Side Effect Update: Hypersensitivity, Heart, Bones
Report 6 from the 44th ICAAC, October 30-November 2, 2004, Washington, DC
Written for NATAP by Mark Mascolini
The 44th ICAAC meeting featured a half-dozen interesting reports on antiretroviral side effects, including a possible way to predict hypersensitivity reactions to abacavir, news on lipids and the heart disease predictor C-reactive protein (CRP), and a study finding a higher risk of bone loss in African-American men with HIV than in men of other races.
Predicting hypersensitivity reactions to abacavir
A few years ago Simon Mallal's group in Perth identified genetic wrinkles that predicted a high risk of hypersensitivity to abacavir (1). Now researchers from four US clinics believe they've spotted a way to predict the reaction without gene testing. Sudden, unexplained drops in CD4 and CD8 counts appeared to herald hypersensitivity in this survey of 5 people starting abacavir (2).
Berrigno Rodriguez (Case Western Reserve University) and colleagues reported that 4 of the 5 people were enrolled in an AIDS Clinical Trials Group (ACTG) study of Trizivir (AZT/3TC/abacavir) with or without efavirenz and cyclosporine, an agent given to boost CD4 cells. Two of the 5 people studied took cyclosporine and 3 took efavirenz; 4 were men. They first had symptoms of hypersensitivity 4 to 22 days after starting abacavir, and the symptoms lasted 2 to 9 days. Unexplained plunges in CD4 and CD8 cells preceded or accompanied symptoms in 4 people and followed the onset of symptoms by 6 days in the fifth person.
• Average (range) CD4-cell drop: -166 cells/µL (-110 to -322 cells/µL)
• Average (range) CD8-cell drop: -465.2 cells/µL (-298.5 to -717.5 cells/µL)
• Average (range) total lymphocyte drop: -754 cells/µL (-465 to -1520 cells/µL)
• Average (range) white blood cell drop: -350 cells/µL (-1150 to +4100 cells/µL)
These T-cell plunges did not reflect loss of virologic control, a fall in platelets, or other blood disorders. Nor did these drops correlate with the severity of hypersensitivity reaction symptoms, time since abacavir therapy began, or use of other drugs including cyclosporine, efavirenz, or trimethoprim-sulfamethoxazole. T-cell counts returned to previous levels when the people stopped taking abacavir.
Whether these T-cell ebbs are a specific feature of abacavir hypersensitivity or a result of the systemic illness that characterizes these reactions remains unclear. Rodriguez and coworkers speculated that the T-cell swoons "might represent cellular redistribution [out of the peripheral circulation and] to the areas affected by the inflammatory response." The quick rebound in circulating T cells after abacavir stops, they add, dovetails with that theory.
This phenomenon must be confirmed in bigger groups of people starting abacavir. Even if further study does confirm this finding, it probably would not be a practical predictive tool in the clinic because it would require intense T-cell monitoring in the first weeks of abacavir therapy. But unexplained T-cell slumps that clinicians do happen to notice in people starting abacavir may alert them to the possibility of hypersensitivity.
Low-down on lipid-lowerers for HIV dyslipidemia
Statins and fibrates can help lower cholesterol and triglycerides in people taking antiretrovirals. But as ACTG researchers reported earlier this year, pravastatin and fenofibrate—alone or together—helped few HIV-infected people reach National Cholesterol Education Program (NCEP) lipid goals (3). A smaller retrospective study of HIV-infected men in the Veterans Administration system confirmed that antilipid drugs rarely push cholesterol and triglycerides down to NCEP targets (4).
Jasmin Bhalodia (Campbell University School of Pharmacy, North Carolina) and colleagues studied 53 men cared for at the Durham VA Medical Center. All had total cholesterol or triglyceride readings above 200 mg/dL, all were taking antiretrovirals, and all took lipid-lowering drugs for more than 2 months. Twenty-seven men had tried pravastatin, 12 simvastatin, 7 gemfibrozil, and 1 atorvastatin. Thirty-seven were taking at least one protease inhibitor (PI) and 31 were taking a nonnucleoside, usually efavirenz.
Ten of the men (19%) had guidance from a specialty lipid management clinic, and 20 got dietary advice. Their ages ranged from 28 to 69 years (median 47 years), and they had a median of three risk factors for coronary artery disease. Twenty-seven of the men (51%) had clinical atherosclerotic disease or the coronary artery disease risk equivalent. Half were white.
Total cholesterol dropped 15% after 3 months of antilipid therapy and 13% after 6 months, both significant changes (P < 0.05). Triglycerides, meanwhile, dwindled by 8.5% after 3 months of therapy and by 19% after 6 months (P < 0.05). But among men with lipid measures at 3 and 6 months, only one quarter reached NCEP goals for total cholesterol or triglycerides.
|Lipid target ||At target in 3 months ||At target in 6 months |
|Total cholesterol <200 mg/dL ||9 of 34 (26%) ||8 of 27 (30%) |
|Triglycerides <200 mg/dL ||5 of 19 (26%) ||4 of 17 (23%) |
Average total cholesterol measured 275 mg/dL before treatment, 233 mg/dL after 3 months of lipid-lowering therapy, and 228 mg/dL after 6 months. Triglycerides averaged 516 mg/dL before treatment, 508 mg/dL after 3 months, and 344 mg/dL after 6 months. Protective high-density lipoprotein (HDL) cholesterol rose slightly from 37 mg/dL before treatment to 40 mg/dL at month 6, while dangerous low-density lipoprotein (LDL) cholesterol dropped from 164 mg/dL to 132 mg/dL in that time.
Fosamprenavir's impact on HDL and triglycerides
Lipid analysis of treatment-naive people enrolled in the NEAT study of fosamprenavir (1400 mg twice daily) versus nelfinavir (1250 mg twice daily) found better gains in "good" HDL cholesterol in the fosamprenavir group (5). While triglycerides climbed over 48 weeks in the nelfinavir group, they stayed the same in the fosamprenavir group.
Jeffrey Nadler (University of South Florida, Tampa) randomized 166 people to start fosamprenavir plus 3TC/abacavir and 83 to start nelfinavir with the same two nucleosides. When treatment began, only 4% in the fosamprenavir arm and 3% in the nelfinavir group had an HDL level above the National Cholesterol Education Program's "negative risk" goal of 60 mg/dL. After 48 weeks of treatment, 27% taking fosamprenavir versus 10% taking nelfinavir had climbed into the salutary over-60 mg/dL bracket. The percentage of people with 40 to 60 mg/dL of HDL rose from 27% at baseline to 49% after 48 weeks in the fosamprenavir group, and from 21% at baseline to 43% in the nelfinavir group. Overall HDL cholesterol rose 37% with fosamprenavir and 22% with nelfinavir.
The two groups started treatment with nearly identical average fasting triglycerides: 150 mg/dL in the fosamprenavir arm and 152 mg/dL in the nelfinavir arm. After 48 weeks of antiretroviral therapy, the average in the fosamprenavir group had inched up to 152 mg/dL, while the nelfinavir group's average clambered to 200 mg/dL, the threshold of the NCEP danger zone.
How well does CRP predict heart disease?
Not very well in people with HIV infection, according to results of a 180-person case-control study presented by Peter Sklar (Drexel University, Philadelphia) and colleagues from five other sites (6). Earlier work by Sklar (7) and others (8) suggested C-reactive protein (CRP) could signal impending heart disease in people with HIV infection. But those studies linked CRP to other risk factors, not to actual cases of heart disease. The new case-control study confirmed that traditional risk factors remain strong predictors of so-called cardiovascular events in people with HIV, but CRP added nothing to their predictive power.
Defining cardiovascular disease as myocardial infarction, stroke, peripheral vascular disease, or coronary artery surgery, Sklar matched 60 HIV-infected men who had cardiovascular disease with 120 HIV-infected men of the same age who did not. Although the heart disease "cases" and matched "controls" had similar average CD4 counts (427 cells/µL for cases and 429 cells/µL for controls), the cases had a significantly shorter duration of HIV infection (4.0 versus 5.5 years, P = 0.008) and a significantly shorter time taking antiretrovirals (1.6 versus 3.4 years, P = 0.007). Not surprisingly, the men with heart disease had higher rates of cardiovascular risk factors than the controls:
• Tobacco use: 72% of cases versus 43% of controls
• Hypertension: 41% of cases versus 21% of controls
• LDL cholesterol: 138 mg/dL versus 112 mg/dL in controls
Median CRP measured 2.0 mg/dL (range 0.3 to 46.5 mg/dL) in cases and 1.3 mg/dL (range 0 to 39.9 mg/dL) in controls. But CRP did not correlate with cardiovascular disease in these men, whereas tobacco use (P < 0.001), LDL cholesterol (P = 0.008), and hypertension (P = 0.005) did.
Breaking the groups into thirds with low (0 to 0.7 mg/dL), medium (>0.7 to 3 mg/dL) or high (>3 mg/dL) CRP, Sklar found no significant differences between numbers of cases and controls in the three clusters. However, compared with low-level CRP, medium- or high-level CRP doubled the risk of cardiovascular disease. An analysis including both cases and controls linked higher CRP with traditional risk factors:
• Older age (P = 0.007)
• Tobacco use (P = 0.04)
• Higher total cholesterol (P = 0.008)
• Higher LDL cholesterol (P = 0.001)
But CRP didn't correlate with HIV-related factors like CD4 count, duration of HIV infection, or time taking antiretrovirals.
Finally, Sklar looked at the impact of CRP on two statistical models that predicted heart disease—one model resting on LDL cholesterol alone, and one model combining LDL cholesterol, tobacco use, and hypertension. CRP didn't enhance the predictive power of either model.
Low bone density in African-American men
A survey of 272 active members of the US military, veterans, and their dependents with HIV infection found a significantly higher rate of bone loss among African Americans than among others in the cohort (9). The findings apply mainly to men, who made up 84% of the study group. Most of them, 62%, were African American, while 28% were Caucasian, and 7% Hispanic. PI therapy appeared to protect against bone loss in these people.
Only 14% of the study group smoked, and only 12% downed more than two alcoholic drinks a day. Duration of HIV infection averaged 3447 days, and the average CD4 count stood at 539 cells/µL. Average age did not differ significantly between whites (44 years), African Americans (40 years), and Hispanics (41 years).
Everyone in the study had a DEXA scan between August 2001 and October 2002. The researchers used World Health Organization criteria to define osteopenia (t score between -1 and -2.5) and osteoporosis (t score less than -2.5). They defined bone loss as osteopenia or osteoporosis of the hip or spine.
In a bivariate analysis three factors correlated with bone loss—being African American, less time with an undetectable viral load, and lower body mass index:
| ||With bone loss ||Without bone loss ||P |
|Caucasian (%) ||21 ||35 ||0.054 |
|African American (%) ||68 ||56 || |
|Other (%) ||10 ||10 || |
|Mean days with undetectable viral load + SD ||1030 + 767 ||1272 + 833 ||0.019 |
|Mean body mass index (kg/m2) + SD ||25 + 4 ||27 + 3 ||0.001 |
SD = standard deviation.
Factors not linked to bone loss in this analysis were gender, age, mean duration or HIV infection or antiretroviral therapy, mean days with a CD4 count under 50 cells/µL, lowest-ever CD4 count, current CD4 count, smoking, and PI use. Possibly because classic bone loss risk factors including steroid use, wasting, smoking, drinking, thyroid disease, and hypogonadism were infrequent in this population, the study did not tie them to osteopenia or osteoporosis.
In a multivariate statistical model controlling for smoking, African Americans were 1.7 times more likely than others to have any bone loss (95% confidence interval [CI] 1.1 to 2.9). Limiting the analysis to spine density, the researchers rated bone loss 2.3 times more likely among African Americans (95% CI 1.3 to 3.8). Hip bone loss alone did not correlate with race in this analysis. But in a model controlling for smoking, hip bone loss proved 4.2 times more likely in people older than 50 (95% CI 2.0 to 8.9) and half as likely in people who had ever taken a PI (95% CI 0.2 to 0.8).
Mark Mascolini writes about HIV infection (firstname.lastname@example.org).
1. Mallal S, Nolan D, Witt C, et al. Association between presence of HLA-B*5701, HLA-DR7, and HLA-DQ3 and hypersensitivity to HIV-1 reverse-transcriptase inhibitor abacavir. Lancet 2002;359:727-732.
2. Rodriguez R, Valdez H, Loupa C, et al. ABC hypersensitivity reactions are heralded by transient declines in circulatory CD4+ cell counts. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-171.
3. Aberg JA, Zackin RA, Evans SR, et al. A prospective, multi-center, randomized trial comparing the efficacy and safety of fenofibrate vs pravastatin in HIV-infected subjects with lipid abnormalities: final results of ACTG 5087. 11th Conference on Retroviruses and Opportunistic Infections. February 8-11, 2004. San Francisco. Abstract 723.
4. Bhalodia Jasmin, Sogol EM, Kaye KS, et al. Effectiveness of lipid-lowering therapy among HIV-infected patients with HAART-associated dyslipidemia. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-155.
5. Nadler J, Rodriguez-French A, Wannamaker P, et al. Favorable increases in HDL-cholesterol in HIV-infected antiretroviral therapy-naive subjects receiving GW433908 (FPV) in the NEAT study. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-156.
6. Sklar PA, Nayak GS, Blackwelder WC, et al. Predictive value of C-reactive protein versus traditional risk factors for cardiovascular events among HIV-infected patients. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-159.
7. Sklar P, Blackwelder W, Csako G, et al. C-reactive protein may be an important biomarker of cardiovascular risk and does not appear to be confounded by antiretroviral use or HIV viremia. 10th Conference on Retroviruses and Opportunistic Infections. February 10-14, 2003. Boston. Abstract 742.
8. Vanig TJ, Williams CE, Guilonard K, et al. Elevated high sensitivity C-reactive protein, a marker for cardiovascular disease, in HIV patients. XV International AIDS Conference. July 11-16, 2004. Bangkok. Abstract MoPeC3525.
9. Aronson N, Howard R, Hammock V, et al. Decreased bone mineral density is seen in an African-American male HIV-infected patient cohort. 44th Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC). October 30-November 2, 2004. Washington, DC. Abstract H-166.