The introduction of safe and effective combination antiretroviral therapy (cART) has dramatically improved the course of HIV infection.1 As new agents are developed and introduced into the existing armamentarium, it is a constant requirement of clinicians to remain up to date. Treatment recommendations, compiled by expert panels, are updated regularly as clinical trial data are published, to guide clinicians in their regimen choice.
Some cART combinations are discouraged owing to toxicity, drug interactions, or suboptimal efficacy. Double-boosted protease inhibitor (DBPI) regimens, for example, although previously used as part of salvage treatment in cases of multidrug resistance because of their high genetic barrier to resistance,2,3 were subsequently shown to have more side effects than new generation single-boosted PIs such as tipranavir and darunavir.4,5 The August 2006 International AIDS Society—USA (IAS-USA) recommendations proposed that DBPI combinations should be avoided accordingly.6 Triple-nucleoside (or nucleotide) reverse transcriptase inhibitor (NRTI) regimens, whose advantages over nonnucleoside reverse transcriptase inhibitor– or PI-based regimens included minimal drug interactions, ease of use and sparing of PIs and nonnucleoside reverse transcriptase inhibitors for future availability, were prescribed as alternatives for initial therapy.7 However, the 2004 IAS-USA recommendations discouraged triple-NRTI regimens because of inferior potency (inferior virological control). Finally, the 2004 IAS-USA recommendations discouraged the combination of didanosine (ddI) plus stavudine (d4T), given synergistic toxicity such as peripheral neuropathy, lipodystrophy, pancreatitis, and hyperlactatemia.3,8
Recommendations are also published by the European AIDS Clinical Society (EACS)9–11 and the US Department of Health and Human Sciences (DHHS).8,12 Considering the cART regimens above, the 2007 EACS guidelines recommend against DBPIs, whereas the DHHS does not mention this combination.8,10 Triple-NRTIs were identified as inferior in both the 2003 EACS and the 2004 DHHS recommendations. Finally, the 2003 DHHS recommendations advise against the combination of ddI plus d4T.8
Not all clinicians observe treatment recommendations, with inappropriate cART reported in 5%–47% of treatment-naive patients.13–16 Patients who are highly treatment experienced, or in whom there are concerns regarding adherence or drug interactions, may be faced with treatment options that are limited or not conforming to recommendations. Regional variation, ethnicity, sex, pretreatment CD4 count, and HIV viral load (VL) have been shown to play a part in guideline nonobservance.13,14,16 Surprisingly, the influence of new recommendations on clinical practice is understudied. To our knowledge, although some studies have examined guideline observance concerning initial cART,13–18 no study has analyzed the clinical impact of guideline updates in patients already well controlled on cART.
In Switzerland, the first national treatment recommendations were published in July 2011.19 Before this, most Swiss HIV physicians, and certainly those at Swiss HIV Cohort Study centers (SHCS; described below), referred to the IAS-USA recommendations (M.C., personal communication). We aimed to determine the impact of the August 2006 IAS-USA recommendations on the prescribing practices of Swiss clinicians by examining patients enrolled in the SHCS who were on nonrecommended regimens at that time.
We conducted a retrospective analysis from September 1, 2006, to December 31, 2010, using data recorded from patients enrolled in the SHCS. The SHCS is an ongoing, open, prospective observational cohort of HIV-infected patients followed at 7 centers in Switzerland, comprising outpatient clinics and their affiliated hospitals, and private practitioners.20 Cohort patients undergo data collection (sociodemographic characteristics, comorbidities, cART regimen, treatment adherence, whether or not on lipid-lowering drugs and clinical course) and blood sampling (CD4 count, VL, and lipid values) at inclusion and approximately every 6 months. We used the SHCS database extract of May 31, 2011. Approval for this study was obtained from the local ethical committees of all participating centers.
The August 2006 IAS-USA recommendations were reviewed, and nonrecommended cART regimens were chosen for analysis: DBPI regimens, defined as those containing booster-dose ritonavir plus 2 other PIs (excluding tipranavir prescribed in the context of a clinical study)2; triple-NRTI regimens [particularly lamivudine (3TC) and zidovudine (AZT) with abacavir (ABC) or tenofovir (TDF)] and regimens containing ddI plus d4T. In the triple-NRTI group, a subanalysis was performed to compare reasons of switch before and after the 2008 Conference on Retroviruses and Opportunistic Infections, when the results of the D:A:D study, showing an association between ABC use and risk of myocardial infarction, were presented.21
Patients enrolled in the SHCS were included if they (1) were receiving cART containing DBPIs, triple-NRTIs, or ddI plus d4T and (2) had at least 2 consecutive undetectable HIV VLs (defined as <50 copies per milliliter) throughout a period of at least 12 weeks before September 1, 2006. SHCS patients on cART not containing DBPIs, triple-NRTIs, or ddI plus d4T, but having at least 2 consecutive undetectable VLs for ≥12 weeks before September 1, 2006, were taken as controls.
The SHCS center location was recorded, and source of follow-up was classified as the SHCS center, other hospital, or private practitioner. For the sake of clarity, we use the term, nonobservance when referring to clinician nonadherence to treatment recommendations and the term, nonadherence when referring to patient nonadherence to the prescribed cART. Drug adherence to cART was recorded as the lowest adherence as reported by the patient during the period with an undetectable VL before September 1, 2006. We defined suboptimal drug adherence as the omission of at least 1 cART dose per month. Framingham 10-year risk score was calculated as previously described22 and classified according to the 10-year risk of developing coronary heart disease (myocardial infarction, angina pectoris, heart failure, and/or coronary death) as low (<10%), intermediate (10–20%), and high (>20%) risk. Cardiovascular (CV) events included myocardial infarction, cerebral infarction, or hemorrhage, other cardiac events (coronary angioplasty/stenting, coronary artery bypass grafting), and other vascular events (deep vein thrombosis, pulmonary embolism, and vascular procedures such as carotid endarterectomy and phlebectomy).
The end of the nonrecommended cART was defined as a switch when involving a change of treatment to another regimen, or as a discontinuation, when cART was stopped for ≥4 weeks as previously described.23 A switch to another regimen was defined as stopping the double PI association in the DBPI group and the triple-NRTI association in the triple-NRTI group and, in the ddI/d4T group, as stopping either of the 2 drugs. Reasons for treatment switch or discontinuation were classified as physician’s decision, patient’s wish/decision, abnormal fat distribution, toxicity (of gastrointestinal tract, liver, pancreas, nervous system, kidneys, endocrine system, hematological toxicity, lactic acidosis, and other), treatment failure (virological, immunological, or clinical), unknown, and other (such as death, loss to follow-up or structured treatment interruption). Physician’s decision encompassed treatment simplification or recommendation observance. These definitions correspond to the official classification used at the 6-monthly analyses of SHCS-enrolled patients.
The nearest CD4 count, VL (as performed by the Roche Amplicor HIV-1 Monitor Assay), and lipid values [total cholesterol, low-density lipoprotein, high-density lipoprotein, triglyceride (TG)] obtained up to 6 months before September 1, 2006, were recorded as inclusion parameters. CD4 count and VL and lipid values obtained at 24 and 48 weeks after any treatment switch were analyzed as, respectively, 24- and 48-week outcomes. HIV RNA suppression (undetectable VL) was defined as a VL <50 copies per milliliter.
The 3 nonrecommended cART regimens were analyzed separately. Patients in >1 treatment group were analyzed once in each group. Baseline characteristics of patients receiving each regimen were compared with that of the control group. Predictors of recommendation nonobservance were evaluated by comparing patients who continued nonrecommended regimens with those who switched/discontinued treatment at 2 different time points: 1 year (September 1, 2007) and 4 years (September 1, 2010) after the August 2006 IAS-USA recommendations. Predictors of receiving nonrecommended cART and of continuing outdated regimens were assessed using multivariable logistic regression analysis.
Drug adherence, lipid values, prescription of lipid-lowering agents, immunological, and virological outcomes at the time of treatment switch were compared with these parameters 24 and 48 weeks postregimen change for each patient using the Wilcoxon–Mann–Whitney test for continuous variables and the χ2-test for categorical variables.
Finally, continuation on each of the 3 nonrecommended cART regimens over time was described using Kaplan–Meier survival analyses. Statistical analyses were performed using Stata software (StataCorp, College Station, TX, version 11.1).
Patients’ Baseline Characteristics
On September 1, 2006, 104 patients were in the DBPI group, 436 in the triple-NRTI group, and 19 patients in the ddI/d4T group, with 3171 patient controls (Table 1). Of the patients on triple-NRTIs, 398/436 (91%) were on Trizivir (3TC ABC AZT), 22/436 (5%) were on 3TC AZT TDF, and the 16/436 (4%) were on other NRTI combinations. Of the 19 patients in the ddI/d4T group, 1 was additionally in the DBPI group, and 1 was in the triple-NRTI group.
TABLE 1-a Characteri...Image Tools
Compared with controls, a higher proportion of patients in the DBPI group had advanced HIV infection (documented AIDS-defining illness) and was highly treatment experienced, 53% having had >8 treatment changes. Treatment adherence was lower, and total cholesterol and TG values were higher, despite the fact that these patients were more frequently on lipid-lowering treatment. Most patients were followed up in SHCS centers.
TABLE 1-b Characteri...Image Tools
In the triple-NRTI group, fewer patients had advanced HIV infection. They were less treatment experienced, had higher Framingham risk scores but lower total cholesterol values compared with control patients. Most were followed up by private practitioners.
In the ddI/d4T group, more patients were followed up in peripheral hospitals compared with control patients. More than half had suboptimal treatment adherence (cART forgotten on ≥1 occasion per month).
Predictors of Nonobservance to the August 2006 IAS-USA Recommendations
Factors associated with recommendation nonobservance were examined at 1 and 4 years post-IAS-USA recommendation publication (September 1, 2007 and 2010, respectively), by comparing patients continued on nonrecommended cART to those who switched or discontinued such treatment (Fig. 1; see Tables S1–S3, Supplemental Digital Content 1, http://links.lww.com/AIE/A1).
By September 2007 and 2010, 75 (73%) and 28 (29%) patients, respectively, were still on DBPIs. After adjustment for relevant covariables, female patients and those with higher CD4 counts were more likely to continue on DBPIs until 2010. All the patients with a history of diabetes or a CV event discontinued DBPIs during the study period (Fig 1; see Table S1, Supplemental Digital Content 1, http://links.lww.com/QAI/A370).
For triple-NRTIs, 366 (86%) and 204 (52%) patients were still on this treatment by September 2007 and 2010, respectively. Patients with a history of a CV event were more likely to discontinue this regimen, whereas those with undetectable VLs were more likely to continue until 2010. Patients with fewer previous treatment changes were also more likely to continue up to 2007, but this characteristic was less relevant in 2010 (Fig. 1; see Table S2, Supplemental Digital Content 2, http://links.lww.com/QAI/A370).
For ddI plus d4T, 9 (47%) and 1 (5%) patients were still on this treatment by September 2007 and 2010, respectively. Patients continuing ddI plus d4T in September 2007 had higher lipid values (total cholesterol, low-density lipoprotein, and TG) than those who switched (Fig 1, see Table S3, Supplemental Digital Content 3, http://links.lww.com/QAI/A370).
Reason for Switching From Nonrecommended cART and Choice of New Regimen(s)
Discontinuation or switching from DBPIs to other regimens was driven mainly by the physician’s decision, including treatment simplification or recommendation observance (Tables 2 and 3). For triple-NRTIs, switching was mainly due to abnormal fat distribution. In the subanalysis of this group, concern of CV disease [odds ratio (OR) 4.64, 95% confidence interval (CI) 0.9 to 24.12, P = 0.07] and the physician's decision (OR 2.22, 95% CI 0.92 to 5.36, P = 0.08) were slightly more frequently reasons for switching or discontinuation after the D:A:D study,21 than abnormal fat distribution. Switching from ddI plus d4T was driven mainly by the physician’s decision. The alternative cART choices are shown in Table 3.
Changes in Treatment Adherence, Lipid values, and Virological and Immunological Outcome at 24 and 48 Weeks After cART Switch
During the study period, 70 (71%) patients switched over from DBPIs to other regimens. Excluding patients not followed up 6 and 12 months later, patients who discontinued cART and patients resumed on a DBPI regimen, 50 and 41 patients were analyzed, respectively (Fig. 2, Table 4, see Table S4, Supplemental Digital Content, http://links.lww.com/QAI/A370). Postswitch, cholesterol values were lower (P = 0.06; Table 4; see Table S4, Supplemental Digital Content, http://links.lww.com/QAI/A370). Two additional patients started lipid-lowering treatment. However, the cholesterol values in these 2 patients increased with time.
Considering triple-NRTIs, 187 (48%) patients switched regimen (Fig. 2). Follow-up 6 and 12 months later occurred in 135 and 121 patients, respectively. Postswitch, all lipid values were higher (P < 0.001) despite the fact that 2 additional patients were receiving lipid-lowering agents. Fewer patients had detectable VLs (P = 0.02; Table 4; see Table S4, Supplemental Digital Content 4, http://links.lww.com/QAI/A370).
Eighteen out of 19 patients (95%) in the ddI/d4T group switched regimen (Fig. 2). Follow-up 6 and 12 months later occurred in 14 and 13 patients, respectively. No parameter was significantly different 6 or 12 months postswitch. Although there was a trend toward improved treatment adherence at 6 months (P = 0.08), by 12 months, this had returned to baseline (P = 1; Table 4; see Table S4, Supplemental Digital Content 4, http://links.lww.com/QAI/A370).
Finally, although concern of CV disease increased as a reason for switching patients on triple-NRTIs after the presentation of the D:A:D data at the 2008 CROI meeting, we observed no increase in treatment switches per se after this time point (Fig. 2).
In our study of patients on nonrecommended cART regimens, we observe that physician’s decision (DBPIs; ddI plus d4T) and concern over abnormal fat distribution (triple-NRTIs) were the main factors driving treatment switches to recommendation-appropriate therapy. Conversely, patients with high CD4 counts (DBPIs) and suppressed VLs (triple-NRTIs) were more likely to continue on nonrecommended regimens.
Patient characteristics differed between the treatment groups. The patients on DBPIs mostly had advanced HIV infection, were heavily treatment experienced, and were followed up in SHCS centers rather than in affiliated hospitals or private practitioners, characteristics related to the salvage therapy nature of this regimen. The slightly higher lipid (notably TG) values observed in this group are probably related to the PIs.2,24 Patients on triple-NRTIs had less advanced HIV disease and were less treatment experienced, as expected for this regimen with a low genetic barrier to resistance. Most patients on ddI plus d4T had low treatment adherence suggesting that physicians were reluctant to change treatment in this population or that patients did not want to change therapy. We observed some cART prescribing differences between the different center types, in keeping with the degree of HIV specialization. SHCS centers are centers of excellence in HIV care and treat more experienced or ‘complicated’ patients such as those on DBPIs; the majority (67%) of the patients on ddI plus d4T were also followed up at SHCS centers. Private practitioners are HIV community physicians who care for less ill patients such as those on triple-NRTIs. In addition to center differences, it is also possible that individual clinicians within a single center had cART preferences, although we did not examine this. A recent SHCS analysis reported large differences in prescription choices among clinicians treating cART-naive patients but no difference in patient outcome.25
The number of patients in each treatment group is very different, as is the slope of the survival analysis curve. Two years after the publication of recommendations advising against the use of ddI plus d4T and triple-NRTIs, we identified only 19 patients on ddI plus d4T and yet 436 on triple-NRTIs. Furthermore, at the end of the study period, 95%, 71%, and 48% of the patients in the ddI plus d4T, DBPI, and triple-NRTI groups, respectively, had switched over to another regimen. These differences could be related to different clinician perceptions regarding the toxicity of each regimen.
Predictors of continued nonrecommended cART prescription also differed according to treatment group: Predictors of continued DBPI prescription were female sex and, as a trend, higher CD4 counts; predictors of continued triple-NRTI prescription were undetectable VL, fewer previous treatment changes, and lack of CV events; patients who continued ddI plus d4T had higher lipid values. The differences observed between the 3 cART groups are understandable given the diverse characteristics of each. As no published study has examined if and how cART regimens are altered in response to updated recommendations, we can only compare our results with those of studies examining the choice of initial cART prescriptions in treatment-naïve patients. A gender association was described in a recent SHCS publication by Wandeler et al,16 in which patients receiving an initial regimen in violation of the IAS-USA recommendations between 1998 and 2007 were more frequently female. This gender-based difference has already been described for prescription of specific types of cART26,27 and could be related to the socioeconomic status that is correlated with gender among HIV-positive patients. The association between recommendation nonobservance and high CD4 count has also been reported by Wandeler et al (4189 patients), and in a smaller US cohort of HIV-infected women (217 patients).13,16 For patients who are undetectable on triple NRTIs, the association between recommendation nonobservance and high CD4 count may be due to a perceived level of security on the part of the prescribing clinician. The clinician may be faced with the risks of continuing with a regimen that seems to work against the risk of starting a new recommended regimen that could precipitate problems with adherence and other adverse effects. We should mention also that, in Switzerland, the first single-tablet regimen (efavirenz/TDF/emtricitabine) became available in August 2010, 4 years after its availability in the United States.
We observe an association between clinical characteristics, such as CV events and diabetes and changes of cART in the DBPI and triple-NRTI groups, suggesting that clinicians observe recommendations when they become ‘relevant’ to patients. Rather than being influenced by CV risk factors such as lipid values or CV risk assessment tools such as the Framingham score, clinicians seem to react to an event once it has happened, rather than switch cART as primary prevention, if patients exhibit good virological and immunological control.
The reason for switching DBPIs or discontinuing cART altogether was driven primarily by the physician’s decision, in some cases specified as observance of treatment recommendations. The main reason for switching triple NRTIs was concern regarding abnormal fat distribution. A subanalysis of this group showed that concern of CV disease and physician’s decision increased slightly as reasons for switching after publication of the D:A:D study data,21 suggesting that large studies may influence cART choice. The switch from ddI plus d4T to another regimen was driven mainly by the physician’s decision and concern of abnormal fat distribution, in accordance with recommendations. The patients’ clinical course posttreatment switch agrees with the IAS-USA recommendations: Cholesterol values were slightly lower after DBPI switch, and the number of patients with detectable VL was lower28 after triple-NRTI switch. These results highlight the importance of treatment recommendation observance, supported further by the Swiss and US cohort studies cited above in which patients on recommendation-appropriate cART obtained significant virological benefit compared with patients on nonrecommended regimens.13,16
Our study has several limitations. Regarding reasons for treatment changes, although we were able to study why patients were switched, we were not able to assess why patients were continued on nonrecommended regimens. For the reasons themselves, we were restricted by the established SHCS classifications, notably physician’s decision, as there is no classification entitled response to treatment recommendations. Although some clinicians annotated physician’s decision by specifying that they were following treatment recommendations, many did not. Indeed, we observed some discrepancy between the ‘reasons for switch,’ as documented in the patient notes, and the ‘factors associated with switch,’ as determined from our analyses, and it is possible that we have overestimated the number of patients who changed cART as a result of treatment recommendations. Another limitation is that patient inclusion began in September 2006, after the August 2006 IAS-USA recommendations. Although this was appropriate for patients on DBPIs, it is likely that some patients on the other 2 nonrecommended regimens were already on different cART by 2006, in response to earlier recommendations, notably the 2004 IAS-USA recommendations. This would explain the small size of the ddI/d4T group, reducing the power of our study of these patients. Finally, although this was not the principal endpoint, it is difficult to draw conclusions regarding the outcome in patients who switched from nonrecommended to other cART: as alternative regimens were diverse, the postswitch patients represent a group that is sufficiently heterogeneous for it to be difficult to make meaningful comparisons between their preswitch and postswitch statuses.
In conclusion, we observe that, for some regimens, clinicians are guided more by the clinical, immunological, and virological course of individual patients than by recommendations, and that patients who are switched to recommended cART regimens have an improved clinical and virological course. Although cART choice can be a complex process, particularly in experienced patients whose treatment history is highly heterogeneous, our results demonstrate the benefit of observing available cART recommendations and suggest that recommendations should be observed even in well-controlled patients.
Ethical approval for this research was obtained from the institutional review boards. The members of the Swiss HIV Cohort Study are: Aubert V., Barth J., Battegay M., Bernasconi E., Böni J., Bucher H.C., Burton-Jeangros C., Calmy A., Cavassini M., Egger M., Elzi L., Fehr J., Fellay J., Francioli P., Furrer H. (Chairman of the Clinical and Laboratory Committee), Fux C.A., Gorgievski M., Günthard H. (President of the SHCS), Haerry D. (Deputy of “Positive Council”), Hasse B., Hirsch H.H., Hirschel B., Hösli I., Kahlert C., Kaiser L., Keiser O., Kind C., Klimkait T., Kovari H., Ledergerber B., Martinetti G., Martinez de Tejada B., Metzner K., Müller N., Nadal D., Pantaleo G., Rauch A. (Chairman of the Scientific Board), Regenass S., Rickenbach M. (Head of Data Center), Rudin C. (Chairman of the Mother & Child Substudy), Schmid P., Schultze D., Schöni-Affolter F., Schüpbach J., Speck R., Taffé P., Tarr P., Telenti A., Trkola A., Vernazza P., Weber R., and Yerly S.
1. Palella FJ Jr, Delaney KM, Moorman AC, et al.. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med. 1998;338:853–860.
2. Osih RB, Taffe P, Rickenbach M, et al.. Outcomes of patients on dual-boosted PI regimens: experience of the Swiss HIV cohort study. AIDS Res Hum Retroviruses. 2010;26:1239–1246.
3. Yeni PG, Hammer SM, Hirsch MS, et al.. Treatment for adult HIV infection: 2004 recommendations of the International AIDS Society-USA Panel. JAMA. 2004;292:251–265.
4. Walmsley SL, Katlama C, Lazzarin A, et al.. Pharmacokinetics, safety, and efficacy of tipranavir boosted with ritonavir alone or in combination with other boosted protease inhibitors as part of optimized combination antiretroviral therapy in highly treatment-experienced patients (BI Study 1182.51). J Acquir Immune Defic Syndr. 2008;47:429–440.
5. Hicks CB, Cahn P, Cooper DA, et al.. Durable efficacy of tipranavir–ritonavir in combination with an optimised background regimen of antiretroviral drugs for treatment-experienced HIV-1-infected patients at 48 weeks in the Randomized Evaluation of Strategic Intervention in multi-drug reSistant patients with Tipranavir (RESIST) studies: an analysis of combined data from two randomised open-label trials. Lancet. 2006;368:466–475.
6. Hammer SM, Saag MS, Schechter M, et al.. Treatment for adult HIV infection: 2006 recommendations of the International AIDS Society-USA panel. [Reprint in Top HIV Med
. 2006;14:827–43; PMID: 17016878]. JAMA
7. Yeni PG, Hammer SM, Carpenter CC, et al.. Antiretroviral treatment for adult HIV infection in 2002: updated recommendations of the International AIDS Society-USA Panel. JAMA. 2002;288:222–235.
9. European guidelines for the clinical management and treatment of HIV-infected adults in Europe. AIDS. 2003;17(suppl 2):S3–S26.
11. Guidelines clinical management and treatment of HIV infected adults in Europe. European Aids Clinical Society (EACS). 2009;5.
13. Cocohoba J, Wang QJ, Cox C, et al.. Consistency of initial antiretroviral therapy with HIV treatment guidelines in a US cohort of HIV-infected women. J Acquir Immune Defic syndr. 2008;47:377–383.
14. Holodniy M, Hornberger J, Rapoport D, et al.. Relationship between antiretroviral prescribing patterns and treatment guidelines in treatment-naive HIV-1-infected US veterans (1992–2004). J Acquir Immune Defic Syndr. 2007;44:20–29.
15. Wood E, Hogg RS, Kerr T, et al.. Rates of inappropriate antiretroviral prescription among injection drug users. Harm Reduct J. 2007;4:2.
16. Wandeler G, Keiser O, Hirschel B, et al.. A Comparison of initial antiretroviral therapy in the Swiss HIV Cohort Study and the recommendations of the International AIDS Society-USA. PLoS One. 2011;6:e27903.
17. Latthaphasavang V, Bouldouyre MA, Rachline A, et al.. Antiretroviral therapy initiation in France: adherence to national guidelines and outcome. J Int Assoc Phys AIDS Care (Chic). 2011;11:40–46.
18. Suarez-Lozano I, Viciana P, Lacalle JR, et al.. The relationship between antiretroviral prescription patterns and treatment guidelines in treatment-naive HIV-1-infected patients. HIV Med. 2009;10:573–579.
19. Elzi L, Flepp M. Recommandations de la Commission d’experts clinique et thérapie VIH/sida (CCT) pour le début d’un traitement antirétroviral chez les adultes infectés par le VIH. Bull OFSP. 2011;27:565–568.
20. Schoeni-Affolter F, Ledergerber B, Rickenbach M, et al.. Cohort profile: the Swiss HIV Cohort study. Int J Epidemiol. 2010;39:1179–1189.
21. Sabin CA, Worm SW, Weber R, et al.. Use of nucleoside reverse transcriptase inhibitors and risk of myocardial infarction in HIV-infected patients enrolled in the D: A:D study: a multi-cohort collaboration. Lancet. 2008;371:1417–1426.
22. Expert Panel on Detection E, Treatment of High Blood Cholesterol in A. Executive summary of the third report of The National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA. 2001;285:2486–2497.
23. Elzi L, Marzolini C, Furrer H, et al.. Treatment modification in human immunodeficiency virus-infected individuals starting combination antiretroviral therapy between 2005 and 2008. Arch intern Med. 2010;170:57–65.
24. Gyalrong-Steur M, Bogner JR, Seybold U. Changes in lipid profiles after switching to a protease inhibitor-containing cART–unfavourable effect of fosamprenavir in obese patients. Eur J Med Res. 2011;16:85–92.
25. Elzi L, Erb S, Furrer H, et al.. Choice of initial combination antiretroviral therapy in individuals with HIV infection: determinants and outcomes. Arch Intern Med. 2012:13:1–9.
26. Gebo KA, Fleishman JA, Conviser R, et al.. Racial and gender disparities in receipt of highly active antiretroviral therapy persist in a multistate sample of HIV patients in 2001. J Acquir Immune Defic Syndr. 2005;38:96–103.
27. Mocroft A, Gill MJ, Davidson W, et al.. Are there gender differences in starting protease inhibitors, HAART, and disease progression despite equal access to care? J Acquir Immune Defic Syndr. 2000;24:475–482.
28. Gulick RM, Ribaudo HJ, Shikuma CM, et al.. Triple-nucleoside regimens versus efavirenz-containing regimens for the initial treatment of HIV-1 infection. N Engl J Med. 2004;350:1850–1861.
© 2013 Lippincott Williams & Wilkins, Inc.