...from the 7th Conferences on Retroviruses and Opportunistic Infections.

... from the 7th Conference on Retroviruses and Opportunistic Infections 

Characterizing primary HIV infection. Several studies investigated primary HIV infection (PHI) in an attempt to characterize the immune response and the effects of antiretroviral intervention during this early phase of infection. One group measured levels of HIV RNA in oral fluids, genital secretions, peripheral blood, and cerebrospinal fluid from subjects with PHI (Abstract 556). High levels of HIV expression were seen in all compartments, especially in seminal plasma and saliva. Antiretroviral treatment (in this case, didanosine/stavudine/hydroxyurea/nevirapine) rapidly reduced the levels of HIV in all compartments. Researchers in France compared CD8 T cell responses from 24 patients with documented PHI to CD8 responses from 30 asymptomatic HIV-infected patients who were antiretroviral treatment naive (Abstract 573). The patients with PHI exhibited a weak CD8 T cell response, which the authors suggested could account for the immune system's failure to control HIV during early infection. A third study showed a strong CD4 T cell response during PHI, which was conserved in patients who received treatment, but was "strongly" reduced in untreated patients (Abstract 583).

Kidney: a reservoir for HIV. Researchers in New York have determined that HIV can infect renal (kidney) epithelial cells. They suggest that direct renal infection may have an important role in the pathogenesis of HIV-related renal disease. However, they do not know if kidney epithelial cell infection is productive. HIV-associated nephropathy is particularly prevalent in the African-American population, where it is the third leading cause of end-stage renal disease in persons between the ages of 25 and 65. (Abstract 161.)

Reservoir rebound. Research suggests that the pool of latently infected, resting CD4 T cells does not completely account for the early rebounding of viral load after highly active antiretroviral therapy (HAART) has been discontinued. One interesting observation was that the HIV envelope of the rebounding plasma virus was genetically distinct from either cell-associated HIV RNA or replication-competent virus in the latently infected CD4 T cells. The authors conclude that other reservoirs of HIV also contribute to the rapid rebound of viral load after discontinuation of HAART. (Abstract 239.)

Phenotypic resistance testing updates. In the VIRA 3001 study (Abstract 237), researchers demonstrated improved virologic outcome when phenotypic resistance testing was used by physicians to guide treatment choices for patients failing their first protease inhibitor containing treatment regimens. The study compared 271 subjects experienced with 2 or more nucleoside reverse transcriptase inhibitors (NRTIs) and 1 protease inhibitor, with viral loads greater than 2000 copies/mL. The subjects were randomized into 2 groups and followed for 16 weeks with treatment decisions based on either phenotypic testing or standard of care. At 16 weeks, 62% of subjects in the phenotypic testing arm had viral loads less than 400 copies/mL, while 33% of subjects in the standard of care arm had viral loads less than 400 copies/mL. Conversely, less benefit was shown in another study (Abstract 786) that evaluated the effect of phenotypic testing on the response of patients failing protease inhibitor containing regimens (no restriction on the number of prior protease inhibitors). Although phenotypic testing helped improve the selection of active drugs, decreased the viral load from baseline (at week 4) and seemed to improve CD4 T cell count, the virologic endpoint in the phenotypic testing group was similar to the standard of care group by week 16. A third abstract described the development of a nonculture-based phenotypic assay for rapid (1 day) detection of resistance to reverse transcriptase inhibitors (Abstract 790).

Improving genotypic resistance testing. Investigators used genotypic antiretroviral resistance testing to identify mutations in plasma versus peripheral blood mononuclear cells (PBMC) in patients with low-level HIV viremia. Greater variability in resistance patterns was seen in sequential samples of PBMC than in plasma. The genotypic concordance between duplicate plasma and PBMC testing was over 95%. The researchers suggest that genotypic testing in plasma may be more reproducible than in PBMC. However, 26% of PBMC samples showed mutations that were not seen in plasma samples. The importance of different mutations in plasma and PBMC compartments has yet to be determined. (Abstract 795.)

Thymic Output. A number of studies attempted to characterize the role of the thymus during HIV infection. Researchers from Montreal and Northwest University in Chicago found that thymic production of naive T cells, or "thymic output," rebounds at the earliest phase of infection, probably compensating for the severe decrease in the number of CD4 T cells during primary HIV infection (Abstract 329). This increase in T cells was independent of treatment, and the authors suggest that early intervention may help with immune reconstitution. On the other hand, a group of researchers from France studied thymic output in patients experiencing virologic rebound (Abstract 453). They found that better thymic output was associated with improved CD4 T cell counts, despite treatment failure. Also, the increase in cell counts decreased with age (probably because thymic output normally decreases with age). A third group from the Gladstone Institute of Virology and Immunology in San Francisco (Abstract 188) found that trends exist between thymic mass (which can increase in some individuals during HIV infection) and levels of T cell receptor excision circles (TRECs). TRECs are used as a measure of the amount of extrachromosomal rearrangement products in the peripheral blood. (T cell receptor gene rearrangements are known to occur in the thymus). Another study (Abstract 328) found that overall thymic output was significantly decreased in HIV-uninfected compared to HIV-infected patients (studied between baseline and week 48 on HAART). Finally, decreased levels of TRECs in PBMC were significantly associated with risk of AIDS and death in a study comparing longterm nonprogressors with progressors (Abstract 123).

Reversing vertical infection? A study by investigators at the University of Massachusetts examined the impact of early combination antiretroviral therapy in 17 vertically infected infants who began therapy at less than 3 months of age and maintained viral loads less than 50 copies/mL for 48 weeks or longer. CD4 T cell counts normalized or remained normal in all infants, and all but 3 infants became HIV seronegative after 48 weeks of therapy. The infants also did not develop HIV specific immune responses. The researchers conclude that early combination antiretroviral therapy for vertical HIV transmission may lead to a cessation of viral replication and could interfere with the establishment of infection. (Abstract 211.)

About those blips. An investigation using genotypic testing revealed that drug-resistant HIV can be selected in vivo during successful antiretroviral treatment and in the absence of sustained viral rebound. Selection of drug resistance occurred in patients who experienced transient viremia (temporary increases in viral load measurable above 50 copies/mL) sometimes called "blips." New resistance mutations did not occur in those patients who did not experience transient viremia. (Abstract 238.)

Lamivudine resistance. A study characterizing viral resistance profiles in over 5000 clinical isolates identified a novel lamivudine resistance pattern. Significant numbers of phenotypically lamivudine-resistant samples lacked the mutation at position 184, suggesting that this mutation does not account for all lamivudine resistance. The authors believe widespread use of phenotypic and genotypic resistance testing will further characterize common and rare resistance profiles and allow for the identification of new resistance pathways. (Abstract 740.)

Neuropathy risks. Data from a longitudinal study of patients at Johns Hopkins AIDS Services indicate that the incidence of neuropathy is higher with the use of some drugs in combination than with individual drugs. The risk of neuropathy was additive or even synergistic for didanosine/stavudine/hydroxyurea compared to didanosine or stavudine alone. The combination of didanosine/stavudine also increased the risk of neuropathy, but less than when hydroxyurea was included. (Abstract 302.)

Indinavir inhibits emivirine metabolism. Emivirine (Coactinon), a non-nucleoside reverse transcriptase inhibitor (NNRTI) in Phase III clinical development, was studied for pharmacokinetic interactions with zidovudine/lamivudine or zidovudine/indinavir. No interactions were seen with emivirine and zidovudine/ lamivudine, so no dosing adjustments are required. Indinavir inhibited the metabolism of emivirine resulting in an 88.5% increase in emivirine levels. This means that the emivirine could not be cleared sufficiently from the body. Therefore, emivirine should not be combined with indinavir. (Abstract 84.)

Nevirapine and oral contraceptives. Nevirapine appears to increase the clearance of an oral contraceptive, ethinyl estradiol/norethindrone (ORTHO-NOVUM 1/35; 21 pack). The study, conducted in HIV-infected women, concluded that "oral contraceptives should not be used as the primary means of contraception when nevirapine is prescribed to women of childbearing potential." (Abstract 89.)

Hydroxyurea may increase toxicity. A randomized trial of continued indinavir/zidovudine/lamivudine versus a switch to indinavir/didanosine/stavudine or indinavir/didanosine/stavudine plus hydroxyurea (Hydrea) was prematurely discontinued after interim review. Primary endpoints that led to discontinuation of antiretroviral therapy were loss of viral suppression or drug toxicity. The group taking hydroxyurea had a shorter overall time to endpoint, and a much shorter time to endpoint specifically caused by toxicity. This group (n=68) also included 3 deaths, 2 from complications of pancreatitis. The researchers conclude that the addition of hydroxyurea to "potent suppressive therapy (in this study) did not enhance efficacy and was associated with excess toxicity." (Abstract 456.)

Amprenavir in protease inhibitor experienced patients. A researcher in Germany reported that amprenavir may be an alternative choice for patients experiencing protease inhibitor cross-resistance. Two thirds of blood samples taken from 132 patients with prior experience of at least one protease inhibitor were not resistant to amprenavir. Over 37% of samples with phenotypic protease inhibitor cross-resistance were still sensitive to the drug. In genotypic analyses, mutations at positions 46, 54, 84 and 90 were significantly associated with amprenavir resistance. (Abstract 726.)

Nelfinavir, efavirenz, or both? A study comparing regimens of nelfinavir or efavirenz or both drugs, with dual NRTIs (one or two new NRTIs assigned) in 195 heavily NRTI experienced subjects reported results up to 48 weeks. The combination of nelfinavir and efavirenz provided significantly greater and longer viral load suppression (less than 50 copies/mL) than other regimens with either drug plus a placebo (all regimens still had dual NRTIs). Also, subjects in the efavirenz or nevirapine arms with viral loads less than 50 copies/mL at week 16 had significantly shorter times to viral loads greater than 200 copies/mL than subjects in the efavirenz plus nelfinavir arm. (Abstract 531.)

Salvage therapy: a very rocky road. Several investigations grappled with the dilemma of salvage therapy. A study from Europe (Abstract 539) found that patients starting a second protease inhibitor containing regimen experienced greater success if the second regimen was initiated at a lower viral load or higher CD4 T cell count, or if new NRTIs were added. Several conclusions came out of a late breaker report on ACTG 398 (Abstract LB7): only 34% of patients achieved viral loads below 200 copies/mL despite implementing 4-5 new agents, dual protease inhibitor regimens were more successful than single protease inhibitor regimens, and previous NNRTI experience was significantly associated with treatment failure while previous protease inhibitor experience was not. Also, a study in amprenavir experienced patients (64% monotherapy, 32% with lamivudine and zidovudine, 4% with other regimens) found that a majority of subjects who switched to a 4-drug regimen (indinavir/nevirapine/ stavudine/lamivudine) had suppressed viremia (less than or equal to 500 copies/mL) for up to 48 weeks (Abstract 526). Finally, a Canadian group reported preliminary results of sustained antiviral responses (less than or equal to 400 copies/mL) in 40-50% of heavily pretreated patients at up to 68 weeks under multidrug rescue therapy, where up to 9 drugs were used. Clearly, such treatments present difficulties with adherence and toxicity.

Non-nucleoside viral susceptibility. A study investigating why NNRTI containing regimens perform well as "salvage" therapy in NRTI experienced patients found a significant correlation between NRTI resistance and NNRTI susceptibility. Increased viral susceptibility to NNRTIs (delavirdine, efavirenz, and nelfinavir) was seen in greater percentages of NRTI experienced patient viruses (29%, 26%, and 21%) than in NRTI naive patients (5%, 9%, and 11%). Researchers also showed a significant inverse correlation between reduced NRTI susceptibility (zidovudine, lamivudine) and increased NNRTI susceptibility. In other words, increased NRTI resistance was associated with greater efficacy of NNRTI therapy against the virus in phenotypic analyses. Viruses with strong susceptibility to NNRTIs had an average of 4.5 NRTI resistance-associated mutations, while those that showed minor NNRTI susceptibility had an average of 0.9 NRTI resistance-associated mutations. (Abstract 234.)

Delavirdine or continued lamivudine. In ACTG 370, 63 patients with dual NRTI experience only (naive to NNRTIs and protease inhibitors) were randomized to one of 3 regimens. Patients experienced with zidovudine/lamivudine were randomized to stavudine/ delavirdine/indinavir. Patients experienced with stavudine/lamivudine or didanosine/lamivudine were randomized to either zidovudine/lamivudine/indinivar or zidovudine/delavirdine/indinavir. In short, one group of patients continued the use of lamivudine while adding delavirdine and indinavir, while the other 2 groups switched both initial NRTIs and added indinavir. As expected the groups that switched both initial NRTIs had better virologic responses. Of the patients who continued lamivudine, 39% had viral loads below 50 copies/mL (compared to 77% and 48% in the zidovudine/lamivudine/indinivar and stavudine/delavirdine/indinavir arms, respectively). This study supports the treatment guidelines (see reference page 5) recommendation of switching all drugs whenever possible when starting a protease inhibitor. (Abstract 525.)

Metabolic and morphologic (M/M) changes: more, not less complex. Alter several years of noting metabolic and morphologic changes in HIV-infected patients, researchers concur that things will get more complicated before they get simpler. Kathleen Mulligan, MD, of the University of California at San Francisco outlined the different changes that have been noted to date. These changes fall into 4 broad categories. Among the metabolic changes are glucose and lipid changes. The morphologic changes include fat accumulation and fat loss. Whereas all these changes were previously lumped into the broad term "lipodystrophy" and blamed on protease inhibitor use, it is now clear that these changes are not completely attributable to protease inhibitor therapy, and in the view of' some, not attributable to antiretroviral therapy at all. Furthermore, certain observational studies show disassociations between these changes, posing the additional question of whether these symptoms represent a single syndrome or a group of related syndromes. (Abstract 20.)

The consequences of M/M changes. The crux of the matter is that these changes--at times resulting in insulin resistance; glucose intolerance, central adiposity, decreased HDL and hypertension--in non-HIV settings generally point to an increased risk of cardiovascular disease and diabetes. Mulligan noted the lack of a clear case definition as a disadvantage to study in this area and emphasized that while treatment ranging from lifestyle changes to lipid-lowering drugs to surgical interventions have been implemented, none has served to ameliorate symptoms completely or to explain the etiology of these changes. (Abstract S20.)

M/M changes: are protease inhibitors to blame? The Australian Prevalence Survey of Lipodystrophy Syndrome is a large study involving 1350 patients followed between November 1998 and June 1999. In this study 15% of patients had never received a protease inhibitor and 63% were receiving at least 1 protease inhibitor. Physicians observed morphologic changes in 55% of subjects. The prevalence of these changes among the different groups was as follows: 81% in protease inhibitor experienced patients, 33% in protease inhibitor naive patients and 5% in antiretroviral naive patients. Another study that more directly implicates protease inhibitors for metabolic complications was a study cited by Mulligan. In that study HIV-uninfected individuals were given ritonavir for 14 days. These subjects experienced immediate elevations in cholesterol, triglycerides, Lp(a) and apoprotein B. Such elevations in metabolic parameters clearly point to protease inhibitors, independent of other factors, as contributors to these metabolic complications. (Abstract S20.)

M/M changes: is stavudine to blame? The ALBI Trial studied 151 patients assigned to 6 months of various combinations of NRTIs. Patients were allowed to switch therapy as required. Patients switched at varying degrees to protease inhibitors and NNRTIs. Thirty-five percent of patients had at least one morphologic change, but researchers noted that prevalence at 30 months was twice as high in the group randomized to the stavudine and didanosine arm than in the groups assigned to the other NRTI combinations (Abstract 19). Another study reported morphologic changes in terms of either fat wasting, fat accumulation or mixed. Among 149 protease inhibitor naive patients taking NRTIs, approximately 39% had one morphologic change. Use of stavudine and duration of treatment were strongly associated with the changes. (Abstract 20.)

M/M changes: is time on drugs or time since diagnosis a factor? One of the more interesting studies presented in this area was one that looked at the incidence of fat redistribution and metabolic abnormalities in patients who were treated during primary HIV infection. Seventeen patients were treated with stavudine, lamivudine and indinavir within 90 days after the onset of symptoms or seroconversion. The patients had received more than 6 months of treatment at this analysis. Two patients switched therapies to stavudine, nevirapine and nelfinavir because of virologic failure. Three patients substituted nevirapine for indinavir due to nephrolithiasis. After a median of 18 months, six patients (35%) developed lipodystrophy (defined here as loss of peripheral fat or increase of central fat). Moderately high cholesterol (greater than 200 mg/dL) was detected in 6 cases and 5 patients had moderate elevations of triglycerides (200-400 mg/dL), while 4 had high elevations of triglycerides (greater than 400 mg/dL). (Abstract 12.)

M/M changes: are nondrug factors to blame? The HIV Outpatient Study (HOPS) attempted to identify factors related to the severity of fat redistribution. Physicians interviewed and assessed 1077 patients. The study found several factors unrelated to drug use associated with fat redistribution. The factors identified included age (over 40 years old), time since HIV diagnosis, time since AIDS diagnosis, time since reaching nadir of CD4 T cells, duration and severity of HIV disease and loss or change in body mass index. Also, the researchers noted an association between moderate to severe fat redistribution with the use of stavudine and indinavir (Abstract 23). Mulligan noted that no patients showed morphologic changes without the presence of nondrug factors (Abstract S20). Finally, a study at Tufts University School of Medicine evaluated the "true" prevalence of fat deposition and fat atrophy by strict anthropometric criteria in a cohort of HIV-infected men and women. (Total number in the cohort is not given.) The study found 63% of the women and 35% of the men had fat deposition; 11% of the women and 19% of the men had fat atrophy. Neither fat deposition nor atrophy was associated with the use of HAART, with or without a protease inhibitor (Abstract 24).

M/M changes: does gender matter? Researchers analyzed a database of case reports of 324 male and 72 female HIV-infected patients. Body shape changes were classified as fat accumulation, fat depletion or both. Reviewers noted histories of hypertriglyceridemia, hypercholesterolemia and hyperglycemia. Fat accumulation occurred in 84% of the males and 97% of the females (a statistically significant difference). Fat depletion was reported in 77% of the males and 61% of the females (also a statistically significant difference). A mix of depletion and accumulation was reported in 65% of the males and 58% of the females. The differences in men and women in rates of mixed depletion and accumulation were not statistically significant. In general men tended to have more fat depletion and more severe metabolic abnormalities, while women tended to have more fat accumulation and less severe metabolic abnormalities (Abstract 26).

Switching to nevirapine. A small study of 40 HIV-infected patients looked at the effects on metabolic complications of switching the protease inhibitor in their regimen to nevirapine. Patients were on their first protease inhibitor containing regimen with undetectable viral loads for at least 6 months. At a median of 24 weeks, fasting triglycerides decreased by 31% and cholesterol decreased by 11%. The change in triglycerides was statistically significant even at this early stage; the change in cholesterol was not statistically significant. There was a decrease in insulin, pro-insulin and C-peptide levels. No changes were observed in fat redistribution. (Abstract 45.)

Benefits of switching to nevirapine. Sixty patients recruited into a study in which one group stayed on their current therapy of stavudine/lamivudine and a protease inhibitor and the second group switched the protease inhibitor for nevirapine and didanosine for lamivudine were analyzed. The first group was subsequently switched to stavudine/didanosine and a protease inhibitor to preserve lamivudine in case of virologic failure. At week 36 nevirapine rash was seen in 3 patients from the switch group; viral load became detectable in 4 of the switch group and 3 of the group that stayed on their protease inhibitor. Cholesterol and triglycerides decreased significantly in the switch group, but there were no significant improvements in body change abnormalities in the nevirapine group. One patient in the switch group died of lactic acidosis. (Abstract 206.)

Switching to efavirenz. A subset of 33 patients who had been on a protease inhibitor and two NRTIs substituted their protease inhibitor for efavirenz. Over 80% of the patients had initially been on indinavir/ stavudine/didanosine. At 10 months no significant differences were seen in weight, fat redistribution, lipid or glucose abnormalities (Abstract 46). Similarly, a French study reported on 39 patients who switched their protease inhibitor for efavirenz. Nucleoside therapy, which was stavudine/lamivudine in 87% of the subjects, was not changed. At 6 months patients subjectively reported partial improvement in body shape and waist to hip ratio decrease. Body mass index did not change and triglyceride, cholesterol, HDL cholesterol and glucose levels all rose (Abstract 48). Finally, another French study looked at changing the protease inhibitor in a regimen for efavirenz in 43 HIV-infected patients who had at least one year of experience on their protease inhibitor. As with the other studies at 6 months, no improvements in lipid profiles or body abnormalities were found (Abstract 49).

Switching to efavirenz, part 2. One Spanish study of 20 patients found some benefit from switching the protease inhibitor to efavirenz. In this study 55% of the patients were on a regimen of indinavir/stavudine/ lamivudine. At 6 months these researchers found a 31% decrease in triglycerides and 28% reduction in the tasting insulin resistance index. No improvement was seen in total cholesterol, HDL cholesterol and glucose levels. (Abstract 50.)

Stopping stavudine. A study of 36 patients who stopped stavudine for 9 months reported improvements in triglycerides and serum lactate levels, as well as improvement in body abnormalities, with 11 patients reporting "a major improvement" in their fat redistribution and 21 patients reporting a partial improvement. Cholesterol, glucose and insulin levels remained unchanged. (Abstract 52.)

Australian switch study. Eighty patients with M/M changes receiving protease inhibitors were randomized in a 2:3 fashion to continue current therapy or switch their protease inhibitor(s) to abacavir/nevirapine/ adefovir/hydroxyurea. CD4 T cell counts fell 70 cells/mm3 in the switch group. Central fat declined more in the switch group, but so did total muscle mass. At week 24 the switch patients rated overall, central and peripheral body abnormalities as less severe. Triglyceride and LDL cholesterol levels fell in the switch group, but C-peptide, insulin and HDL cholesterol levels did not change. (Abstract 205.)

The risk of cardiovascular disease. Several studies looked for evidence of cardiovascular disease or factors that might indicate cardiovascular disease. The reports came to conflicting conclusions depending on the study design. One group studied flow-mediated vasodilation and found evidence that protease inhibitor use was associated with impaired endothelium-dependent vasodilation and endothelial dysfunction, the consequences of which are increased risk for cardiovascular disease (Abstract 29). Another group found the formation of plaques in the carotid and femoral arteries of HIV-infected patients, but these correlated with smoking and age, rather than protease inhibitor use (Abstract 30). Another study of carotid arteries found them slightly thicker in HIV-infected patients than in controls (Abstract 31). A study in 15 women detected no differences in the thickness of carotid intima media by ultrasound in HIV-infected patients compared to HIV-uninfected controls (Abstract 32).

Osteopenia and osteoporosis. Two studies addressed the issue of bone abnormalities observed in HIV-infected patients. In one study 21% of men receiving protease inhibitors had osteopenia (loss of bone density) or osteoporosis (increased porousness of the bones) as compared to 6% of controls. The osteopenia and osteoporosis did not correlate with the presence of morphologic abnormalities (Abstract 207). In the second study 80 men on a protease inhibitor containing regimen with fat loss, but virally suppressed (less than 400 copies/mL) were randomized to stay on their regimen or switch to a nonprotease inhibitor containing regimen. At baseline 28.4% had evidence of osteopenia and 9.5% had evidence of osteoporosis. At week 24 of randomization there was no change in the proportion of subjects with the bone abnormalities. (Abstract 208.)

Mitochondrial toxicity. What causes M/M changes is unclear. One theory in vogue today is that the mitochondrial toxicity of NTRIs, which scientists have long recognized as causing complications such as lactic acidosis, may in fact be responsible for many or all of the M/M changes that are currently observed. David Cooper, MD, in his symposium presentation on lactic acidosis and mitochondrial toxicity postulated that many early factors contributed to obscuring these complications. The late recognition of these complications may be due in part to (1) drugs having been developed quickly without much chance to learn about long-term toxicities or the effects of polypharmacy; (2) survival having been extended and patients having had much longer exposures to the drugs; (3) no intervening opportunistic infections having masked the complications; and finally, (4) co-infection with hepatitis C having played an additional role in increasing the liver's susceptibility to damage. Cooper further theorized that both NRTIs and protease inhibitors may cause overlapping toxicities by different mechanisms and that there may be host factors such as age, gender, HIV infection and genetic predisposition that may add to the complexity of the situation. (Abstract S21.)

Lactic acidosis. The most dangerous and least understood complication of NRTIs is the condition called lactic acidosis. In one study Dutch researchers described the clinical features and predisposing factors of lactic acidosis in 4 patients who died of the complication. All patients had received stavudine. Three patients combined the stavudine with didanosine and one patient combined those two drugs with hydroxyurea. One patient combined the stavudine with lamivudine. All patients presented initially with nausea and vomiting followed by tachypnea (increased rate of respiration). One was treated with riboflavin, but all died within 22 days of admission to a hospital. None of the patients had a history of chronic hepatitis B or C (Abstract 59). Another study in California reported on 20 patients with mildly elevated lactic acid levels and abdominal pain, nausea and distention. Abnormal liver enzymes were noted but not frank acidosis. Five patients had chronic hepatitis B or C, and 3 patients were also diagnosed with pancreatitis. All patients were receiving stavudine and at least one other NRTI and 18 patients were receiving a protease inhibitor or NNRTI. Antiretrovirals were suspended in all patients and the mean time to resolution of the elevated lactate levels was 62 days. All patients survived and only 3 required hospitalization (Abstract 56).

Other M/M hypotheses. Other hypotheses have been proposed to explain some of the M/M changes that some HIV-infected patients have been experiencing. One theory stems from a study conducted at Kaiser Permanente Medical Center in San Francisco. These researchers suggest that white fat cells contain fewer mitochondria than brown fat cells and thus are disproportionately affected by NRTIs. White fat is primarily found in the face, arms, legs and buttocks and, according to this theory, is depleted by the toxicity to its mitochondria. Brown fat, which is primarily found in the stomach, back and neck, has more mitochondria and thus is less affected and expands by necessity to compensate for the loss of white fat. While this theory is intriguing, it requires validation (Abstract 44). Another theory holds that triglyceride and cholesterol elevations are associated with an increase in the cortisol/DHEA ratio (Abstract 203). Finally, a French study examined the impairment of interleukin-2 synthesis by naive CD4 T cells and alteration of TNF alpha T cell homeostasis by memory CD8 T cells under HAART and found a positive correlation with lipodystrophy (Abstract 391).

New drug: lopinavir (ABT-378/r). A study sponsored by Abbott Laboratories examined 72 antiretroviral naive patients with a median baseline CD4 T cell count of approximately 350 cells/mm3 and a median baseline viral load of approximately 100,000 copies/mL. Investigators originally randomized the subjects to receive stavudine/lamivudine plus one of 4 doses of lopinavir (ABT378/r), an experimental protease inhibitor, the pharmacokinetics of which are enhanced by small doses of ritonavir. After 48 weeks, all subjects began 400 mg twice daily of lopinavir. Of the 72 patients out to 72 weeks of therapy, between 93% and 100% had viral loads less than 400 copies/mL, depending on the dose group to which they were initially assigned. (Abstract 515.)

New drug: calanolide A. In a multisite study, researchers randomized 43 HIV-infected subjects to receive one of 3 doses of calanolide A, an investigational NNRTI. Treatment history and baseline laboratory values for the subjects were not stated. Antiretroviral response was dose-dependent, with participants in the 600 mg twice daily group experiencing a mean reduction in viral load of 0.81 log after 14 days of treatment. Despite receiving monotherapy, the participants demonstrated no genotypic evidence of drug resistance. (Abstract 508.)

New drug: DAPD. In a study sponsored by Triangle Pharmaceuticals, researchers examined 20 antiretroviral naive patients with a median baseline CD4 T cell count of approximately 354 cells/[mm.sup.3] and a median baseline viral load of approximately 21,000 copies/mL. The study participants were given one of 4 doses of DAPD, an investigational NRTI. Antiviral response was dose-dependent, with volunteers in the 300 mg twice daily group experiencing a maximum median drop in viral load of 1.45 log. Despite receiving monotherapy, the 18 subjects who underwent genotypic testing demonstrated no evidence of drug resistance. (Abstract 668.)

New drug: tipranavir. In a study sponsored by Pharmacia & Upjohn, investigators examined 31 antiretroviral naive patients with a median baseline CD4 T cell count of approximately 291 cells/mm3 and a median baseline viral load of approximately 104,000 copies/mL. The participants were randomized to receive one of 3 doses of tipranavir, an investigational protease inhibitor. Two of the dosing groups also received 200 mg of ritonavir as a pharmacokinetic booster. Antiviral response was dose-dependent, with volunteers who took 1200 mg of tipranavir plus 200 mg of ritonavir twice daily experiencing a reduction of' 1.6 log in viral load and a CD4 T cell increase of 73 cells/[mm.sup.3] after 2 weeks of treatment. The most common side effects were loose stools or diarrhea, nausea and vomiting. (Abstract 673.)

Structured Treatment Interruptions (STIs). A group of researchers from Spain examined 25 HAART-treated patients with well-controlled viremia for at least 2 years. Thirteen were randomized to remain on therapy and the other 12 to treatment interruption. Upon treatment interruption, 10 of the 12 saw their viral loads rebound to at least 3000 copies/mL within 30 days; 2 remained undetectable. No resistance-conferring mutations emerged as a consequence of the interruption and all patients were able to resume HAART successfully. No significant changes in CD4 T cell counts occurred during the break from therapy. This small study suggests that STIs do not result in the emergence of drug-resistant virus. (Abstract 354.)

STIs from didanosine/hydoxyurea. Researchers led by Franco Lori, MD, examined 2 groups to study the effects of STI. Group I consisted of 8 patients on HAART, 7 of whom had viral loads less than 50 copies/mL. Group II consisted of 9 patients on a 2 drug combination of didanosine/hydroxyurea, 1 of whom had a viral load less than 50 copies/mL. (Mean viral load in Group II was 549 copies/mL.) Treatment was interrupted and viral load rebounded to 10,000 copies/mL in 5 of the HAART patients by week 6; however, none of the patients in the second group rebounded to that level. CD4 T cell counts fell sharply for 5 of the HAART group (to less than 200 cells/mm3), but remained stable in Group II. No clinical evidence of drug resistance emerged as a consequence of the interruption. After restarting therapy, all of the HAART patients were able to push their viral loads back below 50 copies/mL and regain the lost CD4 T cell counts. This small, pilot study suggests that a combination of didanosine/hydroxyurea may serve as a useful platform for STI. (Abstract 352.)

STIs with interleukin-2. Researchers from San Jose and New York examined the effect of interleukin-2 (Proleukin) on STIs. Nine HAART-treated patients who were also using intermittent, low-dose interleukin-2 stopped their drugs but continued to use interleukin-2. Before the treatment interruption, all 9 had viral loads less than 50 copies/mL. Following the break in therapy, viral load became detectable in an average of 19 days and peaked at almost 350,000 copies/mL. However, over the next month, viral load spontaneously dropped in the absence of HAART, but continued in the presence of interleukin-2, to a mean of approximately 26,000 copies/mL. (Pre-HAART viral load averaged approximately 70,000 copies/mL.) During the peak of viremia, CD4 T cell counts dropped an average of 24% and CD8 T cell counts rose an average of 200%. The authors suggest that interleukin-2 may play a role in an HIV-specific CD8 response capable of controlling viral load in the absence of HAART. (Abstract 355.)

Continuing therapy despite failure. Researchers in San Francisco led by Steven Deeks, MD, examined the effect of STI in 18 patients who had failed virologically for more than a year on protease inhibitor therapy. All 18 had a viral load of at least 2500 copies/mL. Following a 3-month treatment interruption, CD4 T cell counts decreased an average of 94 cells/[mm.sup.3] and viral load rose an average of 0.82 log. For 16 of the patients, virus in the plasma reverted from protease inhibitor resistant to wild-type; however, investigators could still culture protease inhibitor resistant virus from peripheral blood mononuclear cells in 4 of the patients for up to 36 weeks after treatment interruption. For all patients, plasma virus resistance to NRTIs generally persisted. The authors conclude that remaining on protease inhibitor-based therapy provides virologic and immunologic benefit despite drug resistance. (Abstract LB10.)

More STI news. Investigators examined 10 HAART-treated patients who had viral loads of less than 20 copies/mL for more than 8 months. Viral load rebounded in all patients following each of 3 treatment interruptions. (The interruptions lasted for 4 weeks; therapy was restarted for 6 months before the next interruption.) In 4 of the patients, viral load rebounded and then fell spontaneously following thc second interruption; this occurred in tandem with in vitro evidence of HIV-specific cellular responses. The so-called "viral set point" (i.e. the viral load before HAART) was lower in 4 of the 7 patients who had completed 3 interruptions. No resistance-conferring mutations emerged, but CD4 T cell counts dropped following the third interruption to their pre-HAART levels. The authors conclude that STI may rally an HIV-specific immune response in at least some patients. (Abstract LB 11.)

Super-infection. A group of Canadian researchers outlined a possible case of "super-infection." Super-infection is the theoretical possibility that an HIV-infected person could be re-infected with a second strain of HIV. In this case Patient A had an 8-year history of nonprogressive, untreated HIV infection with a CD4 T cell count greater than 600 cells/[mm.sup.3.] Patient B had advanced HIV disease and a history of extensive treatment, with a viral load of less than 30,000 copies/mL and a CD4 T cell count of approximately 100 cells/[mm.sup.3] (JB Angel, personal communication). Following the start of a sexual relationship between the two, Patient A experienced a rapid rise in viral load and a decline in CD4 T cell count. Scientists compared stored samples of Patient A's blood that had been drawn in 1989 to samples drawn from both patients in 1998. The samples drawn in 1989 from Patient A had no similarities to the 1998 samples of Patient B. Samples drawn from Patient A in 1998 did, however, demonstrate homology in 5 of 16 protease clones and 4 of 12 reverse transcriptase clones to those of Patient B. Based on these facts scientists concluded that this was "the first report of HIV-1 super-infection" followed by more rapid disease progression. (Abstract LB2.)

Super-hype? Although the report of superinfection is intriguing, many questions remain unanswered. When asked, the lead author of the abstract did not know the particulars of the couple's sexual history, or whether Patient A had engaged in unsafe sex with Patient B or other HIV-infected individuals (JB Angel, personal communication). Second, the researchers did not state why the transmission of drug-resistant virus should produce more rapid disease progression. Other investigators have shown clinical evidence that drug-resistant virus--specifically virus resistant to protease inhibitors--is less fit than wild-type HIV, producing slower disease progression. Finally, the researchers showed clones from Patient A for only 2 time points: 1989 and 1998. It is possible that the 1998 mutations matching those of Patient B evolved during the previous 9 years, before the men ever met. This is only a single case report, giving no indication of how often the phenomenon of super-infection occurs. Contrary to reports in both the popular media and HIV community press, the data presented were inconclusive. (Follow-up on Abstract LB2.)