Changes in Lipid Profiles in Two Groups of HIV-1 Infected Patients in Cameroon on Two Treatment Regimens with Either Efavirenz or Nevirapine, in Association with Reverse Transcriptase Inhibitors

The aim of this study was to determine the effect of two antiretroviral therapy regimens on lipid profiles. Patients were allocated to two treatment regimens: Nevirapine (NVP) + Stavudine (d4T) + Lamivudine (3TC) (n = 197) or Efavirenz (EFV) + Stavudine (d4T) + Lamivudine (3TC) (n = 181). Serum was prepared from blood samples collected before the start of treatment (Month 0) and at 24 months. Lipids and lipoproteins were measured using colorimetric enzyme assays or by calculation. Overall, there was an increase in all lipid parameters in patients on both treatment regimens at 24 months, although there were individual differences with respect to each lipid parameter that affected the atherogenicity indices for both regimens. Increase of high density lipoprotein cholesterol (HDLC) (42.82%) was significantly larger in patients on the NVP than on EFV (24.03%) (p<0.001), as opposed to Total Cholesterol (TC), triglycerides (TG) And Low Density Lipoprotein Cholesterol (LDLC) that were significantly lower in patient on NVP than on EFV; TG, Very Low Density Lipoproteins (VLDL) and LDLC increased in both regimens. These changes were not much affected by changes in viral load and CD₄ cell levels. The changes in the atherogenicity indices showed that the regimen with NVP seems to have less risk of coronary heart disease compared to EFV.

INTRODUCTION

In untreated HIV infected individuals, particularly those with advanced infection, lipid abnormalities are common (Grundfeld et al., 1992; Feingold et al., 1993). Abnormalities include variations in Low Density Lipoprotein Cholesterol (LDLC) and High Density Lipoprotein Cholesterol (HDLC) levels (Feingold et al., 1993; Nguemaïm et al., 2010). Further, metabolic disorders, which include disturbances in lipid metabolism and increases in serum level of triglyceride and cholesterol, have been observed in all stages of HIV-infection (Bernasconi et al., 1998; Mooser and Carr, 2001; Danwe et al., 2005). Before the advent of Highly Active Antiretroviral Therapy (HAART), antiretrowviral drugs such as zidovudine (AZT) were shown to decrease plasma triglycerides levels (Mildvan et al., 1992).

Combination of antiretroviral therapy (ART) for treatment of HIV type 1 infection has been associated with fat redistribution, insulin resistance and changes in plasma lipid concentrations (Mildvan et al., 1992; Carr et al., 1998, 1999, 2000; Behrens et al., 1999; Periard et al., 1999; Mulligan et al., 2000). More dyslipidaemia with different patterns have been observed in patients receiving protease inhibitor based HAART (Henry et al., 1998; Bertold et al., 1999; Carr et al., 1999; Dong et al., 1999; Roberts et al., 1999; Bonnet et al., 2000; Bozzette et al., 2003). A recent study, by Smith et al. (2004) demonstrated an excess of cardiovascular risk factors in HIV patients receiving HAART (Carr et al., 1998). Non Nucleoside Reverse Transcriptase Inhibitor (NNRTI) based regimens differ from protease inhibitor (PI) based regimens by the marked increases of LDLC and TG (Van der valk et al., 2001; Tashima et al., 2003). The NVP-containing regimen increased total-cholesterol, HDLC concentration and particle size and apolipoprotein A₁ (apo A₁) levels at 24 weeks (Clotet et al., 2003). Although, no clinical data have yet been generated to support this hypothesis, these differences between ART regimens raised the expectation that NNRTI-based regimens, particularly in view of their effects on HDLC, may favourably modify Coronary Heart Disease (CHD) risk compared with many PI containing regimens. In addition, patients infected with HIV initiating antiretroviral therapy (ART) containing a Non Nucleoside Reverse Transcriptase Inhibitor (NNRTI) show fewer atherogenic lipid changes than those initiating most ARTs containing protease inhibitors (Van Leth et al., 2004). With respect to the two most commonly used NNRTIs, NVP and EFV, no detailed comparative study has been reported in Cameroon concerning their effect on serum lipid profiles. In the present study, we analysed lipid and lipoprotein changes in two groups of HIV-1 patients on two treatment regimens, both including Stavudine (d4T) and Lamivudine (3TC) with one group containing NVP and the other EFV in order to improve the management of HIV-1 infected patients.

MATERIALS AND METHODS

Participants and treatment allocation: Patients enrolled in the present study were more than 15 years (mean 32.62) old. The main exclusion criteria were pregnancy, breastfeeding, abnormal laboratory results at screening and use of drugs which either affect the immune system or lipid parameters (Table 1). Patients took d4T (one tablet of 40 mg twice daily) and 3TC (one tablet of 150 mg twice daily). The first dose of both drugs was taken at 7 am and the second at 7 pm. In addition, patients were randomly allocated to NVP (one tablet of 200 mg twice daily at 7 am and 7 pm), or EFV (one tablet of 600 mg once daily at bedtime or one tablet of 800 mg once daily at bedtime, depending on whether they had associated opportunistic infections treatment, especially tuberculosis drugs). All the patients were Cameroonians and were recruited during the dermatology consultation at the Yaoundé University Teaching Hospital (Cameroon) from September 2007 to September 2009. The study was approved by the National Ethics Committees of the Ministry of Public Health. All patients gave a written informed consent before they were recruited.

Lipid parameters assessment: Serum samples for determination of lipids and lipoprotein parameters were collected at baseline before start (month 0) of antiretroviral treatments and at months 3,6,9,12,18 and 24 months. Blood was taken from participants after 12 h fasting in the dry tubes. After centrifugation at 3000 g for 10 mn, sera obtained were aliquoted and set aside for analysis of blood lipids.

Table 1:	Sampling (eligibility and exclusion)

All samples were stored at -20°C and analysed within a week in the Biochemistry Laboratory of Yaounde Gynae-Obstetric and Paediatric Hospital according to predefined protocols using colorimetric enzyme methods. Total cholesterol was determined using enzymatic method (Allain et al., 1974) and serum triglyceride was determined as previously described by Buccolo and David (1973). The HDLC was determined using a heparin manganese precipitation of Apo B-containing lipoprotein (Warnick and Alberers, 1978) and serum concentration of LDLC was calculated using the Friedewald equation, but only when the concentration of TG was below 500 mg dL^-1 (Friedewald et al., 1972). The VLDL (mg dL^-1) was calculated as TG/5 according to Friedewald et al. (1972) and the atherogenicity indices were calculated using TC/HDLC and LDLC/HDLC ratios. The Viral Load (VL) was measured at Centre Pasteur du Cameroun, Yaoundé laboratory using ultra sensitive Amplicor 1.5 (Roche Diagnostics, Basel, Switzerland) with a lower limit quantification of 50 copies mL^-1 of blood.

Outcome measurement: The mean percentage changes of TG, TC, HDLC, LDLC, VLDL, TC/HDLC and LDLC/HDLC ratios, between baseline (Month 0) and month X after start of treatment were determined for each individual patient using the Van Leth et al. (2004) formula:

where, X is the time-points of follow-up after the start treatment (baseline or month 0).

Changes were analysed with respect to sex, Body Mass Index (BMI), CD₄ cell counts (<50, 50-200, >200 cells μL^-1) or Viral Load (VL) in log₁₀ (< 2.5, 2.5 - 3.5, or > 3.5 log₁₀).

Statistical analysis: All statistical calculations were done using computer programs Microsoft Excel 2003 and the software SPSS (Statistical Package for the Social Sciences, SPSS Inc., Chicago, IL, USA) version 12.0. For the differences between mean percentage changes level of parameters in the two treatment groups NVP and EFV modelled by repeated measurements were tested for the significance compared using the Student t-test. Student t-test was also used to compare the difference between the mean age and the mean of BMI in the two treatment groups NVP and EFV. Chi-square test was used to compare the percentages between the two groups of patients. Independent risk factors were assessed by multivariable analyses including the factors associated with percentage changes of all predefined variables in the treatment groups. The p-values less than 0.05 were considered statistically significant.

RESULTS

Patients: Of the 540 patients included in this study, 241 (44.63%) were allocated to the NVP treatment group and 230 (42.59%) to the EFV treatment group. Of these, 44 (27.16%) patients in NVP group, 49 (30.25%) in EFV group and 69 (42.59%) who did not start their treatment were excluded from the analysed. This resulted in a final sample size of 197 (52.12%) patients in the NVP group and 181 (47.88%) in the EFV group, making a total of 378 patients at the beginning of t he follow-up period (Table 1). A total of 22 patients (12 on NVP and 10 on EFV) disappeared during the follow-up period, leaving 356 patients at month 24.

The baseline characteristics of the patients included in this study are comparable for patients included in the two treatment regimens (Table 2).

Changes in lipids and lipoproteins: The proportional changes of the different serum lipid concentration and atherogenicity indices, over 24 months are graphically depicted in Fig. 1a (for TG), Fig. 1b (for TC), Fig. 1c (for HDLC), Fig. 1d (for LDLC), Fig. 1e (for VLDL), Fig. 1f (for TC/HDLC), Fig. 1g (for LDLC/HDLC). Changes at 24 months were compared to base line values at month zero. All changes within the treatment groups in lipid and lipoproteins concentration as well as atherogenicity indices (TC/HDLC, LDLC/HDLC ratios) were statistically significant (Table 3).

Table 2:	Baseline characteristic of patients included in the study
Values in brackets indicate percentage. NVP: Nevirapine; EFV: Efavirenz; n: Number; CD4: Cluster of Differentiation 4; Cells μL-1: Cells per microliter; SE: Standard error; CDC: Centers for disease control; BMI: Body mass index; A, B, C categories: Stages of the evolution of the HIV infection as classified by CDC/OMS in 1993. There was no statistically significant difference between the two groups of patients in the Table 2

Fig. 1:	Change in serum concentrations of lipids and lipoproteins

Table 3:	Lipid concentrations (standard error) at baseline and 24 months and mean percentage changes
M0: Month 0; M24: Month 24; n: Effective; SE: Standard error; a: Mean values of lipid parameters and atherogenicity indice (SE); b: Mean percentage change (SE) modelled by repeated measurements; CI: Confidence interval; NVP: Nevirapine; EFV: Efavirenz; TG: Triglycerides; TC: Total cholesterol; HDLC: High Density lipoprotein cholesterol; LDLC: Low density lipoprotein cholesterol; VLDL: Very low density lipoprotein; TC/HDLC and LDLC/HDLC: Atherogenicity indices; mg dL-1: Milligram per decilitre; n: Number. *Statistically significant results when the percent increase in the two group of patients are compared (p<0.05).

The increase of HDLC was significantly larger in the NVP treatment group (42.82%) than in EFV treatment group (24.03%) (p<0.001) (Table 3). In contrast, the increase in TC was significantly smaller in NVP group (26.25%) than the EFV group (31.57%) (p = 0.043). These changes resulted in the significant decreased of TC/HDLC ratio in the NVP group (-12.82%) compared to an increase in the EFV group (10.12%) (p<0.001).

The increase of TG was significantly smaller in the NVP group (29.23%) than in the EFV group (52.50%) (p<0.001), as well as the increase of VLDL (29.06%) in NVP group compared to 48.62% in EFV (p<0.001). The difference in increase of LDLC was also statistically significant (26.34% for NVP group compared to 40.07% for EFV group; p = 0.002) (Table 3).

Effect of sex, Viral Load (VL), Body Mass Index (BMI) and CD₄ cell counts: Some factors associated with changes in lipid concentration were analysed by a multivariable analysis (Table 4, 5). According to sex, the increase of HDLC (p = 0.042) and TC (p<0.001) was significantly smaller in women than in men (Table 4). This resulted in a greater decrease of TC/HDLC ratio for men (-9.84%) compared to women (7.14%) (p<0.001) (Table 5). There was a significant increase in all lipid levels and TC/HDLC ratio in patients a ssociated with a decrease of VL over 24 months except for LDLC and LDLC/HDLC ratio whose increase was larger only when VL remained more than 2.5 log₁₀. There was a significant association between VL (log₁₀) and lipids, except HDLC (p = 0.072) (Table 4) and LDLC (p = 0.064) (Table 5).

The BMI was significantly associated with increase in all lipid parameters and LDLC/HDLC ratio, except TC/HDLC ratio (p = 0.062) (Table 5).

Table 4:	Multivariable analysis of some factors associated with percentage changes in lipid (TG, TC and HDLC) concentrations (SE)
M0: Month 0; M24: month 24; a: Mean values of lipid parameters and atherogenicity indice (SE); b: Mean percentage change (SE) modelled by repeated measurements; CI: Confidence interval; M: Male; F: Female; PI: Percent increase; SE: Standard error; BMI: Body mass index; kg m-2: Kilogram per millimetre square; mg dL-1: Milligram per decilitre; Cells μL-1: Cells per microliter; VL: Viral load; log10: Decimal logarithm; CD4: Cluster of Differentiation 4; TG: Triglycerides; T C: Total cholesterol; HDLC: High density lipoprotein cholesterol; *statistically significant results (p<0.05). **Statistically higher significant results (p<0.0001)

Table 5:	Multivariable analysis of some factors associated with percentage changes in lipid (LDLC and VLDL) concentrations (SE) and atherogenicity index (SE)
M0: Month 0; M24: Month 24; a: Mean values of lipid parameters and atherogenicity indice (SE); b: Mean percentage change (SE) modelled by repeated measurements; CI: Confidence interval; M: Male; F: female; PI: Percent increase; SE: Standard error; BMI: Body mass index; mg dL-1: Milligram per decilitre; Cells μL-1: Cells per microliter; VL: Viral load; CD4: Cluster of differentiation 4; LDLC: Low density lipoprotein cholesterol; VLDL: Very low density lipoprotein; TC/HDLC and LDLC/HDLC: Atherogenicity indices; kg m-2: Kilogram per millimetre square; log10: Decimal logarithm. *Statistically significant results (p<0.05)

The CD₄ cell counts were in general, positively associated with changes in lipid parameters and in atherogenicity indices (Table 4, 5).

DISCUSSION

Initiation of an ART regimen containing NVP or EFV is accompanied by a significant increase of HDLC with concomitant increases of TG, TC, LDLC and VLDL. The increase of HDLC was significantly larger in the NVP treatment group compared to the EFV treatment group as well as the increase of TG, TC, LDLC and VLDL. In the NVP group, the TC/HDLC and LDLC/HDLC ratios decreased, compared to an increase in the EFV group. These observations are different from reported changes in most PI-based ART regimens, in which higher concentration of TC, LDLC and TG were found without concurrent higher levels of HDLC (Stein et al., 2003; Fontas et al., 2004) they also differ from the results of Nunez et al. (2002) that did not show any differences between NVP and EFV treatment groups. Present findings agree with those of Van Leth et al. (2004), who used cases from many countries, including African countries.

HDLC increase and NNRTIs: Present results showed an increase of HDLC in both treatment groups at 24 months. Increase of HDLC with the use of NVP or EFV have been described in some previous studies, in patients switching from a PI-based regimen to a NNRTI-based regimen (Martinez et al., 1999; Negredo et al., 2002), for patients initiating treatment with Didanosine, Stavudine and NVP (Van der valk et al., 2001), in patients treated with EFV and either Zidovudine with 3TC or Indinavir (Tashima et al., 2003) in treatment-naive subjects receiving NVP in combination with the nucleoside analogues Zidovudine/Lamivudine (Fisac et al., 2004) and in naive patient starting a regimen of Didanosine, d4T and EFV (Negredo et al., 2002). Apart from the study by Negredo et al. (2002) which included patients with similar baseline HDLC levels as in the present study and which showed increases in HDLC similar to the ones we report here, the others showed different increase patterns, probably due to differences in viral load, the PI-based regimen and the degree of strict adherence of the patients to the treatment protocol. Further, Riddler et al. (2003) also reported similar variations to ours for TC and LDLC.

Several studies have shown that an HDLC increase is associated with a significant decrease in mortality from Coronary Heart Disease (CHD) independent of changes in LDLC (Manninen et al., 1988; Rubins et al., 1999). Other studies have associated risk of cardiovascular disease (CVD) with low concentrations of HDLC (Lipid Research Clinics Program, 1984; Frick et al., 1987; Castelli, 1988; Gordon et al., 1989; Assmann et al., 1996; Robins, 2001). Taken together, our results show that the NVP regimen leads to lower atherogenicity index and so reduces CHD risk better than the EFV regimen. Other studies have shown that a substitution of IP with Nevirapine in combination of antiretroviral therapy for 6-12 improved and even normalized dyslipidemia, glycemia and insuline resistance, whereas HIV suppression was maintained (Molina et al., 2000; Moyle et al., 2001).

Changes in TG, TC, LDLC, VLDL and atherogenicity indices: Present results show that EFV causes a greater increase in TG levels than NVP. These results corroborate those of other authors (Molina et al., 2000; Negredo et al., 2002; Fisac et al., 2004). The results are unlikely to be explained by the effect of d4T, since the proportion of patients in both groups on d4T was virtually the same (52.12% for the NVP treatment group and 47.88% for the EFV treatment group). The d4T might however be responsible for the apparent acceleration of the increase of TC and TG towards the end of the study (Fig. 1), probably related to the lipodystrophy reported in patients on d4T treatment (Heath et al., 2001; Dube et al., 2002; Nolan et al., 2003; Sattler, 2003; Nolan and Mallal, 2004; McComsey et al., 2004) and the possible association of HIV-1 infection with the reduced TG clearance following food intake (Grundfeld et al., 1992). Concentration of TG had an effect in the calculated values of LDLC and VLDL; however, the smaller increase in levels of LDLC and TG in the NVP regimen than for the EFV might suggest that the LDLC and TG results are valid. Furthermore, elevated cholesterol has been observed with the use of EFV (Martinez et al., 1999) and NRTIs such as Stavudine (d4T) (Mallal et al., 2000). A great improvement of atherogenicity indices (TC/HDLC and LDLC/HDLC) on NVP compared to EFV was observed and this result correlates those of Ngondi et al. (2007). The low atherogenicity indices can be explained by the reduction of TC, LDLC and the elevation of HDLC on NVP.

CONCLUSION

The HDLC increases were higher in patients on the NVP regimen than those on the EFV regimen. The increase of TG, TC, LDLC and VLDL levels are smaller for patient taking NVP than for those taking EFV. The less atherogenic lipid profile of patients taking NVP in comparison to those taking EFV may be among the various factors to consider when selecting the most appropriate initial ART regimen, particularly for those patients with HIV type 1 with a significant a priori CHD risk like diabetes, or a previous cardiovascular event.

ACKNOWLEDGMENTS

The study was materially support by the Biochemistry Laboratory of Yaoundé Gynae-Obstetric and Paediatric Hospital, the Biotechnology and Immunology Laboratory of the Faculty of Medicine and Biomedical Sciences of the University of Yaoundé I and the Yaoundé University Teaching Hospital, Cameroon. The authors thank all those who gave their informed consent for participation in the study.