HIV treatment safe and effective in South African patients with hepatitis B co-infection, but co-infection frequent

South African patients who are co-infected with HIV and hepatitis B virus have a comparable response to HIV treatment to patients who only have HIV, according to a South African study published in the December 1^st edition of Clinical Infectious Diseases. For most of the HIV/hepatitis B co-infected patients, antiretroviral therapy did not cause liver toxicity, and there was no difference in the risk of death for patients with or without hepatitis B co-infection. However the study found a high frequency of hepatitis B/HIV coinfection in an urban cohort - almost one in five were co-infected.

Most of the world’s HIV infections are located in Africa, a region which also has a high prevalence of hepatitis B infection. Antiretroviral therapy is becoming increasingly available in Africa, but there is currently uncertainty about the impact of hepatitis B co-infection on responses to HIV treatment.

Investigators in South Africa therefore designed a study involving 537 HIV-positive individuals receiving antiretroviral therapy through work-place treatment programmes. A fifth of these individuals (106) had evidence of chronic hepatitis B infection (hepatitis B surface antigen-positive, [HBsAg-positive]). A high hepatitis B DNA viral load is considered to be 10,000 copies/ml or above and the investigators tests indicated that 46 patients were in this category and that a further 60 patients had hepatitis B DNA loads below 10,000 copies/ml.

All the patients in the study started antiretroviral treatment between 2001 and early 2006. They were followed until the end of November 2006 or for a maximum of 72 weeks.

The investigators divided the patients into three exclusive groups according to their hepatitis B infection status: those who were HBsAg-negative; patients who were HBsAg-positive with a low hepatitis B DNA load; and individuals who were HBsAg-positive with a high hepatitis B DNA load.

All the patients were antiretroviral-naïve at the start of the study and commenced HIV treatment with a combination that included efavirenz, 3TC (a drug with activity against both HIV and hepatitis B) and AZT.

Four outcomes were monitored by the investigators:

• The achievement and maintenance of a HIV viral load below 400 copies/ml.
• Increases in CD4 cell count.
• Liver toxicity indicated by increases in ALT/AST levels of five times or more the upper limit of normal.
• Death.

Viral load was measured during the study at weeks six, 24, 48 and 72 and at tall these points all three groups of patients were equally likely to achieve and maintain a viral load below 400 copies/ml. For example, at week 48, 84% of patients who only had HIV had a viral load below 400 copies/ml, as did 85% of the hepatitis B co-infected patients with low hepatitis B DNA-levels and 86% of the co-infected patients with high DNA levels.

Next the investigators looked at changes in CD4 cell count amongst the three groups of patients. During the first six weeks of antiretroviral treatment, the average weekly increase in CD4 cell count was 16 cells/mm³ each week and once again this did not differ between the three patient groups. The only factors that predicted CD4 cell count increase during early antiretroviral therapy were baseline CD4 cell count (p = 0.02), baseline haemoglobin (p = 0.05) and baseline HIV viral load (p

In the longer-term, CD4 cell count increased by an average of 1 cell/mm³ per week with no difference between the three groups of patients.

At baseline, patients with hepatitis B coinfection and a high hepatitis B DNA viral load had significantly higher liver-enzyme levels than the other groups of patients (median 56 iu/dl vs. 38 iu/dl for patients who only had HIV and 40 iu/dl for coinfected patients and a low hepatitis B DNA viral load, difference with this two patient groups, p = 0.004).

Although liver enzyme levels decreased in this group of patients after starting HIV therapy, at week 48 it was still significantly higher (median, 46 iu/dl) than in patients with a low hepatitis B DNA load at baseline (29 iu/dl) and HIV-negative patients (33 iu/dl, p = 0.004).

A total of 23 episodes of grade 3 or 4 liver toxicity were observed in 23 patients during the course of the study. The rate of these events was highest in patients with a high hepatitis B DNA load (17 per 100 person) years compared to a rate of 1.7 per 100 patient years in co-infected patients with a low hepatitis B DNA load and 3.8 per 100 person years in patients who only had HIV.

Most of these events occurred during the first six months of antiretroviral therapy, and the investigators believe that immune reconstitution inflammatory syndrome is the most likely explanation in the hepatitis B-infected patients. However, three events occurred after six months of HIV treatment, all involving patients with co-infection and the investigators suggest that the emergence of 3TC-resistant hepatitis B and the subsequent loss of viral control is the most likely cause.

There were a total of 24 deaths, 17 of which were non-traumatic. There was no difference in the risk of death between the three groups of patients.

“Hepatitis B co-infection did not affect early mortality or response to highly active antiretroviral therapy that included agents active against hepatitis B virus” write the investigators, adding, “even the group of patients with the highest pre-[antiretroviral] hepatitis B virus DNA levels achieved increases in their CD4 cell count and HIV [viral load] suppression at rates similar to those among patients who were HBsAg-negative. Furthermore, we demonstrated that the risk of [antiretroviral]-related hepatotoxicity can be stratified by hepatitis B DNA level”.

The investigators conclude, “chronic hepatitis B infection does not alter response to [antiretroviral therapy] or increase mortality during the first 18 months of therapy and, thus, should not deter clinicians from initiating [antiretroviral therapy].”

An accompanying editorial says there are five important messages from this study:

• There is a high prevalence of hepatitis B in South Africa and it is therefore essential to test HIV-positive patients for hepatitis B and vaccine patients who are negative.
• Viral load and CD4 cell count responses to antiretroviral therapy was not affected by hepatitis B co-infection. They note that treatment guidelines, which are largely based on poorer outcomes of antiretroviral therapy in patients co-infected with hepatitis C, recommend the early initiation of HIV therapy in patients with hepatitis co-infection.
• There was a low rate of liver toxicity.
• High hepatitis B DNA levels were associated with an increased risk of liver toxicity suggesting that regular monitoring of hepatitis B viral load would be of value.
• 3TC resistant-hepatitis B may be contributing to liver toxicity.

Research presented earlier this year at the HIV Implementers' Meeting in Kampala, also conducted in South Africa, among 502 patients about to start therapy in the Themba Lethu clinic cohort at Helen Joseph Hospital in Johannesburg, found that 5.6% were hepatitis B SAg-positive, but 10.6% had antibody to hepatitis B core, accompanied by detectable HBV DNA in liver or serum in 91% of cases, indicating that they had an "occult" or silent infection. In one case a patient who was hepatitis B surface antigen-negative had an HBV DNA viral load of 1 million copies/ml. No correlation between occult infection and CD4 count, gender or any other variable could be detected.

The study also found that of those patients who were hepatitis B surface antigen-positive, 57% had grade 1 elevated liver enzyme levels. Yet a chart audit in the clinic found that none of 22 HIV-positive patients with grade 1/2 elevated liver enzymes was referred for hepatitis serology testing, despite the fact that raised liver enzymes should be a cause for suspicion of hepatitis.

Cynthia Firnhaber of WITS Consortium at the University of Witswatersrand, who carried out the research, said that the prevalence of occult hepatitis B infection in the South Africa context was cause for concern because patients may receive inappropriate first-line therapy. In particular she highlighted the risk of giving what amounts to long-term hepatitis B monotherapy with 3TC if that drug is the only anti-HBV drug in an antiretroviral combination. The South African HIV Clinicians Society recommends that people with HIV and hepatitis B coinfection should receive tenofovir in addition to 3TC in first-line treatment, or tenofovir and FTC (a similar drug to 3TC), since both drugs suppress hepatitis B replication.