Occult HBV infection—both hidden and mysterious

Abstract

Besisik F, Karaca C, Akyuz F, Horosanh S, Onel D, Badur S, Sever MS, Danahoglu A, Demir K, Kaymakoglu S, Cakaloglu Y, Okten A (Divisions of Hepatology and Nephrology, Department of Internal Medicine, and Department of Microbiology, Istanbul Medical Facility, Istanbul University, Istanbul, Turkey). Occult HBV infection and YMDD variants in hemodialysis patients with chronic HCV infection. J Hepatol 2003;38:506–510.These authors investigated the prevalence and clinical impact of occult hepatitis B virus (HBV) infection in hemodialysis patients with chronic hepatitis C virus (HCV) infection, and the frequency of YMDD (tyrosine-methionine-aspartate-aspartate amino-acid motif of HBV polymerase) variant in this setting. In 33 HCV RNA-positive, hepatitis B surface antigen (HBsAg)-negative, hemodialysis patients, HBV DNA was measured using a polymerase chain reaction (PCR) technique. YMDD mutation was studied in all HBV DNA-positive patients by the BOOM method. HBV DNA was detected in 12 of the 33 patients. The mean age, dialysis period, and biochemical parameters were not significantly different in these 12 HBV DNA-positive patients compared with the HBV DNA-negative patients. YMDD variants were identified in 6 of the 12 HBV DNA-positive patients, and there were no significant biochemical differences between those patients with or without the variants. The investigators conclude that occult HBV infection, with or without the appearance of YMDD variants, is common in hemodialysis patients with chronic HCV infection.CommentMore than 20 years ago, it was shown that patients who received HBsAg-negative, anti-hepatitis B core antibody (HBcAb)-positive blood were still at risk of developing post-transfusional hepatitis B (N Engl J Med 1978;298:1379–1383). In addition, patients who were HBsAg negative, hepatitis B surface antibody (HBsAb) positive, undergoing cancer chemotherapy, developed reactivation of HBsAg associated with severe, life-threatening hepatitis (Ann Intern Med 1982;96:447–449). With the development of PCR technology, it was clearly shown that not only patients with only hepatitis B antibodies in their serum, but also patients with chronic liver disease and with no serological markers of HBV infection, could still be shown to have HBV DNA present in their serum or liver (N Engl J Med 1985;312:270–276). Thus the concept of “occult” HBV infection emerged. Occult is defined as hidden or mysterious (The Shorter Oxford English Dictionary 1956;3:1355). The initial use of “occult” in this context was to describe this hidden HBV infection, hidden in the sense of levels of HBV that were undetectable except by the most sensitive of methods. HBsAg-positive patients generally have HBV DNA levels of 104 to 108 viral genomes per milliliter. In contrast, in chronic liver disease patients with “occult” infections, the circulating levels of HBV DNA are below 103 viral genomes per milliliter (Hepatology 2001;34:194–203), and liver HBV DNA levels are also low (Hepatology 2000;31:507–512). However, in general, about 33% of patients with HBsAg-negative hepatitis B antibody-positive chronic liver disease have detectable circulating HBV DNA by PCR (Hepatology 1991;13:1044–1051) and 40%–50% liver or mononuclear cell HBV DNA. However, the levels of liver HBV in HBsAg-negative patients are much lower than in HBsAg-positive patients (N Engl J Med 1990;323:80–85, J Hepatol 1998;29:173–183). It has also been shown that patients with serum hepatitis B core antibody (HBcAb) positivity are much more likely to have “occult” HBV DNA present than patients with only HBsAb positivity or no markers at all (Hepatology 1995;22:25–29, J Med Virol 1994;44:293–297).In such patients with severe cryptogenic chronic liver disease, there was a significant percentage, over 50%, with occult HBV infection in the livers (J Hepatol 2001;34:447–454), although the presence of other etiologic agents were not excluded (J Hepatol 2001;34:471–473). In the present series, the percentage of patients with “occult” HBV were the same in HBcAb-positive and -negative cases, although there were fewer “occult” cases in the patients with no HBV serum markers. The present study comes from Turkey, and “occult” HBV infection has been shown to occur in countries with both high and low prevalence of HBV infection.What is the mechanism of “occult” HBV infection? There is no evidence to support the contention that the findings are due to the presence of new viruses that cross react with HBV DNA testing (Lancet 1987;12:1354–1357). Does “occult” HBV infection mean then, that there are no circulating complete viral particles of wild-type HBV? The answer is no because circulating complete viral particles have been demonstrated in such patients (Immunol Rev 2000;174:98–111). Some investigations, but not all, have suggested that “occult” HBV infections are associated with particular genotypes of HBV (J Hepatol 1997;27:973–978). Genotypic studies were not performed in the present investigation but would be of interest. Alternatively, investigators have suggested that certain mutations of HBV might be associated with HBsAg-negative infections (J Med Virol 1999;58:193–195), especially mutations in the S gene (Hepatology 1997;26:1658–1666, J Virol 1998;72:7692–7696). Mutations in the preS1 sequence have recently been reported to be associated with reduced HBsAg expression, possibly explaining the HBsAg-negative, HBV DNA findings in these patients (Hepatology 2000;32:116–123). In the present study, the investigators showed that only half the patients with “occult” infection had a mutant virus present. However, the bulk of evidence is building up that “occult” HBV infections are caused by low viral replication rates (Hepatology 2001;34:194–203). This would explain the initial observations of exacerbations of hepatitis of HBsAg-negative patients undergoing chemotherapy (Ann Intern Med 1982;96:447–449); the several observations of persistence of HBV DNA in serum, mononuclear cells (Nat Med 1996;2:1104–1108), and the liver (PNAS 1981;78:3906–3910) in patients with hepatitis following serum conversion to HBsAg negativity; and the persistence of HBV DNA for up to 30 years, after a self-limited attack of acute hepatitis, without a sequence change (J Hepatol 2000;33:992–997). The question as to whether HBV is ever completely eliminated after a self-limiting attack of acute hepatitis B has been raised again in a recent study. In 14 such HBsAg-negative, HBsAb-positive patients, followed for 2 to 9 years, with a median of 4.2 years, 9 patients underwent liver biopsy at a median of 7 years after resolution of the acute infection. HBV DNA surface and X regions were found in 7 of the 9 livers, of whom 2 had detectable serum HBV DNA, associated with abnormal histology (Hepatology 2003;37:1172–1179). The finding of covalently closed circulating HBV DNA replicative intermediates by PCR suggested that the persistence of HBV DNA was caused by low replication rate rather than integration into the host genome.It has become clear that other exogenous factors may affect the replication rate of HBV. For example, there is evidence that chronic alcohol ingestion may down-regulate HBsAg expression, possibly explaining the increased incidence of HBsAg-negative, HBV DNA-positive infections in alcoholics (J Hepatol 1992;15:118–124). The association of chronic renal disease and dialysis may suppress the immune response in such patients. Nearly 10% of dialysis patients with HCV RNA positivity are anti-HCV antibody negative (J Intern Med 2002;251:119–128). In the present series, all the patients were HCV RNA positive. It has been repeatedly shown that a high percentage of such patients with chronic HCV hepatitis have a coinfection with HBV, manifesting as HBsAg negative, HBV DNA positivity in the serum, in about 20%–30% of cases (J Med Virol 1997;52:399–405, J Med Virol 1999;58:201–207), or in the liver in about 50% of cases (N Engl J Med 1999;341:22–26). Therefore, in the present series, the percentage of HCV-positive dialysis patients is not unexpected. What is the mechanism for the “occult” HBV infections in HCV-positive patients? The evidence is that coinfection results in the down-regulation of one virus, e.g. HBV, by the other virus, HCV, resulting in HBsAg negativity, HBcAb positivity (Blood 2002;99:4245–4246), despite ongoing but reduced viral replication (Gastroenterology 1993;105:1529–1533, J Med Virol 1995;46:258–264). What are the implications of this study? First, as far as the staff of the dialysis unit are concerned, as already stated, transmission of HBV infection has been documented from HBsAg-negative blood (N Engl J Med 1978;298:1379–1383). However, one would assume that all staff members of dialysis units would be vaccinated against HBV. What are the implications for the patients? The evidence that coinfection of occult HBV disease results in more severe HCV-induced liver disease is not strong (Hepatology 2001;34:204–206). Nevertheless, two recent studies have shown that coinfection is more common in patients with HCV-induced cirrhosis than in HCV-associated chronic hepatitis (Dig Dis Sci 1995;40:8–13, N Engl J Med 1999;341:22–26). In the present series, there were no significant biochemical differences, including serum ALT levels between HCV patients with or without occult HBV infection. However, there is evidence that ALT levels are suppressed in HCV-infected dialysis patients with hepatitis (Am J Kidney Dis 2001;38:1009–1115). Therefore, there is still the possibility that the coinfected patients may have more advanced histology. The second issue is that of therapy. There is the suggestion in the literature that the presence of “occult” HBV infection may impair the response of HCV to interferon (Arch Virol 1997;142:535–544). However, we are not aware of any data with coinfected patients involving “state of the art” treatment with PEG interferon and ribavirin (Hepatology 2001;34:204–206). The third issue is that of the development of hepatocellular carcinoma (HCC) in these coinfected patients. Patients with HBsAg-negative serum have nevertheless been found to have selected accumulation of the X transcript of HBV in the tumors of patients with HCC (Hepatology 1995;21:313–321). Studies have shown both: (1) an association of serum HBsAg and anti-HCV positivity increases the relative risk of HCC compared to case controls with anti-HCV positivity alone (Int J Cancer 1999;81:695–699); and (2) an increased rate of HCC occurs in HCV cirrhotic patients whose liver contained HBV DNA (Int J Oncol 1999;14:1153–1156).So should the finding of HBV DNA positivity in patients with HbsAg-negative, HCV RNA-positive serum change the management of such patients in any way? Dialysis patients with HBV positivity, but not those with HCV positivity, are separated from other dialysis patients. Therefore, the question must be raised as to whether HCV patients with “occult” HBV positivity should also be separated? If yes, then HBV DNA must be performed on all HCV-positive dialysis patients. The next issue is one of prognosis. There is evidence that anti-HCV-positive patients have a poorer prognosis because of the increased development of cirrhosis (Nephrol Dial Transplant 2001;16:1669–1674). The question as to the role of “occult” HBV in the prognosis of such patients remains to be determined. HCV and HCV+HBV infections are factors in the poorer outcome of patients with chronic renal failure following renal transplantation (Am J Kidney Dis 2001;38:919–934). The role of “occult” HBV infection in the natural history of such patients is unknown. The issue of anti-viral therapy was not raised by the authors. If this was being considered, one could make out a case for a liver biopsy, in the anticipation of finding unexpectedly severe liver disease. For the same reason, a stronger case for liver biopsy could be made if renal transplantation was being considered. A case could also be made for more intensive surveillance for HCC, perhaps by determining serial serum alpha feto-protein levels and abdominal ultrasonography every 3 months. Until we determine whether “occult” HBV infection in HCV-infected patients is a pathogenic cofactor, or just an innocent bystander, the “hidden” will also remain “mysterious.” In the meantime, it seems reasonable to us to err on the side of caution. Besisik F, Karaca C, Akyuz F, Horosanh S, Onel D, Badur S, Sever MS, Danahoglu A, Demir K, Kaymakoglu S, Cakaloglu Y, Okten A (Divisions of Hepatology and Nephrology, Department of Internal Medicine, and Department of Microbiology, Istanbul Medical Facility, Istanbul University, Istanbul, Turkey). Occult HBV infection and YMDD variants in hemodialysis patients with chronic HCV infection. J Hepatol 2003;38:506–510. These authors investigated the prevalence and clinical impact of occult hepatitis B virus (HBV) infection in hemodialysis patients with chronic hepatitis C virus (HCV) infection, and the frequency of YMDD (tyrosine-methionine-aspartate-aspartate amino-acid motif of HBV polymerase) variant in this setting. In 33 HCV RNA-positive, hepatitis B surface antigen (HBsAg)-negative, hemodialysis patients, HBV DNA was measured using a polymerase chain reaction (PCR) technique. YMDD mutation was studied in all HBV DNA-positive patients by the BOOM method. HBV DNA was detected in 12 of the 33 patients. The mean age, dialysis period, and biochemical parameters were not significantly different in these 12 HBV DNA-positive patients compared with the HBV DNA-negative patients. YMDD variants were identified in 6 of the 12 HBV DNA-positive patients, and there were no significant biochemical differences between those patients with or without the variants. The investigators conclude that occult HBV infection, with or without the appearance of YMDD variants, is common in hemodialysis patients with chronic HCV infection. CommentMore than 20 years ago, it was shown that patients who received HBsAg-negative, anti-hepatitis B core antibody (HBcAb)-positive blood were still at risk of developing post-transfusional hepatitis B (N Engl J Med 1978;298:1379–1383). In addition, patients who were HBsAg negative, hepatitis B surface antibody (HBsAb) positive, undergoing cancer chemotherapy, developed reactivation of HBsAg associated with severe, life-threatening hepatitis (Ann Intern Med 1982;96:447–449). With the development of PCR technology, it was clearly shown that not only patients with only hepatitis B antibodies in their serum, but also patients with chronic liver disease and with no serological markers of HBV infection, could still be shown to have HBV DNA present in their serum or liver (N Engl J Med 1985;312:270–276). Thus the concept of “occult” HBV infection emerged. Occult is defined as hidden or mysterious (The Shorter Oxford English Dictionary 1956;3:1355). The initial use of “occult” in this context was to describe this hidden HBV infection, hidden in the sense of levels of HBV that were undetectable except by the most sensitive of methods. HBsAg-positive patients generally have HBV DNA levels of 104 to 108 viral genomes per milliliter. In contrast, in chronic liver disease patients with “occult” infections, the circulating levels of HBV DNA are below 103 viral genomes per milliliter (Hepatology 2001;34:194–203), and liver HBV DNA levels are also low (Hepatology 2000;31:507–512). However, in general, about 33% of patients with HBsAg-negative hepatitis B antibody-positive chronic liver disease have detectable circulating HBV DNA by PCR (Hepatology 1991;13:1044–1051) and 40%–50% liver or mononuclear cell HBV DNA. However, the levels of liver HBV in HBsAg-negative patients are much lower than in HBsAg-positive patients (N Engl J Med 1990;323:80–85, J Hepatol 1998;29:173–183). It has also been shown that patients with serum hepatitis B core antibody (HBcAb) positivity are much more likely to have “occult” HBV DNA present than patients with only HBsAb positivity or no markers at all (Hepatology 1995;22:25–29, J Med Virol 1994;44:293–297).In such patients with severe cryptogenic chronic liver disease, there was a significant percentage, over 50%, with occult HBV infection in the livers (J Hepatol 2001;34:447–454), although the presence of other etiologic agents were not excluded (J Hepatol 2001;34:471–473). In the present series, the percentage of patients with “occult” HBV were the same in HBcAb-positive and -negative cases, although there were fewer “occult” cases in the patients with no HBV serum markers. The present study comes from Turkey, and “occult” HBV infection has been shown to occur in countries with both high and low prevalence of HBV infection.What is the mechanism of “occult” HBV infection? There is no evidence to support the contention that the findings are due to the presence of new viruses that cross react with HBV DNA testing (Lancet 1987;12:1354–1357). Does “occult” HBV infection mean then, that there are no circulating complete viral particles of wild-type HBV? The answer is no because circulating complete viral particles have been demonstrated in such patients (Immunol Rev 2000;174:98–111). Some investigations, but not all, have suggested that “occult” HBV infections are associated with particular genotypes of HBV (J Hepatol 1997;27:973–978). Genotypic studies were not performed in the present investigation but would be of interest. Alternatively, investigators have suggested that certain mutations of HBV might be associated with HBsAg-negative infections (J Med Virol 1999;58:193–195), especially mutations in the S gene (Hepatology 1997;26:1658–1666, J Virol 1998;72:7692–7696). Mutations in the preS1 sequence have recently been reported to be associated with reduced HBsAg expression, possibly explaining the HBsAg-negative, HBV DNA findings in these patients (Hepatology 2000;32:116–123). In the present study, the investigators showed that only half the patients with “occult” infection had a mutant virus present. However, the bulk of evidence is building up that “occult” HBV infections are caused by low viral replication rates (Hepatology 2001;34:194–203). This would explain the initial observations of exacerbations of hepatitis of HBsAg-negative patients undergoing chemotherapy (Ann Intern Med 1982;96:447–449); the several observations of persistence of HBV DNA in serum, mononuclear cells (Nat Med 1996;2:1104–1108), and the liver (PNAS 1981;78:3906–3910) in patients with hepatitis following serum conversion to HBsAg negativity; and the persistence of HBV DNA for up to 30 years, after a self-limited attack of acute hepatitis, without a sequence change (J Hepatol 2000;33:992–997). The question as to whether HBV is ever completely eliminated after a self-limiting attack of acute hepatitis B has been raised again in a recent study. In 14 such HBsAg-negative, HBsAb-positive patients, followed for 2 to 9 years, with a median of 4.2 years, 9 patients underwent liver biopsy at a median of 7 years after resolution of the acute infection. HBV DNA surface and X regions were found in 7 of the 9 livers, of whom 2 had detectable serum HBV DNA, associated with abnormal histology (Hepatology 2003;37:1172–1179). The finding of covalently closed circulating HBV DNA replicative intermediates by PCR suggested that the persistence of HBV DNA was caused by low replication rate rather than integration into the host genome.It has become clear that other exogenous factors may affect the replication rate of HBV. For example, there is evidence that chronic alcohol ingestion may down-regulate HBsAg expression, possibly explaining the increased incidence of HBsAg-negative, HBV DNA-positive infections in alcoholics (J Hepatol 1992;15:118–124). The association of chronic renal disease and dialysis may suppress the immune response in such patients. Nearly 10% of dialysis patients with HCV RNA positivity are anti-HCV antibody negative (J Intern Med 2002;251:119–128). In the present series, all the patients were HCV RNA positive. It has been repeatedly shown that a high percentage of such patients with chronic HCV hepatitis have a coinfection with HBV, manifesting as HBsAg negative, HBV DNA positivity in the serum, in about 20%–30% of cases (J Med Virol 1997;52:399–405, J Med Virol 1999;58:201–207), or in the liver in about 50% of cases (N Engl J Med 1999;341:22–26). Therefore, in the present series, the percentage of HCV-positive dialysis patients is not unexpected. What is the mechanism for the “occult” HBV infections in HCV-positive patients? The evidence is that coinfection results in the down-regulation of one virus, e.g. HBV, by the other virus, HCV, resulting in HBsAg negativity, HBcAb positivity (Blood 2002;99:4245–4246), despite ongoing but reduced viral replication (Gastroenterology 1993;105:1529–1533, J Med Virol 1995;46:258–264). What are the implications of this study? First, as far as the staff of the dialysis unit are concerned, as already stated, transmission of HBV infection has been documented from HBsAg-negative blood (N Engl J Med 1978;298:1379–1383). However, one would assume that all staff members of dialysis units would be vaccinated against HBV. What are the implications for the patients? The evidence that coinfection of occult HBV disease results in more severe HCV-induced liver disease is not strong (Hepatology 2001;34:204–206). Nevertheless, two recent studies have shown that coinfection is more common in patients with HCV-induced cirrhosis than in HCV-associated chronic hepatitis (Dig Dis Sci 1995;40:8–13, N Engl J Med 1999;341:22–26). In the present series, there were no significant biochemical differences, including serum ALT levels between HCV patients with or without occult HBV infection. However, there is evidence that ALT levels are suppressed in HCV-infected dialysis patients with hepatitis (Am J Kidney Dis 2001;38:1009–1115). Therefore, there is still the possibility that the coinfected patients may have more advanced histology. The second issue is that of therapy. There is the suggestion in the literature that the presence of “occult” HBV infection may impair the response of HCV to interferon (Arch Virol 1997;142:535–544). However, we are not aware of any data with coinfected patients involving “state of the art” treatment with PEG interferon and ribavirin (Hepatology 2001;34:204–206). The third issue is that of the development of hepatocellular carcinoma (HCC) in these coinfected patients. Patients with HBsAg-negative serum have nevertheless been found to have selected accumulation of the X transcript of HBV in the tumors of patients with HCC (Hepatology 1995;21:313–321). Studies have shown both: (1) an association of serum HBsAg and anti-HCV positivity increases the relative risk of HCC compared to case controls with anti-HCV positivity alone (Int J Cancer 1999;81:695–699); and (2) an increased rate of HCC occurs in HCV cirrhotic patients whose liver contained HBV DNA (Int J Oncol 1999;14:1153–1156).So should the finding of HBV DNA positivity in patients with HbsAg-negative, HCV RNA-positive serum change the management of such patients in any way? Dialysis patients with HBV positivity, but not those with HCV positivity, are separated from other dialysis patients. Therefore, the question must be raised as to whether HCV patients with “occult” HBV positivity should also be separated? If yes, then HBV DNA must be performed on all HCV-positive dialysis patients. The next issue is one of prognosis. There is evidence that anti-HCV-positive patients have a poorer prognosis because of the increased development of cirrhosis (Nephrol Dial Transplant 2001;16:1669–1674). The question as to the role of “occult” HBV in the prognosis of such patients remains to be determined. HCV and HCV+HBV infections are factors in the poorer outcome of patients with chronic renal failure following renal transplantation (Am J Kidney Dis 2001;38:919–934). The role of “occult” HBV infection in the natural history of such patients is unknown. The issue of anti-viral therapy was not raised by the authors. If this was being considered, one could make out a case for a liver biopsy, in the anticipation of finding unexpectedly severe liver disease. For the same reason, a stronger case for liver biopsy could be made if renal transplantation was being considered. A case could also be made for more intensive surveillance for HCC, perhaps by determining serial serum alpha feto-protein levels and abdominal ultrasonography every 3 months. Until we determine whether “occult” HBV infection in HCV-infected patients is a pathogenic cofactor, or just an innocent bystander, the “hidden” will also remain “mysterious.” In the meantime, it seems reasonable to us to err on the side of caution. More than 20 years ago, it was shown that patients who received HBsAg-negative, anti-hepatitis B core antibody (HBcAb)-positive blood were still at risk of developing post-transfusional hepatitis B (N Engl J Med 1978;298:1379–1383). In addition, patients who were HBsAg negative, hepatitis B surface antibody (HBsAb) positive, undergoing cancer chemotherapy, developed reactivation of HBsAg associated with severe, life-threatening hepatitis (Ann Intern Med 1982;96:447–449). With the development of PCR technology, it was clearly shown that not only patients with only hepatitis B antibodies in their serum, but also patients with chronic liver disease and with no serological markers of HBV infection, could still be shown to have HBV DNA present in their serum or liver (N Engl J Med 1985;312:270–276). Thus the concept of “occult” HBV infection emerged. Occult is defined as hidden or mysterious (The Shorter Oxford English Dictionary 1956;3:1355). The initial use of “occult” in this context was to describe this hidden HBV infection, hidden in the sense of levels of HBV that were undetectable except by the most sensitive of methods. HBsAg-positive patients generally have HBV DNA levels of 104 to 108 viral genomes per milliliter. In contrast, in chronic liver disease patients with “occult” infections, the circulating levels of HBV DNA are below 103 viral genomes per milliliter (Hepatology 2001;34:194–203), and liver HBV DNA levels are also low (Hepatology 2000;31:507–512). However, in general, about 33% of patients with HBsAg-negative hepatitis B antibody-positive chronic liver disease have detectable circulating HBV DNA by PCR (Hepatology 1991;13:1044–1051) and 40%–50% liver or mononuclear cell HBV DNA. However, the levels of liver HBV in HBsAg-negative patients are much lower than in HBsAg-positive patients (N Engl J Med 1990;323:80–85, J Hepatol 1998;29:173–183). It has also been shown that patients with serum hepatitis B core antibody (HBcAb) positivity are much more likely to have “occult” HBV DNA present than patients with only HBsAb positivity or no markers at all (Hepatology 1995;22:25–29, J Med Virol 1994;44:293–297). In such patients with severe cryptogenic chronic liver disease, there was a significant percentage, over 50%, with occult HBV infection in the livers (J Hepatol 2001;34:447–454), although the presence of other etiologic agents were not excluded (J Hepatol 2001;34:471–473). In the present series, the percentage of patients with “occult” HBV were the same in HBcAb-positive and -negative cases, although there were fewer “occult” cases in the patients with no HBV serum markers. The present study comes from Turkey, and “occult” HBV infection has been shown to occur in countries with both high and low prevalence of HBV infection. What is the mechanism of “occult” HBV infection? There is no evidence to support the contention that the findings are due to the presence of new viruses that cross react with HBV DNA testing (Lancet 1987;12:1354–1357). Does “occult” HBV infection mean then, that there are no circulating complete viral particles of wild-type HBV? The answer is no because circulating complete viral particles have been demonstrated in such patients (Immunol Rev 2000;174:98–111). Some investigations, but not all, have suggested that “occult” HBV infections are associated with particular genotypes of HBV (J Hepatol 1997;27:973–978). Genotypic studies were not performed in the present investigation but would be of interest. Alternatively, investigators have suggested that certain mutations of HBV might be associated with HBsAg-negative infections (J Med Virol 1999;58:193–195), especially mutations in the S gene (Hepatology 1997;26:1658–1666, J Virol 1998;72:7692–7696). Mutations in the preS1 sequence have recently been reported to be associated with reduced HBsAg expression, possibly explaining the HBsAg-negative, HBV DNA findings in these patients (Hepatology 2000;32:116–123). In the present study, the investigators showed that only half the patients with “occult” infection had a mutant virus present. However, the bulk of evidence is building up that “occult” HBV infections are caused by low viral replication rates (Hepatology 2001;34:194–203). This would explain the initial observations of exacerbations of hepatitis of HBsAg-negative patients undergoing chemotherapy (Ann Intern Med 1982;96:447–449); the several observations of persistence of HBV DNA in serum, mononuclear cells (Nat Med 1996;2:1104–1108), and the liver (PNAS 1981;78:3906–3910) in patients with hepatitis following serum conversion to HBsAg negativity; and the persistence of HBV DNA for up to 30 years, after a self-limited attack of acute hepatitis, without a sequence change (J Hepatol 2000;33:992–997). The question as to whether HBV is ever completely eliminated after a self-limiting attack of acute hepatitis B has been raised again in a recent study. In 14 such HBsAg-negative, HBsAb-positive patients, followed for 2 to 9 years, with a median of 4.2 years, 9 patients underwent liver biopsy at a median of 7 years after resolution of the acute infection. HBV DNA surface and X regions were found in 7 of the 9 livers, of whom 2 had detectable serum HBV DNA, associated with abnormal histology (Hepatology 2003;37:1172–1179). The finding of covalently closed circulating HBV DNA replicative intermediates by PCR suggested that the persistence of HBV DNA was caused by low replication rate rather than integration into the host genome. It has become clear that other exogenous factors may affect the replication rate of HBV. For example, there is evidence that chronic alcohol ingestion may down-regulate HBsAg expression, possibly explaining the increased incidence of HBsAg-negative, HBV DNA-positive infections in alcoholics (J Hepatol 1992;15:118–124). The association of chronic renal disease and dialysis may suppress the immune response in such patients. Nearly 10% of dialysis patients with HCV RNA positivity are anti-HCV antibody negative (J Intern Med 2002;251:119–128). In the present series, all the patients were HCV RNA positive. It has been repeatedly shown that a high percentage of such patients with chronic HCV hepatitis have a coinfection with HBV, manifesting as HBsAg negative, HBV DNA positivity in the serum, in about 20%–30% of cases (J Med Virol 1997;52:399–405, J Med Virol 1999;58:201–207), or in the liver in about 50% of cases (N Engl J Med 1999;341:22–26). Therefore, in the present series, the percentage of HCV-positive dialysis patients is not unexpected. What is the mechanism for the “occult” HBV infections in HCV-positive patients? The evidence is that coinfection results in the down-regulation of one virus, e.g. HBV, by the other virus, HCV, resulting in HBsAg negativity, HBcAb positivity (Blood 2002;99:4245–4246), despite ongoing but reduced viral replication (Gastroenterology 1993;105:1529–1533, J Med Virol 1995;46:258–264). What are the implications of this study? First, as far as the staff of the dialysis unit are concerned, as already stated, transmission of HBV infection has been documented from HBsAg-negative blood (N Engl J Med 1978;298:1379–1383). However, one would assume that all staff members of dialysis units would be vaccinated against HBV. What are the implications for the patients? The evidence that coinfection of occult HBV disease results in more severe HCV-induced liver disease is not strong (Hepatology 2001;34:204–206). Nevertheless, two recent studies have shown that coinfection is more common in patients with HCV-induced cirrhosis than in HCV-associated chronic hepatitis (Dig Dis Sci 1995;40:8–13, N Engl J Med 1999;341:22–26). In the present series, there were no significant biochemical differences, including serum ALT levels between HCV patients with or without occult HBV infection. However, there is evidence that ALT levels are suppressed in HCV-infected dialysis patients with hepatitis (Am J Kidney Dis 2001;38:1009–1115). Therefore, there is still the possibility that the coinfected patients may have more advanced histology. The second issue is that of therapy. There is the suggestion in the literature that the presence of “occult” HBV infection may impair the response of HCV to interferon (Arch Virol 1997;142:535–544). However, we are not aware of any data with coinfected patients involving “state of the art” treatment with PEG interferon and ribavirin (Hepatology 2001;34:204–206). The third issue is that of the development of hepatocellular carcinoma (HCC) in these coinfected patients. Patients with HBsAg-negative serum have nevertheless been found to have selected accumulation of the X transcript of HBV in the tumors of patients with HCC (Hepatology 1995;21:313–321). Studies have shown both: (1) an association of serum HBsAg and anti-HCV positivity increases the relative risk of HCC compared to case controls with anti-HCV positivity alone (Int J Cancer 1999;81:695–699); and (2) an increased rate of HCC occurs in HCV cirrhotic patients whose liver contained HBV DNA (Int J Oncol 1999;14:1153–1156). So should the finding of HBV DNA positivity in patients with HbsAg-negative, HCV RNA-positive serum change the management of such patients in any way? Dialysis patients with HBV positivity, but not those with HCV positivity, are separated from other dialysis patients. Therefore, the question must be raised as to whether HCV patients with “occult” HBV positivity should also be separated? If yes, then HBV DNA must be performed on all HCV-positive dialysis patients. The next issue is one of prognosis. There is evidence that anti-HCV-positive patients have a poorer prognosis because of the increased development of cirrhosis (Nephrol Dial Transplant 2001;16:1669–1674). The question as to the role of “occult” HBV in the prognosis of such patients remains to be determined. HCV and HCV+HBV infections are factors in the poorer outcome of patients with chronic renal failure following renal transplantation (Am J Kidney Dis 2001;38:919–934). The role of “occult” HBV infection in the natural history of such patients is unknown. The issue of anti-viral therapy was not raised by the authors. If this was being considered, one could make out a case for a liver biopsy, in the anticipation of finding unexpectedly severe liver disease. For the same reason, a stronger case for liver biopsy could be made if renal transplantation was being considered. A case could also be made for more intensive surveillance for HCC, perhaps by determining serial serum alpha feto-protein levels and abdominal ultrasonography every 3 months. Until we determine whether “occult” HBV infection in HCV-infected patients is a pathogenic cofactor, or just an innocent bystander, the “hidden” will also remain “mysterious.” In the meantime, it seems reasonable to us to err on the side of caution.