Lipodystrophy induced by combination antiretroviral therapy in HIV / AIDS patients : A Belgrade cohort study

Background/Aim. Highly active antiretroviral therapy (HAART) has led to dramatic reductions in mortality and morbidity of HIV/AIDS-patients. Lipodystrophy, a syndrome including peripheral fat wasting and central obesity, is well-documented side effect of HAART. The aim of this study was to evaluate the incidence of lipodystrophy, and to determine its risk ratios in a HIV/AIDS-cohort. Methods. This cross-sectional study included all the antiretroviralnaive HIV/AIDS patients commencing HAART from October 1, 2001 to October 1, 2010, at the HIV/AIDS Center, Institute of Infectious and Tropical Diseases, Belgrade, Serbia. Univariate and stepwise multivariate logistic regression analyses were used to determine the odds ratios (OR) with the confidence interval (CI) of 95%, in order to establish the relative risk for lipodystrophy. The KaplanMeier-method was used to determine the probability of development lipodystrophy over time. All statistical analyses were performed using SPSS software version using 0.05 as a p-treshold for the significance. Results. This study included 840 HIV/AIDS patients, 608 women and 232 men, followed for 5.6 ± 2.8 years. The prevalence of lipodystrophy was 69.2%. Univariate and stepwise multivariate regression analysis identified that the female gender, hepatitis C coinfection, AIDS diagnosis prior to HAART initiation, nucleoside-reverse-transcriptase-inhibitors and proteaseinhibitors based regimens had a high risk for developing lipodystrophy in HIV/AIDS-patients (OR = 1.6, 95% CI = 1.1–3.49, p = 0.04; OR = 3.31, 95% CI = 1.3–6.8, p < 0.01; OR = 3.7, 95% CI = 1.7–6.1, p < 0.01; OR = 2.1, 95% CI = 1.7–3.3, p < 0.01; OR = 6.1, 95% CI = 4.1–9.7, p < 0.01, respectively). Conclusion. Despite much greater life expectancy of HIV/AIDS-patients, treatment-related toxicities still remain a major concern. Monitoring of lipodystrophy, as side effect of HAART, is particularly important.


Introduction
The use of highly active antiretroviral therapy (HAART), in treatment of human immunodeficiency virus (HIV) has led to dramatic reductions in mortality and morbidity of patients 1 .From a very bad prognosis in its beginnings, HIV infection and acquired immunodeficiency syndrome (AIDS) meanwhile became the chronic disease, which can be treated, controlled and supervised.
HAART usually consists of the combination of two drugs from the nucleoside reverse transcriptase inhibitors (NRTI) group, and a single drug from the protease inhibitors (PI) group (2 NRTIs + 1 PI) or of one non-nucleoside reverse transcriptase inhibitor (NNRTI).The latter combination (2 NRTIs + 1 NNRTI) is equally efficient as the first one, so that the PIs are often reserved for the later stages of the treatment due to their higher resistance barrier 2 .
Lipodystrophy, a syndrome including peripheral fat wasting and central obesity, which in some patients may be followed with metabolic changes such as hyperlipidemia and glucose intolerance, or diabetes mellitus, is well-documented side effect of HAART 3,4 .Changes in body composition include lipoatrophy, a complete or partial loss of adipose tissue predominantly in limbs and face and lipohypertrophy, which is pathological accumulation of adipose tissue in the omentum, mesenterium, retroperitoneum and pelvis, dorsocervical region, or breast enlargement in women 5 .Lipoatrophy and lipohypertrophy may or may not coexist.Reported prevalence rates of lipodystrophy range from a few percent to over 80% 4,5 .
The aim of this study was to evaluate the incidence of lipodystrophy, and to determine the relative risk rates of this side effect of HAART, in HIV/AIDS cohort of antiretroviral-naive patients commencing HAART in Belgrade, Serbia.

Methods
This cross-sectional study included all eligible HIVinfected patients receiving HAART at the HIV/AIDS Center of Institute of Infectious and Tropical Diseases, Clinical Center of the Republic of Serbia, Belgrade.
To be eligible in the study, a patient had to be 18 or older, with documented HIV infection 6 , antiretroviralnaive patients commencing HAART from October 1, 2001 to October 1, 2010, and with collected data about the occurrence of lipodystrophy.The clinical diagnosis of lipo-dystrophy was made on the basis of physical examination, by recognizing changes such as fat loss from face, arms and legs, and accumulation of fat in the abdominal and/or dorso-cervical region, including breast enlargement in women, as described in the European AIDS Clinical Society (EACS) guidelines 7 .
The exclusion criteria were: simultaneous therapy with corticosteroids, anabolic steroids, as well as the treatment of some of the AIDS-related opportunistic infections, within 3 months before the lipodystrophy diagnosis.
During the study period, 25 antiretroviral drugs were registered worldwide, of which the following were also registered in Serbia: zidovudine, didanosine, lamivudine, stavudine, zalcitabine and abacavir (from the NRTI group), nevirapine and efavirenz (from the NNRTI group), saquinavir, nelfinavir, indinavir, fosamprenavir, lopinavir/ritonavir, and ritonavir for boosting other PIs (from the PI group).Nelfinavir and indinavir were withdrawn in 2008.Enfuvirtide was introduced in 2007 for salvage regimens.Newer drugs, such as tenofovir/emtricitabine, tipranavir, atazanavir, etravirine, raltegravir, as well as maraviroc, were not included among the drugs reimbursed by the national health insurance system during the study period.
The immunological and virological responses to HAART were evaluated every 4-6 months by measuring plasma viral loads (pVL) and CD4+ T-cell counts.Due to shortages, HAART monitoring was not performed regularly, which led to the delay in the measurements of CD4+ T-cell counts and viral loads.CD4 cells were quantified by flowcytometry.Plasma HIV-1 RNA loads were measured by quantitative reverse transcriptase polymerase chain reaction (ultrasensitive assay version 1.5, Roche Molecular Systems, Branchburg, NJ, USA), with a lower limit of detection of 50 copies/mL (1.7 log 10 ).
All statistical analyses were performed using SPSS software version using 0.05 as a p treshold for the significance.Possible associations between lipodystrophy and gender, age above 40, coinfection with hepatitis C virus, AIDS 7 at HAART initiation and the type of HAART regimen, were tested.Univariate and stepwise multivariate logistic regression analyses were used to determine the odds ratios (OR), with the confidence interval (CI) of 95%, in order to establish the relative risk for the occurrence of lipodystrophy.The Kaplan-Meier product limit method were used to determine the probability of developing lipodystrophy over time.
The study was approved by the Clinical Center of Serbia Ethics Committee, Belgrade, Serbia.
HAART regimens included the following: combinations of two NRTIs as a treatment backbone, with one or two PIs (taken by 14.3% of all the patients), or one NNRTI (33.3%), and multiple combinations of drugs from all the classes (52.4%).
The prevalence of lipodystrophy was 69.2%.The overall estimated probability of developing lipodystrophy increased with time, reaching 100% after ten years of treatment, with median time of 7 years for the development of lipodystrophy in 50% of patients (Figure 1).In the subgroup of patients coinfected with HCV, the prevalence of lipodystrophy was 41%, as opposed to 25.8% among those without HCV (p < 0.01).

Fig. 1 -The estimated probability (Kaplan-Meier product limit method) of developing lipodistrophy over time.
Univariate logistic regression analysis (Table 1) showed that female gender, age above 40, HCV coinfection, AIDS at HAART initiation, and prolonged usage of NRTIs, were all associated with lipodystrophy.Among the NRTIs, multivariate logistic regression analysis indicated that highest risk for lipodystrophy was associated with the usage of stavudine, didanosine and zidovudine.In contrast with this, lamivudine and abacavir were not significant predictors (OR 2.1, 95% CI 1.6-3.9,p = 0.09; OR 3.3, 95% CI 1.7-4.1,p = 0.6, respectively).The usage of NNRTI based regimens carried a lower risk for lipodystrophy (OR 2.1, 95% CI 1.2-3.3,p < 0.01).The same analysis showed that the PI regimens were also associated with the development of lipodystrophy (OR 5.9, 95% CI 3.7-9.6,p < 0.001).The stepwise multivariate regression analysis identified that the PI based regimens had together with NRTI drugs, female gender, HCV coinfection, and AIDS diagnosis prior to HAART initiation, a high risk for developing lipodystrophy in HIV/AIDS patients (Table 1).

Discussion
Within the recognition that HIV production could be suppressed, but not eradicated, the focus of antiretroviral therapy in the mid-1990s was to convert the infection from uniformly fatal disease into a long term, manageable condition.This accomplishment was reported in 1998, with the seminal publication from the Multicenter AIDS Cohort Study (MACS), associating a dramatic drop in death rates from AIDS with the widespread application of HAART in the USA 1 .However, as patients may potentially be exposed to HAART for decades, treatment-associated toxicities, such as lipodystrophy, remains a concern.
The frequency of lipodystrophy in the Belgrade cohort was 69.2%, which was in correlation with the results of Miller et al. 8 .Some previous studies had shown a lower prevalence of lipodystrophy, but that was probably in connection with the lack of precise definition of the syndrome, and the fact that only the most severe cases had been registered [9][10][11] .
We also showed that the probability of developing lipodystrophy increases with time, reaching 100% by 10 years of HAART and with the median time of 7 years for the development of lipodystrophy.Lipodystrophy identification had profound implications for the management of HIV/AIDS.Its recognition as a side effect of HAART led to the reevaluation of the appropriate time to start the therapy and also to various modifications in therapy of HIV.Numerous cross-sectional studies and many prospective cohort studies analyzed the occurrence of lipodystrophy, and attempted to determine the factors that cause it.In our study, both the age above 40 and AIDS diagnosis at HAART initiation were associated with lipodystrophy, but the age above 40 was not an independent predictor.Several studies have demonstrated that the older age and AIDS diagnosis prior to HAART initiation are the risk factors for lipodystrophy [12][13][14] .There are conflicting data regarding the risks associated with gender.Martinez et al. 15 , in their study, found that the female gender is in higher risk for lipodystrophy.Thiebaut et al. 16 , on the other hand, came to the contrary conclusions, that the female gender was not in higher risk for lidodistrophy.Several other studies established HCV coinfection as an independent predictor of lipodistrophy, demonstrating the association of HCV co-infection with higher rates of lipodystrophy [17][18][19] .
NRTIs, especially thymidine analogues, are known to cause mitochondrial toxicity 20 .Given the similarities in function between HIV reverse transcriptase and human DNA polymerase, it is not surprising that nucleoside analogues are competitive inhibitors of human DNA polymerase-gamma, a key enzyme for mitochondrial DNA (mtDNA) replication.The accumulation of mtDNA deficits induces a deficient production of molecules devoted to the intramitochondrial synthesis of adenosine triphosphate (ATP).Once ATP production drops below a certain threshold, sudden mitochondrial and then cellular damage occurs that can lead to cell death.Tissues and organs cannot function properly, and the damage becomes clinically apparent 21 .The clinical use of NRTI has been associated with adverse effects caused by mitochondrial dysfunction, such as acute pancreatitis, myopathy, peripheral neuropathy, anemia, neutropenia, hepatic toxicity and hyperlactataemia/lactic acidosis [20][21][22] .Our data show that the prolonged usage of NRTIs, especially thymidine analogues such as stavudine, didanosine and zidovudine, is likely to be associated with the development of lipodystrophy, which is in accordance with numerous other studies [23][24][25] .NRTIs induce lipodystrophy by way of mito-chondrial toxicity, which is not the case with PIs and NNRTIs.Moreover, NNRTI based HAART is less likely to induce lipodystrophy than the PI based HAART 26 .
Lipodystrophy syndrome, together with the insulin resistance, type 2 diabetes mellitus and hyperlipidemia develop a cluster of metabolic abnormalities, referred to as the metabolic syndrome.Metabolic syndrome, increases the cardiovascular risk and has great impact on life expectancy in HIV/AIDS patients 27 .
Life style modifications and exercise training, antiretroviral switch strategies, pharmacological management (rosiglitazone, metformin, human growth hormone, human growth hormone-releasing factor, leptin, etc) and reconstructive surgery could be possible treatment options for lipodystrophy syndrome 14 .
The initial enthusiasm of clinicians with HAART success, primarily because of the possibility of the long-term control of HIV replication, is suppressed with the new knowledge about the potential toxicity of antiretroviral drugs.For some patients lipodystrophy is only a cosmetic problem, but for others the very same problem may affect their future decisions on the treatment, making them cease the therapy or significantly reduce treatment compliance.As the result, the patients may experience rebound in viral load or the development of viral resistance.This would eventually lead to disease progression and even to an increased risk of disease transmission, possibly by a drug resistant viral strain 28 .

Conclusion
Despite much greater life expectancy of HIV/AIDS patients, treatment related toxicities still remain a major concern.Monitoring of the side effects of highly active antiretroviral therapy, such as lipodystrophy is, particularly important.