Switch to Nevirapine Improves Lipids
"Long-Term Efficacy and Safety of Protease Inhibitor Switching to Nevirapine in HIV-Infected Patients with Undetectable Virus Load"
Clinical Infectious Diseases October 1, 2004; 39:1024-1029
Paloma Gil,1 Miguel de Górgolas,1 Vicente Estrada,4 Alberto Arranz,5 Pablo Rivas,1 Carmen Yera,1 Rosa García,3 Juan J. Granizo,3 and Manuel Fernández-Guerrero1
1Division of Infectious Diseases, 2Department of Immunology, and 3Epidemiology Unit, Fundación Jiménez Díaz, Universidad Autónoma de Madrid, 4Hospital Clínico de San Carlos, Universidad Complutense de Madrid, and 5Hospital Príncipe de Asturias, Univesidad de Alcalá de Henares, Madrid, Spain
Simplified highly active antiretroviral therapy (HAART) regimens are becoming widely used, particularly as a result of the side effects of and difficult compliance with protease inhibitor (PI) therapy. However, the long-term efficacy of HAART has not been properly assessed.
We performed a prospective study of 110 patients infected with human immunodeficiency virus type 1 (HIV-1) with undetectable virus load who discontinued PI therapy and initiated therapy with nevirapine without changing nucleoside analogues. Reasons for switching were treatment simplification (45%), lipodystrophy (24%), renal problems (23%), and dyslipidemia (8%). HIV-1 load, CD4 cell count, and fasting biochemistry profiles were performed at the time of switching (baseline) and every 3--4 months thereafter. The aim of the study was to evaluate the long-term efficacy and safety of this combination.
Sixty-eight patients (61.8%) had a duration of follow-up of 3 years. The mean increase in the CD4 cell count after 3 years was 90 cells/μL (13.8% from baseline). Virus loads remained undetectable in all patients but 9 (8.2%). Triglyceride levels dramatically improved at 12 months (a 75% decrease; P < .02) and remained statistically significant over time (P < .04). The same occurred with serum cholesterol levels: there was an initial reduction of 25% (P < .02) and at the end of the follow-up period (P < .015). However, at the long-term evaluation, complete normalization of mean serum cholesterol and triglyceride levels could not be achieved. Sixteen patients (14.5%) had to stop therapy as a result of nevirapine-associated side effects.
The switching of a PI to nevirapine is a safe and well-tolerated option for maintaining long-term virological suppression and immunological control. Three years after starting nevirapine therapy, rates of hypercholesterolemia and hypertriglyceridemia improved, although normal cholesterol and triglyceride values were not achieved.
CHANGE IN CD4, Cholesterol, & Triglycerides
| ||N ||Baseline ||1-yr ||2-yrs ||3-yrs ||P |
|CD4 ||110 ||628 ||684 ||716 ||745 ||NS |
|Chol mg/dl ||21* ||296 ||224 ||228 ||232 ||<.015 |
|Trig mg/dL ||11** ||1792 ||452 ||421 ||460 ||<.04 |
*patients with serum cholesterol levels >240 mg/dL
**patients with serum TG levbels >400 mg/dL
Antiretroviral drug regimens containing protease inhibitors (PIs) are highly effective in achieving suppression of viral replication and increasing the number of CD4+ lymphocytes and their function. A dramatic reduction in morbidity and mortality among HIV-infected individuals associated with HAART has been reported. However, it is estimated that 30%--70% of patients receiving treatment with PIs develop a wide array of side effects, including gastrointestinal disturbances, hematologic and electrolytic alterations, paresthesias, nephrolithiasis, rash, and metabolic abnormalities (such as insulin-resistant diabetes mellitus, hypertriglyceridemia, and hypercholesterolemia), as well as an increased risk of cardiovascular disease. Lipodystrophy is now a common finding in patients receiving HAART, which has been estimated to occur in 50% of patients after 10 months of treatment with PIs in addition to other as-yet unidentified individual risk factors. The development of lipodystrophy seems to be associated with the duration of antiretroviral therapy, body weight, and total CD4+ cell count before treatment with PIs.
PI-sparing regimens have been used in recent years in an attempt to increase adherence to therapy and to reduce toxicity. PIs have been switched to nevirapine for patients who cannot tolerate PIs or who wish for a simplified regimen. However, the long-term efficacy and safety of this strategy has not been assessed thus far. In this study, we show the results of a study of treatment with nevirapine after discontinuation of PIs in 110 patients who were observed for 3 years.
PATIENTS AND METHODS
This was a prospective, observational, longitudinal study of 110 HIV-1--infected patients receiving HAART that included 2 nucleoside reverse-transcriptase inhibitors (NRTIs) (zidovudine, stavudine, lamivudine, or didanosine) plus 1 PI (indinavir, nelfinavir, or saquinavir), with the following characteristics: (1) virus load suppression (virus load, <200 copies/mL) for ⩾3 months before discontinuation of the PI component of therapy, and (2) a reason for PI discontinuation (presence of hypercholesterolemia, hypertriglyceridemia, increased serum creatinine level or colicky renal pain, or lipodystrophy or voluntary treatment simplification).
The aims of the study were as follows: to evaluate the long-term (3-year) efficacy of the combination of 2 NRTIs plus nevirapine in maintaining virus load suppression and CD4+ lymphocyte counts after switching from a PI-containing regimen, to evaluate the tolerance and side effects of the new combination after long-term use, and to assess the short- and long-term evolution of biochemical abnormalities associated with use of PIs (increased creatinine, cholesterol, and triglyceride serum levels) after withholding therapy.
The purpose of the study was explained to the patients, and after obtaining informed consent, the PI component of the regimen was stopped and nevirapine was started at a dosage of 200 mg once per day for 2 weeks, with a subsequent increase to 200 mg twice per day. The NRTIs used were not changed. Lipid-lowering drugs were not administered during the study period. The study was conducted during a period of 54 months (July 1998 through December 2002) at 3 hospitals in Madrid.
Definitions. Lipodystrophy was defined as redistribution of peripheral fat (lipoatrophy of the face, arms, legs, or buttocks) or central obesity. This was accepted when the patient's belief agreed with the findings from the doctor's physical examination. Dyslipidemia was diagnosed on the basis of the serum level parameters published by National Cholesterol Education Program in 1994: hypercholesterolemia was defined when fasting serum total cholesterol levels were >240 mg/dL, and hypertriglyceridemia was defined when fasting serum triglycerides levels were >400 mg/dL. Renal dysfunction was considered when the serum creatinine level was ⩾1.4 mg/dL (normal range, 0.7--1.3 mg/dL). Nephrolithiasis was defined as repeated episodes of colicky renal pain with abnormalities in the urine sediment (hematuria or crystalluria) and ultrasonography findings. Treatment-naive patients were those HIV-infected patients who had not been provided with antiretroviral drugs before HAART. Efficacy was defined in terms of success in virus suppression (<200 copies/mL). Virus rebound was defined as the presence of ⩾500 copies/mL in 2 consecutive samples obtained at least 1 month apart. Tolerance was defined as the absence of side effects associated with nevirapine. Side effects and drug interactions associated with nevirapine therapy were those that developed after the introduction of the drug as described in the literature and included rash, hepatotoxicity, and methadone-withdrawal syndrome due to induction of cytochrome P-450 system. Hepatotoxicity as a result of nevirapine use was defined as an increase of >3-fold of the upper limit of normal for serum glutamic-oxaloacetic transferase or serum glutamic-pyruvic transaminase levels. Changes in γ-glutamyltransferase levels were not considered.
Patients were evaluated at baseline (at the time of the drug switch) and every 3--4 months thereafter. Physical examination, hematology and biochemistry tests (determination of levels of fasting glycemia, cholesterol, triglycerides, and creatinine), and determination of CD4+ cell count and virus load were performed at each visit. For determination of the virus load, the Amplicor kit (Roche; virus load detection level, 200 copies/mL) was used from 1998 through September 1999, and the Ultrasensitive Amplicor kit (Roche; virus load detection level, 50 copies/mL) was used from October 1999 to 2003.
The baseline demographic, virological, and immunological features of patients are outlined. Seventy-eight patients had received indinavir therapy, 3 had received nelfinavir, 2 had received ritonavir plus saquinavir, 6 had received saquinavir, 6 had received ritonavir, and 1 had received indinavir plus ritonavir; for 14 patients, these data were missing. Patients had received PI-based HAART for ⩾6 months before the switch to nevirapine. Seventy-nine patients (72%) were treatment naive at the time of HAART initiation, but 31 subjects had previously been treated with monotherapy or double-drug therapy. At the time of HAART initiation, 68 patients (62%) had a virus load of <100,000 copies/mL, and 42 patients (38%) had a virus load of >100,000 copies/mL.
Regarding coinfection with hepatitis viruses, serological test results could be determined for all but 8 patients. Six patients were chronically infected with hepatitis B virus, 23 patients had serological tests that were positive for hepatitis C virus, and 74 patients tested negative for both viruses.
Although some patients presented with combined side effects due to PIs (for instance, patients with lipodystrophy and dyslipidemia), we classified patients on the basis of the main reason for the PI switch, as follows: voluntary treatment simplification, 49 patients (44.5%); lipodystrophy, 26 patients (23.6%); renal dysfunction and/or nephrolithiasis associated with indinavir treatment, 25 patients (22.7%); isolated hypertriglyceridemia, 6 patients (5.5%); and hypercholesterolemia, 4 patients (3.6%).
Loss to follow-up.
Nine patients (8.2%; 4 during the first year, 2 during the second year, and 3 during the third year) were lost to follow-up.
Interruption of nevirapine treatment.
Thirty-three patients (30%) stopped treatment with nevirapine for the following reasons: nevirapine-associated side effects, 16 patients (14.5%); voluntary treatment interruption, 3 patients (2.7%); change to another simplified HAART regimen with zidovudine, lamivudine, and abacavir (Trizivir; GlaxoSmithKline), 3 patients (2.7%); concurrent acute viral hepatitis A, 2 patients (1.8%); and rebound in the virus load, 9 patients (8.2%).
Clinical, immunological, and virological outcomes.
After 1 year of follow-up, 88 patients remained in the study, and the CD4+ cell count increased to 684.9 ± 34.5 cells/μL, which represents a gain of 57 cells/μL (9%; P = not significant). After 2 years, 77 patients were still in the study, and the CD4+ cell count had increased to 716.8 ± 39.9 cells/μL, with a mean increase of 84 cells/μL (13.3% from baseline; P = not significant). Sixty-eight patients (61.8%) were observed for 3 years. All of these patients were alive at the time of writing, and no AIDS-defining opportunistic infections have occurred. The mean CD4+ cell count was 745.5 ± 43.9 cells/μL, which represented a mean increase of 90 cells/μL (13.8%), compared with baseline values (P = .105) (table 2).
Virus load remained <20 copies/mL throughout the study in all but 9 patients (8.2%) receiving nevirapine therapy. One of these 9 patients did not adhere to treatment. Virus load rebound occurred in 4 patients 10--12 months, in 2 patients 13--24 months, and in 3 patients 25--36 months after nevirapine treatment was initiated. Treatment analysis revealed that an undetectable virus load was maintained in 83 (95.4%) of 87 nevirapine recipients in the first year, 77 (97.5%) of 79 in the second year, and 68 (95.8%) of 71 in the third year. The results of an intent-to-treat analysis were as follows: 83 (75.5%) of 110 patient, 77 (70%) of 110 patient, and 68 (61.8%) of 110 patients maintained an undetectable virus load in the first, second, and third years, respectively.
For the 9 patients whose illness failed to respond to nevirapine therapy, the reasons for the switching to nevirapine were either treatment simplification or lipodystrophy. Four patients were treatment naive before initiation of HAART. In 7 patients, the baseline virus load was <100,000 copies/mL, and it was >100,000 copies/mL in 2 patients. When virus load rebound occurred, detectable virus load values ranged by 3--4 logs. In only 1 of the patients who experienced virus rebound did genotypic resistance tests show resistance to nevirapine, zidovudine, lamivudine, and PIs. All of these patients had a return to undetectable virus load levels after reintroducing HAART that included 1 PI.
Assessment of metabolic abnormalities.
Of 17 patients who developed renal dysfunction associated with indinavir therapy, 14 were observed and underwent periodic creatinine level determinations. All of these cases evolved favorably, showing a normalization of creatinine values (baseline serum creatinine level, 1.7 ± 0.07 mg/dL; serum creatinine level at 1 year, 1.2 ± 0.04 mg/dL [P < .001, compared with baseline]; and serum creatinine level at 3 years, 1.14 ± 0.05 mg/dL [P < .001, compared with baseline]).
Twenty-one patients had serum cholesterol values of >240 mg/dL at the time of switching. The percentages of reduction of the mean serum cholesterol levels from baseline were 24.2% in the first year, 25.8% in the second year, and 24.5% in the third year. By the same token, the clinical evolution of 11 patients who had serum triglyceride levels of >400 mg/dL showed a 74.8% reduction from the baseline level in the first year, a 76.5% reduction in the second year, and a 74.2% reduction in the third year.
Side effects associated with nevirapine.
Sixteen patients (14.5%) developed side effects associated with nevirapine use that led to treatment discontinuation. Acute icteric hepatitis was observed in 2 patients (1.8%), and another 5 patients (4.5%) had asymptomatic but significant elevations of transaminase levels. Nevirapine-associated hepatotoxicity occurred within the first year of nevirapine use in 6 of 7 cases. After discontinuation, the abnormal biochemical levels returned to normal in all these patients. Hepatotoxicity was significantly associated with serological tests that were positive for hepatitis C virus (P < .03).
Seven patients (6.4%) exhibited skin rashes, all of which occurred within the first 2 weeks of nevirapine therapy. Finally, 2 patients (1.8%) developed acute methadone-withdrawal syndrome after nevirapine was begun.
Switching PIs to nonnucleoside reverse-transcriptase inhibitors may improve the quality of life of patients receiving HAART by reducing toxic and metabolic side effects of the drugs and by facilitating adherence to antiretroviral regimens. Thus far, most studies have looked at the efficacy and lipid changes in the short term, and there has been a lack of studies assessing the efficacy and safety of nevirapine therapy for periods of >52 weeks. Our data confirm that, after 3 years of receipt of nevirapine therapy, 96% of patients maintained an undetectable virus load. In addition, an increase in the CD4+ T cell population was seen, and patients did not experience any adverse clinical event associated with progression to AIDS.
At least 1 previous report has shown that simplification with PI-sparing regimens failed to affect the illness in some experienced patients at the time of HAART initiation. In our series, 31 subjects were not treatment naive, but treatment failures were not more commonly observed in this group of patients, which suggests that switching to nevirapine could be safely recommended, regardless of previous experience with antiretroviral drugs.
Reduction of serum levels of cholesterol and triglycerides is one the main goals of PI-sparing regimens. Previous studies of triglyceride levels have yielded conflicting results, and the data in these studies were collected during the first year of treatment. In our study, a dramatic decrease in serum triglyceride levels, without the aid of concomitant lipid-lowering drugs, was observed both in short- and long-term follow-up. The major decrease of serum triglyceride levels was achieved after 2 years of nevirapine therapy, with a mild increase at 3 years. However, as shown by others, patients with exceedingly high baseline triglyceride values (>1000 mg/dL), despite an initial improvement, never did achieve normal values. This observation seems to be important for the future design of long-term HAART therapies, particularly for young patients who are expected to receive treatment for life.
Like others, we also found a significant decrease in the total cholesterol levels 1, 2, and 3 years after the regimen switch. The major decrease after 1 year was followed by a slight to moderate increase in serum cholesterol levels during the subsequent months. Unfortunately, we did not obtain information about high- and low-density lipoprotein cholesterol levels, but this observation suggests that hyperlipidemic patients switching to nevirapine-based regimens would possibly benefit from adding cholesterol-lowering drugs. These results agreed with a recently published meta-analysis of simplified antiretroviral regimens, which showed that normalization of abnormalities in blood lipids as a result of PI intake after simplification with nonnucleoside reverse-transcriptase inhibitors is not consistently achieved.
To date, there have been no conclusive data addressing possible benefits of PI-sparing regimens for improvement of lipodystrophy. Unfortunately, physical changes related to lipodystrophy could not be quantitatively assessed in this study.
Hepatotoxicity is by far the most serious side effect associated with nevirapine use, and cases of life-threatening hepatitis and acute hepatic failure requiring liver transplantation have been reported. Minor elevations of hepatic enzyme levels are commonly found, but acute icteric hepatitis is rarely seen. In this study, 5 patients (4.5%) had silent significant elevations of liver enzyme levels, and 2 patients (1.8%) developed acute icteric hepatitis associated with nevirapine—findings that correlate well with most studies. It has been shown that hepatic damage is more common in patients with concomitant hepatitis B or C infection, and the safety of nevirapine therapy in these patients has not yet been fully established. In this study, hepatotoxicity was also significantly associated with hepatitis C coinfection. It is of clinical importance that nevirapine-associated hepatotoxicity occurs within the first year, and it occurs even more during the first 12 weeks of therapy in most instances; however, cases may occasionally occur afterward.
In summary, switching PIs to nevirapine is an effective strategy for maintaining long-term virological and immunological control in HIV-infected patients. Significant benefits from lowering triglyceride and cholesterol serum levels can be expected. However, total restoration of normal serum lipids is not achieved after the switch, suggesting that lipid-lowering therapy might also be needed in selected patients.