SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Background  We previously reported high prevalence of hepatitis C virus genotype 5a (HCV 5) (14%) in Central France.

Aim  To identify the risk factors associated with HCV5 infection and to characterize local HCV5 lineages.

Method  A case-control study and phylogenetic analysis were conducted.

Results  In all, 131 HCV5 and 343 HCV non 5 infected patients were enrolled. No HCV5 patient was born in sub-Saharan Africa and only two were injection drug user. HCV5 contamination was associated with living in a rural area called Vic le Comte (VLC) in non-transfused patients (OR = 17.7), with transfusion in patients living outside VLC (OR = 3.8) and with receiving injections in patients from VLC (OR = 3.1). More than 80% of the patients from outside VLC were contaminated by transfusion and those from VLC mainly by an iatrogenic factor – injections performed before 1972 by the local physician. Phylogenetic analysis of HCV5 isolates evidenced no distinct genetic cluster, but close relationships between the isolates of spouse pairs and between blood donors and recipients.

Conclusions  Our results suggest that HCV5 spread in our district by iatrogenic route before 1972 and then via transfusion to the whole district. Collaborative studies are underway to study viral sequences from different parts of Africa and Europe to estimate the origin of our HCV 5a strains.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Hepatitis C virus genotype 5a (HCV 5) is mainly found in northern South Africa, where it accounts for up to 40% of the genotypes of HCV infected patients.1, 2 Elsewhere, HCV 5 infections are rare and very few studies have been published on its geographical distribution,3–7 pathogenicity, treatment response8 and phylogenetic features.7 No case control study has been conducted on the risk factors associated with HCV 5 contamination.

A multicentre study was recently conducted in France9 to assess genotype distribution in patients with chronic hepatitis C. Data on 1769 patients from 26 referral centres scattered throughout the country documented HCV 1 in 57%, HCV 2 in 9%, HCV 3 in 21%, HCV 4 in 9%, HCV 5 in 3%, and mixed infections in 1%.

A prospective typing was performed in the Virology department of the University Hospital of Clermont-Ferrand in more than 1200 unselected HCV-infected and viremic patients, of whom 83 were co-infected with HIV 1. The distribution of HCV genotypes between 1996 and 2002 showed an unusually high rate of genotype 5a (14%) that was stable over time. Its frequency was higher in patients older than 50 years.10

These findings prompted us to conduct a case-control study to assess whether living in Vic le Comte (VLC) was associated with HCV 5 infection and to study the risk factors possibly associated with HCV 5 contamination in this area and in other parts of our district. We also performed phylogenetic analysis to identify the HCV 5 lineages of our district and to compare them with published HCV 5 sequences from other geographical areas, including South Africa.

Patients and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Selection of patients for the case-control study

All the 1260 patients with chronic hepatitis C and an HCV strain typed by the Laboratory of Virology of Clermont-Ferrand teaching hospital, Central France, between January 1991 and June 2002 were included in the database. Overall, only the 1205 patients seen in the Gastroenterology or Infectious Diseases departments were selected, including HIV co-infected subjects. All the patients and their physician were contacted by mail. If no response was received, a second letter was sent to the patients, and/or the physician was called by phone.

All HCV 5 infected patients seen in the two departments were included as cases in the study, except those who declined to take part. Control patients were randomly selected in the database from HCV non 5 infected patients and were matched by age category (<35, 35–49, 50–64 and ≥65 years) and gender.

Two epidemiological research assistants (ERA) interviewed cases and controls by phone using a standardized questionnaire to study the potential risk factors for hepatitis C listed in Table 1. Standardized operational procedures have been given to ERA under supervision of a project manager to homogenize data collection. Time periods of exposure were also recorded. The interviewers were not informed of the patients’ genotype status but it was not possible to blind them for place of residence. Several HCV 5 infected patients were referred to our hospital by physicians from few rural communes around the small town of VLC in the south east of our district which comprises a total population of 25 000. Hence, present and past places of residence were included in the epidemiological questionnaire and selected as an exposure factor to calculate the number of patients required for the study.

Table 1.   Potential risk factors of hepatitis C virus infection recorded in the epidemiological questionnaire
  1. * Mesotherapy is a technique in which small amounts of medication are injected into the connective tissue to stimulate the mesoderm.

  2. † Autohemotherapy is a technique in which a small amount of blood is taken from a vein in the patient’s arm and then reinjected sometimes mixed with a remedy subcutaneously or intramuscularly.

  3. ‡ VLC, Vic le Comte (rural area in the south east of our district). Living in VLC means for a period of at least 12 months.

Alternative medicine (acupuncture, mesotherapy* and autohemotherapy†)
Beauty/cosmetic treatment: tattooing, body/ear piercing, peeling, professional pedicure or manicure, shaving at barbershop
Birth place
Blood transfusion (including platelet or plasma transfusion, previous administration of coagulation factor concentrates) before 1991
Dental care: tooth implant, root curettage, tooth extraction, cyst excision, periodontium treatment, tartar removal, dental caries care, bridge, crown or a pivot tooth
Dermatological lesion treatment (bistoury use, wound suture, suture clip, curettage, pedicure, wounds, whitlows, furuncles, corns, warts and varicose ulcers) and varicose vein sclerotherapy
Endoscopy or catheterization
Gynaecological and obstetrical history (pregnancy associated procedures, abortion, miscarriage, caesarian)
Hemodialysis
Hospitalization for medical reason
Incarceration
Injections (at least five subcutaneous and/or intra-muscular and/or intra-venous injections and/or infiltrations and/or desensitization injections, specific immuno globulin injections): date, number and place
Intra-venous and intra-nasal drug use
Military service in France or abroad
Place and time period of residence (VLC‡ or outside VLC)
Sexuality (number of sexual partners, homo or bisexual, usual or occasional partners, HCV infected partner)
Surgery
Travel abroad (trip, stay and medical care)

The local ethics committee approved this study and all patients gave informed written consent to take part. Database was authorized by the Commission Nationale de l’Informatique et des Libertés.

Statistical analysis

The proportion of HCV non 5 infected patients from the Gastroenterology department database living in VLC was 10% (exposure rate in control patients); that of patients in the Infectious Diseases department was unknown. We estimated therefore that the exposure rate was 7% for controls from the two departments. Thus, to evidence an odds ratio value of 3 (α = 0.05 and 1−β = 90%), 123 cases were required and 369 controls for unmatched patients (EPI-INFO version 6.01). Univariate analyses were stratified according to age and sex (Cochran-Mantel-Haenszel test). Risk factors found to be significantly associated with HCV 5 infection (P value ≤ 0.10) were entered into a conditional logistic regression model. When there was interaction, OR was calculated according to each level of the variable. All the calculations were performed with SAS system version 8.02.

Transfusional investigations

Transfusional investigations were conducted by the French Blood Institution Auvergne Loire.

Virological study of HCV genotype 5

RT-PCR amplification and sequencing

Reverse transcription-polymerase chain reaction (RT-PCR) amplification and sequencing of NS5B and E1–E2 genes were carried out as previously described.9, 11 A specifically HCV 5 antisense primer was designed as follows: 5′-GGGCAGTWCTRTTGATRTGCC-3′ (nucleotide positions 1600–1620, according to the numerotation of H77 isolate).

Nucleotide sequence accession numbers

Twenty sequences described in this study were submitted to the EMBL data bank and assigned accession no. AJ608776 to AJ608785, AJ626926 and AJ626990 to AJ626998 for NS5B and no. AM177433 to AM177452 for E1–E2.

Phylogenetic analyses

Alignments were generated by CLUSTAL X software.12 Phylogenetic analysis was carried out with MEGA version 3.0. The Tamura-Nei algorithm calculated evolutionary distances, and trees were produced by the Neighbour-Joining method. The reliability of the phylogenetic topologies was estimated by the bootstrap resampling test with 1000 replicates.13 In the comparative analyses of the NS5B genes and the E1–E2 sequences, the prototype sequences used were designated as follows: 1a, HCV-1 (M62321); 5a, SA-13 (AF064490), EUH1480 (Y13184); BE95 (L29583), BE96 (L29584), FrSSD104 (AJ291281), MRS83 (AF516051), P10 (AF515976), FR741 (D50467), NOE (AY373483), AZA (AY373468).

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

Epidemiological study

One hundred and seventy HCV 5 infected patients were eligible. All were contacted and 39 (23%) were excluded (20 died, 17 were unable or unwilling to participate, one was lost to follow-up, and one was a genotype mistake). Finally, 131 cases were enrolled in the study. Every case was matched with three controls and only two controls for cases older than 65 years. Four hundred and ninety three controls were randomly selected and 150 (30%) were excluded (42 died, 71 were unable or unwilling to participate, 34 were lost to follow-up, and three were unmatched patients). Finally, 343 control patients were included. The different genotypes were distributed as follows: HCV 1 in 74%, HCV 2 in 10%, HCV 3 in 12%, HCV 4 in 3.5% and HCV 6 in 0.5% of the patients. Overall, 474 patients took part. The participation rate was 77% and 70% respectively in cases and in controls.

Age distribution was not significantly different between HCV 5 infected patients and controls (= 0.39). Ninety seven out of 131 HCV 5 infected patients (74%) and 245/343 HCV non 5 infected patients (71%) were older than 50 years. Sex ratio M/F was 0.90 in HCV 5 infected patients and 0.95 in controls (P = 0.79). Four hundred and four out of 474 patients were born in France, 40 elsewhere in Europe, 20 in North Africa, three in sub-Saharan Africa, five in Asia and two in America. All were Caucasians, except two cases born in Reunion and three HCV non 5 infected patients born in Brazil, Ivory Coast and Central African Republic.

Univariate analysis of potential risk factors recorded in the epidemiological questionnaire identified two risk factors positively associated with HCV 5 infection (Table 2): living in VLC [ORa = 8.7 (5.0–15.3)] and military service [ORa = 2.2 (1.2–4.3)]. Intravenous or intranasal drug use (IDU) [ORa = 0.04 (0.01–0.2)] and incarceration [ORa = 0.2 (0.1–0.8)] were negatively associated with HCV 5 infection.

Table 2.   Risk factors associated with contamination in 131 cases [hepatitis C virus (HCV) 5 infected patients] as compared to 343 controls (HCV non 5 infected patients) on univariate analysis
 Cases (%)Controls (%)ORa [95%CI]P
  1. VLC, Vic le Comte.

  2. ORa = Age and sex adjusted odds ratio.

Birth place
 France111 (84.7)293 (85.4) 0.5
 Sub-Saharian Africa03 (0.9)  
 Others20 (15.3)47 (13.7)  
Place of residence
 Living in VLC51 (38.9)24 (7.0)8.7 [5.0–15.3]<0.0001
Military service43 (69.4)84 (50.3)2.2 [1.2–4.3]<0.02
 In France26 (60.5)37 (44.1)  
 In sub-Saharan Africa0 (0)1 (1.2)  
 In other countries17 (39.5)46 (54.8)  
Travel abroad (trip or medical care)93 (71.0)251 (73.2)0.9 [0.6–1.4]0.7
 Stay/trip in sub-Saharan Africa7 (5.3)35 (10.2)0.5 [0.2–1.2]0.3
 Stay/trip in other countries86 (65.7)216 (63.0)1.0 [0.6–1.6]
Blood transfusion83 (63.4)189 (55.1)1.4 [0.9–2.1]0.1
Intravenous or intranasal drug use2 (1.5)61 (17.8)0.04 [0.01–0.2]<0.0001
Nosocomial exposure129 (98.5)334 (97.4)1.7 [0.4–8.5]0.5
Surgery119 (90.8)316 (92.1)0.8 [0.4–1.6]0.5
Hospitalization for medical reason54 (41.2)147 (43.0)0.9 [0.6–1.4]0.7
Gynaecological and obstetrical history62 (89.9)161 (91.5)0.8 [0.3–2.3]0.7
 Pregnancy associated procedures56 (82.4)138 (78.4)1.3 [0.6–2.8]0.6
 Abortion, miscarriage, caesarian35 (51.5)86 (48.9)1.2 [0.7–2.2]0.5
Hemodialysis3 (2.3)7 (2.0)1.0 [0.5–4.0]1
Endoscopy or catheterization78 (59.5)165 (48.1)1.5 [1.0–2.3]0.05
Iatrogenic exposure (out-patient care)130 (99.2)341 (99.4)0.8 [0.1–9.2]0.9
Alternative medicine59 (45.0)127 (37.0)1.4 [0.9–2.1]0.1
Dermatological treatment and varicose vein sclerotherapy92 (70.2)226 (65.9)1.2 [0.8–1.9]0.3
Having received more than 5 injections79 (60.3)207 (60.3)1.0 [0.6–1.5]0.9
Dental care125 (95.4)337 (98.3)0.4 [0.1–1.3]0.1
Beauty/cosmetic treatment85 (64.9)231 (67.4)0.8 [0.5–1.4]0.5
Incarceration3 (2.2)31 (9.0)0.2 [0.1–0.8]<0.02
High risk sexual behaviour33 (27.1)106 (33.5)0.8 [0.5–1.3]0.3

Logistic regression analysis identified three risk factors associated with HCV 5 contamination (Table 3): living in VLC, transfusion, and IDU.

Table 3.   Logistic regression analysis of risk factors associated with [hepatitis C virus (HCV) 5] infection in 131 HCV 5 infected patients and 343 HCV non 5 infected patients
Risk factorsOR95%CI
  1. VLC, Vic le Comte; OR, odds ratio.

Living in VLC30.4712.01–77.28
Transfusion3.351.71–6.56
Intravenous or intranasal drug use0.120.02–0.64

There was a negative interaction between living in VLC and transfusion. Living in VLC was strongly associated with HCV 5 infection in non-transfused patients [OR = 31.4 (12.3–80.5)] but not in transfused patients [OR = 3.7 (0.4–32.2)]. Transfusion was associated with HCV 5 infection [OR = 3.6 (1.8–7.2)] in patients living outside VLC but not in those living in VLC [OR = 0.4 (0.1–2.9)].

The multivariate analysis confirmed that living in VLC was highly associated with HCV 5 infection. To investigate the risk factors related to VLC, the percentage of patients exposed to transfusion, IDU or other factors in HCV 5 and non 5 infected patients were compared according to place of residence (VLC or outside VLC). Results (Table 4) showed that the risk factor distribution was different between HCV 5 and non 5 infected patients according to place of residence. In patients living outside VLC, exposure to transfusion was significantly more frequent in HCV 5 infected patients than in controls (83.8% vs. 55.5%, < 0.0001). In patients living in VLC, exposure to factors other than transfusion and IDU was significantly more frequent in HCV 5 infected patients than in controls (68.3% vs. 41.7%, < 0.02).

Table 4.   Modes of contamination in cases and controls according to their place of residence
 Cases (%)Controls (%)P
  1. VLC, Vic le Comte; IDU, intravenous or intranasal drug use.

  2. * Two patients were exposed both to transfusion and to IDU.

Risk factors in patients living outside VLCn = 80n = 319<0.0001
 Transfusion67 (83.8)177 (55.5) 
 IDU2 (2.5)*59 (18.5) 
 Other factors13 (16.2)83 (26.0) 
Risk factors in patients living in VLCn = 51n = 24<0.02
 Transfusion16 (31.4)12 (50) 
 IDU02 (8.3) 
 Other factors35 (68.3)10 (41.7) 

We then compared the percentage of patients, in HCV 5 infected patients and in controls, exposed to iatrogenic factors according to place of out-patient care (exposure to iatrogenic factor in VLC or not). Exposure in VLC to each iatrogenic factor (alternative medicine, dermatological treatment, injections, dental care, beauty/cosmetic treatment) was statistically more frequent in HCV 5 infected patients than in controls (Table 5).

Table 5.   Iatrogenic risk factors (out-patient care) associated with infection in 130 [hepatitis C virus (HCV) 5] infected patients as compared to 341 HCV non 5 infected patients on univariate analysis according to place of exposure
 Cases (%)Controls (%)ORa [95%CI]P
  1. ORa = Age and sex adjusted odds ratio.

  2. * Exposure in VLC (Vic le Comte) means the patient had been exposed to iatrogenic factors in VLC. Absence of exposure in VLC means that the patient was not exposed to iatrogenic factors in VLC or was not exposed at all either in VLC or outside VLC.

Iatrogenic exposure (out-patient care) in VLC52 (39.7)44 (12.8)4.1 [2.6–6.6]<0.0001
Alternative medicine in VLC*17 (13.0)8 (2.5)6.4 [2.6–15.4]<0.0001
Dermatological care and varicose vein sclerotherapy in VLC*26 (19.9)13 (3.8)5.8 [2.9–11.8]<0.0001
Injections (more than 5) in VLC*36 (27.5)14 (4.1)8.6 [4.4–16.8]<0.0001
Dental care in VLC*48 (36.6)36 (10.5)4.5 [2.8–7.3]<0.0001

Finally, we performed a second multivariate analysis in which we included the factors associated with HCV contamination in the first multivariate analysis (place of residence, transfusion, IDU) and the five iatrogenic factors significantly associated with HCV 5 infection. Negative interaction was found between living in VLC and transfusion. Living in VLC in non-transfused patients, transfusion in patients living outside VLC, receiving injections in VLC and IDU were independently associated with HCV 5 infection (Table 6).

Table 6.   Final logistic regression analysis of risk factors associated with hepatitis C virus 5 infection
Risk factorsOR95%CI
  1. VLC, Vic le Comte; OR, odds ratio.

Living in VLC in non-transfused patients17.76.3–49.7
Transfusion in patients living outside VLC3.81.9–7.8
Injections (more than five injections) in VLC3.11.1–8.6
Alternative medicine in VLC1.20.4–3.8
Dermatological care and varicose vein sclerotherapy in VLC1.30.5–3.5
Dental care in VLC1.30.5–3.1
Beauty/cosmetic treatment in VLC0.90.3–2.6
Intravenous or intranasal drug use0.10.02–0.4

We then evaluated whether a relationship existed between the risk of acquiring HCV and the date of exposure (before 1972, 1972–1990 and after 1990) to the two identified risk factors. Seventy five percent of HCV 5 infected patients received more than five injections in VLC before 1972 (retirement year of the VLC general practitioner) versus 35.7% of HCV non 5 infected patients (= 0.0056). In contrast, exposure to transfusion was not significantly different between HCV 5 infected patients and controls in the three periods.

Transfusional investigations

Transfusional investigations conducted in our district by the French Blood Institution Auvergne-Loire confirmed that transfusion was the transmission route in 27 patients out of the 131 HCV 5 infected patients and in 17 patients out of the 343 HCV non 5 infected patients. Fifteen donors of 27 blood recipients (56%) with HCV 5 were living in VLC compared to two donors of 17 recipients (12%) with HCV non 5 (< 0.004).

Virological study

HCV genotyping and comparison of sequences

Thirty-four patients out of 131 HCV 5 infected patients (26%) received anti-viral therapy and thus were non-viremic. HCV RNA amplification was successful for both NS5B and E1–E2 genomic fragments in specimens from the remaining 97/131 patients (74%).

Sequence analysis was performed with two genome fragments, the partial NS5B gene encoding the RNA dependent RNA polymerase and the partial E1–E2 genes encoding envelope proteins, including the first hypervariable region (HVR1). The NS5B sequences determined in the 97 patients (sequences designated Clermont-Ferrand (CF)) were compared with the sequences recovered in 10 HCV 5 infected patients living in different geographical areas in France and with five prototype sequences available from the GenBank data base. For the E1–E2 genome region, CF sequences were compared with those of three other French patients and the two sequences available in Genbank (i.e. EUH1480 and the SA13 strains). All CF sequences belonged to genotype 5a according to the current classification.14

Phylogenetic analysis

A phylogenetic analysis was undertaken to determine whether HCV5 strains of our district exhibited significant differences from strains of other geographical areas. A phylogenetic tree for NS5B segment did not differentiate any cluster because of high similarity of all sequences (≥96%) (data not shown). To gain further insights into the clustering of the HCV sequences, the E1–HVR1 genomic region was analysed because this genomic segment is characterized by the highest diversity within the genome. Again, as observed with the NS5B gene, the CF HCV5 sequences did not cluster separately from the other sequences investigated (Figure 1). All the CF sequences were scattered within the phylogenetic tree, indicating that there was no genetic grouping correlated with the risk factors associated with HCV5 infection in the 97 patients (i.e. living in VLC or transfusion or having injections in VLC). However, 10 pairs of sequences were detected with a high phylogenetic significance (Figure 1), suggesting an epidemiological link between the corresponding pairs of patients. For six sequence pairs, the corresponding patients were blood donors and their recipients. For a seventh pair, the two corresponding patients were blood recipients from the same donor as evidenced by transfusional investigations. The donor of these two patients did not take part in the study. Finally, the three other pairs of closely related sequences were observed in three pairs of husband and wife.

image

Figure 1.  Phylogenetic analysis of the E1–HVR1 sequences from hepatitis C virus (HCV) 5 genomes (designated by CF-number of the strain) characterized within the Clermont-Ferrand district, three other French sequences (S2, LIMgas and ANGger) and the two reference sequences. The tree was reconstructed using the Tamura-Nei model and the Neighbour-Joining algorithm, using HCV-1a sequence of Choo4 as the out-group. Bootstrap values (%) are indicated for 1000 replicates above the branches of pairs of blood donor/recipient proved by transfusional investigations and pairs of husband/wife. The year of transfusion is given in brackets. CF 32 and CF 55 were sequenced from two blood recipients who had the same common donor.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

In a previous study, we reported an unusually high prevalence of HCV 5 infected patients in Central France.10 In our hospital, 14.2% of HCV infected patients were infected by HCV 5, compared to 0.7–2.6% nationally.9, 15 HCV 5 prevalence was stable in our hospital between 1991 and 2002, and this genotype was the third most frequent type after genotype 1 (59%) and genotype 3 (15%). In the present study, we found that living in a small delimited rural area, called VLC, was the major risk factor of HCV 5 infection in non-transfused patients. We also found that transfusion in patients living outside VLC and treatment by injections before 1972 in VLC were associated with HCV 5 contamination.

Our study has some limitations. First, HCV infected patients referred to teaching hospitals could be different from HCV patients seen in the general population. A previous study has shown that genotype distribution in blood donors was very similar to that in patients followed up in French referral centres.9, 16 It is then possible that genotype distribution is not different between patients seen in or outside our hospital. Our previous study showed that every physician in our district was following of 3.64 HCV infected patients on average17 as compared to 15 patients for every physician in VLC (personal communication). This difference suggests a higher HCV prevalence in VLC in addition to a better screening by the physicians of this area. Second, we cannot exclude the possibility of a recall bias. Patients were interviewed about events that occurred many years ago, and some of them could have forgotten some events, such as injections or other therapies. Some patients from VLC became aware of the local epidemic before the beginning of this study, but it is unlikely that they knew that HCV 5 was the study hypothesis. Furthermore, it also unlikely that patients infected with HCV 5 had a better recall of investigated exposure than patients infected with other genotypes. To limit this bias, our research assistants were blinded to the viral genotype. The most likely bias in our study is a non-differential misclassification which is well known to reduce associations (toward the null condition) but not to exaggerate them.18

In contrast to other HCV strains, the endemic source and epidemic spread of HCV genotype 5 remain unknown. High prevalence of HCV genotype 5 was first described in the northern part of South Africa1, 2 where transfusion and medical practices are posited as potential transmission routes.19 In 2001, Jover et al. reported 50 HCV 5 infected patients in a region in southeast Spain.3 The origin of infection was unknown in 80% of patients, 16% acquired the infection by blood transfusion and one patient was an IDU. In 2006, a Belgian group identified a cluster of 41 HCV 5 isolates in a rural area in West Flanders province, in patients mainly contaminated by transfusion and hemodialysis.7 Our study is the first case-control study investigating risk factors associated with HCV 5 infection. We first demonstrated that HCV 5 infection was strongly associated with living in a small rural community in the south east of our district (OR > 30). Results of the first multivariate analysis also suggested that transfusion was involved in HCV 5 spread. In contrast, no HCV 5 patient in our study was contaminated by intravenous drug abuse.

No HCV 5 infected patient was born in sub-Saharan Africa (Table 2) and there were not multiple immigrants from sub-Saharian Africa before 1972 in our district. Nosocomial exposure was not associated with HCV 5 infection (Table 2) and there was no health institution in VLC. Hence, iatrogenic exposure emerged as a potential source of contamination. Table 5 shows that HCV 5 infected patients were exposed more frequently to out-patient care in VLC than controls (40% vs. 13%, < 0.0001). In the second multivariate analysis, treatment by more than five injections was the only iatrogenic risk factor identified as significantly associated with HCV 5 infection. Unsafe medical procedures before 1985, such as the use of inadequately sterilized glass syringes, were demonstrated to be responsible for HCV infection in several studies.20–22 In a recent report, the most common risk factor associated with HCV infection in a small rural town in Argentina was the apparent use of reusable syringes by a nurse serving the community.23 In the rural area of our district, most of the injections were performed by one physician before 1972, the date of his retirement, suggesting that the spread of HCV 5 occurred in VLC more than 30 years ago. The HCV status of this physician, who died in 1980, remains unknown. Anecdotal reports provided by patients living in VLC treated by him suggest that the use of inadequately sterilized glass syringes or non-sterilized glass syringes may have contributed to patient-to-patient contamination. Other minor invasive procedures forgotten by patients could also be implicated in this process.

The second multivariate analysis (Table 6) showed that there were three risk factors independently associated with HCV 5 infection: living in VLC in non-transfused patients, transfusion in patients living outside VLC and receiving injections in VLC. This suggests that there were two groups of HCV 5 infected patients: those living in VLC and mainly infected by iatrogenic route, and those living outside VLC infected for the most part by transfusion (Table 6). Sixty seven out of 80 (84%) HCV 5 infected patients living outside VLC were exposed to transfusion and 35 out of 51 (69%) patients living in VLC were exposed neither to transfusion nor to IDU (Table 4). Finally, more than half of the 27 HCV 5 infected patients living outside VLC and infected by transfusion, as shown by transfusional investigations and phylogenetic analysis, received blood products from donors living in VLC. Thus, we can hypothesize that HCV 5 spread first in VLC by out-patient care (injections and probably other iatrogenic factors) undertaken by one physician before 1972 and then by transfusion to the other parts of the district in the 1970s and 1980s.

The usefulness of molecular techniques for investigating HCV transmission has been previously established in studies using phylogenetic analysis to investigate outbreaks of nosocomial HCV infection with a short time lapse between transmission and sampling for analysis.24–26 When analysis is performed several years after transmission, the time that elapsed between the transmission event and sample collection may be critical in the application of sequence analysis to the study of virus transmission.27 For example, the evolution of viral sequences in anti-D immunoglobulin recipients demonstrated that the changes observed in the HVR1 domain were not randomly distributed but rather strongly constrained.28 Moreover, in most studies, infections came from a clearly common identified source. In contrast, in the present study, patients were probably infected more than three decades ago by transfusion and iatrogenic route and the origin of HCV 5 infections is unknown. The NS5B nucleotide sequence was not found to be informative in identifying routes of transmission. Further, the phylogenetic analysis of the envelope region was mainly useful to highlight the close relationship between isolates of blood donors and their recipients, and between isolates of three spouse pairs. Phylogenetic analysis was able to associate the viral sequences of the 6 pairs of donors and recipients included in the study. However, no genetic cluster corresponding to VLC sequences was distinguished when we compared them with others from France or with the two complete sequences from South Africa and Edinburgh. In the recent Belgian study,7 phylogenetic analysis of Belgian and South Africa HCV 5 strains, based on E1 and NS4 viral sequences, revealed two distinct clusters of similar diversity, according to their geographic origin. A coalescent method demonstrated that HCV 5 has been spreading independently in these two countries for more than 120 years and suggested that there has been a common ancestral HCV 5 population in another country still not identified, different from South Africa. These results also suggested that Belgian HCV 5 strains were not introduced recently from South Africa as generally supposed. Further collaborative studies on relationships between viral sequences from a wide range of geographical origins are needed to investigate the origin of HCV 5 strains in our district and more widely in western countries.

In summary, the high prevalence of HCV 5 genotype we found in our district in a previous study could be explained by an outbreak, which started before 1972. Our results suggested that HCV 5 infection was first introduced in a small rural area of our district. The virus has spread in this area by unsafe injections and then, to other parts of our district by transfusion. The singularity of HCV 5 genotype has allowed us to evidence the iatrogenic transmission, certainly underestimated for more frequent genotypes, like genotype 1, and the amplification of transmission by transfusion, inducing an outbreak of hundred patients contaminated by the same HCV genotype.

Sample collection from HCV 5 infected patients born in different countries of Europe (i.e. France, Belgium and Spain) and Africa is in progress for further phylogenetic analyses, to understand the transmission history of HCV genotype 5 and trace the origin of this rare genotype in our district and in other geographical areas.

Acknowledgments

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References

We thank the members of the HCV Group of ANRS AC-11 for providing us with HCV 5 isolates (Alain S and Rogez S – Limoges, Chaix ML – Paris, Necker hospital, Gassin M – Nantes, Izopet J – Toulouse, Payan C - Angers, Roque-Afonso AM – Villejuif, Stoll-Keller F – Strasbourg, Thibault V – Paris, La Pitié Salpétrière hospital), Vanessa Tixier for her excellent technical assistance, Anne-Marie Fontanier for her implication in transfusional investigations, Dr Blanchet, Dr Bonduel, Dr Codegnat, Dr Dupouyet and Dr Thomas general practitioners from Vic le Comte, and Dr Van der Merwe for reviewing the manuscript. Declaration of personal interests: None. Declaration of funding interests: This study was funded in part by the French National Agency for Research on AIDS and Viral Hepatitis, and in part by the Programme Hospitalier de Recherche Clinique régional and the Société Nationale Française de Gastroentérologie.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. References
  • 1
    Ohno T, Mizokami M, Tibbs CJ, et al. New genotype of hepatitis C virus in South Africa. J Med Virol 1994; 42: 40913.
  • 2
    Smuts HE, Kannemeyer J. Genotyping of hepatitis C Virus in South Africa. J Clin Microbiol 1995; 33: 167981.
  • 3
    Jover R, Perez-Serra J, De Vera F, et al. Infection by genotype 5a of HCV in a district of Southeast Spain. Am J Gastroenterol 2001; 96: 30423.
    Direct Link:
  • 4
    Legrand-Abravanel F, Sandres-Sauné K, Barange K, et al. Hepatitis C virus genotype 5: epidemiological characteristics and sensitivity to combination therapy with interferon-alpha plus ribavirin. J Infect Dis 2004; 189: 1397400.
  • 5
    Levi JE, Takaoka DT, Garrini RH, et al. Three cases of infection with hepatitis C virus genotype 5 among Brazilian hepatitis patients. J Clin Microbiol 2002; 40: 26457.
  • 6
    Murphy DG, Willems B, Vincelette J, Bernier L, Cote J, Delage G. Biological and clinicopathological features associated with hepatitis C virus type 5 infections. J Hepatol 1996; 24: 10913.
  • 7
    Verbeeck J, Maes P, Lemey P, et al. Investigating the origin and spread of hepatitis C virus genotype 5a. J Virol 2006; 80: 42206.
  • 8
    Bonny C, Fontaine H, Poynard T, et al. Effectiveness of interferon plus ribavirin combination in the treatment of naïve patients with hepatitis C virus type 5. A French multicentre retrospective study. Aliment Pharmacol Ther 2006; 24: 593600.
  • 9
    Payan C, Roudot-Thoraval F, Marcellin P, et al. Changing of hepatitis C virus genotype patterns in France at the beginning of the third millennium: The GEMHEP GenoCII Study. J Viral Hepat 2005; 12: 40513.
  • 10
    Henquell C, Cartau C, Abergel A, et al. High prevalence of hepatitis C virus type 5 in Central France evidenced by a prospective study from 1996 to 2002. J Clin Microbiol 2004; 42: 30305.
  • 11
    Morice Y, Roulot D, Grando V, et al. Phylogenetic analyses confirm the high prevalence of hepatitis C virus (HCV) type 4 in the Seine-Saint-Denis district (France) and indicate seven different HCV-4 subtypes linked to two different epidemiological patterns. J Gen Virol 2001; 82: 100112.
  • 12
    Thompson JD, Gibson TJ, Plewniak F, Jeenmougin F, Higgins DG. The Clustal X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res 1997; 24: 487682.
  • 13
    Felsenstein J. Confidence limits on phylogenies, an approach using the bootstrap. Evolution 1985; 39: 78391.
  • 14
    Simmonds P, Bukh J, Combet C, et al. Consensus proposals for a unified system of nomenclature of hepatitis C virus genotypes. Hepatology 2005; 42: 96273.
  • 15
    Delarocque-Astagneau E, Campèse C, Bronowicki JP, et al. Surveillance de l’hépatite C à l’échelon national à partir des pôles de références volontaires, 2000–2001. BEH 2003; 17: 903.
  • 16
    Pillonel J, Laperche S. Surveillance épidémiologique des donneurs de sang homologues en france entre 1992 et 2002. Rapport de l’Institut national de Veille Sanitaire 2004.
  • 17
    Bonny C, Rayssiguier R, Ughetto S, et al. Medical practices and expectations of general practitioners in relation to hepatitis C virus infection in the Auvergne region. Gastroenterol Clin Biol 2003; 27: 10215.
  • 18
    Rothman KJ. Modern Epidemiology. Little Brown and Compagny, 1986.
  • 19
    Tanaka Y, Kurbanov F, Mano S, et al. Molecular tracing of the global hepatitis C virus epidemic predicts regional patterns of hepatocellular carcinoma mortality. Gastroenterology 2006; 130: 70314.
  • 20
    Frank C, Mohamed MK, Strickland GT, et al. The role of the parenteral antischistosomal therapy in the spread of hepatitis C virus in Egypt. Lancet 2000; 355: 88791.
  • 21
    Comandini UV, Tossini G, Longo MA, et al. Sporadic hepatitis C virus infection: a case-control study of transmission routes in a selected hospital sample of the general population in Italy. Scand J Infect Dis 1998; 30: 115.
  • 22
    Karmochkine M, Carrat F, Dos Santos O, Cacoub P., Raguin G. The GERMIVIC Study Group. A case-control study of risk factors for hepatitis C infection in patients with unexplained routes of infection. J Viral Hepatitis 2006; 13: 77582.
  • 23
    Picchio GR, Baré PC, Descalzi VI, et al. High prevalence of infection with a single hepatitis C genotype in a small rural community of Argentina. Liver Int 2006; 26: 6605.
  • 24
    Bracho MA, Gosalbes MJ, Blasco D, Moya A, Gonzales-Candelas F. Molecular epidemiology of a hepatitis C virus outbreak in a hemodialysis unit. J Clin Microbiol 2005; 43: 27505.
  • 25
    Bruguera M, Saiz JC, Franco S, et al. Outbreak of nosocomial hepatitis C virus infection resolved by genetic analysis of HCV RNA. J Clin Microbiol 2002; 40: 43636.
  • 26
    Halfon P, Roubicek C, Gerolami V, et al. Use of phylogenetic analysis of hepatitis C virus (HCV) hypervariable region 1 sequences to trace an outbreak of HCV in an autodialysis unit. J Clin Microbiol 2002; 40: 15415.
  • 27
    Lin HJ, Seeff LB, Barbosa L, Hollinger FB. Occurrence of identical hypervariable region 1 sequences of hepatitis C virus in transfusion recipients and their respective blood donors: divergence over time. Hepatology 2001; 34: 4249.
  • 28
    McAllister J, Casino C, Davidson F, et al. Long-term evolution of the hypervariable region of hepatitis C virus in a common-source-infected cohort. J Virol 1998; 72: 4893905.