Durban World AIDS Conference
July 9-14, Durban, South Africa

REPORT 32

Mike Norton, PA, Greenwich House, NYC:

1. Severity of Liver in Biopsies in HCV/HIV Coinfected
2. Developing AZT Mutations After d4T Initial Treatment
3. "Drug Holidays"

Severity of Liver in Biopsies in HCV/HIV Coinfected

J. Kostman and colleagues added to what is now a significant body of knowledge on HCV infection in the HIV co-infected.  This work comes from the AmFAR study of IFN TIW vs. IFN + RBV TIW.  In this abstract they gave the results of 99 liver biopsies performed prior to study entry.  They found 43 patients had moderate fibrosis, 26 had severe fibrosis, 16 had cirrhosis and 57 were found to have some degree of necrosis.  As in other studies, HCV RNA, HIV RNA, CD4+ and ALT were not predictive of finding fibrosis or cirrhosis on liver biopsy.  Abstract # WePeA4028

Commentary:  While not compared to HIV negative, HCV positive controls, this description of the amount of severe liver disease is striking and adds to a body of evidence calling for a prompt addressing and work-up, most importantly liver biopsy, of HCV in HIV+ patients.

Here is another RTV/IDV study from the group that reported one already at Durban.  All the RTV/IDV studies will be gathered together in one report. The German group this time led by A. Wickesberg and colleagues tried IDV 800 bid and RTV 100 bid with 2 or 3 nucs as background HAART.  Once again this study is in HIV antiviral naÔve patients.  In this study 48 patients were enrolled.  Baseline viral load was 320,000 or 5.5 log copies/ml.  Median baseline cd4+ was 50, which is lower then the group reported on yesterday, who were also naÔve, but went on RTV 400 and IDV 400 with impressive results.  At 24 weeks, in this 800/100mg study, 25% of patients had left the study.  There were 8 cases of kidney stones as opposed to 0 cases of kidney stones reported in yesterdayís larger and longer cohort.  There was a significant decrease in viral load of 1.84 log copies/ml at this 24-week juncture.  T-cells increased over 220 to a median of 270.5 cells/mm3.  For those who remained on therapy at 24 weeks, 95.8% had < 400 copies/ml HIV RNA and 75% had <50 copies/ml.

Abstract #WePeB4130

Commentary:  If one compares this study with others attempting to sort out the optimal dose of RTV and IDV to use when combining these agents, it seems increasingly clearer that a lower dose of IDV, to 400mg bid, reduces the incidence of kidney stones seen with IDV significantly.  Reducing IDV to 400mg bid may even eliminate any increased risk of kidney stones.  The larger doses of RTV also probably add to enhanced virologic potency of the regimen allowing greater numbers of individuals to reach undetectability.  However, the tolerability of RTV 400mg bid is less than at 200mg bid and with more RTV one sees a greater increase in lipids.  If the potential added potency of 800/100 or 800/200 is desired, one can change dose to 400/400 if kidney related side effects emerge.

Developing AZT Mutations After d4T Initial Treatment

In a report from C. de Mendoza, from Spain, more information related to the development of D4T resistance.  22 antiretroviral naÔve patients, who started a D4T containing HAART regimen and reached an undetectable viral load but were now rebounding were genotypically characterized.  This genotype was drawn at the first sign of rebound and here are the results:  Amino acid substitutions (mutations) potentially associated to drug resistance were recognized in 10 (45%). 2 patients developed classical AZT related mutations at 215 and 41 without ever seeing AZT.  Another 8 patients (all being exposed to 3TC0 had the 3TC associated M184V mutation, and in 2 of them it was associated with the changes L214F and R211K, however this virus was found to be phenotypically sensitive to AZT. No multi-nucleoside resistance mutations (Q151M, T69SSS) were seen.  Abstract # WePeA4077. The authors concluded 2/22 (9%) developed classic AZT related mutations at 215 and 41.

Commentary:  This is yet another abstract contributing to the understanding of D4T resistance.  It confirms what others have found, that D4T can lead to AZT mutations.  And in general that the nucleoside category has broader cross class resistance then was appreciated when these agents were first introduced.  One of the follow-ups to this study that would be helpful is to track what regimen patients went onto next and whether they responded to AZT containing regimens.  This is an important issue because the nucleoside prescribing patterns are most often AZT leading to D4T or visa versa.  And this practice may become less common as we understand more how to predict resistance to D4T and what this means for future AZT efficacy.

Similar results were obtained by Nicastri and coworkers, reported at Durban.  They studied 56 patients who received ZDV/lamivudine (3TC) or d4T/3TC as initial antiretroviral therapy. Mutations associated with ZDV resistance (principally T215Y and M41L) were found in samples from 13 (44%) of 29 patients treated with ZDV/3TC and 11 (40% ) of 27 patients treated with d4T/3TC after 18 months of therapy. These results reinforce data presented at the 4th International Workshop on HIV-1 Drug Resistance and Treatment Strategies and strongly suggest that ZDV and d4T select a shared set of drug resistance mutations.

"Drug Holidays"

H. Jeager and colleagues from Germany presented this Drug Holiday observational study.  This study can be a bit confusing because of the number of groups that were looked at.  Basically there are two groups and each group also has a matched control that didnít take a drug holiday.  The first group who took a drug holiday was a group of patients who were below the level of detection before they took a drug holiday.  The length of all drug holidays is between 4 to 6 weeks.  During this first groups drug holiday they increased their viral load an average of 2.6 log copies/ml.  They lost an average of 64 T-cells during the drug holiday.  And 58% were able to return to below the level of detection once they restarted therapy.  6 months after they had taken their drug holiday there was still net loss of 11 cd4+ís.  This group was compared to an equally matched group who didnít take a drug holiday and who were below the level of detection at the start of the observational period.  They continued to gain cd4ís and at the end of 6 months they had added another 69 cd4ís on average.  This group did see an increase in viral load over the 6 months, as undoubtedly some patients were not perfectly adherent or developed resistance and thus lost viral control.  The second set of groups were individuals who did or didnít take drug holidays but who did not start at below the level of detection.  The group who took a drug holiday lost 71 cd4ís after the drug holiday and their viral load increased 1.2 log copies.  No data was given on their viral loads 6 months after re-starting therapy.  However they had a net loss of 20 cd4ís 6 months after their holiday.  When compared to matched controls, who also were detectable but didnít take a holiday, the controls lost on average 36 cd4ís over the observational period.  The only other item of note from this study was that those who took a drug holiday saw a significant decrease in lipids during the drug holiday.  Abstract # WePeB 4207

Commentary:  If one had to pick a headline for track B during this conference, drug holidays or STIís would be it.  This presentation did not rate an oral presentation by the organizing committee.  Still it does add to the growing body of data in this area.  Questions I would need to ask this group are: did the individuals who started below the level of detection and took the drug holiday re-start their same regimen when they went back on therapy?  Were any of the individuals who didnít re-suppress to undetectable genotyped?  In both groups who took drug holidays the authors noted a statistically significant decrease in lipids during the drug holiday.  When they re-started therapy did they retain lower lipids or by the end of 6 months had they gone back to baseline/elevated lipid profiles?  What was the baseline median viral loads in the groups who were above the level of detection prior to the holiday or 6 months of observation?  Could this explain why no difference between the drug holiday takers and non- holiday takers were observed?  Or is it that once virus is detectable, drug holidays are less of a risk?

(comments from Jules Levin: Although resistance has not yet been generally seen in individuals who stop and restart therapy in clinical observational studies, there have been a few cases of individuals who developed resistance and had a more difficult time reducing their viral back down to undetectable after restarting therapy. If one's viral load increases to high level or to a level above their previous level, and this has been observed, getting VL baxck down to undetectable may not be achieved. It remains uncertain if resistance will be a problem in stopping and restarting therapy. Several researchers spoke out in the oral session STIs on Thursday in Durban saying they felt the concern about resistance developing was still unresolved. In particular, nevirapine and efavirenz have distinctly longer half-lives than protease inhibitors and NRTIs. In other words, when a person stops a NNRTI regimen, the NNRTI will remain in the blood for several days at decreasing levels. There is concern that this could cause resistance to NNRTIs. As well, although it is generally agreed that PI & NRTI blood levels may not be an issue if a person stops therapy, it remains uncertain if NRTI & PI drug concentrations in cells may not be a concern regarding leading to resistance. Several studies are ongoing and will explore these questions. Some researchers have too easily said, in my opinion, that resistance is not an issue but I think that until additional information on this question is known, the development of drug resistance is a risk.