- Email: [email protected]
Risk factors for HCV infection in two haemodialysis units in the Netherlands
Netherlands Journal of Medicine 50 Ž1997. 97–101
Original article
Risk factors for HCV infection in two haemodialysis units in the Netherlands I. Keur a , P.M. Schneeberger
b,)
, Y. van der Graaf c , J. Vos d , W.C. van Dijk e, L.-J. van Doorn f
a
b
Diatel Haemodialysis Centre, Amsterdam, Netherlands Laboratory of Virology, National Institute of Public Health and EnÕironmental Protection, PO Box 1, 3720 BA BilthoÕen, Netherlands c Department of Clinical Epidemiology, UniÕersity Hospital of Utrecht, Utrecht, Netherlands d Stichting Thuisdialyse, Utrecht, Netherlands e Department of Clinical Microbiology, SloterÕaartziekenhuis, Amsterdam, Netherlands f Diagnostic Centre SSDZ, Department of Molecular Biology, Delft, Netherlands Received 14 May 1996; revised 19 November 1996; accepted 2 December 1996
Abstract Background: In order to assess risk factors for HCV infection during haemodialysis, all patients receiving haemodialysis for more than 6 months in two separate units in the Netherlands were studied retrospectively. Methods: Antibodies to HCV, HCV-RNA and HCV genotypes were determined. Risk factors were identified by analysis of an extensive collection of clinical data. Results: In unit A, 8 out of 75 Ž11%. patients and in unit B 4 out of 122 Ž3%. patients had antibodies to HCV. Eleven out of the 12 anti-HCV-positive patients had detectable HCV-RNA. Genotyping showed the presence of 4 different genotypes in unit A Ž1, 1a, 2b, and 3a.. Three patients in unit B were infected with the same genotype Ž1b., where one of these patients was also infected with genotype 1a. One patient in unit B did not have detectable HCV-RNA. The risk of acquiring a HCV infection in unit A was associated with the number of blood transfusions. However, in unit B this risk was associated with the duration of dialysis. Other factors such as the number of surgical procedures were not associated with HCV infection. Conclusions: Blood transfusions and the dialysis process itself are important and independent risk factors for HCV transmission in dialysis patients. Surgical events do not appear to be important risk factors. However, relative risks may vary considerably between different dialysis centres. Keywords: Hepatitis C virus; Haemodialysis; Epidemiology
1. Introduction Hepatitis C virus ŽHCV. is the major aetiological agent of post-transfusion and community-acquired non-A, non-B hepatitis.w1x In dialysis patients the prevalence of antibodies to HCV ranges from 3 to )
Tel. q31 30 742391; fax q31 30 281168.
42% w2,3x. Blood transfusions play an important role in the transmission of HCV w1x. However, after the introduction of erythropoietin and routine screening for anti-HCV in blood donors, it became clear that the dialysis process itself is an independent risk factor for HCV infection w4–7x. Recent prospective studies on HCV infections in dialysis units, including patients who never received any blood transfusions,
0300-2977r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 3 0 0 - 2 9 7 7 Ž 9 6 . 0 0 0 9 2 - 7
98
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101
showed a yearly incidence of 1.7–4.9% w6,8x. The exact routes of HCV transmission in dialysis units are largely unknown and therefore the efficacy of preventive measures remains obscure w6x. In this study the prevalence of HCV infections was assessed retrospectively in two separate dialysis units, and was related to multiple putative risk factors.
2. Materials and methods 2.1. Patients From December 1991 to January 1992, 75 patients in unit A and 122 in unit B treated with haemodialysis for more than 6 months entered the study after informed consent. Both centres are situated in large cities in the Netherlands. Unit A is a limited-care haemodialysis centre. An average of 50% of the patients of unit B are on home dialysis. For each patient the following risk factors were assessed independently: intravenous drug abuse, exposure to blood products, period of dialysis, number of blood transfusions and operations. All major surgical procedures, such as kidney transplantation, removal of kidney, laparotomy, hysterectomy, herniorrhaphy orthopaedic surgery, Žpara-.thyroidectomy, prostatectomy, and mastectomy were included in the analysis. Minor surgical events such as dental procedures and shunt operations were not included. Blood samples from all patients were centrifuged within 2 h after collection. Serum was aliquotted and stored at y208C.
dinium–phenol–chloroform method w9x. For cDNA synthesis antisense primer HCV19 Žpos. y1 to y20; 5X-GTGCACGGTCTACGAGACCT-3X . was used. The first-round PCR mixture contained both antisense cDNA primer HCV 19 and sense-primer HCV35 Žpos. y318 to y296; TTGGCGGCCGCACTCCACCATGAATCACTCCCC;.. Nested PCR was performed with primers NCR3 Žpos. y314 to y288; GGGGCGGCCGCCACCATARRATCACTCCCCTGTGAGG 3X . and NCR4 Žpos. y66 to y47; AGCACCCTATCAGGCAGTACC.. Amplification products were analyzed by agarose gel electrophoresis and Southern blot hybridization using probe HCV17 Žpos y88 to y69; 5X-GAGTAGTGTTGGGTCGCGAA-3X .. All samples were tested in duplicate with negative and positive controls. 2.4. Genotyping The Inno-LiPA HCV I ŽInnogenetics, Antwerp, Belgium. genotyping assay is based on reverse hybridization of labelled PCR amplification products to specific oligonucleotide probes w10x. The HCV-LiPA II was performed according to the manufacturer’s instructions. It allows detection of genotypes 1, 2, 3, and 4 q 5 w11x. 2.5. Data analysis Differences between groups were analysed with the chi-squared test. To estimate the separate effect of each variable on the risk of HCV, unconditional logistic regression was performed and the results are expressed as odds ratios with 95% CI.
2.2. Serology Anti-HCV antibodies were determined by a second generation antibody assay ŽHCV EIA, UBI, Organon Teknika, Boxtel, The Netherlands.. Positive samples were tested with a second-generation recombinant immunoblot assay ŽRIBA-II, Ortho Diagnostic, Systems, Raritan, NJ, USA.. 2.3. RNA isolation, and reÕerse transcriptase PCR (RT-PCR) RNA was isolated from freshly frozen plasma samples by a modified version of the acid guani-
3. Results A total of 197 patients from two geographically separate dialysis units in the Netherlands were tested for markers of HCV infection. The characteristics of the patients from units A and B are summarised in Table 1. Twelve patients were positive for anti-HCV by ELISA and confirmed by RIBA. The prevalence of anti-HCV was 8 out of 75 Ž11%. in unit A and 4 out of 122 Ž3%. in unit B, which is significantly different Žodds ratio 3.5; 95% confidence interval 1.1–11.4..
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101 Table 1 Base-line characteristics of the patients from dialysis units A and B No. of patients
Unit A 75
Unit B 122
Male Ž%. Female Ž%. Mean age Žs.d.. Anti-HCV-positive Ž%. HCV-RNA-positive Ž%.
47 Ž63. 28 Ž37. 49 Ž16. 8 Ž11. 8 Ž11.
74 Ž61. 48 Ž39. 53 Ž14. 4 Ž3. 3 Ž3.
n.s. a n.s. n.s. P - 0.05 P - 0.05
No. of operations Mean Range
2.3 0–9
0.9 0–5
P s 0.001
No. of transfusions Mean Range
15 0–118
20 0–367
n.s.
Years on dialysis Mean Range
5.9 0–18
6.5 1–20
n.s.
a
Non-significant.
Base-line characteristics of both HCV-positive and HCV-negative patients are shown in Table 2. None of the patients included in the study was an intra-
99
venous drug user or received blood-derived products. Logistic regression showed that the number of blood transfusions was an independent risk factor for HCV infection in unit A Žodds ratio 1.04; 95% confidence limits 1.01–1.07.. However, in unit B there was an independent association between HCV infection and the duration of dialysis Žodds ratio 1.5; 95% confidence bounds 1.06–2.13.. Patients in unit A underwent more operations than patients in unit B, but the number of operations was not associated with the acquisition of HCV in either centre. Viraemia was tested in all anti-HCV-positive patients by RT-PCR. HCV-RNA was detected in 11 of 12 Ž91%. anti-HCV-positive patients. One patient in unit B was anti-HCV-positive but had no detectable HCV-RNA. All HCV-RNA-positive isolates were genotyped. In unit A genotype 1b was detected in 4, 2b in 1, and type 3a in 2 patients. One isolate was typed as genotype 1, but could not be further subtyped. The 3 HCV-positive patients in unit B had genotype 1b, whereas one of these patients also had type 1a. From these HCV-RNA-positive patients, serum samples were also obtained in 1994–1995. All patients were still HCV-RNA-positive and had the same genotype.
Table 2 Baseline characteristics of HCV-negative and HCV-positive patients Risk factor
Unit A
Unit B
Total unit A and B
HCV-neg.rHCV-pos
HCV-neg.rHCV-pos.
HCV-neg.rHCV-pos.
No. of operations F 1 No. of operations G 2
32 3 35 5 Odds ratio s 1.5 Ž95%CI; 0.3–8.9.
92 2 26 2 Odds ratio s 3.5 Ž95%CI; 0.3–37.5.
124 5 61 7 Odds ratio s 2.0 Ž95%CI; 0.5–8.9.
No. of transfusions F 5 No. of transfusions G 6
45 2 22 6 Odds ratio s 6.1 Ž95%CI; 0.9–48.4.
56 0 62 4 Odds ratio s infinite Ž95%CI – .
101 2 84 10 Odds ratio s 9.3 Ž95%CI; 1.5–59.5.
Years on dialysis F 4 Years on dialysis G 5
31 2 36 6 Odds ratio s 2.6 Ž95%CI; 0.4–20.1.
67 0 51 4 Odds ratio s infinite Ž95%CI – .
98 2 87 10 Odds ratio s 4.9 Ž95%CI; 1.0–35.2.
The number of operations, transfusions and duration of dialysis had a skewed distribution and for this reason percentiles were calculated. Differences were tested with a Wilcoxon rank-sum test. The odds ratios with 95% confidence intervals ŽCI. for the risk factors were calculated after dichomatization. The medians were used as cut-off level. The Mantel-Haenszel odds ratio ŽMH odds ratio. was calculated. 95% confidence interval ŽCI. between brackets.
100
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101
4. Discussion The prevalence of HCV infections in dialysis units is considerably higher than in the healthy population w2,3x. Although dialysis patients require intensive medical treatment, the reasons for the increased prevalence of HCV infections remain obscure. We therefore analysed the prevalence of HCV infections in 2 dialysis centres in the Netherlands. The present study is part of an extensive screening program of HCV infections in all dialysis patients in the Netherlands. The mean prevalence of anti-HCV-positive patients Ž"6.1%. is low compared to results from other studies that reported prevalence figures as high as 42% w3,6,12x. However, the prevalence of HCV infection differed significantly between the two units, indicative of differences between these two dialysis centres. The mean number of blood transfusions and the duration of dialysis did not differ between unit A and B. However, there was an independent association between the prevalence of anti-HCV and the number of blood transfusions only in unit A. Duration of dialysis was identified as an independent risk factor only in unit B. Both blood transfusions and the dialysis process itself have already been identified as possible risk factors w13x. The present study shows that local differences may influence the relative importance of these risk factors w5,8x. Remarkably, patients in unit A underwent more surgical procedures and also received more kidney transplants than those in unit B. The reasons for these differences are not known. However, the number of operations and kidney transplantations was not associated with the risk of acquiring HCV infection in either centre. On average 50% of the patients of unit B are on home dialysis, using their own dialysis machine, not shared by other patients. Nosocomial transmission of HCV within dialysis units has been shown to occur although the exact route of transmission remains obscure w5,14,15x. It has been suggested that patients on home dialysis may have a reduced risk of HCV infection compared to patients attending a centralised dialysis facility w16x. This may explain the lower prevalence of HCV in unit B. Ninety-two percent Ž11r12. of the anti-HCVpositive patients also had HCV-RNA. Since dialysis patients have limited immunocompetence, detection
of HCV-RNA may be a more accurate marker of HCV infection w17,18x. All HCV isolates were genotyped. Genotype 1b, which is the most prevalent genotype in Western Europe, was detected in the majority of the HCV-infected patients Ž6r11; 54%.. This is accordance with earlier reports w19x. The finding of 5 different genotypes Ž1, 1a, 1b, 2a, and 3a. within unit A is compatible with transmission from an exogenous source Že.g., blood transfusions.. On the other hand, all HCV-positive patients in unit B had genotype 1b. Since genotype 1b is the most common type, no epidemiological conclusions can be drawn from genotyping alone. Further discrimination between isolates belonging to the same genotype would require nucleotide sequence analysis of variable regions of the HCV-RNA genome w5x. Genotyping results of one patient in unit B are compatible with a mixed infection. Although the frequency of mixed infections in dialysis patients is unknown, mixed infections have been frequently described in haemophiliacs, suggesting exposure to multiple infectious sources. Analysis of new isolates from each HCV-infected patient, obtained in 1994 and 1995, revealed the same genotype as in 1991–1992. This confirms the chronicity of HCV infections. In conclusion, the exact routes of HCV transmission in dialysis patients appear to be complex and remain to be further elucidated. Apparently, differences in local circumstances, practices and conditions have an important impact on the epidemiology of HCV. Therefore, epidemiological analyses should be performed separately for each dialysis unit to elucidate unrecognized confounding factors and to define effective preventive measures.
Acknowledgements The authors would like to thank Dr. A.M van Loon for reviewing the manuscript. We also thank Henk Boswijk, Kitty van Hoek and Walter van der Vliet for their valuable technical assistance.
References w1x Kuo G, Choo QL, Alter HJ, et al. An assay for circulating antibodies to a major etiologic virus of human non-A,non-B hepatitis. Science 1989;244:362–364.
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101 w2x Van der Poel CL, Reesink HW, Mauser-Bunschoten EP, et al. Prevalence of anti-HCV antibodies confirmed by recombinant immunoblot in different population subsets in the Netherlands. Vox Sang 1993;61:30–36. w3x Jemni S, Ikbel K, Mahjoub J, Ghachem L, Bidet JM, Boukef K. Seropositivity to hepatitis C virus in Tunisian haemodialysis patients. Nouv Rev Fr Hematol 1994;36:349–351. ´ w4x Schlipkoter ¨ U, Roggendorf M, Ernst G, et al. Hepatitis C virus antibodies in haemodialysis patients. Lancet 1990;335:1409. w5x Allander T, Medin C, Jacobson SH, Grillner L, Persson MAA. Hepatitis C transmission in a hemodialysis unit: Molecular evidence for spread of virus among patients not sharing equipment. J Med Virol 1994;43:415–419. w6x Jadoul M, Cornu C, Van Ypersele de Strihou, and the UCL Collaborative Group. Incidence and risk factors for seroconversion in hemodialysis: a prospective study. Kidney Int 1993;44:1322–1326. w7x Huang CS, Ho MS, Yang CS, Lee CL, Tan CA. Hepatitis C markers in hemodialysis patients. J Clin Microbiol 1993;31:1764–1769. w8x Chan TM, Lok ASF, Cheng IKP, Chan RT. Prevalence of hepatitis C virus infection in hemodialysis patients: A longitudinal study comparing the results of RNA and antibody assays. Hepatology 1993;17:5–8. w9x Chomczynski P, Sacchi N. Single step method of RNA isolation by guanidinium thiocyanate–phenol–chloroform extraction. Anal Biochem 1987;162:152–159. w10x Stuyver L, Rossau R, Wyseur A, et al. Typing of hepatitis C virus isolates and characterization of new subtypes using a line probe assay. J Gen Virol 1993;74:1093–1102. w11x Simmonds P, Alberti A, Alter HJ, et al. A proposed system
w12x
w13x
w14x
w15x
w16x
w17x
w18x
w19x
101
for nomenclature of hepatitis C viral genotypes. Hepatology 1994;19:1321–1324. Da Silva Cardoso M, Koerner M, Epple S, Kramer R, Bundschu D, Kubanek B. Prevalence of HCV-RNA-positive patients in a dialysis unit in Germany. Nephron 1994;68:517–518. Irie Y, Hayashi H, Yokozeki K, Kashima T, Okuda K. Hepatitis C infection unrelated to blood transfusion in hemodialysis patients. J Hepatol 1994;20:557–559. Corcoran GD, Brink NS, Millar GM, et al. Hepatitis C virus infection in hemodialysis patients: a clinical and virological study. J Infect 1994;28:279–285. Caramelo C, Navas S, Alberola ML, Bermejillo T, Reyero A, Carreno V. Evidence against transmission of hepatitis C virus in ultrafiltrate and peritoneal fluid. Nephron 1994;66:470– 473. Pascual J, Teruel JL, Liano F, Ortuno J. Home dialysis protects against hepatitis C virus transmission. Nephron 1993;64:314. Bukh J, Wantzin P, Krogsgaard K, Knudsen F, Purcell RH, Miller RH. High prevalence of hepatitis C virus ŽHCV. RNA in dialysis patients; failure of commercially available antibody tests to identify a significant number of patients with HCV infection. J Infect Dis 1993;168:1343–1348. Medin C, Allender T, Roll M, Jacobson SH, Grillner L. Seroconversion to hepatitis C virus in dialysis patients: a retrospective study. Nephron 1993;65:40–45. Van Doorn LJ, Kleter GEM, Stuyver L, et al. Sequence analysis of hepatitis C virus genotypes 1 to 5 reveals multiple novel subtypes in the BENELUX countries. J Gen Virol 1996;76:1871–1876.
Original article
Risk factors for HCV infection in two haemodialysis units in the Netherlands I. Keur a , P.M. Schneeberger
b,)
, Y. van der Graaf c , J. Vos d , W.C. van Dijk e, L.-J. van Doorn f
a
b
Diatel Haemodialysis Centre, Amsterdam, Netherlands Laboratory of Virology, National Institute of Public Health and EnÕironmental Protection, PO Box 1, 3720 BA BilthoÕen, Netherlands c Department of Clinical Epidemiology, UniÕersity Hospital of Utrecht, Utrecht, Netherlands d Stichting Thuisdialyse, Utrecht, Netherlands e Department of Clinical Microbiology, SloterÕaartziekenhuis, Amsterdam, Netherlands f Diagnostic Centre SSDZ, Department of Molecular Biology, Delft, Netherlands Received 14 May 1996; revised 19 November 1996; accepted 2 December 1996
Abstract Background: In order to assess risk factors for HCV infection during haemodialysis, all patients receiving haemodialysis for more than 6 months in two separate units in the Netherlands were studied retrospectively. Methods: Antibodies to HCV, HCV-RNA and HCV genotypes were determined. Risk factors were identified by analysis of an extensive collection of clinical data. Results: In unit A, 8 out of 75 Ž11%. patients and in unit B 4 out of 122 Ž3%. patients had antibodies to HCV. Eleven out of the 12 anti-HCV-positive patients had detectable HCV-RNA. Genotyping showed the presence of 4 different genotypes in unit A Ž1, 1a, 2b, and 3a.. Three patients in unit B were infected with the same genotype Ž1b., where one of these patients was also infected with genotype 1a. One patient in unit B did not have detectable HCV-RNA. The risk of acquiring a HCV infection in unit A was associated with the number of blood transfusions. However, in unit B this risk was associated with the duration of dialysis. Other factors such as the number of surgical procedures were not associated with HCV infection. Conclusions: Blood transfusions and the dialysis process itself are important and independent risk factors for HCV transmission in dialysis patients. Surgical events do not appear to be important risk factors. However, relative risks may vary considerably between different dialysis centres. Keywords: Hepatitis C virus; Haemodialysis; Epidemiology
1. Introduction Hepatitis C virus ŽHCV. is the major aetiological agent of post-transfusion and community-acquired non-A, non-B hepatitis.w1x In dialysis patients the prevalence of antibodies to HCV ranges from 3 to )
Tel. q31 30 742391; fax q31 30 281168.
42% w2,3x. Blood transfusions play an important role in the transmission of HCV w1x. However, after the introduction of erythropoietin and routine screening for anti-HCV in blood donors, it became clear that the dialysis process itself is an independent risk factor for HCV infection w4–7x. Recent prospective studies on HCV infections in dialysis units, including patients who never received any blood transfusions,
0300-2977r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved. PII S 0 3 0 0 - 2 9 7 7 Ž 9 6 . 0 0 0 9 2 - 7
98
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101
showed a yearly incidence of 1.7–4.9% w6,8x. The exact routes of HCV transmission in dialysis units are largely unknown and therefore the efficacy of preventive measures remains obscure w6x. In this study the prevalence of HCV infections was assessed retrospectively in two separate dialysis units, and was related to multiple putative risk factors.
2. Materials and methods 2.1. Patients From December 1991 to January 1992, 75 patients in unit A and 122 in unit B treated with haemodialysis for more than 6 months entered the study after informed consent. Both centres are situated in large cities in the Netherlands. Unit A is a limited-care haemodialysis centre. An average of 50% of the patients of unit B are on home dialysis. For each patient the following risk factors were assessed independently: intravenous drug abuse, exposure to blood products, period of dialysis, number of blood transfusions and operations. All major surgical procedures, such as kidney transplantation, removal of kidney, laparotomy, hysterectomy, herniorrhaphy orthopaedic surgery, Žpara-.thyroidectomy, prostatectomy, and mastectomy were included in the analysis. Minor surgical events such as dental procedures and shunt operations were not included. Blood samples from all patients were centrifuged within 2 h after collection. Serum was aliquotted and stored at y208C.
dinium–phenol–chloroform method w9x. For cDNA synthesis antisense primer HCV19 Žpos. y1 to y20; 5X-GTGCACGGTCTACGAGACCT-3X . was used. The first-round PCR mixture contained both antisense cDNA primer HCV 19 and sense-primer HCV35 Žpos. y318 to y296; TTGGCGGCCGCACTCCACCATGAATCACTCCCC;.. Nested PCR was performed with primers NCR3 Žpos. y314 to y288; GGGGCGGCCGCCACCATARRATCACTCCCCTGTGAGG 3X . and NCR4 Žpos. y66 to y47; AGCACCCTATCAGGCAGTACC.. Amplification products were analyzed by agarose gel electrophoresis and Southern blot hybridization using probe HCV17 Žpos y88 to y69; 5X-GAGTAGTGTTGGGTCGCGAA-3X .. All samples were tested in duplicate with negative and positive controls. 2.4. Genotyping The Inno-LiPA HCV I ŽInnogenetics, Antwerp, Belgium. genotyping assay is based on reverse hybridization of labelled PCR amplification products to specific oligonucleotide probes w10x. The HCV-LiPA II was performed according to the manufacturer’s instructions. It allows detection of genotypes 1, 2, 3, and 4 q 5 w11x. 2.5. Data analysis Differences between groups were analysed with the chi-squared test. To estimate the separate effect of each variable on the risk of HCV, unconditional logistic regression was performed and the results are expressed as odds ratios with 95% CI.
2.2. Serology Anti-HCV antibodies were determined by a second generation antibody assay ŽHCV EIA, UBI, Organon Teknika, Boxtel, The Netherlands.. Positive samples were tested with a second-generation recombinant immunoblot assay ŽRIBA-II, Ortho Diagnostic, Systems, Raritan, NJ, USA.. 2.3. RNA isolation, and reÕerse transcriptase PCR (RT-PCR) RNA was isolated from freshly frozen plasma samples by a modified version of the acid guani-
3. Results A total of 197 patients from two geographically separate dialysis units in the Netherlands were tested for markers of HCV infection. The characteristics of the patients from units A and B are summarised in Table 1. Twelve patients were positive for anti-HCV by ELISA and confirmed by RIBA. The prevalence of anti-HCV was 8 out of 75 Ž11%. in unit A and 4 out of 122 Ž3%. in unit B, which is significantly different Žodds ratio 3.5; 95% confidence interval 1.1–11.4..
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101 Table 1 Base-line characteristics of the patients from dialysis units A and B No. of patients
Unit A 75
Unit B 122
Male Ž%. Female Ž%. Mean age Žs.d.. Anti-HCV-positive Ž%. HCV-RNA-positive Ž%.
47 Ž63. 28 Ž37. 49 Ž16. 8 Ž11. 8 Ž11.
74 Ž61. 48 Ž39. 53 Ž14. 4 Ž3. 3 Ž3.
n.s. a n.s. n.s. P - 0.05 P - 0.05
No. of operations Mean Range
2.3 0–9
0.9 0–5
P s 0.001
No. of transfusions Mean Range
15 0–118
20 0–367
n.s.
Years on dialysis Mean Range
5.9 0–18
6.5 1–20
n.s.
a
Non-significant.
Base-line characteristics of both HCV-positive and HCV-negative patients are shown in Table 2. None of the patients included in the study was an intra-
99
venous drug user or received blood-derived products. Logistic regression showed that the number of blood transfusions was an independent risk factor for HCV infection in unit A Žodds ratio 1.04; 95% confidence limits 1.01–1.07.. However, in unit B there was an independent association between HCV infection and the duration of dialysis Žodds ratio 1.5; 95% confidence bounds 1.06–2.13.. Patients in unit A underwent more operations than patients in unit B, but the number of operations was not associated with the acquisition of HCV in either centre. Viraemia was tested in all anti-HCV-positive patients by RT-PCR. HCV-RNA was detected in 11 of 12 Ž91%. anti-HCV-positive patients. One patient in unit B was anti-HCV-positive but had no detectable HCV-RNA. All HCV-RNA-positive isolates were genotyped. In unit A genotype 1b was detected in 4, 2b in 1, and type 3a in 2 patients. One isolate was typed as genotype 1, but could not be further subtyped. The 3 HCV-positive patients in unit B had genotype 1b, whereas one of these patients also had type 1a. From these HCV-RNA-positive patients, serum samples were also obtained in 1994–1995. All patients were still HCV-RNA-positive and had the same genotype.
Table 2 Baseline characteristics of HCV-negative and HCV-positive patients Risk factor
Unit A
Unit B
Total unit A and B
HCV-neg.rHCV-pos
HCV-neg.rHCV-pos.
HCV-neg.rHCV-pos.
No. of operations F 1 No. of operations G 2
32 3 35 5 Odds ratio s 1.5 Ž95%CI; 0.3–8.9.
92 2 26 2 Odds ratio s 3.5 Ž95%CI; 0.3–37.5.
124 5 61 7 Odds ratio s 2.0 Ž95%CI; 0.5–8.9.
No. of transfusions F 5 No. of transfusions G 6
45 2 22 6 Odds ratio s 6.1 Ž95%CI; 0.9–48.4.
56 0 62 4 Odds ratio s infinite Ž95%CI – .
101 2 84 10 Odds ratio s 9.3 Ž95%CI; 1.5–59.5.
Years on dialysis F 4 Years on dialysis G 5
31 2 36 6 Odds ratio s 2.6 Ž95%CI; 0.4–20.1.
67 0 51 4 Odds ratio s infinite Ž95%CI – .
98 2 87 10 Odds ratio s 4.9 Ž95%CI; 1.0–35.2.
The number of operations, transfusions and duration of dialysis had a skewed distribution and for this reason percentiles were calculated. Differences were tested with a Wilcoxon rank-sum test. The odds ratios with 95% confidence intervals ŽCI. for the risk factors were calculated after dichomatization. The medians were used as cut-off level. The Mantel-Haenszel odds ratio ŽMH odds ratio. was calculated. 95% confidence interval ŽCI. between brackets.
100
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101
4. Discussion The prevalence of HCV infections in dialysis units is considerably higher than in the healthy population w2,3x. Although dialysis patients require intensive medical treatment, the reasons for the increased prevalence of HCV infections remain obscure. We therefore analysed the prevalence of HCV infections in 2 dialysis centres in the Netherlands. The present study is part of an extensive screening program of HCV infections in all dialysis patients in the Netherlands. The mean prevalence of anti-HCV-positive patients Ž"6.1%. is low compared to results from other studies that reported prevalence figures as high as 42% w3,6,12x. However, the prevalence of HCV infection differed significantly between the two units, indicative of differences between these two dialysis centres. The mean number of blood transfusions and the duration of dialysis did not differ between unit A and B. However, there was an independent association between the prevalence of anti-HCV and the number of blood transfusions only in unit A. Duration of dialysis was identified as an independent risk factor only in unit B. Both blood transfusions and the dialysis process itself have already been identified as possible risk factors w13x. The present study shows that local differences may influence the relative importance of these risk factors w5,8x. Remarkably, patients in unit A underwent more surgical procedures and also received more kidney transplants than those in unit B. The reasons for these differences are not known. However, the number of operations and kidney transplantations was not associated with the risk of acquiring HCV infection in either centre. On average 50% of the patients of unit B are on home dialysis, using their own dialysis machine, not shared by other patients. Nosocomial transmission of HCV within dialysis units has been shown to occur although the exact route of transmission remains obscure w5,14,15x. It has been suggested that patients on home dialysis may have a reduced risk of HCV infection compared to patients attending a centralised dialysis facility w16x. This may explain the lower prevalence of HCV in unit B. Ninety-two percent Ž11r12. of the anti-HCVpositive patients also had HCV-RNA. Since dialysis patients have limited immunocompetence, detection
of HCV-RNA may be a more accurate marker of HCV infection w17,18x. All HCV isolates were genotyped. Genotype 1b, which is the most prevalent genotype in Western Europe, was detected in the majority of the HCV-infected patients Ž6r11; 54%.. This is accordance with earlier reports w19x. The finding of 5 different genotypes Ž1, 1a, 1b, 2a, and 3a. within unit A is compatible with transmission from an exogenous source Že.g., blood transfusions.. On the other hand, all HCV-positive patients in unit B had genotype 1b. Since genotype 1b is the most common type, no epidemiological conclusions can be drawn from genotyping alone. Further discrimination between isolates belonging to the same genotype would require nucleotide sequence analysis of variable regions of the HCV-RNA genome w5x. Genotyping results of one patient in unit B are compatible with a mixed infection. Although the frequency of mixed infections in dialysis patients is unknown, mixed infections have been frequently described in haemophiliacs, suggesting exposure to multiple infectious sources. Analysis of new isolates from each HCV-infected patient, obtained in 1994 and 1995, revealed the same genotype as in 1991–1992. This confirms the chronicity of HCV infections. In conclusion, the exact routes of HCV transmission in dialysis patients appear to be complex and remain to be further elucidated. Apparently, differences in local circumstances, practices and conditions have an important impact on the epidemiology of HCV. Therefore, epidemiological analyses should be performed separately for each dialysis unit to elucidate unrecognized confounding factors and to define effective preventive measures.
Acknowledgements The authors would like to thank Dr. A.M van Loon for reviewing the manuscript. We also thank Henk Boswijk, Kitty van Hoek and Walter van der Vliet for their valuable technical assistance.
References w1x Kuo G, Choo QL, Alter HJ, et al. An assay for circulating antibodies to a major etiologic virus of human non-A,non-B hepatitis. Science 1989;244:362–364.
I. Keur et al.r Netherlands Journal of Medicine 50 (1997) 97–101 w2x Van der Poel CL, Reesink HW, Mauser-Bunschoten EP, et al. Prevalence of anti-HCV antibodies confirmed by recombinant immunoblot in different population subsets in the Netherlands. Vox Sang 1993;61:30–36. w3x Jemni S, Ikbel K, Mahjoub J, Ghachem L, Bidet JM, Boukef K. Seropositivity to hepatitis C virus in Tunisian haemodialysis patients. Nouv Rev Fr Hematol 1994;36:349–351. ´ w4x Schlipkoter ¨ U, Roggendorf M, Ernst G, et al. Hepatitis C virus antibodies in haemodialysis patients. Lancet 1990;335:1409. w5x Allander T, Medin C, Jacobson SH, Grillner L, Persson MAA. Hepatitis C transmission in a hemodialysis unit: Molecular evidence for spread of virus among patients not sharing equipment. J Med Virol 1994;43:415–419. w6x Jadoul M, Cornu C, Van Ypersele de Strihou, and the UCL Collaborative Group. Incidence and risk factors for seroconversion in hemodialysis: a prospective study. Kidney Int 1993;44:1322–1326. w7x Huang CS, Ho MS, Yang CS, Lee CL, Tan CA. Hepatitis C markers in hemodialysis patients. J Clin Microbiol 1993;31:1764–1769. w8x Chan TM, Lok ASF, Cheng IKP, Chan RT. Prevalence of hepatitis C virus infection in hemodialysis patients: A longitudinal study comparing the results of RNA and antibody assays. Hepatology 1993;17:5–8. w9x Chomczynski P, Sacchi N. Single step method of RNA isolation by guanidinium thiocyanate–phenol–chloroform extraction. Anal Biochem 1987;162:152–159. w10x Stuyver L, Rossau R, Wyseur A, et al. Typing of hepatitis C virus isolates and characterization of new subtypes using a line probe assay. J Gen Virol 1993;74:1093–1102. w11x Simmonds P, Alberti A, Alter HJ, et al. A proposed system
w12x
w13x
w14x
w15x
w16x
w17x
w18x
w19x
101
for nomenclature of hepatitis C viral genotypes. Hepatology 1994;19:1321–1324. Da Silva Cardoso M, Koerner M, Epple S, Kramer R, Bundschu D, Kubanek B. Prevalence of HCV-RNA-positive patients in a dialysis unit in Germany. Nephron 1994;68:517–518. Irie Y, Hayashi H, Yokozeki K, Kashima T, Okuda K. Hepatitis C infection unrelated to blood transfusion in hemodialysis patients. J Hepatol 1994;20:557–559. Corcoran GD, Brink NS, Millar GM, et al. Hepatitis C virus infection in hemodialysis patients: a clinical and virological study. J Infect 1994;28:279–285. Caramelo C, Navas S, Alberola ML, Bermejillo T, Reyero A, Carreno V. Evidence against transmission of hepatitis C virus in ultrafiltrate and peritoneal fluid. Nephron 1994;66:470– 473. Pascual J, Teruel JL, Liano F, Ortuno J. Home dialysis protects against hepatitis C virus transmission. Nephron 1993;64:314. Bukh J, Wantzin P, Krogsgaard K, Knudsen F, Purcell RH, Miller RH. High prevalence of hepatitis C virus ŽHCV. RNA in dialysis patients; failure of commercially available antibody tests to identify a significant number of patients with HCV infection. J Infect Dis 1993;168:1343–1348. Medin C, Allender T, Roll M, Jacobson SH, Grillner L. Seroconversion to hepatitis C virus in dialysis patients: a retrospective study. Nephron 1993;65:40–45. Van Doorn LJ, Kleter GEM, Stuyver L, et al. Sequence analysis of hepatitis C virus genotypes 1 to 5 reveals multiple novel subtypes in the BENELUX countries. J Gen Virol 1996;76:1871–1876.