now.jpg (1640 bytes)

January 1998 Volume 1, Number 3-4, Section 8

Hydroxyurea: a new approach to treating HIV

The use of hydroxyurea (HU) has been receiving increased attention. This article updates new information and data which appeared in NATAP Reports July 1997 issue. Although, hydroxyurea in combination with ddI or ddI/d4T has been shown to significantly decrease viral load, there are some concerns about its effects. In some studies where viral load was appreciably reduced when using HU with ddI or ddI+d4T, the CD4 counts did not proportionately increase; although, in some other studies CD4 increases were larger. Individuals with higher CD4 prior to using HU therapy may be more likely to achieve a better CD4 increase. HU acts in a way that may prevent CD4s from increasing. The mechanism by which that may occur was described in the article on HU in our July issue. HU therapy also may cause bone marrow suppression and neutropenia, a reduction in white blood cell count, which appears to be reversible upon stopping of therapy. Following are several reports issued over the last several months, including data from additional studies.

In the August 30, 1997 issue of Lancet, Dr. Jorge Vila reported in a research letter about a study of hydroxyurea+ddI addressing the issue of induction/maintenance, whether or not the proviral DNA they detected is infectious, the combination’s effect on a variety of virus activity measures in plasma and lymph nodes, and the combination’s activity in resting cells as a key to its benefit. It discussed the use of hydroxyurea+ddI as a first line therapy for treatment naive individuals, as an eradication therapy (if eradication is possible; it has not yet been proven), or as a therapy that may be useful during primary infection. The recently reported findings by Dr. Robert Siliciano has created doubt that eradication is possible.

You may have read press coverage about these 2 study participants, whose plasma viral load was undetectable (>200 copies/ml) after one year of therapy with ddI+hydroxyurea. Both agreed to completely stop therapy and remained undetectable at 1 year of follow-up. These two patients started therapy at 3 and 12 months, respectively, after HIV infection. Seropositivity was confirmed by ELISA and western blot. Neither patient had symptoms after infection. However, there may be reasons not to have overly high expectations from these developments. The baseline viral loads for these 2 persons were low (676 and 1120 copies/ml) and their CD4s were normal prior to study treatment, they were recently just infected, and I have heard a number of other anecdotal reports of individuals stopping a potent protease therapy and remaining undetectable.

In this study conducted by Vila, 25 ddI-naive individuals without symptoms with CD4 >200 were treated for up to one year with 200 mg ddI twice daily, and 500 mg hydroxyurea twice daily. Baseline HIV RNA was 29,396 copies/ml. Eight individuals received lymph node biopsies. Investigators said treatment was well tolerated. There were no treatment interruptions due to side effects. There were a few minor transient symptoms such as digestive disorders, and an increase in mean corpuscular volume. After one year, leucopenia (reduction in WBCs) was seen in four patients (three grade 1, and one patient Grade II). No changes in platelets, hemoglobin, amylases, lipases, or LFTs were recorded.

Results. 13/24 and 10/20 individuals suppressed their plasma viral load to undetectable (<200 copies/ml) at 6 months and one year, respectively. Most of these undetectable individuals at one year also had no detectable infectious virus in their lymph node MNCs and CD4s after one year’s therapy. The average CD4 increased from 525 to 601 cells, although in other studies the CD4 increase can be slight because hydroxyurea inhibits cell replication. Of the 8 individuals who had their lymph node biopsied, 1/8 had a detectable plasma viral load (1017 copies/ml), undetectable virus in lymph node mononuclear cells (MNCs), but had detectable infectious virus in their lymph node CD4s; 5/8 had undetectable plasma viral load, and no infectious virus in both their lymph node MNCs and CD4s; 1/8 had undetectable plasma viral load, but had detectable infectious virus in both the lymph node MNCs and CD4s; data was not available for one person.

For the two individuals undetectable after stopping therapy, they both reached undetectable in their plasma HIV RNA by one year on therapy and remained undetectable one year after stopping therapy. In addition, extracellular RNA was undetectable after 2 years in both plasma and lymph nodes; and intracellular HIV RNA was below the level of detection (undetectable) in PBMC (peripheral blood mononuclear cells) and LNMC (lymph node mononuclear cells). Culture yielded no infectious virus from either LNMC or purified CD4 t-cells.

At both 1 and 2 years proviral DNA was non-quantifiable in both patients who remained undetectable for a year after stopping therapy. Qualitative evaluation of proviral DNA at 2 years in PBMC showed low levels for both patients but Vila said it "seems" to be non-infectious or incompetent for the release of infectious virions. Siliciano’s findings are that there is a small pool of proviral DNA that is replication competent and persists or will be long lasting. Some researchers have suggested that Vila just may have not found the replication competent proviral DNA.

Investigators said that "an explanation for the absence of viral rebound (for the 2 patients who stopped therapy) could be that this combination may exercise its anti-HIV activity in resting cells." But according to anecdotal reports other PWAs on treatment without hydroxyurea may have also remained undetectable after stopping therapy. Resting cells are an important reservoir for proviral DNA. It has been suggested that ddI, as well as 3TC and ddC, work in resting cells, while d4T and AZT work in activated cells. Hydroxyurea’s mechanism of action may, theoretically, increase the chances of ddI in terminating the successful growth of proviral DNA competent to produce infectious virus. Furthermore, the authors say these findings suggest the combination of ddI and hydroxyurea may be applied for potential eradication in early disease, although Siliciano’s findings might suggest otherwise.

Bristol Myers Squibb has started three trials exploring hydroxyurea in HIV treatment. A large trial comparing ddI+ d4T to hydroxyurea+ddI+d4T in 200 individuals with 200-500 CD4s started in December 1996. The second trial of 20 individuals will explore a 4-drug regimen of nelfinavir+ddI+d4T+hydroxyurea in individuals with >500 CD4. The third trial will explore the potential for eradication in 20 individuals during primary infection with the same 4-drug regimen:nelfinavir+ddI+d4T+hydroxyurea.

DDI Resistance. The authors stated that cellular proteins are less prone to mutations than viral proteins and that resistance of tumor cells to HU has not been reported after 35 years of clinical experience.

Investigators in the study found that virus with genotypic resistance to ddI were suppressed when HU was used with ddI. If resistance to HU does not develop and the development of genotypic resistance to an accompanying drug in a regimen does not effect the combination’s efficacy, then it may not be necessary to suppress viral load to below detection to prevent the development of resistance from causing therapy failure.

Hydroxyurea + d4T/ddI vs d4T/ddI.

In Hamburg, OT Ruschmann, Bernard Hirschel and others from the Swiss HIV Cohort Study reported findings from this 24 week study comparing the triple regimen of hydroxyurea (HU) and d4T/ddI to the double nucleoside regimen of d4T+ddI. The purpose as defined by the investigators of the study was to determine the short term effects of the HU triple regimen.

144 participants were d4T naive, ddI naive or with <6 months prior experience. They were randomized to HU+d4T/ddI or d4T/ddI. The dosing regimens were: HU- 500 mg bid, ddI - 200 mg bid, d4T- 40 mg bid. After 12 weeks, those receiving ddI/d4T were given the option of adding HU.

72 individuals were randomized to each group. Prior to reaching 12 weeks 7 discontinued from the HU group (1 due to an adverse event); and 3 discontinued from the d4T/ddI group (none due to adverse event). At week 12, there were 65 evaluable participants remaining in the HU arm and 69 in the placebo arm. See Table 22 and 23

Table 22. Hydroxurea - d4T/ddI: Results - Adverse Events



HU + d4T/ddI













Neuropathy grade 1



grade 2



grade 3



Neutropenia grade 1



grade 2



grade 3



Elevated lipase



Elevated ALT/AST (LFTs)



Table 23. Week 12 Changes from Baseline

  d4T/ddI HU+ d4t/ddI
CD4 +107 +28
CD4% +2.5 +3.0
CD8 +5 -124
Total lymphocytes +196 -124
HIV RNA (200 copy test) -1.5 log -1.9 log
HIV RNA (20 copy test) -1.7 log -2.3 log
%<200 copies/ml 29% (20/69)* 60% (39/65)*
%<20 copies/ml 8% (6/72)* 19% (14/72)*

* These percentages are based on the number of evaluable participants. It is not an intent-to-treat analysis which would base the percentage on the 72 patients that started in each arm. But, the %<20 copies/ml in based upon the entire 72 patient groups that started the study. In the HU arm 26 had >200 copies/ml and in the placebo arm 49 had > 200 copies/ml. These individuals had viral loads ranging from 201 copies/ml on up.

Investigators reported 27 participants withdrew at week 12 due to nausea (8), neuropathies (4), depression (3), lost to follow-up (8), and patient choice (4). Adverse events were more frequent in the HU arm (p<0.05).

The mean baseline CD4 and viral load were 367 cells and 4.53 log (about 33,900 copies/ml). The mean baseline CD8 was 1017, the total lymphocytes were 1774.

After 24 weeks. At week 24, there were 19 evaluable patients still in the d4T/ddI arm, 24 in the group that added HU to d4T/ddI at week 12, and 34 who started and were continuing the triple regimen of HU+d4T/ddI (the dropouts were mostly individuals who did not reach <200 copies/ml, 34/39 who reached <200 copies/ml stayed on HU triple therapy.)

Those who had a limited reduction in viral load from d4T/ddI were characterized as poor responders, and added HU at week 12. They were able to produce a nice decrease in viral load approaching those observed in patients initially randomized to the HU arm. But only 55% of these individuals who added HU had <200 copies/ml at week 24.

Generally, those initially randomized to the HU arm were able to sustain their viral load reductions at week 24. 84% who started the HU regimen and stayed with it remained <200 copies/ml at week 24.

Generally, the lower a person’s viral load was at baseline, the more likely they were to reach undetectable. From a graph shown by the investigators it appeared as though most of those individuals with between 1,000 and 10,000 copies/ml at baseline were able to reach <20 copies/ml. However, those with >100,000 copies/ml at baseline were less likely to reach <200 copies/ml (and more unlikely to reach the 20 copy test) than those with <100,000 copies/ml viral load at baseline. Commentary - This may be a factor to consider when deciding when to start antiretroviral therapy. If you accept the idea that lowering viral load to undetectable or to as low as possible should be the goal of therapy (when possible), beginning therapy when one’s viral load is no more than about 10,000 copies/ml may be a superior approach because it could be easier to reach <20 copies/ml and possibly easier to retain that reduction over the longer term.

Safety and Antiviral Activity of Hydroxyurea (HU) with ddI. 80 individuals without prior ddI, ddC and 3TC experience were enrolled in this multi-center, randomized, open-label AMFAR study in which participants were randomized to receive ddI alone for the first 12 weeks or ddI+HU. Individuals randomized to ddI alone were permitted to add HU after week 12. The delayed HU group received ddI alone for 12 weeks and the immediate HU group received HU+ddI from the start of the study. Participants were stratified by baseline CD4 (50-300, 301-600) to detect if there were differences in CD4 responses to HU therapy based on baseline CD4, since prior HU studies have shown that despite viral load reductions CD4 increases may not be proportionate and sometimes there were only slight increases in CD4 counts.

The HU dose was 500 mg bid (twice daily); the ddI dose was adjusted for body weight greater than or equal to 60 kg, 2 100 mg tablets bid; less than 60kg 125 mg bid (a 100 mg tablet and a 25 mg tablet). Plasma HIV RNA was evaluated by Chiron’s bDNA assay. Exclusion criteria included: absolute neutrophil count less than 1500 cells/mm cubed; SGOT, SGPT, alkaline phospatase, or total billirubin less than 2.5 times upper limit of normal; evidence of pancreatiitis (assessed by serum amaylase, pancreatitis amylase isoenzyme, and/or serum lipase); evidence of peripheral neuropathy of grade 2 (moderate) or higher.

Safety. Discontinued prior to week 12: 2/38 in the immediate HU group; 8/42 in the delayed HU group (the dropouts in the delayed group were primarily due to suboptimal plasma HIV RNA suppression and persisting gastrointestinal adverse events.)

Discontinuations weeks 13-24: 14 individuals in the immediate group dropped out primarily characterized as due to physician/participant decision; 6 discontinued during weeks 13-24 in the delayed group also primarily due to physician/participant decision.

A total of 12 serious adverse events were reported but only 3 were determined as possibly study drug related.

Thrombocytopenia was reported in 1% of patients receiving ddI in phase I studies. Thrombocytopenia (less than50,000/uL) was reported in 1-4% of patients receiving recommended or high dose ddI in ACTG 116B/117 and 116A. Thrombocytopenia has been reported as associated with HU and is usually preceded by leukopenia. See Table 24.

Table 24. Hydroxyurea - ddI:
Week 12 HIV RNA and CD4 changes from Baseline by CD4 Strata

Immediate HU

Delayed HU

CD4 strata










median base VL





median wk 12 VL





mean wk 12 log decrease

-1.1 log

-1.2 log

-0.6 log

-1.1 log

wk 12 less than500 copies/ml

6 (27%)

7 (58%)

2 (11%)

8 (53%)

CD4 (n=68)





median base CD4





mediam wk 12 CD4





median wk 12 CD4 change





Week 24. After week 12 HU was added to ddl monotherapy and resulted in a mean 0.4 log reduction in HIV RNA beyond that seen with ddl alone (n=33). At week 24 the HIV RNA reduction for the delayed HUgroup was similar to that seen at week 24 for the immediate HU group (n=30).

Protease Inhibitor + HU and ddI

Franco Lori, Heiko Jessen and others reported in Hamburg, results from a group of 17 individuals receiving indinavir or nelfinavir plus HU and ddI for an average of 8 months (range 1-12 months). Five were treated before sero-conversion; 5 within one year after sero-conversion; and 7 more than 1 year after sero-conversion. The investigators mentioned an appeal for this combination was that HU and ddI inhibit HIV in macrophages (resting cells), while protease inhibitors are effective in stimulated cells.

The average viral load and CD4 before treatment was reported to be 698.000 copies/ml and 450 cells. Investigators reported 17/17 achieved undetectable plasma viral load; semen viral load undetectable in 6/6; CD4 increases were 159 cells; treatments were reported as well as tolerated, 7/8 became undetectable for HIV RNA in lymph nodes. However, as I mentioned above, a question that remains, do they have latently infected integrated proviral DNA?

Investigators reported that for individuals treated before sero-conversion Western Blot remained negative during treatment. One of these patients discontinued treatment while their plasma viral load was undetectable. Seven months later they were still undetectable and their Western Blot was still negative. As mentioned earlier, there have been other anecdotal reports of remaining undetectable after stopping treatment with HAART. Several immune system tests suggested the patients immune system may have improved. Partial immune system restoration has also been observed for some individuals following successful treatment with HAART (potent protease inhibitor+two nucleosides) which doesn’t include HU.

back to top