|
|
|
||
| BMJ
1998;316:553-554 (14 February) Letters Polymerase chain reaction as marker of infectivity in people with hepatitis CSummary vertical transmission rates may be misleading Editor–We recently completed a systematic review of worldwide published and unpublished data on vertical transmission of hepatitis C virus.1 We agree with Dore et al that the probability of transmission of the virus from a mother with antibody to hepatitis C virus but who is negative for the virus by polymerase chain reaction is very low,2 although such transmissions have been documented.1 We are concerned, however, about their calculation of summary vertical transmission rates (their table 1). Firstly, they have not included all the published studies that are eligible by their stated criteria (references available on request), including one that reported 100% transmission from mothers positive by polymerase chain reaction.3 Secondly, they have included two partially duplicated studies: most of the infants reported on by Paccagnini et al were included in the collaborative study by Zanetti et al (A Zanetti, unpublished communication). Thirdly, we think it inappropriate to include in these calculations the two large retrospective studies of mothers infected from contaminated anti-D immunoglobulin. Fourthly, we found in our review that use of reported transmission rates was problematic, as studies varied considerably in the diagnostic criteria used for infection and non-infection with hepatitis C virus and in the duration of follow up of the infants. Finally, we found considerable heterogeneity in transmission rates, even after applying standardised diagnostic criteria and stratifying by maternal hepatitis C viraemia status and HIV serostatus. This may have been due to varying distributions of other risk factors for transmission and also to differences in polymerase chain reaction methodologies. Dore et al suggest that developments in polymerase chain reaction technology have overcome the initial unreliability of the technique. Many of the studies included in the authors' table 1, however, were carried out before the second EUROHEP hepatitis C virus study, which showed continuing problems with both false positive and false negative results.4 It is important that all relevant data are included in systematic reviews and that the results of separate studies are not combined uncritically.5 In our opinion the summary transmission estimates reported by Dore et al are misleading. Variations among studies could be lessened by the development of standardised diagnostic criteria for vertical transmission of hepatitis C virus and by improvements in the reliability of polymerase chain reaction for hepatitis C virus RNA. Summary transmission risks could then be estimated and women of childbearing age who have hepatitis C could be more effectively counselled. Sara Thomas, Research student,a Andrew Hall, Reader in communicable disease epidemiology,a Marie-Louise Newell, Senior lecturer in epidemiology b a Infectious Disease Epidemiology Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT, b Department of Epidemiology and Biostatistics, Institute of Child Health, London WC1N 1EE
Authors' reply Editor–The main objective of our review was to assess whether hepatitis C virus RNA as detected by polymerase chain reaction could be used as a marker of infectiousness in a person with antibodies to the virus. The secondary objective was to estimate the risk of transmission of the virus through various modes. Thomas et al report that cases of transmission of the virus have been documented from people with undetectable hepatitis C virus RNA.1 We did not claim that such cases could not occur. An analogy can be drawn with transmission of hepatitis B virus, which has been documented from people with no evidence of hepatitis B surface antigen or hepatitis B e antigen.2 3 Nevertheless, absence of these markers is considered to indicate that a person is unlikely to be infectious. We concluded that hepatitis C virus RNA is a strong marker of infectiousness, and our conclusion that "negative results by polymerase chain reaction indicate an extremely low probability of transmission of hepatitis C" remains unchanged. We also raised the possibility of false positive results of polymerase chain reaction and recommended that a person with hepatitis C should be counselled on the basis of a persistently positive or negative polymerase chain reaction rather than a single assessment. Thomas et al are concerned that we did not include all published studies in our review, and they cite in evidence an early study of vertical transmission in 10 mother-infant pairs.4 The 100% (8/8) transmission rate from mothers positive for hepatitis C on polymerase chain reaction in this study may have been due to a lack of specificity, as seven of the eight infants who were positive on polymerase chain reaction were reported to have lost their hepatitis C antibodies by 12 months. Inclusion of this study would not have altered the pooled vertical transmission rate. Exclusion of Paccagnini et al's study (of 37 mother-infant pairs)5 would likewise have had little effect on our estimated rate of vertical transmission. We expressed concern in our discussion about including the two large retrospective studies of mothers infected by contaminated anti-D immunoglobulin and for this reason calculated vertical transmission rates with and without these studies. We see no reason to attribute the heterogeneity in hepatitis C transmission rates to variation in diagnostic criteria. Differences in the populations and sample sizes in the studies are a much more plausible explanation. In either case, estimation of pooled hepatitis C virus transmission rates, despite substantial heterogeneity, remains valid. Thus we strongly reject Thomas et al's claim that our estimates are misleading. Gregory J Dore, Lecturer in epidemiology,c John M Kaldor, Professor of epidemiology,c Geoffrey W McCaughan, Clinical associate professor d c National Centre in HIV Epidemiology and Clinical Research, University of New South Wales, Sydney, Australia, d A W Morrow Gastroenterology and Liver Centre, Royal Prince Alfred Hospital, Sydney, Australia
|
