STIs and Effect on Lipids, HIV Drug Resistance, and Viral Load
"Long-Cycle Structured Intermittent versus Continuous Highly Active
Antiretroviral Therapy for the Treatment of Chronic Infection with Human
Immunodeficiency Virus: Effects on Drug Toxicity and on Immunologic and
Journal of Infectious Diseases, August 1, 2003, Volume 188, Number 3
Mark Dybul,1 Elizabeth Nies-Kraske,1 Marybeth Daucher,1 Kurt Hertogs,3 Claire W. Hallahan,1 Gyorgy Csako,2 Christian Yoder,1 Linda Ehler,1 Peter A. Sklar,1 Michael Belson,1 Bertha Hidalgo,1 Julia A. Metcalf,1 Richard T. Davey,1 Diane M. Rock Kress,1 April Powers,1 and Anthony S. Fauci1
1Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, and 2Department of Laboratory Medicine, W. G. Magnuson Clinical Center, National Institutes of Health, Bethesda, Maryland; 3Tibotec-Virco, Mechelen, Belgium
Summary: We evaluated the effect of long-cycle structured intermittent therapy (SIT; 4 weeks without highly active antiretroviral therapy [HAART] followed by 8 weeks with HAART) versus continuous HAART. The study was prematurely terminated to new enrollment because of the emergence of genetic mutations associated with resistance to antiretroviral drugs in 5 patients.
After 48 weeks, there was no significant difference between groups in lipid, hepatic transaminase, and C-reactive protein levels in 41 patients.
Although there were no differences in CD4+ or CD8+ T cell counts or the percentage of cells that were CD4+CD25+, CD8+CD25+, or CD4+DR+, patients who received SIT had a significantly higher percentage of CD8+CD38+ and CD8+DR+ cells.
There was no clear autoimmunization effect by immunologic or virologic parameters.
There was no benefit to long-cycle SIT versus continuous HAART with regard to certain toxicity, immunologic, or virologic parameters.
It has been suggested that repeated "autoimmunization" cycles could ultimately lead to control of plasma viremia during extended discontinuation of HAART in certain individuals who were initially treated during acute infection. However, in individuals who began antiretroviral therapy during the chronic stage of infection, the results have been inconsistent, and a significant reduction in HIV plasma viremia from the first to subsequent cycles without HAART occurred in a much smaller proportion of patients than that observed with treatment during acute infection. Our approach has been to evaluate the possibility that structured intermittent therapy (SIT) with predetermined cyclic periods of
time with and without HAART could maintain suppression of HIV and preserve levels of CD4+T cell counts while reducing certain markers of toxicity in individuals with chronic HIV infection.
HIV-infected patients with a CD4 count >300 cells who were receiving at least
a 3-drug HAART regimen with HIV viral load levels <500 copies/ml for >6
months and <50 copies/ml during screening were eligible for enrollment. Patients
could not have a treatment history consistent with drug resistance; however,
individuals who had received antiretroviral drugs before initiating HAART were
eligible for inclusion in the study.
The study was designed for 90 patients randomly assigned to continue
HAART or to receive 7 cycles of 4 weeks without HAART followed by 8 weeks with HAART. Patients could not receive abacavir or nevirapine during the study as a result of potential problems with cycling with and without these drugs.
The primary end point of the study was the proportion of patients in each arm who had plasma HIV RNA levels <50 copies/mL. Secondary end points included CD4+ T cell counts, levels of serum lipids and liver enzymes, and immunologic
Patients underwent laboratory evaluations at least every 4 weeks, including plasma HIV RNA determined by branched-chain DNA (limit of detection, 50 copies/mL) and lymphocyte subsets, including analysis of lymphocyte
activation, by standard flow cytometric analysis. Patients had the option to have more-frequent plasma HIV RNA assays performed during without-HAART periods.
Fifty-two patients were enrolled in the study; 26 were randomly assigned to receive long-cycle SIT of 4 weeks without HAART followed by 8 weeks with HAART, and 26 were randomly assigned to receive continuous HAART. Although it was originally planned to enroll 90 patients, the study was prematurely terminated to new enrollment after 52 patients had enrolled.
New enrollment was terminated as a result of evidence for newly emergent resistance in 3 of 8 patients in the intermittent arm receiving an efavirenz-based regimen. Two patients had both K103N, which is associated with resistance to efavirenz, and M184V, which is associated with resistance to lamivudine, genetic mutations during the fourth or sixth cycle without HAART that were not present during previous cycles of SIT. One patient had a M184V
mutation during the fourth cycle without HAART that was not present during previous cycles of SIT. Two patients randomly assigned to SIT who received a protease inhibitorbased regimen had evidence for emergence of mutations that
are associated with resistance to nucleoside reverse-transcriptase inhibitors during cycle 2 without HAART; one patient had an M184V mutation and one had a T215Y mutation.
There was no evidence of the emergence of resistance in patients who received continuous HAART. Three of the 5 individuals in the SIT arm with evidence for the emergence of resistance had received suboptimal therapy with antiretroviral drugs before the initiation of HAART. Of the 26 patients randomly assigned to receive SIT, 11 were receiving their first HAART regimen at the time of enrollment, 8 had received previous therapy with HAART, and 7 had received suboptimal antiretroviral therapy before HAART. Of the 26 patients randomly assigned to receive continuous therapy, 11 were receiving their first HAART regimen at the time of enrollment, 9 had received previous therapy with HAART, and 6 had received suboptimal antiretroviral therapy before HAART.
To evaluate the effect of long-cycle SIT, we evaluated these parameters of toxicity at baseline, week 40 (after the fourth without-HAART period for patients receiving intermittent HAART), and week 48 (after an 8-week resumption of HAART for patients receiving intermittent HAART) in
patients receiving long-cycle versus continuous HAART. There was no difference in the 2 groups at baseline with respect to these 3 markers, and although there was a significant difference in total serum cholesterol and triglyceride levels at week 40 (P = .04 for both) between patients who received long-cycle SIT versus continuous HAART, there was no significant difference in these parameters at week 48. There was no difference between the groups with respect to LDL cholesterol at any time point evaluated.
In addition, there was a significant reduction at week 40 in the intermittent therapy arm in triglyceride levels, compared with baseline values (median, -21%; P = .04), total serum cholesterol levels (median, -19%; P = .001), and LDL
cholesterol levels (median, -15%; P = .001). However, at week 48, there was no significant change compared with baseline in serum triglycerides (median, +2%; P > .5), total serum cholesterol (median, -5%; P = .13), and LDL (median,
-2%; P > .5). There was no significant change at week 40 or week 48, compared with baseline in patients who received continuous therapy in triglyceride (week 40 median, +20%; P = .17; week 48 median, +10%; P = .15), total cholesterol
(week 40 median, +5%; P = .22; week 48 median, +5%; P = .26), or LDL cholesterol levels (week 40 median, +4%; P > .5; week 48 median, -3%, P > .5) (data not shown).
It is of note that 1 patient who received long-cycle SIT and 3 patients who received continuous HAART had been receiving lipid-lowering agents before enrollment. In addition, no one who received long-cycle SIT and 1 patient who
received continuous HAART initiated therapy with lipid-lowering agents during the 48 weeks of the study that were evaluated. When we excluded these patients from the analysis, the results were similar; there was no significant difference in median levels of serum triglyceride, total cholesterol, or LDL cholesterol at baseline or week 48 between patients who received SIT or continuous HAART (P = .15 to P > .5). At week 40, there was a significant difference in median levels of serum triglyceride (P = .05) and total cholesterol (P = .05) between patients who received SIT or continuous HAART.
Plasma hsCRP levels
It has been demonstrated that hsCRP is a highly sensitive marker for determining the risk of cardiovascular disease. Therefore, we evaluated the effects of long-cycle intermittent versus continuous HAART on plasma hsCRP levels. There was no significant difference in the baseline, week 40, or week 48 median levels of hsCRP between patients receiving long-cycle SIT versus those receiving continuous HAART. In addition, there was no significant change in levels of hsCRP, compared with baseline, in patients receiving long-cycle SIT or continuous HAART at week 40 (median, -26%; P > .5; median, +4%, P > .5) or week 48 (median, -7%; P > .5; median, +9%, P > .5) (data not shown).
Certain antiretroviral drugs are known to cause significant elevations in hepatic enzymes in HIV-infected individuals. Therefore, we evaluated the effects of long-cycle intermittent therapy on serum levels of ALT or AST. Patients were ineligible if they were positive for hepatitis B surface antigen. However, 9 of 19 patients in the SIT arm and 12 of 22 patients in the continuous arm were reactive to hepatitis B core and surface antibody. Three patients in the SIT arm and no patients in the continuous arm were reactive to hepatitis C.
There were no significant differences in serum levels of ALT and AST in patients receiving intermittent versus continuous HAART determined at baseline, week 40, or week 48 (table 2). In addition, there was no significant difference
in the change in ALT levels, compared with baseline levels, in patients receiving long-cycle SIT or continuous HAART at week 40 (median, 9%; P > .5; median, -2%; P > .5, respectively) or week 48 (median, -8%; P = .5; median, -6%; P >.5). There was no significant change in AST levels, compared with baseline levels, in patients receiving long-cycle SIT or continuous HAART at week 40 (median, -12%; P > .5; median, +8%; P > .5, respectively) or week 48 (median, -10%; P= .44; median, 0%; P > .5, respectively) (data not shown). One patient in the SIT arm and 8 patients in the continuous HAART arm had abnormal (>34 IU/L) ALT levels at baseline. At week 40, 3 patients in each arm, and at week 48, 3 patients in the SIT arm and 5 patients in the continuous arm, had abnormal ALT levels (data not shown). For AST, 1, 1, and 0 and 3, 3, and 1 patients in the intermittent and continuous arms had abnormal (>41 IU/L) levels of AST at baseline, week 40, and week 48, respectively (data not shown).
Effect of long-cycle intermittent versus continuous HAART on plasma HIV RNA levels. The primary end point of the study was the number of individuals receiving long-cycle intermittent versus continuous HAART who had plasma HIV RNA levels <50 copies/mL as determined at 92 weeks of study participation. Because new enrollment was prematurely terminated, this end point could not be evaluated at 92 weeks. However, 5 (26%) of 19 patients receiving long-cycle intermittent HAART and 2 (9%) of 22 patients receiving continuous HAART had plasma HIV RNA levels >50 copies/mL at week 48 (range, 69265 copies/mL for patients receiving intermittent HAART and 94228 copies/mL for patients receiving continuous HAART; data not shown). Of the 5 patients receiving intermittent HAART who had plasma HIV RNA levels >50 copies/mL at week 48, 3 had evidence for resistance to antiretroviral medications that they
were receiving (data not shown). In addition, 19 of 22 patients receiving continuous HAART experienced at least 1 plasma HIV RNA level >50 copies/mL for a total of 43 times over 48 weeks (mean, 325 copies/mL; range, 502520 copies/mL); 3 patients experienced levels >1000 copies/mL.
The present study was not designed to evaluate the kinetics of rebound plasma HIV RNA after sequential interruptions of HAART; patients were not required to undergo measurements of plasma HIV RNA other than at week 4 of each
interruption. However, of the 114 cycles without HAART among the 26 patients randomly assigned to SIT, 3 patients had plasma HIV RNA levels <50 copies/mL at week 4 without HAART during 5 of 14 cycles without HAART. In addition,
there were data to evaluate the level of plasma HIV RNA at 7 days without HAART during the first interruption cycle in 19 patients with levels of plasma HIV RNA <50 copies/mL at baseline. Two (11%) of 19 patients randomly assigned to receive SIT had >50 copies/mL (336 and 955 copies/mL) of plasma HIV RNA on day 7 of the first without-HAART period (data not shown).
Although this study was not designed to evaluate the potential role of autoimmunization in the control of levels of plasma HIV RNA, the data suggest that multiple interruptions of HAART do not result in immunological control of levels of plasma HIV RNA in patients with chronic HIV infection. The mean log difference in plasma HIV RNA levels from the first to the second, third, and fourth cycles of treatment interruptions of -0.33, -0.27, and -0.46, respectively, were not significantly different (P > .5 for all cycles) (data not shown).
Effect of long-cycle intermittent versus continuous HAART on immunologic parameters
There was no significant difference in the median absolute CD4+ or CD8+ T cell counts measured at baseline, week 40, or week 48 in patients who received long-cycle SIT versus continuous HAART (table 4). In addition, there was no difference in the median percentage of certain general markers of immune activation, CD4+CD25+, CD8+CD25+, and CD4+DR+, as determined at baseline, week 40, and week 48 in patients who received long-cycle intermittent versus continuous HAART. However, there was a significant increase in the median percentage of CD8+DR+ and CD8+CD38+ T cells at week 40 that persisted through week 48 in patients who received long-cycle SIT versus continuous HAART.
DISCUSSION BY AUTHORS
The present study demonstrates that long intermittent therapy with HAART does not result in a diminution of toxicities associated with administration of HAART, nor does it enhance or decrease measurable immunologic parameters. In addition, long-cycle intermittent therapy that results in rebound of plasma viremia during the without-HAART periods had no significant beneficial effect on the control of plasma viremia, which suggests that clinically significant autoimmunization against autologous virus does not occur in this setting.
We have previously demonstrated that short-cycle SIT of 7 days with
HAART, followed by 7 days without HAART, resulted in significant reductions in serum total cholesterol, LDL cholesterol, and triglyceride levels.
In contrast, in the present study, we demonstrated that long-cycle intermittent HAART, administered as 4 weeks without HAART followed by 8 weeks with HAART, did not significantly reduce levels of these serum lipid levels after 48 weeks (4 cycles) of intermittent therapy. The lack of a reduction in serum lipid levels with long-cycle SIT is likely the result of the lengthy period of time required to suppress plasma HIV RNA after an interruption that results in rebound plasma viremia. In this regard, the short-cycle strategy allowed for 50% less drug exposure than continuous HAART versus 30% less drug
exposure than continuous HAART with the long-cycle strategy used in this study.
It is of note that median serum total cholesterol and triglyceride levels were significantly reduced in patients who received long-cycle SIT at week 40, the end of the fourth without-HAART period, but not at week 48, the end of the fourth with-HAART period, compared with patients who received continuous HAART. The pattern of decreased median serum lipid levels during the without-HAART periods that were abrogated during the with-HAART periods was observed with each cycle of intermittent therapy (data not shown). It is unclear whether these periodic reductions in lipid levels could ultimately result in a clinical benefit to patients over time. In this regard, there may be a difference in the number of patients who develop high levels of serum lipid levels over time. Clinical end point studies are necessary to evaluate these
It has been demonstrated that hsCRP levels are highly predictive of the risk of cardiovascular disease and that hsCRP levels can be elevated in patients receiving HAART. In contrast to serum lipid levels, in which fluctuations in
mean levels between with-HAART and without-HAART levels were observed, there was no significant difference between patients receiving long-cycle intermittent versus continuous HAART in hsCRP levels at any time point that was evaluated. Because hsCRP can be a marker of immune activation, a potential decrease in plasma levels that result from an interruption of HAART, which we did not observe, could have been balanced by an increase in immune activation that was observed during the without-HAART periods as a result of a rebound in plasma HIV RNA. In this regard, long-cycle SIT resulted in a significant increase in the percentage of CD8+ T cells that expressed DR and CD38, markers of immune activation. Furthermore, there was no significant difference in mean levels of the hepatic enzymes ALT or AST from
determinations made during the fourth with- or without-HAART period in patients who received long-cycle SIT. Therefore, there was no difference in ALT, AST, or hsCRP levels between patients who received long-cycle SIT versus continuous HAART.
This study had a primary end point of number of patients with plasma HIV RNA levels <50 copies/mL at the end of the seventh with-HAART period of SIT. Because the study was prematurely terminated to new enrollment, it was not possible to evaluate this primary end point. Of note, 14 (74%) of 19 of patients randomly assigned to receive SIT achieved <50 copies/mL of plasma HIV RNA after the fourth with-HAART period. Therefore, it may be possible to design strategies whereby patients continue to receive HAART for >8 weeks to achieve low levels of HIV replication before subsequent cycles of SIT. In this regard, it may be possible to design long-cycle SIT strategies that do not result in the emergence of resistance to antiretroviral drugs.
Although the strategy that we have investigated in this trial was not designed to evaluate autoimmunization (patients resumed therapy at a predetermined time point after interruption, rather than after reaching a certain predetermined level of rebound plasma viremia), the lack of a significant change in plasma HIV RNA levels from the first through the fourth cycles of SIT did not provide support for a significant autoimmunization effect in the patients who were evaluated. These results are consistent with some previous reports of patients who were treated with HAART during chronic HIV infection. In addition, although some reports have demonstrated a persistent increase in CD4+ HIV-specific immune responses as a result of treatment interruptions in patients treated with HAART during chronic HIV infection, our results were consistent with reports in which there was no significant enhancement of these responses as a result of multiple interruptions in HAART.
The lack of a significant decrease in median absolute CD4+ T cell count at week 40 of SIT (after the fourth without-HAART period), compared with continuous HAART, likely occurred because levels of rebound plasma HIV RNA
after only 4 weeks of treatment interruptions were relatively low and quite variable. Patients with relatively low plasma HIV RNA plasma after a without-HAART period may not be expected to have a substantial decrease in peripheral CD4+ T cell counts. In this regard, the mean plasma HIV RNA level at week 40 was log10 3.57 among the patients who received long-cycle SIT. Substantial decreases in CD4+ T cell counts were observed in certain patients with relatively high levels of plasma HIV RNA after a 4-week interruption in HAART (data not shown).
Thus, we have demonstrated that there was a lack of clear benefit in terms of virologic and immunologic effects, as well as of multiple parameters of toxicity, after 48 weeks of long-cycle SIT versus continuous HAART. The emergence of genetic mutations in HIV that are associated with resistance to antiretroviral drugs led to the premature termination of new enrollment in the study. Potential benefits to long-cycle intermittent therapy should be weighed against this risk. In this regard, because a reduction in markers of toxicity, such as serum lipid and liver enzyme and hsCRP levels evaluated in this study, is one of the major reasons for patient and clinician interest in treatment interruption approaches, the lack of a reduction in the long-cycle SIT versus continuous groups in these levels at 48 weeks may be important for assessing the
relative advantages of certain treatment strategies. However, because periodic reductions in serum total cholesterol and triglyceride levels were observed during the without-HAART intervals, there may be a long-term advantage to intermittent decreases in certain markers of toxicity. In addition, if safe regimens can be identified, long-cycle SIT for the sole purpose of less time receiving HAART to reduce the cost of treatment or to provide respites from continuous therapy may be an important clinical strategy; such approaches may have particular relevance in resource-limited settings. Larger studies with
longer follow-up are necessary to evaluate these possibilities.