Once Daily ddI as Effective as Twice DailyThe complexity of many HIV drug regimens and related side-effects are at least partially responsible for non-compliance on the part of some patients. Some of this complexity stems from drug interactions; for example, the buffering agent in didanosine (ddI) interferes with the absorption of certain drugs, such as indinavir, so that the two agents have to be taken at least an hour apart. This means that patients on indinavir and ddI have to dose themselves at least 5 times a day (thrice for indinavir, and twice for ddI). However, a recent study (Mobley et al., AIDS (13)11: pages F87-93) shows that one daily dose of didanosine (ddI) is just as effective at reducing plasma HIV RNA and increasing CD4 counts as the currently approved twice daily dose, thereby making it easier for more patients to fit into their treatment regimens.
The double-blind, multicenter trial randomized 87 antiretroviral-naive adults to once or twice daily ddI both in combination with d4T given twice daily. At baseline, participants had viral loads >10,000 copies/ml and CD4 cell counts >100. The daily dose of each drug was adjusted depending upon weight (those weighing 60 kg or more received ddI 400 mg/d and d4T 80 mg/d, those who weighed less were given ddI 250 mg/d and d4T 60 mg/d).
At the beginning of the study the treatment group and control group were similar with respect to demographic characteristics, virological and immunological parameters, and baseline CD4 counts; the baseline HIV RNA level was slightly higher in the control group. Nine of 44 (20%) in the control group discontinued therapy by week 12, as compared to seven of 43 (16%) in the treatment group.
At week 12, median HIV RNA changes were -1.83 log copies/ml in the treatment group and -1.80 log in the control group; 18 (41%) and 17 (40%), respectively, had levels below 400 copies/ml. The results at week 24 were similar. At week 12, those in the treatment group had a change from baseline in CD4 cells of 146.4 cells/mm3 while those in the control group had a change of 135.3. At week 24, the changes from baseline were 90.9 and 136.6, respectively. None of the differences in CD4 counts were significant. Subgroup analysis (HIV RNA level of <30,000 and >30,000) yielded similar results. Adverse events and laboratory abnormalities were mild to moderate and occurred with similar frequencies in the groups.
The authors conclude that the currently approved twice-a-day ddI dosing schedule shows no identifiable advantage over once-a-day ddI dosing, when combined with d4T twice-a-day. Moreover, the once-a-day ddI schedule may be better tolerated.
Further Doubts about the Ability of Antiviral Techniques to Eradicate HIVDespite early hopes of eradicating HIV from an infected person, it was discovered that even people with long-term suppressed HIV replication had latent virus that was capable of replication in resting memory CD4 lymphocytes. Moreover, the latent reservoir of HIV gets a foothold early in infection and persists despite two years of apparently effective antiretroviral therapy that begins during acute infection. Nevertheless, although it might take longer than originally hoped, if therapy has halted replication, that reservoir should gradually become depleted, and viral eradication is only a matter of time. The time to eradication would be a function of the life span of the latently infected cells and the size of the reservoir. If however, viral replication persists despite therapy, albeit at a low level, this reservoir may become replenished, and eradication would be much more difficult to achieve. Two recently published studies have sought determine which scenario is more likely.
Zhang et al., (New England Journal of Medicine, May 27, 1999; pages 1605-13) set out to determine whether drug therapy completely stops viral replication in order to assess the true size of the latent reservoir. The study involved eight men who had complete suppression of HIV in plasma and who began HAART within 90 days of acute infection. The base line mean CD4 count was 533 cell/mm3 and the average viral load (PCR assay) was 174,000 copies/ml. After treatment began, all patients quickly reached a viral load below 50 copies/ml, which was sustained for 20-35 months of therapy. No patient had detectable viremia in the plasma after month 5, and there was an average increase in CD4 count of about 300 cells/mm3 for the last four measurements.
Samples of peripheral-blood mononuclear cells were taken from each patient at baseline and at three successive timepoints. These were examined for changes in DNA sequences that might evolve over time, a clear indication of ongoing HIV replication. Natural history studies have demonstrated that shortly after infection, a person's viral population is relatively homogeneous but over time, with ongoing replication and viral evolution, diverse strains emerge. Conversely, if therapy brings replication to a complete halt (presumably sometime after becoming undetectable -- although how long afterward is unclear), there should be no new mutations in the viral DNA.
As would be expected, at baseline each sample contained very homogenous virus. In subsequent samples, three patients showed no significant increases in sequence variation. In three other patients, there was some divergence, but this was the result of hypermutations (multiple mutations introduced during a single replication cycle), and therefore not evidence of gradual sequence evolution. In other words, there was no genetic evidence of continued HIV replication in these three patients. However, two patients did show significant sequential divergence, exclusive of hypermutations, which indicated continued viral replication. The greatest variation was found by the second sampling, indicating much of the replication happened early on in treatment. The divergence in these two patients was not due to drug-resistant virus, suggesting that the virus might be replicating in compartments, either a specific anatomical location or diffusely scattered cell populations, that are not affected by the antiretroviral drugs.
In six of the eight patients, then, the viral sequence did not evolve during treatment. While one would hope that HIV replication was completely suppressed, it is impossible to draw such a conclusion based upon such limited number of samples. That the other two patients did show evidence of ongoing residual replication although at very low levels, casts doubt on such optimism. Furthermore, it should be noted that these eight subjects represented the best case scenario; they had sustained suppression of plasma greater than observed in most patients in a clinical setting, so most patients would probably show greater residual viral activity.
However, the authors concluded that the extent of the persistence of a latent reservoir of virus has perhaps been overestimated. Nevertheless, HAART does not always appear to be active enough, and there should be a focus on intensifying treatment despite the fact that current regimens are toxic, expensive, and complex. They estimated that it would take seven to ten years of continuous, truly effective antiretroviral therapy to eliminate the reservoir.
A related study led to even less promising results (Furtado et al., New England Journal of Medicine, May 27, 1999; pages 1614-22). This study involved five men who were selected on the basis of compliance with potent antiretroviral therapy and suppression of HIV in the plasma to undetectable levels (RT-PCR assay). Four received two of four nucleoside analogues (zidovudine, stavudine, didanosine, and lamivudine) and a protease inhibitor (ritonavir, indinavir, or saquinavir), while the fifth received an NNRTI (nevirapine) and a protease inhibitor for 20 to 31 months. Four had HIV for at least ten years prior to the study, and one had HIV for two years prior. Only one patient was treatment-naive. The baseline mean level of plasma HIV was just over 64,000 copies/ml, and it decreased to less than 50 copies/ml, with an average half-life of 30 days.
Although all five patients had undetectable levels of plasma HIV RNA for 20 months or more, all also had evidence of viral replication in peripheral-blood mononuclear cells, which indicates the persistent presence of reservoirs of HIV. During treatment, a two-phase decrease was noted in the levels of integrated and unintegrated proviral DNA and unspliced viral mRNA. The first phase, which lasted through the first 500 days of treatment, showed substantial decreases in cell-associated HIV DNA and unspliced viral mRNA (approximately 1.5 log). However, in the second phase, an additional 300 days or more, the values remained at a quasi-steady state. Since the patients showed no new viral resistance during treatment (some had resistance due to previously unsuccessful monotherapy), the findings were consistent with both (i) occasional stimulation by antigens that allow persistently infected cells with replication-competent HIV DNA to make progeny virus transiently and then die, thereby replenishing the pool of HIV-infected CD4 cells, and (ii) the production of infectious virus in cells or parts of the body where medications do not reach.
The authors conclude that a previous mathematical model showing that the HIV virus could be eliminated from long-lived infected CD4 cells after two to three years of a completely inhibitory antiretroviral regimen is not entirely accurate. The reason is that the model did not take into account the persistence of cells carrying replication-competent integrated forms of proviral DNA; nor did it address the effects of incomplete suppression of viral replication. There are not, they continued, sufficient data to allow an estimate of the length of therapy required to eradicate HIV infection and, unless the residual replication ceases or can be stopped, infection may never be eradicated.
The prospective, randomized, open-label study enrolled 108 patients with viral loads of >10,000 copies/ml despite at least six months of treatment with nucleoside analogues and at least three months with a protease inhibitor.
Patients were randomly assigned to the control group (n=43) and treatment group (n=65). The drug regimen of those in the control group was determined according to published guidelines on current optimal care. Physicians were not given the results of their genotypic results. For the treatment group, therapy was not changed until the physician received the genotypic results. Clinical status and adverse effects were assessed at each visit, and viral load, CD4 counts, and genotypic-resistance assays were measured every three months for patients in both groups. If, after three months on a new regimen, HIV RNA remained above 10,000 copies/mL or less than .5 log lower than baseline, treatment was modified. The primary endpoint was the variation of HIV RNA from baseline to month three and month six. The study ended after six months, rather than the expected twelve, because significant virological advantage was shown in the treatment group at that time.
Both groups had similar baseline characteristics, including prior treatment. The overall prevalence of primary mutations for the reverse transcriptase gene was 90%; all patients had one or more secondary mutations. All patients also had at least one secondary mutation in the protease gene (mean of 6.2 mutations/patient). In addition, all patients were treatment-experienced (mean of 3.9 nucleoside analogues for mean duration of 39.5 months, and a mean of 1.8 protease inhibitors for a mean duration of 11.6 months). At entry, the most common drug combinations were (i) AZT/3TC, and indinavir, (ii) d4T/3TC and indinavir, and (iii) d4T/3TC and saquinavir.
The mean change in HIV RNA at month three was -1.04 log in the treatment group and -0.46 log in the control group; HIV RNA was undetectable (i.e., <200 copies/mL) in 19 (29%) of 65 patients in the treatment group and 6 (14%) of 43 in the control group. At month six, the change was -1.15 log for the treatment group and -0.67 for the control group; HIV RNA was undetectable for 21 (32%) of 65 in the treatment group and 6 (14%) of 43 in the control group. The result of combining the viral loads at months three and six was significant (p=0.015). The mean change in CD4 count did not differ significantly between the two groups.
After the study began, preferred regimens for those in the control group were nelfinavir plus either (i) d4T/ddI or (ii) d4T/3TC (30%); and ritonavir/saquinavir plus either (i) d4T/3TC or (ii) d4T/ddI (37%). The regimens for the treatment group were more diverse, involving more combinations of reverse-transcriptase inhibitors and protease inhibitors, resulting in more individualized treatment patterns.
Published guidelines, which suggest changing to at least two new drugs, were not followed in 19% of the control group and 27% of the treatment group. In the former case, this was due to a lack of another option, whereas in the treatment group it was due to HIV RNA sensitivity to one or two drugs in an earlier regimen. When a change was made to ritonavir and saquinavir combinations, a greater decrease in viral load was found in the treatment group.
The authors concluded that the study shows promise for the use of genotypic-resistance assays. However, while genotypic technology is applicable to clinical practice, more studies are required to find optimum indications for genotyping in the general population of HIV-positive persons. Long-term effects of genotypic strategies also needs to be studied. Furthermore, the assays may suggest the best first switch regimen, but the value of resistance testing may decrease with time as a patient becomes more resistant to available therapies.
In a commentary on this study, however, some other questions are raised (Falloon, Lancet, June 26, 1999; 352, pages 2173-4). The primary problem that Dr. Falloon identified is interpreting the results of genotypic resistant assays. While she says that computer analysis would be helpful, a physician would still have to make judgments based on the patient's drug history, viral-load responses, drug and regimen preferences, and past and expected toxicity and tolerance. In the study above, the regimens were not determined by an algorithm taking such factors into account but rather by a table that indicated which drugs should be used when certain mutations were observed. Yet, if strictly followed, this table would have led to some misrecommendations about new drugs. Nevertheless, Falloon sees this as an important study and a first step toward determining the appropriate clinical application of genotyping.
Many experts feel that both genotypic and phenotypic assays provide information that is useful for making the best treatment decisions; however their relative cost-effectiveness versus merely relying on treatment history remains to be determined.
Improved Tolerability and Adherence with Reduced Mass Didanosine Tablets in HIV-1 Infected Adults Receiving Highly Active Antiretroviral Therapy
Session 84, Abstract 661: A Randomized Double-Blind Study of d4T + ddI versus ZDV + ddI as Initial Treatment in Subjects with CD4 Count <= 500 cells/mm3
This article was provided by Gay Men's Health Crisis. It is a part of the publication GMHC Treatment Issues. Visit GMHC's website to find out more about their activities, publications and services.