SemOpenAlex |

SemOpenAlex

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2007174639> ?p ?o ?g. }

Showing items 1 to 100 of ±128 with 100 items per page.

W2007174639 endingPage "S201" @default.
W2007174639 startingPage "S193" @default.
W2007174639 abstract "Introduction Highly active antiretroviral therapy (HAART), a combination of at least three drugs that typically includes either a protease inhibitor (PI) or a non-nucleoside analogue reverse transcriptase inhibitor (NNRTI) and two nucleoside analogues, has led to substantial reductions of the morbidity and mortality of HIV infection. HAART regimens that include at least three drugs and aim to reduce and maintain viral plasma levels below detectable levels represent the current standard of antiretroviral therapy [1,2]. The toxicities and other adverse effects of HAART have received much attention recently [3]. This is not surprising considering that cure of the infection is unlikely with current regimens, and that treatment will probably have to be continued indefinitely [4]. In addition to hepatitis, pancreatitis, neuropathy and gastrointestinal symptoms, there is growing concern about a syndrome of lipodystrophy that is accompanied by metabolic complications, including hypercholesterolaemia, hypertriglyceridaemia and insulin resistance [3,5]. These metabolic effects may increase cardiovascular risk; however, their clinical significance is unclear at present. In this article, we present a brief overview of cardiovascular epidemiology, with an emphasis on coronary heart disease (CHD), and review the metabolic complications associated with the lipodystrophy syndrome. We then analyse data from cardiovascular cohort studies to estimate the relative and absolute CHD risks that are likely to be associated with the lipodystrophy syndrome. Using data from prospective studies of HIV-1 infection, we put these risks into perspective by comparing the benefits in terms of AIDS-related complications prevented with the risk of coronary events. We conclude by discussing the limitations of this approach, and outlining the need for further research. The epidemiology of CHD Cardiovascular disease is a leading cause of morbidity and mortality in most of the industrialized Western nations where HAART is currently widely used, and is predicted to achieve that status in many less developed countries within the next decades [6]. Here we shall focus on the epidemiology of CHD, which accounts for about one-half of all atherosclerotic cardiovascular complications. What is the incidence of CHD in different countries? The WHO MONICA Project (monitoring trends and determinants in cardiovascular disease) used standardized methods to measure the incidence of fatal and non-fatal CHD events over a 10-year period from the early 1980s in 37 populations from 21 countries [7]. The results showed a heterogeneous picture (Fig. 1) with important differences in CHD event rates between countries, and in trends over time [8]. For example, among men from Catalonia, Spain, the incidence was only about one-quarter of the rate in Glasgow, United Kingdom. In most populations, rates declined over the study period with the greatest falls observed in Northern European populations. Important increases in CHD rates were, however, observed in central and Eastern Europe, and in Asia [8].Fig. 1: Population rankings of coronary event rates in selected populations from the WHO MONICA Project, by sex. Annual rates per 100 000 population with 95% confidence intervals are shown. Adapted from Tunstall-Pedoe et al. [8].Age, sex and genes CHD is the consequence of coronary atherosclerosis, a chronic process that is only partly understood but progresses from the deposition of lipid-laden macrophages (foam cells) early in life to fatty streaks and atherosclerotic plaques with lipid core and calcium deposits [9]. The impact of age on CHD risk is reminiscent of the CD4 cell count in HIV infection: CHD rates increase exponentially with age (Fig. 2). Age is a marker of increasing atherosclerotic burden, which in men starts to manifest itself with plaque ruptures and clinical events in middle age. The incidence of CHD in women lags about 10-15 years behind and, at any age, is considerably lower than in men. This may be related to a protective effect of female hormones in pre-menopausal women, although there is little evidence that CHD rates in women accelerate after the menopause [10]. A large number of susceptibility genes and mutant alleles have been identified, but only a few relatively rare genetic disorders such as familial hypercholesterolaemia are attributable to a single gene effect [11]. Environmental factors are the dominant determinants of CHD at the population level.Fig. 2: Mortality from coronary heart disease by sex in England and Wales, 1989-1993.Modifiable risk factors Epidemiological research based on cohort studies with long follow-up has contributed to a vast body of information on the factors that predispose to the occurrence of cardiovascular disease. These factors have come to be known as 'risk factors', a term popularized by the Framingham Heart Study Group in the 1960s [12]. This work led to the definition of multivariate risk profiles for different manifestations of cardiovascular disease that facilitate the identification of people most likely to benefit from preventive measures [13]. The classical risk factors for CHD include cigarette smoking, serum cholesterol, blood pressure, obesity and diabetes. For atherosclerotic stroke, hypertension and diabetes are of particular importance; cigarette smoking and diabetes, and diabetes and hypertension, respectively, dominate in the case of peripheral artery disease and heart failure. The atherogenic component of serum cholesterol is its low density lipoprotein component, whereas a high density lipoprotein (HDL), which is involved in the removal of cholesterol from tissues and the clearance of triglyceride-rich particles, is a recognized protective factor [14]. The role of hypertriglyceridaemia is controversial [15,16]. Although univariable analyses have consistently found an association with CHD, in many studies this association was greatly attenuated when the effect was adjusted for other cardiovascular risk factors [16,17]. A recent analysis of the Caerphilly Heart Disease Study concluded that, while hypertriglyceridaemia may exert an influence independent of other lipid factors, insulin resistance is probably the underlying metabolic disturbance [18]. Insulin resistance clusters with a number of related risk factors ('syndrome X') [19], including hypertriglyceridaemia, hypercholesterolaemia and low levels of HDL cholesterol, a combination of risk factors that is also observed in patients with HAART-associated lipodystrophy. The lipodystrophy syndrome The term 'lipodystrophy syndrome' has been introduced [5] to describe abnormal fat distribution (peripheral lipoatrophy and central fat accumulation), dyslipidaemia and insulin resistance, a combination observed in a substantial proportion of patients treated with HAART. A study to develop a precise case definition has recently been initiated and is ongoing (Andrew Carr, personal communication, 2000). There is uncertainty regarding the role played by the different agents and classes of agents combined in HAART. The use of PI has been associated with the presence of lipodystrophy in numerous studies [20]. Nucleoside analogue reverse transcriptase inhibitors, in particular stavudine, also appear to contribute to the development of lipodystrophy [21-23]. More recently, lipodystrophy has also been associated with the use of the NNRTI efavirenz [24]. Current understanding is hampered by the lack of randomized trials directly comparing different drugs and drug classes. Dyslipidaemia A randomized, placebo-controlled trial in healthy, HIV-seronegative volunteers showed that ritonavir led to an increase in fasting total cholesterol by 1 mmol/l. This increase was mainly due to an increase in very low density lipoprotein cholesterol, whereas HDL cholesterol decreased by 0.1 mmol/l, resulting in an unfavourable increase in the ratio of total cholesterol to HDL cholesterol [25] [see Table 1 for conversion factors (mmol/l to mg/dl) for cholesterol, triglycerides and glucose]. Triglycerides also increased substantially, by 1.8 mmol/l. In an age-matched comparison of HIV-infected lipodystrophy patients with participants of the Framingham Offspring Study, total cholesterol was 0.9 mmol/l higher and HDL cholesterol 0.3 mmol/l lower in patients with lipodystrophy [26]. The cholesterol to HDL ratio was 6.6 among lipodystrophy patients compared with 4.4 among Framingham controls.Table 1: Converting units for cholesterol, triglycerides and glucose.Are there differences between different PI? Cross-sectional and cohort studies found comparable changes in lipid factors in patients on different PI, either compared with pre-treatment levels [27,28] or with other patients not receiving PI [29-31]. An analysis from the Swiss HIV Cohort Study [27], based on 93 patients, suggested the following ranking in terms of propensity to cause hypercholesterolaemia: ritonavir ≫ nelfinavir > indinavir > saquinavir. Interestingly, in dyslipidaemic patients being switched from ritonavir to other PI, the cholesterol levels decreased [27]. Also, the ritonavir effect appears to be dose dependent. In a subgroup analysis of a randomized trial [32], ritonavir, but not indinavir, induced hypertriglycerideamia. The effects of lopinavir and amprenavir are not well documented at present. Efavirenz, similar to the PI, increased cholesterol levels in healthy volunteers [33], and data presented at recent meetings indicate that NNRTI-containing HAART regimens also lead to dyslipidaemia [34,35]. Insulin resistance, impaired glucose tolerance and diabetes mellitus Insulin resistance may be the underlying disturbance induced by HAART, with dyslipidaemia and hypertriglyceridaemia developing later as a consequence of this basic abnormality [19,36]. Indeed, a recent study in HIV-seronegative volunteers showed that administration of indinavir during 4 weeks led to insulin resistance, with little change in lipid factors [37]. This appears to be the result of decreased sensitivity to insulin in the peripheral tissues, whereas insulin release from the pancreas is not affected [36,38]. A study in HIV-1-infected patients found that none of 13 untreated controls but 41 (61.2%) of 67 PI-treated patients had reduced peripheral insulin sensitivity [29]. Insulin resistance may lead to impaired glucose tolerance, and in some patients to overt type 2 diabetes. In Carr et al.'s cohort study of PI-treated patients [30], diabetes mellitus was diagnosed in 7% and impaired glucose tolerance in a further 16%, whereas rates in two smaller, cross-sectional studies were 38 and 17% [29], and 7 and 35% [26], respectively. Absolutely relative: estimating the increase in coronary risk We used two existing cardiovascular cohort studies from non-HIV-infected populations to estimate what increase in coronary risk is likely to be associated with HAART-induced lipodystrophy. The characteristics of the Caerphilly Heart Disease Study and the Framingham Heart Study are summarized in Table 2, together with the definitions used for CHD. Figure 3 shows relative risks of CHD associated with the metabolic disturbances that are typical for the lipodystrophy syndrome in Caerphilly men. Depending on the severity of the metabolic disturbances present, the relative risk of CHD ranged from 1.4 when only one lipid factor was abnormal to 5.2 when both total cholesterol and triglycerides were increased and diabetes was present. These estimates were derived from a cohort exclusively based on middle-aged men, but relationships with lipid factors are known to be similar in women whereas diabetes is a somewhat stronger risk factor in women [39]. Findings from the Chicago Heart Association Detection Project in Industry [40] indicate that relative risks for most risk factors are of similar magnitude in young and middle-aged men. The association with cholesterol is, however, stronger in younger men [40-42].Table 2: Description of cardiovascular cohort studies and outcome variables used in analyses.Fig. 3: Relative risk of coronary heart disease associated with the metabolic disturbances that are typical for the lipodystrophy syndrome. Results from multivariable models adjusted for age, smoking, HDL cholesterol and all variables shown. Hypercholesterolaemia, > 5.5 mmol/l; hypertriglyceridaemia, > 2.0 mmol/l; diabetes, fasting blood glucose > 7.0 mmol/l or known diabetes.Relative risks do not reflect underlying absolute risk of CHD. The absolute adverse effect of HAART on the risk of CHD will vary considerably depending on the underlying risk in the absence of HAART. Indeed, there are pronounced differences in absolute risk between men and women, and between the young and the old (Fig. 2). Even a fivefold increase in the risk of CHD will translate into a modest increase if the underlying risk is very low, for example in young women. The same principle does of course apply to the benefit of HAART: if the risk of progressing to AIDS is small without treatment, then the absolute reduction in risk with treatment is also bound to be small. Numbers needed to treat to benefit or harm The number of patients needed to be treated for one patient to benefit (NNTbenefit), or one patient to be harmed (NNTharm), are useful measures both for clinicians attempting to determine the optimal care for an individual patient and for those planning health services [43,44]. The NNTbenefit or NNTharm value corresponds to the reciprocal of the absolute risk difference. For example, if the risk of developing AIDS over 3 years is reduced from 0.5 (50%) to 0.3 (30%), then the absolute risk reduction is 0.2 (20%) and NNTbenefit is 5 (1/0.2 or 100/20%). In this example, the relative risk is 0.6 (0.3/0.5), for a relative risk reduction of 40%. Applying the same relative reduction to a patient at much lower risk of progressing to AIDS, then NNTbenefit will increase correspondingly (to 50 if the baseline risk is 0.05 or 5%). Similarly, a threefold increase in the risk of CHD associated with HAART will result in NNTharm of 50 if the baseline risk is 1% but produce NNTharm of 5 if pre-treatment risk is 10%. What is the likely NNTharm associated with the metabolic complications observed with HAART? We used the updated Framingham risk equation [45] to estimate the absolute increase in the 3-year, 5-year and 10-year risk of CHD associated with more severe metabolic complications and to calculate NNTharm in patient groups defined by sex, smoking status and age. As shown in Table 3, the predicted risk is low, even over 10 years, in young men and women who do not smoke (< 1.5%). It increases substantially when modelling the effect of metabolic complications but the absolute risk remains relatively low (< 5% over 10 years), and the NNTharm is large (77 over 10 years for women). The picture changes dramatically in smokers and when setting age to 50 years: the predicted risk is already fairly high in the absence of therapy-induced metabolic complications and increases to levels as high as 26% over 10 years if complications develop, with the NNTharm falling below 30 over 3 years and below 10 over 10 years. For example, in women aged 50 years who smoke, the risk over 3 years will increase from an estimated 2.1% to an estimated 8.2%, for a NNTharm of 16 (Table 3).Table 3: Estimates of the risk of coronary heart disease (CHD) over 3, 5 and 10 years for patients with no metabolic complications and patients with antiretroviral treatment induced metabolic complications, and the number of patients that need to be treated to produce one additional CHD event (NNTharm), in patient groups defined by sex, smoking status and age.How does harm from CHD compare with the benefits of HAART? We examined this question by comparing the risk of progression to AIDS in the pre-HAART and HAART eras, stratifying by CD4 cell count and viral load. Figure 4a is adapted from the Mellors et al. analysis of the Multicenter AIDS Cohort Study [46] that enrolled patients in the mid-1980s and followed them until 1995, before HAART became available. The analysis is based on 1604 men, 998 of whom developed AIDS during follow-up. Figure 4b is based on analyses of the Swiss HIV Cohort Study of 2622 men and women who initiated HAART from October 1995 onwards and were followed-up, on average, for 3 years. There were 194 AIDS events. Kaplan-Meier estimates of the probability of AIDS at 3 years are shown. It is clear that few patients need to be treated for one person to benefit if CD4 cell counts are low or viral load is high. For example, an estimated two out three patients with CD4 cell counts below 200 × 106/l and viral load above 55 000 copies/ml will benefit from antiretroviral treatment (NNTbenefit = 1 / (0.855-0.191) = 1.5). NNTbenefit values are between 1.5 and 3 if viral load is above 55 000 copies/ml, 4-11 if viral load is between 20 000 and 55 000 copies, 16-31 for viral loads in the range of 7000-20 000 copies, and 50-200 in patients with viral loads below 7000 copies/ml.Fig. 4: Probability of AIDS in (a) the pre-HAART era (1985-1995; data from the Multicenter AIDS Cohort Study [46]) and (b) the HAART era (1996-2000; Swiss HIV Cohort Study [69]). Kaplan-Meier estimates of the probability of AIDS at 3 years are shown.Limitations of the proposed approach There are important limitations to our estimation of both the risks of CHD and AIDS. The wisdom, or otherwise, of extrapolating estimates of coronary risk from prospective studies of non-HIV-infected populations to HIV-infected patients with drug-induced metabolic complications remains to be determined. Several widely used drugs induce secondary dyslipidaemia, and an increase in cardiovascular disease has been documented for some of them [47]. However, whether changes in CHD risk factors induced by drugs predict risk as accurately as they predict the risk in the general population remains uncertain. The risk of cardiovascular complications will probably depend on duration of drug use, as in weight lifters who used anabolic agents [48]. It seems plausible that HAART-induced dyslipidaemia and insulin resistance will increase the risk of CHD, but the significance of duration of use and of possible interactions with age, sex and pre-existing risk factors are unclear at present. The risk of cardiovascular disease in HIV-infected patients may differ from the general population for reasons other than HAART, including differences in relevant lifestyles, dyslipidaemia due to chronic HIV infection [49] and HIV-induced heart disease such as myocarditis, cardiomyopathy or endocarditis [50]. The Framingham risk equation used in our calculations has been shown to be applicable to Northern Europe [51]. It will, however, overestimate absolute risk in Southern European populations (the 'French paradox') [52,53]. The Framingham data reflect the experience in past decades and will overestimate present levels of absolute coronary risk even in the United States and Northern Europe [54]. Our calculations of absolute risk are therefore likely to overestimate CHD risk in many patients. On the contrary, the greater excess CHD risks in young men with hypercholesterolaemia [40-42] are not taken into account in the Framingham equation. Also, we ignored the risk of other cardiovascular events that are also likely to be affected by metabolic complications, including cerebrovascular and peripheral vascular disease, and congestive heart failure. Our estimates of the benefit of HAART stem from observational cohort studies, rather than randomized, controlled trials, which means that they are vulnerable to bias and confounding [55]. Furthermore, estimates for the progression rates in the era of HAART are based on small numbers of events and are therefore imprecise. For example, the confidence interval for the estimated risk of 2.7% in patients starting HAART with a CD4 cell count of 201-350 cells × 106/l and a viral load of 7000-20 000 copies/ml (Fig. 4) ranged from 0.1 to 8.0%. Also, the analysis included both drug-naïve and pre-treated patients, but pre-treatment is known to affect prognosis [56]. Need for further research These concerns illustrate the need for prospective data from HIV-infected populations. A meta-analysis of four randomized trials comparing a PI containing a triple regimen with dual nucleoside analogue reverse transcriptase inhibitor therapy showed no increase in CHD risk among patients treated with PI [57]. However, the study lacked power and patients were followed for 1 year only. As already mentioned, longer periods of exposure may be required for an increase in CHD risk to become evident. A retrospective analysis of the French Hospital Database on HIV identified 54 incident cases of myocardial infarction among 19 795 men exposed to protease inhibitor containing regimens [58]. The standardized mortality ratios relative to the general population were 1.7 (95% confidence interval = 1.0-2.7) for patients exposed to HAART for 18-29 months and 3.1 (95% confidence interval = 1.7-5.4) among those exposed 30 months or longer [58]. These results should be interpreted with caution considering the retrospective nature of the study. For example, ascertainment of events may have been less complete in earlier calendar years, before the metabolic complications of HAART became widely known. A major prospective study, the DAD cohort (Data Collection on Adverse events of Anti-HIV Drugs) is ongoing. The DAD study is a collaboration between 11 HIV cohort studies from Europe, Australia and the United States (see http://www.cphiv.dk/dad for details) that, supported by regulatory authorities and all pharmaceutical companies with licensed anti-HIV agents, has enrolled over 20 000 patients. Cardiovascular risk factors at baseline and the incidence of myocardial infarction are assessed according to a standardized protocol. The study was set up in 1999; incidence data will become available in 2002. Further research is also required to describe and model the risk of progression to AIDS and death of patients treated with HAART. The ART Cohort Collaboration was established in 2000 to provide precise estimates of the probability of progression to a new AIDS event or death for drug-naïve patients starting potent antiretroviral combination therapy [59]. The database currently includes 13 000 individuals from 12 cohorts. Large and well-designed prospective studies using appropriate methods to measure glucose tolerance are needed to define the risk of impaired glucose tolerance and overt diabetes. Finally, a cardiovascular risk equation that is tailored to the metabolic complications associated with HAART and based on data from HIV-infected populations would be useful. Conclusions The approach outlined herein allows direct comparison of likely benefits and harm in patient groups characterized by virological, immunological and cardiovascular risk factors, and should inform decision-making in individual cases, taking into account estimates of the risk of progression both to AIDS and CHD. Our calculations illustrate the potential limitations of current treatment recommendations. For example, the updated guidelines of the International AIDS Society, USA Panel, recommend HAART for patients with a viral load in the range of 5000-30 000 copies/ml and a CD4 cell count between 200 and 350 cells × 106/l [2]. A middle-aged, male smoker with a viral load below 10 000 copies who starts HAART and develops metabolic complications may, however, increase his/her cardiovascular risk in the short term more than reducing the risk of HIV-related complications. The same data illustrate the importance of smoking as a risk factor for CHD. Smoking cessation should be discussed routinely in clinical practice and, together with a healthy diet and exercise, should generally take precedence over the administration of additional drugs to lower cholesterol or increase insulin sensitivity. We acknowledge that our assessments of CHD risk are not based on direct evidence from HIV-infected patients, and are therefore hypothetical at present. Prospective data from HIV-infected populations are urgently required to define the nature of the increase in cardiovascular risk that is likely to be associated with the metabolic complications of HAART. Finally, it is clear that it will always be difficult to predict effects of interventions with any certainty for individual patients, a problem that has plagued medicine for centuries [60-62]. Acknowledgements The authors are grateful to the Swiss HIV Cohort Study and the Caerphilly Heart Disease Study groups for allowing access to their data. The Swiss HIV Cohort Study is supported by the Swiss National Science Foundation (Grant number 3345-062041). The Caerphilly study was undertaken by the former MRC Epidemiology Unit (South Wales) and was funded by the Medical Research Council of the United Kingdom. They also thank George Davey Smith, Caroline Sabin, Bruno Ledergerber, Andrew Phillips and the editors for helpful comments on an earlier draft of this paper." @default.
W2007174639 created "2016-06-24" @default.
W2007174639 creator A5037887037 @default.
W2007174639 creator A5064632852 @default.
W2007174639 creator A5077161809 @default.
W2007174639 creator A5088502098 @default.
W2007174639 date "2001-01-01" @default.
W2007174639 modified "2023-10-16" @default.
W2007174639 title "Highly active antiretroviral therapy and coronary heart disease: the need for perspective" @default.
W2007174639 cites W142854403 @default.
W2007174639 cites W1735143085 @default.
W2007174639 cites W1963894977 @default.
W2007174639 cites W1967962375 @default.
W2007174639 cites W1978444406 @default.
W2007174639 cites W1989258056 @default.
W2007174639 cites W1992229080 @default.
W2007174639 cites W1998660799 @default.
W2007174639 cites W2002697160 @default.
W2007174639 cites W2003358339 @default.
W2007174639 cites W2007061978 @default.
W2007174639 cites W2007142745 @default.
W2007174639 cites W2010844841 @default.
W2007174639 cites W2017766892 @default.
W2007174639 cites W2024720867 @default.
W2007174639 cites W2028643246 @default.
W2007174639 cites W2029255052 @default.
W2007174639 cites W2031968478 @default.
W2007174639 cites W2032941190 @default.
W2007174639 cites W2039907887 @default.
W2007174639 cites W2052176508 @default.
W2007174639 cites W2052578635 @default.
W2007174639 cites W2053382469 @default.
W2007174639 cites W2056629334 @default.
W2007174639 cites W2058217770 @default.
W2007174639 cites W2060210361 @default.
W2007174639 cites W2072353912 @default.
W2007174639 cites W2076562284 @default.
W2007174639 cites W2077038891 @default.
W2007174639 cites W2090443364 @default.
W2007174639 cites W2094768566 @default.
W2007174639 cites W2103395470 @default.
W2007174639 cites W2105886042 @default.
W2007174639 cites W2108361166 @default.
W2007174639 cites W2115999476 @default.
W2007174639 cites W2126223811 @default.
W2007174639 cites W2133315239 @default.
W2007174639 cites W2136385005 @default.
W2007174639 cites W2137570167 @default.
W2007174639 cites W2138990735 @default.
W2007174639 cites W2142662148 @default.
W2007174639 cites W2149559097 @default.
W2007174639 cites W2151975228 @default.
W2007174639 cites W2155449170 @default.
W2007174639 cites W2171141994 @default.
W2007174639 cites W2319657764 @default.
W2007174639 cites W2324473093 @default.
W2007174639 cites W2329192296 @default.
W2007174639 cites W4214927700 @default.
W2007174639 cites W4234806062 @default.
W2007174639 cites W4244919073 @default.
W2007174639 cites W4248142985 @default.
W2007174639 cites W4248391713 @default.
W2007174639 cites W4252910375 @default.
W2007174639 doi "https://doi.org/10.1097/00002030-200100005-00023" @default.
W2007174639 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/11816168" @default.
W2007174639 hasPublicationYear "2001" @default.
W2007174639 type Work @default.
W2007174639 sameAs 2007174639 @default.
W2007174639 citedByCount "26" @default.
W2007174639 countsByYear W20071746392013 @default.
W2007174639 countsByYear W20071746392020 @default.
W2007174639 countsByYear W20071746392021 @default.
W2007174639 crossrefType "journal-article" @default.
W2007174639 hasAuthorship W2007174639A5037887037 @default.
W2007174639 hasAuthorship W2007174639A5064632852 @default.
W2007174639 hasAuthorship W2007174639A5077161809 @default.
W2007174639 hasAuthorship W2007174639A5088502098 @default.
W2007174639 hasBestOaLocation W20071746391 @default.
W2007174639 hasConcept C126322002 @default.
W2007174639 hasConcept C12713177 @default.
W2007174639 hasConcept C142462285 @default.
W2007174639 hasConcept C154945302 @default.
W2007174639 hasConcept C159047783 @default.
W2007174639 hasConcept C164705383 @default.
W2007174639 hasConcept C177713679 @default.
W2007174639 hasConcept C2776452011 @default.
W2007174639 hasConcept C2780727368 @default.
W2007174639 hasConcept C2993143319 @default.
W2007174639 hasConcept C3013748606 @default.
W2007174639 hasConcept C3018906752 @default.
W2007174639 hasConcept C41008148 @default.
W2007174639 hasConcept C71924100 @default.
W2007174639 hasConceptScore W2007174639C126322002 @default.
W2007174639 hasConceptScore W2007174639C12713177 @default.
W2007174639 hasConceptScore W2007174639C142462285 @default.
W2007174639 hasConceptScore W2007174639C154945302 @default.
W2007174639 hasConceptScore W2007174639C159047783 @default.
W2007174639 hasConceptScore W2007174639C164705383 @default.