Hepatitis Virus Infections, Including HIV Co-Infection
Part III:

6%-20% of HCV/HIV coinfected found to have false HCV negativity, may be related to low CD4 count; HBV false negatives also seen
     By Harvey S. Bartnof, MD

Dr. Bartnof is a Staff Physician at the AIDS Virus Education and Research Institute (AVERI) in San Francisco

"CCR5" Gene Mutation Increases HCV Viral Load

New research presented at the 8th Annual Retrovirus Conference indicates a gene mutation on immune "macrophage" cells is associated with an increased HCV viral load in patients with HCV mono-infection. And, the rate of having two "doses" ("homozygous") of the "CCR5 delta-32" mutation is much higher in this population than would be expected by chance alone. R. P. Woitas, MD of the University of Bonn in Germany also reported that among HIV/HCV co-infected patients, having one dose of the mutation ("heterozygous") was associated with a somewhat higher mean HCV viral load than among co-infected patients without the mutation ("wild-type"), but this was not statistically significant. Regarding HIV viral loads, those with one dose of the mutation had a lower mean level than those with wild-type, but this only reached statistical significance for those with HIV mono-infection, as has been reported previously. A similar trend occurred for HIV/HCV co-infected patients, but this was not statistically significant.

The "CCR5 delta 32" mutation gained much interest in the area of HIV/AIDS research when it was reported that being homozygous afforded near 100% protection against HIV infection (transmission). Being heterozygous is associated with a slower rate of HIV disease progression in many studies and a milder set of symptoms with primary HIV infection, both of which were confirmed or reported at this Conference. Approximately 1% of all Caucasians have one dose of the mutation, with higher rates in northern Europe. When tested, several race-ethnic minority groups have a lower rate of the mutation. The effect of "CCR5 delta 32" in HCV mono-infected and HIV/HCV co-infected patients had not been investigated previously. CCR5 is a major co-receptor for HIV entry into immune cell "macrophages."

Dr. Woitas investigated the rates of "CCR5 delta 32" and the resulting effects on HCV and HIV viral loads in several groups of patients: (1) 153 HCV mono-infected patients (11% women); (2) 130 HIV/HCV co-infected patients (5% women); (3) 102 HIV mono-infected patients (15% women); and (4) 102 healthy control blood donors. Hemophilia (males born with genetic blood clotting deficiency) was the risk factor for infection in 85% of the HCV-mono-infected patients and 93% of the HIV/HCV co-infected patients.

HCV genotype 1 (the most difficult to treat) was present in 68% of the HCV-mono-infected and 73% of the HIV/HCV co-infected patients. Treatment for HIV had been undertaken in 82% of the HIV-mono-infected, but only 65% of the HIV/HCV co-infected patients. HAART (highly active antiretroviral therapy) was prescribed for only 45% of the HIV/HCV co-infected patients, versus 70% of the HIV mono-infected patients. In contrast, only 2 NRTI (nucleoside reverse transcriptase inhibitor) drugs were prescribed for 12% of the HIV patients, but for 21% of the co-infected patients. ALT (liver enzyme, alanine aminotransferase) was the highest (34 international units per liter) in HIV/HCV co-infected patients, intermediate in HCV mono-infected patients (21 IU/L), and lowest in the HIV-mono-infected patients (16 IU/L).

CCR5 delta 32 "homozygosity" (double dose) was present in 7% of HCV-mono-infected patients, and was significantly higher than would be expected by chance alone (using "Hardy-Weinberg" equation). A double dose was not present in any of the HIV/HCV co-infected or HIV mono-infected patients, but was present in 1% of controls. One dose ("heterozygous") was present in 16% of HCV-mono-infected, 25% of HIV/HCV co-infected, 19% of HIV-mono-infected, and 15% of control patients.

What are the real-life implications of these findings?

When examining the HCV viral loads, having a double dose of the mutation was associated with a significantly higher level (approximately 50 million copies per milliliter) among HCV mono-infected than among those with one dose (approximately 30 million) or having wild-type (approximately 25 million). HIV/HCV-co-infected patients had even greater HCV viral loads than HCV-mono-infected patients. Even though the HCV level among co-infected patients with one dose was greater (approximately 60 million) than wild-type (approximately 45 million), this difference was not statistically significant. The HCV viral load in HCV mono-infected patients with CCR5 delta-32 homozygosity (two doses) was rather similar to that in HIV/HCV co-infected patients with either one dose or no dose (wild-type).

Since HIV/HCV co-infected patients in several other "cross-sectional" studies have higher HCV viral loads than HCV mono-infected patients, and since HIV/HCV co-infected patients have a more accelerated HCV disease course than HCV mono-infected patients in several other studies (without anti-HIV treatment), these new findings might represent a subset of HCV mono-infected patients with more accelerated liver disease. However, the current study did not show this. The other significance is that higher HCV viral loads have been associated with a lower response rate to anti-HCV treatment.

Regarding HIV viral loads, there were no significant differences when comparing levels in HIV mono-infected with HIV/HCV co-infected patients when stratified by CCR5 wild type or CCR5 heterozygosity. As has been shown in other studies of HIV mono-infection without anti-HIV treatment, CCR5 heterozygotes had a significantly lower HIV viral load than those with wild-type. There was a similar trend in the current study for HIV/HCV co-infected patients, when comparing heterozygotes to wild-type, but the difference was not significant.

The implications and conclusions of the study require additional statistical "multivariate" analyses and confirmation in other patient cohorts. The "cross-sectional" nature also limits the findings somewhat. Nonetheless, the research adds to our knowledge of a new potential genetic co-factor that effects HCV viral load levels in HCV mono-infected and HIV/HCV co-infected patients. Also, it would appear that being homozygous for the CCR5 delta-32 mutation protects against HIV/HCV co-infection, as it does for HIV mono-infection, but not against HCV mono-infection and might even be associated with an increased risk for HCV mono-infection.

Beware of False Negative HCV Antibody Tests In HIV/HCV Co-Infected Patients

• If CD4 count is low, HCV antibody test may be negative, but chronic HCV infection might be present (positive HCV RNA).
• As CD4 count decreases, antibodies to HCV may no longer be detectable.
• As HAART leads to increases in the CD4 count, HCV antibodies may appear or reappear.

There were two presentations at the 8th Annual Retrovirus Conference that addressed the issue of "false negative" HCV antibody tests in HIV/HCV co-infected patients. This was the case for patients with low CD4 count (in 1 study, less than 150 cells per microliter; in the other, the exact level was not reported). The issue is an important one, since a physician and patient might incorrectly believe that there is no HCV infection in someone with HIV infection and a low CD4 count. This false impression would be due to a "false negative" HCV antibody test, but with a detectable (positive) HCV RNA viral load test. Yet, HCV infection could be present that might need treatment. Untreated HCV infection might lead to liver disease and liver-related death.

The first report was a series of ten HIV/HCV patients "haphazardly" identified at the Amelia Court HIV clinic at Parkland Hospital in Dallas, Texas. The lead author was Dr. R. Berggren.

The ten HIV/HCV co-infected patients were all men, 50% Caucasian, 30% Hispanic and 20% African-American. The behavioral risk profile for transmission was gay/bisexual men in 50%, injection drug use with or without gay/bisexual male behavior in 40% and blood transfusion in 10%. The median (half greater than, half less than) CD4 count was 44 cells per microliter (range 1-150), while the median HIV RNA was 5.5 log (314,1249) copies per milliliter (range 3.5-5.8 log or 3,029-645,806). A CD4 count less than 100 cells was present in 90%. The median HCV RNA was 6 log (1 million) copies per milliliter (range 3.3-6.4 log or 2,190-2.4 million). Only two had HCV genotype testing: both were 1a (type 1 is the most difficult to treat). The median ALT (liver enzyme, alanine aminotransferase) was abnormal at 54 units per liter.

The "second generation" Abbott 2.0 EIA (enzyme immunoassay) antibody test was used and represents a commonly used antibody test in the US currently. For HCV RNA, Roche "Amplicor" testing was used, with a lower limit or 3 log or 1,000 copies per milliliter.

Past infection with or vaccine to hepatitis B was present in 80%, as indicated by a positive test to hepatitis B core. The other two patients did have antibodies to hepatitis A and syphilis. These various positive test results represent evidence by the authors that at the time of a negative antibody test to HCV, there were antibodies present to other infections. Therefore, they suggest that the lack of HCV antibodies is not due to total immune "burnout," i.e. not producing any antibodies.

Based upon stored blood samples, the researchers were able to document that 20% of the patients previously had HCV antibodies, but they became undetectable as their CD4 counts decreased. Interestingly, three other patients with undetectable HCV antibody tests while having low CD4 counts developed detectable HCV antibodies after HAART led to CD4 count increases and HIV RNA decreases. The authors concluded, "Physicians should consider HCV PCR [viral load testing] in [HIV] patients with low CD4 [counts] to accurately screen for HCV infection."

In the second report, Michael P. Busch, MD, PhD of Blood Centers of the Pacific in San Francisco documented false negative antibody tests for hepatitis C in HIV/HCV co-infected patients (they had a positive HCV RNA test). He also documented "false negative hepatitis B surface antigen" tests in HIV/HBV co-infected patients.

A total of 531 HIV-infected patients in the Virus Activation Transfusion Study were tested for HCV and HBV infection. For HCV antibodies, the "Ortho version 3.0 EIA" test was used. For HCV RNA, the Chiron/GenProbe "TMA" (transcription-mediated amplification) test (lower limit of detection, 50 g-equivalents per milliliter) used by many US Blood Centers was employed.

The results were a false negative HCV antibody rate of 4% among HIV positive patients, meaning 4% of those with a negative HCV antibody test had a positive HCV RNA test. When considering all HCV RNA positive patients who were HIV positive, 12% had a negative HCV antibody test, meaning 12% of HIV patients were not classified as HCV-infected by antibody testing, when they truly were HCV positive, as determined by HCV RNA. The median CD4 count for all 531 tested patients was approximately 14 cells per microliter (range 3-76). In a personal conversation with Dr. Busch, he said that he did not yet have the average or range of CD4 counts for those patients with a false negative HCV antibody test.

In a separate analysis, Dr. Busch found that the standard test to screen for active hepatitis B infection ("surface antigen") among the HIV positive patients carried an 8% false negative rate. All of them did have a positive antibody test to hepatitis B "core." Chiron/GenProbe "TMA" testing was used to detect HBV DNA. Those HBV DNA positive, yet surface antigen negative patients represented 26% of all active HBV infections among their HIV positive transfusion patients.

In other studies, the reported "sensitivity" (ability to detect antibodies when present) of the Abbott HCV EIA 2.0 is 86%, while the "specificity" (ability to predict a 'negative' result, when antibodies are not present) is 99.8%. At last year's 10th International Symposium on Viral Hepatitis and Liver Disease in Atlanta, Dr. S. George reported that 6% of HIV/HCV co-infected patients from the Iowa City Veterans' Medical Center had a negative Abbott 2.0 antibody test (with positive HCV RNA test). However, Dr. George found no association between false negative antibody test results and CD4 counts.

A report published in last year's Journal of Clinical Microbiology by Dr. C.L. Thio indicated that the sensitivity of the Ortho HCV version 3.0 antibody test was 88% in HIV/HCV co-infected patients. That study included more than 150 patients with a CD4 count less than 200 cells per microliter.

Note that the transfusion patients in Dr. Busch's report might represent a different group and natural history of disease than other HIV/HCV co-infected patients. However, the median age in his study was only 37 years and therefore does not represent an elderly population with an "older" and weaker immune system.

Also to be considered in these discussions is the issue of "seroconverting" to HCV after initially infected ("window period"), HCV antibody loss that has been documented in HCV-mono-infected patients (possibly a part of the natural history in a minority of patients) and "seroreversion" (becoming antibody positive after being antibody undetectable that was documented after being positive prior to that, unrelated to HAART) that has been documented in both HCV mono-infected and HIV/HCV co-infected patients (Morishima reference).

Hepatitis A Vaccine Effective in HIV Patients When CD4 is Greater than 200 Cells

Acute hepatitis A can be lethal for HIV or HCV positive patients. A safe and effective vaccine is available that prevents hepatitis A after exposure. C. Kemper, MD of the California Collaborative Treatment Group has reported that the standard 2 injections of Havrix (hepatitis A vaccine at time zero and 6 months) leads to effective antibody levels in 68% of HIV positive patients ("as treated" analysis, including those who were able to be evaluated 3 months after the second injection, excluding drop-outs), when the CD4 count was greater than 200 cells per microliter. In contrast, only 9% of those with a CD4 count less than 200 cells achieved a similarly effective antibody level.

Anti-HIV medication was taken by 92% of the 133 enrolled patients (11% women, 18% non-Caucasian). The study was randomized, placebo-controlled, and double-blind (medication or placebo not known by physician or patient). The mean CD4 count was approximately 350 cells per microliter, with HIV RNA approximately 3.3 log (1,995 copies per milliliter). Neither CD4 counts nor HIV RNA changed significantly different when comparing the two study arms. However, each study arm achieved CD4 count increases (31 cells increase in Havrix and 52 cells per microliter count increase in placebo) and HIV RNA decreases (-0.18 log and -0.35 log copies per milliliter, respectively) during the 9 months of observation. The only side effect was mild soreness at the upper arm injection sites. The most common reason for not completing the study was "lost to follow-up" (13%).

Dr. Kemper concluded that, "There probably is little value in providing Havrix to HIV positive patients with CD4 counts [less than or equal] to 200 cells per microliter. The authors also commented, The study did not examine whether it is beneficial to delay vaccination in HIV positive patients with lower CD4 counts until their immune status improves with more potent antiretroviral regimens."

Hepatitis A is transmitted by fecal contamination of food or water or when fecal material is transmitted to the mouth during sexual contact. After recovery from hepatitis A, lifelong immunity generally is the ruleamong HIV negative persons.

Is HCV Transmitted Sexually Among Gay/Bisexual Men?

HCV (hepatitis C virus) is known to be transmitted by routes similar to those of HIV. However, most studies have shown that sexual transmission of HCV appears to be much less efficient than for HIV. Such studies include long-term monogamous "discordant" (one person HCV positive, partner negative), heterosexual partners (less than 5% transmission rate). Studies of "men who have sex with men" reveal that 10% or less are HCV positive, if there is no history of injection drug use. Now, researchers from Vancouver have concluded that sexual transmission of HCV occurs between gay/bisexual men and is associated with specific sexual practices. Note in the results below that "associated with" does not mean "causes." The lead author was Dr. K.J.P. Craib of the British Columbia Center for Excellence in HIV/AIDS.

A total of 662 gay/bisexual men participated in the Vancouver Lymphadenopathy-AIDS Study. Blood tests and self-administered questionnaires from 1982-1985 were used for the analysis. The HCV positivity rate was 5.9%. A significantly higher percentage of HIV positive men were HCV positive (9%), when compared to HIV negative men (3%). HCV positive men reported significantly greater number of male sexual partners and a significantly higher rate of using injection and non-injection, illicit recreational drugs.

However, 49% of the 39 HCV positive men "reported never injecting drugs." Among those men, the HCV positive ones were significantly more likely to report "insertive" oral-anal contact ("rimming") and "insertive fisting" (placing a hand into partner's rectum). In a "multivariate" analysis, the following factors were found to be significantly associated with HCV positivity: lifetime injection drug use ("odds ratio" [OR] 25), HIV infection (OR 2.5) and number of male sex partners in the last year (OR 2.2).

Whether the above sexual practices represent the true cause and mechanism of HCV transmission for non-injecting gay/bisexual men is not known. They might simply represent behaviors associated with the actual mechanism of transmission that might be more likely to occur with increasing numbers of sexual partners.

One could hypothesize that small scrapes on one's fingers or hands might represent a portal of entry if stretching of the partner's rectal opening led to a minute amount of bleeding that contained HCV (for transmission during 'insertive fisting'). One could also hypothesize that if small amounts of blood with HCV were present in stool, then 'rimming' might lead to either ingesting of infectious virus or if there were scrapes or lesions in the mouth (herpes or canker sores, gum disease), that could be a portal of entry for HCV directly across the mouth mucosa (moist lining).

A similar line of reasoning has been hypothesized for a history of intranasal cocaine "insufflation" ("snorting"), which has emerged as a risk factor for HCV infection. Is it the actual blood-tinged (with HCV) device that is shared among "snorters" that transmits HCV through the nasal mucosa (moist lining)? Or is the behavior of intranasal cocaine usage merely a marker for or associated with unsafe sexual practices associated with bleeding and scrapes that transmits HCV between partners?

Since the real answers to the above questions are unlikely to be determined easily, it would seem reasonable that safer sexual practices may help to limit HCV transmission by the sexual route between gay/bisexual men or even between heterosexual partners or lesbian partners. Such practices would include barrier protection and limiting "disinhibiting" recreational drug usage during sex that might otherwise lead to unsafe practices associated with even minute exchanges of blood potentially infected with HCV. These discussions are strikingly similar to those for sexual behaviors between men previously associated with having Kaposi's sarcoma, an AIDS opportunistic cancer, later found to be caused by human herpes virus type 8 ("KSHV").

Infection with HCV has been associated with "autoimmune" abnormalities (when the immune system makes antibodies to normal human proteins and antigens in the body) and other manifestations outside of the liver. However, whether HIV/HCV co-infection magnifies or mitigates these effects has not been studied. Now, R.P. Woitas, MD of the University of Bonn in Germany has reported the rates of several of these abnormalities in HIV/HCV co-infected (98 patients), HCV mono-infected (78 patients) and HIV control (45 patients) groups. The study was "cross-sectional" (one time "snap-shot"). Women more commonly have autoimmune conditions than men, yet women comprise a minority of Dr. Woitas' HCV and HIV/HCV patients

"Arthralgias" (joint aches), symptoms of "vasculitis" (abnormal inflammation of blood vessels) and "Raynaud's phenomenon" (abnormal excess constriction of blood vessels in fingers causing bluish color, tingling, pain after exposure to cold) were significantly more common among HIV/HCV co-infected patients (76%) and HCV mono-infected patients (56%) than among HIV mono-infected patients (none). In blood, "rheumatoid factor" (arthritis autoimmune antibody) and other antibodies to normal human proteins ("ANA" or "anti-nuclear" antibodies, anti-smooth muscle antibodies, others) were significantly more common in HCV mono-infected patients than among either HIV/HCV co-infected or HIV mono-infected patients ("uncommon" in latter 2 groups). However, "cryoglobulins" (abnormal "precipitation" or clustering of proteins from blood liquid, i.e., no longer in solution, after exposure to the cold) was present in 21% of both HIV/HCV co-infected and HCV mono-infected, compared to only 5% in HIV mono-infected patients.

"Synergy" (enhanced effects when 2 viruses combined in co-infection, compared to adding the individual effects of each virus) was noted for the rate of "thyroglobulin [thyroid gland protein] antibodies" and "cardiolipin" antibodies (increases the risk of abnormal clots). Thyroglobulin antibodies were present in 31% of HIV/HCV co-infected patients, compared to 9% in HCV mono-infected and 15% of HIV mono-infected patients. Somewhat similarly, cardiolipin antibodies were present in 46% of HIV/HCV co-infected patients, but in only 12% of HCV positive and 31% of HIV positive patients.

While these differing trends of autoimmune abnormalities would need to be confirmed in a greater number of patients followed prospectively, Dr. Woitas concluded, "HIV co-infection modulates the frequency of HCV-related extrahepatic manifestations differentially."