to PHA* PWM* ConA*
Table 7. Correlation between serum albumin, Niderman's index, and immunological parameters in hemodialysis patients.
3.54 + 0.91 P < 0.05 1.38 + 0.84 NS 3.24 + 0.83 NS
9stimulation index = cpm with mitogen/cpm without mitogen.
DISCUSSION Infection is an important cause of death in patients with E S R D , 4,18-20 however, most data comes from studies conducted prior to the early 1980s. In Japan, infection is an important cause of death, contributing to 12.6% of deaths in 1994, ranking it as the third leading cause of death following heart diseases (35.3%) and cerebrovascular diseases (14.1%).1 Keane et al. 3 reported in 1977 that infection of the respiratory system and at the vascular access site were the most frequent problems encountered with hemodialysis patients. In contrast, this study showed that the rate of respiratory infections was higher (36.8% vs. 9.0%), while bacteremia and infection of the vascular access site were less comm o n (bacteremia, 12.4% vs. 20.4%; vascular access site, 6.2% vs. 80.5%) than those reported by Keane et al. 3 Although the reason for these differences is not known,
Neutrophil phagocytosis bacterial killing superoxide production chemotaxis Monocyte phagocytosis bacterial killing Blastoid transformation response to PHA
Serum albumin
0.55 0.52 0.22 0.43
0.91 0.65 0.60 0.19
0.20 0.11
0.26 0.13
0.04*
0.02*
ConA PWM Number of neutrophils
0.22 O.58 0.07
0.04* 0.49 0.006*
Number of lymphocytes Surface marker CD3 CD4 CD8 CD16/56
0.001 *
0.001 *
0.90 0.27 0.50 0.25
0.26 0.28 0.24 0.18
Serum IgG Serum transferrin
0.13 0.003*
0.04* 0.01 *
*P< 0.05 by Spearman's correlation.
several possibilities exist. During the last 2 decades, several technical changes have improved the process of dialysis, such as the use of different dialysis membranes and dialysis equipment, thus potentially contributing to the reduced rate of bacteremia. In addition, the use of arteriovenous fistulae is n o w commonplace while the use of intravenous cannulae has become less c o m m o n in Japan, thus contributing to the diminished rate of infection at the site of vascular access. Additional supporting evidence includes a published decrease in the incidence of infection of the vascular access site between 1965 and 1978 and between 1979 and 1985. 2 The increased susceptibility of patients on hemodialysis
251
J Infect Chemother 1996;2:247-253
to infection should be carefully considered. This study s h o w e d t h a t o n e - t h i r d to o n e - h a l f of p n e u m o n i a , bacteremia and tuberculosis cases occurred within the first 6 m o n t h s of hemodialysis, indicating that this period is the m o s t critical for m a n a g e m e n t and prophylaxis. W h y these patients are at higher risk of developing infection during the first 6 months of hemodialysis is not clear at the present time and further studies are warranted to establish the mechanisms of enhanced susceptibility to infection in these patients. Age h a d a m a r k e d i m p a c t on the p r o g n o s i s o f hemodialysis patients with pneumonia. T h e mortality rate in patients older than 60 years of age who develo p e d p n e u m o n i a was p o o r c o m p a r e d with that o f younger patients. Considering that 56.4% of Japanese patients who started dialysis therapy in 1994 were older than 60 years of age, and that 45% of the patients undergoing dialysis were at that time over 60 years of age, 1 it b e c o m e s evident that early recognition of infection and the administration of appropriate therapy as early as possible should be the strategy of m a n a g e m e n t of elderly patients on dialysis. Identification of the causative organisms of infection in patients on dialysis is clinically i m p o r t a n t . It is r e c o m m e n d e d that c o m m e n c e m e n t of a p p r o p r i a t e antibiotic therapy be based on suspicion of the type of pathogen in each infection until the results of culture and sensitivity are obtained. A few studies have e x a m ined the type of p a t h o g e n s causing p n e u m o n i a in hemodialysis patients. In 1977, it was reported that Streptococcuspneumoniae was present in 53% of patients, gram-negative organisms in 33% and S. aureus in 13% of patients with pneumonia.3 In contrast, in this study, the m o s t c o m m o n pathogen isolated in cases with p n e u m o n i a was S. aureus, followed by E. faecalis and K. pneumoniae. In particular, M R S A was isolated from a b o u t one-third of cases with S. aureus p n e u m o n i a . Therefore, possible infection with S. aureus in patients on dialysis should be suspected in order to provide adequate therapy. Nasal carriage of S. aureus is considered to be an i m p o r t a n t contributing factor, 21,22 and the prophylactic use of antibiotics reduced the incidence o f this infection. 21 It should also be n o t e d that no p a t h o g e n was isolated f r o m 31.5% of patients with p n e u m o n i a in this study. T h e s e results suggest that the causative organism of p n e u m o n i a may be changing and that a m o r e continuous effort to detect the causative organisms may be necessary in clinical situations. In contrast, the pathogens detected in hemodialysis patients with bacteremia, including S. aureus and E. coli, were identical to those observed in previous studies. 2 4 T h e estimated incidence of tuberculous infection in this study was 2.90 per 1000 patients per year, while previous studies have shown this incidence to vary from 0.86 tO 5.81.2,23 T h e s e differences are likely related to the incidence of tuberculosis in the general population,
252
since reactivation of latent infections occurred in most cases f r o m dialysis patients. A clinical feature of tuberculous infections in our study is the relatively high prevalence of extrapulmonary involvement, which is in agreem e n t with previous reports. 2,23,24 Taken together with the incidence of tuberculosis in Japan, tuberculous infection should always be considered in the differential diagnosis of fever of u n k n o w n origin in patients on dialysis with or without chest x-ray abnormalities. Recent advances in molecular biology and use of the polymerase chain reaction are also helpful in the early detection and confirmation of diagnosis in these patients. 25 T h e enhanced susceptibility to infection m a y be the result of an acquired i m m u n e deficiency in dialysis patients. 26 A l t h o u g h several studies have e x a m i n e d neutrophil and monocyte function in hemodialysis patients, conflicting results have b e e n r e p o r t e d . 2,27-29 T h e s e differences may be due to several factors, including different testing methodologies and variations in the patient population. 2,27 Although the exact mechanisms contributing to diminished immunological function rem a i n to be clarified, malnutrition is known to have a negative impact on the i m m u n e system in the general population, 1~ and probably also in hemodialysis patients. However, only a small n u m b e r of studies have examined the relationship between malnutrition and immunological abnormalities in patients on h e m o d i alysis. In this study, we demonstrated that the nutritional index was significantly lower in h e m o d i a l y s i s patients and that some of these indices correlated with a decrease in immunological function. T h e s e results allow us to hypothesize that i m p r o v e m e n t of the nutritional status may lead to better i m m u n e function in dialysis patients and that a u g m e n t e d defense mechanisms against infection m a y reduce the morbidity and mortality in these patients, particularly those with infection. In this study, immunological function was assessed in malnourished hemodialysis patients who did not have infection, and even in these patients, i m m u n e function was impaired. Further studies are necessary to identify the mechanisms of impaired i m m u n e function and the association between nutrition and immunological function in these patients. In conclusion, we have shown that p n e u m o n i a was the most common infection among patients on hemodialysis, followed by bacteremia. T h e main or ganisms causing p n e u m o n i a in these patients were S. aureus, E. faecalis and Is pneumoniae. T h e prognosis of patients with pneumonia is quite p o o r in elderly patients and impaired immunity was observed a m o n g patients which was associated with malnutrition. It is suggested that surveillance for infection, recognition of the type of infection, and identification of the causative organisms are of p a r a m o u n t i m p o r t a n c e for better clinical m a n agement of these patients.
Infection in hemodialysis patients
ACKNOWLEDGMENTS The authors wish to thank Drs Kohichi Taura, Masahiro Kawatomi, Masato Tadokoro, Kohichi Yamaguchi, Keiko Sasagawa, Yasuhide Kanamoto, Satoru Namie for their assistance in analyzing the data. REFERENCES 1. An overview of regular dialysis treatment in Japan (as of December 31, 1994). JJpn Soc DialTher 1996;29:1-22. 2. Goldman M, Vanherweghem JL. Bacterial infections in chronic hemodialysis patients: epidemiologic and pathophysiologic aspects. Adv Nephrol 1990; 19:315-332. 3. Keane WF, Shapiro FL, Raij L. Incidence and type of infections occurring in 445 chronic hemodialysis patients. Trans Am Soc Artif Intern Organs 1977;23:41-47. 4. Nsouli KA, Lazarus JM, Schoenbaum SC, Gottlieb MN, Lowrie EG, Shocair M. Bacteremic infection in hemodialysis. Arch Intern Med 1979;139:1255-1258. 5. Lewis SL, Van Epps DE. Neutrophil and monocyte alterations in chronic dialysis patients. Am J Kid Dis 1987;9:381-395. 6. Vanholder R, Ringoir S, Dhondt A, Hakim R. Phagocytosis in uremic and hemodialysis patients: a prospective and cross sectional study. Kidney Int 1991;39:320-327. 7. Goldblum SE, PeedWP. Host defenses and immunologic alterations associated with chronic hemodialysis. Ann Int Med 1980;93:597-613. 8. Hosking CS, Atkins RC, Scott DF, Holdsworth SR, Fitzgerald MG, Shelton MJ. Immune and phagocytic function in patients on maintenance dialysis and posttransplantation. Clin Nephrol 1976;6:501-505. 9. Vanholder R, Ringoir S. Polymorphonuclear cell function and infection in dialysis. Kidney Int 1992;42(suppl 38):$91-S95. 10. Good RA. Nutrition and immunity. J Clin Immunol 1981;1:3-11. 11. Chandra RK, Sarchielli P. Nutritional status and immune responses. Clin Lab Med 1993;13:455-461. 12. Bergstrom J, Lindholm B. Nutrition and adequacy of dialysis. How do hemodialysis and CAPD compare? Kidney Int 1993;43(suppl 40):$39-$50. 13. Niederman MS, MerrillWW, Ferranti RD, Pagano YdV[, Palmer LB, Reynolds HY. Nutritional status and bacterial binding in the lower respiratory tract in patients with chronic tracheostomy. Ann Int Med 1984;100:795-800. 14. Werner F, Richard HG, Jr., Edward JL. A 48-well micro chemotaxis assembly for rapid and accurate measurement of leukocyte migration. J Immunol Methods 1980;33:239247.
15. Kitagawa S, Ohta M, Nojiri H, Kakinuma K, Saito M, Takaku F, et al. Functional maturation of membrane potential changes and superoxide producing capacity during differentiation of human granulocytes. ]" Clin Invest 1984;73:1062-1071. 16. Steinkamp JA, Wilson JS, Saunders GC, Stewart CC. Phagocytosis: flow cytometric quantitation with fluorescent microspheres. Science 1982;215:64-66. 17. Murasko DM, Nelson BJ, Silver R, Matour D, Kaye D. Immunologic response in an elderly population with a mean age of 85. Ann Int Med 1986;81:612-618. 18. Izhar HI(, Gatto GRD. Long-term complications of dialysis: infection. Kidney Int 1993;43(suppl 41):S143S148. 19. Lowrie EG, Lazarus JM, Mocelin AJ, Bailey GL, Hampers CL, Wilson RE, et al. Survival of patients undergoing chronic hemodialysis and renal transplantation. N Engl J Med 1973;288:863-867. 20. Neff MS, Eiser AR, Slifkin RF, Baum M, Baez A, Gupta S, et al. Patients surviving 10 years ofhaemodialysis. Am J Med 1983;74:996-1004. 21. YuVL, Goetz A, Wagener M, Smith PB, Rihs JD, Hanchett J, et al. Staphylococcus aureus nasal carriage and infection in patients on hemodialysis. Efficacy of antibiotic prophylaxis. N Engl J Med 1986;315:91-96. 22. Kirmani N, Tuazon CU, Murray HW, Parrish AE, Sheagren JN. Staphylococcus aureus carriage rate of patients receiving long-term hemodialysis. Arch Inter n Med 1978;138:1657-1659. 23. Andrew OT, Schoenfeld PY, Hopewell PG, Humphreys MH. Tuberculosis in patients with end-stage renal disease. Am J Med 1980;68:59-65. 24. Harada T, Tanaka T, Matuo S, Ozono Y, Tomonaga A, Hara K, et al. Tuberculosis in patients undergoing maintenance dialysis. Kekkaku (Tub erculosis) 1985; 60: 53-58. 25. Miyazaki Y, Koga H, Kohno S, Kaku M. Nested polymerase chain reaction for detection of Mycobacterium tuberculosis in clinical samples. J Clin Microbiol 1994;32:2283 2232. 26. Vanholder R, Dell'Aquila R, Ringoir S. Phagocytic function in uremia. Adv Exp Med Biol 1991;297:193-205. 27. Vanholder R, Ringoir S. Infectious morbidity and defects of phagocytic function in end-stage renal disease: a review. JAm Soc Nephrol 1992;3:1541-1554. 28. Sutowicz W, Komorowska Z, Hanicki Z, Cichocki T, Smolenski O. The enzymatic activities and NBT reduction test of granulocytes in untreated and dialysed uraemic patients. Acta Haematol 1979;61:144-149. 29. HirabayashiY, KobayashiT~ Nishikawa A, Okazaki H, Aoki T, Takaya J, et al. Oxidative metabolism and phagocytosis of polymorphonuclear leukocytes in patients with chronic renal failure. Nephron 1988;49:305-312.
253
Infection is a major complication associated with increased morbidity and mortality in patients on hemodialysis. We analyzed the incidence and type of infection occurring in 4841 patients on hemodialysis between 1986 and 1993 in our hospital and 11 other hemodialysis centers. Infection was noted in 193 patients (4.98 infections/1000 patients/year). Pneumonia (n = 71 ) and bacteremia (n -- 24) were the 2 most common infections, followed by tuberculosis (n = 14), herpes zoster infections (n = 12) and infections at the vascular access site (n = 12). The most commonly isolated organism in pneumonia, bacteremia and vascular access site infections was Staphylococcus aureus. Analysis of the prognosis of patients with pneumonia showed a mortality rate of 50% in patients greater than 60 years old, which was significantly higher than that of younger patients (6.7%, P < 0.01), whereas the mortality rate in patients with bacteremia was not different between the 2 age groups (60.0% vs. 57.9%, respectively). We also analyzed changes in immunological function and nutritional status in 16 patients on hemodialysis and 21 healthy control subjects. Although the phagocytic and bactericidal activities of neutrophils and monocytes were not different between the groups, superoxide production, the percentage of natural killer cells and the degree of blastoid transformation with phytohemagglutinin stimulation were significantly lower in hemodialysis patients. Low levels of Niderman's index and serum albumin and transferrin indicated poor nutritional status in these patients. Furthermore, the degree of Niderman's index and serum albumin significantly correlated with impairment of immunological function, such as reduced blastoid transformation and the number of lymphocytes. Our results suggest that analysis of the patterns of infection in patients on hemodialysis should provide better management and that improvement of malnutrition may ameliorate impaired immunity in hemodialysis patients. J Infect Chemother 1996;2:247-253 Key words: cellular immunity, chronic renal failure, hemodialysis, infection, neutrophil function, nutritional index
INTRODUCTION
T h e n u m b e r of patients with end-stage renal failure (ESRD) has been increasing in recent years. For example, dialysis was performed in 53,017 Japanese patients in 1983, but the figure had increased by 270% at the end of 1994 to 143,709.1 T h e 3 major causes of death in patients on dialysis therapy are infection, cardiovascular and cerebrovascular diseases. Although there has been extensive d e v e l o p m e n t of c h e m o t h e r a p y against infectious organisms in the last several years, infection remains a major p r o b l e m in dialysis patients. 2 It is obvious that identification of the type of microorganisms causing such infections is important in the prevention and management of these patients. Several stud-
Received Feb. 8, 1996; accepted for publication in revised form Oct. 14, 1996. *Correspondence and requests for reprints to: Second Department of Internal Medicine, Nagasaki University, School of Medicine, 1-7-1 Sakamoto, Nagasaki 852, Japan.
1341-321X/96/O204-O247/US$3.00 9 JSC/CLJ 1996
ies examined the type of infections and the causative pathogens in patients with ESRD, 3,4 but these studies were performed 10 to 20 years ago. During this period, remarkable changes in patient populations, dialysis membranes and methods of dialysis have taken place. T h e development of infectious diseases in dialysis patients is thought to be associated with impaired immune function. Changes in chemotaxis and phagocytosis of polymorphonuclear cells 5-9 and reduced cellular immunity have been reported, although results from different laboratories have conflicted. In addition, increased susceptibility to infection in these patients may be linked to malnutrition, 1~ 12 however, the relationship between nutritional status and immunological abnormalities remains poorly understood in patients on hemodialysis. In this study, we examined the incidence of infectious diseases f r o m 1986 to 1993 in p a t i e n t s o n hemodialysis in our hospital and affiliated hospitals in Nagasaki prefecture. We also examined the pattern of immunological abnormalities, including the phagocytic and bactericidal functions ofneutrophils and monocytes,
247
J Infect Chemother 1996;2:247-253
superoxide production and chemotaxis of neutrophils, and lymphocyte subsets in patients on hemodialysis. In addition, the relationship between these immunological tests and nutritional status was analyzed.
1000, 1 = 1000 to 2000 and 2 = total lymphocytes greater than 2000. Based on the N I values, patients were classified into 3 groups, those with good ( N I less than 39), m o d e r a t e (greater than 40 but less than 49), and p o o r ( N I greater than 50) nutritional status.
PATIENTS A N D METHODS
Immunological function Patients A total of 4841 patients were studied between January 1986 and D e c e m b e r 1993 at the Second D e p a r t m e n t of Internal Medicine, Nagasaki University School of Medicine and 11 affiliated institutions. These patients were on regular intermittent hemodialysis 3 times a week throughout the year. T h e diagnosis of infection was based on the presence of leukocytosis (leukocyte count greater than 10,000), neutrophilia (percentage ofneutrophils greater than 60%), positive C R P reaction (greater than 0.2 mg/dL), and elevated m u c o p r o t e i n levels (110 mg/dL). Infectious diseases were classified based on the clinical findings and results of culture. Isolation and identification of organisms were performed using standard clinical bacteriological methods. Bacterial counts exceeding 107 in s p u t u m samples and 105 in the urine were considered significant. T h e diagnosis of tuberculosis was established by Ziehl~Neelsen staining and/or a positive culture for Mycobacterium tuberculosis. We u s e d N i d e r m a n ' s i n d e x (NI) to evaluate the nutritional status. 13 T h e index was calculated as follows: N I = 158 - (16.6 x s e r u m albumin, in mg/dL) (0.78 • skinfold thickness, in millimeters) - (0.2 x serum transferrin value, in mg/dL) - (5.8 • lymphocyte score, LS). LS was expressed on a scale of 0 to 2, with 0 representing a total lymphocyte count of less than
Heparinized blood samples were collected f r o m 16 patients on hemodialysis to evaluate their immunological and nutritional status. All subjects gave informed consent to the experiment. N o n e had local or systemic infection at the time of the study and there was no history of infection in these patients. Table 1 summarizes the clinical findings in 16 hemodialysis patients as well as of 21 age-matched healthy adults. T h e total s e r u m protein, s e r u m albumin, serum transferrin and the n u m b e r of lymphocytes were significantly lower in patients on hemodialysis, while the white blood cells count, n u m b e r of neutrophils, and serum I g G did not differ between the 2 groups. After mixing the blood sample with 3% dextran, the n e u t r o p h i l s were s e p a r a t e d b y FicollH y p a q u e density gradient centrifugation. A suspension of 3 • 106 cells/mL in H a n k s solution ( p H 7.2) was used for the chemotaxis assay as described b y W e r n e r et al. 14 A 48-pore chamber was used to evaluate Chemotaxis and the results expressed as the n u m b e r of neutrophils migrating towards N-formyl-methionyl-leucyl phenylalanine ( F M L P ) in 10 fields (magnification x 1000). T h e production of superoxide by neutrophils was also assessed in these patients. 1 x 107 neutrophils were susp e n d e d in a Krebs Ringer phosphate solution ( p H 7.4), as previously described./5 T h e neutrophils were stimulated by 100 p g / m L concanavalin A (Con A) and 5 #g/
Table 1. Clinical parameters in dialysis patients and controls. Hemodialysis patients
Control
n Age (years) Sex (M/F) Duration of HD (years) Underlying disease*
16 62.1 + 12.7 7/9 3.5 + 2.3 CGN 12, DM3, RCD1
21 64.8 + 3.1 10/11 ---
Serum t~tn! prnt~in (R/d! / albumin (g/dL) IgG (mg/dL) transferrin (mg/dL) BUN (mg/dL) creatinine (mg/dL)
B a + n.n 3.4 + 0.5 1538 + 641 211.9 + 49.3 65.3 + 12.9 8.1 + 2.2
72 + 4.1 + 1479 + 270.3 + ---
WBC (/ram 3)
4462 Z 1162
5552 _+] 922
Neutrophils (/mm 3)
2744 + 427
3292 + 560
NS
Lymphocytes (/mm 3)
1155 + 470
1660 + 579
P < 0.005
AA 0.3 274 36.7
P
NS NS
P.~ n nan! P< 0.0001 NS P< 0.0001
NS
*Underlying diseases, causative disease of end-stage renal failure; CGN, chronic glomerulonephritis; DM, non-insulin dependent diabete< meflitu<; RCD, renal cystic disease; HD, hemodialysis.
248
Infection in hemodialysis patients
m L of cytochalasin E. Superoxide production was measured by the reduction ofcytochrome C using a two-wavelength spectrophotometer. T h e results were expressed as the a m o u n t of superoxide produced (in nM) per 5 x 105 neutrophils per minute. T h e phagocytic and bactericidal activity ofneutrophils and monocytes were measured using the m e t h o d des c r i b e d b y S t e i n k a m p et al. 16 T e n m L o f 2 ' - 7 ' dichlorofluorescein diacetate (DCFDA) was added to 1.0 m L of heparinized whole blood, incubated for 10 minutes at 37~ and 20 m L of fluorescein-conjugated Staphylococcus aureus was then added to the mixture and incubated for 15 minutes at 37~ A fluorescence activating cell sorter (FACS) lysing solution was added to the specimen and left to stand for 5 minutes. After centrifugation, the sample was washed and fixed with 1% paraformaldehyde and phosphate-buffered saline. T h e phagocytic and bactericidal activities of these cells were measured after the neutrophil and monocytic cell fractions were gated by FACS scanning. To assess cellular immunity in the test group, peripheral b l o o d m o n o n u c l e a r cells w e r e s t a i n e d w i t h monoclonal antibodies against various surface markers including C D 3 (T cells), CD4 (helper T cells), CD8 (suppressor T ceils), C D 4 5 R A (2H4, naive T cells), C D 4 5 R O (memory T cells) and CD3 16/56 + (natural killer, N K cells). T h e H L A - D R antigen and the interleukin-2 receptor (Tac antigen, CD25) were also examined as activation markers. After incubation with these antibodies, the samples were washed and fixed with 0.5 % paraformaldehyde and phosphate-buffered saline. T h e lymphocyte fractions were gated and cells positive for each surface marker were analyzed by F A C S c a n (Nippon Becton Dickinson Co, Tokyo, Japan). In addition, blastoid transformation was assessed by incorporation o f 3 H - T d R after stimulation with 4 y g / m L phytohemagglutinin (PHA, Sigma Chemical Co, St. Louis, MO, USA), 4 #g/mL of ConA and 1.0 p g / m L of pokeweed mitogen (PWM, Sigma). 17 T h e 3H uptake was measured by a liquid scintillation counter, and the results were expressed as the stimulation index using the following formula; stimulation index = average counts per minute with mitogen/average counts per minute with medium alone.
Statistical
analysis
Data were expressed as the m e a n _+ SD and compared for statistical differences using the unpaired Student's r test. T h e correlation between nutritional status and i m m u n o l o g i c a l s t a t u s was e x a m i n e d u s i n g t h e Spearman's rank correlation test. A P value less than 0.05 was considered to be statistically significant. RESULTS
Clinical features of infection in patients on hemodialysis Among the 4841 patients on hemodialysis, 193 (4.98 infections per 1000 patients per year) developed infection (Table 2). T h e incidence of infection ranged from 2.7% to 6.0%. Respiratory tract infections were the most c o m m o n among these patients, followed by bacteremia, urinary and gastrointestinal tract infections (Table 3). Among the respiratory tract infections, pneumonia was the most common, but tuberculosis, including pulmonary and extrapulmonary tuberculosis, was also noted, and herpes zoster infections and infections at the site of vascular access were also frequently observed. We also analyzed the time interval between the c o m m e n c e m e n t of hemodialysis and the development of each type of infection. This analysis showed that 34.7% of pneumonias, 46.8% ofbacteremias and 57.9% of cases with tuberculosis developed within the first 6 months ofhemodialysis therapy. T h e incidence decreased from 4% to 13% per year thereafter. T h e causative pathogenic microorganisms from cases of pneumonia and bacteremia are shown inTable 4. In 67.6% of pneumonia cases, causative pathogens were confirmed with S. aureus and Enterococcusfaecalis being the 2 predominant pathogens, followed by Klebsiella p n e u m o n i a e a n d P s e u d o m o n a s aeruginosa. Also, methicillin-resistant S. aureus was isolated from one-third of patients infected with S. aureus. In patients who developed bacteremia, gram-positive bacteria were detected in 13 of 24 episodes, with S. aureus the most frequently detected organism. Gram-negative bacteria, such as P. aeruginosa and Escherichia coli also caused bacteremia, and Candida albicans was the causative p a t h o g e n o f s y s t e m i c i n f e c t i o n in 2 p a t i e n t s o n hemodialysis.
Table 2. The number and rate of infection in hemodialysis patients. Year Number of patients on hemodialysis Number ot patients developing infections Rate of infection (%/patient/year)
1986
1987
1988
1989
1990
1991
1992
1993
Total
307
431
515
628
546
703
717
994
4841
10
12
14
17
21
25
43
51
193
3.3
2.8
2.7
2.7
3.9
3.6
6.0
5.1
4.0
249
J Infect Chemother 1996;2:247-253 Table 4. Causative organisms in pneumonia and bacteremia in hemodialysis patients.
Table 3. Type of infections in hemodialysis patients. Number of patients
%*
5 71 6
5.4 76.3 6.5
3
3.2
5
5.4
atypical mycobacterial infection others
2
2.1
1
1.1
Gastrointestinal infections (n = 20)
infective colitis peritonitis esophageal candidiasis cholecystitis liver abscess others
11 3 2 2 1 1
55.0 15.0 10.0 10.0 5.0 5.0
Urinary tract infection (n = 21)
cystitis pyelonephritis renal abscess
14 5 2
66.7 23.8 9.5
Nervous system (n = 3)
meningitis encephalitis
2 1
66.7 33.3
Systemic infections (n = 26)
bacteremia fever of unknown origin
24 2
92.3 7.7
Others (n = 30)
herpes zoster infection vascular site infection others
12 12 6
40.0 40.0 20.0
Disease Respiratory tract bronchial infection infection pneumonia (n = 93) pulmonary tuberculosis tuberculous pleuritis tuberculous cervical
lymphadenitis
Total n = 193; *percentage of each infection.
We also examined the prognosis of patients who developed pneumonia or bacteremia. T h e mortality rate in patients less than 60 years of age who developed pneumonia (6.7%) was significantly lower than in patients over 60 years of age (50%, P < 0.01), and was very high among patients over 80 years of age (83.3%). In contrast, the mortality rate in patients with bacteremia was not influenced by age (60% vs. 57.9%, for patients less than vs. greater than 60 years of age, respectively).
Immunological and nutritional status in patients on hemodialysis Neutrophil and m o n o c y t e function were analyzed in h e m o d i a l y s i s p a t i e n t s ( T a b l e 5). A l t h o u g h the phagocytic, bactericidal, and chemotactic activities of neutrophils from patients on hemodialysis were similar to those of the control subjects, the p r o d u c t i o n of superoxide by neutrophils was significantly less in hemodialysis patients (P < 0.05), while the phagocytic
250
Pneumonia (n= 71)
Bacteremia (n =24)
Gram positive bacteria Streptococcus sp. S. pneumoniae S. aureus S. epidermidis E. faecalis E. faecium Others
1 (2.1)* 1 (2.1) 14 (29.1) t 3 (6.3) 8 (16.6) 0 (0) 0 (0)
1 (4.2) 0 (0) 5 (20.9) 2 (8.3) 2 (8.3) 0 (0) 3 (12.5)
Gram negative bacteria M. catarrhalis H. influenzae E. coli K. pneumoniae S. marcescens E. cloacae P aeruginosa Others
1 (2.1) 2 (4.2) 1 (2.1) 5 (10.3) 3 (6.3) 1 (2.1) 4 (8.3) 3 (6.3)
0 (0) 0 (0) 3 (12.5) 2 (8.3) 0 (0) 1 (4.2) 3 (12.5) 0 (0)
Fungus C. albicans
1 (2.1)
2 (8.3)
48 (100) 23
24 (100) 0
Total Unknown
*Numbers in parentheses represent the percentage of patients in whom the specified organism was detected; t5 cases caused by methicillin-resistant S. aureus (MRSA).
and bactericidal activities of monocytes were similar between the 2 groups. FACScan analysis of the various surface markers showed that t h e T cell subsets a n d T cell activation markers such as the D R antigen and IL-2R were not significantly different between the 2 groups (Table 6), however, the percentage of C D 16/56-positive lymphocytes, recognized as N K cells, was significantly lower in hemodialysis patients (P < 0.01). Blastoid transformation in response to P H A was also significantly reduced in hemodialysis patients (P < 0.05). T h e mean Niderman's nutritional index was higher than 50 in patients on hemodialysis (patients, 50.7 + 17.3; control, 21.7 _+ 14.5; P < 0.0001), indicating that the nutritional status in these patients was poor. T h e relationship between nutritional status and the various immunological parameters was also analyzed (Table 7). Serum albumin correlated with the degree of blastoid transformation with P H A and ConA stimulation, the n u m b e r of peripheral neutrophils and lymphocytes, and serum levels of IgG and transferrin. T h e Niderman's index also correlated with the degree of PHA-blastoid transformation, the n u m b e r of lymphocytes and serum transferrin.
Infection in hemodialysis patients Table 5. Results of immunological functions of neutrophils and monocytes.
Neutrophits phagocytosis (%) bacterial killing (%) superoxide production (nM/5 x 105 PMN/min) chemotaxis (number of PMN/OIF) Monocytes phagocytosis (%) bacterial killing (%)
Patients (n = 16)
Control (n = 21 )
P
95.1 + 4.6 92.5 + 5.1
96.3 + 3.4 94.4 _+4.6
NS NS
2.61 + 1.14
3.54 + 1.12
P< 0.05
101.9 + 11.5
108.6 + 17.5
NS
65.6 + 24.7 63.7 + 25,3
70.9 + 16.5 66.9 + 15.7
NS NS
PMN, polymorphonuclear cell; OIF, oil immersion field = ten fields o f x 1000 power fields; NS, not significant.
Table 6. Results of surface markers of peripheral blood mononuclear cells and blastoid formation. Patients (n= 16) Surface marker CD3 (%) CD3 + DR + (%) CD3 + CD25 + (%) CD3-16/56 + (%) CD4 (%) CD8 (%) CD45RA (%) CD45RO (%)
65.9 + 21.1 + 8.6 + 17.2 + 41.3 + 38.1 + 33.1 + 32.9 +
9.8 13.2 5.6 10.8 13.7 12.5 8.6 10.3
Control (n=21)
2.78 + 1.02 1.20 + 0.91 2.59 + 1.58
Niderman's index
P
60.1 + 10.2 NS 15.0 + 5.8 NS 8.6 + 4.7 NS 28.1 + 8.4 P < 0.01 41.8 + 9.7 NS 30.2 +_10.3 NS 22.6 + 7.8 NS 30.0 + 8.5 NS
Blastoid transformation Response to PHA* PWM* ConA*
Table 7. Correlation between serum albumin, Niderman's index, and immunological parameters in hemodialysis patients.
3.54 + 0.91 P < 0.05 1.38 + 0.84 NS 3.24 + 0.83 NS
9stimulation index = cpm with mitogen/cpm without mitogen.
DISCUSSION Infection is an important cause of death in patients with E S R D , 4,18-20 however, most data comes from studies conducted prior to the early 1980s. In Japan, infection is an important cause of death, contributing to 12.6% of deaths in 1994, ranking it as the third leading cause of death following heart diseases (35.3%) and cerebrovascular diseases (14.1%).1 Keane et al. 3 reported in 1977 that infection of the respiratory system and at the vascular access site were the most frequent problems encountered with hemodialysis patients. In contrast, this study showed that the rate of respiratory infections was higher (36.8% vs. 9.0%), while bacteremia and infection of the vascular access site were less comm o n (bacteremia, 12.4% vs. 20.4%; vascular access site, 6.2% vs. 80.5%) than those reported by Keane et al. 3 Although the reason for these differences is not known,
Neutrophil phagocytosis bacterial killing superoxide production chemotaxis Monocyte phagocytosis bacterial killing Blastoid transformation response to PHA
Serum albumin
0.55 0.52 0.22 0.43
0.91 0.65 0.60 0.19
0.20 0.11
0.26 0.13
0.04*
0.02*
ConA PWM Number of neutrophils
0.22 O.58 0.07
0.04* 0.49 0.006*
Number of lymphocytes Surface marker CD3 CD4 CD8 CD16/56
0.001 *
0.001 *
0.90 0.27 0.50 0.25
0.26 0.28 0.24 0.18
Serum IgG Serum transferrin
0.13 0.003*
0.04* 0.01 *
*P< 0.05 by Spearman's correlation.
several possibilities exist. During the last 2 decades, several technical changes have improved the process of dialysis, such as the use of different dialysis membranes and dialysis equipment, thus potentially contributing to the reduced rate of bacteremia. In addition, the use of arteriovenous fistulae is n o w commonplace while the use of intravenous cannulae has become less c o m m o n in Japan, thus contributing to the diminished rate of infection at the site of vascular access. Additional supporting evidence includes a published decrease in the incidence of infection of the vascular access site between 1965 and 1978 and between 1979 and 1985. 2 The increased susceptibility of patients on hemodialysis
251
J Infect Chemother 1996;2:247-253
to infection should be carefully considered. This study s h o w e d t h a t o n e - t h i r d to o n e - h a l f of p n e u m o n i a , bacteremia and tuberculosis cases occurred within the first 6 m o n t h s of hemodialysis, indicating that this period is the m o s t critical for m a n a g e m e n t and prophylaxis. W h y these patients are at higher risk of developing infection during the first 6 months of hemodialysis is not clear at the present time and further studies are warranted to establish the mechanisms of enhanced susceptibility to infection in these patients. Age h a d a m a r k e d i m p a c t on the p r o g n o s i s o f hemodialysis patients with pneumonia. T h e mortality rate in patients older than 60 years of age who develo p e d p n e u m o n i a was p o o r c o m p a r e d with that o f younger patients. Considering that 56.4% of Japanese patients who started dialysis therapy in 1994 were older than 60 years of age, and that 45% of the patients undergoing dialysis were at that time over 60 years of age, 1 it b e c o m e s evident that early recognition of infection and the administration of appropriate therapy as early as possible should be the strategy of m a n a g e m e n t of elderly patients on dialysis. Identification of the causative organisms of infection in patients on dialysis is clinically i m p o r t a n t . It is r e c o m m e n d e d that c o m m e n c e m e n t of a p p r o p r i a t e antibiotic therapy be based on suspicion of the type of pathogen in each infection until the results of culture and sensitivity are obtained. A few studies have e x a m ined the type of p a t h o g e n s causing p n e u m o n i a in hemodialysis patients. In 1977, it was reported that Streptococcuspneumoniae was present in 53% of patients, gram-negative organisms in 33% and S. aureus in 13% of patients with pneumonia.3 In contrast, in this study, the m o s t c o m m o n pathogen isolated in cases with p n e u m o n i a was S. aureus, followed by E. faecalis and K. pneumoniae. In particular, M R S A was isolated from a b o u t one-third of cases with S. aureus p n e u m o n i a . Therefore, possible infection with S. aureus in patients on dialysis should be suspected in order to provide adequate therapy. Nasal carriage of S. aureus is considered to be an i m p o r t a n t contributing factor, 21,22 and the prophylactic use of antibiotics reduced the incidence o f this infection. 21 It should also be n o t e d that no p a t h o g e n was isolated f r o m 31.5% of patients with p n e u m o n i a in this study. T h e s e results suggest that the causative organism of p n e u m o n i a may be changing and that a m o r e continuous effort to detect the causative organisms may be necessary in clinical situations. In contrast, the pathogens detected in hemodialysis patients with bacteremia, including S. aureus and E. coli, were identical to those observed in previous studies. 2 4 T h e estimated incidence of tuberculous infection in this study was 2.90 per 1000 patients per year, while previous studies have shown this incidence to vary from 0.86 tO 5.81.2,23 T h e s e differences are likely related to the incidence of tuberculosis in the general population,
252
since reactivation of latent infections occurred in most cases f r o m dialysis patients. A clinical feature of tuberculous infections in our study is the relatively high prevalence of extrapulmonary involvement, which is in agreem e n t with previous reports. 2,23,24 Taken together with the incidence of tuberculosis in Japan, tuberculous infection should always be considered in the differential diagnosis of fever of u n k n o w n origin in patients on dialysis with or without chest x-ray abnormalities. Recent advances in molecular biology and use of the polymerase chain reaction are also helpful in the early detection and confirmation of diagnosis in these patients. 25 T h e enhanced susceptibility to infection m a y be the result of an acquired i m m u n e deficiency in dialysis patients. 26 A l t h o u g h several studies have e x a m i n e d neutrophil and monocyte function in hemodialysis patients, conflicting results have b e e n r e p o r t e d . 2,27-29 T h e s e differences may be due to several factors, including different testing methodologies and variations in the patient population. 2,27 Although the exact mechanisms contributing to diminished immunological function rem a i n to be clarified, malnutrition is known to have a negative impact on the i m m u n e system in the general population, 1~ and probably also in hemodialysis patients. However, only a small n u m b e r of studies have examined the relationship between malnutrition and immunological abnormalities in patients on h e m o d i alysis. In this study, we demonstrated that the nutritional index was significantly lower in h e m o d i a l y s i s patients and that some of these indices correlated with a decrease in immunological function. T h e s e results allow us to hypothesize that i m p r o v e m e n t of the nutritional status may lead to better i m m u n e function in dialysis patients and that a u g m e n t e d defense mechanisms against infection m a y reduce the morbidity and mortality in these patients, particularly those with infection. In this study, immunological function was assessed in malnourished hemodialysis patients who did not have infection, and even in these patients, i m m u n e function was impaired. Further studies are necessary to identify the mechanisms of impaired i m m u n e function and the association between nutrition and immunological function in these patients. In conclusion, we have shown that p n e u m o n i a was the most common infection among patients on hemodialysis, followed by bacteremia. T h e main or ganisms causing p n e u m o n i a in these patients were S. aureus, E. faecalis and Is pneumoniae. T h e prognosis of patients with pneumonia is quite p o o r in elderly patients and impaired immunity was observed a m o n g patients which was associated with malnutrition. It is suggested that surveillance for infection, recognition of the type of infection, and identification of the causative organisms are of p a r a m o u n t i m p o r t a n c e for better clinical m a n agement of these patients.
Infection in hemodialysis patients
ACKNOWLEDGMENTS The authors wish to thank Drs Kohichi Taura, Masahiro Kawatomi, Masato Tadokoro, Kohichi Yamaguchi, Keiko Sasagawa, Yasuhide Kanamoto, Satoru Namie for their assistance in analyzing the data. REFERENCES 1. An overview of regular dialysis treatment in Japan (as of December 31, 1994). JJpn Soc DialTher 1996;29:1-22. 2. Goldman M, Vanherweghem JL. Bacterial infections in chronic hemodialysis patients: epidemiologic and pathophysiologic aspects. Adv Nephrol 1990; 19:315-332. 3. Keane WF, Shapiro FL, Raij L. Incidence and type of infections occurring in 445 chronic hemodialysis patients. Trans Am Soc Artif Intern Organs 1977;23:41-47. 4. Nsouli KA, Lazarus JM, Schoenbaum SC, Gottlieb MN, Lowrie EG, Shocair M. Bacteremic infection in hemodialysis. Arch Intern Med 1979;139:1255-1258. 5. Lewis SL, Van Epps DE. Neutrophil and monocyte alterations in chronic dialysis patients. Am J Kid Dis 1987;9:381-395. 6. Vanholder R, Ringoir S, Dhondt A, Hakim R. Phagocytosis in uremic and hemodialysis patients: a prospective and cross sectional study. Kidney Int 1991;39:320-327. 7. Goldblum SE, PeedWP. Host defenses and immunologic alterations associated with chronic hemodialysis. Ann Int Med 1980;93:597-613. 8. Hosking CS, Atkins RC, Scott DF, Holdsworth SR, Fitzgerald MG, Shelton MJ. Immune and phagocytic function in patients on maintenance dialysis and posttransplantation. Clin Nephrol 1976;6:501-505. 9. Vanholder R, Ringoir S. Polymorphonuclear cell function and infection in dialysis. Kidney Int 1992;42(suppl 38):$91-S95. 10. Good RA. Nutrition and immunity. J Clin Immunol 1981;1:3-11. 11. Chandra RK, Sarchielli P. Nutritional status and immune responses. Clin Lab Med 1993;13:455-461. 12. Bergstrom J, Lindholm B. Nutrition and adequacy of dialysis. How do hemodialysis and CAPD compare? Kidney Int 1993;43(suppl 40):$39-$50. 13. Niederman MS, MerrillWW, Ferranti RD, Pagano YdV[, Palmer LB, Reynolds HY. Nutritional status and bacterial binding in the lower respiratory tract in patients with chronic tracheostomy. Ann Int Med 1984;100:795-800. 14. Werner F, Richard HG, Jr., Edward JL. A 48-well micro chemotaxis assembly for rapid and accurate measurement of leukocyte migration. J Immunol Methods 1980;33:239247.
15. Kitagawa S, Ohta M, Nojiri H, Kakinuma K, Saito M, Takaku F, et al. Functional maturation of membrane potential changes and superoxide producing capacity during differentiation of human granulocytes. ]" Clin Invest 1984;73:1062-1071. 16. Steinkamp JA, Wilson JS, Saunders GC, Stewart CC. Phagocytosis: flow cytometric quantitation with fluorescent microspheres. Science 1982;215:64-66. 17. Murasko DM, Nelson BJ, Silver R, Matour D, Kaye D. Immunologic response in an elderly population with a mean age of 85. Ann Int Med 1986;81:612-618. 18. Izhar HI(, Gatto GRD. Long-term complications of dialysis: infection. Kidney Int 1993;43(suppl 41):S143S148. 19. Lowrie EG, Lazarus JM, Mocelin AJ, Bailey GL, Hampers CL, Wilson RE, et al. Survival of patients undergoing chronic hemodialysis and renal transplantation. N Engl J Med 1973;288:863-867. 20. Neff MS, Eiser AR, Slifkin RF, Baum M, Baez A, Gupta S, et al. Patients surviving 10 years ofhaemodialysis. Am J Med 1983;74:996-1004. 21. YuVL, Goetz A, Wagener M, Smith PB, Rihs JD, Hanchett J, et al. Staphylococcus aureus nasal carriage and infection in patients on hemodialysis. Efficacy of antibiotic prophylaxis. N Engl J Med 1986;315:91-96. 22. Kirmani N, Tuazon CU, Murray HW, Parrish AE, Sheagren JN. Staphylococcus aureus carriage rate of patients receiving long-term hemodialysis. Arch Inter n Med 1978;138:1657-1659. 23. Andrew OT, Schoenfeld PY, Hopewell PG, Humphreys MH. Tuberculosis in patients with end-stage renal disease. Am J Med 1980;68:59-65. 24. Harada T, Tanaka T, Matuo S, Ozono Y, Tomonaga A, Hara K, et al. Tuberculosis in patients undergoing maintenance dialysis. Kekkaku (Tub erculosis) 1985; 60: 53-58. 25. Miyazaki Y, Koga H, Kohno S, Kaku M. Nested polymerase chain reaction for detection of Mycobacterium tuberculosis in clinical samples. J Clin Microbiol 1994;32:2283 2232. 26. Vanholder R, Dell'Aquila R, Ringoir S. Phagocytic function in uremia. Adv Exp Med Biol 1991;297:193-205. 27. Vanholder R, Ringoir S. Infectious morbidity and defects of phagocytic function in end-stage renal disease: a review. JAm Soc Nephrol 1992;3:1541-1554. 28. Sutowicz W, Komorowska Z, Hanicki Z, Cichocki T, Smolenski O. The enzymatic activities and NBT reduction test of granulocytes in untreated and dialysed uraemic patients. Acta Haematol 1979;61:144-149. 29. HirabayashiY, KobayashiT~ Nishikawa A, Okazaki H, Aoki T, Takaya J, et al. Oxidative metabolism and phagocytosis of polymorphonuclear leukocytes in patients with chronic renal failure. Nephron 1988;49:305-312.
253