Atherogenic risk assessment of naive HIV-infected patients attending Infectious Diseases Service of Kinshasa University Teaching Hospital, Democratic Republic of the Congo (DRC)

Background and aim: Metabolic abnormalities are common in HIV/AIDS. Increasingly, lipid ratios are used as screening tools for dyslipidaemia in these medical conditions. The aim of this study was to assess the ability of 4 lipid ratios to predict cardiovascular risks. Methods: This is a cross-sectional and analytical study included 105 HIV+ patients followed in Kinshasa University Teaching Hospital (KUTH). Four indices [Atherogenic Index of Plasma (AIP), Castelli Risk Index (CRI) I and II, Atherogenic coeﬃ cient (AC)] were compared. Statistical analyzis consisted of measuring frequencies and means, Student’s t-tests, ANOVA and Ficher’s exact test, and the calculation of the Kappa value. Results: Lipid ratios predicted respectively the risk in 62% (AIP), 28.6% (CRI-I) and 23.8% (CRI-II). CRI-I and II were elevated, especially in women. The AIP appeared to be a better predictor than CRI-I and II to assess dyslipidaemia in general and the high-risk frequency. The cholesterol detected risk in 66.7% (Low HDL-C), 50% (High LDL-C), 38.9% (High TC and/or TG). The atherogenic risk was higher with age, advanced WHO stage, HIV-TB, HBV-HCV co-infections, smoking and alcohol intake. Haemoglobin (Hb) and CD4 counts were low when the risk was high. Age ≥ 50 years, stage 4 (WHO), CD4s


Introduction
HIV infection is currently a chronic condition since the use of highly active antiretroviral therapy (HAART) [1,2]. Infections in general [3] and HIV infection particularly, through immune activation and in lammation that characterize them, lead to dyslipidaemia, which is the cause of atherosclerosis [2,4]. HIV per se is responsible for the onset of this medical condition, but opportunistic infections (OIs), antiretroviral therapy, traditional atherogenic risk factors [smoking, hypertension, total cholesterol, decrease in HDL-C , increased TG, diabetes mellitus, family history of coronary artery disease (CAD)] and many other factors are implicated in the occurrence of atherosclerosis during HIV infection [4][5][6]. From the irst descriptions of HIV/AIDS, heart attacks had been described: dilated cardiomyopathy, myocarditis, infectious endocarditis, pericarditis, pericardial effusion, tamponade, arterial hypertension [5]. Since the introduction of HAART in 1996, CAD including myocardial infarction has been observed frequently [7]. HAART, dyslipidaemia, and other traditional cardiovascular risk factors have been implicated in the development of cardiovascular complications [8,9]. In HIV infection, HDL hypocholesterolaemia, hypertriglyceridemia and signi icantly increased visceral fat were noted [10,11]. The decrease in HDL-C and the increase in TG are two factors which act synergistically and have been responsible for the occurrence of myocardial infarction [1]. HAART has reduced morbidity and mortality from HIV, thereby improving the patient quality of life. Nevertheless, the propensity of protease inhibitors (PIs), especially those of the irst generation and other classes of ART, to induce dyslipidaemias and insulin resistance has been observed, although these results have not been observed and veri ied in all studies [12]. During HIV infection, the risk of CAD increases because the risk of atherosclerosis increases. New complications associated with HIV infection leading to chronic pathology causing cardiovascular disease and cancer are emerging [1]. The toxicity of the ARTs to which the patient is exposed generates metabolic disturbances and premature aging. These abnormalities related to HIV infection itself, ARTs (particularly PIS), restoration of health, body modi ications and increased risk of cardiovascular disease, independent of traditional risk factors put strain on health systems especially in resourcelimited areas where many patients are still reaching the stage of advanced HIV disease despite WHO recommendations that patients be treated immediately after testing (test and treat strategy). Providers in these regions, like those in regions with more resources, will increasingly have to deal with multimorbidity for several years. Health systems will need to organize the management and prevention of these medical conditions associated with HIV disease. It is therefore essential to assess the atherogenic risk. This evaluation can be carried out by various methods, in this case by scores such as the European Systematic Coronary Risk Evaluation (SCORE), the Framingham Risk Score (FRS), the Data on Adverse Events of Anti-HIV Drugs (DAD) Score to predict cardiovascular risks over 10 or 5 years [13][14][15][16][17]. Assessment can also be done by looking for subclinical atherosclerosis, obtained by measuring the thickness of the common carotid intima-media. This is done using Doppler ultrasound [17][18][19]. Currently, indices obtained from the lipid ratios are increasingly used. They predict cardiovascular risk for cardiovascular disease and can be applied effectively in PLWHIVs [20][21][22][23]. Very few African studies [20][21][22][23][24] have evaluated the cardiovascular risk during HIV infection. Africa has the highest number of PLHIVs and does not have the means to properly care for them given limited resources. As the life expectancy of PLWHIVs increases with HAART, providers should be aware of the cardiovascular complications of HIV and should include evaluation of risk factors and cardiovascular risk in clinical routine. The burden of non-communicable diseases associated with HIV is increasing, and there is a need to use non-biological factors in patient management and to predict cardiovascular risk in PLWHIVs. In addition to prevention, care, HAART, programs to ight HIV/AIDS in Africa must develop innovative approaches to deal with these new medical conditions. It is therefore of interest for us to take stock of these questions to which many providers are not prepared in the context of HIV and which have been little studied in Africa and DRC. It is in this context that this study was initiated. The aim of this study was to assess the cardiovascular risks of our patients by using indices derived from lipid ratios used as tools for screening and predicting cardiovascular risks.

Methods
This is a cross-sectional and analytical study that was carrying out in the Infectious Diseases Service, Department of Internal Medicine, KUTH, DRC. It covered a period of 6 years (from January 1, 2008 to December 31, 2014). It concerns a consecutive series of HIV positive patients followed in outpatient and hospitalization. The parameters of interest for this study are listed below.
The average values of the lipid balance and the atherogenic risk prediction indices in the groups were studied and the comparison of atherogenic risk according to the indices [AIP, CRI-I and II, AC].
The diagnosis concordance of high risk of atherogenicity between various indices.
Clinical characteristics according to atherogenic risk.
Biological characteristics as a function of atherogenic risk.
Factors associated with high atherogenic risk.
The abnormal values of the indices used in the study are listed in table 1.
The reference values of biological parameters are listed in table 2.

Statistical analysis
Data were analyzed using Excel 2010 and SPSS for windows software version 22.0. Descriptive statistics were used to calculate frequencies, means and standard deviations. The use of Student's t tests and ANOVA made it possible to compare the means of 2 groups for the Student's t test and of more than 2 groups for the ANOVA test. Fisher's x 2 and exact test compared the proportions. The diagnostic concordance of atherogenic risk prediction according to the indices was evaluated by calculating the kappa value. The factors associated with the elevated atherogenic risk were sought using logistic regression tests in univariate and multivariate analysis; calculations of OR and 95% con idence intervals were used to estimate the degree of association between high atherogenic risk and independent factors. p < 0.05 was the threshold of statistical signi icance for all the tests used.

Ethical considerations
As the data had been de-identi ied, the use of informed consent and an ethics committee was not required. The staff of the service enforced ethical rules when collecting and processing data. The study had respected the rules of con identiality, justice, and charity of PLWHIVs during all the processes leading to the publication of this article.

Results
The general characteristics of the study population are in table 3.
Only 38.8% of our patients had performed the lipid assessment (105 out of 270 patients).   Of the 105 patients in the study, 61.9% were female (Sex ratio: 2 F/1 M). The mean age was 44.5 ± 11.5 years. The majority were married (47.4%), without profession (47.1%), of secondary education (58.3%), attending revival churches (50%). The average socioeconomic level predominated (87.5%) for patients for which data was available and a high frequency of the Kongo ethnicity in half of the cases (Table 3). Table 4 shows the lipid balance and the predictive indices of atherogenic risk in the groups studied. The mean total cholesterol ± SD is higher in women (138.8 ± 42.4 for men and 163.0 ± 59.6 for women; p = 0.027). More women than men experimented hypercholesterolemia (38.5% of women and 12.5% of men; p = 0.003). Low HDL-C levels are statistically high in women (53.8%) than in men (25.0 %; p = 0.003). In general, dyslipidaemia is more frequent in women (46.2%) than in men (25.0%), p = 0.024. Overall CRI-I and CRI-2 are high, and they are statistically higher in women than in men.
Dyslipidaemia in general had a frequency of 38.1% in the study group (Table 4).
Risk of AIP atherogenicity in three categories Figure 1 shows 62% of patients had a high atherogenic risk.
Comparison of atherogenic risk according to the lipid ratios and fractions Figure 2 shows that the AIP and low HDL-C predicted a high frequency of risk compared to CRI-I, CRI-II and dyslipidaemia in general, but HDL-C scores better.
The high risk of atherogenicity according to the indices considered was completely consistent with the low HDL-C, CRI-I and CRI-II index (kappa = 1) on the one hand and fairly well concordant (kappa = 0.863) between TC and / or high TG and High LDL-C (kappa = 0.895) on the other hand (Table 5). Table 6 shows that the atherogenic risk signi icantly increased linearly with the age of the patients and is higher (p < 0.001) when patients are elder. The risk increased also with the WHO clinical stages of infection. In the other hand, the frequency of high risk was signi icantly increased in patients with a history of TB, with HBV-HCV coinfection, tobacco, and alcohol users (p < 0.05) and the frequency of fever was higher in patients with high atherogenic risk (p = 0.013). However, the frequency of weight loss decreased when the risk was high (p = 0.015). The mean of SBP and DBP was signi icantly higher in patients with high atherogenic risk. Patients with TB and anaemic PLWHIV had a signi icantly high frequency of elevated atherogenic risk (Table 6).
At high risk, Hb and CD4s + levels decreased while the WBC count increased signi icantly when the risk was high (Table 7).

Discussion
This study evaluated the prediction of cardiovascular risk in HAART-naive HIV patients attending the KUTH using lipid ratios. The general characteristics and the lipid abnormalities observed in the study population have been described and discussed previously in another study. Dyslipidaemia was present even before the initiation of HAART [25]. In this study, lipid abnormalities were found more in women than in men (Statistically higher mean TC value, more frequent hypercholesterolemia, statistically signi icant low HDL-C, more frequent presence of dyslipidaemia). The reasons for this variation are not known. Lipid ratios (AIP, CRI-I, CRI-II and AC) have been studied as predictors of cardiovascular risk in this paper. Other authors have also used these tools to predict the atherogenic risk in HIV infection. [20][21][22][23][24]26] Apart from HIV infection, these indices have also been used to assess cardiovascular risk in pre-eclampsia, type 2 diabetes mellitus, in stroke, among university staff in Malaysia, in dwellers in Nigeria etc... [28][29][30]31]. In our study, comparisons between these indices have been made and have shown that they can predict this risk reliably. By comparing the 4 indices, AIP seems to be the best marker of atherogenic risk. AIP can be considered as an independent factor impacting on cardiovascular risk [21]. AIP has been found to be frequently abnormal in HAART-naïve PLWHIVs. It is recommended to make it an index to be calculated in a baseline assessment of PLWHIVs before starting HAART [22]. AIP, AC, CRI are predictors of emerging cardiovascular risk in treated PLWHIVs [23]. In this series, CRI-I and CRI-II were also statistically higher in females. These indices can therefore be calculated before the initiation of HAART and be used in the evaluation of cardiovascular risk apart from cholesterol and its fractions as well as AIP. Dyslipidaemia in general was only detected in 38.1% while with AIP the risk was predicted in 62% of cases. Nevertheless, CRI-I and CRI-II had a low predictive ability, even when compared to TC and/ or high TG. In our series, the cardiovascular risk is high in the elderly, elderly women, people with advanced HIV disease. For various reasons, the diagnosis of HIV infection is made late in older adults. Hence the late initiation of HAART when their immune system deteriorates over time [32]. The reason why cardiovascular risk is higher in women with older age is not known. In advanced HIV disease, the risk is extremely high as immune activation and chronic in lammation are particularly important. PLWHIVs with TB and HBV-HCB coinfection, high blood pressure, smoking and alcohol use were also at high risk. These results support the recommendations of the National AIDS and STI Control Program (PNLS) which stipulate that tuberculosis must be actively sought out in PLHIVs in order to prevent and treat it. The health care providers should also check for HIV-hepatitis, HCV, and HBV co-infections. These comorbidities can have a negative impact on the health of PLWHIVs and eventually lead to serious complications. While collecting data on the history of the disease, providers should look for traditional cardiovascular risk factors that are otherwise frequently encountered in elderly PLWHIVs. A few clinical signs have played a predictive role in the onset of cardiovascular risk. These are fever, weight loss, elevated SBP and DBP, and respiratory rate. Fever in HIV infection is caused by different factors (HIV itself, infectious, cancerous, and other comorbidities). It therefore occurs in a context of immune activation and chronic in lammation and frequently at an advanced stage of the disease. Many PWLHIVs experience signi icant weight loss in the late stages of HIV infection. In a context where the number of elderly PLWHIVs is increasing, pathologies such as hypertension and diabetes will also increase. Hence the need to take vital signs at each consultation in order to monitor the appearance of hypertension if it is not already present and the weight, even if in the context of low resources, the equipment may be lacking in the care structures. In co-morbidities, TB and anaemia were medical conditions predicting high cardiovascular risk. In DRC, HIV infection occurs in a context of poverty, malnutrition, and low income for the population. It inevitably leads to anaemia. In the biological assessment, the blood count gave a statistically signi icant indication (Hb, WBC, neutrophils, lymphocytes). The blood count is among the basic examinations recommended by the PNLS. The CD4s + count which gives us an idea of the immune status shows that the risk is high in advanced HIV disease. In univariate and multivariate analysis, this study shows that advanced age (from 50 years), advanced HIV disease (WHO Stage 4 and CD4s+ < 200 cells/μL) and TB were factors associated with high atherogenic risk. In univariate analysis alone, anaemia was also a factor associated with atherogenic risk. The results of this study indicate that health structures caring for PLWHIV must realize that they will increasingly have to manage medical conditions with patients with multimorbidities and polypharmacies. When collecting data on the history of affection, providers must increasingly consider personal and family cardiovascular risk factors (hypertension, smoking, alcohol intake, diabetes mellitus, history of CAD, obesity, hypercholesterolemia, etc). These data also give an indication to the PNLS on the importance of achieving the lipid pro ile. This means that providers need to be trained and their knowledge updated regularly. They should learn to estimate cardiovascular risk by lipid ratios. The lipid pro ile and lipid ratios must be part of the routine care of PLWHIVs. The PNLS must advocate to obtain funding to subsidize the biological examinations. CD4s+ counting remains relevant and funding must be obtained for this examination, because it allows to assess the immunological status at the initial treatment, but also to determine if the PLWHIV is already at an advanced stage of the disease. Although the number of patients was not large, this study gives an indication of the place of lipid ratios in the evaluation of cardiovascular risk in PLWHIVs in addition to the lipid pro ile.
Lipid ratios have been studied by other authors and we compare our results to theirs in table 9.

Conclusion
Both Lipid ratios, CRI, CRII, and AC might be used as reliable and precise markers for diagnosis of high atherogenic risk among naive HIV patients from Kinshasa University Hospital.
Furthermore, age > = 50 years, Stage 4 WHO, tuberculosis, anaemia, and CD4s < 200 Cells/μL were identi ied as the most important signi icant and independent determinants of high atherogenic risk in those Congolese naive HIV patients.