Spousal Transmission of the Hepatitis C Virus?

A vast amount has been learned about hepatitis C virus (HCV) in the short time since its discovery through molecular cloning [1]. Much still remains to be defined, however, particularly in regard to clinical outcome, pathogenesis, and modes of transmission. Our discussion focuses on the last item.

When first identified, non-A, non-B (C) hepatitis was believed to be acquired almost exclusively through percutaneous exposure. This early view derived naturally from the fact that the virus was first recognized in the context of blood transfusion studies. Subsequent investigations moved from the transfusion setting to that of community-acquired hepatitis, focusing on an analysis of risk factors for viral acquisition. An early study involving five Baltimore hospitals described 295 patients, 42% of whom fulfilled the then-existing criteria for non-A, non-B hepatitis [2]. Risk factors associated with the illness included previous receipt of blood transfusions, parenteral drug use, and hospital employment. Also showing a strong correlation, however, was “personal contact with other persons with hepatitis.” Investigators at the Centers for Disease Control and Prevention enlarged on these observations with a series of epidemiologic “sentinel county” surveys and also concluded that sexual contact was a factor in transmission, although a small one [3-5]. Of these, the most compelling involved a comparison between 52 patients with non-A, non-B hepatitis and 104 matched controls [5]. Multivariate analysis showed a significant association between sexual contact with two or more partners in the preceding 6 months and the presence of HCV infection and a near-significant association between general household exposure to hepatitis and HCV infection.

Although these epidemiologic studies seemed to support the concept of sexual transmission of HCV, in none of them did heterosexual contact represent more than 6% to 8% of all identifiable risk factors. Most perplexing in these surveys was the fact that the mode of transmission remained undefined in 30% to 50% of cases.

With the advent of increasingly sensitive serologic assays for HCV, it was anticipated that sexual contact as a basis for transmission would now be more clearly established. Instead, most serosurveys involving homosexual men [6, 7], prostitutes and their clients [8], persons attending clinics for sexually transmitted disease [9], and, until recently, spouses of HCV-positive persons [10, 11] have identified HCV antibodies in only 4% to 8% after correcting for coexisting parenteral drug abuse. This relatively low frequency of HCV infection contrasts sharply with the high frequency of infection with the hepatitis B virus, the human immunodeficiency virus (HIV), and other proven sexually transmitted agents in these same promiscuous populations. Nevertheless, it is generally agreed that HCV can be transmitted sexually, albeit inefficiently. This conclusion derives from the fact that the background prevalence of HCV antibodies, based on screening volunteer Red Cross blood donors, is about 0.6%, although it is argued that this figure may not be representative of the general population because blood donors are a highly select, prescreened group. Confounding the issue further is that screening for HCV RNA of body fluids that might be vehicles for transmission, such as saliva and semen, has yielded conflicting results [12, 13].

In this issue, Akahane and colleagues [14] now present more convincing data in support of sexual transmission of HCV. Spouses (52 men and 102 women) of 154 HCV-infected persons were screened with sophisticated assays for HCV antibodies and HCV RNA. Care was taken to exclude spouses with known parenteral risk factors for HCV infection or who might have had extramarital sexual relationships. Antibodies for HCV were detected in 27% and HCV RNA in 18% of the heterosexual contacts of HCV-positive index patients, and no significant difference in frequency was found between men and women. Moreover, the frequency of HCV infection correlated with the duration of marriage. No HCV markers were detected among those married for less than 10 years, whereas markers were found with increasing frequency with each additional decade of marriage, reaching a striking 60% among the five spouses who had been married for more than 50 years. The authors also found a 90% concordance of HCV genotype between the index patients and the exposed spouses, a finding not duplicated among infected spouses excluded from analysis because they were thought to have been exposed by other means. That sexual transmission has clinical relevance is underscored by the fact that 10 of 24 infected spouses had biochemical or histologic evidence of chronic liver disease.

These results are not dissimilar from those of five other recent surveys of sexual or household contacts of HCV-infected persons [15-19], three done in Asian countries [17-19]. Combined data show that 50 of 242 spouses (20.7%) had evidence of HCV infection compared with 31 of 409 (7.6%) nonsexual household contacts. In contrast, previous surveys had concluded that evidence was minimal for sexual transmission of HCV. In two studies [20, 21] using first-generation HCV assays, only 1 of 79 (1.3%) heterosexual partners of infected index patients was HCV-positive. Five recent European studies identified HCV antibodies in 9 of 248 (3.6%) heterosexual partners [22-25], and an ongoing National Institutes of Health study using second-generation assays has found only 1 of 73 (1.4%) partners to be HCV-positive [26].

How do we reconcile these widely divergent data? Akahane and colleagues suggest that in view of generally low circulating levels of HCV (in contrast to hepatitis B virus), transmission is more likely to occur if exposure is repetitive and extends over time. Many of the earlier surveys involved less repetitively exposed sexual contacts. Recent unpublished studies by one of us at the National Institutes of Health, using branched DNA technology, show widely fluctuating levels of HCV RNA during the course of the chronic HCV infection, and it is possible that sexual transmission would occur only at times of peak viral replication. Indirect support for the viral concentration theory derives from the results of a study by Eyster and colleagues [27] showing that 7 of 140 (4%) female partners of persons with hemophilia positive for both HCV and HIV were infected with HCV, compared with none of 30 female partners of persons with hemophilia positive for HCV in the absence of HIV. The assumption is that HCV concentrations are higher in the immunocompromised host, a view bolstered by the evidence that maternal-fetal transmission of HCV is more likely to occur if the mother is infected with both HCV and HIV [28].

Another intriguing consideration is that nonpercutaneous transmission is somehow modified by the genotypic expression of the virus. Genotype I predominates in the United States, whereas most cases in Japan consist of genotypes II and III. Indeed, chronic hepatitis C in Japan seems to differ in several regards from chronic hepatitis C in the United States. The disease in Japan appears to progress more rapidly, to culminate more commonly in primary hepatocellular carcinoma, and to respond better to interferon treatment. Might the rate and efficiency of sexual transmission also differ between the two countries? Certainly, measurement of HCV concentration among the different genotypes seems warranted.

A third possible explanation for the apparent disparate sexual transmission rates is that the higher prevalence in the Japanese study may not, in fact, reflect sexual transmission but rather may result from covert parenteral exposure that would less likely occur in Western populations. Kiyosawa has traced the high HCV prevalence (20%) in one Japanese community to the practice of folk medicine involving skin puncture with shared instruments [29]. Presumably, such practices would be engaged in by both the index patient and the spouse and perhaps even other members of the household. The study by Akahane and colleagues [14] does not provide the data necessary to exclude this possibility or provide prevalence rates in nonsexual household contacts. Hence, although sexual transmission remains the most viable explanation for the findings in the Akahane study, covert parenteral exposure, particularly the practice of folk medicine, has not been fully excluded.

In the face of this uncertainty, how should the HCV-infected person be counseled regarding routine behavior in the household setting and with respect to sexual activities? Undoubtedly, any practice that promotes transfer of blood or blood-contaminated body fluids should be entirely proscribed. This would include the sharing of toothbrushes, dental appliances, razors, and nail-grooming equipment. Little evidence implicates shared eating utensils, glassware, or crockery in the transmission of this virus. By far the greater concern is what advice to offer regarding sexual practices. Until the issue is better resolved, we suggest the following approach:

1. The HCV “carrier” should be fully informed of the currently available data that suggest that a small but real possibility exists for sexual transmission of HCV. This would apply especially to immunocompromised persons.

2. Despite the questionable detection of HCV markers in saliva, no convincing evidence has shown that the disease is transmitted by casual kissing [30].

3. Sexual practices in long-standing, monogamous relationships probably need not be modified other than to advise abstention if overt bleeding is present, such as during menstrual periods or hematuria or in the presence of overt genital ulceration. It is wise, however, to examine the partner for HCV infection to determine whether medical evaluation and therapy are required.

4. Promiscuous sexual behavior requires safe sex practice, including the use of condoms, which is already mandated for the prevention of HIV transmission.

5. The more difficult circumstance on which to advise regards the initial sexual contacts between a couple who hope ultimately to establish a long-term, monogamous relationship. The most reasonable approach currently is to present the existing facts clearly, indicating that sexual transmission is uncommon but not impossible. The couple could then decide whether they perceived the risk sufficient to warrant routine condom use.

Clearly, much is yet to be learned about modes of transmission of HCV in general and about the hazard of sexual transmission in particular. Ultimately, prevention will rest on the development of a specific immune globulin and vaccine, prospects that are remote because of HCV heterogeneity and subpar immune responses.

• Copyright ©2004 by the American College of Physicians